{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 70000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.0981595516204834, "learning_rate": 1.8e-07, "loss": 1.666, "step": 10 }, { "grad_norm": 1.1748125553131104, "learning_rate": 3.8e-07, "loss": 1.6754, "step": 20 }, { "grad_norm": 1.083774209022522, "learning_rate": 5.8e-07, "loss": 1.6706, "step": 30 }, { "grad_norm": 1.0464515686035156, "learning_rate": 7.8e-07, "loss": 1.6619, "step": 40 }, { "grad_norm": 1.1835473775863647, "learning_rate": 9.8e-07, "loss": 1.6416, "step": 50 }, { "grad_norm": 2.0714001655578613, "learning_rate": 1.18e-06, "loss": 1.553, "step": 60 }, { "grad_norm": 1.1343004703521729, "learning_rate": 1.3800000000000001e-06, "loss": 1.4618, "step": 70 }, { "grad_norm": 1.2518393993377686, "learning_rate": 1.5800000000000003e-06, "loss": 1.4226, "step": 80 }, { "grad_norm": 1.5689958333969116, "learning_rate": 1.7800000000000001e-06, "loss": 1.3708, "step": 90 }, { "grad_norm": 2.0473012924194336, "learning_rate": 1.98e-06, "loss": 1.3201, "step": 100 }, { "grad_norm": 2.328078031539917, "learning_rate": 2.1800000000000003e-06, "loss": 1.2639, "step": 110 }, { "grad_norm": 1.0677911043167114, "learning_rate": 2.38e-06, "loss": 1.2008, "step": 120 }, { "grad_norm": 1.095163345336914, "learning_rate": 2.5800000000000003e-06, "loss": 1.1555, "step": 130 }, { "grad_norm": 1.3630059957504272, "learning_rate": 2.78e-06, "loss": 1.1194, "step": 140 }, { "grad_norm": 3.42504620552063, "learning_rate": 2.9800000000000003e-06, "loss": 1.1111, "step": 150 }, { "grad_norm": 1.026049017906189, "learning_rate": 3.1800000000000005e-06, "loss": 1.0947, "step": 160 }, { "grad_norm": 1.191851019859314, "learning_rate": 3.38e-06, "loss": 1.0886, "step": 170 }, { "grad_norm": 1.1898078918457031, "learning_rate": 3.58e-06, "loss": 1.08, "step": 180 }, { "grad_norm": 0.6012499332427979, "learning_rate": 3.7800000000000002e-06, "loss": 1.077, "step": 190 }, { "grad_norm": 0.8238781094551086, "learning_rate": 3.98e-06, "loss": 1.0753, "step": 200 }, { "grad_norm": 0.540116548538208, "learning_rate": 4.18e-06, "loss": 1.0647, "step": 210 }, { "grad_norm": 0.9415344595909119, "learning_rate": 4.38e-06, "loss": 1.0676, "step": 220 }, { "grad_norm": 0.89017653465271, "learning_rate": 4.58e-06, "loss": 1.0622, "step": 230 }, { "grad_norm": 0.7420381903648376, "learning_rate": 4.780000000000001e-06, "loss": 1.0569, "step": 240 }, { "grad_norm": 0.8182544112205505, "learning_rate": 4.98e-06, "loss": 1.05, "step": 250 }, { "grad_norm": 0.6720947027206421, "learning_rate": 5.18e-06, "loss": 1.051, "step": 260 }, { "grad_norm": 1.5585936307907104, "learning_rate": 5.38e-06, "loss": 1.0434, "step": 270 }, { "grad_norm": 0.6578641533851624, "learning_rate": 5.580000000000001e-06, "loss": 1.0375, "step": 280 }, { "grad_norm": 0.7728487253189087, "learning_rate": 5.78e-06, "loss": 1.024, "step": 290 }, { "grad_norm": 1.449583888053894, "learning_rate": 5.98e-06, "loss": 1.0029, "step": 300 }, { "grad_norm": 1.1525822877883911, "learning_rate": 6.18e-06, "loss": 0.9609, "step": 310 }, { "grad_norm": 1.2075045108795166, "learning_rate": 6.38e-06, "loss": 0.9196, "step": 320 }, { "grad_norm": 1.1753730773925781, "learning_rate": 6.58e-06, "loss": 0.8844, "step": 330 }, { "grad_norm": 1.2754079103469849, "learning_rate": 6.78e-06, "loss": 0.8467, "step": 340 }, { "grad_norm": 0.9708346128463745, "learning_rate": 6.98e-06, "loss": 0.7829, "step": 350 }, { "grad_norm": 1.2854881286621094, "learning_rate": 7.180000000000001e-06, "loss": 0.7257, "step": 360 }, { "grad_norm": 1.6023991107940674, "learning_rate": 7.3800000000000005e-06, "loss": 0.6628, "step": 370 }, { "grad_norm": 1.5819013118743896, "learning_rate": 7.580000000000001e-06, "loss": 0.6122, "step": 380 }, { "grad_norm": 1.6555638313293457, "learning_rate": 7.78e-06, "loss": 0.5657, "step": 390 }, { "grad_norm": 1.231196403503418, "learning_rate": 7.98e-06, "loss": 0.5183, "step": 400 }, { "grad_norm": 1.254557728767395, "learning_rate": 8.18e-06, "loss": 0.4797, "step": 410 }, { "grad_norm": 1.516592025756836, "learning_rate": 8.380000000000001e-06, "loss": 0.4349, "step": 420 }, { "grad_norm": 1.3813271522521973, "learning_rate": 8.580000000000001e-06, "loss": 0.4069, "step": 430 }, { "grad_norm": 1.764649510383606, "learning_rate": 8.78e-06, "loss": 0.3635, "step": 440 }, { "grad_norm": 1.669900894165039, "learning_rate": 8.98e-06, "loss": 0.333, "step": 450 }, { "grad_norm": 1.712644338607788, "learning_rate": 9.180000000000002e-06, "loss": 0.3061, "step": 460 }, { "grad_norm": 1.4138727188110352, "learning_rate": 9.38e-06, "loss": 0.2675, "step": 470 }, { "grad_norm": 1.8411630392074585, "learning_rate": 9.58e-06, "loss": 0.2601, "step": 480 }, { "grad_norm": 1.8448125123977661, "learning_rate": 9.78e-06, "loss": 0.2372, "step": 490 }, { "grad_norm": 1.8637194633483887, "learning_rate": 9.980000000000001e-06, "loss": 0.2234, "step": 500 }, { "grad_norm": 1.797864556312561, "learning_rate": 1.018e-05, "loss": 0.1874, "step": 510 }, { "grad_norm": 1.3534449338912964, "learning_rate": 1.038e-05, "loss": 0.1817, "step": 520 }, { "grad_norm": 1.9229480028152466, "learning_rate": 1.058e-05, "loss": 0.1794, "step": 530 }, { "grad_norm": 1.4626076221466064, "learning_rate": 1.0780000000000002e-05, "loss": 0.1659, "step": 540 }, { "grad_norm": 1.5099800825119019, "learning_rate": 1.098e-05, "loss": 0.1488, "step": 550 }, { "grad_norm": 1.4870288372039795, "learning_rate": 1.118e-05, "loss": 0.1584, "step": 560 }, { "grad_norm": 1.3798811435699463, "learning_rate": 1.1380000000000001e-05, "loss": 0.159, "step": 570 }, { "grad_norm": 1.4397133588790894, "learning_rate": 1.1580000000000001e-05, "loss": 0.1458, "step": 580 }, { "grad_norm": 1.3889820575714111, "learning_rate": 1.178e-05, "loss": 0.1351, "step": 590 }, { "grad_norm": 1.7743124961853027, "learning_rate": 1.198e-05, "loss": 0.1448, "step": 600 }, { "grad_norm": 1.4062581062316895, "learning_rate": 1.2180000000000002e-05, "loss": 0.1317, "step": 610 }, { "grad_norm": 1.3358649015426636, "learning_rate": 1.238e-05, "loss": 0.1267, "step": 620 }, { "grad_norm": 1.4395185708999634, "learning_rate": 1.258e-05, "loss": 0.1237, "step": 630 }, { "grad_norm": 1.6668097972869873, "learning_rate": 1.278e-05, "loss": 0.1297, "step": 640 }, { "grad_norm": 1.4436836242675781, "learning_rate": 1.2980000000000001e-05, "loss": 0.124, "step": 650 }, { "grad_norm": 1.1975443363189697, "learning_rate": 1.3180000000000001e-05, "loss": 0.1231, "step": 660 }, { "grad_norm": 2.0019848346710205, "learning_rate": 1.338e-05, "loss": 0.1145, "step": 670 }, { "grad_norm": 1.42374587059021, "learning_rate": 1.358e-05, "loss": 0.1142, "step": 680 }, { "grad_norm": 1.6116094589233398, "learning_rate": 1.3780000000000002e-05, "loss": 0.1051, "step": 690 }, { "grad_norm": 1.3906396627426147, "learning_rate": 1.3980000000000002e-05, "loss": 0.104, "step": 700 }, { "grad_norm": 1.4092786312103271, "learning_rate": 1.4180000000000001e-05, "loss": 0.1013, "step": 710 }, { "grad_norm": 1.2959115505218506, "learning_rate": 1.4380000000000001e-05, "loss": 0.1049, "step": 720 }, { "grad_norm": 1.355329155921936, "learning_rate": 1.4580000000000003e-05, "loss": 0.1011, "step": 730 }, { "grad_norm": 1.497562289237976, "learning_rate": 1.4779999999999999e-05, "loss": 0.1052, "step": 740 }, { "grad_norm": 1.5350676774978638, "learning_rate": 1.4979999999999999e-05, "loss": 0.0998, "step": 750 }, { "grad_norm": 1.654044508934021, "learning_rate": 1.518e-05, "loss": 0.0958, "step": 760 }, { "grad_norm": 1.3451182842254639, "learning_rate": 1.538e-05, "loss": 0.1013, "step": 770 }, { "grad_norm": 1.3671021461486816, "learning_rate": 1.558e-05, "loss": 0.095, "step": 780 }, { "grad_norm": 1.393040418624878, "learning_rate": 1.578e-05, "loss": 0.0856, "step": 790 }, { "grad_norm": 1.595890760421753, "learning_rate": 1.598e-05, "loss": 0.0954, "step": 800 }, { "grad_norm": 1.4862169027328491, "learning_rate": 1.618e-05, "loss": 0.0856, "step": 810 }, { "grad_norm": 1.5412952899932861, "learning_rate": 1.6380000000000002e-05, "loss": 0.0857, "step": 820 }, { "grad_norm": 1.2183696031570435, "learning_rate": 1.658e-05, "loss": 0.0851, "step": 830 }, { "grad_norm": 1.3772952556610107, "learning_rate": 1.6780000000000002e-05, "loss": 0.0891, "step": 840 }, { "grad_norm": 1.6511870622634888, "learning_rate": 1.698e-05, "loss": 0.088, "step": 850 }, { "grad_norm": 1.4692013263702393, "learning_rate": 1.718e-05, "loss": 0.0855, "step": 860 }, { "grad_norm": 1.5530446767807007, "learning_rate": 1.7380000000000003e-05, "loss": 0.0817, "step": 870 }, { "grad_norm": 1.7136114835739136, "learning_rate": 1.758e-05, "loss": 0.0843, "step": 880 }, { "grad_norm": 1.1873081922531128, "learning_rate": 1.7780000000000003e-05, "loss": 0.0808, "step": 890 }, { "grad_norm": 1.3465912342071533, "learning_rate": 1.798e-05, "loss": 0.0837, "step": 900 }, { "grad_norm": 1.3957197666168213, "learning_rate": 1.818e-05, "loss": 0.0761, "step": 910 }, { "grad_norm": 1.6671689748764038, "learning_rate": 1.838e-05, "loss": 0.0747, "step": 920 }, { "grad_norm": 1.2347415685653687, "learning_rate": 1.858e-05, "loss": 0.0818, "step": 930 }, { "grad_norm": 1.215998888015747, "learning_rate": 1.878e-05, "loss": 0.0765, "step": 940 }, { "grad_norm": 1.2381305694580078, "learning_rate": 1.898e-05, "loss": 0.0783, "step": 950 }, { "grad_norm": 1.4403188228607178, "learning_rate": 1.918e-05, "loss": 0.0764, "step": 960 }, { "grad_norm": 1.4157181978225708, "learning_rate": 1.938e-05, "loss": 0.0718, "step": 970 }, { "grad_norm": 1.732481837272644, "learning_rate": 1.9580000000000002e-05, "loss": 0.0736, "step": 980 }, { "grad_norm": 1.798630714416504, "learning_rate": 1.978e-05, "loss": 0.0861, "step": 990 }, { "grad_norm": 1.0528844594955444, "learning_rate": 1.9980000000000002e-05, "loss": 0.0719, "step": 1000 }, { "grad_norm": 1.1852469444274902, "learning_rate": 2.0180000000000003e-05, "loss": 0.0699, "step": 1010 }, { "grad_norm": 1.2871661186218262, "learning_rate": 2.038e-05, "loss": 0.0782, "step": 1020 }, { "grad_norm": 1.3425966501235962, "learning_rate": 2.0580000000000003e-05, "loss": 0.0669, "step": 1030 }, { "grad_norm": 1.4823933839797974, "learning_rate": 2.078e-05, "loss": 0.0622, "step": 1040 }, { "grad_norm": 0.9770628213882446, "learning_rate": 2.098e-05, "loss": 0.0639, "step": 1050 }, { "grad_norm": 1.0395350456237793, "learning_rate": 2.118e-05, "loss": 0.0629, "step": 1060 }, { "grad_norm": 1.4697076082229614, "learning_rate": 2.138e-05, "loss": 0.067, "step": 1070 }, { "grad_norm": 1.2311962842941284, "learning_rate": 2.158e-05, "loss": 0.0702, "step": 1080 }, { "grad_norm": 1.4791337251663208, "learning_rate": 2.178e-05, "loss": 0.0627, "step": 1090 }, { "grad_norm": 1.5390149354934692, "learning_rate": 2.198e-05, "loss": 0.0592, "step": 1100 }, { "grad_norm": 1.1831820011138916, "learning_rate": 2.218e-05, "loss": 0.0586, "step": 1110 }, { "grad_norm": 1.1605489253997803, "learning_rate": 2.2380000000000003e-05, "loss": 0.0626, "step": 1120 }, { "grad_norm": 1.102067470550537, "learning_rate": 2.258e-05, "loss": 0.067, "step": 1130 }, { "grad_norm": 1.3766251802444458, "learning_rate": 2.2780000000000002e-05, "loss": 0.0635, "step": 1140 }, { "grad_norm": 1.4389090538024902, "learning_rate": 2.298e-05, "loss": 0.062, "step": 1150 }, { "grad_norm": 1.2968324422836304, "learning_rate": 2.318e-05, "loss": 0.0649, "step": 1160 }, { "grad_norm": 1.2080411911010742, "learning_rate": 2.3380000000000003e-05, "loss": 0.065, "step": 1170 }, { "grad_norm": 1.3099950551986694, "learning_rate": 2.358e-05, "loss": 0.0642, "step": 1180 }, { "grad_norm": 1.6132113933563232, "learning_rate": 2.3780000000000003e-05, "loss": 0.0626, "step": 1190 }, { "grad_norm": 1.4161043167114258, "learning_rate": 2.398e-05, "loss": 0.0664, "step": 1200 }, { "grad_norm": 0.8655836582183838, "learning_rate": 2.418e-05, "loss": 0.0564, "step": 1210 }, { "grad_norm": 1.1864510774612427, "learning_rate": 2.438e-05, "loss": 0.0606, "step": 1220 }, { "grad_norm": 1.4448095560073853, "learning_rate": 2.4580000000000002e-05, "loss": 0.0652, "step": 1230 }, { "grad_norm": 1.5856877565383911, "learning_rate": 2.478e-05, "loss": 0.0612, "step": 1240 }, { "grad_norm": 0.865655779838562, "learning_rate": 2.498e-05, "loss": 0.0552, "step": 1250 }, { "grad_norm": 1.2102159261703491, "learning_rate": 2.5180000000000003e-05, "loss": 0.0583, "step": 1260 }, { "grad_norm": 1.194739818572998, "learning_rate": 2.5380000000000004e-05, "loss": 0.0576, "step": 1270 }, { "grad_norm": 1.2699246406555176, "learning_rate": 2.5580000000000002e-05, "loss": 0.0622, "step": 1280 }, { "grad_norm": 1.154442310333252, "learning_rate": 2.5779999999999997e-05, "loss": 0.058, "step": 1290 }, { "grad_norm": 1.3044975996017456, "learning_rate": 2.598e-05, "loss": 0.0565, "step": 1300 }, { "grad_norm": 0.8417702913284302, "learning_rate": 2.618e-05, "loss": 0.0567, "step": 1310 }, { "grad_norm": 1.3872730731964111, "learning_rate": 2.6379999999999998e-05, "loss": 0.0559, "step": 1320 }, { "grad_norm": 1.711880087852478, "learning_rate": 2.658e-05, "loss": 0.0655, "step": 1330 }, { "grad_norm": 1.1182583570480347, "learning_rate": 2.678e-05, "loss": 0.0626, "step": 1340 }, { "grad_norm": 1.1944183111190796, "learning_rate": 2.698e-05, "loss": 0.0581, "step": 1350 }, { "grad_norm": 1.6460931301116943, "learning_rate": 2.718e-05, "loss": 0.0565, "step": 1360 }, { "grad_norm": 1.294204592704773, "learning_rate": 2.738e-05, "loss": 0.0541, "step": 1370 }, { "grad_norm": 1.1357831954956055, "learning_rate": 2.758e-05, "loss": 0.048, "step": 1380 }, { "grad_norm": 1.165892243385315, "learning_rate": 2.778e-05, "loss": 0.0523, "step": 1390 }, { "grad_norm": 1.0497643947601318, "learning_rate": 2.798e-05, "loss": 0.0516, "step": 1400 }, { "grad_norm": 1.4770041704177856, "learning_rate": 2.818e-05, "loss": 0.056, "step": 1410 }, { "grad_norm": 0.9275290369987488, "learning_rate": 2.8380000000000003e-05, "loss": 0.0582, "step": 1420 }, { "grad_norm": 1.1991444826126099, "learning_rate": 2.858e-05, "loss": 0.0564, "step": 1430 }, { "grad_norm": 1.1844748258590698, "learning_rate": 2.8780000000000002e-05, "loss": 0.0563, "step": 1440 }, { "grad_norm": 1.6104141473770142, "learning_rate": 2.898e-05, "loss": 0.0582, "step": 1450 }, { "grad_norm": 1.2599174976348877, "learning_rate": 2.9180000000000002e-05, "loss": 0.0639, "step": 1460 }, { "grad_norm": 1.138422966003418, "learning_rate": 2.9380000000000003e-05, "loss": 0.054, "step": 1470 }, { "grad_norm": 1.2027629613876343, "learning_rate": 2.958e-05, "loss": 0.0571, "step": 1480 }, { "grad_norm": 1.1384162902832031, "learning_rate": 2.9780000000000003e-05, "loss": 0.0553, "step": 1490 }, { "grad_norm": 1.2283378839492798, "learning_rate": 2.998e-05, "loss": 0.0515, "step": 1500 }, { "grad_norm": 1.0853252410888672, "learning_rate": 3.0180000000000002e-05, "loss": 0.0568, "step": 1510 }, { "grad_norm": 1.4757615327835083, "learning_rate": 3.0380000000000004e-05, "loss": 0.0512, "step": 1520 }, { "grad_norm": 1.1209266185760498, "learning_rate": 3.058e-05, "loss": 0.0526, "step": 1530 }, { "grad_norm": 1.4083342552185059, "learning_rate": 3.078e-05, "loss": 0.0528, "step": 1540 }, { "grad_norm": 1.017566442489624, "learning_rate": 3.0980000000000005e-05, "loss": 0.052, "step": 1550 }, { "grad_norm": 0.8404308557510376, "learning_rate": 3.118e-05, "loss": 0.0503, "step": 1560 }, { "grad_norm": 1.0298374891281128, "learning_rate": 3.138e-05, "loss": 0.0502, "step": 1570 }, { "grad_norm": 1.3170708417892456, "learning_rate": 3.1580000000000006e-05, "loss": 0.0483, "step": 1580 }, { "grad_norm": 0.9448978304862976, "learning_rate": 3.1780000000000004e-05, "loss": 0.0491, "step": 1590 }, { "grad_norm": 1.4348145723342896, "learning_rate": 3.198e-05, "loss": 0.052, "step": 1600 }, { "grad_norm": 1.0491156578063965, "learning_rate": 3.218e-05, "loss": 0.0472, "step": 1610 }, { "grad_norm": 1.3141669034957886, "learning_rate": 3.238e-05, "loss": 0.0452, "step": 1620 }, { "grad_norm": 0.9949411153793335, "learning_rate": 3.2579999999999996e-05, "loss": 0.0507, "step": 1630 }, { "grad_norm": 1.1432642936706543, "learning_rate": 3.278e-05, "loss": 0.0456, "step": 1640 }, { "grad_norm": 1.2709522247314453, "learning_rate": 3.298e-05, "loss": 0.0473, "step": 1650 }, { "grad_norm": 0.9455405473709106, "learning_rate": 3.318e-05, "loss": 0.0489, "step": 1660 }, { "grad_norm": 0.9978872537612915, "learning_rate": 3.338e-05, "loss": 0.0477, "step": 1670 }, { "grad_norm": 1.0148335695266724, "learning_rate": 3.358e-05, "loss": 0.0506, "step": 1680 }, { "grad_norm": 1.1567243337631226, "learning_rate": 3.378e-05, "loss": 0.0597, "step": 1690 }, { "grad_norm": 0.8598361611366272, "learning_rate": 3.398e-05, "loss": 0.0518, "step": 1700 }, { "grad_norm": 0.8911224007606506, "learning_rate": 3.418e-05, "loss": 0.047, "step": 1710 }, { "grad_norm": 0.9990471601486206, "learning_rate": 3.438e-05, "loss": 0.047, "step": 1720 }, { "grad_norm": 0.9072572588920593, "learning_rate": 3.4580000000000004e-05, "loss": 0.0478, "step": 1730 }, { "grad_norm": 1.1190738677978516, "learning_rate": 3.478e-05, "loss": 0.0457, "step": 1740 }, { "grad_norm": 0.9113914370536804, "learning_rate": 3.498e-05, "loss": 0.0496, "step": 1750 }, { "grad_norm": 0.9672986268997192, "learning_rate": 3.518e-05, "loss": 0.0472, "step": 1760 }, { "grad_norm": 1.1792596578598022, "learning_rate": 3.5380000000000003e-05, "loss": 0.0538, "step": 1770 }, { "grad_norm": 0.7285626530647278, "learning_rate": 3.558e-05, "loss": 0.0499, "step": 1780 }, { "grad_norm": 1.0047026872634888, "learning_rate": 3.578e-05, "loss": 0.0477, "step": 1790 }, { "grad_norm": 0.9896399974822998, "learning_rate": 3.5980000000000004e-05, "loss": 0.0444, "step": 1800 }, { "grad_norm": 1.09730064868927, "learning_rate": 3.618e-05, "loss": 0.0465, "step": 1810 }, { "grad_norm": 1.1739137172698975, "learning_rate": 3.638e-05, "loss": 0.0513, "step": 1820 }, { "grad_norm": 1.043871283531189, "learning_rate": 3.6580000000000006e-05, "loss": 0.048, "step": 1830 }, { "grad_norm": 1.003370761871338, "learning_rate": 3.6780000000000004e-05, "loss": 0.0462, "step": 1840 }, { "grad_norm": 0.9374337792396545, "learning_rate": 3.698e-05, "loss": 0.0447, "step": 1850 }, { "grad_norm": 1.005472183227539, "learning_rate": 3.7180000000000007e-05, "loss": 0.0515, "step": 1860 }, { "grad_norm": 0.7386491298675537, "learning_rate": 3.7380000000000005e-05, "loss": 0.0472, "step": 1870 }, { "grad_norm": 0.9624290466308594, "learning_rate": 3.758e-05, "loss": 0.044, "step": 1880 }, { "grad_norm": 0.9921140670776367, "learning_rate": 3.778000000000001e-05, "loss": 0.0449, "step": 1890 }, { "grad_norm": 0.8156343102455139, "learning_rate": 3.7980000000000006e-05, "loss": 0.0436, "step": 1900 }, { "grad_norm": 1.156186580657959, "learning_rate": 3.818e-05, "loss": 0.0424, "step": 1910 }, { "grad_norm": 0.9208691120147705, "learning_rate": 3.838e-05, "loss": 0.044, "step": 1920 }, { "grad_norm": 1.131684422492981, "learning_rate": 3.858e-05, "loss": 0.0452, "step": 1930 }, { "grad_norm": 0.8476451635360718, "learning_rate": 3.878e-05, "loss": 0.0452, "step": 1940 }, { "grad_norm": 0.899394690990448, "learning_rate": 3.898e-05, "loss": 0.0443, "step": 1950 }, { "grad_norm": 1.182052493095398, "learning_rate": 3.918e-05, "loss": 0.0448, "step": 1960 }, { "grad_norm": 0.9176322221755981, "learning_rate": 3.938e-05, "loss": 0.0445, "step": 1970 }, { "grad_norm": 0.824549674987793, "learning_rate": 3.958e-05, "loss": 0.0444, "step": 1980 }, { "grad_norm": 1.0005178451538086, "learning_rate": 3.978e-05, "loss": 0.0461, "step": 1990 }, { "grad_norm": 0.9171388745307922, "learning_rate": 3.998e-05, "loss": 0.0445, "step": 2000 }, { "grad_norm": 1.0091439485549927, "learning_rate": 4.018e-05, "loss": 0.0384, "step": 2010 }, { "grad_norm": 0.9137564301490784, "learning_rate": 4.038e-05, "loss": 0.0451, "step": 2020 }, { "grad_norm": 0.9268901348114014, "learning_rate": 4.058e-05, "loss": 0.042, "step": 2030 }, { "grad_norm": 0.7605022192001343, "learning_rate": 4.078e-05, "loss": 0.0422, "step": 2040 }, { "grad_norm": 0.962983250617981, "learning_rate": 4.0980000000000004e-05, "loss": 0.0423, "step": 2050 }, { "grad_norm": 0.9885984063148499, "learning_rate": 4.118e-05, "loss": 0.042, "step": 2060 }, { "grad_norm": 1.0472898483276367, "learning_rate": 4.138e-05, "loss": 0.0413, "step": 2070 }, { "grad_norm": 0.9535574316978455, "learning_rate": 4.1580000000000005e-05, "loss": 0.0408, "step": 2080 }, { "grad_norm": 0.9931361675262451, "learning_rate": 4.178e-05, "loss": 0.046, "step": 2090 }, { "grad_norm": 0.8580541014671326, "learning_rate": 4.198e-05, "loss": 0.0441, "step": 2100 }, { "grad_norm": 1.1096606254577637, "learning_rate": 4.2180000000000006e-05, "loss": 0.04, "step": 2110 }, { "grad_norm": 0.7948083281517029, "learning_rate": 4.2380000000000004e-05, "loss": 0.0405, "step": 2120 }, { "grad_norm": 0.865608274936676, "learning_rate": 4.258e-05, "loss": 0.0433, "step": 2130 }, { "grad_norm": 0.8211621642112732, "learning_rate": 4.278e-05, "loss": 0.0379, "step": 2140 }, { "grad_norm": 0.7556817531585693, "learning_rate": 4.2980000000000005e-05, "loss": 0.0356, "step": 2150 }, { "grad_norm": 0.6994597911834717, "learning_rate": 4.318e-05, "loss": 0.034, "step": 2160 }, { "grad_norm": 0.9436121582984924, "learning_rate": 4.338e-05, "loss": 0.0379, "step": 2170 }, { "grad_norm": 0.8988975882530212, "learning_rate": 4.3580000000000006e-05, "loss": 0.0391, "step": 2180 }, { "grad_norm": 0.8228256106376648, "learning_rate": 4.3780000000000004e-05, "loss": 0.0399, "step": 2190 }, { "grad_norm": 1.3318077325820923, "learning_rate": 4.398e-05, "loss": 0.05, "step": 2200 }, { "grad_norm": 1.2019710540771484, "learning_rate": 4.418000000000001e-05, "loss": 0.0505, "step": 2210 }, { "grad_norm": 1.1239675283432007, "learning_rate": 4.438e-05, "loss": 0.0533, "step": 2220 }, { "grad_norm": 0.8025117516517639, "learning_rate": 4.458e-05, "loss": 0.0416, "step": 2230 }, { "grad_norm": 1.1386997699737549, "learning_rate": 4.478e-05, "loss": 0.0392, "step": 2240 }, { "grad_norm": 0.6321112513542175, "learning_rate": 4.498e-05, "loss": 0.0378, "step": 2250 }, { "grad_norm": 0.7859548330307007, "learning_rate": 4.518e-05, "loss": 0.0395, "step": 2260 }, { "grad_norm": 0.6838705539703369, "learning_rate": 4.538e-05, "loss": 0.0361, "step": 2270 }, { "grad_norm": 0.8296784162521362, "learning_rate": 4.558e-05, "loss": 0.0355, "step": 2280 }, { "grad_norm": 0.7238032221794128, "learning_rate": 4.578e-05, "loss": 0.0367, "step": 2290 }, { "grad_norm": 0.9613937735557556, "learning_rate": 4.5980000000000004e-05, "loss": 0.0354, "step": 2300 }, { "grad_norm": 0.5165014266967773, "learning_rate": 4.618e-05, "loss": 0.0377, "step": 2310 }, { "grad_norm": 0.690712034702301, "learning_rate": 4.638e-05, "loss": 0.035, "step": 2320 }, { "grad_norm": 0.7384291291236877, "learning_rate": 4.6580000000000005e-05, "loss": 0.0388, "step": 2330 }, { "grad_norm": 1.0140254497528076, "learning_rate": 4.678e-05, "loss": 0.0322, "step": 2340 }, { "grad_norm": 0.6200680732727051, "learning_rate": 4.698e-05, "loss": 0.0408, "step": 2350 }, { "grad_norm": 0.7198208570480347, "learning_rate": 4.718e-05, "loss": 0.0384, "step": 2360 }, { "grad_norm": 0.7433923482894897, "learning_rate": 4.7380000000000004e-05, "loss": 0.0347, "step": 2370 }, { "grad_norm": 0.8213379979133606, "learning_rate": 4.758e-05, "loss": 0.0375, "step": 2380 }, { "grad_norm": 0.7975394129753113, "learning_rate": 4.778e-05, "loss": 0.0366, "step": 2390 }, { "grad_norm": 0.9348559379577637, "learning_rate": 4.7980000000000005e-05, "loss": 0.0379, "step": 2400 }, { "grad_norm": 0.7708485722541809, "learning_rate": 4.818e-05, "loss": 0.0327, "step": 2410 }, { "grad_norm": 0.5926417112350464, "learning_rate": 4.838e-05, "loss": 0.0388, "step": 2420 }, { "grad_norm": 1.0047569274902344, "learning_rate": 4.8580000000000006e-05, "loss": 0.039, "step": 2430 }, { "grad_norm": 0.6719374060630798, "learning_rate": 4.8780000000000004e-05, "loss": 0.0406, "step": 2440 }, { "grad_norm": 0.7119841575622559, "learning_rate": 4.898e-05, "loss": 0.036, "step": 2450 }, { "grad_norm": 0.8891400694847107, "learning_rate": 4.918000000000001e-05, "loss": 0.032, "step": 2460 }, { "grad_norm": 0.7229396104812622, "learning_rate": 4.9380000000000005e-05, "loss": 0.0359, "step": 2470 }, { "grad_norm": 0.4989955723285675, "learning_rate": 4.958e-05, "loss": 0.0317, "step": 2480 }, { "grad_norm": 0.8929285407066345, "learning_rate": 4.978e-05, "loss": 0.0424, "step": 2490 }, { "grad_norm": 0.6892445087432861, "learning_rate": 4.9980000000000006e-05, "loss": 0.0391, "step": 2500 }, { "grad_norm": 0.6316903233528137, "learning_rate": 5.0180000000000004e-05, "loss": 0.0378, "step": 2510 }, { "grad_norm": 0.7365928888320923, "learning_rate": 5.038e-05, "loss": 0.0358, "step": 2520 }, { "grad_norm": 0.6830778121948242, "learning_rate": 5.058000000000001e-05, "loss": 0.0368, "step": 2530 }, { "grad_norm": 0.7513113617897034, "learning_rate": 5.0780000000000005e-05, "loss": 0.0332, "step": 2540 }, { "grad_norm": 0.5417002439498901, "learning_rate": 5.098e-05, "loss": 0.0345, "step": 2550 }, { "grad_norm": 0.6839695572853088, "learning_rate": 5.118000000000001e-05, "loss": 0.0377, "step": 2560 }, { "grad_norm": 0.7591220140457153, "learning_rate": 5.1380000000000006e-05, "loss": 0.0322, "step": 2570 }, { "grad_norm": 0.6824414134025574, "learning_rate": 5.1580000000000004e-05, "loss": 0.0343, "step": 2580 }, { "grad_norm": 0.8412578701972961, "learning_rate": 5.178000000000001e-05, "loss": 0.0335, "step": 2590 }, { "grad_norm": 1.0605273246765137, "learning_rate": 5.198000000000001e-05, "loss": 0.0375, "step": 2600 }, { "grad_norm": 0.7210973501205444, "learning_rate": 5.2180000000000005e-05, "loss": 0.0361, "step": 2610 }, { "grad_norm": 0.7081848978996277, "learning_rate": 5.238000000000001e-05, "loss": 0.0353, "step": 2620 }, { "grad_norm": 0.9603499174118042, "learning_rate": 5.258000000000001e-05, "loss": 0.035, "step": 2630 }, { "grad_norm": 0.9469519257545471, "learning_rate": 5.2780000000000006e-05, "loss": 0.041, "step": 2640 }, { "grad_norm": 0.708386242389679, "learning_rate": 5.2980000000000004e-05, "loss": 0.0394, "step": 2650 }, { "grad_norm": 0.8074994683265686, "learning_rate": 5.318000000000001e-05, "loss": 0.0369, "step": 2660 }, { "grad_norm": 0.7278133034706116, "learning_rate": 5.338000000000001e-05, "loss": 0.0371, "step": 2670 }, { "grad_norm": 0.7524117231369019, "learning_rate": 5.3580000000000005e-05, "loss": 0.0394, "step": 2680 }, { "grad_norm": 0.644686758518219, "learning_rate": 5.378e-05, "loss": 0.0375, "step": 2690 }, { "grad_norm": 0.7016997933387756, "learning_rate": 5.3979999999999995e-05, "loss": 0.0304, "step": 2700 }, { "grad_norm": 0.44945383071899414, "learning_rate": 5.418e-05, "loss": 0.0287, "step": 2710 }, { "grad_norm": 0.6435315012931824, "learning_rate": 5.438e-05, "loss": 0.0293, "step": 2720 }, { "grad_norm": 0.560971736907959, "learning_rate": 5.4579999999999996e-05, "loss": 0.0314, "step": 2730 }, { "grad_norm": 0.9628281593322754, "learning_rate": 5.478e-05, "loss": 0.0333, "step": 2740 }, { "grad_norm": 0.7214536070823669, "learning_rate": 5.498e-05, "loss": 0.0297, "step": 2750 }, { "grad_norm": 0.6018235087394714, "learning_rate": 5.518e-05, "loss": 0.0312, "step": 2760 }, { "grad_norm": 0.652439296245575, "learning_rate": 5.538e-05, "loss": 0.0328, "step": 2770 }, { "grad_norm": 0.8577497005462646, "learning_rate": 5.558e-05, "loss": 0.0319, "step": 2780 }, { "grad_norm": 0.6926858425140381, "learning_rate": 5.578e-05, "loss": 0.0333, "step": 2790 }, { "grad_norm": 0.8299049735069275, "learning_rate": 5.5979999999999996e-05, "loss": 0.0367, "step": 2800 }, { "grad_norm": 0.5393622517585754, "learning_rate": 5.618e-05, "loss": 0.034, "step": 2810 }, { "grad_norm": 0.5885865092277527, "learning_rate": 5.638e-05, "loss": 0.0321, "step": 2820 }, { "grad_norm": 0.5534732341766357, "learning_rate": 5.658e-05, "loss": 0.0298, "step": 2830 }, { "grad_norm": 0.9866615533828735, "learning_rate": 5.678e-05, "loss": 0.0318, "step": 2840 }, { "grad_norm": 0.6535489559173584, "learning_rate": 5.698e-05, "loss": 0.0353, "step": 2850 }, { "grad_norm": 0.6989291310310364, "learning_rate": 5.718e-05, "loss": 0.0363, "step": 2860 }, { "grad_norm": 0.6578271985054016, "learning_rate": 5.738e-05, "loss": 0.035, "step": 2870 }, { "grad_norm": 0.7954059839248657, "learning_rate": 5.758e-05, "loss": 0.0308, "step": 2880 }, { "grad_norm": 0.9117987155914307, "learning_rate": 5.778e-05, "loss": 0.0357, "step": 2890 }, { "grad_norm": 0.706235945224762, "learning_rate": 5.7980000000000004e-05, "loss": 0.0339, "step": 2900 }, { "grad_norm": 0.572581946849823, "learning_rate": 5.818e-05, "loss": 0.0294, "step": 2910 }, { "grad_norm": 0.6471242308616638, "learning_rate": 5.838e-05, "loss": 0.0307, "step": 2920 }, { "grad_norm": 0.5479424595832825, "learning_rate": 5.858e-05, "loss": 0.0347, "step": 2930 }, { "grad_norm": 0.5997476577758789, "learning_rate": 5.878e-05, "loss": 0.0369, "step": 2940 }, { "grad_norm": 0.7462559938430786, "learning_rate": 5.898e-05, "loss": 0.0302, "step": 2950 }, { "grad_norm": 0.5933098196983337, "learning_rate": 5.918e-05, "loss": 0.0298, "step": 2960 }, { "grad_norm": 0.6737980246543884, "learning_rate": 5.9380000000000004e-05, "loss": 0.0331, "step": 2970 }, { "grad_norm": 0.6923305988311768, "learning_rate": 5.958e-05, "loss": 0.0328, "step": 2980 }, { "grad_norm": 0.5705406665802002, "learning_rate": 5.978e-05, "loss": 0.0302, "step": 2990 }, { "grad_norm": 0.6358248591423035, "learning_rate": 5.9980000000000005e-05, "loss": 0.0306, "step": 3000 }, { "grad_norm": 0.6332859396934509, "learning_rate": 6.018e-05, "loss": 0.0281, "step": 3010 }, { "grad_norm": 0.7649080157279968, "learning_rate": 6.038e-05, "loss": 0.0297, "step": 3020 }, { "grad_norm": 0.6221539378166199, "learning_rate": 6.0580000000000006e-05, "loss": 0.0326, "step": 3030 }, { "grad_norm": 0.6510138511657715, "learning_rate": 6.0780000000000004e-05, "loss": 0.0334, "step": 3040 }, { "grad_norm": 0.3819681704044342, "learning_rate": 6.098e-05, "loss": 0.0302, "step": 3050 }, { "grad_norm": 0.889785885810852, "learning_rate": 6.118000000000001e-05, "loss": 0.0336, "step": 3060 }, { "grad_norm": 0.5044720768928528, "learning_rate": 6.138e-05, "loss": 0.0303, "step": 3070 }, { "grad_norm": 0.5941210389137268, "learning_rate": 6.158e-05, "loss": 0.0289, "step": 3080 }, { "grad_norm": 0.6734341382980347, "learning_rate": 6.178000000000001e-05, "loss": 0.0313, "step": 3090 }, { "grad_norm": 0.5975148677825928, "learning_rate": 6.198e-05, "loss": 0.0358, "step": 3100 }, { "grad_norm": 0.725938618183136, "learning_rate": 6.218e-05, "loss": 0.0344, "step": 3110 }, { "grad_norm": 0.710146427154541, "learning_rate": 6.238000000000001e-05, "loss": 0.0322, "step": 3120 }, { "grad_norm": 0.5901218056678772, "learning_rate": 6.258e-05, "loss": 0.0367, "step": 3130 }, { "grad_norm": 0.3942064046859741, "learning_rate": 6.278e-05, "loss": 0.032, "step": 3140 }, { "grad_norm": 0.4872140884399414, "learning_rate": 6.298000000000001e-05, "loss": 0.029, "step": 3150 }, { "grad_norm": 0.3403817415237427, "learning_rate": 6.318e-05, "loss": 0.026, "step": 3160 }, { "grad_norm": 0.6652414798736572, "learning_rate": 6.338e-05, "loss": 0.0284, "step": 3170 }, { "grad_norm": 0.7223265171051025, "learning_rate": 6.358000000000001e-05, "loss": 0.0278, "step": 3180 }, { "grad_norm": 0.8326142430305481, "learning_rate": 6.378e-05, "loss": 0.0334, "step": 3190 }, { "grad_norm": 0.7251725792884827, "learning_rate": 6.398000000000001e-05, "loss": 0.0321, "step": 3200 }, { "grad_norm": 0.6617615818977356, "learning_rate": 6.418000000000001e-05, "loss": 0.0307, "step": 3210 }, { "grad_norm": 0.6117438077926636, "learning_rate": 6.438e-05, "loss": 0.0269, "step": 3220 }, { "grad_norm": 0.5717401504516602, "learning_rate": 6.458000000000001e-05, "loss": 0.0244, "step": 3230 }, { "grad_norm": 0.6399633288383484, "learning_rate": 6.478000000000001e-05, "loss": 0.029, "step": 3240 }, { "grad_norm": 0.5776907801628113, "learning_rate": 6.498e-05, "loss": 0.0252, "step": 3250 }, { "grad_norm": 0.5947620272636414, "learning_rate": 6.518000000000001e-05, "loss": 0.0288, "step": 3260 }, { "grad_norm": 0.6413024663925171, "learning_rate": 6.538000000000001e-05, "loss": 0.0268, "step": 3270 }, { "grad_norm": 0.637661337852478, "learning_rate": 6.558e-05, "loss": 0.0371, "step": 3280 }, { "grad_norm": 0.7051257491111755, "learning_rate": 6.578000000000001e-05, "loss": 0.0377, "step": 3290 }, { "grad_norm": 0.47957292199134827, "learning_rate": 6.598e-05, "loss": 0.0295, "step": 3300 }, { "grad_norm": 0.6549534201622009, "learning_rate": 6.618e-05, "loss": 0.0301, "step": 3310 }, { "grad_norm": 0.49720513820648193, "learning_rate": 6.638e-05, "loss": 0.0287, "step": 3320 }, { "grad_norm": 0.4570195972919464, "learning_rate": 6.658e-05, "loss": 0.0241, "step": 3330 }, { "grad_norm": 0.9223978519439697, "learning_rate": 6.678e-05, "loss": 0.0303, "step": 3340 }, { "grad_norm": 0.698764979839325, "learning_rate": 6.698e-05, "loss": 0.0343, "step": 3350 }, { "grad_norm": 0.6458147764205933, "learning_rate": 6.718e-05, "loss": 0.0307, "step": 3360 }, { "grad_norm": 0.5980828404426575, "learning_rate": 6.738e-05, "loss": 0.0339, "step": 3370 }, { "grad_norm": 0.6140214204788208, "learning_rate": 6.758e-05, "loss": 0.0317, "step": 3380 }, { "grad_norm": 0.6344859004020691, "learning_rate": 6.778e-05, "loss": 0.0293, "step": 3390 }, { "grad_norm": 0.7402243614196777, "learning_rate": 6.798e-05, "loss": 0.0341, "step": 3400 }, { "grad_norm": 0.5136579871177673, "learning_rate": 6.818e-05, "loss": 0.0287, "step": 3410 }, { "grad_norm": 0.598612368106842, "learning_rate": 6.838e-05, "loss": 0.0281, "step": 3420 }, { "grad_norm": 0.6137620806694031, "learning_rate": 6.858e-05, "loss": 0.0278, "step": 3430 }, { "grad_norm": 0.5715351104736328, "learning_rate": 6.878e-05, "loss": 0.0331, "step": 3440 }, { "grad_norm": 0.46528691053390503, "learning_rate": 6.898e-05, "loss": 0.0278, "step": 3450 }, { "grad_norm": 0.524677574634552, "learning_rate": 6.918e-05, "loss": 0.0301, "step": 3460 }, { "grad_norm": 0.45118075609207153, "learning_rate": 6.938e-05, "loss": 0.0229, "step": 3470 }, { "grad_norm": 0.600976288318634, "learning_rate": 6.958e-05, "loss": 0.0286, "step": 3480 }, { "grad_norm": 0.5734143257141113, "learning_rate": 6.978e-05, "loss": 0.0288, "step": 3490 }, { "grad_norm": 0.496779203414917, "learning_rate": 6.998e-05, "loss": 0.0279, "step": 3500 }, { "grad_norm": 0.4380475580692291, "learning_rate": 7.018e-05, "loss": 0.0258, "step": 3510 }, { "grad_norm": 0.5873051285743713, "learning_rate": 7.038e-05, "loss": 0.029, "step": 3520 }, { "grad_norm": 0.49277955293655396, "learning_rate": 7.058e-05, "loss": 0.028, "step": 3530 }, { "grad_norm": 0.5317379236221313, "learning_rate": 7.078e-05, "loss": 0.0241, "step": 3540 }, { "grad_norm": 0.5515751242637634, "learning_rate": 7.098e-05, "loss": 0.0254, "step": 3550 }, { "grad_norm": 0.5748841166496277, "learning_rate": 7.118e-05, "loss": 0.0287, "step": 3560 }, { "grad_norm": 0.4087177813053131, "learning_rate": 7.138e-05, "loss": 0.0281, "step": 3570 }, { "grad_norm": 0.4325161278247833, "learning_rate": 7.158e-05, "loss": 0.0306, "step": 3580 }, { "grad_norm": 0.6810705661773682, "learning_rate": 7.178000000000001e-05, "loss": 0.0257, "step": 3590 }, { "grad_norm": 0.4843975007534027, "learning_rate": 7.198e-05, "loss": 0.0295, "step": 3600 }, { "grad_norm": 0.48425501585006714, "learning_rate": 7.218e-05, "loss": 0.0245, "step": 3610 }, { "grad_norm": 0.6250986456871033, "learning_rate": 7.238000000000001e-05, "loss": 0.0257, "step": 3620 }, { "grad_norm": 0.5939698815345764, "learning_rate": 7.258e-05, "loss": 0.0281, "step": 3630 }, { "grad_norm": 0.4309736490249634, "learning_rate": 7.278e-05, "loss": 0.024, "step": 3640 }, { "grad_norm": 0.4115956425666809, "learning_rate": 7.298000000000001e-05, "loss": 0.0235, "step": 3650 }, { "grad_norm": 0.3764101564884186, "learning_rate": 7.318e-05, "loss": 0.0249, "step": 3660 }, { "grad_norm": 0.5296189785003662, "learning_rate": 7.338e-05, "loss": 0.0251, "step": 3670 }, { "grad_norm": 0.438326895236969, "learning_rate": 7.358000000000001e-05, "loss": 0.0234, "step": 3680 }, { "grad_norm": 0.6391661167144775, "learning_rate": 7.378e-05, "loss": 0.0276, "step": 3690 }, { "grad_norm": 0.41852325201034546, "learning_rate": 7.398e-05, "loss": 0.0278, "step": 3700 }, { "grad_norm": 0.42059171199798584, "learning_rate": 7.418000000000001e-05, "loss": 0.0248, "step": 3710 }, { "grad_norm": 0.30289891362190247, "learning_rate": 7.438e-05, "loss": 0.0226, "step": 3720 }, { "grad_norm": 0.43075835704803467, "learning_rate": 7.458000000000001e-05, "loss": 0.0239, "step": 3730 }, { "grad_norm": 0.556752622127533, "learning_rate": 7.478e-05, "loss": 0.0243, "step": 3740 }, { "grad_norm": 0.6728338003158569, "learning_rate": 7.498e-05, "loss": 0.0308, "step": 3750 }, { "grad_norm": 0.6405932307243347, "learning_rate": 7.518000000000001e-05, "loss": 0.0293, "step": 3760 }, { "grad_norm": 0.4937228858470917, "learning_rate": 7.538e-05, "loss": 0.0305, "step": 3770 }, { "grad_norm": 0.474018394947052, "learning_rate": 7.558e-05, "loss": 0.0248, "step": 3780 }, { "grad_norm": 0.5468189120292664, "learning_rate": 7.578000000000001e-05, "loss": 0.0284, "step": 3790 }, { "grad_norm": 0.538583517074585, "learning_rate": 7.598e-05, "loss": 0.0245, "step": 3800 }, { "grad_norm": 0.42049941420555115, "learning_rate": 7.618e-05, "loss": 0.0204, "step": 3810 }, { "grad_norm": 0.5444878935813904, "learning_rate": 7.638000000000001e-05, "loss": 0.0225, "step": 3820 }, { "grad_norm": 0.45845529437065125, "learning_rate": 7.658e-05, "loss": 0.0257, "step": 3830 }, { "grad_norm": 0.4994812607765198, "learning_rate": 7.678000000000001e-05, "loss": 0.0275, "step": 3840 }, { "grad_norm": 0.5147393345832825, "learning_rate": 7.698000000000001e-05, "loss": 0.0248, "step": 3850 }, { "grad_norm": 0.4168397784233093, "learning_rate": 7.718e-05, "loss": 0.025, "step": 3860 }, { "grad_norm": 0.39506784081459045, "learning_rate": 7.738000000000001e-05, "loss": 0.0258, "step": 3870 }, { "grad_norm": 0.5629037618637085, "learning_rate": 7.758000000000001e-05, "loss": 0.0265, "step": 3880 }, { "grad_norm": 0.4680747985839844, "learning_rate": 7.778e-05, "loss": 0.0247, "step": 3890 }, { "grad_norm": 0.4062015116214752, "learning_rate": 7.798000000000001e-05, "loss": 0.0253, "step": 3900 }, { "grad_norm": 0.617353618144989, "learning_rate": 7.818000000000001e-05, "loss": 0.0281, "step": 3910 }, { "grad_norm": 0.5497764945030212, "learning_rate": 7.838e-05, "loss": 0.0287, "step": 3920 }, { "grad_norm": 0.45421963930130005, "learning_rate": 7.858000000000001e-05, "loss": 0.0236, "step": 3930 }, { "grad_norm": 0.3541388511657715, "learning_rate": 7.878e-05, "loss": 0.0224, "step": 3940 }, { "grad_norm": 0.34936946630477905, "learning_rate": 7.897999999999999e-05, "loss": 0.0206, "step": 3950 }, { "grad_norm": 0.4011842906475067, "learning_rate": 7.918e-05, "loss": 0.0215, "step": 3960 }, { "grad_norm": 0.5677523016929626, "learning_rate": 7.938e-05, "loss": 0.0247, "step": 3970 }, { "grad_norm": 0.5341632962226868, "learning_rate": 7.958e-05, "loss": 0.0306, "step": 3980 }, { "grad_norm": 0.6608514785766602, "learning_rate": 7.978e-05, "loss": 0.0314, "step": 3990 }, { "grad_norm": 0.6420552134513855, "learning_rate": 7.998e-05, "loss": 0.0259, "step": 4000 }, { "grad_norm": 0.5039737820625305, "learning_rate": 8.018e-05, "loss": 0.0281, "step": 4010 }, { "grad_norm": 0.4083591401576996, "learning_rate": 8.038e-05, "loss": 0.026, "step": 4020 }, { "grad_norm": 0.504033625125885, "learning_rate": 8.058e-05, "loss": 0.0259, "step": 4030 }, { "grad_norm": 0.4180271029472351, "learning_rate": 8.078e-05, "loss": 0.0223, "step": 4040 }, { "grad_norm": 0.3613048791885376, "learning_rate": 8.098e-05, "loss": 0.0193, "step": 4050 }, { "grad_norm": 0.4639596939086914, "learning_rate": 8.118e-05, "loss": 0.0225, "step": 4060 }, { "grad_norm": 0.6144746541976929, "learning_rate": 8.138e-05, "loss": 0.0255, "step": 4070 }, { "grad_norm": 0.49204498529434204, "learning_rate": 8.158e-05, "loss": 0.0262, "step": 4080 }, { "grad_norm": 0.6307779550552368, "learning_rate": 8.178e-05, "loss": 0.027, "step": 4090 }, { "grad_norm": 0.46632054448127747, "learning_rate": 8.198e-05, "loss": 0.0255, "step": 4100 }, { "grad_norm": 0.37334367632865906, "learning_rate": 8.218e-05, "loss": 0.0254, "step": 4110 }, { "grad_norm": 0.341823935508728, "learning_rate": 8.238000000000001e-05, "loss": 0.0246, "step": 4120 }, { "grad_norm": 0.3904557526111603, "learning_rate": 8.258e-05, "loss": 0.0202, "step": 4130 }, { "grad_norm": 0.49292951822280884, "learning_rate": 8.278e-05, "loss": 0.0249, "step": 4140 }, { "grad_norm": 0.4197901785373688, "learning_rate": 8.298000000000001e-05, "loss": 0.0213, "step": 4150 }, { "grad_norm": 0.5803514122962952, "learning_rate": 8.318e-05, "loss": 0.0233, "step": 4160 }, { "grad_norm": 0.3577316999435425, "learning_rate": 8.338e-05, "loss": 0.0205, "step": 4170 }, { "grad_norm": 0.6489518880844116, "learning_rate": 8.358e-05, "loss": 0.0213, "step": 4180 }, { "grad_norm": 0.39904916286468506, "learning_rate": 8.378e-05, "loss": 0.0227, "step": 4190 }, { "grad_norm": 0.5226292014122009, "learning_rate": 8.398e-05, "loss": 0.0256, "step": 4200 }, { "grad_norm": 0.46362513303756714, "learning_rate": 8.418e-05, "loss": 0.0269, "step": 4210 }, { "grad_norm": 0.4689826965332031, "learning_rate": 8.438e-05, "loss": 0.024, "step": 4220 }, { "grad_norm": 0.5248680710792542, "learning_rate": 8.458e-05, "loss": 0.02, "step": 4230 }, { "grad_norm": 0.48570483922958374, "learning_rate": 8.478e-05, "loss": 0.0228, "step": 4240 }, { "grad_norm": 0.4977204203605652, "learning_rate": 8.498e-05, "loss": 0.0204, "step": 4250 }, { "grad_norm": 0.3681754767894745, "learning_rate": 8.518000000000001e-05, "loss": 0.0212, "step": 4260 }, { "grad_norm": 0.5219210982322693, "learning_rate": 8.538e-05, "loss": 0.022, "step": 4270 }, { "grad_norm": 0.5552477836608887, "learning_rate": 8.558e-05, "loss": 0.0237, "step": 4280 }, { "grad_norm": 0.48745694756507874, "learning_rate": 8.578000000000001e-05, "loss": 0.0259, "step": 4290 }, { "grad_norm": 0.34131917357444763, "learning_rate": 8.598e-05, "loss": 0.0227, "step": 4300 }, { "grad_norm": 0.4231133460998535, "learning_rate": 8.618e-05, "loss": 0.0238, "step": 4310 }, { "grad_norm": 0.4362500309944153, "learning_rate": 8.638000000000001e-05, "loss": 0.0213, "step": 4320 }, { "grad_norm": 0.3767835199832916, "learning_rate": 8.658e-05, "loss": 0.0215, "step": 4330 }, { "grad_norm": 0.5135570168495178, "learning_rate": 8.678e-05, "loss": 0.0218, "step": 4340 }, { "grad_norm": 0.4377560019493103, "learning_rate": 8.698000000000001e-05, "loss": 0.0206, "step": 4350 }, { "grad_norm": 0.4707851707935333, "learning_rate": 8.718e-05, "loss": 0.02, "step": 4360 }, { "grad_norm": 0.4994048476219177, "learning_rate": 8.738000000000001e-05, "loss": 0.0229, "step": 4370 }, { "grad_norm": 0.4044944941997528, "learning_rate": 8.758000000000001e-05, "loss": 0.0245, "step": 4380 }, { "grad_norm": 0.4314010441303253, "learning_rate": 8.778e-05, "loss": 0.0222, "step": 4390 }, { "grad_norm": 0.3293844759464264, "learning_rate": 8.798000000000001e-05, "loss": 0.022, "step": 4400 }, { "grad_norm": 0.44393235445022583, "learning_rate": 8.818000000000001e-05, "loss": 0.0244, "step": 4410 }, { "grad_norm": 0.4528575837612152, "learning_rate": 8.838e-05, "loss": 0.0205, "step": 4420 }, { "grad_norm": 0.38441580533981323, "learning_rate": 8.858000000000001e-05, "loss": 0.0237, "step": 4430 }, { "grad_norm": 0.38463616371154785, "learning_rate": 8.878000000000001e-05, "loss": 0.0218, "step": 4440 }, { "grad_norm": 0.38466885685920715, "learning_rate": 8.898e-05, "loss": 0.023, "step": 4450 }, { "grad_norm": 0.31920406222343445, "learning_rate": 8.918000000000001e-05, "loss": 0.0179, "step": 4460 }, { "grad_norm": 0.4025540351867676, "learning_rate": 8.938e-05, "loss": 0.0214, "step": 4470 }, { "grad_norm": 0.49419400095939636, "learning_rate": 8.958e-05, "loss": 0.0238, "step": 4480 }, { "grad_norm": 0.4537307918071747, "learning_rate": 8.978000000000001e-05, "loss": 0.0223, "step": 4490 }, { "grad_norm": 0.3611532151699066, "learning_rate": 8.998e-05, "loss": 0.0226, "step": 4500 }, { "grad_norm": 0.33598077297210693, "learning_rate": 9.018000000000001e-05, "loss": 0.0183, "step": 4510 }, { "grad_norm": 0.42014771699905396, "learning_rate": 9.038000000000001e-05, "loss": 0.0196, "step": 4520 }, { "grad_norm": 0.4670145511627197, "learning_rate": 9.058e-05, "loss": 0.0232, "step": 4530 }, { "grad_norm": 0.37349048256874084, "learning_rate": 9.078000000000001e-05, "loss": 0.0219, "step": 4540 }, { "grad_norm": 0.42039257287979126, "learning_rate": 9.098000000000001e-05, "loss": 0.0243, "step": 4550 }, { "grad_norm": 0.8203719854354858, "learning_rate": 9.118e-05, "loss": 0.0311, "step": 4560 }, { "grad_norm": 0.5355848073959351, "learning_rate": 9.138e-05, "loss": 0.0346, "step": 4570 }, { "grad_norm": 0.4511075019836426, "learning_rate": 9.158e-05, "loss": 0.0329, "step": 4580 }, { "grad_norm": 0.465806782245636, "learning_rate": 9.178e-05, "loss": 0.0271, "step": 4590 }, { "grad_norm": 0.37722817063331604, "learning_rate": 9.198e-05, "loss": 0.0228, "step": 4600 }, { "grad_norm": 0.6056216359138489, "learning_rate": 9.218e-05, "loss": 0.0248, "step": 4610 }, { "grad_norm": 0.4560282826423645, "learning_rate": 9.238e-05, "loss": 0.0238, "step": 4620 }, { "grad_norm": 0.5303015112876892, "learning_rate": 9.258e-05, "loss": 0.025, "step": 4630 }, { "grad_norm": 0.45912060141563416, "learning_rate": 9.278e-05, "loss": 0.0238, "step": 4640 }, { "grad_norm": 0.49959132075309753, "learning_rate": 9.298e-05, "loss": 0.0263, "step": 4650 }, { "grad_norm": 0.41596072912216187, "learning_rate": 9.318e-05, "loss": 0.0285, "step": 4660 }, { "grad_norm": 0.5346922874450684, "learning_rate": 9.338e-05, "loss": 0.031, "step": 4670 }, { "grad_norm": 0.4807937443256378, "learning_rate": 9.358e-05, "loss": 0.0254, "step": 4680 }, { "grad_norm": 0.47110870480537415, "learning_rate": 9.378e-05, "loss": 0.0297, "step": 4690 }, { "grad_norm": 0.4722379446029663, "learning_rate": 9.398e-05, "loss": 0.0262, "step": 4700 }, { "grad_norm": 0.5071490406990051, "learning_rate": 9.418e-05, "loss": 0.0263, "step": 4710 }, { "grad_norm": 0.45311617851257324, "learning_rate": 9.438e-05, "loss": 0.0341, "step": 4720 }, { "grad_norm": 0.4925312101840973, "learning_rate": 9.458e-05, "loss": 0.0292, "step": 4730 }, { "grad_norm": 0.4019123613834381, "learning_rate": 9.478e-05, "loss": 0.0264, "step": 4740 }, { "grad_norm": 0.5312362313270569, "learning_rate": 9.498e-05, "loss": 0.0286, "step": 4750 }, { "grad_norm": 0.503860592842102, "learning_rate": 9.518000000000001e-05, "loss": 0.0294, "step": 4760 }, { "grad_norm": 0.33204928040504456, "learning_rate": 9.538e-05, "loss": 0.0273, "step": 4770 }, { "grad_norm": 0.3185102045536041, "learning_rate": 9.558e-05, "loss": 0.028, "step": 4780 }, { "grad_norm": 0.4000436067581177, "learning_rate": 9.578000000000001e-05, "loss": 0.0269, "step": 4790 }, { "grad_norm": 0.576178789138794, "learning_rate": 9.598e-05, "loss": 0.0265, "step": 4800 }, { "grad_norm": 0.4806734323501587, "learning_rate": 9.618e-05, "loss": 0.0255, "step": 4810 }, { "grad_norm": 0.3837417960166931, "learning_rate": 9.638000000000001e-05, "loss": 0.0285, "step": 4820 }, { "grad_norm": 0.4284833073616028, "learning_rate": 9.658e-05, "loss": 0.0314, "step": 4830 }, { "grad_norm": 0.46648043394088745, "learning_rate": 9.678e-05, "loss": 0.0331, "step": 4840 }, { "grad_norm": 0.40389078855514526, "learning_rate": 9.698000000000001e-05, "loss": 0.0243, "step": 4850 }, { "grad_norm": 0.46299025416374207, "learning_rate": 9.718e-05, "loss": 0.0317, "step": 4860 }, { "grad_norm": 0.4533017873764038, "learning_rate": 9.738e-05, "loss": 0.029, "step": 4870 }, { "grad_norm": 0.4882132411003113, "learning_rate": 9.758000000000001e-05, "loss": 0.0259, "step": 4880 }, { "grad_norm": 0.625675618648529, "learning_rate": 9.778e-05, "loss": 0.0227, "step": 4890 }, { "grad_norm": 0.4292561411857605, "learning_rate": 9.798000000000001e-05, "loss": 0.0311, "step": 4900 }, { "grad_norm": 0.48107782006263733, "learning_rate": 9.818000000000001e-05, "loss": 0.0285, "step": 4910 }, { "grad_norm": 0.3697425425052643, "learning_rate": 9.838e-05, "loss": 0.0281, "step": 4920 }, { "grad_norm": 0.37744635343551636, "learning_rate": 9.858000000000001e-05, "loss": 0.0263, "step": 4930 }, { "grad_norm": 0.32905906438827515, "learning_rate": 9.878e-05, "loss": 0.0226, "step": 4940 }, { "grad_norm": 0.586070716381073, "learning_rate": 9.898e-05, "loss": 0.0239, "step": 4950 }, { "grad_norm": 0.5445890426635742, "learning_rate": 9.918000000000001e-05, "loss": 0.0292, "step": 4960 }, { "grad_norm": 0.5319381356239319, "learning_rate": 9.938e-05, "loss": 0.0303, "step": 4970 }, { "grad_norm": 0.4946310520172119, "learning_rate": 9.958e-05, "loss": 0.0318, "step": 4980 }, { "grad_norm": 0.396771639585495, "learning_rate": 9.978000000000001e-05, "loss": 0.0293, "step": 4990 }, { "grad_norm": 0.5106689929962158, "learning_rate": 9.998e-05, "loss": 0.0288, "step": 5000 }, { "grad_norm": 0.46806883811950684, "learning_rate": 9.999999778549045e-05, "loss": 0.0262, "step": 5010 }, { "grad_norm": 0.49085932970046997, "learning_rate": 9.999999013039593e-05, "loss": 0.0284, "step": 5020 }, { "grad_norm": 0.30878397822380066, "learning_rate": 9.999997700737766e-05, "loss": 0.0229, "step": 5030 }, { "grad_norm": 0.3396780788898468, "learning_rate": 9.999995841643709e-05, "loss": 0.0242, "step": 5040 }, { "grad_norm": 0.4874900281429291, "learning_rate": 9.999993435757623e-05, "loss": 0.0215, "step": 5050 }, { "grad_norm": 0.6320798397064209, "learning_rate": 9.999990483079773e-05, "loss": 0.0246, "step": 5060 }, { "grad_norm": 0.5191620588302612, "learning_rate": 9.999986983610481e-05, "loss": 0.0251, "step": 5070 }, { "grad_norm": 0.46208977699279785, "learning_rate": 9.99998293735013e-05, "loss": 0.0253, "step": 5080 }, { "grad_norm": 0.47997140884399414, "learning_rate": 9.999978344299161e-05, "loss": 0.0231, "step": 5090 }, { "grad_norm": 0.3160593509674072, "learning_rate": 9.99997320445808e-05, "loss": 0.0239, "step": 5100 }, { "grad_norm": 0.3570175766944885, "learning_rate": 9.999967517827444e-05, "loss": 0.0333, "step": 5110 }, { "grad_norm": 0.3414735794067383, "learning_rate": 9.999961284407879e-05, "loss": 0.0262, "step": 5120 }, { "grad_norm": 0.32839253544807434, "learning_rate": 9.999954504200067e-05, "loss": 0.0232, "step": 5130 }, { "grad_norm": 0.4136643707752228, "learning_rate": 9.999947177204744e-05, "loss": 0.0258, "step": 5140 }, { "grad_norm": 0.49542784690856934, "learning_rate": 9.999939303422718e-05, "loss": 0.0307, "step": 5150 }, { "grad_norm": 0.3329619765281677, "learning_rate": 9.999930882854847e-05, "loss": 0.0321, "step": 5160 }, { "grad_norm": 0.39945438504219055, "learning_rate": 9.999921915502051e-05, "loss": 0.0239, "step": 5170 }, { "grad_norm": 0.3511650860309601, "learning_rate": 9.99991240136531e-05, "loss": 0.0202, "step": 5180 }, { "grad_norm": 0.36695775389671326, "learning_rate": 9.999902340445668e-05, "loss": 0.0211, "step": 5190 }, { "grad_norm": 0.3836784362792969, "learning_rate": 9.999891732744224e-05, "loss": 0.0211, "step": 5200 }, { "grad_norm": 0.42754170298576355, "learning_rate": 9.999880578262135e-05, "loss": 0.024, "step": 5210 }, { "grad_norm": 0.26776957511901855, "learning_rate": 9.999868877000624e-05, "loss": 0.0228, "step": 5220 }, { "grad_norm": 0.2709443271160126, "learning_rate": 9.99985662896097e-05, "loss": 0.0197, "step": 5230 }, { "grad_norm": 0.26034829020500183, "learning_rate": 9.999843834144513e-05, "loss": 0.0198, "step": 5240 }, { "grad_norm": 0.2985207736492157, "learning_rate": 9.99983049255265e-05, "loss": 0.022, "step": 5250 }, { "grad_norm": 0.4415627717971802, "learning_rate": 9.999816604186843e-05, "loss": 0.0264, "step": 5260 }, { "grad_norm": 0.5872815847396851, "learning_rate": 9.999802169048609e-05, "loss": 0.0303, "step": 5270 }, { "grad_norm": 0.4163254201412201, "learning_rate": 9.999787187139527e-05, "loss": 0.0236, "step": 5280 }, { "grad_norm": 0.27182602882385254, "learning_rate": 9.999771658461234e-05, "loss": 0.0196, "step": 5290 }, { "grad_norm": 0.34723490476608276, "learning_rate": 9.999755583015431e-05, "loss": 0.0204, "step": 5300 }, { "grad_norm": 0.35240939259529114, "learning_rate": 9.999738960803874e-05, "loss": 0.0268, "step": 5310 }, { "grad_norm": 0.3901340961456299, "learning_rate": 9.99972179182838e-05, "loss": 0.0261, "step": 5320 }, { "grad_norm": 0.39656302332878113, "learning_rate": 9.99970407609083e-05, "loss": 0.0218, "step": 5330 }, { "grad_norm": 0.3107569217681885, "learning_rate": 9.999685813593159e-05, "loss": 0.0233, "step": 5340 }, { "grad_norm": 0.3312112092971802, "learning_rate": 9.999667004337362e-05, "loss": 0.0199, "step": 5350 }, { "grad_norm": 0.4755190312862396, "learning_rate": 9.9996476483255e-05, "loss": 0.0246, "step": 5360 }, { "grad_norm": 0.4241868555545807, "learning_rate": 9.999627745559688e-05, "loss": 0.0237, "step": 5370 }, { "grad_norm": 0.43817847967147827, "learning_rate": 9.999607296042101e-05, "loss": 0.0241, "step": 5380 }, { "grad_norm": 0.4763711392879486, "learning_rate": 9.99958629977498e-05, "loss": 0.021, "step": 5390 }, { "grad_norm": 0.638861358165741, "learning_rate": 9.999564756760615e-05, "loss": 0.0288, "step": 5400 }, { "grad_norm": 0.3937637507915497, "learning_rate": 9.999542667001366e-05, "loss": 0.0289, "step": 5410 }, { "grad_norm": 0.43036088347435, "learning_rate": 9.999520030499647e-05, "loss": 0.0244, "step": 5420 }, { "grad_norm": 0.2977253794670105, "learning_rate": 9.999496847257936e-05, "loss": 0.0225, "step": 5430 }, { "grad_norm": 0.39687034487724304, "learning_rate": 9.999473117278764e-05, "loss": 0.0215, "step": 5440 }, { "grad_norm": 0.3553573489189148, "learning_rate": 9.999448840564731e-05, "loss": 0.0213, "step": 5450 }, { "grad_norm": 0.2944963574409485, "learning_rate": 9.999424017118488e-05, "loss": 0.0184, "step": 5460 }, { "grad_norm": 0.24880331754684448, "learning_rate": 9.999398646942751e-05, "loss": 0.0196, "step": 5470 }, { "grad_norm": 0.285370409488678, "learning_rate": 9.999372730040296e-05, "loss": 0.0227, "step": 5480 }, { "grad_norm": 0.38702696561813354, "learning_rate": 9.999346266413953e-05, "loss": 0.0202, "step": 5490 }, { "grad_norm": 0.4749315083026886, "learning_rate": 9.99931925606662e-05, "loss": 0.0256, "step": 5500 }, { "grad_norm": 0.39953383803367615, "learning_rate": 9.99929169900125e-05, "loss": 0.0207, "step": 5510 }, { "grad_norm": 0.24804310500621796, "learning_rate": 9.999263595220855e-05, "loss": 0.0196, "step": 5520 }, { "grad_norm": 0.325897753238678, "learning_rate": 9.99923494472851e-05, "loss": 0.0169, "step": 5530 }, { "grad_norm": 0.37113210558891296, "learning_rate": 9.999205747527348e-05, "loss": 0.0177, "step": 5540 }, { "grad_norm": 0.34263402223587036, "learning_rate": 9.999176003620561e-05, "loss": 0.0158, "step": 5550 }, { "grad_norm": 0.22056053578853607, "learning_rate": 9.999145713011405e-05, "loss": 0.0185, "step": 5560 }, { "grad_norm": 0.2291717380285263, "learning_rate": 9.999114875703186e-05, "loss": 0.0183, "step": 5570 }, { "grad_norm": 0.3864569365978241, "learning_rate": 9.999083491699281e-05, "loss": 0.0216, "step": 5580 }, { "grad_norm": 0.2557288408279419, "learning_rate": 9.999051561003123e-05, "loss": 0.017, "step": 5590 }, { "grad_norm": 0.4527098834514618, "learning_rate": 9.999019083618202e-05, "loss": 0.0221, "step": 5600 }, { "grad_norm": 0.38070911169052124, "learning_rate": 9.99898605954807e-05, "loss": 0.027, "step": 5610 }, { "grad_norm": 0.33062323927879333, "learning_rate": 9.998952488796338e-05, "loss": 0.0211, "step": 5620 }, { "grad_norm": 0.2425406128168106, "learning_rate": 9.998918371366676e-05, "loss": 0.0215, "step": 5630 }, { "grad_norm": 0.29708749055862427, "learning_rate": 9.99888370726282e-05, "loss": 0.0169, "step": 5640 }, { "grad_norm": 0.3812570571899414, "learning_rate": 9.998848496488556e-05, "loss": 0.0229, "step": 5650 }, { "grad_norm": 0.34488147497177124, "learning_rate": 9.998812739047736e-05, "loss": 0.0209, "step": 5660 }, { "grad_norm": 0.27377405762672424, "learning_rate": 9.99877643494427e-05, "loss": 0.0211, "step": 5670 }, { "grad_norm": 0.34671953320503235, "learning_rate": 9.998739584182128e-05, "loss": 0.0193, "step": 5680 }, { "grad_norm": 0.37238001823425293, "learning_rate": 9.998702186765342e-05, "loss": 0.0206, "step": 5690 }, { "grad_norm": 0.2813030183315277, "learning_rate": 9.998664242698e-05, "loss": 0.0182, "step": 5700 }, { "grad_norm": 0.34642550349235535, "learning_rate": 9.998625751984251e-05, "loss": 0.021, "step": 5710 }, { "grad_norm": 0.25382402539253235, "learning_rate": 9.998586714628307e-05, "loss": 0.0173, "step": 5720 }, { "grad_norm": 0.41643330454826355, "learning_rate": 9.998547130634432e-05, "loss": 0.0253, "step": 5730 }, { "grad_norm": 0.36095646023750305, "learning_rate": 9.99850700000696e-05, "loss": 0.0202, "step": 5740 }, { "grad_norm": 0.45488667488098145, "learning_rate": 9.998466322750278e-05, "loss": 0.0239, "step": 5750 }, { "grad_norm": 0.35828471183776855, "learning_rate": 9.998425098868834e-05, "loss": 0.0218, "step": 5760 }, { "grad_norm": 0.3265201449394226, "learning_rate": 9.998383328367136e-05, "loss": 0.0181, "step": 5770 }, { "grad_norm": 0.38894200325012207, "learning_rate": 9.99834101124975e-05, "loss": 0.0194, "step": 5780 }, { "grad_norm": 0.3059667646884918, "learning_rate": 9.998298147521309e-05, "loss": 0.0157, "step": 5790 }, { "grad_norm": 0.3948608338832855, "learning_rate": 9.998254737186496e-05, "loss": 0.0187, "step": 5800 }, { "grad_norm": 0.40458595752716064, "learning_rate": 9.99821078025006e-05, "loss": 0.0191, "step": 5810 }, { "grad_norm": 0.24252288043498993, "learning_rate": 9.998166276716807e-05, "loss": 0.0192, "step": 5820 }, { "grad_norm": 0.4054071009159088, "learning_rate": 9.998121226591606e-05, "loss": 0.0165, "step": 5830 }, { "grad_norm": 0.27798691391944885, "learning_rate": 9.998075629879382e-05, "loss": 0.0164, "step": 5840 }, { "grad_norm": 0.2566308379173279, "learning_rate": 9.99802948658512e-05, "loss": 0.0174, "step": 5850 }, { "grad_norm": 0.32421156764030457, "learning_rate": 9.99798279671387e-05, "loss": 0.0189, "step": 5860 }, { "grad_norm": 0.26689231395721436, "learning_rate": 9.997935560270734e-05, "loss": 0.0172, "step": 5870 }, { "grad_norm": 0.49600285291671753, "learning_rate": 9.997887777260879e-05, "loss": 0.0241, "step": 5880 }, { "grad_norm": 0.28667038679122925, "learning_rate": 9.997839447689532e-05, "loss": 0.0173, "step": 5890 }, { "grad_norm": 0.31485095620155334, "learning_rate": 9.997790571561978e-05, "loss": 0.0185, "step": 5900 }, { "grad_norm": 0.4640851318836212, "learning_rate": 9.99774114888356e-05, "loss": 0.0218, "step": 5910 }, { "grad_norm": 0.3864789605140686, "learning_rate": 9.997691179659684e-05, "loss": 0.0216, "step": 5920 }, { "grad_norm": 0.5019938349723816, "learning_rate": 9.997640663895815e-05, "loss": 0.0231, "step": 5930 }, { "grad_norm": 0.477200448513031, "learning_rate": 9.997589601597477e-05, "loss": 0.0202, "step": 5940 }, { "grad_norm": 0.2794519364833832, "learning_rate": 9.997537992770252e-05, "loss": 0.0197, "step": 5950 }, { "grad_norm": 0.3585629165172577, "learning_rate": 9.997485837419788e-05, "loss": 0.0188, "step": 5960 }, { "grad_norm": 0.3763025403022766, "learning_rate": 9.997433135551786e-05, "loss": 0.0182, "step": 5970 }, { "grad_norm": 0.22828133404254913, "learning_rate": 9.997379887172009e-05, "loss": 0.0193, "step": 5980 }, { "grad_norm": 0.33693796396255493, "learning_rate": 9.997326092286281e-05, "loss": 0.0177, "step": 5990 }, { "grad_norm": 0.28562435507774353, "learning_rate": 9.997271750900486e-05, "loss": 0.0162, "step": 6000 }, { "grad_norm": 0.29076066613197327, "learning_rate": 9.997216863020565e-05, "loss": 0.0194, "step": 6010 }, { "grad_norm": 0.29150524735450745, "learning_rate": 9.99716142865252e-05, "loss": 0.0174, "step": 6020 }, { "grad_norm": 0.32884663343429565, "learning_rate": 9.997105447802415e-05, "loss": 0.0195, "step": 6030 }, { "grad_norm": 0.2870180606842041, "learning_rate": 9.997048920476373e-05, "loss": 0.0171, "step": 6040 }, { "grad_norm": 0.3591146469116211, "learning_rate": 9.996991846680572e-05, "loss": 0.0169, "step": 6050 }, { "grad_norm": 0.24721670150756836, "learning_rate": 9.996934226421257e-05, "loss": 0.016, "step": 6060 }, { "grad_norm": 0.20441651344299316, "learning_rate": 9.996876059704726e-05, "loss": 0.0142, "step": 6070 }, { "grad_norm": 0.16219273209571838, "learning_rate": 9.996817346537343e-05, "loss": 0.0172, "step": 6080 }, { "grad_norm": 0.23250271379947662, "learning_rate": 9.996758086925526e-05, "loss": 0.0156, "step": 6090 }, { "grad_norm": 0.2682386338710785, "learning_rate": 9.996698280875759e-05, "loss": 0.0147, "step": 6100 }, { "grad_norm": 0.30239683389663696, "learning_rate": 9.99663792839458e-05, "loss": 0.0154, "step": 6110 }, { "grad_norm": 0.31320327520370483, "learning_rate": 9.99657702948859e-05, "loss": 0.0163, "step": 6120 }, { "grad_norm": 0.1810259222984314, "learning_rate": 9.996515584164448e-05, "loss": 0.0166, "step": 6130 }, { "grad_norm": 0.36500754952430725, "learning_rate": 9.996453592428873e-05, "loss": 0.018, "step": 6140 }, { "grad_norm": 0.30559563636779785, "learning_rate": 9.996391054288646e-05, "loss": 0.0162, "step": 6150 }, { "grad_norm": 0.22682185471057892, "learning_rate": 9.996327969750605e-05, "loss": 0.0157, "step": 6160 }, { "grad_norm": 0.2210545688867569, "learning_rate": 9.996264338821649e-05, "loss": 0.0126, "step": 6170 }, { "grad_norm": 0.278077632188797, "learning_rate": 9.996200161508735e-05, "loss": 0.0138, "step": 6180 }, { "grad_norm": 0.31677180528640747, "learning_rate": 9.996135437818885e-05, "loss": 0.0136, "step": 6190 }, { "grad_norm": 0.353982150554657, "learning_rate": 9.996070167759175e-05, "loss": 0.0148, "step": 6200 }, { "grad_norm": 0.4331282377243042, "learning_rate": 9.996004351336743e-05, "loss": 0.0157, "step": 6210 }, { "grad_norm": 0.2910541892051697, "learning_rate": 9.995937988558785e-05, "loss": 0.0166, "step": 6220 }, { "grad_norm": 0.2696931064128876, "learning_rate": 9.995871079432561e-05, "loss": 0.0146, "step": 6230 }, { "grad_norm": 0.34993934631347656, "learning_rate": 9.995803623965389e-05, "loss": 0.0152, "step": 6240 }, { "grad_norm": 0.2799660861492157, "learning_rate": 9.995735622164641e-05, "loss": 0.016, "step": 6250 }, { "grad_norm": 0.4241600036621094, "learning_rate": 9.995667074037758e-05, "loss": 0.0271, "step": 6260 }, { "grad_norm": 0.4713914692401886, "learning_rate": 9.995597979592232e-05, "loss": 0.0205, "step": 6270 }, { "grad_norm": 0.3680015504360199, "learning_rate": 9.995528338835625e-05, "loss": 0.0228, "step": 6280 }, { "grad_norm": 0.30621686577796936, "learning_rate": 9.995458151775547e-05, "loss": 0.0196, "step": 6290 }, { "grad_norm": 0.21515825390815735, "learning_rate": 9.995387418419677e-05, "loss": 0.0148, "step": 6300 }, { "grad_norm": 0.32048699259757996, "learning_rate": 9.99531613877575e-05, "loss": 0.0157, "step": 6310 }, { "grad_norm": 0.3358544409275055, "learning_rate": 9.995244312851559e-05, "loss": 0.0159, "step": 6320 }, { "grad_norm": 0.2292274385690689, "learning_rate": 9.995171940654961e-05, "loss": 0.015, "step": 6330 }, { "grad_norm": 0.22032791376113892, "learning_rate": 9.995099022193871e-05, "loss": 0.0147, "step": 6340 }, { "grad_norm": 0.3017629086971283, "learning_rate": 9.995025557476261e-05, "loss": 0.0173, "step": 6350 }, { "grad_norm": 0.2223232090473175, "learning_rate": 9.994951546510165e-05, "loss": 0.0134, "step": 6360 }, { "grad_norm": 0.4480077624320984, "learning_rate": 9.994876989303679e-05, "loss": 0.0154, "step": 6370 }, { "grad_norm": 0.32045069336891174, "learning_rate": 9.994801885864955e-05, "loss": 0.0175, "step": 6380 }, { "grad_norm": 0.28558456897735596, "learning_rate": 9.994726236202205e-05, "loss": 0.0211, "step": 6390 }, { "grad_norm": 0.28010696172714233, "learning_rate": 9.994650040323704e-05, "loss": 0.0194, "step": 6400 }, { "grad_norm": 0.3362777531147003, "learning_rate": 9.994573298237784e-05, "loss": 0.0186, "step": 6410 }, { "grad_norm": 0.25897857546806335, "learning_rate": 9.994496009952837e-05, "loss": 0.0166, "step": 6420 }, { "grad_norm": 0.5184109210968018, "learning_rate": 9.994418175477316e-05, "loss": 0.0164, "step": 6430 }, { "grad_norm": 0.39059221744537354, "learning_rate": 9.994339794819733e-05, "loss": 0.0186, "step": 6440 }, { "grad_norm": 0.3812291920185089, "learning_rate": 9.994260867988658e-05, "loss": 0.0199, "step": 6450 }, { "grad_norm": 0.5765135884284973, "learning_rate": 9.994181394992723e-05, "loss": 0.0189, "step": 6460 }, { "grad_norm": 0.24981817603111267, "learning_rate": 9.994101375840618e-05, "loss": 0.0181, "step": 6470 }, { "grad_norm": 0.3442535698413849, "learning_rate": 9.994020810541098e-05, "loss": 0.0208, "step": 6480 }, { "grad_norm": 0.21051858365535736, "learning_rate": 9.99393969910297e-05, "loss": 0.02, "step": 6490 }, { "grad_norm": 0.34333816170692444, "learning_rate": 9.993858041535104e-05, "loss": 0.0192, "step": 6500 }, { "grad_norm": 0.384826123714447, "learning_rate": 9.99377583784643e-05, "loss": 0.0212, "step": 6510 }, { "grad_norm": 0.30864208936691284, "learning_rate": 9.993693088045939e-05, "loss": 0.0165, "step": 6520 }, { "grad_norm": 0.5421059727668762, "learning_rate": 9.99360979214268e-05, "loss": 0.0215, "step": 6530 }, { "grad_norm": 0.4012059271335602, "learning_rate": 9.99352595014576e-05, "loss": 0.0203, "step": 6540 }, { "grad_norm": 0.3232114911079407, "learning_rate": 9.993441562064354e-05, "loss": 0.0211, "step": 6550 }, { "grad_norm": 0.2394786775112152, "learning_rate": 9.993356627907685e-05, "loss": 0.0159, "step": 6560 }, { "grad_norm": 0.2593953311443329, "learning_rate": 9.99327114768504e-05, "loss": 0.0165, "step": 6570 }, { "grad_norm": 0.22513937950134277, "learning_rate": 9.99318512140577e-05, "loss": 0.0133, "step": 6580 }, { "grad_norm": 0.2339996099472046, "learning_rate": 9.993098549079284e-05, "loss": 0.018, "step": 6590 }, { "grad_norm": 0.2847902774810791, "learning_rate": 9.993011430715047e-05, "loss": 0.0189, "step": 6600 }, { "grad_norm": 0.30212825536727905, "learning_rate": 9.992923766322586e-05, "loss": 0.0163, "step": 6610 }, { "grad_norm": 0.3168274760246277, "learning_rate": 9.99283555591149e-05, "loss": 0.0182, "step": 6620 }, { "grad_norm": 0.24779583513736725, "learning_rate": 9.992746799491404e-05, "loss": 0.0177, "step": 6630 }, { "grad_norm": 0.2611415386199951, "learning_rate": 9.992657497072033e-05, "loss": 0.0186, "step": 6640 }, { "grad_norm": 0.33824703097343445, "learning_rate": 9.992567648663147e-05, "loss": 0.0201, "step": 6650 }, { "grad_norm": 0.24690550565719604, "learning_rate": 9.992477254274568e-05, "loss": 0.0142, "step": 6660 }, { "grad_norm": 0.28805261850357056, "learning_rate": 9.992386313916183e-05, "loss": 0.0158, "step": 6670 }, { "grad_norm": 0.28322747349739075, "learning_rate": 9.992294827597934e-05, "loss": 0.0162, "step": 6680 }, { "grad_norm": 0.30428674817085266, "learning_rate": 9.992202795329831e-05, "loss": 0.0165, "step": 6690 }, { "grad_norm": 0.24074390530586243, "learning_rate": 9.992110217121936e-05, "loss": 0.0145, "step": 6700 }, { "grad_norm": 0.47108474373817444, "learning_rate": 9.992017092984372e-05, "loss": 0.0175, "step": 6710 }, { "grad_norm": 0.38665956258773804, "learning_rate": 9.991923422927326e-05, "loss": 0.0224, "step": 6720 }, { "grad_norm": 0.27064865827560425, "learning_rate": 9.991829206961037e-05, "loss": 0.0183, "step": 6730 }, { "grad_norm": 0.28373682498931885, "learning_rate": 9.991734445095813e-05, "loss": 0.0159, "step": 6740 }, { "grad_norm": 0.34279295802116394, "learning_rate": 9.991639137342015e-05, "loss": 0.0199, "step": 6750 }, { "grad_norm": 0.2648771405220032, "learning_rate": 9.991543283710064e-05, "loss": 0.0163, "step": 6760 }, { "grad_norm": 0.2809069752693176, "learning_rate": 9.991446884210445e-05, "loss": 0.0179, "step": 6770 }, { "grad_norm": 0.3966277241706848, "learning_rate": 9.9913499388537e-05, "loss": 0.0174, "step": 6780 }, { "grad_norm": 0.3176640272140503, "learning_rate": 9.99125244765043e-05, "loss": 0.0139, "step": 6790 }, { "grad_norm": 0.33510076999664307, "learning_rate": 9.991154410611296e-05, "loss": 0.017, "step": 6800 }, { "grad_norm": 0.25116321444511414, "learning_rate": 9.99105582774702e-05, "loss": 0.0166, "step": 6810 }, { "grad_norm": 0.20127956569194794, "learning_rate": 9.990956699068384e-05, "loss": 0.0161, "step": 6820 }, { "grad_norm": 0.2591211795806885, "learning_rate": 9.990857024586224e-05, "loss": 0.0156, "step": 6830 }, { "grad_norm": 0.25172558426856995, "learning_rate": 9.990756804311446e-05, "loss": 0.0157, "step": 6840 }, { "grad_norm": 0.366792231798172, "learning_rate": 9.990656038255006e-05, "loss": 0.0157, "step": 6850 }, { "grad_norm": 0.28901320695877075, "learning_rate": 9.990554726427926e-05, "loss": 0.0166, "step": 6860 }, { "grad_norm": 0.32565319538116455, "learning_rate": 9.990452868841284e-05, "loss": 0.019, "step": 6870 }, { "grad_norm": 0.32619741559028625, "learning_rate": 9.99035046550622e-05, "loss": 0.0198, "step": 6880 }, { "grad_norm": 0.3044421970844269, "learning_rate": 9.99024751643393e-05, "loss": 0.0184, "step": 6890 }, { "grad_norm": 0.25814566016197205, "learning_rate": 9.990144021635677e-05, "loss": 0.0205, "step": 6900 }, { "grad_norm": 0.3141074776649475, "learning_rate": 9.990039981122775e-05, "loss": 0.0209, "step": 6910 }, { "grad_norm": 0.31739580631256104, "learning_rate": 9.989935394906602e-05, "loss": 0.0194, "step": 6920 }, { "grad_norm": 0.2817141115665436, "learning_rate": 9.989830262998598e-05, "loss": 0.0181, "step": 6930 }, { "grad_norm": 0.3052203357219696, "learning_rate": 9.989724585410259e-05, "loss": 0.0168, "step": 6940 }, { "grad_norm": 0.30307647585868835, "learning_rate": 9.989618362153139e-05, "loss": 0.0179, "step": 6950 }, { "grad_norm": 0.2869158387184143, "learning_rate": 9.989511593238859e-05, "loss": 0.0149, "step": 6960 }, { "grad_norm": 0.2844948172569275, "learning_rate": 9.98940427867909e-05, "loss": 0.017, "step": 6970 }, { "grad_norm": 0.3450637459754944, "learning_rate": 9.989296418485573e-05, "loss": 0.0193, "step": 6980 }, { "grad_norm": 0.23289790749549866, "learning_rate": 9.989188012670101e-05, "loss": 0.0141, "step": 6990 }, { "grad_norm": 0.19724862277507782, "learning_rate": 9.989079061244528e-05, "loss": 0.0138, "step": 7000 }, { "grad_norm": 0.27938422560691833, "learning_rate": 9.988969564220769e-05, "loss": 0.0156, "step": 7010 }, { "grad_norm": 0.22075650095939636, "learning_rate": 9.988859521610801e-05, "loss": 0.0145, "step": 7020 }, { "grad_norm": 0.2845039367675781, "learning_rate": 9.988748933426656e-05, "loss": 0.0178, "step": 7030 }, { "grad_norm": 0.4067605137825012, "learning_rate": 9.988637799680428e-05, "loss": 0.0173, "step": 7040 }, { "grad_norm": 0.29925036430358887, "learning_rate": 9.98852612038427e-05, "loss": 0.0151, "step": 7050 }, { "grad_norm": 0.24178756773471832, "learning_rate": 9.988413895550397e-05, "loss": 0.0135, "step": 7060 }, { "grad_norm": 0.2422114461660385, "learning_rate": 9.98830112519108e-05, "loss": 0.0156, "step": 7070 }, { "grad_norm": 0.21958957612514496, "learning_rate": 9.98818780931865e-05, "loss": 0.0144, "step": 7080 }, { "grad_norm": 0.2880597710609436, "learning_rate": 9.988073947945502e-05, "loss": 0.0168, "step": 7090 }, { "grad_norm": 0.30587267875671387, "learning_rate": 9.987959541084087e-05, "loss": 0.0149, "step": 7100 }, { "grad_norm": 0.2394077628850937, "learning_rate": 9.987844588746915e-05, "loss": 0.0184, "step": 7110 }, { "grad_norm": 0.2826347053050995, "learning_rate": 9.987729090946558e-05, "loss": 0.0163, "step": 7120 }, { "grad_norm": 0.25536051392555237, "learning_rate": 9.987613047695647e-05, "loss": 0.0169, "step": 7130 }, { "grad_norm": 0.3303028345108032, "learning_rate": 9.987496459006871e-05, "loss": 0.013, "step": 7140 }, { "grad_norm": 0.19895286858081818, "learning_rate": 9.987379324892982e-05, "loss": 0.0139, "step": 7150 }, { "grad_norm": 0.236467644572258, "learning_rate": 9.987261645366788e-05, "loss": 0.0159, "step": 7160 }, { "grad_norm": 0.2825152277946472, "learning_rate": 9.987143420441158e-05, "loss": 0.0165, "step": 7170 }, { "grad_norm": 0.25952351093292236, "learning_rate": 9.987024650129022e-05, "loss": 0.0173, "step": 7180 }, { "grad_norm": 0.34721726179122925, "learning_rate": 9.986905334443368e-05, "loss": 0.0162, "step": 7190 }, { "grad_norm": 0.2785630524158478, "learning_rate": 9.986785473397245e-05, "loss": 0.0169, "step": 7200 }, { "grad_norm": 0.20103847980499268, "learning_rate": 9.98666506700376e-05, "loss": 0.0147, "step": 7210 }, { "grad_norm": 0.3180676996707916, "learning_rate": 9.986544115276081e-05, "loss": 0.0118, "step": 7220 }, { "grad_norm": 0.39567193388938904, "learning_rate": 9.986422618227433e-05, "loss": 0.0201, "step": 7230 }, { "grad_norm": 0.3105962872505188, "learning_rate": 9.986300575871106e-05, "loss": 0.024, "step": 7240 }, { "grad_norm": 0.2861376404762268, "learning_rate": 9.986177988220444e-05, "loss": 0.016, "step": 7250 }, { "grad_norm": 0.28855660557746887, "learning_rate": 9.986054855288856e-05, "loss": 0.0142, "step": 7260 }, { "grad_norm": 0.2561110854148865, "learning_rate": 9.985931177089802e-05, "loss": 0.0196, "step": 7270 }, { "grad_norm": 0.32368502020835876, "learning_rate": 9.985806953636814e-05, "loss": 0.0188, "step": 7280 }, { "grad_norm": 0.33856436610221863, "learning_rate": 9.985682184943471e-05, "loss": 0.017, "step": 7290 }, { "grad_norm": 0.34523308277130127, "learning_rate": 9.98555687102342e-05, "loss": 0.0162, "step": 7300 }, { "grad_norm": 0.26444485783576965, "learning_rate": 9.985431011890367e-05, "loss": 0.0163, "step": 7310 }, { "grad_norm": 0.4086281657218933, "learning_rate": 9.985304607558075e-05, "loss": 0.0179, "step": 7320 }, { "grad_norm": 0.24875298142433167, "learning_rate": 9.985177658040364e-05, "loss": 0.0151, "step": 7330 }, { "grad_norm": 0.24018077552318573, "learning_rate": 9.985050163351119e-05, "loss": 0.0144, "step": 7340 }, { "grad_norm": 0.2860602140426636, "learning_rate": 9.984922123504286e-05, "loss": 0.0139, "step": 7350 }, { "grad_norm": 0.34075498580932617, "learning_rate": 9.984793538513862e-05, "loss": 0.02, "step": 7360 }, { "grad_norm": 0.2717236578464508, "learning_rate": 9.984664408393912e-05, "loss": 0.0177, "step": 7370 }, { "grad_norm": 0.23238734900951385, "learning_rate": 9.984534733158556e-05, "loss": 0.0136, "step": 7380 }, { "grad_norm": 0.30514633655548096, "learning_rate": 9.984404512821977e-05, "loss": 0.0154, "step": 7390 }, { "grad_norm": 0.28789857029914856, "learning_rate": 9.984273747398411e-05, "loss": 0.0128, "step": 7400 }, { "grad_norm": 0.20226159691810608, "learning_rate": 9.984142436902165e-05, "loss": 0.0127, "step": 7410 }, { "grad_norm": 0.21096046268939972, "learning_rate": 9.984010581347596e-05, "loss": 0.013, "step": 7420 }, { "grad_norm": 0.3392312526702881, "learning_rate": 9.983878180749121e-05, "loss": 0.0127, "step": 7430 }, { "grad_norm": 0.33523643016815186, "learning_rate": 9.983745235121222e-05, "loss": 0.013, "step": 7440 }, { "grad_norm": 0.28541359305381775, "learning_rate": 9.983611744478438e-05, "loss": 0.0125, "step": 7450 }, { "grad_norm": 0.24144074320793152, "learning_rate": 9.983477708835365e-05, "loss": 0.0167, "step": 7460 }, { "grad_norm": 0.2597450911998749, "learning_rate": 9.983343128206664e-05, "loss": 0.0139, "step": 7470 }, { "grad_norm": 0.241965189576149, "learning_rate": 9.983208002607049e-05, "loss": 0.0132, "step": 7480 }, { "grad_norm": 0.20115967094898224, "learning_rate": 9.9830723320513e-05, "loss": 0.0126, "step": 7490 }, { "grad_norm": 0.3145284056663513, "learning_rate": 9.982936116554254e-05, "loss": 0.0154, "step": 7500 }, { "grad_norm": 0.22100676596164703, "learning_rate": 9.982799356130803e-05, "loss": 0.0156, "step": 7510 }, { "grad_norm": 0.3120085299015045, "learning_rate": 9.982662050795908e-05, "loss": 0.0162, "step": 7520 }, { "grad_norm": 0.29351988434791565, "learning_rate": 9.982524200564583e-05, "loss": 0.0151, "step": 7530 }, { "grad_norm": 0.2900771498680115, "learning_rate": 9.982385805451901e-05, "loss": 0.0134, "step": 7540 }, { "grad_norm": 0.2849946618080139, "learning_rate": 9.982246865472998e-05, "loss": 0.0152, "step": 7550 }, { "grad_norm": 0.35997143387794495, "learning_rate": 9.982107380643069e-05, "loss": 0.0174, "step": 7560 }, { "grad_norm": 0.3049889802932739, "learning_rate": 9.981967350977368e-05, "loss": 0.0135, "step": 7570 }, { "grad_norm": 0.29618656635284424, "learning_rate": 9.981826776491208e-05, "loss": 0.0136, "step": 7580 }, { "grad_norm": 0.24635356664657593, "learning_rate": 9.98168565719996e-05, "loss": 0.0139, "step": 7590 }, { "grad_norm": 0.2550266683101654, "learning_rate": 9.98154399311906e-05, "loss": 0.0157, "step": 7600 }, { "grad_norm": 0.26798927783966064, "learning_rate": 9.981401784263997e-05, "loss": 0.0131, "step": 7610 }, { "grad_norm": 0.2991575300693512, "learning_rate": 9.981259030650326e-05, "loss": 0.0128, "step": 7620 }, { "grad_norm": 0.2544022798538208, "learning_rate": 9.981115732293655e-05, "loss": 0.0137, "step": 7630 }, { "grad_norm": 0.33998969197273254, "learning_rate": 9.980971889209659e-05, "loss": 0.0129, "step": 7640 }, { "grad_norm": 0.17317301034927368, "learning_rate": 9.980827501414064e-05, "loss": 0.0144, "step": 7650 }, { "grad_norm": 0.2553466260433197, "learning_rate": 9.980682568922663e-05, "loss": 0.0169, "step": 7660 }, { "grad_norm": 0.301842600107193, "learning_rate": 9.980537091751304e-05, "loss": 0.0151, "step": 7670 }, { "grad_norm": 0.21082603931427002, "learning_rate": 9.980391069915897e-05, "loss": 0.013, "step": 7680 }, { "grad_norm": 0.30611371994018555, "learning_rate": 9.98024450343241e-05, "loss": 0.0143, "step": 7690 }, { "grad_norm": 0.5632011294364929, "learning_rate": 9.980097392316872e-05, "loss": 0.0232, "step": 7700 }, { "grad_norm": 0.33049705624580383, "learning_rate": 9.97994973658537e-05, "loss": 0.0199, "step": 7710 }, { "grad_norm": 0.2757904529571533, "learning_rate": 9.979801536254054e-05, "loss": 0.0157, "step": 7720 }, { "grad_norm": 0.26855936646461487, "learning_rate": 9.979652791339127e-05, "loss": 0.0166, "step": 7730 }, { "grad_norm": 0.3247511088848114, "learning_rate": 9.97950350185686e-05, "loss": 0.0143, "step": 7740 }, { "grad_norm": 0.24506399035453796, "learning_rate": 9.979353667823574e-05, "loss": 0.0182, "step": 7750 }, { "grad_norm": 0.2684948742389679, "learning_rate": 9.979203289255658e-05, "loss": 0.0161, "step": 7760 }, { "grad_norm": 0.3632759153842926, "learning_rate": 9.979052366169557e-05, "loss": 0.0193, "step": 7770 }, { "grad_norm": 0.2890929877758026, "learning_rate": 9.978900898581775e-05, "loss": 0.0163, "step": 7780 }, { "grad_norm": 0.2836230397224426, "learning_rate": 9.978748886508875e-05, "loss": 0.0144, "step": 7790 }, { "grad_norm": 0.2146635353565216, "learning_rate": 9.978596329967484e-05, "loss": 0.0141, "step": 7800 }, { "grad_norm": 0.24431434273719788, "learning_rate": 9.978443228974284e-05, "loss": 0.0134, "step": 7810 }, { "grad_norm": 0.2838689982891083, "learning_rate": 9.978289583546015e-05, "loss": 0.0113, "step": 7820 }, { "grad_norm": 0.40135136246681213, "learning_rate": 9.978135393699484e-05, "loss": 0.0168, "step": 7830 }, { "grad_norm": 0.2383560836315155, "learning_rate": 9.977980659451548e-05, "loss": 0.0147, "step": 7840 }, { "grad_norm": 0.3966743052005768, "learning_rate": 9.977825380819135e-05, "loss": 0.0151, "step": 7850 }, { "grad_norm": 0.28650596737861633, "learning_rate": 9.97766955781922e-05, "loss": 0.0163, "step": 7860 }, { "grad_norm": 0.30796000361442566, "learning_rate": 9.977513190468848e-05, "loss": 0.0146, "step": 7870 }, { "grad_norm": 0.29928943514823914, "learning_rate": 9.977356278785116e-05, "loss": 0.0143, "step": 7880 }, { "grad_norm": 0.2598132789134979, "learning_rate": 9.977198822785184e-05, "loss": 0.0151, "step": 7890 }, { "grad_norm": 0.3376540541648865, "learning_rate": 9.977040822486273e-05, "loss": 0.0145, "step": 7900 }, { "grad_norm": 0.25965067744255066, "learning_rate": 9.97688227790566e-05, "loss": 0.0151, "step": 7910 }, { "grad_norm": 0.2505318820476532, "learning_rate": 9.976723189060684e-05, "loss": 0.0145, "step": 7920 }, { "grad_norm": 0.2687638998031616, "learning_rate": 9.976563555968742e-05, "loss": 0.014, "step": 7930 }, { "grad_norm": 0.2881550192832947, "learning_rate": 9.976403378647292e-05, "loss": 0.0156, "step": 7940 }, { "grad_norm": 0.3016267716884613, "learning_rate": 9.97624265711385e-05, "loss": 0.0159, "step": 7950 }, { "grad_norm": 0.25278347730636597, "learning_rate": 9.976081391385993e-05, "loss": 0.0195, "step": 7960 }, { "grad_norm": 0.24402952194213867, "learning_rate": 9.975919581481356e-05, "loss": 0.0179, "step": 7970 }, { "grad_norm": 0.3247901201248169, "learning_rate": 9.975757227417634e-05, "loss": 0.0195, "step": 7980 }, { "grad_norm": 0.35133567452430725, "learning_rate": 9.975594329212586e-05, "loss": 0.0192, "step": 7990 }, { "grad_norm": 0.2690684497356415, "learning_rate": 9.97543088688402e-05, "loss": 0.0152, "step": 8000 }, { "grad_norm": 0.27191218733787537, "learning_rate": 9.975266900449814e-05, "loss": 0.0124, "step": 8010 }, { "grad_norm": 0.33858466148376465, "learning_rate": 9.975102369927898e-05, "loss": 0.0149, "step": 8020 }, { "grad_norm": 0.39171501994132996, "learning_rate": 9.974937295336269e-05, "loss": 0.0181, "step": 8030 }, { "grad_norm": 0.29706788063049316, "learning_rate": 9.974771676692975e-05, "loss": 0.0175, "step": 8040 }, { "grad_norm": 0.27190101146698, "learning_rate": 9.974605514016131e-05, "loss": 0.0162, "step": 8050 }, { "grad_norm": 0.23318074643611908, "learning_rate": 9.974438807323907e-05, "loss": 0.0183, "step": 8060 }, { "grad_norm": 0.4528290629386902, "learning_rate": 9.974271556634535e-05, "loss": 0.0196, "step": 8070 }, { "grad_norm": 0.25770819187164307, "learning_rate": 9.974103761966302e-05, "loss": 0.0188, "step": 8080 }, { "grad_norm": 0.22396163642406464, "learning_rate": 9.973935423337563e-05, "loss": 0.0159, "step": 8090 }, { "grad_norm": 0.31590700149536133, "learning_rate": 9.973766540766722e-05, "loss": 0.0228, "step": 8100 }, { "grad_norm": 0.45245596766471863, "learning_rate": 9.97359711427225e-05, "loss": 0.0206, "step": 8110 }, { "grad_norm": 0.38643506169319153, "learning_rate": 9.973427143872677e-05, "loss": 0.02, "step": 8120 }, { "grad_norm": 0.41388142108917236, "learning_rate": 9.973256629586589e-05, "loss": 0.0195, "step": 8130 }, { "grad_norm": 0.514032781124115, "learning_rate": 9.973085571432632e-05, "loss": 0.021, "step": 8140 }, { "grad_norm": 0.45046645402908325, "learning_rate": 9.972913969429513e-05, "loss": 0.028, "step": 8150 }, { "grad_norm": 0.3172313868999481, "learning_rate": 9.972741823596e-05, "loss": 0.0262, "step": 8160 }, { "grad_norm": 0.327244371175766, "learning_rate": 9.972569133950917e-05, "loss": 0.0198, "step": 8170 }, { "grad_norm": 0.2623292803764343, "learning_rate": 9.972395900513151e-05, "loss": 0.0219, "step": 8180 }, { "grad_norm": 0.2915143668651581, "learning_rate": 9.972222123301645e-05, "loss": 0.022, "step": 8190 }, { "grad_norm": 0.2613820731639862, "learning_rate": 9.972047802335403e-05, "loss": 0.0216, "step": 8200 }, { "grad_norm": 0.2384902536869049, "learning_rate": 9.971872937633488e-05, "loss": 0.0212, "step": 8210 }, { "grad_norm": 0.3346947133541107, "learning_rate": 9.971697529215024e-05, "loss": 0.0173, "step": 8220 }, { "grad_norm": 0.36041274666786194, "learning_rate": 9.971521577099192e-05, "loss": 0.0167, "step": 8230 }, { "grad_norm": 0.26696836948394775, "learning_rate": 9.971345081305236e-05, "loss": 0.0196, "step": 8240 }, { "grad_norm": 0.22905786335468292, "learning_rate": 9.971168041852456e-05, "loss": 0.0167, "step": 8250 }, { "grad_norm": 0.3161394000053406, "learning_rate": 9.970990458760215e-05, "loss": 0.013, "step": 8260 }, { "grad_norm": 0.25163736939430237, "learning_rate": 9.970812332047929e-05, "loss": 0.0145, "step": 8270 }, { "grad_norm": 0.3739457428455353, "learning_rate": 9.97063366173508e-05, "loss": 0.0151, "step": 8280 }, { "grad_norm": 0.3742094039916992, "learning_rate": 9.970454447841207e-05, "loss": 0.0165, "step": 8290 }, { "grad_norm": 0.35616958141326904, "learning_rate": 9.970274690385909e-05, "loss": 0.0201, "step": 8300 }, { "grad_norm": 0.3430787920951843, "learning_rate": 9.970094389388844e-05, "loss": 0.018, "step": 8310 }, { "grad_norm": 0.3349452316761017, "learning_rate": 9.969913544869728e-05, "loss": 0.0164, "step": 8320 }, { "grad_norm": 0.3033505976200104, "learning_rate": 9.96973215684834e-05, "loss": 0.0176, "step": 8330 }, { "grad_norm": 0.410997211933136, "learning_rate": 9.969550225344513e-05, "loss": 0.0181, "step": 8340 }, { "grad_norm": 0.35769525170326233, "learning_rate": 9.969367750378147e-05, "loss": 0.0185, "step": 8350 }, { "grad_norm": 0.18587970733642578, "learning_rate": 9.969184731969194e-05, "loss": 0.0145, "step": 8360 }, { "grad_norm": 0.3411788046360016, "learning_rate": 9.96900117013767e-05, "loss": 0.0159, "step": 8370 }, { "grad_norm": 0.2450026273727417, "learning_rate": 9.96881706490365e-05, "loss": 0.0151, "step": 8380 }, { "grad_norm": 0.3249128460884094, "learning_rate": 9.968632416287265e-05, "loss": 0.0182, "step": 8390 }, { "grad_norm": 0.24781182408332825, "learning_rate": 9.96844722430871e-05, "loss": 0.018, "step": 8400 }, { "grad_norm": 0.26827123761177063, "learning_rate": 9.968261488988235e-05, "loss": 0.0209, "step": 8410 }, { "grad_norm": 0.2928686738014221, "learning_rate": 9.968075210346155e-05, "loss": 0.0173, "step": 8420 }, { "grad_norm": 0.27897387742996216, "learning_rate": 9.967888388402839e-05, "loss": 0.0169, "step": 8430 }, { "grad_norm": 0.23474374413490295, "learning_rate": 9.967701023178717e-05, "loss": 0.0173, "step": 8440 }, { "grad_norm": 0.24531544744968414, "learning_rate": 9.967513114694282e-05, "loss": 0.019, "step": 8450 }, { "grad_norm": 0.3245459198951721, "learning_rate": 9.967324662970079e-05, "loss": 0.0132, "step": 8460 }, { "grad_norm": 0.3866824507713318, "learning_rate": 9.96713566802672e-05, "loss": 0.0153, "step": 8470 }, { "grad_norm": 0.21944279968738556, "learning_rate": 9.966946129884873e-05, "loss": 0.0165, "step": 8480 }, { "grad_norm": 0.285833477973938, "learning_rate": 9.966756048565265e-05, "loss": 0.0177, "step": 8490 }, { "grad_norm": 0.23665152490139008, "learning_rate": 9.966565424088681e-05, "loss": 0.0169, "step": 8500 }, { "grad_norm": 0.2370656132698059, "learning_rate": 9.96637425647597e-05, "loss": 0.018, "step": 8510 }, { "grad_norm": 0.3282519578933716, "learning_rate": 9.966182545748038e-05, "loss": 0.0159, "step": 8520 }, { "grad_norm": 0.3665575385093689, "learning_rate": 9.96599029192585e-05, "loss": 0.0149, "step": 8530 }, { "grad_norm": 0.2566502094268799, "learning_rate": 9.965797495030428e-05, "loss": 0.0164, "step": 8540 }, { "grad_norm": 0.2511953115463257, "learning_rate": 9.96560415508286e-05, "loss": 0.0152, "step": 8550 }, { "grad_norm": 0.325973778963089, "learning_rate": 9.965410272104286e-05, "loss": 0.0164, "step": 8560 }, { "grad_norm": 0.1833496242761612, "learning_rate": 9.96521584611591e-05, "loss": 0.0201, "step": 8570 }, { "grad_norm": 0.31379398703575134, "learning_rate": 9.965020877138994e-05, "loss": 0.0169, "step": 8580 }, { "grad_norm": 0.18577533960342407, "learning_rate": 9.964825365194861e-05, "loss": 0.0153, "step": 8590 }, { "grad_norm": 0.26682958006858826, "learning_rate": 9.96462931030489e-05, "loss": 0.0152, "step": 8600 }, { "grad_norm": 0.2757103145122528, "learning_rate": 9.96443271249052e-05, "loss": 0.0182, "step": 8610 }, { "grad_norm": 0.27474871277809143, "learning_rate": 9.964235571773255e-05, "loss": 0.0164, "step": 8620 }, { "grad_norm": 0.2134350836277008, "learning_rate": 9.96403788817465e-05, "loss": 0.0171, "step": 8630 }, { "grad_norm": 0.28391310572624207, "learning_rate": 9.963839661716325e-05, "loss": 0.0148, "step": 8640 }, { "grad_norm": 0.21064570546150208, "learning_rate": 9.963640892419958e-05, "loss": 0.0162, "step": 8650 }, { "grad_norm": 0.21912361681461334, "learning_rate": 9.963441580307286e-05, "loss": 0.0161, "step": 8660 }, { "grad_norm": 0.3003552258014679, "learning_rate": 9.963241725400104e-05, "loss": 0.015, "step": 8670 }, { "grad_norm": 0.3420783281326294, "learning_rate": 9.963041327720271e-05, "loss": 0.0141, "step": 8680 }, { "grad_norm": 0.1862717717885971, "learning_rate": 9.962840387289697e-05, "loss": 0.0107, "step": 8690 }, { "grad_norm": 0.16795925796031952, "learning_rate": 9.962638904130363e-05, "loss": 0.0103, "step": 8700 }, { "grad_norm": 0.2933952212333679, "learning_rate": 9.962436878264298e-05, "loss": 0.0136, "step": 8710 }, { "grad_norm": 0.17514538764953613, "learning_rate": 9.962234309713598e-05, "loss": 0.0135, "step": 8720 }, { "grad_norm": 0.2055303156375885, "learning_rate": 9.962031198500414e-05, "loss": 0.0125, "step": 8730 }, { "grad_norm": 0.27110859751701355, "learning_rate": 9.961827544646958e-05, "loss": 0.0131, "step": 8740 }, { "grad_norm": 0.21925655007362366, "learning_rate": 9.961623348175501e-05, "loss": 0.0128, "step": 8750 }, { "grad_norm": 0.29946696758270264, "learning_rate": 9.961418609108377e-05, "loss": 0.0131, "step": 8760 }, { "grad_norm": 0.26184332370758057, "learning_rate": 9.961213327467971e-05, "loss": 0.015, "step": 8770 }, { "grad_norm": 0.24054640531539917, "learning_rate": 9.961007503276736e-05, "loss": 0.0143, "step": 8780 }, { "grad_norm": 0.24395452439785004, "learning_rate": 9.960801136557179e-05, "loss": 0.014, "step": 8790 }, { "grad_norm": 0.28609898686408997, "learning_rate": 9.960594227331866e-05, "loss": 0.0116, "step": 8800 }, { "grad_norm": 0.3089383542537689, "learning_rate": 9.960386775623429e-05, "loss": 0.0138, "step": 8810 }, { "grad_norm": 0.20885807275772095, "learning_rate": 9.96017878145455e-05, "loss": 0.0135, "step": 8820 }, { "grad_norm": 0.27070823311805725, "learning_rate": 9.959970244847977e-05, "loss": 0.0128, "step": 8830 }, { "grad_norm": 0.20717939734458923, "learning_rate": 9.959761165826518e-05, "loss": 0.0148, "step": 8840 }, { "grad_norm": 0.23323401808738708, "learning_rate": 9.959551544413033e-05, "loss": 0.0158, "step": 8850 }, { "grad_norm": 0.24635380506515503, "learning_rate": 9.959341380630448e-05, "loss": 0.0129, "step": 8860 }, { "grad_norm": 0.28887471556663513, "learning_rate": 9.959130674501746e-05, "loss": 0.0142, "step": 8870 }, { "grad_norm": 0.207117959856987, "learning_rate": 9.958919426049968e-05, "loss": 0.0115, "step": 8880 }, { "grad_norm": 0.2217278778553009, "learning_rate": 9.958707635298219e-05, "loss": 0.0158, "step": 8890 }, { "grad_norm": 0.2915216386318207, "learning_rate": 9.958495302269657e-05, "loss": 0.0136, "step": 8900 }, { "grad_norm": 0.2410089373588562, "learning_rate": 9.958282426987503e-05, "loss": 0.0129, "step": 8910 }, { "grad_norm": 0.26434504985809326, "learning_rate": 9.95806900947504e-05, "loss": 0.0125, "step": 8920 }, { "grad_norm": 0.19089139997959137, "learning_rate": 9.957855049755604e-05, "loss": 0.0146, "step": 8930 }, { "grad_norm": 0.26146095991134644, "learning_rate": 9.957640547852593e-05, "loss": 0.0118, "step": 8940 }, { "grad_norm": 0.20751230418682098, "learning_rate": 9.957425503789466e-05, "loss": 0.0125, "step": 8950 }, { "grad_norm": 0.2786296606063843, "learning_rate": 9.957209917589738e-05, "loss": 0.0146, "step": 8960 }, { "grad_norm": 0.3467561602592468, "learning_rate": 9.956993789276987e-05, "loss": 0.014, "step": 8970 }, { "grad_norm": 0.24016329646110535, "learning_rate": 9.956777118874847e-05, "loss": 0.0121, "step": 8980 }, { "grad_norm": 0.22165843844413757, "learning_rate": 9.956559906407016e-05, "loss": 0.0154, "step": 8990 }, { "grad_norm": 0.2698768675327301, "learning_rate": 9.956342151897245e-05, "loss": 0.0132, "step": 9000 }, { "grad_norm": 0.19573241472244263, "learning_rate": 9.956123855369346e-05, "loss": 0.0133, "step": 9010 }, { "grad_norm": 0.2674793303012848, "learning_rate": 9.955905016847196e-05, "loss": 0.0146, "step": 9020 }, { "grad_norm": 0.18347512185573578, "learning_rate": 9.955685636354723e-05, "loss": 0.0151, "step": 9030 }, { "grad_norm": 0.1937624216079712, "learning_rate": 9.95546571391592e-05, "loss": 0.0122, "step": 9040 }, { "grad_norm": 0.23477104306221008, "learning_rate": 9.955245249554837e-05, "loss": 0.0123, "step": 9050 }, { "grad_norm": 0.1416858285665512, "learning_rate": 9.955024243295582e-05, "loss": 0.0112, "step": 9060 }, { "grad_norm": 0.3132907748222351, "learning_rate": 9.954802695162328e-05, "loss": 0.0112, "step": 9070 }, { "grad_norm": 0.32980701327323914, "learning_rate": 9.954580605179302e-05, "loss": 0.0143, "step": 9080 }, { "grad_norm": 0.27827736735343933, "learning_rate": 9.954357973370788e-05, "loss": 0.0163, "step": 9090 }, { "grad_norm": 0.2478659749031067, "learning_rate": 9.954134799761135e-05, "loss": 0.0134, "step": 9100 }, { "grad_norm": 0.2672819197177887, "learning_rate": 9.953911084374748e-05, "loss": 0.0123, "step": 9110 }, { "grad_norm": 0.2754737138748169, "learning_rate": 9.953686827236093e-05, "loss": 0.0148, "step": 9120 }, { "grad_norm": 0.23772020637989044, "learning_rate": 9.953462028369695e-05, "loss": 0.0145, "step": 9130 }, { "grad_norm": 0.27359578013420105, "learning_rate": 9.953236687800136e-05, "loss": 0.0137, "step": 9140 }, { "grad_norm": 0.3237113356590271, "learning_rate": 9.95301080555206e-05, "loss": 0.017, "step": 9150 }, { "grad_norm": 0.25713643431663513, "learning_rate": 9.952784381650171e-05, "loss": 0.0133, "step": 9160 }, { "grad_norm": 0.17752860486507416, "learning_rate": 9.952557416119226e-05, "loss": 0.0146, "step": 9170 }, { "grad_norm": 0.32225286960601807, "learning_rate": 9.95232990898405e-05, "loss": 0.0162, "step": 9180 }, { "grad_norm": 0.2327519953250885, "learning_rate": 9.95210186026952e-05, "loss": 0.0186, "step": 9190 }, { "grad_norm": 0.3349674642086029, "learning_rate": 9.951873270000576e-05, "loss": 0.0198, "step": 9200 }, { "grad_norm": 0.273930162191391, "learning_rate": 9.951644138202216e-05, "loss": 0.0183, "step": 9210 }, { "grad_norm": 0.2406584620475769, "learning_rate": 9.951414464899498e-05, "loss": 0.0155, "step": 9220 }, { "grad_norm": 0.2763700485229492, "learning_rate": 9.951184250117538e-05, "loss": 0.0165, "step": 9230 }, { "grad_norm": 0.39099961519241333, "learning_rate": 9.950953493881513e-05, "loss": 0.0144, "step": 9240 }, { "grad_norm": 0.21896101534366608, "learning_rate": 9.950722196216658e-05, "loss": 0.016, "step": 9250 }, { "grad_norm": 0.16185538470745087, "learning_rate": 9.950490357148265e-05, "loss": 0.0181, "step": 9260 }, { "grad_norm": 0.39514803886413574, "learning_rate": 9.950257976701692e-05, "loss": 0.0201, "step": 9270 }, { "grad_norm": 0.26806455850601196, "learning_rate": 9.950025054902348e-05, "loss": 0.0186, "step": 9280 }, { "grad_norm": 0.3048141896724701, "learning_rate": 9.949791591775706e-05, "loss": 0.0207, "step": 9290 }, { "grad_norm": 0.3074019253253937, "learning_rate": 9.949557587347298e-05, "loss": 0.019, "step": 9300 }, { "grad_norm": 0.2796652913093567, "learning_rate": 9.949323041642713e-05, "loss": 0.016, "step": 9310 }, { "grad_norm": 0.2789228856563568, "learning_rate": 9.949087954687602e-05, "loss": 0.0122, "step": 9320 }, { "grad_norm": 0.25437164306640625, "learning_rate": 9.948852326507672e-05, "loss": 0.0192, "step": 9330 }, { "grad_norm": 0.22343376278877258, "learning_rate": 9.948616157128694e-05, "loss": 0.0208, "step": 9340 }, { "grad_norm": 0.3107854425907135, "learning_rate": 9.948379446576493e-05, "loss": 0.0197, "step": 9350 }, { "grad_norm": 0.3534637987613678, "learning_rate": 9.948142194876952e-05, "loss": 0.023, "step": 9360 }, { "grad_norm": 0.3241763114929199, "learning_rate": 9.947904402056024e-05, "loss": 0.0249, "step": 9370 }, { "grad_norm": 0.30371150374412537, "learning_rate": 9.947666068139708e-05, "loss": 0.0207, "step": 9380 }, { "grad_norm": 0.26051580905914307, "learning_rate": 9.947427193154071e-05, "loss": 0.0214, "step": 9390 }, { "grad_norm": 0.3054509460926056, "learning_rate": 9.947187777125233e-05, "loss": 0.0201, "step": 9400 }, { "grad_norm": 0.2765798568725586, "learning_rate": 9.946947820079377e-05, "loss": 0.0209, "step": 9410 }, { "grad_norm": 0.49112141132354736, "learning_rate": 9.946707322042747e-05, "loss": 0.0289, "step": 9420 }, { "grad_norm": 0.2647642493247986, "learning_rate": 9.94646628304164e-05, "loss": 0.023, "step": 9430 }, { "grad_norm": 0.3844645321369171, "learning_rate": 9.946224703102418e-05, "loss": 0.0216, "step": 9440 }, { "grad_norm": 0.20580799877643585, "learning_rate": 9.945982582251498e-05, "loss": 0.0166, "step": 9450 }, { "grad_norm": 0.27227404713630676, "learning_rate": 9.94573992051536e-05, "loss": 0.0173, "step": 9460 }, { "grad_norm": 0.2396056205034256, "learning_rate": 9.94549671792054e-05, "loss": 0.0148, "step": 9470 }, { "grad_norm": 0.37168556451797485, "learning_rate": 9.945252974493635e-05, "loss": 0.015, "step": 9480 }, { "grad_norm": 0.23860624432563782, "learning_rate": 9.9450086902613e-05, "loss": 0.0188, "step": 9490 }, { "grad_norm": 0.20068363845348358, "learning_rate": 9.944763865250248e-05, "loss": 0.0135, "step": 9500 }, { "grad_norm": 0.3320692777633667, "learning_rate": 9.944518499487254e-05, "loss": 0.017, "step": 9510 }, { "grad_norm": 0.2381013184785843, "learning_rate": 9.944272592999151e-05, "loss": 0.0131, "step": 9520 }, { "grad_norm": 0.12308555841445923, "learning_rate": 9.94402614581283e-05, "loss": 0.0108, "step": 9530 }, { "grad_norm": 0.2967275381088257, "learning_rate": 9.943779157955244e-05, "loss": 0.0132, "step": 9540 }, { "grad_norm": 0.33537623286247253, "learning_rate": 9.943531629453403e-05, "loss": 0.0139, "step": 9550 }, { "grad_norm": 0.22844187915325165, "learning_rate": 9.943283560334375e-05, "loss": 0.0117, "step": 9560 }, { "grad_norm": 0.19647110998630524, "learning_rate": 9.943034950625288e-05, "loss": 0.0145, "step": 9570 }, { "grad_norm": 0.214546799659729, "learning_rate": 9.942785800353332e-05, "loss": 0.0143, "step": 9580 }, { "grad_norm": 0.22903765738010406, "learning_rate": 9.942536109545751e-05, "loss": 0.0131, "step": 9590 }, { "grad_norm": 0.17920711636543274, "learning_rate": 9.942285878229853e-05, "loss": 0.0125, "step": 9600 }, { "grad_norm": 0.31093066930770874, "learning_rate": 9.942035106433001e-05, "loss": 0.0119, "step": 9610 }, { "grad_norm": 0.22234076261520386, "learning_rate": 9.94178379418262e-05, "loss": 0.0129, "step": 9620 }, { "grad_norm": 0.24905391037464142, "learning_rate": 9.941531941506194e-05, "loss": 0.0148, "step": 9630 }, { "grad_norm": 0.44659531116485596, "learning_rate": 9.941279548431263e-05, "loss": 0.016, "step": 9640 }, { "grad_norm": 0.17533747851848602, "learning_rate": 9.941026614985431e-05, "loss": 0.0153, "step": 9650 }, { "grad_norm": 0.21096478402614594, "learning_rate": 9.940773141196357e-05, "loss": 0.0134, "step": 9660 }, { "grad_norm": 0.1753935068845749, "learning_rate": 9.94051912709176e-05, "loss": 0.0129, "step": 9670 }, { "grad_norm": 0.18819020688533783, "learning_rate": 9.940264572699421e-05, "loss": 0.0126, "step": 9680 }, { "grad_norm": 0.1857440173625946, "learning_rate": 9.940009478047174e-05, "loss": 0.0117, "step": 9690 }, { "grad_norm": 0.24355654418468475, "learning_rate": 9.939753843162918e-05, "loss": 0.0118, "step": 9700 }, { "grad_norm": 0.2656756341457367, "learning_rate": 9.939497668074609e-05, "loss": 0.015, "step": 9710 }, { "grad_norm": 0.25384393334388733, "learning_rate": 9.93924095281026e-05, "loss": 0.0126, "step": 9720 }, { "grad_norm": 0.20678994059562683, "learning_rate": 9.938983697397948e-05, "loss": 0.0109, "step": 9730 }, { "grad_norm": 0.17685499787330627, "learning_rate": 9.938725901865805e-05, "loss": 0.0137, "step": 9740 }, { "grad_norm": 0.1890147179365158, "learning_rate": 9.93846756624202e-05, "loss": 0.0141, "step": 9750 }, { "grad_norm": 0.23930047452449799, "learning_rate": 9.938208690554849e-05, "loss": 0.0122, "step": 9760 }, { "grad_norm": 0.25994542241096497, "learning_rate": 9.9379492748326e-05, "loss": 0.0139, "step": 9770 }, { "grad_norm": 0.3298279345035553, "learning_rate": 9.937689319103641e-05, "loss": 0.0218, "step": 9780 }, { "grad_norm": 0.2863154709339142, "learning_rate": 9.937428823396404e-05, "loss": 0.0156, "step": 9790 }, { "grad_norm": 0.1939290165901184, "learning_rate": 9.937167787739372e-05, "loss": 0.0127, "step": 9800 }, { "grad_norm": 0.22792081534862518, "learning_rate": 9.936906212161095e-05, "loss": 0.0114, "step": 9810 }, { "grad_norm": 0.2276717722415924, "learning_rate": 9.936644096690176e-05, "loss": 0.0116, "step": 9820 }, { "grad_norm": 0.24423544108867645, "learning_rate": 9.936381441355282e-05, "loss": 0.0105, "step": 9830 }, { "grad_norm": 0.2591060996055603, "learning_rate": 9.936118246185136e-05, "loss": 0.0119, "step": 9840 }, { "grad_norm": 0.19467571377754211, "learning_rate": 9.935854511208518e-05, "loss": 0.0128, "step": 9850 }, { "grad_norm": 0.19363342225551605, "learning_rate": 9.935590236454272e-05, "loss": 0.0102, "step": 9860 }, { "grad_norm": 0.2130424827337265, "learning_rate": 9.935325421951298e-05, "loss": 0.012, "step": 9870 }, { "grad_norm": 0.2184283286333084, "learning_rate": 9.935060067728557e-05, "loss": 0.0112, "step": 9880 }, { "grad_norm": 0.25058674812316895, "learning_rate": 9.934794173815067e-05, "loss": 0.0117, "step": 9890 }, { "grad_norm": 0.21481983363628387, "learning_rate": 9.934527740239906e-05, "loss": 0.0139, "step": 9900 }, { "grad_norm": 0.2113254964351654, "learning_rate": 9.934260767032209e-05, "loss": 0.0127, "step": 9910 }, { "grad_norm": 0.3339419662952423, "learning_rate": 9.933993254221172e-05, "loss": 0.0314, "step": 9920 }, { "grad_norm": 0.3452908992767334, "learning_rate": 9.933725201836053e-05, "loss": 0.0265, "step": 9930 }, { "grad_norm": 0.33856987953186035, "learning_rate": 9.933456609906162e-05, "loss": 0.0181, "step": 9940 }, { "grad_norm": 0.32439300417900085, "learning_rate": 9.933187478460875e-05, "loss": 0.019, "step": 9950 }, { "grad_norm": 0.2742484211921692, "learning_rate": 9.93291780752962e-05, "loss": 0.0211, "step": 9960 }, { "grad_norm": 0.2228974550962448, "learning_rate": 9.932647597141893e-05, "loss": 0.0174, "step": 9970 }, { "grad_norm": 0.305730938911438, "learning_rate": 9.932376847327239e-05, "loss": 0.0194, "step": 9980 }, { "grad_norm": 0.37023991346359253, "learning_rate": 9.932105558115268e-05, "loss": 0.0192, "step": 9990 }, { "grad_norm": 0.2564316689968109, "learning_rate": 9.931833729535651e-05, "loss": 0.0158, "step": 10000 }, { "grad_norm": 0.16419345140457153, "learning_rate": 9.931561361618111e-05, "loss": 0.0133, "step": 10010 }, { "grad_norm": 0.1684248447418213, "learning_rate": 9.931288454392435e-05, "loss": 0.0125, "step": 10020 }, { "grad_norm": 0.24861733615398407, "learning_rate": 9.931015007888467e-05, "loss": 0.0119, "step": 10030 }, { "grad_norm": 0.20048461854457855, "learning_rate": 9.930741022136112e-05, "loss": 0.0124, "step": 10040 }, { "grad_norm": 0.16538085043430328, "learning_rate": 9.930466497165333e-05, "loss": 0.0134, "step": 10050 }, { "grad_norm": 0.14653506875038147, "learning_rate": 9.93019143300615e-05, "loss": 0.0123, "step": 10060 }, { "grad_norm": 0.3167251944541931, "learning_rate": 9.929915829688644e-05, "loss": 0.0127, "step": 10070 }, { "grad_norm": 0.2696905732154846, "learning_rate": 9.929639687242955e-05, "loss": 0.0164, "step": 10080 }, { "grad_norm": 0.21866635978221893, "learning_rate": 9.929363005699281e-05, "loss": 0.0149, "step": 10090 }, { "grad_norm": 0.33137255907058716, "learning_rate": 9.92908578508788e-05, "loss": 0.0156, "step": 10100 }, { "grad_norm": 0.3900640606880188, "learning_rate": 9.928808025439069e-05, "loss": 0.0199, "step": 10110 }, { "grad_norm": 0.30194181203842163, "learning_rate": 9.928529726783223e-05, "loss": 0.0166, "step": 10120 }, { "grad_norm": 0.23519688844680786, "learning_rate": 9.928250889150774e-05, "loss": 0.0138, "step": 10130 }, { "grad_norm": 0.23419789969921112, "learning_rate": 9.92797151257222e-05, "loss": 0.0127, "step": 10140 }, { "grad_norm": 0.2006024271249771, "learning_rate": 9.927691597078108e-05, "loss": 0.0117, "step": 10150 }, { "grad_norm": 0.16667750477790833, "learning_rate": 9.927411142699053e-05, "loss": 0.0108, "step": 10160 }, { "grad_norm": 0.2216326892375946, "learning_rate": 9.927130149465725e-05, "loss": 0.0154, "step": 10170 }, { "grad_norm": 0.21564216911792755, "learning_rate": 9.92684861740885e-05, "loss": 0.0155, "step": 10180 }, { "grad_norm": 0.21959742903709412, "learning_rate": 9.926566546559217e-05, "loss": 0.0139, "step": 10190 }, { "grad_norm": 0.22112058103084564, "learning_rate": 9.926283936947673e-05, "loss": 0.0151, "step": 10200 }, { "grad_norm": 0.2039620727300644, "learning_rate": 9.926000788605126e-05, "loss": 0.0153, "step": 10210 }, { "grad_norm": 0.25692018866539, "learning_rate": 9.92571710156254e-05, "loss": 0.0152, "step": 10220 }, { "grad_norm": 0.2506529688835144, "learning_rate": 9.925432875850936e-05, "loss": 0.0142, "step": 10230 }, { "grad_norm": 0.19300781190395355, "learning_rate": 9.925148111501396e-05, "loss": 0.0132, "step": 10240 }, { "grad_norm": 0.249482199549675, "learning_rate": 9.924862808545066e-05, "loss": 0.0124, "step": 10250 }, { "grad_norm": 0.17114050686359406, "learning_rate": 9.924576967013141e-05, "loss": 0.0133, "step": 10260 }, { "grad_norm": 0.22455058991909027, "learning_rate": 9.924290586936887e-05, "loss": 0.0124, "step": 10270 }, { "grad_norm": 0.204708531498909, "learning_rate": 9.924003668347614e-05, "loss": 0.0125, "step": 10280 }, { "grad_norm": 0.2853237986564636, "learning_rate": 9.923716211276704e-05, "loss": 0.0125, "step": 10290 }, { "grad_norm": 0.22236202657222748, "learning_rate": 9.923428215755594e-05, "loss": 0.013, "step": 10300 }, { "grad_norm": 0.21516898274421692, "learning_rate": 9.923139681815775e-05, "loss": 0.0124, "step": 10310 }, { "grad_norm": 0.24964682757854462, "learning_rate": 9.922850609488801e-05, "loss": 0.0098, "step": 10320 }, { "grad_norm": 0.2737549841403961, "learning_rate": 9.922560998806287e-05, "loss": 0.0141, "step": 10330 }, { "grad_norm": 0.31785720586776733, "learning_rate": 9.922270849799905e-05, "loss": 0.0133, "step": 10340 }, { "grad_norm": 0.2444351315498352, "learning_rate": 9.92198016250138e-05, "loss": 0.0137, "step": 10350 }, { "grad_norm": 0.2039456069469452, "learning_rate": 9.921688936942506e-05, "loss": 0.012, "step": 10360 }, { "grad_norm": 0.1695556938648224, "learning_rate": 9.921397173155129e-05, "loss": 0.0121, "step": 10370 }, { "grad_norm": 0.2602848708629608, "learning_rate": 9.921104871171157e-05, "loss": 0.011, "step": 10380 }, { "grad_norm": 0.23214195668697357, "learning_rate": 9.920812031022554e-05, "loss": 0.0134, "step": 10390 }, { "grad_norm": 0.2177918702363968, "learning_rate": 9.920518652741348e-05, "loss": 0.0129, "step": 10400 }, { "grad_norm": 0.19306723773479462, "learning_rate": 9.920224736359618e-05, "loss": 0.0125, "step": 10410 }, { "grad_norm": 0.3038657605648041, "learning_rate": 9.91993028190951e-05, "loss": 0.0115, "step": 10420 }, { "grad_norm": 0.2552294135093689, "learning_rate": 9.919635289423222e-05, "loss": 0.0104, "step": 10430 }, { "grad_norm": 0.27272897958755493, "learning_rate": 9.919339758933015e-05, "loss": 0.0117, "step": 10440 }, { "grad_norm": 0.2092997282743454, "learning_rate": 9.919043690471209e-05, "loss": 0.0131, "step": 10450 }, { "grad_norm": 0.30459415912628174, "learning_rate": 9.91874708407018e-05, "loss": 0.0142, "step": 10460 }, { "grad_norm": 0.30529168248176575, "learning_rate": 9.918449939762367e-05, "loss": 0.0243, "step": 10470 }, { "grad_norm": 0.30123746395111084, "learning_rate": 9.91815225758026e-05, "loss": 0.0179, "step": 10480 }, { "grad_norm": 0.3551059067249298, "learning_rate": 9.917854037556419e-05, "loss": 0.0173, "step": 10490 }, { "grad_norm": 0.2598206698894501, "learning_rate": 9.917555279723454e-05, "loss": 0.0219, "step": 10500 }, { "grad_norm": 0.1682259738445282, "learning_rate": 9.917255984114036e-05, "loss": 0.0139, "step": 10510 }, { "grad_norm": 0.2554153800010681, "learning_rate": 9.916956150760896e-05, "loss": 0.0141, "step": 10520 }, { "grad_norm": 0.28103798627853394, "learning_rate": 9.916655779696826e-05, "loss": 0.0133, "step": 10530 }, { "grad_norm": 0.26814016699790955, "learning_rate": 9.916354870954671e-05, "loss": 0.0142, "step": 10540 }, { "grad_norm": 0.31190115213394165, "learning_rate": 9.91605342456734e-05, "loss": 0.0147, "step": 10550 }, { "grad_norm": 0.3594316244125366, "learning_rate": 9.915751440567795e-05, "loss": 0.018, "step": 10560 }, { "grad_norm": 0.27373385429382324, "learning_rate": 9.915448918989066e-05, "loss": 0.0186, "step": 10570 }, { "grad_norm": 0.36190399527549744, "learning_rate": 9.915145859864232e-05, "loss": 0.019, "step": 10580 }, { "grad_norm": 0.28749898076057434, "learning_rate": 9.914842263226437e-05, "loss": 0.0266, "step": 10590 }, { "grad_norm": 0.18503417074680328, "learning_rate": 9.914538129108882e-05, "loss": 0.0208, "step": 10600 }, { "grad_norm": 0.21508893370628357, "learning_rate": 9.914233457544825e-05, "loss": 0.0165, "step": 10610 }, { "grad_norm": 0.2111719697713852, "learning_rate": 9.913928248567586e-05, "loss": 0.0154, "step": 10620 }, { "grad_norm": 0.28600236773490906, "learning_rate": 9.913622502210542e-05, "loss": 0.0151, "step": 10630 }, { "grad_norm": 0.14692813158035278, "learning_rate": 9.913316218507128e-05, "loss": 0.0138, "step": 10640 }, { "grad_norm": 0.2272130846977234, "learning_rate": 9.91300939749084e-05, "loss": 0.0156, "step": 10650 }, { "grad_norm": 0.2501775920391083, "learning_rate": 9.91270203919523e-05, "loss": 0.0122, "step": 10660 }, { "grad_norm": 0.24398231506347656, "learning_rate": 9.912394143653912e-05, "loss": 0.0114, "step": 10670 }, { "grad_norm": 0.1782490462064743, "learning_rate": 9.912085710900555e-05, "loss": 0.0136, "step": 10680 }, { "grad_norm": 0.2303052842617035, "learning_rate": 9.911776740968892e-05, "loss": 0.0133, "step": 10690 }, { "grad_norm": 0.24568040668964386, "learning_rate": 9.911467233892709e-05, "loss": 0.0124, "step": 10700 }, { "grad_norm": 0.24156953394412994, "learning_rate": 9.911157189705853e-05, "loss": 0.0131, "step": 10710 }, { "grad_norm": 0.354289174079895, "learning_rate": 9.910846608442229e-05, "loss": 0.0145, "step": 10720 }, { "grad_norm": 0.2516837418079376, "learning_rate": 9.910535490135805e-05, "loss": 0.0215, "step": 10730 }, { "grad_norm": 0.2501070201396942, "learning_rate": 9.910223834820603e-05, "loss": 0.0188, "step": 10740 }, { "grad_norm": 0.23132754862308502, "learning_rate": 9.909911642530703e-05, "loss": 0.0155, "step": 10750 }, { "grad_norm": 0.20377452671527863, "learning_rate": 9.909598913300249e-05, "loss": 0.0131, "step": 10760 }, { "grad_norm": 0.2524181306362152, "learning_rate": 9.909285647163438e-05, "loss": 0.0107, "step": 10770 }, { "grad_norm": 0.2717682421207428, "learning_rate": 9.908971844154531e-05, "loss": 0.0106, "step": 10780 }, { "grad_norm": 0.21323491632938385, "learning_rate": 9.908657504307843e-05, "loss": 0.0135, "step": 10790 }, { "grad_norm": 0.31888526678085327, "learning_rate": 9.908342627657751e-05, "loss": 0.0127, "step": 10800 }, { "grad_norm": 0.16357752680778503, "learning_rate": 9.908027214238689e-05, "loss": 0.0144, "step": 10810 }, { "grad_norm": 0.20973123610019684, "learning_rate": 9.90771126408515e-05, "loss": 0.0121, "step": 10820 }, { "grad_norm": 0.19528602063655853, "learning_rate": 9.907394777231685e-05, "loss": 0.0105, "step": 10830 }, { "grad_norm": 0.1523207277059555, "learning_rate": 9.907077753712905e-05, "loss": 0.0098, "step": 10840 }, { "grad_norm": 0.2849341034889221, "learning_rate": 9.906760193563482e-05, "loss": 0.0112, "step": 10850 }, { "grad_norm": 0.2515259087085724, "learning_rate": 9.906442096818139e-05, "loss": 0.0099, "step": 10860 }, { "grad_norm": 0.2465417981147766, "learning_rate": 9.906123463511665e-05, "loss": 0.0111, "step": 10870 }, { "grad_norm": 0.23922637104988098, "learning_rate": 9.905804293678907e-05, "loss": 0.0118, "step": 10880 }, { "grad_norm": 0.19982922077178955, "learning_rate": 9.905484587354766e-05, "loss": 0.0107, "step": 10890 }, { "grad_norm": 0.25449538230895996, "learning_rate": 9.905164344574205e-05, "loss": 0.0116, "step": 10900 }, { "grad_norm": 0.22932924330234528, "learning_rate": 9.904843565372248e-05, "loss": 0.0147, "step": 10910 }, { "grad_norm": 0.18670274317264557, "learning_rate": 9.904522249783972e-05, "loss": 0.013, "step": 10920 }, { "grad_norm": 0.3931998014450073, "learning_rate": 9.904200397844517e-05, "loss": 0.0143, "step": 10930 }, { "grad_norm": 0.2784362733364105, "learning_rate": 9.903878009589078e-05, "loss": 0.0145, "step": 10940 }, { "grad_norm": 0.2458958476781845, "learning_rate": 9.903555085052915e-05, "loss": 0.0134, "step": 10950 }, { "grad_norm": 0.20244038105010986, "learning_rate": 9.903231624271338e-05, "loss": 0.0106, "step": 10960 }, { "grad_norm": 0.19228056073188782, "learning_rate": 9.902907627279724e-05, "loss": 0.0125, "step": 10970 }, { "grad_norm": 0.1957712024450302, "learning_rate": 9.902583094113504e-05, "loss": 0.0114, "step": 10980 }, { "grad_norm": 0.1694270819425583, "learning_rate": 9.902258024808168e-05, "loss": 0.014, "step": 10990 }, { "grad_norm": 0.22756518423557281, "learning_rate": 9.901932419399264e-05, "loss": 0.0133, "step": 11000 }, { "grad_norm": 0.20924212038516998, "learning_rate": 9.9016062779224e-05, "loss": 0.0122, "step": 11010 }, { "grad_norm": 0.20166070759296417, "learning_rate": 9.901279600413242e-05, "loss": 0.0112, "step": 11020 }, { "grad_norm": 0.1860986202955246, "learning_rate": 9.900952386907518e-05, "loss": 0.0106, "step": 11030 }, { "grad_norm": 0.2927027642726898, "learning_rate": 9.90062463744101e-05, "loss": 0.0151, "step": 11040 }, { "grad_norm": 0.178673654794693, "learning_rate": 9.900296352049558e-05, "loss": 0.0143, "step": 11050 }, { "grad_norm": 0.4017442762851715, "learning_rate": 9.899967530769065e-05, "loss": 0.0145, "step": 11060 }, { "grad_norm": 0.3037525415420532, "learning_rate": 9.899638173635489e-05, "loss": 0.0142, "step": 11070 }, { "grad_norm": 0.27036499977111816, "learning_rate": 9.899308280684849e-05, "loss": 0.0132, "step": 11080 }, { "grad_norm": 0.2548007071018219, "learning_rate": 9.898977851953222e-05, "loss": 0.011, "step": 11090 }, { "grad_norm": 0.23491127789020538, "learning_rate": 9.898646887476741e-05, "loss": 0.0129, "step": 11100 }, { "grad_norm": 0.19607877731323242, "learning_rate": 9.898315387291603e-05, "loss": 0.0131, "step": 11110 }, { "grad_norm": 0.28823044896125793, "learning_rate": 9.89798335143406e-05, "loss": 0.0152, "step": 11120 }, { "grad_norm": 0.3025262653827667, "learning_rate": 9.897650779940419e-05, "loss": 0.0158, "step": 11130 }, { "grad_norm": 0.25660762190818787, "learning_rate": 9.897317672847054e-05, "loss": 0.0145, "step": 11140 }, { "grad_norm": 0.2655591368675232, "learning_rate": 9.89698403019039e-05, "loss": 0.0132, "step": 11150 }, { "grad_norm": 0.22491541504859924, "learning_rate": 9.896649852006917e-05, "loss": 0.0146, "step": 11160 }, { "grad_norm": 0.28715264797210693, "learning_rate": 9.896315138333177e-05, "loss": 0.0142, "step": 11170 }, { "grad_norm": 0.27737849950790405, "learning_rate": 9.895979889205774e-05, "loss": 0.0151, "step": 11180 }, { "grad_norm": 0.17178930342197418, "learning_rate": 9.895644104661372e-05, "loss": 0.0133, "step": 11190 }, { "grad_norm": 0.20421473681926727, "learning_rate": 9.895307784736691e-05, "loss": 0.0126, "step": 11200 }, { "grad_norm": 0.2579026520252228, "learning_rate": 9.894970929468512e-05, "loss": 0.0129, "step": 11210 }, { "grad_norm": 0.22352294623851776, "learning_rate": 9.89463353889367e-05, "loss": 0.0146, "step": 11220 }, { "grad_norm": 0.2982726991176605, "learning_rate": 9.894295613049065e-05, "loss": 0.0195, "step": 11230 }, { "grad_norm": 0.23292505741119385, "learning_rate": 9.893957151971649e-05, "loss": 0.0121, "step": 11240 }, { "grad_norm": 0.25943973660469055, "learning_rate": 9.893618155698436e-05, "loss": 0.0115, "step": 11250 }, { "grad_norm": 0.19855603575706482, "learning_rate": 9.8932786242665e-05, "loss": 0.0119, "step": 11260 }, { "grad_norm": 0.1906764805316925, "learning_rate": 9.89293855771297e-05, "loss": 0.0105, "step": 11270 }, { "grad_norm": 0.2294672578573227, "learning_rate": 9.892597956075036e-05, "loss": 0.0134, "step": 11280 }, { "grad_norm": 0.2377382218837738, "learning_rate": 9.892256819389947e-05, "loss": 0.0135, "step": 11290 }, { "grad_norm": 0.19480568170547485, "learning_rate": 9.891915147695006e-05, "loss": 0.0129, "step": 11300 }, { "grad_norm": 0.3438933491706848, "learning_rate": 9.891572941027577e-05, "loss": 0.0133, "step": 11310 }, { "grad_norm": 0.17022976279258728, "learning_rate": 9.89123019942509e-05, "loss": 0.0117, "step": 11320 }, { "grad_norm": 0.19550691545009613, "learning_rate": 9.89088692292502e-05, "loss": 0.0126, "step": 11330 }, { "grad_norm": 0.1586330085992813, "learning_rate": 9.89054311156491e-05, "loss": 0.0154, "step": 11340 }, { "grad_norm": 0.2567533552646637, "learning_rate": 9.890198765382357e-05, "loss": 0.0143, "step": 11350 }, { "grad_norm": 0.23671504855155945, "learning_rate": 9.889853884415021e-05, "loss": 0.0147, "step": 11360 }, { "grad_norm": 0.24345436692237854, "learning_rate": 9.889508468700614e-05, "loss": 0.0154, "step": 11370 }, { "grad_norm": 0.363232284784317, "learning_rate": 9.889162518276915e-05, "loss": 0.0146, "step": 11380 }, { "grad_norm": 0.2700692415237427, "learning_rate": 9.888816033181752e-05, "loss": 0.015, "step": 11390 }, { "grad_norm": 0.20039454102516174, "learning_rate": 9.888469013453018e-05, "loss": 0.0132, "step": 11400 }, { "grad_norm": 0.2220679223537445, "learning_rate": 9.888121459128663e-05, "loss": 0.013, "step": 11410 }, { "grad_norm": 0.2149966061115265, "learning_rate": 9.887773370246693e-05, "loss": 0.0112, "step": 11420 }, { "grad_norm": 0.22424152493476868, "learning_rate": 9.887424746845177e-05, "loss": 0.0113, "step": 11430 }, { "grad_norm": 0.132783442735672, "learning_rate": 9.887075588962239e-05, "loss": 0.0103, "step": 11440 }, { "grad_norm": 0.1810733526945114, "learning_rate": 9.88672589663606e-05, "loss": 0.0125, "step": 11450 }, { "grad_norm": 0.20136666297912598, "learning_rate": 9.886375669904886e-05, "loss": 0.0134, "step": 11460 }, { "grad_norm": 0.27011412382125854, "learning_rate": 9.886024908807014e-05, "loss": 0.0113, "step": 11470 }, { "grad_norm": 0.2415134757757187, "learning_rate": 9.885673613380806e-05, "loss": 0.0105, "step": 11480 }, { "grad_norm": 0.2944153845310211, "learning_rate": 9.885321783664676e-05, "loss": 0.0119, "step": 11490 }, { "grad_norm": 0.2514844536781311, "learning_rate": 9.884969419697101e-05, "loss": 0.0118, "step": 11500 }, { "grad_norm": 0.16631926596164703, "learning_rate": 9.884616521516614e-05, "loss": 0.0118, "step": 11510 }, { "grad_norm": 0.23755955696105957, "learning_rate": 9.88426308916181e-05, "loss": 0.011, "step": 11520 }, { "grad_norm": 0.2286119908094406, "learning_rate": 9.883909122671335e-05, "loss": 0.0126, "step": 11530 }, { "grad_norm": 0.21439233422279358, "learning_rate": 9.883554622083904e-05, "loss": 0.0117, "step": 11540 }, { "grad_norm": 0.19651095569133759, "learning_rate": 9.88319958743828e-05, "loss": 0.0129, "step": 11550 }, { "grad_norm": 0.22122205793857574, "learning_rate": 9.882844018773291e-05, "loss": 0.0121, "step": 11560 }, { "grad_norm": 0.1861579269170761, "learning_rate": 9.882487916127823e-05, "loss": 0.0105, "step": 11570 }, { "grad_norm": 0.19966746866703033, "learning_rate": 9.882131279540815e-05, "loss": 0.013, "step": 11580 }, { "grad_norm": 0.20963993668556213, "learning_rate": 9.881774109051271e-05, "loss": 0.0118, "step": 11590 }, { "grad_norm": 0.15706202387809753, "learning_rate": 9.881416404698252e-05, "loss": 0.0133, "step": 11600 }, { "grad_norm": 0.22447845339775085, "learning_rate": 9.881058166520873e-05, "loss": 0.0155, "step": 11610 }, { "grad_norm": 0.27259477972984314, "learning_rate": 9.880699394558311e-05, "loss": 0.0167, "step": 11620 }, { "grad_norm": 0.24119620025157928, "learning_rate": 9.880340088849801e-05, "loss": 0.0178, "step": 11630 }, { "grad_norm": 0.2087341845035553, "learning_rate": 9.879980249434637e-05, "loss": 0.0135, "step": 11640 }, { "grad_norm": 0.2751704454421997, "learning_rate": 9.879619876352168e-05, "loss": 0.0137, "step": 11650 }, { "grad_norm": 0.2762129008769989, "learning_rate": 9.879258969641809e-05, "loss": 0.0134, "step": 11660 }, { "grad_norm": 0.28641483187675476, "learning_rate": 9.878897529343023e-05, "loss": 0.0142, "step": 11670 }, { "grad_norm": 0.27041956782341003, "learning_rate": 9.878535555495338e-05, "loss": 0.013, "step": 11680 }, { "grad_norm": 0.2604629397392273, "learning_rate": 9.87817304813834e-05, "loss": 0.0149, "step": 11690 }, { "grad_norm": 0.24762418866157532, "learning_rate": 9.877810007311671e-05, "loss": 0.0145, "step": 11700 }, { "grad_norm": 0.2214679718017578, "learning_rate": 9.877446433055035e-05, "loss": 0.0117, "step": 11710 }, { "grad_norm": 0.20460769534111023, "learning_rate": 9.877082325408191e-05, "loss": 0.0109, "step": 11720 }, { "grad_norm": 0.17377042770385742, "learning_rate": 9.876717684410954e-05, "loss": 0.0129, "step": 11730 }, { "grad_norm": 0.25703078508377075, "learning_rate": 9.876352510103204e-05, "loss": 0.0131, "step": 11740 }, { "grad_norm": 0.15667377412319183, "learning_rate": 9.875986802524875e-05, "loss": 0.0127, "step": 11750 }, { "grad_norm": 0.237854465842247, "learning_rate": 9.87562056171596e-05, "loss": 0.011, "step": 11760 }, { "grad_norm": 0.16552071273326874, "learning_rate": 9.875253787716511e-05, "loss": 0.0101, "step": 11770 }, { "grad_norm": 0.1662503033876419, "learning_rate": 9.874886480566637e-05, "loss": 0.0096, "step": 11780 }, { "grad_norm": 0.18458417057991028, "learning_rate": 9.874518640306507e-05, "loss": 0.0129, "step": 11790 }, { "grad_norm": 0.43931254744529724, "learning_rate": 9.874150266976347e-05, "loss": 0.0161, "step": 11800 }, { "grad_norm": 0.23342956602573395, "learning_rate": 9.873781360616443e-05, "loss": 0.0163, "step": 11810 }, { "grad_norm": 0.22645337879657745, "learning_rate": 9.873411921267137e-05, "loss": 0.0154, "step": 11820 }, { "grad_norm": 0.2251751571893692, "learning_rate": 9.873041948968829e-05, "loss": 0.0146, "step": 11830 }, { "grad_norm": 0.2481146901845932, "learning_rate": 9.872671443761981e-05, "loss": 0.0135, "step": 11840 }, { "grad_norm": 0.21566905081272125, "learning_rate": 9.872300405687109e-05, "loss": 0.0152, "step": 11850 }, { "grad_norm": 0.2795197367668152, "learning_rate": 9.871928834784792e-05, "loss": 0.0129, "step": 11860 }, { "grad_norm": 0.19997140765190125, "learning_rate": 9.871556731095661e-05, "loss": 0.0193, "step": 11870 }, { "grad_norm": 0.2390233874320984, "learning_rate": 9.871184094660411e-05, "loss": 0.0158, "step": 11880 }, { "grad_norm": 0.312682569026947, "learning_rate": 9.870810925519791e-05, "loss": 0.0151, "step": 11890 }, { "grad_norm": 0.27051493525505066, "learning_rate": 9.870437223714612e-05, "loss": 0.014, "step": 11900 }, { "grad_norm": 0.18106766045093536, "learning_rate": 9.87006298928574e-05, "loss": 0.0131, "step": 11910 }, { "grad_norm": 0.16554835438728333, "learning_rate": 9.869688222274103e-05, "loss": 0.0144, "step": 11920 }, { "grad_norm": 0.2270713597536087, "learning_rate": 9.869312922720681e-05, "loss": 0.0144, "step": 11930 }, { "grad_norm": 0.205253466963768, "learning_rate": 9.868937090666521e-05, "loss": 0.0149, "step": 11940 }, { "grad_norm": 0.38961970806121826, "learning_rate": 9.86856072615272e-05, "loss": 0.0163, "step": 11950 }, { "grad_norm": 0.37010785937309265, "learning_rate": 9.868183829220438e-05, "loss": 0.0259, "step": 11960 }, { "grad_norm": 0.23292602598667145, "learning_rate": 9.867806399910893e-05, "loss": 0.0172, "step": 11970 }, { "grad_norm": 0.22787627577781677, "learning_rate": 9.867428438265356e-05, "loss": 0.0147, "step": 11980 }, { "grad_norm": 0.22948312759399414, "learning_rate": 9.867049944325165e-05, "loss": 0.0138, "step": 11990 }, { "grad_norm": 0.19099970161914825, "learning_rate": 9.86667091813171e-05, "loss": 0.0142, "step": 12000 }, { "grad_norm": 0.224313884973526, "learning_rate": 9.866291359726438e-05, "loss": 0.0141, "step": 12010 }, { "grad_norm": 0.2462412565946579, "learning_rate": 9.865911269150861e-05, "loss": 0.016, "step": 12020 }, { "grad_norm": 0.1874728500843048, "learning_rate": 9.865530646446544e-05, "loss": 0.0156, "step": 12030 }, { "grad_norm": 0.2888890206813812, "learning_rate": 9.86514949165511e-05, "loss": 0.0165, "step": 12040 }, { "grad_norm": 0.23338459432125092, "learning_rate": 9.864767804818243e-05, "loss": 0.0167, "step": 12050 }, { "grad_norm": 0.3425542116165161, "learning_rate": 9.86438558597768e-05, "loss": 0.0165, "step": 12060 }, { "grad_norm": 0.1809963434934616, "learning_rate": 9.864002835175225e-05, "loss": 0.0155, "step": 12070 }, { "grad_norm": 0.11758827418088913, "learning_rate": 9.863619552452734e-05, "loss": 0.0135, "step": 12080 }, { "grad_norm": 0.16921547055244446, "learning_rate": 9.863235737852119e-05, "loss": 0.0155, "step": 12090 }, { "grad_norm": 0.19867339730262756, "learning_rate": 9.862851391415356e-05, "loss": 0.0137, "step": 12100 }, { "grad_norm": 0.23058421909809113, "learning_rate": 9.862466513184477e-05, "loss": 0.0119, "step": 12110 }, { "grad_norm": 0.2563457787036896, "learning_rate": 9.86208110320157e-05, "loss": 0.0133, "step": 12120 }, { "grad_norm": 0.29156550765037537, "learning_rate": 9.861695161508784e-05, "loss": 0.0165, "step": 12130 }, { "grad_norm": 0.17639905214309692, "learning_rate": 9.861308688148324e-05, "loss": 0.0146, "step": 12140 }, { "grad_norm": 0.20930419862270355, "learning_rate": 9.860921683162455e-05, "loss": 0.0121, "step": 12150 }, { "grad_norm": 0.3103286325931549, "learning_rate": 9.860534146593499e-05, "loss": 0.0132, "step": 12160 }, { "grad_norm": 0.20000693202018738, "learning_rate": 9.860146078483836e-05, "loss": 0.0114, "step": 12170 }, { "grad_norm": 0.19021505117416382, "learning_rate": 9.859757478875905e-05, "loss": 0.0122, "step": 12180 }, { "grad_norm": 0.1818147748708725, "learning_rate": 9.859368347812204e-05, "loss": 0.0098, "step": 12190 }, { "grad_norm": 0.2698618769645691, "learning_rate": 9.858978685335285e-05, "loss": 0.013, "step": 12200 }, { "grad_norm": 0.2680658996105194, "learning_rate": 9.858588491487763e-05, "loss": 0.0132, "step": 12210 }, { "grad_norm": 0.17595575749874115, "learning_rate": 9.858197766312308e-05, "loss": 0.011, "step": 12220 }, { "grad_norm": 0.15580879151821136, "learning_rate": 9.857806509851649e-05, "loss": 0.0134, "step": 12230 }, { "grad_norm": 0.3059719502925873, "learning_rate": 9.857414722148574e-05, "loss": 0.0163, "step": 12240 }, { "grad_norm": 0.2815920412540436, "learning_rate": 9.857022403245928e-05, "loss": 0.0127, "step": 12250 }, { "grad_norm": 0.1655087023973465, "learning_rate": 9.856629553186615e-05, "loss": 0.0112, "step": 12260 }, { "grad_norm": 0.1736418753862381, "learning_rate": 9.856236172013595e-05, "loss": 0.0119, "step": 12270 }, { "grad_norm": 0.12636613845825195, "learning_rate": 9.85584225976989e-05, "loss": 0.0089, "step": 12280 }, { "grad_norm": 0.15374262630939484, "learning_rate": 9.855447816498575e-05, "loss": 0.009, "step": 12290 }, { "grad_norm": 0.23446723818778992, "learning_rate": 9.855052842242787e-05, "loss": 0.0141, "step": 12300 }, { "grad_norm": 0.1946077048778534, "learning_rate": 9.85465733704572e-05, "loss": 0.0131, "step": 12310 }, { "grad_norm": 0.19492721557617188, "learning_rate": 9.854261300950624e-05, "loss": 0.0123, "step": 12320 }, { "grad_norm": 0.14542928338050842, "learning_rate": 9.853864734000813e-05, "loss": 0.0106, "step": 12330 }, { "grad_norm": 0.19002185761928558, "learning_rate": 9.85346763623965e-05, "loss": 0.0108, "step": 12340 }, { "grad_norm": 0.18818441033363342, "learning_rate": 9.853070007710564e-05, "loss": 0.0121, "step": 12350 }, { "grad_norm": 0.15269482135772705, "learning_rate": 9.85267184845704e-05, "loss": 0.0101, "step": 12360 }, { "grad_norm": 0.11972050368785858, "learning_rate": 9.852273158522616e-05, "loss": 0.0088, "step": 12370 }, { "grad_norm": 0.19252640008926392, "learning_rate": 9.851873937950896e-05, "loss": 0.0096, "step": 12380 }, { "grad_norm": 0.2208021581172943, "learning_rate": 9.851474186785537e-05, "loss": 0.0116, "step": 12390 }, { "grad_norm": 0.24767838418483734, "learning_rate": 9.851073905070254e-05, "loss": 0.0123, "step": 12400 }, { "grad_norm": 0.2885057330131531, "learning_rate": 9.850673092848824e-05, "loss": 0.0099, "step": 12410 }, { "grad_norm": 0.23195694386959076, "learning_rate": 9.850271750165077e-05, "loss": 0.0125, "step": 12420 }, { "grad_norm": 0.14925344288349152, "learning_rate": 9.849869877062902e-05, "loss": 0.0133, "step": 12430 }, { "grad_norm": 0.18720997869968414, "learning_rate": 9.849467473586252e-05, "loss": 0.011, "step": 12440 }, { "grad_norm": 0.2411753535270691, "learning_rate": 9.849064539779127e-05, "loss": 0.0108, "step": 12450 }, { "grad_norm": 0.17826558649539948, "learning_rate": 9.848661075685594e-05, "loss": 0.0102, "step": 12460 }, { "grad_norm": 0.18673387169837952, "learning_rate": 9.848257081349778e-05, "loss": 0.0137, "step": 12470 }, { "grad_norm": 0.16251260042190552, "learning_rate": 9.847852556815856e-05, "loss": 0.011, "step": 12480 }, { "grad_norm": 0.15665338933467865, "learning_rate": 9.847447502128067e-05, "loss": 0.0145, "step": 12490 }, { "grad_norm": 0.3496224880218506, "learning_rate": 9.847041917330708e-05, "loss": 0.0099, "step": 12500 }, { "grad_norm": 0.2148694396018982, "learning_rate": 9.846635802468132e-05, "loss": 0.0123, "step": 12510 }, { "grad_norm": 0.23879824578762054, "learning_rate": 9.84622915758475e-05, "loss": 0.0104, "step": 12520 }, { "grad_norm": 0.17844875156879425, "learning_rate": 9.845821982725034e-05, "loss": 0.0108, "step": 12530 }, { "grad_norm": 0.19929185509681702, "learning_rate": 9.845414277933514e-05, "loss": 0.0097, "step": 12540 }, { "grad_norm": 0.15159310400485992, "learning_rate": 9.845006043254771e-05, "loss": 0.0104, "step": 12550 }, { "grad_norm": 0.1827370822429657, "learning_rate": 9.844597278733451e-05, "loss": 0.0124, "step": 12560 }, { "grad_norm": 0.21277903020381927, "learning_rate": 9.844187984414259e-05, "loss": 0.0117, "step": 12570 }, { "grad_norm": 0.1941697895526886, "learning_rate": 9.84377816034195e-05, "loss": 0.013, "step": 12580 }, { "grad_norm": 0.2581079602241516, "learning_rate": 9.843367806561345e-05, "loss": 0.0104, "step": 12590 }, { "grad_norm": 0.2553401589393616, "learning_rate": 9.842956923117317e-05, "loss": 0.0102, "step": 12600 }, { "grad_norm": 0.15643568336963654, "learning_rate": 9.842545510054802e-05, "loss": 0.0091, "step": 12610 }, { "grad_norm": 0.22049231827259064, "learning_rate": 9.842133567418792e-05, "loss": 0.0107, "step": 12620 }, { "grad_norm": 0.14937473833560944, "learning_rate": 9.841721095254333e-05, "loss": 0.0131, "step": 12630 }, { "grad_norm": 0.18830302357673645, "learning_rate": 9.841308093606537e-05, "loss": 0.0137, "step": 12640 }, { "grad_norm": 0.1327027678489685, "learning_rate": 9.840894562520565e-05, "loss": 0.0108, "step": 12650 }, { "grad_norm": 0.13571059703826904, "learning_rate": 9.840480502041642e-05, "loss": 0.01, "step": 12660 }, { "grad_norm": 0.19506755471229553, "learning_rate": 9.840065912215049e-05, "loss": 0.0099, "step": 12670 }, { "grad_norm": 0.175992950797081, "learning_rate": 9.839650793086124e-05, "loss": 0.0098, "step": 12680 }, { "grad_norm": 0.21072693169116974, "learning_rate": 9.839235144700265e-05, "loss": 0.0133, "step": 12690 }, { "grad_norm": 0.1538083553314209, "learning_rate": 9.838818967102926e-05, "loss": 0.0108, "step": 12700 }, { "grad_norm": 0.20639942586421967, "learning_rate": 9.83840226033962e-05, "loss": 0.0122, "step": 12710 }, { "grad_norm": 0.14372995495796204, "learning_rate": 9.837985024455918e-05, "loss": 0.0121, "step": 12720 }, { "grad_norm": 0.18390406668186188, "learning_rate": 9.837567259497447e-05, "loss": 0.0105, "step": 12730 }, { "grad_norm": 0.22159312665462494, "learning_rate": 9.837148965509894e-05, "loss": 0.0085, "step": 12740 }, { "grad_norm": 0.1600940227508545, "learning_rate": 9.836730142539001e-05, "loss": 0.0099, "step": 12750 }, { "grad_norm": 0.13252738118171692, "learning_rate": 9.836310790630574e-05, "loss": 0.0099, "step": 12760 }, { "grad_norm": 0.22113557159900665, "learning_rate": 9.83589090983047e-05, "loss": 0.0106, "step": 12770 }, { "grad_norm": 0.3049660325050354, "learning_rate": 9.835470500184605e-05, "loss": 0.0115, "step": 12780 }, { "grad_norm": 0.19674278795719147, "learning_rate": 9.835049561738957e-05, "loss": 0.0109, "step": 12790 }, { "grad_norm": 0.17295381426811218, "learning_rate": 9.834628094539558e-05, "loss": 0.0114, "step": 12800 }, { "grad_norm": 0.19368286430835724, "learning_rate": 9.834206098632499e-05, "loss": 0.0104, "step": 12810 }, { "grad_norm": 0.23641693592071533, "learning_rate": 9.833783574063931e-05, "loss": 0.0107, "step": 12820 }, { "grad_norm": 0.23551815748214722, "learning_rate": 9.833360520880058e-05, "loss": 0.0112, "step": 12830 }, { "grad_norm": 0.18807092308998108, "learning_rate": 9.832936939127144e-05, "loss": 0.0119, "step": 12840 }, { "grad_norm": 0.2653006911277771, "learning_rate": 9.832512828851515e-05, "loss": 0.0119, "step": 12850 }, { "grad_norm": 0.23895367980003357, "learning_rate": 9.832088190099546e-05, "loss": 0.0104, "step": 12860 }, { "grad_norm": 0.2923926115036011, "learning_rate": 9.831663022917679e-05, "loss": 0.0121, "step": 12870 }, { "grad_norm": 0.24376283586025238, "learning_rate": 9.831237327352407e-05, "loss": 0.0128, "step": 12880 }, { "grad_norm": 0.2852049171924591, "learning_rate": 9.830811103450286e-05, "loss": 0.012, "step": 12890 }, { "grad_norm": 0.20624803006649017, "learning_rate": 9.830384351257924e-05, "loss": 0.0096, "step": 12900 }, { "grad_norm": 0.1999959796667099, "learning_rate": 9.829957070821993e-05, "loss": 0.0099, "step": 12910 }, { "grad_norm": 0.15585915744304657, "learning_rate": 9.829529262189218e-05, "loss": 0.0112, "step": 12920 }, { "grad_norm": 0.18297339975833893, "learning_rate": 9.829100925406385e-05, "loss": 0.0117, "step": 12930 }, { "grad_norm": 0.1344749480485916, "learning_rate": 9.828672060520333e-05, "loss": 0.0102, "step": 12940 }, { "grad_norm": 0.25106316804885864, "learning_rate": 9.828242667577966e-05, "loss": 0.0107, "step": 12950 }, { "grad_norm": 0.12839598953723907, "learning_rate": 9.82781274662624e-05, "loss": 0.0096, "step": 12960 }, { "grad_norm": 0.18070706725120544, "learning_rate": 9.82738229771217e-05, "loss": 0.0104, "step": 12970 }, { "grad_norm": 0.17635315656661987, "learning_rate": 9.826951320882829e-05, "loss": 0.0092, "step": 12980 }, { "grad_norm": 0.20186153054237366, "learning_rate": 9.826519816185351e-05, "loss": 0.0087, "step": 12990 }, { "grad_norm": 0.22713495790958405, "learning_rate": 9.826087783666921e-05, "loss": 0.0106, "step": 13000 }, { "grad_norm": 0.15593765676021576, "learning_rate": 9.825655223374787e-05, "loss": 0.0098, "step": 13010 }, { "grad_norm": 0.19873633980751038, "learning_rate": 9.825222135356253e-05, "loss": 0.0105, "step": 13020 }, { "grad_norm": 0.2515781819820404, "learning_rate": 9.82478851965868e-05, "loss": 0.0096, "step": 13030 }, { "grad_norm": 0.20747701823711395, "learning_rate": 9.82435437632949e-05, "loss": 0.0105, "step": 13040 }, { "grad_norm": 0.14175982773303986, "learning_rate": 9.823919705416158e-05, "loss": 0.0147, "step": 13050 }, { "grad_norm": 0.24454207718372345, "learning_rate": 9.82348450696622e-05, "loss": 0.0135, "step": 13060 }, { "grad_norm": 0.2625347673892975, "learning_rate": 9.823048781027268e-05, "loss": 0.0108, "step": 13070 }, { "grad_norm": 0.1610487401485443, "learning_rate": 9.822612527646953e-05, "loss": 0.0125, "step": 13080 }, { "grad_norm": 0.20517520606517792, "learning_rate": 9.822175746872984e-05, "loss": 0.0106, "step": 13090 }, { "grad_norm": 0.20949465036392212, "learning_rate": 9.821738438753123e-05, "loss": 0.0112, "step": 13100 }, { "grad_norm": 0.17376106977462769, "learning_rate": 9.821300603335196e-05, "loss": 0.01, "step": 13110 }, { "grad_norm": 0.156681090593338, "learning_rate": 9.820862240667085e-05, "loss": 0.0093, "step": 13120 }, { "grad_norm": 0.16448557376861572, "learning_rate": 9.820423350796726e-05, "loss": 0.0099, "step": 13130 }, { "grad_norm": 0.21235163509845734, "learning_rate": 9.819983933772118e-05, "loss": 0.0084, "step": 13140 }, { "grad_norm": 0.15472091734409332, "learning_rate": 9.819543989641314e-05, "loss": 0.0094, "step": 13150 }, { "grad_norm": 0.19528715312480927, "learning_rate": 9.819103518452423e-05, "loss": 0.0103, "step": 13160 }, { "grad_norm": 0.14514955878257751, "learning_rate": 9.818662520253618e-05, "loss": 0.0096, "step": 13170 }, { "grad_norm": 0.1839916855096817, "learning_rate": 9.818220995093126e-05, "loss": 0.0095, "step": 13180 }, { "grad_norm": 0.1869998425245285, "learning_rate": 9.817778943019228e-05, "loss": 0.0129, "step": 13190 }, { "grad_norm": 0.23478208482265472, "learning_rate": 9.81733636408027e-05, "loss": 0.0125, "step": 13200 }, { "grad_norm": 0.25286516547203064, "learning_rate": 9.816893258324649e-05, "loss": 0.0135, "step": 13210 }, { "grad_norm": 0.18661989271640778, "learning_rate": 9.816449625800823e-05, "loss": 0.0097, "step": 13220 }, { "grad_norm": 0.22138914465904236, "learning_rate": 9.816005466557308e-05, "loss": 0.013, "step": 13230 }, { "grad_norm": 0.16664648056030273, "learning_rate": 9.815560780642674e-05, "loss": 0.0093, "step": 13240 }, { "grad_norm": 0.21933266520500183, "learning_rate": 9.815115568105555e-05, "loss": 0.0104, "step": 13250 }, { "grad_norm": 0.25786730647087097, "learning_rate": 9.814669828994638e-05, "loss": 0.0142, "step": 13260 }, { "grad_norm": 0.22464042901992798, "learning_rate": 9.814223563358665e-05, "loss": 0.0125, "step": 13270 }, { "grad_norm": 0.1699770838022232, "learning_rate": 9.813776771246443e-05, "loss": 0.0104, "step": 13280 }, { "grad_norm": 0.2781293988227844, "learning_rate": 9.813329452706829e-05, "loss": 0.0123, "step": 13290 }, { "grad_norm": 0.22597889602184296, "learning_rate": 9.812881607788744e-05, "loss": 0.0117, "step": 13300 }, { "grad_norm": 0.17625920474529266, "learning_rate": 9.812433236541163e-05, "loss": 0.0105, "step": 13310 }, { "grad_norm": 0.19720813632011414, "learning_rate": 9.811984339013116e-05, "loss": 0.0118, "step": 13320 }, { "grad_norm": 0.2113656997680664, "learning_rate": 9.811534915253698e-05, "loss": 0.0115, "step": 13330 }, { "grad_norm": 0.21422524750232697, "learning_rate": 9.811084965312056e-05, "loss": 0.0127, "step": 13340 }, { "grad_norm": 0.2555205225944519, "learning_rate": 9.810634489237396e-05, "loss": 0.0118, "step": 13350 }, { "grad_norm": 0.2283104807138443, "learning_rate": 9.81018348707898e-05, "loss": 0.0124, "step": 13360 }, { "grad_norm": 0.22860416769981384, "learning_rate": 9.809731958886131e-05, "loss": 0.0128, "step": 13370 }, { "grad_norm": 0.17560705542564392, "learning_rate": 9.809279904708224e-05, "loss": 0.0124, "step": 13380 }, { "grad_norm": 0.18832632899284363, "learning_rate": 9.808827324594699e-05, "loss": 0.0095, "step": 13390 }, { "grad_norm": 0.17050619423389435, "learning_rate": 9.808374218595046e-05, "loss": 0.0119, "step": 13400 }, { "grad_norm": 0.14745648205280304, "learning_rate": 9.80792058675882e-05, "loss": 0.0106, "step": 13410 }, { "grad_norm": 0.16393771767616272, "learning_rate": 9.807466429135627e-05, "loss": 0.0101, "step": 13420 }, { "grad_norm": 0.20761427283287048, "learning_rate": 9.807011745775132e-05, "loss": 0.0123, "step": 13430 }, { "grad_norm": 0.15347932279109955, "learning_rate": 9.806556536727061e-05, "loss": 0.0117, "step": 13440 }, { "grad_norm": 0.17154806852340698, "learning_rate": 9.806100802041193e-05, "loss": 0.01, "step": 13450 }, { "grad_norm": 0.16990581154823303, "learning_rate": 9.805644541767368e-05, "loss": 0.009, "step": 13460 }, { "grad_norm": 0.16867011785507202, "learning_rate": 9.805187755955478e-05, "loss": 0.0131, "step": 13470 }, { "grad_norm": 0.21115116775035858, "learning_rate": 9.804730444655483e-05, "loss": 0.0109, "step": 13480 }, { "grad_norm": 0.24133624136447906, "learning_rate": 9.804272607917388e-05, "loss": 0.0101, "step": 13490 }, { "grad_norm": 0.17959649860858917, "learning_rate": 9.803814245791265e-05, "loss": 0.0121, "step": 13500 }, { "grad_norm": 0.14846760034561157, "learning_rate": 9.803355358327239e-05, "loss": 0.011, "step": 13510 }, { "grad_norm": 0.19266405701637268, "learning_rate": 9.802895945575492e-05, "loss": 0.0103, "step": 13520 }, { "grad_norm": 0.16540387272834778, "learning_rate": 9.802436007586266e-05, "loss": 0.009, "step": 13530 }, { "grad_norm": 0.19137458503246307, "learning_rate": 9.801975544409858e-05, "loss": 0.0108, "step": 13540 }, { "grad_norm": 0.14836126565933228, "learning_rate": 9.801514556096625e-05, "loss": 0.0139, "step": 13550 }, { "grad_norm": 0.1521512120962143, "learning_rate": 9.801053042696977e-05, "loss": 0.0104, "step": 13560 }, { "grad_norm": 0.17585450410842896, "learning_rate": 9.800591004261388e-05, "loss": 0.0113, "step": 13570 }, { "grad_norm": 0.15905146300792694, "learning_rate": 9.800128440840385e-05, "loss": 0.0114, "step": 13580 }, { "grad_norm": 0.17541860044002533, "learning_rate": 9.799665352484552e-05, "loss": 0.0134, "step": 13590 }, { "grad_norm": 0.21812397241592407, "learning_rate": 9.799201739244532e-05, "loss": 0.0119, "step": 13600 }, { "grad_norm": 0.25742000341415405, "learning_rate": 9.798737601171025e-05, "loss": 0.0111, "step": 13610 }, { "grad_norm": 0.2744276821613312, "learning_rate": 9.79827293831479e-05, "loss": 0.0108, "step": 13620 }, { "grad_norm": 0.18172065913677216, "learning_rate": 9.797807750726638e-05, "loss": 0.0128, "step": 13630 }, { "grad_norm": 0.17301522195339203, "learning_rate": 9.797342038457446e-05, "loss": 0.0095, "step": 13640 }, { "grad_norm": 0.16688215732574463, "learning_rate": 9.796875801558141e-05, "loss": 0.0108, "step": 13650 }, { "grad_norm": 0.17544448375701904, "learning_rate": 9.79640904007971e-05, "loss": 0.0106, "step": 13660 }, { "grad_norm": 0.16743549704551697, "learning_rate": 9.795941754073199e-05, "loss": 0.0102, "step": 13670 }, { "grad_norm": 0.17766447365283966, "learning_rate": 9.795473943589705e-05, "loss": 0.0121, "step": 13680 }, { "grad_norm": 0.2778059244155884, "learning_rate": 9.795005608680394e-05, "loss": 0.014, "step": 13690 }, { "grad_norm": 0.24181459844112396, "learning_rate": 9.794536749396477e-05, "loss": 0.0142, "step": 13700 }, { "grad_norm": 0.19673049449920654, "learning_rate": 9.79406736578923e-05, "loss": 0.0156, "step": 13710 }, { "grad_norm": 0.1408630907535553, "learning_rate": 9.793597457909984e-05, "loss": 0.0124, "step": 13720 }, { "grad_norm": 0.13811790943145752, "learning_rate": 9.793127025810127e-05, "loss": 0.0137, "step": 13730 }, { "grad_norm": 0.16561010479927063, "learning_rate": 9.792656069541104e-05, "loss": 0.0104, "step": 13740 }, { "grad_norm": 0.17769204080104828, "learning_rate": 9.79218458915442e-05, "loss": 0.012, "step": 13750 }, { "grad_norm": 0.13917042315006256, "learning_rate": 9.791712584701634e-05, "loss": 0.0097, "step": 13760 }, { "grad_norm": 0.19814042747020721, "learning_rate": 9.791240056234364e-05, "loss": 0.0087, "step": 13770 }, { "grad_norm": 0.18798492848873138, "learning_rate": 9.790767003804283e-05, "loss": 0.0082, "step": 13780 }, { "grad_norm": 0.16282378137111664, "learning_rate": 9.790293427463126e-05, "loss": 0.0092, "step": 13790 }, { "grad_norm": 0.19280242919921875, "learning_rate": 9.789819327262684e-05, "loss": 0.0075, "step": 13800 }, { "grad_norm": 0.1510276198387146, "learning_rate": 9.7893447032548e-05, "loss": 0.0081, "step": 13810 }, { "grad_norm": 0.20844322443008423, "learning_rate": 9.78886955549138e-05, "loss": 0.0095, "step": 13820 }, { "grad_norm": 0.22965222597122192, "learning_rate": 9.788393884024387e-05, "loss": 0.0115, "step": 13830 }, { "grad_norm": 0.16965535283088684, "learning_rate": 9.787917688905836e-05, "loss": 0.0108, "step": 13840 }, { "grad_norm": 0.4755784869194031, "learning_rate": 9.787440970187807e-05, "loss": 0.0131, "step": 13850 }, { "grad_norm": 0.27475568652153015, "learning_rate": 9.786963727922429e-05, "loss": 0.0146, "step": 13860 }, { "grad_norm": 0.2452457845211029, "learning_rate": 9.786485962161897e-05, "loss": 0.0155, "step": 13870 }, { "grad_norm": 0.30267664790153503, "learning_rate": 9.786007672958455e-05, "loss": 0.0129, "step": 13880 }, { "grad_norm": 0.19683554768562317, "learning_rate": 9.78552886036441e-05, "loss": 0.012, "step": 13890 }, { "grad_norm": 0.24859954416751862, "learning_rate": 9.785049524432124e-05, "loss": 0.0104, "step": 13900 }, { "grad_norm": 0.26513153314590454, "learning_rate": 9.784569665214016e-05, "loss": 0.0118, "step": 13910 }, { "grad_norm": 0.19400469958782196, "learning_rate": 9.784089282762563e-05, "loss": 0.0106, "step": 13920 }, { "grad_norm": 0.222533717751503, "learning_rate": 9.7836083771303e-05, "loss": 0.0129, "step": 13930 }, { "grad_norm": 0.159382164478302, "learning_rate": 9.783126948369817e-05, "loss": 0.01, "step": 13940 }, { "grad_norm": 0.1986059546470642, "learning_rate": 9.78264499653376e-05, "loss": 0.0106, "step": 13950 }, { "grad_norm": 0.13198885321617126, "learning_rate": 9.782162521674838e-05, "loss": 0.0101, "step": 13960 }, { "grad_norm": 0.24981249868869781, "learning_rate": 9.781679523845812e-05, "loss": 0.0118, "step": 13970 }, { "grad_norm": 0.16419583559036255, "learning_rate": 9.781196003099502e-05, "loss": 0.0108, "step": 13980 }, { "grad_norm": 0.17871622741222382, "learning_rate": 9.780711959488786e-05, "loss": 0.0106, "step": 13990 }, { "grad_norm": 0.16928856074810028, "learning_rate": 9.780227393066599e-05, "loss": 0.0093, "step": 14000 }, { "grad_norm": 0.1738281399011612, "learning_rate": 9.77974230388593e-05, "loss": 0.0111, "step": 14010 }, { "grad_norm": 0.19562217593193054, "learning_rate": 9.779256691999829e-05, "loss": 0.0111, "step": 14020 }, { "grad_norm": 0.1784169226884842, "learning_rate": 9.778770557461403e-05, "loss": 0.0098, "step": 14030 }, { "grad_norm": 0.20192411541938782, "learning_rate": 9.778283900323812e-05, "loss": 0.0087, "step": 14040 }, { "grad_norm": 0.16162212193012238, "learning_rate": 9.777796720640277e-05, "loss": 0.0096, "step": 14050 }, { "grad_norm": 0.18420609831809998, "learning_rate": 9.777309018464078e-05, "loss": 0.0077, "step": 14060 }, { "grad_norm": 0.13422735035419464, "learning_rate": 9.776820793848547e-05, "loss": 0.0089, "step": 14070 }, { "grad_norm": 0.1503877341747284, "learning_rate": 9.776332046847075e-05, "loss": 0.0083, "step": 14080 }, { "grad_norm": 0.21528476476669312, "learning_rate": 9.775842777513111e-05, "loss": 0.0081, "step": 14090 }, { "grad_norm": 0.13916443288326263, "learning_rate": 9.775352985900163e-05, "loss": 0.0087, "step": 14100 }, { "grad_norm": 0.2903209328651428, "learning_rate": 9.774862672061791e-05, "loss": 0.0122, "step": 14110 }, { "grad_norm": 0.14850673079490662, "learning_rate": 9.774371836051616e-05, "loss": 0.0117, "step": 14120 }, { "grad_norm": 0.17271669209003448, "learning_rate": 9.773880477923315e-05, "loss": 0.0103, "step": 14130 }, { "grad_norm": 0.18895584344863892, "learning_rate": 9.773388597730623e-05, "loss": 0.0101, "step": 14140 }, { "grad_norm": 0.1675473153591156, "learning_rate": 9.77289619552733e-05, "loss": 0.0081, "step": 14150 }, { "grad_norm": 0.14042095839977264, "learning_rate": 9.772403271367285e-05, "loss": 0.0101, "step": 14160 }, { "grad_norm": 0.1212959736585617, "learning_rate": 9.771909825304396e-05, "loss": 0.0125, "step": 14170 }, { "grad_norm": 0.19649957120418549, "learning_rate": 9.771415857392619e-05, "loss": 0.0092, "step": 14180 }, { "grad_norm": 0.21045120060443878, "learning_rate": 9.770921367685978e-05, "loss": 0.0124, "step": 14190 }, { "grad_norm": 0.18109826743602753, "learning_rate": 9.770426356238551e-05, "loss": 0.0099, "step": 14200 }, { "grad_norm": 0.21660946309566498, "learning_rate": 9.769930823104469e-05, "loss": 0.01, "step": 14210 }, { "grad_norm": 0.1938304603099823, "learning_rate": 9.769434768337926e-05, "loss": 0.0092, "step": 14220 }, { "grad_norm": 0.17185041308403015, "learning_rate": 9.768938191993164e-05, "loss": 0.0097, "step": 14230 }, { "grad_norm": 0.1672002375125885, "learning_rate": 9.768441094124494e-05, "loss": 0.009, "step": 14240 }, { "grad_norm": 0.18118326365947723, "learning_rate": 9.767943474786275e-05, "loss": 0.0097, "step": 14250 }, { "grad_norm": 0.1958937644958496, "learning_rate": 9.767445334032923e-05, "loss": 0.0087, "step": 14260 }, { "grad_norm": 0.1771288961172104, "learning_rate": 9.766946671918919e-05, "loss": 0.012, "step": 14270 }, { "grad_norm": 0.19303639233112335, "learning_rate": 9.766447488498796e-05, "loss": 0.0084, "step": 14280 }, { "grad_norm": 0.16867728531360626, "learning_rate": 9.765947783827139e-05, "loss": 0.0083, "step": 14290 }, { "grad_norm": 0.1852671355009079, "learning_rate": 9.765447557958599e-05, "loss": 0.0108, "step": 14300 }, { "grad_norm": 0.19099697470664978, "learning_rate": 9.764946810947879e-05, "loss": 0.0116, "step": 14310 }, { "grad_norm": 0.16137610375881195, "learning_rate": 9.764445542849738e-05, "loss": 0.0111, "step": 14320 }, { "grad_norm": 0.19291122257709503, "learning_rate": 9.763943753718998e-05, "loss": 0.0105, "step": 14330 }, { "grad_norm": 0.14516916871070862, "learning_rate": 9.76344144361053e-05, "loss": 0.0102, "step": 14340 }, { "grad_norm": 0.18689607083797455, "learning_rate": 9.762938612579269e-05, "loss": 0.0097, "step": 14350 }, { "grad_norm": 0.15493352711200714, "learning_rate": 9.762435260680202e-05, "loss": 0.0091, "step": 14360 }, { "grad_norm": 0.23306263983249664, "learning_rate": 9.761931387968373e-05, "loss": 0.0093, "step": 14370 }, { "grad_norm": 0.21361678838729858, "learning_rate": 9.76142699449889e-05, "loss": 0.0091, "step": 14380 }, { "grad_norm": 0.21374869346618652, "learning_rate": 9.760922080326908e-05, "loss": 0.0137, "step": 14390 }, { "grad_norm": 0.20133498311042786, "learning_rate": 9.760416645507644e-05, "loss": 0.0107, "step": 14400 }, { "grad_norm": 0.22288228571414948, "learning_rate": 9.759910690096375e-05, "loss": 0.0095, "step": 14410 }, { "grad_norm": 0.1568160355091095, "learning_rate": 9.759404214148429e-05, "loss": 0.0083, "step": 14420 }, { "grad_norm": 0.13590215146541595, "learning_rate": 9.758897217719191e-05, "loss": 0.0098, "step": 14430 }, { "grad_norm": 0.23305615782737732, "learning_rate": 9.758389700864113e-05, "loss": 0.0109, "step": 14440 }, { "grad_norm": 0.12958258390426636, "learning_rate": 9.757881663638688e-05, "loss": 0.0096, "step": 14450 }, { "grad_norm": 0.19745375216007233, "learning_rate": 9.757373106098478e-05, "loss": 0.0098, "step": 14460 }, { "grad_norm": 0.1725863665342331, "learning_rate": 9.756864028299097e-05, "loss": 0.0088, "step": 14470 }, { "grad_norm": 0.22393645346164703, "learning_rate": 9.75635443029622e-05, "loss": 0.0095, "step": 14480 }, { "grad_norm": 0.18556775152683258, "learning_rate": 9.755844312145572e-05, "loss": 0.0104, "step": 14490 }, { "grad_norm": 0.17191781103610992, "learning_rate": 9.755333673902941e-05, "loss": 0.01, "step": 14500 }, { "grad_norm": 0.16161753237247467, "learning_rate": 9.75482251562417e-05, "loss": 0.0104, "step": 14510 }, { "grad_norm": 0.21417026221752167, "learning_rate": 9.754310837365155e-05, "loss": 0.0086, "step": 14520 }, { "grad_norm": 0.1632300168275833, "learning_rate": 9.753798639181856e-05, "loss": 0.0102, "step": 14530 }, { "grad_norm": 0.17652587592601776, "learning_rate": 9.753285921130286e-05, "loss": 0.0084, "step": 14540 }, { "grad_norm": 0.17741116881370544, "learning_rate": 9.752772683266512e-05, "loss": 0.0109, "step": 14550 }, { "grad_norm": 0.12239072471857071, "learning_rate": 9.752258925646665e-05, "loss": 0.0099, "step": 14560 }, { "grad_norm": 0.15462473034858704, "learning_rate": 9.751744648326926e-05, "loss": 0.0083, "step": 14570 }, { "grad_norm": 0.1925172209739685, "learning_rate": 9.751229851363536e-05, "loss": 0.0092, "step": 14580 }, { "grad_norm": 0.2292294055223465, "learning_rate": 9.750714534812793e-05, "loss": 0.0089, "step": 14590 }, { "grad_norm": 0.23780293762683868, "learning_rate": 9.750198698731053e-05, "loss": 0.0108, "step": 14600 }, { "grad_norm": 0.1891891211271286, "learning_rate": 9.749682343174722e-05, "loss": 0.0103, "step": 14610 }, { "grad_norm": 0.15641388297080994, "learning_rate": 9.749165468200272e-05, "loss": 0.008, "step": 14620 }, { "grad_norm": 0.17887446284294128, "learning_rate": 9.748648073864229e-05, "loss": 0.0097, "step": 14630 }, { "grad_norm": 0.16557960212230682, "learning_rate": 9.748130160223168e-05, "loss": 0.0101, "step": 14640 }, { "grad_norm": 0.14230963587760925, "learning_rate": 9.747611727333734e-05, "loss": 0.0086, "step": 14650 }, { "grad_norm": 0.12038149684667587, "learning_rate": 9.74709277525262e-05, "loss": 0.0092, "step": 14660 }, { "grad_norm": 0.18660250306129456, "learning_rate": 9.746573304036576e-05, "loss": 0.0105, "step": 14670 }, { "grad_norm": 0.1757165640592575, "learning_rate": 9.746053313742412e-05, "loss": 0.0097, "step": 14680 }, { "grad_norm": 0.15470989048480988, "learning_rate": 9.745532804426994e-05, "loss": 0.0089, "step": 14690 }, { "grad_norm": 0.141976460814476, "learning_rate": 9.745011776147242e-05, "loss": 0.0099, "step": 14700 }, { "grad_norm": 0.1812726855278015, "learning_rate": 9.744490228960138e-05, "loss": 0.008, "step": 14710 }, { "grad_norm": 0.161473348736763, "learning_rate": 9.743968162922713e-05, "loss": 0.0078, "step": 14720 }, { "grad_norm": 0.20487064123153687, "learning_rate": 9.743445578092064e-05, "loss": 0.0087, "step": 14730 }, { "grad_norm": 0.19837601482868195, "learning_rate": 9.742922474525338e-05, "loss": 0.0087, "step": 14740 }, { "grad_norm": 0.15211710333824158, "learning_rate": 9.742398852279741e-05, "loss": 0.0086, "step": 14750 }, { "grad_norm": 0.1867186278104782, "learning_rate": 9.741874711412535e-05, "loss": 0.0116, "step": 14760 }, { "grad_norm": 0.16920877993106842, "learning_rate": 9.741350051981042e-05, "loss": 0.0089, "step": 14770 }, { "grad_norm": 0.21723820269107819, "learning_rate": 9.740824874042633e-05, "loss": 0.0103, "step": 14780 }, { "grad_norm": 0.1704173982143402, "learning_rate": 9.740299177654746e-05, "loss": 0.0101, "step": 14790 }, { "grad_norm": 0.1567949503660202, "learning_rate": 9.739772962874867e-05, "loss": 0.0102, "step": 14800 }, { "grad_norm": 0.17952457070350647, "learning_rate": 9.739246229760541e-05, "loss": 0.0088, "step": 14810 }, { "grad_norm": 0.18171311914920807, "learning_rate": 9.738718978369376e-05, "loss": 0.0108, "step": 14820 }, { "grad_norm": 0.2377317249774933, "learning_rate": 9.738191208759025e-05, "loss": 0.0117, "step": 14830 }, { "grad_norm": 0.20886175334453583, "learning_rate": 9.73766292098721e-05, "loss": 0.0114, "step": 14840 }, { "grad_norm": 0.1813526600599289, "learning_rate": 9.737134115111699e-05, "loss": 0.0097, "step": 14850 }, { "grad_norm": 0.1964183747768402, "learning_rate": 9.736604791190323e-05, "loss": 0.0099, "step": 14860 }, { "grad_norm": 0.17087778449058533, "learning_rate": 9.73607494928097e-05, "loss": 0.0099, "step": 14870 }, { "grad_norm": 0.1833486557006836, "learning_rate": 9.735544589441581e-05, "loss": 0.0102, "step": 14880 }, { "grad_norm": 0.17024829983711243, "learning_rate": 9.735013711730154e-05, "loss": 0.0098, "step": 14890 }, { "grad_norm": 0.36066481471061707, "learning_rate": 9.734482316204747e-05, "loss": 0.015, "step": 14900 }, { "grad_norm": 0.3160833716392517, "learning_rate": 9.733950402923473e-05, "loss": 0.0203, "step": 14910 }, { "grad_norm": 0.38347139954566956, "learning_rate": 9.7334179719445e-05, "loss": 0.0172, "step": 14920 }, { "grad_norm": 0.2577243447303772, "learning_rate": 9.732885023326053e-05, "loss": 0.0147, "step": 14930 }, { "grad_norm": 0.3541359603404999, "learning_rate": 9.732351557126418e-05, "loss": 0.0161, "step": 14940 }, { "grad_norm": 0.32119524478912354, "learning_rate": 9.731817573403929e-05, "loss": 0.015, "step": 14950 }, { "grad_norm": 0.24546566605567932, "learning_rate": 9.731283072216985e-05, "loss": 0.014, "step": 14960 }, { "grad_norm": 0.42534977197647095, "learning_rate": 9.730748053624039e-05, "loss": 0.0208, "step": 14970 }, { "grad_norm": 0.24034275114536285, "learning_rate": 9.730212517683598e-05, "loss": 0.0192, "step": 14980 }, { "grad_norm": 0.2561550736427307, "learning_rate": 9.729676464454228e-05, "loss": 0.0159, "step": 14990 }, { "grad_norm": 0.2517593502998352, "learning_rate": 9.72913989399455e-05, "loss": 0.0209, "step": 15000 }, { "grad_norm": 0.24518898129463196, "learning_rate": 9.728602806363242e-05, "loss": 0.0196, "step": 15010 }, { "grad_norm": 0.31507524847984314, "learning_rate": 9.728065201619043e-05, "loss": 0.0193, "step": 15020 }, { "grad_norm": 0.2886887490749359, "learning_rate": 9.727527079820742e-05, "loss": 0.0143, "step": 15030 }, { "grad_norm": 0.21879513561725616, "learning_rate": 9.726988441027186e-05, "loss": 0.0138, "step": 15040 }, { "grad_norm": 0.2024773508310318, "learning_rate": 9.726449285297281e-05, "loss": 0.0153, "step": 15050 }, { "grad_norm": 0.18801546096801758, "learning_rate": 9.72590961268999e-05, "loss": 0.0145, "step": 15060 }, { "grad_norm": 0.16152003407478333, "learning_rate": 9.725369423264328e-05, "loss": 0.0141, "step": 15070 }, { "grad_norm": 0.19841912388801575, "learning_rate": 9.72482871707937e-05, "loss": 0.0138, "step": 15080 }, { "grad_norm": 0.1955830454826355, "learning_rate": 9.724287494194247e-05, "loss": 0.0122, "step": 15090 }, { "grad_norm": 0.14565768837928772, "learning_rate": 9.723745754668147e-05, "loss": 0.0111, "step": 15100 }, { "grad_norm": 0.2111363410949707, "learning_rate": 9.723203498560313e-05, "loss": 0.0115, "step": 15110 }, { "grad_norm": 0.18363997340202332, "learning_rate": 9.722660725930046e-05, "loss": 0.01, "step": 15120 }, { "grad_norm": 0.15317508578300476, "learning_rate": 9.722117436836702e-05, "loss": 0.0097, "step": 15130 }, { "grad_norm": 0.19812361896038055, "learning_rate": 9.721573631339696e-05, "loss": 0.0106, "step": 15140 }, { "grad_norm": 0.22423793375492096, "learning_rate": 9.721029309498494e-05, "loss": 0.0123, "step": 15150 }, { "grad_norm": 0.20554915070533752, "learning_rate": 9.720484471372627e-05, "loss": 0.0129, "step": 15160 }, { "grad_norm": 0.17004212737083435, "learning_rate": 9.719939117021673e-05, "loss": 0.0127, "step": 15170 }, { "grad_norm": 0.21662279963493347, "learning_rate": 9.719393246505275e-05, "loss": 0.0121, "step": 15180 }, { "grad_norm": 0.2717790901660919, "learning_rate": 9.718846859883128e-05, "loss": 0.0198, "step": 15190 }, { "grad_norm": 0.3164443373680115, "learning_rate": 9.718299957214982e-05, "loss": 0.0146, "step": 15200 }, { "grad_norm": 0.23871253430843353, "learning_rate": 9.717752538560646e-05, "loss": 0.013, "step": 15210 }, { "grad_norm": 0.16636183857917786, "learning_rate": 9.717204603979986e-05, "loss": 0.0111, "step": 15220 }, { "grad_norm": 0.13305269181728363, "learning_rate": 9.716656153532922e-05, "loss": 0.011, "step": 15230 }, { "grad_norm": 0.16865751147270203, "learning_rate": 9.716107187279434e-05, "loss": 0.0107, "step": 15240 }, { "grad_norm": 0.24588672816753387, "learning_rate": 9.715557705279555e-05, "loss": 0.0122, "step": 15250 }, { "grad_norm": 0.2554486393928528, "learning_rate": 9.715007707593372e-05, "loss": 0.0188, "step": 15260 }, { "grad_norm": 0.28947713971138, "learning_rate": 9.714457194281036e-05, "loss": 0.0176, "step": 15270 }, { "grad_norm": 0.23825952410697937, "learning_rate": 9.713906165402751e-05, "loss": 0.0125, "step": 15280 }, { "grad_norm": 0.22974412143230438, "learning_rate": 9.713354621018774e-05, "loss": 0.0126, "step": 15290 }, { "grad_norm": 0.17061102390289307, "learning_rate": 9.712802561189422e-05, "loss": 0.0127, "step": 15300 }, { "grad_norm": 0.15476687252521515, "learning_rate": 9.712249985975069e-05, "loss": 0.0095, "step": 15310 }, { "grad_norm": 0.13297320902347565, "learning_rate": 9.71169689543614e-05, "loss": 0.0093, "step": 15320 }, { "grad_norm": 0.19743303954601288, "learning_rate": 9.711143289633123e-05, "loss": 0.0115, "step": 15330 }, { "grad_norm": 0.18098580837249756, "learning_rate": 9.710589168626561e-05, "loss": 0.0108, "step": 15340 }, { "grad_norm": 0.15963247418403625, "learning_rate": 9.710034532477048e-05, "loss": 0.0104, "step": 15350 }, { "grad_norm": 0.17538976669311523, "learning_rate": 9.709479381245239e-05, "loss": 0.011, "step": 15360 }, { "grad_norm": 0.1532077044248581, "learning_rate": 9.708923714991847e-05, "loss": 0.0112, "step": 15370 }, { "grad_norm": 0.17685160040855408, "learning_rate": 9.708367533777638e-05, "loss": 0.0101, "step": 15380 }, { "grad_norm": 0.20472189784049988, "learning_rate": 9.707810837663431e-05, "loss": 0.0128, "step": 15390 }, { "grad_norm": 0.19521939754486084, "learning_rate": 9.707253626710113e-05, "loss": 0.0114, "step": 15400 }, { "grad_norm": 0.14106512069702148, "learning_rate": 9.706695900978613e-05, "loss": 0.0082, "step": 15410 }, { "grad_norm": 0.15820728242397308, "learning_rate": 9.706137660529926e-05, "loss": 0.012, "step": 15420 }, { "grad_norm": 0.18327495455741882, "learning_rate": 9.705578905425101e-05, "loss": 0.0089, "step": 15430 }, { "grad_norm": 0.20298226177692413, "learning_rate": 9.705019635725241e-05, "loss": 0.0108, "step": 15440 }, { "grad_norm": 0.17176678776741028, "learning_rate": 9.704459851491508e-05, "loss": 0.0079, "step": 15450 }, { "grad_norm": 0.1701579988002777, "learning_rate": 9.703899552785118e-05, "loss": 0.0083, "step": 15460 }, { "grad_norm": 0.1639123260974884, "learning_rate": 9.703338739667346e-05, "loss": 0.0075, "step": 15470 }, { "grad_norm": 0.2148457020521164, "learning_rate": 9.70277741219952e-05, "loss": 0.0111, "step": 15480 }, { "grad_norm": 0.18422651290893555, "learning_rate": 9.702215570443027e-05, "loss": 0.0097, "step": 15490 }, { "grad_norm": 0.22782769799232483, "learning_rate": 9.701653214459309e-05, "loss": 0.0119, "step": 15500 }, { "grad_norm": 0.2830345928668976, "learning_rate": 9.701090344309865e-05, "loss": 0.0125, "step": 15510 }, { "grad_norm": 0.19084054231643677, "learning_rate": 9.700526960056247e-05, "loss": 0.012, "step": 15520 }, { "grad_norm": 0.22961482405662537, "learning_rate": 9.699963061760068e-05, "loss": 0.0106, "step": 15530 }, { "grad_norm": 0.24599884450435638, "learning_rate": 9.699398649482997e-05, "loss": 0.01, "step": 15540 }, { "grad_norm": 0.20444008708000183, "learning_rate": 9.698833723286753e-05, "loss": 0.0104, "step": 15550 }, { "grad_norm": 0.1881866753101349, "learning_rate": 9.698268283233118e-05, "loss": 0.0123, "step": 15560 }, { "grad_norm": 0.1678108125925064, "learning_rate": 9.697702329383929e-05, "loss": 0.0118, "step": 15570 }, { "grad_norm": 0.3156561553478241, "learning_rate": 9.697135861801074e-05, "loss": 0.0126, "step": 15580 }, { "grad_norm": 0.17248637974262238, "learning_rate": 9.696568880546505e-05, "loss": 0.0107, "step": 15590 }, { "grad_norm": 0.17006231844425201, "learning_rate": 9.696001385682223e-05, "loss": 0.0114, "step": 15600 }, { "grad_norm": 0.17622070014476776, "learning_rate": 9.695433377270291e-05, "loss": 0.0099, "step": 15610 }, { "grad_norm": 0.264059841632843, "learning_rate": 9.694864855372824e-05, "loss": 0.0121, "step": 15620 }, { "grad_norm": 0.11246710270643234, "learning_rate": 9.694295820051995e-05, "loss": 0.0086, "step": 15630 }, { "grad_norm": 0.1425141543149948, "learning_rate": 9.693726271370032e-05, "loss": 0.009, "step": 15640 }, { "grad_norm": 0.20135322213172913, "learning_rate": 9.693156209389221e-05, "loss": 0.0107, "step": 15650 }, { "grad_norm": 0.18273811042308807, "learning_rate": 9.692585634171905e-05, "loss": 0.011, "step": 15660 }, { "grad_norm": 0.17756016552448273, "learning_rate": 9.692014545780476e-05, "loss": 0.0093, "step": 15670 }, { "grad_norm": 0.13603192567825317, "learning_rate": 9.691442944277393e-05, "loss": 0.0095, "step": 15680 }, { "grad_norm": 0.14682555198669434, "learning_rate": 9.690870829725162e-05, "loss": 0.0089, "step": 15690 }, { "grad_norm": 0.20610545575618744, "learning_rate": 9.69029820218635e-05, "loss": 0.0106, "step": 15700 }, { "grad_norm": 0.15953895449638367, "learning_rate": 9.689725061723579e-05, "loss": 0.0121, "step": 15710 }, { "grad_norm": 0.1832672655582428, "learning_rate": 9.689151408399527e-05, "loss": 0.0098, "step": 15720 }, { "grad_norm": 0.2265627533197403, "learning_rate": 9.688577242276924e-05, "loss": 0.0095, "step": 15730 }, { "grad_norm": 0.15490394830703735, "learning_rate": 9.688002563418566e-05, "loss": 0.0093, "step": 15740 }, { "grad_norm": 0.1823331117630005, "learning_rate": 9.687427371887293e-05, "loss": 0.0101, "step": 15750 }, { "grad_norm": 0.15924188494682312, "learning_rate": 9.686851667746012e-05, "loss": 0.0088, "step": 15760 }, { "grad_norm": 0.1316547691822052, "learning_rate": 9.686275451057677e-05, "loss": 0.0093, "step": 15770 }, { "grad_norm": 0.14458943903446198, "learning_rate": 9.685698721885308e-05, "loss": 0.0084, "step": 15780 }, { "grad_norm": 0.20172487199306488, "learning_rate": 9.68512148029197e-05, "loss": 0.0095, "step": 15790 }, { "grad_norm": 0.21123608946800232, "learning_rate": 9.684543726340791e-05, "loss": 0.0115, "step": 15800 }, { "grad_norm": 0.19263190031051636, "learning_rate": 9.683965460094952e-05, "loss": 0.0099, "step": 15810 }, { "grad_norm": 0.24501048028469086, "learning_rate": 9.683386681617694e-05, "loss": 0.0099, "step": 15820 }, { "grad_norm": 0.16973888874053955, "learning_rate": 9.68280739097231e-05, "loss": 0.0106, "step": 15830 }, { "grad_norm": 0.14687293767929077, "learning_rate": 9.682227588222148e-05, "loss": 0.0088, "step": 15840 }, { "grad_norm": 0.18933828175067902, "learning_rate": 9.681647273430618e-05, "loss": 0.0113, "step": 15850 }, { "grad_norm": 0.16607846319675446, "learning_rate": 9.681066446661182e-05, "loss": 0.0146, "step": 15860 }, { "grad_norm": 0.192406564950943, "learning_rate": 9.680485107977357e-05, "loss": 0.011, "step": 15870 }, { "grad_norm": 0.18828466534614563, "learning_rate": 9.679903257442716e-05, "loss": 0.0103, "step": 15880 }, { "grad_norm": 0.17601168155670166, "learning_rate": 9.679320895120891e-05, "loss": 0.01, "step": 15890 }, { "grad_norm": 0.1703299880027771, "learning_rate": 9.67873802107557e-05, "loss": 0.0093, "step": 15900 }, { "grad_norm": 0.1801305115222931, "learning_rate": 9.67815463537049e-05, "loss": 0.0084, "step": 15910 }, { "grad_norm": 0.12995003163814545, "learning_rate": 9.677570738069457e-05, "loss": 0.0078, "step": 15920 }, { "grad_norm": 0.14852945506572723, "learning_rate": 9.676986329236318e-05, "loss": 0.0076, "step": 15930 }, { "grad_norm": 0.16402247548103333, "learning_rate": 9.676401408934987e-05, "loss": 0.0085, "step": 15940 }, { "grad_norm": 0.19388413429260254, "learning_rate": 9.675815977229428e-05, "loss": 0.0103, "step": 15950 }, { "grad_norm": 0.14425630867481232, "learning_rate": 9.675230034183664e-05, "loss": 0.0091, "step": 15960 }, { "grad_norm": 0.3870322108268738, "learning_rate": 9.674643579861773e-05, "loss": 0.0099, "step": 15970 }, { "grad_norm": 0.2100217640399933, "learning_rate": 9.674056614327886e-05, "loss": 0.0119, "step": 15980 }, { "grad_norm": 0.17840218544006348, "learning_rate": 9.673469137646198e-05, "loss": 0.0109, "step": 15990 }, { "grad_norm": 0.2840651273727417, "learning_rate": 9.67288114988095e-05, "loss": 0.0106, "step": 16000 }, { "grad_norm": 0.13219502568244934, "learning_rate": 9.672292651096447e-05, "loss": 0.0087, "step": 16010 }, { "grad_norm": 0.17964304983615875, "learning_rate": 9.671703641357042e-05, "loss": 0.0105, "step": 16020 }, { "grad_norm": 0.12256563454866409, "learning_rate": 9.67111412072715e-05, "loss": 0.009, "step": 16030 }, { "grad_norm": 0.1788829267024994, "learning_rate": 9.670524089271242e-05, "loss": 0.0093, "step": 16040 }, { "grad_norm": 0.2397153526544571, "learning_rate": 9.669933547053842e-05, "loss": 0.0122, "step": 16050 }, { "grad_norm": 0.1809842586517334, "learning_rate": 9.669342494139531e-05, "loss": 0.0139, "step": 16060 }, { "grad_norm": 0.16963396966457367, "learning_rate": 9.668750930592943e-05, "loss": 0.01, "step": 16070 }, { "grad_norm": 0.18331056833267212, "learning_rate": 9.668158856478775e-05, "loss": 0.01, "step": 16080 }, { "grad_norm": 0.17593054473400116, "learning_rate": 9.66756627186177e-05, "loss": 0.0128, "step": 16090 }, { "grad_norm": 0.17198523879051208, "learning_rate": 9.666973176806737e-05, "loss": 0.0071, "step": 16100 }, { "grad_norm": 0.19065654277801514, "learning_rate": 9.666379571378534e-05, "loss": 0.0085, "step": 16110 }, { "grad_norm": 0.1492450088262558, "learning_rate": 9.665785455642076e-05, "loss": 0.0089, "step": 16120 }, { "grad_norm": 0.18017259240150452, "learning_rate": 9.665190829662337e-05, "loss": 0.0082, "step": 16130 }, { "grad_norm": 0.15334448218345642, "learning_rate": 9.664595693504342e-05, "loss": 0.0089, "step": 16140 }, { "grad_norm": 0.15273229777812958, "learning_rate": 9.664000047233175e-05, "loss": 0.0083, "step": 16150 }, { "grad_norm": 0.11871496587991714, "learning_rate": 9.663403890913976e-05, "loss": 0.0089, "step": 16160 }, { "grad_norm": 0.1667008101940155, "learning_rate": 9.662807224611938e-05, "loss": 0.0086, "step": 16170 }, { "grad_norm": 0.17212167382240295, "learning_rate": 9.662210048392311e-05, "loss": 0.0085, "step": 16180 }, { "grad_norm": 0.26685693860054016, "learning_rate": 9.661612362320405e-05, "loss": 0.0134, "step": 16190 }, { "grad_norm": 0.24812282621860504, "learning_rate": 9.661014166461579e-05, "loss": 0.0173, "step": 16200 }, { "grad_norm": 0.22625254094600677, "learning_rate": 9.66041546088125e-05, "loss": 0.0142, "step": 16210 }, { "grad_norm": 0.19663690030574799, "learning_rate": 9.659816245644895e-05, "loss": 0.0122, "step": 16220 }, { "grad_norm": 0.5736169219017029, "learning_rate": 9.65921652081804e-05, "loss": 0.0115, "step": 16230 }, { "grad_norm": 0.23424194753170013, "learning_rate": 9.658616286466271e-05, "loss": 0.0146, "step": 16240 }, { "grad_norm": 0.2092261016368866, "learning_rate": 9.65801554265523e-05, "loss": 0.0152, "step": 16250 }, { "grad_norm": 0.26539143919944763, "learning_rate": 9.657414289450612e-05, "loss": 0.0139, "step": 16260 }, { "grad_norm": 0.21242155134677887, "learning_rate": 9.656812526918171e-05, "loss": 0.0144, "step": 16270 }, { "grad_norm": 0.21389196813106537, "learning_rate": 9.656210255123712e-05, "loss": 0.0143, "step": 16280 }, { "grad_norm": 0.19899475574493408, "learning_rate": 9.6556074741331e-05, "loss": 0.0159, "step": 16290 }, { "grad_norm": 0.2428925484418869, "learning_rate": 9.655004184012256e-05, "loss": 0.0128, "step": 16300 }, { "grad_norm": 0.18066450953483582, "learning_rate": 9.654400384827152e-05, "loss": 0.0117, "step": 16310 }, { "grad_norm": 0.15272437036037445, "learning_rate": 9.653796076643818e-05, "loss": 0.0111, "step": 16320 }, { "grad_norm": 0.25407886505126953, "learning_rate": 9.653191259528344e-05, "loss": 0.0111, "step": 16330 }, { "grad_norm": 0.24773693084716797, "learning_rate": 9.65258593354687e-05, "loss": 0.0116, "step": 16340 }, { "grad_norm": 0.48803600668907166, "learning_rate": 9.651980098765591e-05, "loss": 0.0131, "step": 16350 }, { "grad_norm": 0.3050592243671417, "learning_rate": 9.651373755250765e-05, "loss": 0.0185, "step": 16360 }, { "grad_norm": 0.18773579597473145, "learning_rate": 9.650766903068697e-05, "loss": 0.0157, "step": 16370 }, { "grad_norm": 0.2624327838420868, "learning_rate": 9.650159542285753e-05, "loss": 0.0136, "step": 16380 }, { "grad_norm": 0.21875089406967163, "learning_rate": 9.649551672968353e-05, "loss": 0.0162, "step": 16390 }, { "grad_norm": 0.2800912857055664, "learning_rate": 9.648943295182973e-05, "loss": 0.017, "step": 16400 }, { "grad_norm": 0.2685825228691101, "learning_rate": 9.648334408996144e-05, "loss": 0.0156, "step": 16410 }, { "grad_norm": 0.17702364921569824, "learning_rate": 9.647725014474452e-05, "loss": 0.0136, "step": 16420 }, { "grad_norm": 0.17694883048534393, "learning_rate": 9.64711511168454e-05, "loss": 0.0126, "step": 16430 }, { "grad_norm": 0.20332200825214386, "learning_rate": 9.646504700693108e-05, "loss": 0.0114, "step": 16440 }, { "grad_norm": 0.16952070593833923, "learning_rate": 9.645893781566907e-05, "loss": 0.0133, "step": 16450 }, { "grad_norm": 0.17106367647647858, "learning_rate": 9.645282354372744e-05, "loss": 0.01, "step": 16460 }, { "grad_norm": 0.18795382976531982, "learning_rate": 9.644670419177491e-05, "loss": 0.0096, "step": 16470 }, { "grad_norm": 0.19563862681388855, "learning_rate": 9.644057976048062e-05, "loss": 0.011, "step": 16480 }, { "grad_norm": 0.16053466498851776, "learning_rate": 9.643445025051435e-05, "loss": 0.0099, "step": 16490 }, { "grad_norm": 0.2501298487186432, "learning_rate": 9.642831566254641e-05, "loss": 0.0088, "step": 16500 }, { "grad_norm": 0.16754935681819916, "learning_rate": 9.642217599724769e-05, "loss": 0.0123, "step": 16510 }, { "grad_norm": 0.2287687212228775, "learning_rate": 9.64160312552896e-05, "loss": 0.0103, "step": 16520 }, { "grad_norm": 0.21931912004947662, "learning_rate": 9.64098814373441e-05, "loss": 0.0113, "step": 16530 }, { "grad_norm": 0.20176362991333008, "learning_rate": 9.640372654408374e-05, "loss": 0.0087, "step": 16540 }, { "grad_norm": 0.28718554973602295, "learning_rate": 9.639756657618162e-05, "loss": 0.0097, "step": 16550 }, { "grad_norm": 0.19159115850925446, "learning_rate": 9.639140153431138e-05, "loss": 0.0109, "step": 16560 }, { "grad_norm": 0.2592039108276367, "learning_rate": 9.638523141914721e-05, "loss": 0.014, "step": 16570 }, { "grad_norm": 0.1981893628835678, "learning_rate": 9.637905623136388e-05, "loss": 0.0158, "step": 16580 }, { "grad_norm": 0.23743930459022522, "learning_rate": 9.637287597163669e-05, "loss": 0.0134, "step": 16590 }, { "grad_norm": 0.23605792224407196, "learning_rate": 9.63666906406415e-05, "loss": 0.0149, "step": 16600 }, { "grad_norm": 0.20276199281215668, "learning_rate": 9.636050023905473e-05, "loss": 0.0121, "step": 16610 }, { "grad_norm": 0.17196041345596313, "learning_rate": 9.635430476755336e-05, "loss": 0.0097, "step": 16620 }, { "grad_norm": 0.16449980437755585, "learning_rate": 9.63481042268149e-05, "loss": 0.012, "step": 16630 }, { "grad_norm": 0.17496608197689056, "learning_rate": 9.634189861751745e-05, "loss": 0.0133, "step": 16640 }, { "grad_norm": 0.21638022363185883, "learning_rate": 9.633568794033967e-05, "loss": 0.0097, "step": 16650 }, { "grad_norm": 0.2316838949918747, "learning_rate": 9.63294721959607e-05, "loss": 0.0106, "step": 16660 }, { "grad_norm": 0.15593965351581573, "learning_rate": 9.63232513850603e-05, "loss": 0.0115, "step": 16670 }, { "grad_norm": 0.1617666482925415, "learning_rate": 9.631702550831878e-05, "loss": 0.0124, "step": 16680 }, { "grad_norm": 0.16442184150218964, "learning_rate": 9.631079456641698e-05, "loss": 0.0159, "step": 16690 }, { "grad_norm": 0.2134637087583542, "learning_rate": 9.630455856003632e-05, "loss": 0.0114, "step": 16700 }, { "grad_norm": 0.17280839383602142, "learning_rate": 9.629831748985876e-05, "loss": 0.01, "step": 16710 }, { "grad_norm": 0.12368383258581161, "learning_rate": 9.629207135656679e-05, "loss": 0.0086, "step": 16720 }, { "grad_norm": 0.16987723112106323, "learning_rate": 9.628582016084353e-05, "loss": 0.0076, "step": 16730 }, { "grad_norm": 0.14454065263271332, "learning_rate": 9.627956390337254e-05, "loss": 0.0076, "step": 16740 }, { "grad_norm": 0.2024308145046234, "learning_rate": 9.627330258483802e-05, "loss": 0.0137, "step": 16750 }, { "grad_norm": 0.17499688267707825, "learning_rate": 9.62670362059247e-05, "loss": 0.0136, "step": 16760 }, { "grad_norm": 0.1862587332725525, "learning_rate": 9.626076476731786e-05, "loss": 0.0113, "step": 16770 }, { "grad_norm": 0.1968700885772705, "learning_rate": 9.625448826970336e-05, "loss": 0.0104, "step": 16780 }, { "grad_norm": 0.1459048092365265, "learning_rate": 9.624820671376755e-05, "loss": 0.0098, "step": 16790 }, { "grad_norm": 0.14940321445465088, "learning_rate": 9.62419201001974e-05, "loss": 0.0096, "step": 16800 }, { "grad_norm": 0.1401943862438202, "learning_rate": 9.623562842968037e-05, "loss": 0.0124, "step": 16810 }, { "grad_norm": 0.16804999113082886, "learning_rate": 9.622933170290454e-05, "loss": 0.0126, "step": 16820 }, { "grad_norm": 0.14813165366649628, "learning_rate": 9.622302992055849e-05, "loss": 0.011, "step": 16830 }, { "grad_norm": 0.1533757895231247, "learning_rate": 9.62167230833314e-05, "loss": 0.01, "step": 16840 }, { "grad_norm": 0.16298989951610565, "learning_rate": 9.621041119191295e-05, "loss": 0.0106, "step": 16850 }, { "grad_norm": 0.17990775406360626, "learning_rate": 9.620409424699342e-05, "loss": 0.0085, "step": 16860 }, { "grad_norm": 0.2985980212688446, "learning_rate": 9.619777224926359e-05, "loss": 0.0096, "step": 16870 }, { "grad_norm": 0.2917957901954651, "learning_rate": 9.619144519941485e-05, "loss": 0.011, "step": 16880 }, { "grad_norm": 0.2634165287017822, "learning_rate": 9.618511309813912e-05, "loss": 0.0111, "step": 16890 }, { "grad_norm": 0.14148277044296265, "learning_rate": 9.617877594612886e-05, "loss": 0.0097, "step": 16900 }, { "grad_norm": 0.1280996948480606, "learning_rate": 9.617243374407707e-05, "loss": 0.0093, "step": 16910 }, { "grad_norm": 0.2560521364212036, "learning_rate": 9.616608649267736e-05, "loss": 0.0097, "step": 16920 }, { "grad_norm": 0.1464127004146576, "learning_rate": 9.615973419262385e-05, "loss": 0.0087, "step": 16930 }, { "grad_norm": 0.23799413442611694, "learning_rate": 9.615337684461119e-05, "loss": 0.0096, "step": 16940 }, { "grad_norm": 0.19229570031166077, "learning_rate": 9.614701444933465e-05, "loss": 0.0104, "step": 16950 }, { "grad_norm": 0.20253513753414154, "learning_rate": 9.614064700748997e-05, "loss": 0.0086, "step": 16960 }, { "grad_norm": 0.15116970241069794, "learning_rate": 9.613427451977352e-05, "loss": 0.0074, "step": 16970 }, { "grad_norm": 0.1592608541250229, "learning_rate": 9.612789698688216e-05, "loss": 0.0095, "step": 16980 }, { "grad_norm": 0.1669839322566986, "learning_rate": 9.612151440951334e-05, "loss": 0.0107, "step": 16990 }, { "grad_norm": 0.19536352157592773, "learning_rate": 9.611512678836506e-05, "loss": 0.0101, "step": 17000 }, { "grad_norm": 0.15384313464164734, "learning_rate": 9.610873412413584e-05, "loss": 0.0104, "step": 17010 }, { "grad_norm": 0.18034185469150543, "learning_rate": 9.610233641752476e-05, "loss": 0.0113, "step": 17020 }, { "grad_norm": 0.1770647019147873, "learning_rate": 9.609593366923151e-05, "loss": 0.0105, "step": 17030 }, { "grad_norm": 0.21679508686065674, "learning_rate": 9.608952587995625e-05, "loss": 0.0141, "step": 17040 }, { "grad_norm": 0.1638001948595047, "learning_rate": 9.608311305039972e-05, "loss": 0.0112, "step": 17050 }, { "grad_norm": 0.16642752289772034, "learning_rate": 9.607669518126326e-05, "loss": 0.0119, "step": 17060 }, { "grad_norm": 0.14583338797092438, "learning_rate": 9.607027227324866e-05, "loss": 0.0093, "step": 17070 }, { "grad_norm": 0.19347599148750305, "learning_rate": 9.606384432705837e-05, "loss": 0.0097, "step": 17080 }, { "grad_norm": 0.16839534044265747, "learning_rate": 9.60574113433953e-05, "loss": 0.0087, "step": 17090 }, { "grad_norm": 0.1517496258020401, "learning_rate": 9.6050973322963e-05, "loss": 0.0087, "step": 17100 }, { "grad_norm": 0.15078642964363098, "learning_rate": 9.604453026646547e-05, "loss": 0.0084, "step": 17110 }, { "grad_norm": 0.22053632140159607, "learning_rate": 9.603808217460735e-05, "loss": 0.0081, "step": 17120 }, { "grad_norm": 0.15519987046718597, "learning_rate": 9.603162904809377e-05, "loss": 0.0092, "step": 17130 }, { "grad_norm": 0.15393559634685516, "learning_rate": 9.602517088763045e-05, "loss": 0.0088, "step": 17140 }, { "grad_norm": 0.21439112722873688, "learning_rate": 9.601870769392365e-05, "loss": 0.0078, "step": 17150 }, { "grad_norm": 0.20398585498332977, "learning_rate": 9.601223946768017e-05, "loss": 0.0108, "step": 17160 }, { "grad_norm": 0.14353658258914948, "learning_rate": 9.600576620960734e-05, "loss": 0.0111, "step": 17170 }, { "grad_norm": 0.14752936363220215, "learning_rate": 9.599928792041308e-05, "loss": 0.0132, "step": 17180 }, { "grad_norm": 0.1413697451353073, "learning_rate": 9.599280460080587e-05, "loss": 0.0083, "step": 17190 }, { "grad_norm": 0.1894388198852539, "learning_rate": 9.59863162514947e-05, "loss": 0.0079, "step": 17200 }, { "grad_norm": 0.14881564676761627, "learning_rate": 9.597982287318911e-05, "loss": 0.0088, "step": 17210 }, { "grad_norm": 0.2075960636138916, "learning_rate": 9.597332446659923e-05, "loss": 0.0108, "step": 17220 }, { "grad_norm": 0.22279614210128784, "learning_rate": 9.59668210324357e-05, "loss": 0.0112, "step": 17230 }, { "grad_norm": 0.17212817072868347, "learning_rate": 9.596031257140974e-05, "loss": 0.0082, "step": 17240 }, { "grad_norm": 0.24842101335525513, "learning_rate": 9.59537990842331e-05, "loss": 0.0105, "step": 17250 }, { "grad_norm": 0.23370710015296936, "learning_rate": 9.594728057161806e-05, "loss": 0.0116, "step": 17260 }, { "grad_norm": 0.1365707963705063, "learning_rate": 9.594075703427752e-05, "loss": 0.01, "step": 17270 }, { "grad_norm": 0.22670409083366394, "learning_rate": 9.593422847292486e-05, "loss": 0.0142, "step": 17280 }, { "grad_norm": 0.24661552906036377, "learning_rate": 9.592769488827402e-05, "loss": 0.0126, "step": 17290 }, { "grad_norm": 0.1317395567893982, "learning_rate": 9.592115628103952e-05, "loss": 0.0129, "step": 17300 }, { "grad_norm": 0.33041298389434814, "learning_rate": 9.591461265193643e-05, "loss": 0.0123, "step": 17310 }, { "grad_norm": 0.1798882633447647, "learning_rate": 9.590806400168032e-05, "loss": 0.0095, "step": 17320 }, { "grad_norm": 0.257777601480484, "learning_rate": 9.590151033098735e-05, "loss": 0.0113, "step": 17330 }, { "grad_norm": 0.13827450573444366, "learning_rate": 9.589495164057423e-05, "loss": 0.0091, "step": 17340 }, { "grad_norm": 0.13892072439193726, "learning_rate": 9.58883879311582e-05, "loss": 0.0098, "step": 17350 }, { "grad_norm": 0.1657390594482422, "learning_rate": 9.588181920345705e-05, "loss": 0.0092, "step": 17360 }, { "grad_norm": 0.16257105767726898, "learning_rate": 9.587524545818913e-05, "loss": 0.0108, "step": 17370 }, { "grad_norm": 0.17780357599258423, "learning_rate": 9.586866669607335e-05, "loss": 0.0107, "step": 17380 }, { "grad_norm": 0.27169209718704224, "learning_rate": 9.586208291782915e-05, "loss": 0.01, "step": 17390 }, { "grad_norm": 0.17865349352359772, "learning_rate": 9.58554941241765e-05, "loss": 0.0132, "step": 17400 }, { "grad_norm": 0.24869224429130554, "learning_rate": 9.584890031583596e-05, "loss": 0.0083, "step": 17410 }, { "grad_norm": 0.33955690264701843, "learning_rate": 9.584230149352861e-05, "loss": 0.0115, "step": 17420 }, { "grad_norm": 0.1491425782442093, "learning_rate": 9.58356976579761e-05, "loss": 0.0106, "step": 17430 }, { "grad_norm": 0.1674460470676422, "learning_rate": 9.58290888099006e-05, "loss": 0.0109, "step": 17440 }, { "grad_norm": 0.14201858639717102, "learning_rate": 9.582247495002486e-05, "loss": 0.0101, "step": 17450 }, { "grad_norm": 0.17584773898124695, "learning_rate": 9.581585607907214e-05, "loss": 0.0123, "step": 17460 }, { "grad_norm": 0.13217900693416595, "learning_rate": 9.580923219776628e-05, "loss": 0.0102, "step": 17470 }, { "grad_norm": 0.163141131401062, "learning_rate": 9.580260330683167e-05, "loss": 0.009, "step": 17480 }, { "grad_norm": 0.1476038545370102, "learning_rate": 9.579596940699322e-05, "loss": 0.0103, "step": 17490 }, { "grad_norm": 0.14302615821361542, "learning_rate": 9.578933049897643e-05, "loss": 0.007, "step": 17500 }, { "grad_norm": 0.16085825860500336, "learning_rate": 9.578268658350728e-05, "loss": 0.0088, "step": 17510 }, { "grad_norm": 0.10574034601449966, "learning_rate": 9.577603766131235e-05, "loss": 0.0086, "step": 17520 }, { "grad_norm": 0.20604243874549866, "learning_rate": 9.576938373311878e-05, "loss": 0.0093, "step": 17530 }, { "grad_norm": 0.178636372089386, "learning_rate": 9.576272479965421e-05, "loss": 0.0088, "step": 17540 }, { "grad_norm": 0.15704913437366486, "learning_rate": 9.575606086164687e-05, "loss": 0.0105, "step": 17550 }, { "grad_norm": 0.14132073521614075, "learning_rate": 9.57493919198255e-05, "loss": 0.0099, "step": 17560 }, { "grad_norm": 0.12374085187911987, "learning_rate": 9.57427179749194e-05, "loss": 0.0112, "step": 17570 }, { "grad_norm": 0.16455259919166565, "learning_rate": 9.573603902765846e-05, "loss": 0.0095, "step": 17580 }, { "grad_norm": 0.1317022740840912, "learning_rate": 9.572935507877304e-05, "loss": 0.0101, "step": 17590 }, { "grad_norm": 0.20831423997879028, "learning_rate": 9.57226661289941e-05, "loss": 0.011, "step": 17600 }, { "grad_norm": 0.12206508964300156, "learning_rate": 9.571597217905315e-05, "loss": 0.0077, "step": 17610 }, { "grad_norm": 0.14987856149673462, "learning_rate": 9.57092732296822e-05, "loss": 0.0074, "step": 17620 }, { "grad_norm": 0.15806561708450317, "learning_rate": 9.570256928161385e-05, "loss": 0.0098, "step": 17630 }, { "grad_norm": 0.18527351319789886, "learning_rate": 9.569586033558126e-05, "loss": 0.0087, "step": 17640 }, { "grad_norm": 0.17170308530330658, "learning_rate": 9.568914639231807e-05, "loss": 0.0096, "step": 17650 }, { "grad_norm": 0.17149658501148224, "learning_rate": 9.568242745255852e-05, "loss": 0.0083, "step": 17660 }, { "grad_norm": 0.11726391315460205, "learning_rate": 9.567570351703739e-05, "loss": 0.0073, "step": 17670 }, { "grad_norm": 0.17296268045902252, "learning_rate": 9.566897458649001e-05, "loss": 0.0077, "step": 17680 }, { "grad_norm": 0.13575288653373718, "learning_rate": 9.566224066165221e-05, "loss": 0.0094, "step": 17690 }, { "grad_norm": 0.11392894387245178, "learning_rate": 9.565550174326043e-05, "loss": 0.0073, "step": 17700 }, { "grad_norm": 0.11021541059017181, "learning_rate": 9.564875783205162e-05, "loss": 0.0065, "step": 17710 }, { "grad_norm": 0.12367618083953857, "learning_rate": 9.564200892876328e-05, "loss": 0.0075, "step": 17720 }, { "grad_norm": 0.11123467236757278, "learning_rate": 9.563525503413348e-05, "loss": 0.0076, "step": 17730 }, { "grad_norm": 0.1479250192642212, "learning_rate": 9.562849614890079e-05, "loss": 0.0066, "step": 17740 }, { "grad_norm": 0.15452030301094055, "learning_rate": 9.562173227380436e-05, "loss": 0.0069, "step": 17750 }, { "grad_norm": 0.1314118355512619, "learning_rate": 9.561496340958389e-05, "loss": 0.007, "step": 17760 }, { "grad_norm": 0.1590728759765625, "learning_rate": 9.560818955697959e-05, "loss": 0.0081, "step": 17770 }, { "grad_norm": 0.1958295702934265, "learning_rate": 9.560141071673228e-05, "loss": 0.0085, "step": 17780 }, { "grad_norm": 0.2578970789909363, "learning_rate": 9.559462688958323e-05, "loss": 0.0086, "step": 17790 }, { "grad_norm": 0.16440850496292114, "learning_rate": 9.558783807627434e-05, "loss": 0.0084, "step": 17800 }, { "grad_norm": 0.12647786736488342, "learning_rate": 9.558104427754801e-05, "loss": 0.0082, "step": 17810 }, { "grad_norm": 0.1609116941690445, "learning_rate": 9.557424549414722e-05, "loss": 0.0091, "step": 17820 }, { "grad_norm": 0.17918992042541504, "learning_rate": 9.556744172681546e-05, "loss": 0.008, "step": 17830 }, { "grad_norm": 0.13003180921077728, "learning_rate": 9.556063297629677e-05, "loss": 0.0085, "step": 17840 }, { "grad_norm": 0.13006052374839783, "learning_rate": 9.555381924333578e-05, "loss": 0.0088, "step": 17850 }, { "grad_norm": 0.1795496642589569, "learning_rate": 9.554700052867758e-05, "loss": 0.0093, "step": 17860 }, { "grad_norm": 0.16379544138908386, "learning_rate": 9.554017683306789e-05, "loss": 0.0114, "step": 17870 }, { "grad_norm": 0.16991205513477325, "learning_rate": 9.553334815725294e-05, "loss": 0.0101, "step": 17880 }, { "grad_norm": 0.19036710262298584, "learning_rate": 9.552651450197949e-05, "loss": 0.009, "step": 17890 }, { "grad_norm": 0.15101538598537445, "learning_rate": 9.551967586799486e-05, "loss": 0.0082, "step": 17900 }, { "grad_norm": 0.1732707917690277, "learning_rate": 9.551283225604692e-05, "loss": 0.0087, "step": 17910 }, { "grad_norm": 0.2078806459903717, "learning_rate": 9.550598366688406e-05, "loss": 0.0096, "step": 17920 }, { "grad_norm": 0.3035357892513275, "learning_rate": 9.549913010125526e-05, "loss": 0.0118, "step": 17930 }, { "grad_norm": 0.18802639842033386, "learning_rate": 9.549227155990999e-05, "loss": 0.0103, "step": 17940 }, { "grad_norm": 0.19157712161540985, "learning_rate": 9.548540804359828e-05, "loss": 0.0086, "step": 17950 }, { "grad_norm": 0.1340658813714981, "learning_rate": 9.547853955307077e-05, "loss": 0.0079, "step": 17960 }, { "grad_norm": 0.18062971532344818, "learning_rate": 9.547166608907853e-05, "loss": 0.0109, "step": 17970 }, { "grad_norm": 0.2384687215089798, "learning_rate": 9.546478765237326e-05, "loss": 0.008, "step": 17980 }, { "grad_norm": 0.21984444558620453, "learning_rate": 9.545790424370715e-05, "loss": 0.0095, "step": 17990 }, { "grad_norm": 0.1941738724708557, "learning_rate": 9.5451015863833e-05, "loss": 0.0093, "step": 18000 }, { "grad_norm": 0.1835738867521286, "learning_rate": 9.544412251350408e-05, "loss": 0.01, "step": 18010 }, { "grad_norm": 0.1493169665336609, "learning_rate": 9.543722419347422e-05, "loss": 0.0086, "step": 18020 }, { "grad_norm": 0.125152125954628, "learning_rate": 9.543032090449788e-05, "loss": 0.0085, "step": 18030 }, { "grad_norm": 0.1397864818572998, "learning_rate": 9.542341264732992e-05, "loss": 0.0078, "step": 18040 }, { "grad_norm": 0.12953852117061615, "learning_rate": 9.541649942272585e-05, "loss": 0.0088, "step": 18050 }, { "grad_norm": 0.10652283579111099, "learning_rate": 9.54095812314417e-05, "loss": 0.0067, "step": 18060 }, { "grad_norm": 0.14969661831855774, "learning_rate": 9.540265807423401e-05, "loss": 0.009, "step": 18070 }, { "grad_norm": 0.20193958282470703, "learning_rate": 9.53957299518599e-05, "loss": 0.0081, "step": 18080 }, { "grad_norm": 0.1633274406194687, "learning_rate": 9.5388796865077e-05, "loss": 0.0081, "step": 18090 }, { "grad_norm": 0.2660605311393738, "learning_rate": 9.538185881464353e-05, "loss": 0.0092, "step": 18100 }, { "grad_norm": 0.16471903026103973, "learning_rate": 9.537491580131821e-05, "loss": 0.0088, "step": 18110 }, { "grad_norm": 0.21397298574447632, "learning_rate": 9.53679678258603e-05, "loss": 0.0087, "step": 18120 }, { "grad_norm": 0.09332120418548584, "learning_rate": 9.536101488902966e-05, "loss": 0.0084, "step": 18130 }, { "grad_norm": 0.15281374752521515, "learning_rate": 9.535405699158663e-05, "loss": 0.0104, "step": 18140 }, { "grad_norm": 0.12745507061481476, "learning_rate": 9.53470941342921e-05, "loss": 0.0082, "step": 18150 }, { "grad_norm": 0.14187107980251312, "learning_rate": 9.534012631790756e-05, "loss": 0.0096, "step": 18160 }, { "grad_norm": 0.1463308483362198, "learning_rate": 9.533315354319494e-05, "loss": 0.0118, "step": 18170 }, { "grad_norm": 0.14109846949577332, "learning_rate": 9.532617581091682e-05, "loss": 0.0081, "step": 18180 }, { "grad_norm": 0.13566187024116516, "learning_rate": 9.531919312183629e-05, "loss": 0.0076, "step": 18190 }, { "grad_norm": 0.15872445702552795, "learning_rate": 9.531220547671688e-05, "loss": 0.0072, "step": 18200 }, { "grad_norm": 0.12365014851093292, "learning_rate": 9.530521287632285e-05, "loss": 0.0089, "step": 18210 }, { "grad_norm": 0.18692679703235626, "learning_rate": 9.529821532141884e-05, "loss": 0.0121, "step": 18220 }, { "grad_norm": 0.1285250335931778, "learning_rate": 9.52912128127701e-05, "loss": 0.0122, "step": 18230 }, { "grad_norm": 0.14687907695770264, "learning_rate": 9.528420535114244e-05, "loss": 0.009, "step": 18240 }, { "grad_norm": 0.14079293608665466, "learning_rate": 9.527719293730215e-05, "loss": 0.0086, "step": 18250 }, { "grad_norm": 0.17353607714176178, "learning_rate": 9.527017557201611e-05, "loss": 0.0076, "step": 18260 }, { "grad_norm": 0.20594662427902222, "learning_rate": 9.526315325605176e-05, "loss": 0.0077, "step": 18270 }, { "grad_norm": 0.14725619554519653, "learning_rate": 9.525612599017699e-05, "loss": 0.009, "step": 18280 }, { "grad_norm": 0.1275157630443573, "learning_rate": 9.524909377516033e-05, "loss": 0.0077, "step": 18290 }, { "grad_norm": 0.1249048113822937, "learning_rate": 9.524205661177081e-05, "loss": 0.0087, "step": 18300 }, { "grad_norm": 0.16863873600959778, "learning_rate": 9.523501450077801e-05, "loss": 0.0111, "step": 18310 }, { "grad_norm": 0.16979333758354187, "learning_rate": 9.522796744295202e-05, "loss": 0.0089, "step": 18320 }, { "grad_norm": 0.18821464478969574, "learning_rate": 9.522091543906352e-05, "loss": 0.0099, "step": 18330 }, { "grad_norm": 0.21621674299240112, "learning_rate": 9.521385848988369e-05, "loss": 0.0094, "step": 18340 }, { "grad_norm": 0.18244220316410065, "learning_rate": 9.520679659618428e-05, "loss": 0.008, "step": 18350 }, { "grad_norm": 0.15632568299770355, "learning_rate": 9.519972975873754e-05, "loss": 0.0076, "step": 18360 }, { "grad_norm": 0.14299294352531433, "learning_rate": 9.519265797831633e-05, "loss": 0.0077, "step": 18370 }, { "grad_norm": 0.13638924062252045, "learning_rate": 9.518558125569399e-05, "loss": 0.0082, "step": 18380 }, { "grad_norm": 0.11717309802770615, "learning_rate": 9.517849959164442e-05, "loss": 0.0071, "step": 18390 }, { "grad_norm": 0.1684451848268509, "learning_rate": 9.517141298694205e-05, "loss": 0.007, "step": 18400 }, { "grad_norm": 0.1462511569261551, "learning_rate": 9.516432144236188e-05, "loss": 0.0102, "step": 18410 }, { "grad_norm": 0.1464066505432129, "learning_rate": 9.515722495867941e-05, "loss": 0.0113, "step": 18420 }, { "grad_norm": 0.23103134334087372, "learning_rate": 9.515012353667072e-05, "loss": 0.0133, "step": 18430 }, { "grad_norm": 0.18159310519695282, "learning_rate": 9.51430171771124e-05, "loss": 0.0095, "step": 18440 }, { "grad_norm": 0.16696731746196747, "learning_rate": 9.513590588078159e-05, "loss": 0.0104, "step": 18450 }, { "grad_norm": 0.3938882648944855, "learning_rate": 9.512878964845597e-05, "loss": 0.0104, "step": 18460 }, { "grad_norm": 0.178971529006958, "learning_rate": 9.512166848091377e-05, "loss": 0.0143, "step": 18470 }, { "grad_norm": 0.241012305021286, "learning_rate": 9.511454237893376e-05, "loss": 0.0153, "step": 18480 }, { "grad_norm": 0.24551965296268463, "learning_rate": 9.51074113432952e-05, "loss": 0.0143, "step": 18490 }, { "grad_norm": 0.23073551058769226, "learning_rate": 9.510027537477797e-05, "loss": 0.0124, "step": 18500 }, { "grad_norm": 0.17409290373325348, "learning_rate": 9.509313447416242e-05, "loss": 0.0104, "step": 18510 }, { "grad_norm": 0.1789768487215042, "learning_rate": 9.508598864222949e-05, "loss": 0.0099, "step": 18520 }, { "grad_norm": 0.15040668845176697, "learning_rate": 9.507883787976062e-05, "loss": 0.0072, "step": 18530 }, { "grad_norm": 0.139931783080101, "learning_rate": 9.507168218753781e-05, "loss": 0.0095, "step": 18540 }, { "grad_norm": 0.15787965059280396, "learning_rate": 9.506452156634362e-05, "loss": 0.0083, "step": 18550 }, { "grad_norm": 0.11084190011024475, "learning_rate": 9.505735601696109e-05, "loss": 0.0092, "step": 18560 }, { "grad_norm": 0.16849201917648315, "learning_rate": 9.505018554017385e-05, "loss": 0.011, "step": 18570 }, { "grad_norm": 0.14618054032325745, "learning_rate": 9.504301013676604e-05, "loss": 0.008, "step": 18580 }, { "grad_norm": 0.21751950681209564, "learning_rate": 9.503582980752238e-05, "loss": 0.009, "step": 18590 }, { "grad_norm": 0.15989916026592255, "learning_rate": 9.502864455322809e-05, "loss": 0.009, "step": 18600 }, { "grad_norm": 0.1202705055475235, "learning_rate": 9.502145437466891e-05, "loss": 0.0088, "step": 18610 }, { "grad_norm": 0.10629197210073471, "learning_rate": 9.501425927263116e-05, "loss": 0.0075, "step": 18620 }, { "grad_norm": 0.2168872058391571, "learning_rate": 9.500705924790172e-05, "loss": 0.0085, "step": 18630 }, { "grad_norm": 0.15036951005458832, "learning_rate": 9.499985430126794e-05, "loss": 0.009, "step": 18640 }, { "grad_norm": 0.1362268328666687, "learning_rate": 9.499264443351775e-05, "loss": 0.0072, "step": 18650 }, { "grad_norm": 0.14095982909202576, "learning_rate": 9.498542964543961e-05, "loss": 0.0072, "step": 18660 }, { "grad_norm": 0.12493220716714859, "learning_rate": 9.497820993782252e-05, "loss": 0.0064, "step": 18670 }, { "grad_norm": 0.16130301356315613, "learning_rate": 9.497098531145601e-05, "loss": 0.0097, "step": 18680 }, { "grad_norm": 0.15407758951187134, "learning_rate": 9.496375576713017e-05, "loss": 0.0087, "step": 18690 }, { "grad_norm": 0.3265714645385742, "learning_rate": 9.49565213056356e-05, "loss": 0.009, "step": 18700 }, { "grad_norm": 0.19572792947292328, "learning_rate": 9.494928192776342e-05, "loss": 0.0084, "step": 18710 }, { "grad_norm": 0.16312098503112793, "learning_rate": 9.494203763430538e-05, "loss": 0.009, "step": 18720 }, { "grad_norm": 0.1582420915365219, "learning_rate": 9.493478842605366e-05, "loss": 0.0072, "step": 18730 }, { "grad_norm": 0.13823872804641724, "learning_rate": 9.492753430380105e-05, "loss": 0.0098, "step": 18740 }, { "grad_norm": 0.1656852662563324, "learning_rate": 9.492027526834083e-05, "loss": 0.0081, "step": 18750 }, { "grad_norm": 0.14142537117004395, "learning_rate": 9.491301132046684e-05, "loss": 0.0073, "step": 18760 }, { "grad_norm": 0.13864970207214355, "learning_rate": 9.490574246097345e-05, "loss": 0.0072, "step": 18770 }, { "grad_norm": 0.16248250007629395, "learning_rate": 9.48984686906556e-05, "loss": 0.0082, "step": 18780 }, { "grad_norm": 0.10756318271160126, "learning_rate": 9.489119001030871e-05, "loss": 0.0075, "step": 18790 }, { "grad_norm": 0.11824604868888855, "learning_rate": 9.488390642072878e-05, "loss": 0.0081, "step": 18800 }, { "grad_norm": 0.14891093969345093, "learning_rate": 9.48766179227123e-05, "loss": 0.008, "step": 18810 }, { "grad_norm": 0.20202665030956268, "learning_rate": 9.486932451705636e-05, "loss": 0.0086, "step": 18820 }, { "grad_norm": 0.21673160791397095, "learning_rate": 9.486202620455857e-05, "loss": 0.0097, "step": 18830 }, { "grad_norm": 0.15868914127349854, "learning_rate": 9.485472298601704e-05, "loss": 0.0097, "step": 18840 }, { "grad_norm": 0.19421708583831787, "learning_rate": 9.484741486223043e-05, "loss": 0.0103, "step": 18850 }, { "grad_norm": 0.12559208273887634, "learning_rate": 9.484010183399797e-05, "loss": 0.0089, "step": 18860 }, { "grad_norm": 0.1225619837641716, "learning_rate": 9.483278390211938e-05, "loss": 0.0069, "step": 18870 }, { "grad_norm": 0.15294720232486725, "learning_rate": 9.482546106739496e-05, "loss": 0.0071, "step": 18880 }, { "grad_norm": 0.1377153992652893, "learning_rate": 9.48181333306255e-05, "loss": 0.008, "step": 18890 }, { "grad_norm": 0.1637921780347824, "learning_rate": 9.481080069261237e-05, "loss": 0.0092, "step": 18900 }, { "grad_norm": 0.12888476252555847, "learning_rate": 9.480346315415745e-05, "loss": 0.0093, "step": 18910 }, { "grad_norm": 0.19242680072784424, "learning_rate": 9.479612071606314e-05, "loss": 0.0106, "step": 18920 }, { "grad_norm": 0.12486868351697922, "learning_rate": 9.478877337913244e-05, "loss": 0.0094, "step": 18930 }, { "grad_norm": 0.15576805174350739, "learning_rate": 9.478142114416881e-05, "loss": 0.0083, "step": 18940 }, { "grad_norm": 0.15928269922733307, "learning_rate": 9.47740640119763e-05, "loss": 0.0085, "step": 18950 }, { "grad_norm": 0.151119664311409, "learning_rate": 9.476670198335947e-05, "loss": 0.0072, "step": 18960 }, { "grad_norm": 0.10758601874113083, "learning_rate": 9.47593350591234e-05, "loss": 0.0095, "step": 18970 }, { "grad_norm": 0.151702880859375, "learning_rate": 9.475196324007376e-05, "loss": 0.0081, "step": 18980 }, { "grad_norm": 0.14429375529289246, "learning_rate": 9.474458652701669e-05, "loss": 0.0099, "step": 18990 }, { "grad_norm": 0.1425401121377945, "learning_rate": 9.473720492075892e-05, "loss": 0.0083, "step": 19000 }, { "grad_norm": 0.16309593617916107, "learning_rate": 9.472981842210768e-05, "loss": 0.0067, "step": 19010 }, { "grad_norm": 0.2115437090396881, "learning_rate": 9.472242703187074e-05, "loss": 0.0074, "step": 19020 }, { "grad_norm": 0.09776989370584488, "learning_rate": 9.471503075085643e-05, "loss": 0.0082, "step": 19030 }, { "grad_norm": 0.22723901271820068, "learning_rate": 9.470762957987359e-05, "loss": 0.0098, "step": 19040 }, { "grad_norm": 0.1280614137649536, "learning_rate": 9.470022351973158e-05, "loss": 0.0093, "step": 19050 }, { "grad_norm": 0.15426650643348694, "learning_rate": 9.469281257124034e-05, "loss": 0.0098, "step": 19060 }, { "grad_norm": 0.14393453299999237, "learning_rate": 9.46853967352103e-05, "loss": 0.0085, "step": 19070 }, { "grad_norm": 0.17398633062839508, "learning_rate": 9.467797601245246e-05, "loss": 0.0089, "step": 19080 }, { "grad_norm": 0.15696431696414948, "learning_rate": 9.467055040377834e-05, "loss": 0.0082, "step": 19090 }, { "grad_norm": 0.22391144931316376, "learning_rate": 9.466311990999999e-05, "loss": 0.0103, "step": 19100 }, { "grad_norm": 0.19803529977798462, "learning_rate": 9.465568453193e-05, "loss": 0.012, "step": 19110 }, { "grad_norm": 0.17957258224487305, "learning_rate": 9.464824427038148e-05, "loss": 0.0128, "step": 19120 }, { "grad_norm": 0.2430708110332489, "learning_rate": 9.46407991261681e-05, "loss": 0.0115, "step": 19130 }, { "grad_norm": 0.1365572065114975, "learning_rate": 9.463334910010404e-05, "loss": 0.012, "step": 19140 }, { "grad_norm": 0.14784161746501923, "learning_rate": 9.462589419300403e-05, "loss": 0.0098, "step": 19150 }, { "grad_norm": 0.1420236974954605, "learning_rate": 9.461843440568333e-05, "loss": 0.0087, "step": 19160 }, { "grad_norm": 0.17485903203487396, "learning_rate": 9.461096973895773e-05, "loss": 0.0075, "step": 19170 }, { "grad_norm": 0.23026740550994873, "learning_rate": 9.460350019364355e-05, "loss": 0.0079, "step": 19180 }, { "grad_norm": 0.2180488258600235, "learning_rate": 9.459602577055764e-05, "loss": 0.0109, "step": 19190 }, { "grad_norm": 0.1664429008960724, "learning_rate": 9.45885464705174e-05, "loss": 0.0106, "step": 19200 }, { "grad_norm": 0.13623405992984772, "learning_rate": 9.458106229434076e-05, "loss": 0.0091, "step": 19210 }, { "grad_norm": 0.16534513235092163, "learning_rate": 9.457357324284617e-05, "loss": 0.0085, "step": 19220 }, { "grad_norm": 0.15283438563346863, "learning_rate": 9.456607931685262e-05, "loss": 0.0079, "step": 19230 }, { "grad_norm": 0.30464157462120056, "learning_rate": 9.455858051717965e-05, "loss": 0.011, "step": 19240 }, { "grad_norm": 0.16737325489521027, "learning_rate": 9.45510768446473e-05, "loss": 0.0117, "step": 19250 }, { "grad_norm": 0.2104412317276001, "learning_rate": 9.454356830007618e-05, "loss": 0.0094, "step": 19260 }, { "grad_norm": 0.19013236463069916, "learning_rate": 9.45360548842874e-05, "loss": 0.0079, "step": 19270 }, { "grad_norm": 0.2154286950826645, "learning_rate": 9.452853659810261e-05, "loss": 0.0116, "step": 19280 }, { "grad_norm": 0.18269868195056915, "learning_rate": 9.452101344234401e-05, "loss": 0.009, "step": 19290 }, { "grad_norm": 0.1286584585905075, "learning_rate": 9.451348541783431e-05, "loss": 0.0093, "step": 19300 }, { "grad_norm": 0.2600973844528198, "learning_rate": 9.450595252539678e-05, "loss": 0.0089, "step": 19310 }, { "grad_norm": 0.2014167308807373, "learning_rate": 9.449841476585518e-05, "loss": 0.0108, "step": 19320 }, { "grad_norm": 0.17814378440380096, "learning_rate": 9.449087214003384e-05, "loss": 0.0081, "step": 19330 }, { "grad_norm": 0.13995672762393951, "learning_rate": 9.448332464875765e-05, "loss": 0.0075, "step": 19340 }, { "grad_norm": 0.13370555639266968, "learning_rate": 9.447577229285192e-05, "loss": 0.0074, "step": 19350 }, { "grad_norm": 0.12371748685836792, "learning_rate": 9.446821507314261e-05, "loss": 0.0063, "step": 19360 }, { "grad_norm": 0.1828944832086563, "learning_rate": 9.446065299045617e-05, "loss": 0.0106, "step": 19370 }, { "grad_norm": 0.16090676188468933, "learning_rate": 9.445308604561955e-05, "loss": 0.0082, "step": 19380 }, { "grad_norm": 0.14015983045101166, "learning_rate": 9.444551423946028e-05, "loss": 0.0094, "step": 19390 }, { "grad_norm": 0.11189118772745132, "learning_rate": 9.443793757280638e-05, "loss": 0.0082, "step": 19400 }, { "grad_norm": 0.2827378511428833, "learning_rate": 9.443035604648646e-05, "loss": 0.01, "step": 19410 }, { "grad_norm": 0.14686864614486694, "learning_rate": 9.44227696613296e-05, "loss": 0.0089, "step": 19420 }, { "grad_norm": 0.16429835557937622, "learning_rate": 9.441517841816542e-05, "loss": 0.0089, "step": 19430 }, { "grad_norm": 0.17136968672275543, "learning_rate": 9.440758231782413e-05, "loss": 0.0119, "step": 19440 }, { "grad_norm": 0.16574357450008392, "learning_rate": 9.439998136113639e-05, "loss": 0.0107, "step": 19450 }, { "grad_norm": 0.14820808172225952, "learning_rate": 9.439237554893344e-05, "loss": 0.0112, "step": 19460 }, { "grad_norm": 0.1672959178686142, "learning_rate": 9.438476488204705e-05, "loss": 0.0112, "step": 19470 }, { "grad_norm": 0.1572006195783615, "learning_rate": 9.43771493613095e-05, "loss": 0.0097, "step": 19480 }, { "grad_norm": 0.14987751841545105, "learning_rate": 9.436952898755362e-05, "loss": 0.0085, "step": 19490 }, { "grad_norm": 0.16698646545410156, "learning_rate": 9.436190376161276e-05, "loss": 0.0076, "step": 19500 }, { "grad_norm": 0.1540224701166153, "learning_rate": 9.43542736843208e-05, "loss": 0.011, "step": 19510 }, { "grad_norm": 0.1471915990114212, "learning_rate": 9.434663875651216e-05, "loss": 0.0082, "step": 19520 }, { "grad_norm": 0.1873619109392166, "learning_rate": 9.433899897902177e-05, "loss": 0.0085, "step": 19530 }, { "grad_norm": 0.12288257479667664, "learning_rate": 9.433135435268511e-05, "loss": 0.0068, "step": 19540 }, { "grad_norm": 0.13536229729652405, "learning_rate": 9.432370487833819e-05, "loss": 0.0087, "step": 19550 }, { "grad_norm": 0.12242397665977478, "learning_rate": 9.431605055681756e-05, "loss": 0.0101, "step": 19560 }, { "grad_norm": 0.13212111592292786, "learning_rate": 9.430839138896026e-05, "loss": 0.0075, "step": 19570 }, { "grad_norm": 0.29955580830574036, "learning_rate": 9.43007273756039e-05, "loss": 0.0076, "step": 19580 }, { "grad_norm": 0.10396184772253036, "learning_rate": 9.429305851758658e-05, "loss": 0.006, "step": 19590 }, { "grad_norm": 0.2923489809036255, "learning_rate": 9.428538481574699e-05, "loss": 0.0098, "step": 19600 }, { "grad_norm": 0.28029799461364746, "learning_rate": 9.42777062709243e-05, "loss": 0.0107, "step": 19610 }, { "grad_norm": 0.23792517185211182, "learning_rate": 9.427002288395821e-05, "loss": 0.0104, "step": 19620 }, { "grad_norm": 0.1788988560438156, "learning_rate": 9.426233465568898e-05, "loss": 0.0125, "step": 19630 }, { "grad_norm": 0.147507444024086, "learning_rate": 9.42546415869574e-05, "loss": 0.0091, "step": 19640 }, { "grad_norm": 0.22117577493190765, "learning_rate": 9.424694367860473e-05, "loss": 0.0096, "step": 19650 }, { "grad_norm": 0.1462218016386032, "learning_rate": 9.423924093147284e-05, "loss": 0.0096, "step": 19660 }, { "grad_norm": 0.3590726852416992, "learning_rate": 9.423153334640407e-05, "loss": 0.0094, "step": 19670 }, { "grad_norm": 0.1968192756175995, "learning_rate": 9.42238209242413e-05, "loss": 0.0104, "step": 19680 }, { "grad_norm": 0.19349552690982819, "learning_rate": 9.421610366582798e-05, "loss": 0.0093, "step": 19690 }, { "grad_norm": 0.1428815871477127, "learning_rate": 9.420838157200803e-05, "loss": 0.0086, "step": 19700 }, { "grad_norm": 0.22051990032196045, "learning_rate": 9.420065464362594e-05, "loss": 0.0084, "step": 19710 }, { "grad_norm": 0.16272298991680145, "learning_rate": 9.419292288152673e-05, "loss": 0.011, "step": 19720 }, { "grad_norm": 0.20245838165283203, "learning_rate": 9.418518628655588e-05, "loss": 0.0146, "step": 19730 }, { "grad_norm": 0.18412871658802032, "learning_rate": 9.417744485955951e-05, "loss": 0.0123, "step": 19740 }, { "grad_norm": 0.1510406732559204, "learning_rate": 9.41696986013842e-05, "loss": 0.013, "step": 19750 }, { "grad_norm": 0.16647212207317352, "learning_rate": 9.416194751287705e-05, "loss": 0.0138, "step": 19760 }, { "grad_norm": 0.22448350489139557, "learning_rate": 9.415419159488572e-05, "loss": 0.0109, "step": 19770 }, { "grad_norm": 0.14303189516067505, "learning_rate": 9.414643084825837e-05, "loss": 0.0088, "step": 19780 }, { "grad_norm": 0.24544258415699005, "learning_rate": 9.413866527384372e-05, "loss": 0.0099, "step": 19790 }, { "grad_norm": 0.1991521418094635, "learning_rate": 9.4130894872491e-05, "loss": 0.0106, "step": 19800 }, { "grad_norm": 0.1581270694732666, "learning_rate": 9.412311964504998e-05, "loss": 0.0093, "step": 19810 }, { "grad_norm": 0.11709524691104889, "learning_rate": 9.411533959237091e-05, "loss": 0.0087, "step": 19820 }, { "grad_norm": 0.14036880433559418, "learning_rate": 9.410755471530464e-05, "loss": 0.0098, "step": 19830 }, { "grad_norm": 0.15294605493545532, "learning_rate": 9.40997650147025e-05, "loss": 0.0089, "step": 19840 }, { "grad_norm": 0.1806178241968155, "learning_rate": 9.409197049141637e-05, "loss": 0.0081, "step": 19850 }, { "grad_norm": 0.1763848215341568, "learning_rate": 9.408417114629863e-05, "loss": 0.0094, "step": 19860 }, { "grad_norm": 0.14679518342018127, "learning_rate": 9.40763669802022e-05, "loss": 0.0088, "step": 19870 }, { "grad_norm": 0.16121028363704681, "learning_rate": 9.406855799398056e-05, "loss": 0.0089, "step": 19880 }, { "grad_norm": 0.1927756816148758, "learning_rate": 9.406074418848767e-05, "loss": 0.0077, "step": 19890 }, { "grad_norm": 0.13082270324230194, "learning_rate": 9.405292556457805e-05, "loss": 0.0086, "step": 19900 }, { "grad_norm": 0.12007151544094086, "learning_rate": 9.404510212310671e-05, "loss": 0.0098, "step": 19910 }, { "grad_norm": 0.21617354452610016, "learning_rate": 9.403727386492924e-05, "loss": 0.0081, "step": 19920 }, { "grad_norm": 0.2118595391511917, "learning_rate": 9.40294407909017e-05, "loss": 0.0118, "step": 19930 }, { "grad_norm": 0.203443244099617, "learning_rate": 9.40216029018807e-05, "loss": 0.011, "step": 19940 }, { "grad_norm": 0.18622919917106628, "learning_rate": 9.401376019872338e-05, "loss": 0.0113, "step": 19950 }, { "grad_norm": 0.17423515021800995, "learning_rate": 9.400591268228746e-05, "loss": 0.0134, "step": 19960 }, { "grad_norm": 0.14208102226257324, "learning_rate": 9.399806035343106e-05, "loss": 0.0098, "step": 19970 }, { "grad_norm": 0.14041723310947418, "learning_rate": 9.399020321301294e-05, "loss": 0.01, "step": 19980 }, { "grad_norm": 0.15417514741420746, "learning_rate": 9.398234126189234e-05, "loss": 0.0095, "step": 19990 }, { "grad_norm": 0.3278934359550476, "learning_rate": 9.397447450092902e-05, "loss": 0.0089, "step": 20000 }, { "grad_norm": 0.1006150096654892, "learning_rate": 9.39666029309833e-05, "loss": 0.0073, "step": 20010 }, { "grad_norm": 0.12803401052951813, "learning_rate": 9.395872655291596e-05, "loss": 0.0076, "step": 20020 }, { "grad_norm": 0.11609634757041931, "learning_rate": 9.395084536758838e-05, "loss": 0.0079, "step": 20030 }, { "grad_norm": 0.1636081039905548, "learning_rate": 9.394295937586243e-05, "loss": 0.0085, "step": 20040 }, { "grad_norm": 0.1442139446735382, "learning_rate": 9.393506857860052e-05, "loss": 0.011, "step": 20050 }, { "grad_norm": 0.13053743541240692, "learning_rate": 9.392717297666555e-05, "loss": 0.0093, "step": 20060 }, { "grad_norm": 0.12872633337974548, "learning_rate": 9.391927257092101e-05, "loss": 0.0083, "step": 20070 }, { "grad_norm": 0.16463951766490936, "learning_rate": 9.391136736223085e-05, "loss": 0.0072, "step": 20080 }, { "grad_norm": 0.14344199001789093, "learning_rate": 9.390345735145956e-05, "loss": 0.0077, "step": 20090 }, { "grad_norm": 0.14074358344078064, "learning_rate": 9.389554253947219e-05, "loss": 0.0065, "step": 20100 }, { "grad_norm": 0.17594045400619507, "learning_rate": 9.388762292713428e-05, "loss": 0.0075, "step": 20110 }, { "grad_norm": 0.18368422985076904, "learning_rate": 9.38796985153119e-05, "loss": 0.0067, "step": 20120 }, { "grad_norm": 0.15992063283920288, "learning_rate": 9.387176930487169e-05, "loss": 0.0092, "step": 20130 }, { "grad_norm": 0.13725236058235168, "learning_rate": 9.386383529668072e-05, "loss": 0.0094, "step": 20140 }, { "grad_norm": 0.12625358998775482, "learning_rate": 9.385589649160669e-05, "loss": 0.0101, "step": 20150 }, { "grad_norm": 0.16064728796482086, "learning_rate": 9.384795289051775e-05, "loss": 0.0096, "step": 20160 }, { "grad_norm": 0.13799786567687988, "learning_rate": 9.384000449428261e-05, "loss": 0.0095, "step": 20170 }, { "grad_norm": 0.1863483041524887, "learning_rate": 9.383205130377048e-05, "loss": 0.0104, "step": 20180 }, { "grad_norm": 0.15018479526042938, "learning_rate": 9.382409331985114e-05, "loss": 0.0092, "step": 20190 }, { "grad_norm": 0.1608133167028427, "learning_rate": 9.381613054339482e-05, "loss": 0.0096, "step": 20200 }, { "grad_norm": 0.1323542445898056, "learning_rate": 9.380816297527235e-05, "loss": 0.0099, "step": 20210 }, { "grad_norm": 0.2139693647623062, "learning_rate": 9.380019061635506e-05, "loss": 0.0096, "step": 20220 }, { "grad_norm": 0.24222834408283234, "learning_rate": 9.379221346751474e-05, "loss": 0.0091, "step": 20230 }, { "grad_norm": 0.16264912486076355, "learning_rate": 9.378423152962382e-05, "loss": 0.0095, "step": 20240 }, { "grad_norm": 0.1310873180627823, "learning_rate": 9.377624480355517e-05, "loss": 0.0068, "step": 20250 }, { "grad_norm": 0.12553223967552185, "learning_rate": 9.376825329018219e-05, "loss": 0.0073, "step": 20260 }, { "grad_norm": 0.16043943166732788, "learning_rate": 9.376025699037884e-05, "loss": 0.0069, "step": 20270 }, { "grad_norm": 0.13015592098236084, "learning_rate": 9.37522559050196e-05, "loss": 0.0073, "step": 20280 }, { "grad_norm": 0.25506770610809326, "learning_rate": 9.37442500349794e-05, "loss": 0.0131, "step": 20290 }, { "grad_norm": 0.23289872705936432, "learning_rate": 9.373623938113381e-05, "loss": 0.0172, "step": 20300 }, { "grad_norm": 0.1910189390182495, "learning_rate": 9.372822394435883e-05, "loss": 0.0137, "step": 20310 }, { "grad_norm": 0.15114596486091614, "learning_rate": 9.372020372553102e-05, "loss": 0.0093, "step": 20320 }, { "grad_norm": 0.19173309206962585, "learning_rate": 9.371217872552746e-05, "loss": 0.0095, "step": 20330 }, { "grad_norm": 0.1861177533864975, "learning_rate": 9.370414894522576e-05, "loss": 0.014, "step": 20340 }, { "grad_norm": 0.1811930537223816, "learning_rate": 9.369611438550406e-05, "loss": 0.0108, "step": 20350 }, { "grad_norm": 0.25647860765457153, "learning_rate": 9.368807504724095e-05, "loss": 0.0119, "step": 20360 }, { "grad_norm": 0.18495917320251465, "learning_rate": 9.368003093131565e-05, "loss": 0.0109, "step": 20370 }, { "grad_norm": 0.1984940618276596, "learning_rate": 9.367198203860785e-05, "loss": 0.01, "step": 20380 }, { "grad_norm": 0.13401807844638824, "learning_rate": 9.366392836999774e-05, "loss": 0.008, "step": 20390 }, { "grad_norm": 0.156942218542099, "learning_rate": 9.365586992636607e-05, "loss": 0.009, "step": 20400 }, { "grad_norm": 0.12201724946498871, "learning_rate": 9.364780670859412e-05, "loss": 0.0081, "step": 20410 }, { "grad_norm": 0.15217280387878418, "learning_rate": 9.363973871756364e-05, "loss": 0.0078, "step": 20420 }, { "grad_norm": 0.12205091863870621, "learning_rate": 9.363166595415696e-05, "loss": 0.0076, "step": 20430 }, { "grad_norm": 0.14054909348487854, "learning_rate": 9.362358841925686e-05, "loss": 0.0092, "step": 20440 }, { "grad_norm": 0.2961197793483734, "learning_rate": 9.361550611374674e-05, "loss": 0.0117, "step": 20450 }, { "grad_norm": 0.1636262685060501, "learning_rate": 9.360741903851043e-05, "loss": 0.0112, "step": 20460 }, { "grad_norm": 0.14567799866199493, "learning_rate": 9.359932719443236e-05, "loss": 0.0099, "step": 20470 }, { "grad_norm": 0.17393645644187927, "learning_rate": 9.35912305823974e-05, "loss": 0.0113, "step": 20480 }, { "grad_norm": 0.25145235657691956, "learning_rate": 9.358312920329101e-05, "loss": 0.0158, "step": 20490 }, { "grad_norm": 0.44268694519996643, "learning_rate": 9.357502305799914e-05, "loss": 0.0191, "step": 20500 }, { "grad_norm": 0.2533116638660431, "learning_rate": 9.356691214740824e-05, "loss": 0.0175, "step": 20510 }, { "grad_norm": 0.1439150720834732, "learning_rate": 9.355879647240535e-05, "loss": 0.0138, "step": 20520 }, { "grad_norm": 0.2358372062444687, "learning_rate": 9.355067603387798e-05, "loss": 0.0139, "step": 20530 }, { "grad_norm": 0.16579988598823547, "learning_rate": 9.354255083271412e-05, "loss": 0.0124, "step": 20540 }, { "grad_norm": 0.18373417854309082, "learning_rate": 9.353442086980239e-05, "loss": 0.0093, "step": 20550 }, { "grad_norm": 0.18004879355430603, "learning_rate": 9.352628614603185e-05, "loss": 0.0099, "step": 20560 }, { "grad_norm": 0.17558053135871887, "learning_rate": 9.351814666229209e-05, "loss": 0.0102, "step": 20570 }, { "grad_norm": 0.20130421221256256, "learning_rate": 9.351000241947324e-05, "loss": 0.0133, "step": 20580 }, { "grad_norm": 0.2690895199775696, "learning_rate": 9.350185341846594e-05, "loss": 0.0222, "step": 20590 }, { "grad_norm": 0.21680694818496704, "learning_rate": 9.349369966016134e-05, "loss": 0.0166, "step": 20600 }, { "grad_norm": 0.4513136148452759, "learning_rate": 9.348554114545117e-05, "loss": 0.014, "step": 20610 }, { "grad_norm": 0.18501362204551697, "learning_rate": 9.347737787522758e-05, "loss": 0.0123, "step": 20620 }, { "grad_norm": 0.16165025532245636, "learning_rate": 9.346920985038332e-05, "loss": 0.0126, "step": 20630 }, { "grad_norm": 0.14125286042690277, "learning_rate": 9.346103707181162e-05, "loss": 0.0101, "step": 20640 }, { "grad_norm": 0.13797469437122345, "learning_rate": 9.345285954040626e-05, "loss": 0.0088, "step": 20650 }, { "grad_norm": 0.1481485217809677, "learning_rate": 9.34446772570615e-05, "loss": 0.0086, "step": 20660 }, { "grad_norm": 0.11901234090328217, "learning_rate": 9.343649022267214e-05, "loss": 0.0085, "step": 20670 }, { "grad_norm": 0.11742707341909409, "learning_rate": 9.342829843813353e-05, "loss": 0.0084, "step": 20680 }, { "grad_norm": 0.13436876237392426, "learning_rate": 9.342010190434149e-05, "loss": 0.0083, "step": 20690 }, { "grad_norm": 0.13776108622550964, "learning_rate": 9.34119006221924e-05, "loss": 0.008, "step": 20700 }, { "grad_norm": 0.16159994900226593, "learning_rate": 9.340369459258313e-05, "loss": 0.0095, "step": 20710 }, { "grad_norm": 0.15883122384548187, "learning_rate": 9.339548381641106e-05, "loss": 0.0114, "step": 20720 }, { "grad_norm": 0.10620509088039398, "learning_rate": 9.338726829457413e-05, "loss": 0.0086, "step": 20730 }, { "grad_norm": 0.1408083438873291, "learning_rate": 9.337904802797078e-05, "loss": 0.0086, "step": 20740 }, { "grad_norm": 0.13438594341278076, "learning_rate": 9.337082301749993e-05, "loss": 0.0079, "step": 20750 }, { "grad_norm": 0.11020972579717636, "learning_rate": 9.336259326406109e-05, "loss": 0.0077, "step": 20760 }, { "grad_norm": 0.17406727373600006, "learning_rate": 9.335435876855427e-05, "loss": 0.0082, "step": 20770 }, { "grad_norm": 0.14679847657680511, "learning_rate": 9.334611953187994e-05, "loss": 0.0082, "step": 20780 }, { "grad_norm": 0.12854838371276855, "learning_rate": 9.333787555493914e-05, "loss": 0.0071, "step": 20790 }, { "grad_norm": 0.10210227221250534, "learning_rate": 9.332962683863345e-05, "loss": 0.0068, "step": 20800 }, { "grad_norm": 0.12438026070594788, "learning_rate": 9.332137338386489e-05, "loss": 0.0074, "step": 20810 }, { "grad_norm": 0.11221524327993393, "learning_rate": 9.33131151915361e-05, "loss": 0.0074, "step": 20820 }, { "grad_norm": 0.15815909206867218, "learning_rate": 9.330485226255012e-05, "loss": 0.0085, "step": 20830 }, { "grad_norm": 0.12155668437480927, "learning_rate": 9.329658459781061e-05, "loss": 0.008, "step": 20840 }, { "grad_norm": 0.1058306097984314, "learning_rate": 9.328831219822172e-05, "loss": 0.0079, "step": 20850 }, { "grad_norm": 0.1324317902326584, "learning_rate": 9.328003506468808e-05, "loss": 0.0079, "step": 20860 }, { "grad_norm": 0.12476840615272522, "learning_rate": 9.327175319811488e-05, "loss": 0.009, "step": 20870 }, { "grad_norm": 0.07666981965303421, "learning_rate": 9.326346659940781e-05, "loss": 0.0065, "step": 20880 }, { "grad_norm": 0.10932741314172745, "learning_rate": 9.325517526947308e-05, "loss": 0.0074, "step": 20890 }, { "grad_norm": 0.1423187553882599, "learning_rate": 9.32468792092174e-05, "loss": 0.007, "step": 20900 }, { "grad_norm": 0.11217010766267776, "learning_rate": 9.323857841954803e-05, "loss": 0.0072, "step": 20910 }, { "grad_norm": 0.16779501736164093, "learning_rate": 9.323027290137276e-05, "loss": 0.009, "step": 20920 }, { "grad_norm": 0.15290167927742004, "learning_rate": 9.322196265559981e-05, "loss": 0.0068, "step": 20930 }, { "grad_norm": 0.1417829543352127, "learning_rate": 9.321364768313803e-05, "loss": 0.008, "step": 20940 }, { "grad_norm": 0.13967767357826233, "learning_rate": 9.32053279848967e-05, "loss": 0.0076, "step": 20950 }, { "grad_norm": 0.13890480995178223, "learning_rate": 9.319700356178567e-05, "loss": 0.0067, "step": 20960 }, { "grad_norm": 0.11320392787456512, "learning_rate": 9.318867441471527e-05, "loss": 0.0074, "step": 20970 }, { "grad_norm": 0.11731240153312683, "learning_rate": 9.318034054459637e-05, "loss": 0.0063, "step": 20980 }, { "grad_norm": 0.10442376136779785, "learning_rate": 9.317200195234034e-05, "loss": 0.0071, "step": 20990 }, { "grad_norm": 0.18392252922058105, "learning_rate": 9.316365863885909e-05, "loss": 0.0069, "step": 21000 }, { "grad_norm": 0.16530288755893707, "learning_rate": 9.315531060506502e-05, "loss": 0.0072, "step": 21010 }, { "grad_norm": 0.13385045528411865, "learning_rate": 9.314695785187108e-05, "loss": 0.0068, "step": 21020 }, { "grad_norm": 0.13961906731128693, "learning_rate": 9.313860038019069e-05, "loss": 0.0063, "step": 21030 }, { "grad_norm": 0.21674221754074097, "learning_rate": 9.313023819093782e-05, "loss": 0.008, "step": 21040 }, { "grad_norm": 0.1873546838760376, "learning_rate": 9.312187128502695e-05, "loss": 0.0077, "step": 21050 }, { "grad_norm": 0.11857907474040985, "learning_rate": 9.311349966337307e-05, "loss": 0.0064, "step": 21060 }, { "grad_norm": 0.12893952429294586, "learning_rate": 9.310512332689169e-05, "loss": 0.0064, "step": 21070 }, { "grad_norm": 0.1696704477071762, "learning_rate": 9.309674227649883e-05, "loss": 0.0085, "step": 21080 }, { "grad_norm": 0.1250547170639038, "learning_rate": 9.308835651311103e-05, "loss": 0.0069, "step": 21090 }, { "grad_norm": 0.1524982452392578, "learning_rate": 9.307996603764533e-05, "loss": 0.0064, "step": 21100 }, { "grad_norm": 0.13272607326507568, "learning_rate": 9.307157085101932e-05, "loss": 0.006, "step": 21110 }, { "grad_norm": 0.13113704323768616, "learning_rate": 9.306317095415109e-05, "loss": 0.0065, "step": 21120 }, { "grad_norm": 0.17655618488788605, "learning_rate": 9.305476634795922e-05, "loss": 0.0068, "step": 21130 }, { "grad_norm": 0.1460326611995697, "learning_rate": 9.304635703336284e-05, "loss": 0.0062, "step": 21140 }, { "grad_norm": 0.1388712376356125, "learning_rate": 9.303794301128157e-05, "loss": 0.0069, "step": 21150 }, { "grad_norm": 0.1549384593963623, "learning_rate": 9.302952428263555e-05, "loss": 0.0065, "step": 21160 }, { "grad_norm": 0.12340370565652847, "learning_rate": 9.302110084834545e-05, "loss": 0.0063, "step": 21170 }, { "grad_norm": 0.15899163484573364, "learning_rate": 9.301267270933245e-05, "loss": 0.0068, "step": 21180 }, { "grad_norm": 0.21415400505065918, "learning_rate": 9.300423986651823e-05, "loss": 0.0069, "step": 21190 }, { "grad_norm": 0.09414616972208023, "learning_rate": 9.299580232082501e-05, "loss": 0.0066, "step": 21200 }, { "grad_norm": 0.1529362052679062, "learning_rate": 9.298736007317547e-05, "loss": 0.0068, "step": 21210 }, { "grad_norm": 0.12270878255367279, "learning_rate": 9.297891312449288e-05, "loss": 0.0076, "step": 21220 }, { "grad_norm": 0.16994282603263855, "learning_rate": 9.297046147570094e-05, "loss": 0.0078, "step": 21230 }, { "grad_norm": 0.14940646290779114, "learning_rate": 9.296200512772396e-05, "loss": 0.0078, "step": 21240 }, { "grad_norm": 0.15336169302463531, "learning_rate": 9.295354408148668e-05, "loss": 0.0066, "step": 21250 }, { "grad_norm": 0.11602906137704849, "learning_rate": 9.294507833791441e-05, "loss": 0.0061, "step": 21260 }, { "grad_norm": 0.1462554633617401, "learning_rate": 9.293660789793295e-05, "loss": 0.0064, "step": 21270 }, { "grad_norm": 0.2186046838760376, "learning_rate": 9.292813276246858e-05, "loss": 0.0071, "step": 21280 }, { "grad_norm": 0.14762359857559204, "learning_rate": 9.291965293244816e-05, "loss": 0.0073, "step": 21290 }, { "grad_norm": 0.12373524159193039, "learning_rate": 9.291116840879904e-05, "loss": 0.0074, "step": 21300 }, { "grad_norm": 0.12209364771842957, "learning_rate": 9.290267919244904e-05, "loss": 0.0076, "step": 21310 }, { "grad_norm": 0.1606045514345169, "learning_rate": 9.289418528432655e-05, "loss": 0.0075, "step": 21320 }, { "grad_norm": 0.11317729949951172, "learning_rate": 9.288568668536045e-05, "loss": 0.0071, "step": 21330 }, { "grad_norm": 0.1428021341562271, "learning_rate": 9.287718339648013e-05, "loss": 0.0068, "step": 21340 }, { "grad_norm": 0.1466469168663025, "learning_rate": 9.28686754186155e-05, "loss": 0.0071, "step": 21350 }, { "grad_norm": 0.1509382426738739, "learning_rate": 9.286016275269698e-05, "loss": 0.0069, "step": 21360 }, { "grad_norm": 0.13560418784618378, "learning_rate": 9.285164539965551e-05, "loss": 0.0066, "step": 21370 }, { "grad_norm": 0.12972374260425568, "learning_rate": 9.284312336042251e-05, "loss": 0.0067, "step": 21380 }, { "grad_norm": 0.1066850796341896, "learning_rate": 9.283459663592996e-05, "loss": 0.0063, "step": 21390 }, { "grad_norm": 0.10998302698135376, "learning_rate": 9.282606522711033e-05, "loss": 0.0078, "step": 21400 }, { "grad_norm": 0.13957326114177704, "learning_rate": 9.281752913489657e-05, "loss": 0.0076, "step": 21410 }, { "grad_norm": 0.17560698091983795, "learning_rate": 9.280898836022222e-05, "loss": 0.01, "step": 21420 }, { "grad_norm": 0.13809829950332642, "learning_rate": 9.280044290402126e-05, "loss": 0.0085, "step": 21430 }, { "grad_norm": 0.13541999459266663, "learning_rate": 9.279189276722821e-05, "loss": 0.0071, "step": 21440 }, { "grad_norm": 0.10133214294910431, "learning_rate": 9.278333795077812e-05, "loss": 0.0064, "step": 21450 }, { "grad_norm": 0.10158152133226395, "learning_rate": 9.27747784556065e-05, "loss": 0.0067, "step": 21460 }, { "grad_norm": 0.11582990735769272, "learning_rate": 9.276621428264942e-05, "loss": 0.0076, "step": 21470 }, { "grad_norm": 0.13992524147033691, "learning_rate": 9.275764543284345e-05, "loss": 0.008, "step": 21480 }, { "grad_norm": 0.14029768109321594, "learning_rate": 9.274907190712566e-05, "loss": 0.0087, "step": 21490 }, { "grad_norm": 0.16509728133678436, "learning_rate": 9.274049370643363e-05, "loss": 0.0075, "step": 21500 }, { "grad_norm": 0.2108699083328247, "learning_rate": 9.273191083170547e-05, "loss": 0.009, "step": 21510 }, { "grad_norm": 0.18690897524356842, "learning_rate": 9.27233232838798e-05, "loss": 0.0132, "step": 21520 }, { "grad_norm": 0.17312179505825043, "learning_rate": 9.27147310638957e-05, "loss": 0.0092, "step": 21530 }, { "grad_norm": 0.18372467160224915, "learning_rate": 9.270613417269286e-05, "loss": 0.0081, "step": 21540 }, { "grad_norm": 0.18688400089740753, "learning_rate": 9.269753261121138e-05, "loss": 0.0091, "step": 21550 }, { "grad_norm": 0.11881575733423233, "learning_rate": 9.268892638039194e-05, "loss": 0.0082, "step": 21560 }, { "grad_norm": 0.13328544795513153, "learning_rate": 9.268031548117569e-05, "loss": 0.0081, "step": 21570 }, { "grad_norm": 0.12550917267799377, "learning_rate": 9.26716999145043e-05, "loss": 0.0078, "step": 21580 }, { "grad_norm": 0.09351325780153275, "learning_rate": 9.266307968131998e-05, "loss": 0.0069, "step": 21590 }, { "grad_norm": 0.16976776719093323, "learning_rate": 9.26544547825654e-05, "loss": 0.0064, "step": 21600 }, { "grad_norm": 0.16187642514705658, "learning_rate": 9.264582521918376e-05, "loss": 0.008, "step": 21610 }, { "grad_norm": 0.12245144695043564, "learning_rate": 9.263719099211881e-05, "loss": 0.0067, "step": 21620 }, { "grad_norm": 0.14923588931560516, "learning_rate": 9.262855210231476e-05, "loss": 0.0068, "step": 21630 }, { "grad_norm": 0.1331331580877304, "learning_rate": 9.261990855071633e-05, "loss": 0.0062, "step": 21640 }, { "grad_norm": 0.11313740909099579, "learning_rate": 9.261126033826878e-05, "loss": 0.0058, "step": 21650 }, { "grad_norm": 0.14055144786834717, "learning_rate": 9.260260746591786e-05, "loss": 0.0067, "step": 21660 }, { "grad_norm": 0.1642121970653534, "learning_rate": 9.259394993460985e-05, "loss": 0.0069, "step": 21670 }, { "grad_norm": 0.0937051773071289, "learning_rate": 9.258528774529151e-05, "loss": 0.0057, "step": 21680 }, { "grad_norm": 0.1439393013715744, "learning_rate": 9.257662089891013e-05, "loss": 0.0051, "step": 21690 }, { "grad_norm": 0.17804281413555145, "learning_rate": 9.25679493964135e-05, "loss": 0.0076, "step": 21700 }, { "grad_norm": 0.13774622976779938, "learning_rate": 9.255927323874994e-05, "loss": 0.0064, "step": 21710 }, { "grad_norm": 0.15505774319171906, "learning_rate": 9.255059242686822e-05, "loss": 0.007, "step": 21720 }, { "grad_norm": 0.1747245192527771, "learning_rate": 9.254190696171769e-05, "loss": 0.0058, "step": 21730 }, { "grad_norm": 0.12057976424694061, "learning_rate": 9.25332168442482e-05, "loss": 0.0056, "step": 21740 }, { "grad_norm": 0.1614835411310196, "learning_rate": 9.252452207541004e-05, "loss": 0.0076, "step": 21750 }, { "grad_norm": 0.1222984716296196, "learning_rate": 9.251582265615409e-05, "loss": 0.0062, "step": 21760 }, { "grad_norm": 0.16214130818843842, "learning_rate": 9.250711858743169e-05, "loss": 0.0061, "step": 21770 }, { "grad_norm": 0.15248101949691772, "learning_rate": 9.24984098701947e-05, "loss": 0.0078, "step": 21780 }, { "grad_norm": 0.11614740639925003, "learning_rate": 9.248969650539552e-05, "loss": 0.0057, "step": 21790 }, { "grad_norm": 0.20736220479011536, "learning_rate": 9.2480978493987e-05, "loss": 0.0075, "step": 21800 }, { "grad_norm": 0.14850927889347076, "learning_rate": 9.247225583692256e-05, "loss": 0.0077, "step": 21810 }, { "grad_norm": 0.11274724453687668, "learning_rate": 9.246352853515607e-05, "loss": 0.0078, "step": 21820 }, { "grad_norm": 0.13259299099445343, "learning_rate": 9.245479658964194e-05, "loss": 0.0071, "step": 21830 }, { "grad_norm": 0.11582872271537781, "learning_rate": 9.244606000133507e-05, "loss": 0.0073, "step": 21840 }, { "grad_norm": 0.11976059526205063, "learning_rate": 9.24373187711909e-05, "loss": 0.0083, "step": 21850 }, { "grad_norm": 0.16642209887504578, "learning_rate": 9.242857290016537e-05, "loss": 0.0075, "step": 21860 }, { "grad_norm": 0.17705999314785004, "learning_rate": 9.241982238921488e-05, "loss": 0.0085, "step": 21870 }, { "grad_norm": 0.18188445270061493, "learning_rate": 9.24110672392964e-05, "loss": 0.008, "step": 21880 }, { "grad_norm": 0.15792644023895264, "learning_rate": 9.240230745136737e-05, "loss": 0.0099, "step": 21890 }, { "grad_norm": 0.18545950949192047, "learning_rate": 9.239354302638575e-05, "loss": 0.0082, "step": 21900 }, { "grad_norm": 0.13760793209075928, "learning_rate": 9.238477396531e-05, "loss": 0.0084, "step": 21910 }, { "grad_norm": 0.15970249474048615, "learning_rate": 9.23760002690991e-05, "loss": 0.0062, "step": 21920 }, { "grad_norm": 0.1341264545917511, "learning_rate": 9.236722193871252e-05, "loss": 0.0079, "step": 21930 }, { "grad_norm": 0.11199303716421127, "learning_rate": 9.235843897511023e-05, "loss": 0.0078, "step": 21940 }, { "grad_norm": 0.11408074200153351, "learning_rate": 9.234965137925276e-05, "loss": 0.007, "step": 21950 }, { "grad_norm": 0.1446293294429779, "learning_rate": 9.234085915210108e-05, "loss": 0.0067, "step": 21960 }, { "grad_norm": 0.15240755677223206, "learning_rate": 9.23320622946167e-05, "loss": 0.0103, "step": 21970 }, { "grad_norm": 0.1677679717540741, "learning_rate": 9.232326080776163e-05, "loss": 0.0091, "step": 21980 }, { "grad_norm": 0.1690087914466858, "learning_rate": 9.23144546924984e-05, "loss": 0.0079, "step": 21990 }, { "grad_norm": 0.13339711725711823, "learning_rate": 9.230564394979e-05, "loss": 0.0081, "step": 22000 }, { "grad_norm": 0.1530826985836029, "learning_rate": 9.22968285806e-05, "loss": 0.0083, "step": 22010 }, { "grad_norm": 0.15892481803894043, "learning_rate": 9.228800858589242e-05, "loss": 0.0091, "step": 22020 }, { "grad_norm": 0.16265130043029785, "learning_rate": 9.227918396663179e-05, "loss": 0.0073, "step": 22030 }, { "grad_norm": 0.17868617177009583, "learning_rate": 9.227035472378319e-05, "loss": 0.009, "step": 22040 }, { "grad_norm": 0.13840921223163605, "learning_rate": 9.226152085831213e-05, "loss": 0.0078, "step": 22050 }, { "grad_norm": 0.13385801017284393, "learning_rate": 9.22526823711847e-05, "loss": 0.0063, "step": 22060 }, { "grad_norm": 0.11329860985279083, "learning_rate": 9.224383926336745e-05, "loss": 0.0067, "step": 22070 }, { "grad_norm": 0.12996183335781097, "learning_rate": 9.223499153582744e-05, "loss": 0.0064, "step": 22080 }, { "grad_norm": 0.13906189799308777, "learning_rate": 9.222613918953226e-05, "loss": 0.0061, "step": 22090 }, { "grad_norm": 0.10956253856420517, "learning_rate": 9.221728222544999e-05, "loss": 0.0063, "step": 22100 }, { "grad_norm": 0.16004300117492676, "learning_rate": 9.22084206445492e-05, "loss": 0.0064, "step": 22110 }, { "grad_norm": 0.10039234906435013, "learning_rate": 9.2199554447799e-05, "loss": 0.0052, "step": 22120 }, { "grad_norm": 0.11471423506736755, "learning_rate": 9.219068363616897e-05, "loss": 0.0064, "step": 22130 }, { "grad_norm": 0.1764424592256546, "learning_rate": 9.218180821062919e-05, "loss": 0.0071, "step": 22140 }, { "grad_norm": 0.1512899249792099, "learning_rate": 9.21729281721503e-05, "loss": 0.0064, "step": 22150 }, { "grad_norm": 0.19400379061698914, "learning_rate": 9.216404352170339e-05, "loss": 0.0087, "step": 22160 }, { "grad_norm": 0.15007343888282776, "learning_rate": 9.215515426026007e-05, "loss": 0.0075, "step": 22170 }, { "grad_norm": 0.1727759689092636, "learning_rate": 9.214626038879246e-05, "loss": 0.0075, "step": 22180 }, { "grad_norm": 0.20532527565956116, "learning_rate": 9.21373619082732e-05, "loss": 0.0095, "step": 22190 }, { "grad_norm": 0.17509229481220245, "learning_rate": 9.212845881967535e-05, "loss": 0.0104, "step": 22200 }, { "grad_norm": 0.1316119283437729, "learning_rate": 9.211955112397262e-05, "loss": 0.0093, "step": 22210 }, { "grad_norm": 0.15678934752941132, "learning_rate": 9.211063882213909e-05, "loss": 0.0076, "step": 22220 }, { "grad_norm": 0.1963956654071808, "learning_rate": 9.210172191514942e-05, "loss": 0.0099, "step": 22230 }, { "grad_norm": 0.14895574748516083, "learning_rate": 9.209280040397874e-05, "loss": 0.0099, "step": 22240 }, { "grad_norm": 0.13525570929050446, "learning_rate": 9.208387428960268e-05, "loss": 0.0094, "step": 22250 }, { "grad_norm": 0.1333514302968979, "learning_rate": 9.20749435729974e-05, "loss": 0.0068, "step": 22260 }, { "grad_norm": 0.14161299169063568, "learning_rate": 9.206600825513957e-05, "loss": 0.0076, "step": 22270 }, { "grad_norm": 0.09671173989772797, "learning_rate": 9.20570683370063e-05, "loss": 0.0072, "step": 22280 }, { "grad_norm": 0.12353810667991638, "learning_rate": 9.204812381957528e-05, "loss": 0.0084, "step": 22290 }, { "grad_norm": 0.4471836984157562, "learning_rate": 9.203917470382465e-05, "loss": 0.0112, "step": 22300 }, { "grad_norm": 0.2067767083644867, "learning_rate": 9.203022099073309e-05, "loss": 0.0144, "step": 22310 }, { "grad_norm": 0.22651949524879456, "learning_rate": 9.202126268127976e-05, "loss": 0.0206, "step": 22320 }, { "grad_norm": 0.16994832456111908, "learning_rate": 9.20122997764443e-05, "loss": 0.0175, "step": 22330 }, { "grad_norm": 0.18301373720169067, "learning_rate": 9.200333227720692e-05, "loss": 0.0127, "step": 22340 }, { "grad_norm": 0.1822933405637741, "learning_rate": 9.199436018454826e-05, "loss": 0.0122, "step": 22350 }, { "grad_norm": 0.20619864761829376, "learning_rate": 9.198538349944952e-05, "loss": 0.0086, "step": 22360 }, { "grad_norm": 0.2794857919216156, "learning_rate": 9.197640222289234e-05, "loss": 0.0107, "step": 22370 }, { "grad_norm": 0.2124301791191101, "learning_rate": 9.196741635585895e-05, "loss": 0.0094, "step": 22380 }, { "grad_norm": 0.12014969438314438, "learning_rate": 9.195842589933199e-05, "loss": 0.0095, "step": 22390 }, { "grad_norm": 0.14716245234012604, "learning_rate": 9.194943085429466e-05, "loss": 0.0104, "step": 22400 }, { "grad_norm": 0.14884845912456512, "learning_rate": 9.194043122173065e-05, "loss": 0.0098, "step": 22410 }, { "grad_norm": 0.15689145028591156, "learning_rate": 9.193142700262413e-05, "loss": 0.01, "step": 22420 }, { "grad_norm": 0.11692081391811371, "learning_rate": 9.192241819795979e-05, "loss": 0.0086, "step": 22430 }, { "grad_norm": 0.14096690714359283, "learning_rate": 9.191340480872284e-05, "loss": 0.0114, "step": 22440 }, { "grad_norm": 0.1672000139951706, "learning_rate": 9.190438683589895e-05, "loss": 0.0096, "step": 22450 }, { "grad_norm": 0.12934225797653198, "learning_rate": 9.189536428047432e-05, "loss": 0.0092, "step": 22460 }, { "grad_norm": 0.12571580708026886, "learning_rate": 9.188633714343564e-05, "loss": 0.0079, "step": 22470 }, { "grad_norm": 0.11164580285549164, "learning_rate": 9.18773054257701e-05, "loss": 0.0083, "step": 22480 }, { "grad_norm": 0.10088763386011124, "learning_rate": 9.18682691284654e-05, "loss": 0.006, "step": 22490 }, { "grad_norm": 0.10552465915679932, "learning_rate": 9.185922825250974e-05, "loss": 0.008, "step": 22500 }, { "grad_norm": 0.18767304718494415, "learning_rate": 9.185018279889181e-05, "loss": 0.0095, "step": 22510 }, { "grad_norm": 0.10796798020601273, "learning_rate": 9.184113276860082e-05, "loss": 0.0101, "step": 22520 }, { "grad_norm": 0.09295390546321869, "learning_rate": 9.183207816262645e-05, "loss": 0.0076, "step": 22530 }, { "grad_norm": 0.18628191947937012, "learning_rate": 9.182301898195891e-05, "loss": 0.0086, "step": 22540 }, { "grad_norm": 0.13358713686466217, "learning_rate": 9.181395522758889e-05, "loss": 0.0097, "step": 22550 }, { "grad_norm": 0.14156508445739746, "learning_rate": 9.180488690050759e-05, "loss": 0.0074, "step": 22560 }, { "grad_norm": 0.09143716096878052, "learning_rate": 9.179581400170671e-05, "loss": 0.0083, "step": 22570 }, { "grad_norm": 0.12738487124443054, "learning_rate": 9.178673653217845e-05, "loss": 0.0084, "step": 22580 }, { "grad_norm": 0.11936715990304947, "learning_rate": 9.177765449291551e-05, "loss": 0.0081, "step": 22590 }, { "grad_norm": 0.11972411721944809, "learning_rate": 9.176856788491109e-05, "loss": 0.0093, "step": 22600 }, { "grad_norm": 0.16031306982040405, "learning_rate": 9.175947670915887e-05, "loss": 0.0111, "step": 22610 }, { "grad_norm": 0.1946476250886917, "learning_rate": 9.175038096665309e-05, "loss": 0.0103, "step": 22620 }, { "grad_norm": 0.1552681177854538, "learning_rate": 9.17412806583884e-05, "loss": 0.0111, "step": 22630 }, { "grad_norm": 0.14633600413799286, "learning_rate": 9.173217578536002e-05, "loss": 0.0097, "step": 22640 }, { "grad_norm": 0.17692367732524872, "learning_rate": 9.172306634856362e-05, "loss": 0.0107, "step": 22650 }, { "grad_norm": 0.11255714297294617, "learning_rate": 9.171395234899545e-05, "loss": 0.0097, "step": 22660 }, { "grad_norm": 0.1573738157749176, "learning_rate": 9.170483378765214e-05, "loss": 0.01, "step": 22670 }, { "grad_norm": 0.1629123091697693, "learning_rate": 9.169571066553091e-05, "loss": 0.0076, "step": 22680 }, { "grad_norm": 0.1038072481751442, "learning_rate": 9.168658298362946e-05, "loss": 0.0087, "step": 22690 }, { "grad_norm": 0.09343370795249939, "learning_rate": 9.167745074294598e-05, "loss": 0.0079, "step": 22700 }, { "grad_norm": 0.1774098128080368, "learning_rate": 9.166831394447913e-05, "loss": 0.0093, "step": 22710 }, { "grad_norm": 0.11409668624401093, "learning_rate": 9.165917258922812e-05, "loss": 0.0077, "step": 22720 }, { "grad_norm": 0.13284365832805634, "learning_rate": 9.165002667819262e-05, "loss": 0.0069, "step": 22730 }, { "grad_norm": 0.13467666506767273, "learning_rate": 9.164087621237282e-05, "loss": 0.0076, "step": 22740 }, { "grad_norm": 0.1862022429704666, "learning_rate": 9.163172119276942e-05, "loss": 0.0071, "step": 22750 }, { "grad_norm": 0.1637917459011078, "learning_rate": 9.162256162038358e-05, "loss": 0.0077, "step": 22760 }, { "grad_norm": 0.13828125596046448, "learning_rate": 9.161339749621698e-05, "loss": 0.0083, "step": 22770 }, { "grad_norm": 0.14586351811885834, "learning_rate": 9.160422882127177e-05, "loss": 0.0075, "step": 22780 }, { "grad_norm": 0.10729104280471802, "learning_rate": 9.159505559655069e-05, "loss": 0.007, "step": 22790 }, { "grad_norm": 0.16163083910942078, "learning_rate": 9.158587782305684e-05, "loss": 0.0061, "step": 22800 }, { "grad_norm": 0.14565321803092957, "learning_rate": 9.157669550179391e-05, "loss": 0.0071, "step": 22810 }, { "grad_norm": 0.13045835494995117, "learning_rate": 9.156750863376609e-05, "loss": 0.0083, "step": 22820 }, { "grad_norm": 0.13985352218151093, "learning_rate": 9.155831721997801e-05, "loss": 0.0075, "step": 22830 }, { "grad_norm": 0.12738433480262756, "learning_rate": 9.154912126143484e-05, "loss": 0.0066, "step": 22840 }, { "grad_norm": 0.15096165239810944, "learning_rate": 9.153992075914224e-05, "loss": 0.007, "step": 22850 }, { "grad_norm": 0.1198379397392273, "learning_rate": 9.153071571410635e-05, "loss": 0.0066, "step": 22860 }, { "grad_norm": 0.10718231648206711, "learning_rate": 9.152150612733384e-05, "loss": 0.0062, "step": 22870 }, { "grad_norm": 0.13111205399036407, "learning_rate": 9.151229199983184e-05, "loss": 0.0074, "step": 22880 }, { "grad_norm": 0.15236249566078186, "learning_rate": 9.150307333260802e-05, "loss": 0.0065, "step": 22890 }, { "grad_norm": 0.10533840954303741, "learning_rate": 9.149385012667048e-05, "loss": 0.0066, "step": 22900 }, { "grad_norm": 0.16699673235416412, "learning_rate": 9.148462238302788e-05, "loss": 0.0075, "step": 22910 }, { "grad_norm": 0.15784607827663422, "learning_rate": 9.147539010268936e-05, "loss": 0.007, "step": 22920 }, { "grad_norm": 0.1160474643111229, "learning_rate": 9.14661532866645e-05, "loss": 0.0083, "step": 22930 }, { "grad_norm": 0.12931658327579498, "learning_rate": 9.145691193596348e-05, "loss": 0.007, "step": 22940 }, { "grad_norm": 0.09615453332662582, "learning_rate": 9.144766605159691e-05, "loss": 0.0057, "step": 22950 }, { "grad_norm": 0.08373869210481644, "learning_rate": 9.14384156345759e-05, "loss": 0.0065, "step": 22960 }, { "grad_norm": 0.11836820840835571, "learning_rate": 9.142916068591204e-05, "loss": 0.0066, "step": 22970 }, { "grad_norm": 0.1202104389667511, "learning_rate": 9.141990120661746e-05, "loss": 0.0079, "step": 22980 }, { "grad_norm": 0.15243156254291534, "learning_rate": 9.141063719770475e-05, "loss": 0.0073, "step": 22990 }, { "grad_norm": 0.0980241447687149, "learning_rate": 9.140136866018704e-05, "loss": 0.0067, "step": 23000 }, { "grad_norm": 0.14667710661888123, "learning_rate": 9.139209559507788e-05, "loss": 0.0092, "step": 23010 }, { "grad_norm": 0.1423598825931549, "learning_rate": 9.13828180033914e-05, "loss": 0.0079, "step": 23020 }, { "grad_norm": 0.11524704098701477, "learning_rate": 9.137353588614212e-05, "loss": 0.0068, "step": 23030 }, { "grad_norm": 0.11501014977693558, "learning_rate": 9.136424924434519e-05, "loss": 0.0065, "step": 23040 }, { "grad_norm": 0.12023493647575378, "learning_rate": 9.135495807901615e-05, "loss": 0.0072, "step": 23050 }, { "grad_norm": 0.10703840106725693, "learning_rate": 9.134566239117108e-05, "loss": 0.0067, "step": 23060 }, { "grad_norm": 0.18487438559532166, "learning_rate": 9.13363621818265e-05, "loss": 0.0077, "step": 23070 }, { "grad_norm": 0.2166164070367813, "learning_rate": 9.132705745199953e-05, "loss": 0.0086, "step": 23080 }, { "grad_norm": 0.16968344151973724, "learning_rate": 9.131774820270768e-05, "loss": 0.009, "step": 23090 }, { "grad_norm": 0.1568552404642105, "learning_rate": 9.130843443496901e-05, "loss": 0.0083, "step": 23100 }, { "grad_norm": 0.14110715687274933, "learning_rate": 9.129911614980206e-05, "loss": 0.0096, "step": 23110 }, { "grad_norm": 0.20887941122055054, "learning_rate": 9.128979334822584e-05, "loss": 0.0088, "step": 23120 }, { "grad_norm": 0.13220153748989105, "learning_rate": 9.128046603125992e-05, "loss": 0.0087, "step": 23130 }, { "grad_norm": 0.1291963756084442, "learning_rate": 9.12711341999243e-05, "loss": 0.0081, "step": 23140 }, { "grad_norm": 0.1893642395734787, "learning_rate": 9.12617978552395e-05, "loss": 0.0106, "step": 23150 }, { "grad_norm": 0.12784028053283691, "learning_rate": 9.12524569982265e-05, "loss": 0.008, "step": 23160 }, { "grad_norm": 0.17091791331768036, "learning_rate": 9.124311162990684e-05, "loss": 0.0076, "step": 23170 }, { "grad_norm": 0.1330336034297943, "learning_rate": 9.12337617513025e-05, "loss": 0.0076, "step": 23180 }, { "grad_norm": 0.16294054687023163, "learning_rate": 9.122440736343596e-05, "loss": 0.0076, "step": 23190 }, { "grad_norm": 0.10959187150001526, "learning_rate": 9.12150484673302e-05, "loss": 0.0062, "step": 23200 }, { "grad_norm": 0.11494715511798859, "learning_rate": 9.120568506400873e-05, "loss": 0.0074, "step": 23210 }, { "grad_norm": 0.13578519225120544, "learning_rate": 9.119631715449548e-05, "loss": 0.0065, "step": 23220 }, { "grad_norm": 0.1165890097618103, "learning_rate": 9.118694473981493e-05, "loss": 0.0063, "step": 23230 }, { "grad_norm": 0.15677814185619354, "learning_rate": 9.117756782099203e-05, "loss": 0.0065, "step": 23240 }, { "grad_norm": 0.1322413831949234, "learning_rate": 9.11681863990522e-05, "loss": 0.0069, "step": 23250 }, { "grad_norm": 0.13292211294174194, "learning_rate": 9.115880047502142e-05, "loss": 0.0075, "step": 23260 }, { "grad_norm": 0.11676987260580063, "learning_rate": 9.114941004992609e-05, "loss": 0.0065, "step": 23270 }, { "grad_norm": 0.10834596306085587, "learning_rate": 9.114001512479317e-05, "loss": 0.0092, "step": 23280 }, { "grad_norm": 0.11221576482057571, "learning_rate": 9.113061570065003e-05, "loss": 0.0054, "step": 23290 }, { "grad_norm": 0.12317612022161484, "learning_rate": 9.112121177852459e-05, "loss": 0.0062, "step": 23300 }, { "grad_norm": 0.10535067319869995, "learning_rate": 9.111180335944527e-05, "loss": 0.0059, "step": 23310 }, { "grad_norm": 0.09881430119276047, "learning_rate": 9.110239044444093e-05, "loss": 0.0054, "step": 23320 }, { "grad_norm": 0.08907444030046463, "learning_rate": 9.109297303454099e-05, "loss": 0.0055, "step": 23330 }, { "grad_norm": 0.14545899629592896, "learning_rate": 9.108355113077526e-05, "loss": 0.0072, "step": 23340 }, { "grad_norm": 0.13383905589580536, "learning_rate": 9.107412473417419e-05, "loss": 0.0055, "step": 23350 }, { "grad_norm": 0.17774082720279694, "learning_rate": 9.106469384576858e-05, "loss": 0.0067, "step": 23360 }, { "grad_norm": 0.10887420922517776, "learning_rate": 9.105525846658978e-05, "loss": 0.0074, "step": 23370 }, { "grad_norm": 0.12241212278604507, "learning_rate": 9.104581859766965e-05, "loss": 0.007, "step": 23380 }, { "grad_norm": 0.1395062506198883, "learning_rate": 9.10363742400405e-05, "loss": 0.0062, "step": 23390 }, { "grad_norm": 0.10229120403528214, "learning_rate": 9.102692539473518e-05, "loss": 0.0058, "step": 23400 }, { "grad_norm": 0.1319909393787384, "learning_rate": 9.101747206278697e-05, "loss": 0.0062, "step": 23410 }, { "grad_norm": 0.16166149079799652, "learning_rate": 9.100801424522968e-05, "loss": 0.0095, "step": 23420 }, { "grad_norm": 0.17526991665363312, "learning_rate": 9.099855194309762e-05, "loss": 0.0091, "step": 23430 }, { "grad_norm": 0.15234288573265076, "learning_rate": 9.098908515742554e-05, "loss": 0.0077, "step": 23440 }, { "grad_norm": 0.1594197154045105, "learning_rate": 9.097961388924873e-05, "loss": 0.0072, "step": 23450 }, { "grad_norm": 0.16872571408748627, "learning_rate": 9.097013813960298e-05, "loss": 0.0091, "step": 23460 }, { "grad_norm": 0.12706917524337769, "learning_rate": 9.09606579095245e-05, "loss": 0.008, "step": 23470 }, { "grad_norm": 0.15164849162101746, "learning_rate": 9.095117320005008e-05, "loss": 0.0076, "step": 23480 }, { "grad_norm": 0.1608472466468811, "learning_rate": 9.094168401221691e-05, "loss": 0.0076, "step": 23490 }, { "grad_norm": 0.1546970158815384, "learning_rate": 9.093219034706273e-05, "loss": 0.0075, "step": 23500 }, { "grad_norm": 0.13587890565395355, "learning_rate": 9.092269220562577e-05, "loss": 0.0068, "step": 23510 }, { "grad_norm": 0.14962266385555267, "learning_rate": 9.09131895889447e-05, "loss": 0.0078, "step": 23520 }, { "grad_norm": 0.13560114800930023, "learning_rate": 9.090368249805873e-05, "loss": 0.0079, "step": 23530 }, { "grad_norm": 0.12222881615161896, "learning_rate": 9.089417093400754e-05, "loss": 0.0077, "step": 23540 }, { "grad_norm": 0.11942017823457718, "learning_rate": 9.088465489783131e-05, "loss": 0.0073, "step": 23550 }, { "grad_norm": 0.10527737438678741, "learning_rate": 9.087513439057068e-05, "loss": 0.0065, "step": 23560 }, { "grad_norm": 0.15119054913520813, "learning_rate": 9.08656094132668e-05, "loss": 0.0093, "step": 23570 }, { "grad_norm": 0.10731984674930573, "learning_rate": 9.085607996696134e-05, "loss": 0.0086, "step": 23580 }, { "grad_norm": 0.11602286249399185, "learning_rate": 9.084654605269639e-05, "loss": 0.008, "step": 23590 }, { "grad_norm": 0.17227889597415924, "learning_rate": 9.083700767151457e-05, "loss": 0.008, "step": 23600 }, { "grad_norm": 0.22404102981090546, "learning_rate": 9.082746482445898e-05, "loss": 0.007, "step": 23610 }, { "grad_norm": 0.16937275230884552, "learning_rate": 9.081791751257325e-05, "loss": 0.0071, "step": 23620 }, { "grad_norm": 0.19482792913913727, "learning_rate": 9.080836573690142e-05, "loss": 0.0076, "step": 23630 }, { "grad_norm": 0.12101708352565765, "learning_rate": 9.079880949848805e-05, "loss": 0.0065, "step": 23640 }, { "grad_norm": 0.12385652959346771, "learning_rate": 9.078924879837822e-05, "loss": 0.0063, "step": 23650 }, { "grad_norm": 0.11857346445322037, "learning_rate": 9.077968363761747e-05, "loss": 0.0075, "step": 23660 }, { "grad_norm": 0.09472250938415527, "learning_rate": 9.077011401725182e-05, "loss": 0.006, "step": 23670 }, { "grad_norm": 0.07548431307077408, "learning_rate": 9.07605399383278e-05, "loss": 0.0046, "step": 23680 }, { "grad_norm": 0.1020861268043518, "learning_rate": 9.075096140189243e-05, "loss": 0.0054, "step": 23690 }, { "grad_norm": 0.13908343017101288, "learning_rate": 9.074137840899318e-05, "loss": 0.0051, "step": 23700 }, { "grad_norm": 0.12131751328706741, "learning_rate": 9.073179096067804e-05, "loss": 0.0055, "step": 23710 }, { "grad_norm": 0.10815561562776566, "learning_rate": 9.072219905799549e-05, "loss": 0.0061, "step": 23720 }, { "grad_norm": 0.08535411953926086, "learning_rate": 9.071260270199447e-05, "loss": 0.0062, "step": 23730 }, { "grad_norm": 0.09931673109531403, "learning_rate": 9.070300189372441e-05, "loss": 0.0052, "step": 23740 }, { "grad_norm": 0.12370508164167404, "learning_rate": 9.069339663423528e-05, "loss": 0.0064, "step": 23750 }, { "grad_norm": 0.12441319227218628, "learning_rate": 9.068378692457747e-05, "loss": 0.0082, "step": 23760 }, { "grad_norm": 0.11891334503889084, "learning_rate": 9.067417276580189e-05, "loss": 0.006, "step": 23770 }, { "grad_norm": 0.13251250982284546, "learning_rate": 9.066455415895993e-05, "loss": 0.0061, "step": 23780 }, { "grad_norm": 0.13101910054683685, "learning_rate": 9.065493110510346e-05, "loss": 0.0058, "step": 23790 }, { "grad_norm": 0.13275863230228424, "learning_rate": 9.064530360528484e-05, "loss": 0.0058, "step": 23800 }, { "grad_norm": 0.15198060870170593, "learning_rate": 9.063567166055695e-05, "loss": 0.0071, "step": 23810 }, { "grad_norm": 0.1308038979768753, "learning_rate": 9.062603527197308e-05, "loss": 0.0063, "step": 23820 }, { "grad_norm": 0.1485861986875534, "learning_rate": 9.06163944405871e-05, "loss": 0.0068, "step": 23830 }, { "grad_norm": 0.1563904583454132, "learning_rate": 9.060674916745327e-05, "loss": 0.0085, "step": 23840 }, { "grad_norm": 0.14901438355445862, "learning_rate": 9.05970994536264e-05, "loss": 0.008, "step": 23850 }, { "grad_norm": 0.18339982628822327, "learning_rate": 9.05874453001618e-05, "loss": 0.0077, "step": 23860 }, { "grad_norm": 0.17683029174804688, "learning_rate": 9.057778670811517e-05, "loss": 0.0072, "step": 23870 }, { "grad_norm": 0.13978953659534454, "learning_rate": 9.056812367854281e-05, "loss": 0.0063, "step": 23880 }, { "grad_norm": 0.12672843039035797, "learning_rate": 9.055845621250143e-05, "loss": 0.0067, "step": 23890 }, { "grad_norm": 0.45035406947135925, "learning_rate": 9.054878431104825e-05, "loss": 0.0327, "step": 23900 }, { "grad_norm": 0.3121587932109833, "learning_rate": 9.0539107975241e-05, "loss": 0.0256, "step": 23910 }, { "grad_norm": 0.2956344485282898, "learning_rate": 9.052942720613784e-05, "loss": 0.0164, "step": 23920 }, { "grad_norm": 0.24451956152915955, "learning_rate": 9.051974200479745e-05, "loss": 0.0155, "step": 23930 }, { "grad_norm": 0.14867718517780304, "learning_rate": 9.051005237227901e-05, "loss": 0.0137, "step": 23940 }, { "grad_norm": 0.23927131295204163, "learning_rate": 9.050035830964215e-05, "loss": 0.0115, "step": 23950 }, { "grad_norm": 0.24563077092170715, "learning_rate": 9.049065981794698e-05, "loss": 0.0171, "step": 23960 }, { "grad_norm": 0.2005382925271988, "learning_rate": 9.048095689825414e-05, "loss": 0.0176, "step": 23970 }, { "grad_norm": 0.11699552088975906, "learning_rate": 9.047124955162472e-05, "loss": 0.0116, "step": 23980 }, { "grad_norm": 0.10970669239759445, "learning_rate": 9.046153777912028e-05, "loss": 0.0106, "step": 23990 }, { "grad_norm": 0.12910248339176178, "learning_rate": 9.045182158180292e-05, "loss": 0.009, "step": 24000 }, { "grad_norm": 0.14125606417655945, "learning_rate": 9.044210096073516e-05, "loss": 0.0101, "step": 24010 }, { "grad_norm": 0.203143909573555, "learning_rate": 9.043237591698004e-05, "loss": 0.0101, "step": 24020 }, { "grad_norm": 0.17580589652061462, "learning_rate": 9.04226464516011e-05, "loss": 0.0077, "step": 24030 }, { "grad_norm": 0.10400719940662384, "learning_rate": 9.041291256566229e-05, "loss": 0.0078, "step": 24040 }, { "grad_norm": 0.1660243272781372, "learning_rate": 9.040317426022814e-05, "loss": 0.0085, "step": 24050 }, { "grad_norm": 0.17151592671871185, "learning_rate": 9.03934315363636e-05, "loss": 0.0074, "step": 24060 }, { "grad_norm": 0.10922721773386002, "learning_rate": 9.038368439513409e-05, "loss": 0.0075, "step": 24070 }, { "grad_norm": 0.18568545579910278, "learning_rate": 9.03739328376056e-05, "loss": 0.007, "step": 24080 }, { "grad_norm": 0.14489597082138062, "learning_rate": 9.036417686484451e-05, "loss": 0.0069, "step": 24090 }, { "grad_norm": 0.1831553727388382, "learning_rate": 9.035441647791773e-05, "loss": 0.0084, "step": 24100 }, { "grad_norm": 0.1427176296710968, "learning_rate": 9.034465167789263e-05, "loss": 0.0098, "step": 24110 }, { "grad_norm": 0.15330812335014343, "learning_rate": 9.033488246583706e-05, "loss": 0.0078, "step": 24120 }, { "grad_norm": 0.14914283156394958, "learning_rate": 9.032510884281941e-05, "loss": 0.0106, "step": 24130 }, { "grad_norm": 0.27331697940826416, "learning_rate": 9.031533080990848e-05, "loss": 0.0094, "step": 24140 }, { "grad_norm": 0.16585887968540192, "learning_rate": 9.030554836817358e-05, "loss": 0.0094, "step": 24150 }, { "grad_norm": 0.1673828661441803, "learning_rate": 9.029576151868451e-05, "loss": 0.0115, "step": 24160 }, { "grad_norm": 0.13738885521888733, "learning_rate": 9.028597026251155e-05, "loss": 0.0079, "step": 24170 }, { "grad_norm": 0.1276814192533493, "learning_rate": 9.027617460072547e-05, "loss": 0.0079, "step": 24180 }, { "grad_norm": 0.14547541737556458, "learning_rate": 9.026637453439745e-05, "loss": 0.0058, "step": 24190 }, { "grad_norm": 0.1265784054994583, "learning_rate": 9.025657006459927e-05, "loss": 0.0092, "step": 24200 }, { "grad_norm": 0.11464472115039825, "learning_rate": 9.024676119240311e-05, "loss": 0.0073, "step": 24210 }, { "grad_norm": 0.1061423122882843, "learning_rate": 9.023694791888166e-05, "loss": 0.0074, "step": 24220 }, { "grad_norm": 0.13098497688770294, "learning_rate": 9.022713024510808e-05, "loss": 0.0088, "step": 24230 }, { "grad_norm": 0.11949979513883591, "learning_rate": 9.021730817215601e-05, "loss": 0.0071, "step": 24240 }, { "grad_norm": 0.1503986269235611, "learning_rate": 9.02074817010996e-05, "loss": 0.0075, "step": 24250 }, { "grad_norm": 0.12870372831821442, "learning_rate": 9.019765083301342e-05, "loss": 0.0074, "step": 24260 }, { "grad_norm": 0.11290659755468369, "learning_rate": 9.01878155689726e-05, "loss": 0.0083, "step": 24270 }, { "grad_norm": 0.18805401027202606, "learning_rate": 9.017797591005268e-05, "loss": 0.0082, "step": 24280 }, { "grad_norm": 0.17563749849796295, "learning_rate": 9.016813185732972e-05, "loss": 0.0105, "step": 24290 }, { "grad_norm": 0.2037011981010437, "learning_rate": 9.015828341188027e-05, "loss": 0.0117, "step": 24300 }, { "grad_norm": 0.13550898432731628, "learning_rate": 9.01484305747813e-05, "loss": 0.0095, "step": 24310 }, { "grad_norm": 0.11987348645925522, "learning_rate": 9.013857334711033e-05, "loss": 0.0073, "step": 24320 }, { "grad_norm": 0.11145565658807755, "learning_rate": 9.012871172994534e-05, "loss": 0.0065, "step": 24330 }, { "grad_norm": 0.1483203023672104, "learning_rate": 9.011884572436476e-05, "loss": 0.008, "step": 24340 }, { "grad_norm": 0.16227637231349945, "learning_rate": 9.010897533144754e-05, "loss": 0.0067, "step": 24350 }, { "grad_norm": 0.16179685294628143, "learning_rate": 9.009910055227306e-05, "loss": 0.0066, "step": 24360 }, { "grad_norm": 0.13669291138648987, "learning_rate": 9.008922138792124e-05, "loss": 0.0062, "step": 24370 }, { "grad_norm": 0.11948449164628983, "learning_rate": 9.007933783947244e-05, "loss": 0.0059, "step": 24380 }, { "grad_norm": 0.19629760086536407, "learning_rate": 9.006944990800752e-05, "loss": 0.0078, "step": 24390 }, { "grad_norm": 0.13700777292251587, "learning_rate": 9.005955759460779e-05, "loss": 0.0091, "step": 24400 }, { "grad_norm": 0.17542020976543427, "learning_rate": 9.004966090035508e-05, "loss": 0.01, "step": 24410 }, { "grad_norm": 0.1686135232448578, "learning_rate": 9.003975982633166e-05, "loss": 0.0077, "step": 24420 }, { "grad_norm": 0.4256242513656616, "learning_rate": 9.00298543736203e-05, "loss": 0.0085, "step": 24430 }, { "grad_norm": 0.14504562318325043, "learning_rate": 9.001994454330427e-05, "loss": 0.0101, "step": 24440 }, { "grad_norm": 0.14794236421585083, "learning_rate": 9.001003033646727e-05, "loss": 0.0118, "step": 24450 }, { "grad_norm": 0.1585019826889038, "learning_rate": 9.00001117541935e-05, "loss": 0.0081, "step": 24460 }, { "grad_norm": 0.14234279096126556, "learning_rate": 8.999018879756764e-05, "loss": 0.0072, "step": 24470 }, { "grad_norm": 0.09363389015197754, "learning_rate": 8.998026146767487e-05, "loss": 0.0088, "step": 24480 }, { "grad_norm": 0.13719205558300018, "learning_rate": 8.99703297656008e-05, "loss": 0.0066, "step": 24490 }, { "grad_norm": 0.1337050199508667, "learning_rate": 8.996039369243156e-05, "loss": 0.0092, "step": 24500 }, { "grad_norm": 0.14033013582229614, "learning_rate": 8.995045324925378e-05, "loss": 0.0107, "step": 24510 }, { "grad_norm": 0.1176680326461792, "learning_rate": 8.994050843715448e-05, "loss": 0.0083, "step": 24520 }, { "grad_norm": 0.12653173506259918, "learning_rate": 8.993055925722121e-05, "loss": 0.0069, "step": 24530 }, { "grad_norm": 0.11101453006267548, "learning_rate": 8.992060571054202e-05, "loss": 0.0087, "step": 24540 }, { "grad_norm": 0.0913928672671318, "learning_rate": 8.991064779820542e-05, "loss": 0.0095, "step": 24550 }, { "grad_norm": 0.14129826426506042, "learning_rate": 8.990068552130036e-05, "loss": 0.0077, "step": 24560 }, { "grad_norm": 0.13402268290519714, "learning_rate": 8.989071888091634e-05, "loss": 0.0066, "step": 24570 }, { "grad_norm": 0.17478524148464203, "learning_rate": 8.988074787814329e-05, "loss": 0.0078, "step": 24580 }, { "grad_norm": 0.12451521307229996, "learning_rate": 8.987077251407158e-05, "loss": 0.0062, "step": 24590 }, { "grad_norm": 0.08016195148229599, "learning_rate": 8.986079278979216e-05, "loss": 0.0063, "step": 24600 }, { "grad_norm": 0.10570038110017776, "learning_rate": 8.985080870639635e-05, "loss": 0.0058, "step": 24610 }, { "grad_norm": 0.15766702592372894, "learning_rate": 8.984082026497603e-05, "loss": 0.0075, "step": 24620 }, { "grad_norm": 0.11403068900108337, "learning_rate": 8.98308274666235e-05, "loss": 0.0075, "step": 24630 }, { "grad_norm": 0.10242313891649246, "learning_rate": 8.982083031243155e-05, "loss": 0.0064, "step": 24640 }, { "grad_norm": 0.12190364301204681, "learning_rate": 8.98108288034935e-05, "loss": 0.0068, "step": 24650 }, { "grad_norm": 0.1271304190158844, "learning_rate": 8.980082294090305e-05, "loss": 0.007, "step": 24660 }, { "grad_norm": 0.4149022102355957, "learning_rate": 8.979081272575443e-05, "loss": 0.0057, "step": 24670 }, { "grad_norm": 0.14127792418003082, "learning_rate": 8.978079815914236e-05, "loss": 0.0064, "step": 24680 }, { "grad_norm": 0.15020155906677246, "learning_rate": 8.977077924216202e-05, "loss": 0.0069, "step": 24690 }, { "grad_norm": 0.1793440729379654, "learning_rate": 8.976075597590905e-05, "loss": 0.0072, "step": 24700 }, { "grad_norm": 0.17005638778209686, "learning_rate": 8.975072836147958e-05, "loss": 0.0091, "step": 24710 }, { "grad_norm": 0.10303844511508942, "learning_rate": 8.974069639997025e-05, "loss": 0.0063, "step": 24720 }, { "grad_norm": 0.17015887796878815, "learning_rate": 8.973066009247808e-05, "loss": 0.0062, "step": 24730 }, { "grad_norm": 0.06581411510705948, "learning_rate": 8.972061944010066e-05, "loss": 0.0063, "step": 24740 }, { "grad_norm": 0.14752864837646484, "learning_rate": 8.971057444393603e-05, "loss": 0.007, "step": 24750 }, { "grad_norm": 0.09715969860553741, "learning_rate": 8.970052510508268e-05, "loss": 0.0059, "step": 24760 }, { "grad_norm": 0.10934031009674072, "learning_rate": 8.969047142463959e-05, "loss": 0.0072, "step": 24770 }, { "grad_norm": 0.10753252357244492, "learning_rate": 8.968041340370621e-05, "loss": 0.0061, "step": 24780 }, { "grad_norm": 0.11216410249471664, "learning_rate": 8.96703510433825e-05, "loss": 0.0063, "step": 24790 }, { "grad_norm": 0.15535607933998108, "learning_rate": 8.966028434476883e-05, "loss": 0.0065, "step": 24800 }, { "grad_norm": 0.13480542600154877, "learning_rate": 8.96502133089661e-05, "loss": 0.0077, "step": 24810 }, { "grad_norm": 0.10943946987390518, "learning_rate": 8.964013793707564e-05, "loss": 0.009, "step": 24820 }, { "grad_norm": 0.10555145144462585, "learning_rate": 8.963005823019932e-05, "loss": 0.0075, "step": 24830 }, { "grad_norm": 0.11332305520772934, "learning_rate": 8.961997418943939e-05, "loss": 0.0078, "step": 24840 }, { "grad_norm": 0.09676569700241089, "learning_rate": 8.960988581589865e-05, "loss": 0.0066, "step": 24850 }, { "grad_norm": 0.15880052745342255, "learning_rate": 8.959979311068037e-05, "loss": 0.0087, "step": 24860 }, { "grad_norm": 0.1127621978521347, "learning_rate": 8.958969607488823e-05, "loss": 0.0064, "step": 24870 }, { "grad_norm": 0.18975473940372467, "learning_rate": 8.957959470962647e-05, "loss": 0.0089, "step": 24880 }, { "grad_norm": 0.11201585084199905, "learning_rate": 8.956948901599971e-05, "loss": 0.0076, "step": 24890 }, { "grad_norm": 0.08285154402256012, "learning_rate": 8.955937899511315e-05, "loss": 0.0058, "step": 24900 }, { "grad_norm": 0.12132681161165237, "learning_rate": 8.954926464807238e-05, "loss": 0.008, "step": 24910 }, { "grad_norm": 0.15531371533870697, "learning_rate": 8.953914597598347e-05, "loss": 0.0071, "step": 24920 }, { "grad_norm": 0.08356307446956635, "learning_rate": 8.952902297995303e-05, "loss": 0.0059, "step": 24930 }, { "grad_norm": 0.1225881278514862, "learning_rate": 8.951889566108804e-05, "loss": 0.0076, "step": 24940 }, { "grad_norm": 0.0919528678059578, "learning_rate": 8.950876402049606e-05, "loss": 0.005, "step": 24950 }, { "grad_norm": 0.12270626425743103, "learning_rate": 8.949862805928504e-05, "loss": 0.0068, "step": 24960 }, { "grad_norm": 0.1301615983247757, "learning_rate": 8.948848777856343e-05, "loss": 0.0055, "step": 24970 }, { "grad_norm": 0.08295132219791412, "learning_rate": 8.947834317944017e-05, "loss": 0.0061, "step": 24980 }, { "grad_norm": 0.10218722373247147, "learning_rate": 8.946819426302466e-05, "loss": 0.0051, "step": 24990 }, { "grad_norm": 0.17010067403316498, "learning_rate": 8.945804103042676e-05, "loss": 0.0079, "step": 25000 }, { "grad_norm": 0.1329234093427658, "learning_rate": 8.944788348275681e-05, "loss": 0.0056, "step": 25010 }, { "grad_norm": 0.16279256343841553, "learning_rate": 8.943772162112565e-05, "loss": 0.0091, "step": 25020 }, { "grad_norm": 0.15054579079151154, "learning_rate": 8.942755544664454e-05, "loss": 0.009, "step": 25030 }, { "grad_norm": 0.12100110948085785, "learning_rate": 8.941738496042525e-05, "loss": 0.0086, "step": 25040 }, { "grad_norm": 0.141637921333313, "learning_rate": 8.940721016357999e-05, "loss": 0.0066, "step": 25050 }, { "grad_norm": 0.09686225652694702, "learning_rate": 8.939703105722148e-05, "loss": 0.0062, "step": 25060 }, { "grad_norm": 0.10421869903802872, "learning_rate": 8.93868476424629e-05, "loss": 0.0059, "step": 25070 }, { "grad_norm": 0.14298206567764282, "learning_rate": 8.937665992041786e-05, "loss": 0.0111, "step": 25080 }, { "grad_norm": 0.1034500002861023, "learning_rate": 8.93664678922005e-05, "loss": 0.0067, "step": 25090 }, { "grad_norm": 0.1291128545999527, "learning_rate": 8.93562715589254e-05, "loss": 0.0067, "step": 25100 }, { "grad_norm": 0.16043585538864136, "learning_rate": 8.934607092170762e-05, "loss": 0.0073, "step": 25110 }, { "grad_norm": 0.10584166646003723, "learning_rate": 8.933586598166266e-05, "loss": 0.0057, "step": 25120 }, { "grad_norm": 0.12514559924602509, "learning_rate": 8.932565673990655e-05, "loss": 0.0061, "step": 25130 }, { "grad_norm": 0.145785853266716, "learning_rate": 8.931544319755574e-05, "loss": 0.0067, "step": 25140 }, { "grad_norm": 0.14333944022655487, "learning_rate": 8.930522535572718e-05, "loss": 0.0064, "step": 25150 }, { "grad_norm": 0.15817999839782715, "learning_rate": 8.929500321553826e-05, "loss": 0.0069, "step": 25160 }, { "grad_norm": 0.08710937201976776, "learning_rate": 8.928477677810686e-05, "loss": 0.0062, "step": 25170 }, { "grad_norm": 0.1089831292629242, "learning_rate": 8.927454604455137e-05, "loss": 0.0063, "step": 25180 }, { "grad_norm": 0.09117243438959122, "learning_rate": 8.926431101599053e-05, "loss": 0.0057, "step": 25190 }, { "grad_norm": 0.09103097766637802, "learning_rate": 8.925407169354369e-05, "loss": 0.0055, "step": 25200 }, { "grad_norm": 0.14570564031600952, "learning_rate": 8.92438280783306e-05, "loss": 0.0067, "step": 25210 }, { "grad_norm": 0.12319418042898178, "learning_rate": 8.923358017147146e-05, "loss": 0.0073, "step": 25220 }, { "grad_norm": 0.10598481446504593, "learning_rate": 8.922332797408697e-05, "loss": 0.007, "step": 25230 }, { "grad_norm": 0.09150014817714691, "learning_rate": 8.921307148729831e-05, "loss": 0.0075, "step": 25240 }, { "grad_norm": 0.12140679359436035, "learning_rate": 8.920281071222712e-05, "loss": 0.0063, "step": 25250 }, { "grad_norm": 0.11773635447025299, "learning_rate": 8.919254564999548e-05, "loss": 0.0048, "step": 25260 }, { "grad_norm": 0.08916568011045456, "learning_rate": 8.918227630172598e-05, "loss": 0.0058, "step": 25270 }, { "grad_norm": 0.10338445007801056, "learning_rate": 8.917200266854165e-05, "loss": 0.0047, "step": 25280 }, { "grad_norm": 0.09357922524213791, "learning_rate": 8.9161724751566e-05, "loss": 0.006, "step": 25290 }, { "grad_norm": 0.10906552523374557, "learning_rate": 8.915144255192302e-05, "loss": 0.0057, "step": 25300 }, { "grad_norm": 0.11542797088623047, "learning_rate": 8.914115607073714e-05, "loss": 0.0067, "step": 25310 }, { "grad_norm": 0.18133607506752014, "learning_rate": 8.913086530913327e-05, "loss": 0.0097, "step": 25320 }, { "grad_norm": 0.13590896129608154, "learning_rate": 8.912057026823681e-05, "loss": 0.01, "step": 25330 }, { "grad_norm": 0.26509904861450195, "learning_rate": 8.91102709491736e-05, "loss": 0.0105, "step": 25340 }, { "grad_norm": 0.2055283486843109, "learning_rate": 8.909996735306996e-05, "loss": 0.0092, "step": 25350 }, { "grad_norm": 0.21239449083805084, "learning_rate": 8.908965948105268e-05, "loss": 0.0101, "step": 25360 }, { "grad_norm": 0.17790810763835907, "learning_rate": 8.907934733424901e-05, "loss": 0.0112, "step": 25370 }, { "grad_norm": 0.09891059994697571, "learning_rate": 8.906903091378666e-05, "loss": 0.0075, "step": 25380 }, { "grad_norm": 0.20380161702632904, "learning_rate": 8.905871022079384e-05, "loss": 0.0086, "step": 25390 }, { "grad_norm": 0.15406399965286255, "learning_rate": 8.90483852563992e-05, "loss": 0.0075, "step": 25400 }, { "grad_norm": 0.19643144309520721, "learning_rate": 8.903805602173185e-05, "loss": 0.0095, "step": 25410 }, { "grad_norm": 0.1464192271232605, "learning_rate": 8.902772251792137e-05, "loss": 0.0086, "step": 25420 }, { "grad_norm": 0.11759446561336517, "learning_rate": 8.901738474609786e-05, "loss": 0.0078, "step": 25430 }, { "grad_norm": 0.1382470577955246, "learning_rate": 8.900704270739179e-05, "loss": 0.0067, "step": 25440 }, { "grad_norm": 0.14194783568382263, "learning_rate": 8.89966964029342e-05, "loss": 0.0075, "step": 25450 }, { "grad_norm": 0.08486741036176682, "learning_rate": 8.898634583385652e-05, "loss": 0.0065, "step": 25460 }, { "grad_norm": 0.10983921587467194, "learning_rate": 8.897599100129065e-05, "loss": 0.0065, "step": 25470 }, { "grad_norm": 0.1292782574892044, "learning_rate": 8.896563190636903e-05, "loss": 0.0073, "step": 25480 }, { "grad_norm": 0.14518137276172638, "learning_rate": 8.895526855022448e-05, "loss": 0.0084, "step": 25490 }, { "grad_norm": 0.1480627954006195, "learning_rate": 8.894490093399033e-05, "loss": 0.0097, "step": 25500 }, { "grad_norm": 0.16481894254684448, "learning_rate": 8.893452905880035e-05, "loss": 0.0088, "step": 25510 }, { "grad_norm": 0.12436044961214066, "learning_rate": 8.892415292578883e-05, "loss": 0.0097, "step": 25520 }, { "grad_norm": 0.10070884227752686, "learning_rate": 8.891377253609046e-05, "loss": 0.0102, "step": 25530 }, { "grad_norm": 0.1631312221288681, "learning_rate": 8.890338789084043e-05, "loss": 0.0085, "step": 25540 }, { "grad_norm": 0.17833666503429413, "learning_rate": 8.88929989911744e-05, "loss": 0.0102, "step": 25550 }, { "grad_norm": 0.1889190971851349, "learning_rate": 8.888260583822847e-05, "loss": 0.0121, "step": 25560 }, { "grad_norm": 0.17462162673473358, "learning_rate": 8.887220843313921e-05, "loss": 0.0099, "step": 25570 }, { "grad_norm": 0.19795657694339752, "learning_rate": 8.88618067770437e-05, "loss": 0.0085, "step": 25580 }, { "grad_norm": 0.13661330938339233, "learning_rate": 8.885140087107942e-05, "loss": 0.0102, "step": 25590 }, { "grad_norm": 0.12709522247314453, "learning_rate": 8.884099071638436e-05, "loss": 0.0073, "step": 25600 }, { "grad_norm": 0.12666817009449005, "learning_rate": 8.883057631409695e-05, "loss": 0.0072, "step": 25610 }, { "grad_norm": 0.2623928189277649, "learning_rate": 8.882015766535608e-05, "loss": 0.0089, "step": 25620 }, { "grad_norm": 0.176285982131958, "learning_rate": 8.880973477130115e-05, "loss": 0.0116, "step": 25630 }, { "grad_norm": 0.1118612065911293, "learning_rate": 8.879930763307197e-05, "loss": 0.0091, "step": 25640 }, { "grad_norm": 0.15126220881938934, "learning_rate": 8.878887625180884e-05, "loss": 0.0079, "step": 25650 }, { "grad_norm": 0.13586269319057465, "learning_rate": 8.877844062865253e-05, "loss": 0.0071, "step": 25660 }, { "grad_norm": 0.1415441781282425, "learning_rate": 8.876800076474424e-05, "loss": 0.0085, "step": 25670 }, { "grad_norm": 0.14020049571990967, "learning_rate": 8.875755666122568e-05, "loss": 0.007, "step": 25680 }, { "grad_norm": 0.1301334798336029, "learning_rate": 8.8747108319239e-05, "loss": 0.0065, "step": 25690 }, { "grad_norm": 0.12328092753887177, "learning_rate": 8.87366557399268e-05, "loss": 0.0065, "step": 25700 }, { "grad_norm": 0.11549562960863113, "learning_rate": 8.872619892443217e-05, "loss": 0.0083, "step": 25710 }, { "grad_norm": 0.10875890403985977, "learning_rate": 8.871573787389865e-05, "loss": 0.0094, "step": 25720 }, { "grad_norm": 0.11910989880561829, "learning_rate": 8.870527258947024e-05, "loss": 0.0071, "step": 25730 }, { "grad_norm": 0.11097840219736099, "learning_rate": 8.869480307229143e-05, "loss": 0.0076, "step": 25740 }, { "grad_norm": 0.1058402732014656, "learning_rate": 8.868432932350712e-05, "loss": 0.0072, "step": 25750 }, { "grad_norm": 0.10852372646331787, "learning_rate": 8.867385134426272e-05, "loss": 0.0073, "step": 25760 }, { "grad_norm": 0.14910554885864258, "learning_rate": 8.866336913570407e-05, "loss": 0.0094, "step": 25770 }, { "grad_norm": 0.14032317698001862, "learning_rate": 8.865288269897751e-05, "loss": 0.0078, "step": 25780 }, { "grad_norm": 0.15292102098464966, "learning_rate": 8.864239203522981e-05, "loss": 0.009, "step": 25790 }, { "grad_norm": 0.1327805072069168, "learning_rate": 8.863189714560822e-05, "loss": 0.0061, "step": 25800 }, { "grad_norm": 0.09700029343366623, "learning_rate": 8.862139803126043e-05, "loss": 0.0055, "step": 25810 }, { "grad_norm": 0.08869752287864685, "learning_rate": 8.861089469333463e-05, "loss": 0.0064, "step": 25820 }, { "grad_norm": 0.20904970169067383, "learning_rate": 8.860038713297944e-05, "loss": 0.0051, "step": 25830 }, { "grad_norm": 0.15854589641094208, "learning_rate": 8.858987535134394e-05, "loss": 0.0065, "step": 25840 }, { "grad_norm": 0.09094732999801636, "learning_rate": 8.857935934957769e-05, "loss": 0.0057, "step": 25850 }, { "grad_norm": 0.13716650009155273, "learning_rate": 8.856883912883071e-05, "loss": 0.0062, "step": 25860 }, { "grad_norm": 0.1426178216934204, "learning_rate": 8.855831469025346e-05, "loss": 0.007, "step": 25870 }, { "grad_norm": 0.11974912881851196, "learning_rate": 8.854778603499689e-05, "loss": 0.0073, "step": 25880 }, { "grad_norm": 0.16261136531829834, "learning_rate": 8.85372531642124e-05, "loss": 0.0081, "step": 25890 }, { "grad_norm": 0.21089358627796173, "learning_rate": 8.852671607905185e-05, "loss": 0.0075, "step": 25900 }, { "grad_norm": 0.1877056509256363, "learning_rate": 8.851617478066754e-05, "loss": 0.0076, "step": 25910 }, { "grad_norm": 0.30480748414993286, "learning_rate": 8.850562927021227e-05, "loss": 0.0102, "step": 25920 }, { "grad_norm": 0.15641237795352936, "learning_rate": 8.849507954883928e-05, "loss": 0.0194, "step": 25930 }, { "grad_norm": 0.1548752635717392, "learning_rate": 8.848452561770226e-05, "loss": 0.0129, "step": 25940 }, { "grad_norm": 0.2263067215681076, "learning_rate": 8.847396747795538e-05, "loss": 0.0096, "step": 25950 }, { "grad_norm": 0.1778947114944458, "learning_rate": 8.846340513075327e-05, "loss": 0.0117, "step": 25960 }, { "grad_norm": 0.1773536503314972, "learning_rate": 8.845283857725099e-05, "loss": 0.0093, "step": 25970 }, { "grad_norm": 0.1997239887714386, "learning_rate": 8.844226781860409e-05, "loss": 0.0109, "step": 25980 }, { "grad_norm": 0.15450121462345123, "learning_rate": 8.84316928559686e-05, "loss": 0.0111, "step": 25990 }, { "grad_norm": 0.16750116646289825, "learning_rate": 8.842111369050094e-05, "loss": 0.0091, "step": 26000 }, { "grad_norm": 0.15128272771835327, "learning_rate": 8.841053032335808e-05, "loss": 0.0084, "step": 26010 }, { "grad_norm": 0.16429965198040009, "learning_rate": 8.839994275569735e-05, "loss": 0.0087, "step": 26020 }, { "grad_norm": 0.1480948030948639, "learning_rate": 8.838935098867662e-05, "loss": 0.0085, "step": 26030 }, { "grad_norm": 0.11998945474624634, "learning_rate": 8.837875502345418e-05, "loss": 0.008, "step": 26040 }, { "grad_norm": 0.12933287024497986, "learning_rate": 8.83681548611888e-05, "loss": 0.0084, "step": 26050 }, { "grad_norm": 0.10104888677597046, "learning_rate": 8.835755050303969e-05, "loss": 0.0088, "step": 26060 }, { "grad_norm": 0.10722129791975021, "learning_rate": 8.834694195016653e-05, "loss": 0.0062, "step": 26070 }, { "grad_norm": 0.07395298779010773, "learning_rate": 8.833632920372942e-05, "loss": 0.0068, "step": 26080 }, { "grad_norm": 0.09949087351560593, "learning_rate": 8.832571226488903e-05, "loss": 0.0074, "step": 26090 }, { "grad_norm": 0.12003261595964432, "learning_rate": 8.831509113480634e-05, "loss": 0.0061, "step": 26100 }, { "grad_norm": 0.11384882777929306, "learning_rate": 8.83044658146429e-05, "loss": 0.0082, "step": 26110 }, { "grad_norm": 0.17478756606578827, "learning_rate": 8.829383630556067e-05, "loss": 0.0082, "step": 26120 }, { "grad_norm": 0.17681634426116943, "learning_rate": 8.828320260872207e-05, "loss": 0.007, "step": 26130 }, { "grad_norm": 0.11655397713184357, "learning_rate": 8.827256472529e-05, "loss": 0.0065, "step": 26140 }, { "grad_norm": 0.13833944499492645, "learning_rate": 8.826192265642778e-05, "loss": 0.0072, "step": 26150 }, { "grad_norm": 0.1105741485953331, "learning_rate": 8.825127640329923e-05, "loss": 0.0061, "step": 26160 }, { "grad_norm": 0.09162295609712601, "learning_rate": 8.824062596706861e-05, "loss": 0.0071, "step": 26170 }, { "grad_norm": 0.08154502511024475, "learning_rate": 8.822997134890062e-05, "loss": 0.0055, "step": 26180 }, { "grad_norm": 0.1100768968462944, "learning_rate": 8.821931254996044e-05, "loss": 0.0072, "step": 26190 }, { "grad_norm": 0.1428341418504715, "learning_rate": 8.82086495714137e-05, "loss": 0.0074, "step": 26200 }, { "grad_norm": 0.11914908140897751, "learning_rate": 8.81979824144265e-05, "loss": 0.0053, "step": 26210 }, { "grad_norm": 0.12823723256587982, "learning_rate": 8.818731108016536e-05, "loss": 0.0069, "step": 26220 }, { "grad_norm": 0.10156946629285812, "learning_rate": 8.81766355697973e-05, "loss": 0.0068, "step": 26230 }, { "grad_norm": 0.21336574852466583, "learning_rate": 8.816595588448977e-05, "loss": 0.0082, "step": 26240 }, { "grad_norm": 0.11183802038431168, "learning_rate": 8.81552720254107e-05, "loss": 0.0066, "step": 26250 }, { "grad_norm": 0.14774738252162933, "learning_rate": 8.814458399372842e-05, "loss": 0.0097, "step": 26260 }, { "grad_norm": 0.14143648743629456, "learning_rate": 8.813389179061181e-05, "loss": 0.0068, "step": 26270 }, { "grad_norm": 0.09907852858304977, "learning_rate": 8.812319541723012e-05, "loss": 0.0066, "step": 26280 }, { "grad_norm": 0.1265621930360794, "learning_rate": 8.811249487475309e-05, "loss": 0.0071, "step": 26290 }, { "grad_norm": 0.10701841861009598, "learning_rate": 8.810179016435092e-05, "loss": 0.0058, "step": 26300 }, { "grad_norm": 0.11647503823041916, "learning_rate": 8.809108128719428e-05, "loss": 0.0085, "step": 26310 }, { "grad_norm": 0.10182558745145798, "learning_rate": 8.808036824445424e-05, "loss": 0.0068, "step": 26320 }, { "grad_norm": 0.09100115299224854, "learning_rate": 8.806965103730238e-05, "loss": 0.006, "step": 26330 }, { "grad_norm": 0.15012165904045105, "learning_rate": 8.805892966691074e-05, "loss": 0.0089, "step": 26340 }, { "grad_norm": 0.13273632526397705, "learning_rate": 8.804820413445175e-05, "loss": 0.0062, "step": 26350 }, { "grad_norm": 0.08563567698001862, "learning_rate": 8.803747444109837e-05, "loss": 0.0053, "step": 26360 }, { "grad_norm": 0.10381972789764404, "learning_rate": 8.802674058802399e-05, "loss": 0.0053, "step": 26370 }, { "grad_norm": 0.09790708869695663, "learning_rate": 8.801600257640241e-05, "loss": 0.0066, "step": 26380 }, { "grad_norm": 0.17541758716106415, "learning_rate": 8.800526040740795e-05, "loss": 0.0104, "step": 26390 }, { "grad_norm": 0.17099672555923462, "learning_rate": 8.799451408221535e-05, "loss": 0.0086, "step": 26400 }, { "grad_norm": 0.11388708651065826, "learning_rate": 8.798376360199982e-05, "loss": 0.0076, "step": 26410 }, { "grad_norm": 0.10303670912981033, "learning_rate": 8.797300896793701e-05, "loss": 0.0058, "step": 26420 }, { "grad_norm": 0.12391842156648636, "learning_rate": 8.796225018120302e-05, "loss": 0.0059, "step": 26430 }, { "grad_norm": 0.15769678354263306, "learning_rate": 8.795148724297444e-05, "loss": 0.0075, "step": 26440 }, { "grad_norm": 0.09601952880620956, "learning_rate": 8.794072015442825e-05, "loss": 0.0058, "step": 26450 }, { "grad_norm": 0.10422777384519577, "learning_rate": 8.792994891674198e-05, "loss": 0.0058, "step": 26460 }, { "grad_norm": 0.09583377838134766, "learning_rate": 8.79191735310935e-05, "loss": 0.0061, "step": 26470 }, { "grad_norm": 0.07973944395780563, "learning_rate": 8.790839399866122e-05, "loss": 0.0061, "step": 26480 }, { "grad_norm": 0.0997391864657402, "learning_rate": 8.789761032062397e-05, "loss": 0.006, "step": 26490 }, { "grad_norm": 0.0996568575501442, "learning_rate": 8.788682249816103e-05, "loss": 0.0058, "step": 26500 }, { "grad_norm": 0.07861094921827316, "learning_rate": 8.787603053245215e-05, "loss": 0.0073, "step": 26510 }, { "grad_norm": 0.09748488664627075, "learning_rate": 8.78652344246775e-05, "loss": 0.0056, "step": 26520 }, { "grad_norm": 0.12467802315950394, "learning_rate": 8.785443417601776e-05, "loss": 0.0078, "step": 26530 }, { "grad_norm": 0.1353979855775833, "learning_rate": 8.784362978765401e-05, "loss": 0.0069, "step": 26540 }, { "grad_norm": 0.15078090131282806, "learning_rate": 8.783282126076779e-05, "loss": 0.0061, "step": 26550 }, { "grad_norm": 0.12689486145973206, "learning_rate": 8.782200859654112e-05, "loss": 0.0064, "step": 26560 }, { "grad_norm": 0.11248613893985748, "learning_rate": 8.781119179615646e-05, "loss": 0.0071, "step": 26570 }, { "grad_norm": 0.14422345161437988, "learning_rate": 8.780037086079674e-05, "loss": 0.0087, "step": 26580 }, { "grad_norm": 0.10881442576646805, "learning_rate": 8.778954579164527e-05, "loss": 0.0065, "step": 26590 }, { "grad_norm": 0.14526237547397614, "learning_rate": 8.777871658988588e-05, "loss": 0.0091, "step": 26600 }, { "grad_norm": 0.1510900855064392, "learning_rate": 8.776788325670285e-05, "loss": 0.007, "step": 26610 }, { "grad_norm": 0.2049652338027954, "learning_rate": 8.775704579328089e-05, "loss": 0.008, "step": 26620 }, { "grad_norm": 0.11826125532388687, "learning_rate": 8.774620420080517e-05, "loss": 0.0085, "step": 26630 }, { "grad_norm": 0.1394089311361313, "learning_rate": 8.773535848046131e-05, "loss": 0.0075, "step": 26640 }, { "grad_norm": 0.11450022459030151, "learning_rate": 8.772450863343538e-05, "loss": 0.0068, "step": 26650 }, { "grad_norm": 0.5671424269676208, "learning_rate": 8.77136546609139e-05, "loss": 0.0102, "step": 26660 }, { "grad_norm": 0.15785463154315948, "learning_rate": 8.770279656408385e-05, "loss": 0.0109, "step": 26670 }, { "grad_norm": 0.16076956689357758, "learning_rate": 8.769193434413265e-05, "loss": 0.0092, "step": 26680 }, { "grad_norm": 0.11907277256250381, "learning_rate": 8.76810680022482e-05, "loss": 0.0087, "step": 26690 }, { "grad_norm": 0.13736414909362793, "learning_rate": 8.767019753961878e-05, "loss": 0.0078, "step": 26700 }, { "grad_norm": 0.12105780839920044, "learning_rate": 8.765932295743321e-05, "loss": 0.0076, "step": 26710 }, { "grad_norm": 0.12503065168857574, "learning_rate": 8.764844425688068e-05, "loss": 0.0065, "step": 26720 }, { "grad_norm": 0.173686683177948, "learning_rate": 8.763756143915092e-05, "loss": 0.01, "step": 26730 }, { "grad_norm": 0.1113879531621933, "learning_rate": 8.7626674505434e-05, "loss": 0.008, "step": 26740 }, { "grad_norm": 0.1128518208861351, "learning_rate": 8.761578345692053e-05, "loss": 0.0064, "step": 26750 }, { "grad_norm": 0.1802881360054016, "learning_rate": 8.760488829480156e-05, "loss": 0.0089, "step": 26760 }, { "grad_norm": 0.22366315126419067, "learning_rate": 8.759398902026854e-05, "loss": 0.0085, "step": 26770 }, { "grad_norm": 0.1349668949842453, "learning_rate": 8.758308563451339e-05, "loss": 0.0083, "step": 26780 }, { "grad_norm": 0.13949310779571533, "learning_rate": 8.75721781387285e-05, "loss": 0.0096, "step": 26790 }, { "grad_norm": 0.1527024358510971, "learning_rate": 8.75612665341067e-05, "loss": 0.0087, "step": 26800 }, { "grad_norm": 0.13291414082050323, "learning_rate": 8.755035082184126e-05, "loss": 0.0072, "step": 26810 }, { "grad_norm": 0.11894219368696213, "learning_rate": 8.753943100312592e-05, "loss": 0.0071, "step": 26820 }, { "grad_norm": 0.12229418009519577, "learning_rate": 8.752850707915484e-05, "loss": 0.0067, "step": 26830 }, { "grad_norm": 0.1892842948436737, "learning_rate": 8.751757905112264e-05, "loss": 0.0058, "step": 26840 }, { "grad_norm": 0.10695743560791016, "learning_rate": 8.75066469202244e-05, "loss": 0.0065, "step": 26850 }, { "grad_norm": 0.0864701047539711, "learning_rate": 8.749571068765567e-05, "loss": 0.0072, "step": 26860 }, { "grad_norm": 0.09158168733119965, "learning_rate": 8.748477035461238e-05, "loss": 0.0063, "step": 26870 }, { "grad_norm": 0.12182426452636719, "learning_rate": 8.747382592229095e-05, "loss": 0.0057, "step": 26880 }, { "grad_norm": 0.08602140098810196, "learning_rate": 8.746287739188828e-05, "loss": 0.0055, "step": 26890 }, { "grad_norm": 0.08946720510721207, "learning_rate": 8.745192476460165e-05, "loss": 0.0061, "step": 26900 }, { "grad_norm": 0.09459126740694046, "learning_rate": 8.744096804162882e-05, "loss": 0.0061, "step": 26910 }, { "grad_norm": 0.09597020596265793, "learning_rate": 8.743000722416804e-05, "loss": 0.0051, "step": 26920 }, { "grad_norm": 0.09982280433177948, "learning_rate": 8.741904231341793e-05, "loss": 0.0054, "step": 26930 }, { "grad_norm": 0.0822678655385971, "learning_rate": 8.740807331057762e-05, "loss": 0.0052, "step": 26940 }, { "grad_norm": 0.08956585079431534, "learning_rate": 8.739710021684667e-05, "loss": 0.0053, "step": 26950 }, { "grad_norm": 0.1205177754163742, "learning_rate": 8.738612303342503e-05, "loss": 0.0048, "step": 26960 }, { "grad_norm": 0.12507687509059906, "learning_rate": 8.73751417615132e-05, "loss": 0.0039, "step": 26970 }, { "grad_norm": 0.12611107528209686, "learning_rate": 8.736415640231208e-05, "loss": 0.0057, "step": 26980 }, { "grad_norm": 0.11390677094459534, "learning_rate": 8.735316695702297e-05, "loss": 0.0053, "step": 26990 }, { "grad_norm": 0.13673372566699982, "learning_rate": 8.734217342684769e-05, "loss": 0.0057, "step": 27000 }, { "grad_norm": 0.10131809860467911, "learning_rate": 8.733117581298847e-05, "loss": 0.0056, "step": 27010 }, { "grad_norm": 0.1188313364982605, "learning_rate": 8.732017411664796e-05, "loss": 0.006, "step": 27020 }, { "grad_norm": 0.08968942612409592, "learning_rate": 8.730916833902936e-05, "loss": 0.008, "step": 27030 }, { "grad_norm": 0.12128722667694092, "learning_rate": 8.729815848133618e-05, "loss": 0.0066, "step": 27040 }, { "grad_norm": 0.10593883693218231, "learning_rate": 8.728714454477247e-05, "loss": 0.0065, "step": 27050 }, { "grad_norm": 0.14646659791469574, "learning_rate": 8.727612653054269e-05, "loss": 0.008, "step": 27060 }, { "grad_norm": 0.12054923176765442, "learning_rate": 8.726510443985176e-05, "loss": 0.0082, "step": 27070 }, { "grad_norm": 0.11202564090490341, "learning_rate": 8.725407827390503e-05, "loss": 0.0072, "step": 27080 }, { "grad_norm": 0.14348755776882172, "learning_rate": 8.724304803390833e-05, "loss": 0.0064, "step": 27090 }, { "grad_norm": 0.09685664623975754, "learning_rate": 8.723201372106788e-05, "loss": 0.0059, "step": 27100 }, { "grad_norm": 0.14783677458763123, "learning_rate": 8.722097533659038e-05, "loss": 0.0062, "step": 27110 }, { "grad_norm": 0.13156777620315552, "learning_rate": 8.720993288168299e-05, "loss": 0.0059, "step": 27120 }, { "grad_norm": 0.11006385087966919, "learning_rate": 8.719888635755327e-05, "loss": 0.0068, "step": 27130 }, { "grad_norm": 0.10814523696899414, "learning_rate": 8.718783576540928e-05, "loss": 0.0059, "step": 27140 }, { "grad_norm": 0.12029953300952911, "learning_rate": 8.717678110645948e-05, "loss": 0.0053, "step": 27150 }, { "grad_norm": 0.12221744656562805, "learning_rate": 8.716572238191279e-05, "loss": 0.0045, "step": 27160 }, { "grad_norm": 0.0891849473118782, "learning_rate": 8.715465959297857e-05, "loss": 0.0048, "step": 27170 }, { "grad_norm": 0.13931745290756226, "learning_rate": 8.714359274086665e-05, "loss": 0.0056, "step": 27180 }, { "grad_norm": 0.11762450635433197, "learning_rate": 8.713252182678726e-05, "loss": 0.0055, "step": 27190 }, { "grad_norm": 0.134361132979393, "learning_rate": 8.712144685195112e-05, "loss": 0.0063, "step": 27200 }, { "grad_norm": 0.11791511625051498, "learning_rate": 8.711036781756936e-05, "loss": 0.0061, "step": 27210 }, { "grad_norm": 0.1338060349225998, "learning_rate": 8.709928472485357e-05, "loss": 0.006, "step": 27220 }, { "grad_norm": 0.12019756436347961, "learning_rate": 8.708819757501579e-05, "loss": 0.0067, "step": 27230 }, { "grad_norm": 0.12730708718299866, "learning_rate": 8.707710636926846e-05, "loss": 0.0076, "step": 27240 }, { "grad_norm": 0.2236880511045456, "learning_rate": 8.706601110882455e-05, "loss": 0.0133, "step": 27250 }, { "grad_norm": 0.24902692437171936, "learning_rate": 8.705491179489738e-05, "loss": 0.0101, "step": 27260 }, { "grad_norm": 0.13359707593917847, "learning_rate": 8.704380842870077e-05, "loss": 0.0083, "step": 27270 }, { "grad_norm": 0.13815003633499146, "learning_rate": 8.703270101144895e-05, "loss": 0.008, "step": 27280 }, { "grad_norm": 0.14214900135993958, "learning_rate": 8.702158954435664e-05, "loss": 0.0068, "step": 27290 }, { "grad_norm": 0.12367979437112808, "learning_rate": 8.701047402863896e-05, "loss": 0.0078, "step": 27300 }, { "grad_norm": 0.11544149369001389, "learning_rate": 8.699935446551148e-05, "loss": 0.006, "step": 27310 }, { "grad_norm": 0.11410960555076599, "learning_rate": 8.698823085619022e-05, "loss": 0.0067, "step": 27320 }, { "grad_norm": 0.08892063051462173, "learning_rate": 8.697710320189166e-05, "loss": 0.0059, "step": 27330 }, { "grad_norm": 0.10802832245826721, "learning_rate": 8.696597150383268e-05, "loss": 0.0068, "step": 27340 }, { "grad_norm": 0.10507161170244217, "learning_rate": 8.695483576323063e-05, "loss": 0.0055, "step": 27350 }, { "grad_norm": 0.07874419540166855, "learning_rate": 8.69436959813033e-05, "loss": 0.0062, "step": 27360 }, { "grad_norm": 0.09444262832403183, "learning_rate": 8.693255215926892e-05, "loss": 0.0065, "step": 27370 }, { "grad_norm": 0.21678435802459717, "learning_rate": 8.692140429834617e-05, "loss": 0.0068, "step": 27380 }, { "grad_norm": 0.11952652037143707, "learning_rate": 8.691025239975415e-05, "loss": 0.0061, "step": 27390 }, { "grad_norm": 0.13724489510059357, "learning_rate": 8.689909646471243e-05, "loss": 0.007, "step": 27400 }, { "grad_norm": 0.09540851414203644, "learning_rate": 8.688793649444099e-05, "loss": 0.0078, "step": 27410 }, { "grad_norm": 0.15406861901283264, "learning_rate": 8.687677249016029e-05, "loss": 0.0066, "step": 27420 }, { "grad_norm": 0.14153103530406952, "learning_rate": 8.686560445309118e-05, "loss": 0.008, "step": 27430 }, { "grad_norm": 0.2475711554288864, "learning_rate": 8.685443238445499e-05, "loss": 0.01, "step": 27440 }, { "grad_norm": 0.17776599526405334, "learning_rate": 8.68432562854735e-05, "loss": 0.0108, "step": 27450 }, { "grad_norm": 0.22104781866073608, "learning_rate": 8.683207615736887e-05, "loss": 0.0114, "step": 27460 }, { "grad_norm": 0.168935626745224, "learning_rate": 8.682089200136379e-05, "loss": 0.0108, "step": 27470 }, { "grad_norm": 0.1439049392938614, "learning_rate": 8.680970381868132e-05, "loss": 0.0074, "step": 27480 }, { "grad_norm": 0.09815579652786255, "learning_rate": 8.679851161054498e-05, "loss": 0.0095, "step": 27490 }, { "grad_norm": 0.11034809052944183, "learning_rate": 8.678731537817873e-05, "loss": 0.0082, "step": 27500 }, { "grad_norm": 0.10356298834085464, "learning_rate": 8.677611512280697e-05, "loss": 0.0068, "step": 27510 }, { "grad_norm": 0.10867080092430115, "learning_rate": 8.676491084565457e-05, "loss": 0.0075, "step": 27520 }, { "grad_norm": 0.24335747957229614, "learning_rate": 8.675370254794678e-05, "loss": 0.008, "step": 27530 }, { "grad_norm": 0.15782582759857178, "learning_rate": 8.674249023090935e-05, "loss": 0.0088, "step": 27540 }, { "grad_norm": 0.20072799921035767, "learning_rate": 8.673127389576843e-05, "loss": 0.0086, "step": 27550 }, { "grad_norm": 0.25436529517173767, "learning_rate": 8.67200535437506e-05, "loss": 0.0119, "step": 27560 }, { "grad_norm": 0.20411641895771027, "learning_rate": 8.670882917608296e-05, "loss": 0.012, "step": 27570 }, { "grad_norm": 0.5763986706733704, "learning_rate": 8.669760079399292e-05, "loss": 0.0107, "step": 27580 }, { "grad_norm": 0.15181556344032288, "learning_rate": 8.668636839870845e-05, "loss": 0.0124, "step": 27590 }, { "grad_norm": 0.28107720613479614, "learning_rate": 8.667513199145789e-05, "loss": 0.018, "step": 27600 }, { "grad_norm": 0.18135450780391693, "learning_rate": 8.666389157347002e-05, "loss": 0.0128, "step": 27610 }, { "grad_norm": 0.14336246252059937, "learning_rate": 8.66526471459741e-05, "loss": 0.0097, "step": 27620 }, { "grad_norm": 0.11732470244169235, "learning_rate": 8.66413987101998e-05, "loss": 0.0087, "step": 27630 }, { "grad_norm": 0.19798974692821503, "learning_rate": 8.663014626737723e-05, "loss": 0.0104, "step": 27640 }, { "grad_norm": 0.16302873194217682, "learning_rate": 8.661888981873691e-05, "loss": 0.011, "step": 27650 }, { "grad_norm": 0.13406699895858765, "learning_rate": 8.660762936550988e-05, "loss": 0.0088, "step": 27660 }, { "grad_norm": 0.4531153738498688, "learning_rate": 8.659636490892753e-05, "loss": 0.0114, "step": 27670 }, { "grad_norm": 0.17877736687660217, "learning_rate": 8.658509645022174e-05, "loss": 0.0132, "step": 27680 }, { "grad_norm": 0.15296699106693268, "learning_rate": 8.657382399062481e-05, "loss": 0.0122, "step": 27690 }, { "grad_norm": 0.12527413666248322, "learning_rate": 8.656254753136946e-05, "loss": 0.0093, "step": 27700 }, { "grad_norm": 0.1938420981168747, "learning_rate": 8.655126707368891e-05, "loss": 0.012, "step": 27710 }, { "grad_norm": 0.1835276186466217, "learning_rate": 8.653998261881672e-05, "loss": 0.0118, "step": 27720 }, { "grad_norm": 0.15525904297828674, "learning_rate": 8.652869416798699e-05, "loss": 0.0095, "step": 27730 }, { "grad_norm": 0.1288290023803711, "learning_rate": 8.651740172243417e-05, "loss": 0.0096, "step": 27740 }, { "grad_norm": 0.10524071007966995, "learning_rate": 8.65061052833932e-05, "loss": 0.0067, "step": 27750 }, { "grad_norm": 0.12485368549823761, "learning_rate": 8.649480485209945e-05, "loss": 0.0072, "step": 27760 }, { "grad_norm": 0.11821868270635605, "learning_rate": 8.64835004297887e-05, "loss": 0.0068, "step": 27770 }, { "grad_norm": 0.1307949274778366, "learning_rate": 8.64721920176972e-05, "loss": 0.0084, "step": 27780 }, { "grad_norm": 0.12886692583560944, "learning_rate": 8.646087961706164e-05, "loss": 0.0075, "step": 27790 }, { "grad_norm": 0.1628750115633011, "learning_rate": 8.644956322911908e-05, "loss": 0.0096, "step": 27800 }, { "grad_norm": 0.1319960057735443, "learning_rate": 8.643824285510709e-05, "loss": 0.0089, "step": 27810 }, { "grad_norm": 0.16332285106182098, "learning_rate": 8.642691849626364e-05, "loss": 0.0089, "step": 27820 }, { "grad_norm": 0.11460389941930771, "learning_rate": 8.641559015382717e-05, "loss": 0.0078, "step": 27830 }, { "grad_norm": 0.12893636524677277, "learning_rate": 8.640425782903649e-05, "loss": 0.0063, "step": 27840 }, { "grad_norm": 0.6195265650749207, "learning_rate": 8.639292152313091e-05, "loss": 0.0069, "step": 27850 }, { "grad_norm": 0.2740634083747864, "learning_rate": 8.638158123735015e-05, "loss": 0.0083, "step": 27860 }, { "grad_norm": 0.22297874093055725, "learning_rate": 8.637023697293436e-05, "loss": 0.0115, "step": 27870 }, { "grad_norm": 0.2317165732383728, "learning_rate": 8.635888873112414e-05, "loss": 0.0142, "step": 27880 }, { "grad_norm": 0.12556509673595428, "learning_rate": 8.634753651316052e-05, "loss": 0.0134, "step": 27890 }, { "grad_norm": 0.1614890843629837, "learning_rate": 8.633618032028496e-05, "loss": 0.0099, "step": 27900 }, { "grad_norm": 0.16473199427127838, "learning_rate": 8.632482015373934e-05, "loss": 0.0098, "step": 27910 }, { "grad_norm": 0.13238349556922913, "learning_rate": 8.6313456014766e-05, "loss": 0.0092, "step": 27920 }, { "grad_norm": 0.17371173202991486, "learning_rate": 8.630208790460771e-05, "loss": 0.007, "step": 27930 }, { "grad_norm": 0.14581969380378723, "learning_rate": 8.629071582450768e-05, "loss": 0.0086, "step": 27940 }, { "grad_norm": 0.1306581050157547, "learning_rate": 8.62793397757095e-05, "loss": 0.0068, "step": 27950 }, { "grad_norm": 0.11024607717990875, "learning_rate": 8.626795975945729e-05, "loss": 0.0077, "step": 27960 }, { "grad_norm": 0.12787586450576782, "learning_rate": 8.625657577699551e-05, "loss": 0.0065, "step": 27970 }, { "grad_norm": 0.12453877925872803, "learning_rate": 8.624518782956914e-05, "loss": 0.0063, "step": 27980 }, { "grad_norm": 0.11750174313783646, "learning_rate": 8.62337959184235e-05, "loss": 0.008, "step": 27990 }, { "grad_norm": 0.11664518713951111, "learning_rate": 8.622240004480441e-05, "loss": 0.0071, "step": 28000 }, { "grad_norm": 0.08766645938158035, "learning_rate": 8.621100020995814e-05, "loss": 0.0059, "step": 28010 }, { "grad_norm": 0.1488117277622223, "learning_rate": 8.619959641513132e-05, "loss": 0.0079, "step": 28020 }, { "grad_norm": 0.15173545479774475, "learning_rate": 8.618818866157105e-05, "loss": 0.0055, "step": 28030 }, { "grad_norm": 0.09311085939407349, "learning_rate": 8.617677695052487e-05, "loss": 0.007, "step": 28040 }, { "grad_norm": 0.09144808351993561, "learning_rate": 8.616536128324078e-05, "loss": 0.006, "step": 28050 }, { "grad_norm": 0.14339730143547058, "learning_rate": 8.615394166096712e-05, "loss": 0.0069, "step": 28060 }, { "grad_norm": 0.2315031886100769, "learning_rate": 8.614251808495279e-05, "loss": 0.0059, "step": 28070 }, { "grad_norm": 0.10546281933784485, "learning_rate": 8.6131090556447e-05, "loss": 0.0072, "step": 28080 }, { "grad_norm": 0.1297224909067154, "learning_rate": 8.611965907669947e-05, "loss": 0.0074, "step": 28090 }, { "grad_norm": 0.07499151676893234, "learning_rate": 8.610822364696034e-05, "loss": 0.0067, "step": 28100 }, { "grad_norm": 0.11815523356199265, "learning_rate": 8.609678426848015e-05, "loss": 0.008, "step": 28110 }, { "grad_norm": 0.14651383459568024, "learning_rate": 8.60853409425099e-05, "loss": 0.0064, "step": 28120 }, { "grad_norm": 0.11373255401849747, "learning_rate": 8.607389367030104e-05, "loss": 0.0061, "step": 28130 }, { "grad_norm": 0.10479322820901871, "learning_rate": 8.606244245310538e-05, "loss": 0.0077, "step": 28140 }, { "grad_norm": 0.11849235743284225, "learning_rate": 8.605098729217525e-05, "loss": 0.0077, "step": 28150 }, { "grad_norm": 0.1431727111339569, "learning_rate": 8.603952818876335e-05, "loss": 0.0071, "step": 28160 }, { "grad_norm": 0.15690866112709045, "learning_rate": 8.602806514412281e-05, "loss": 0.0075, "step": 28170 }, { "grad_norm": 0.10959208756685257, "learning_rate": 8.601659815950726e-05, "loss": 0.0072, "step": 28180 }, { "grad_norm": 0.12499158829450607, "learning_rate": 8.600512723617067e-05, "loss": 0.0065, "step": 28190 }, { "grad_norm": 0.07018215209245682, "learning_rate": 8.59936523753675e-05, "loss": 0.0043, "step": 28200 }, { "grad_norm": 0.11679545044898987, "learning_rate": 8.598217357835264e-05, "loss": 0.0051, "step": 28210 }, { "grad_norm": 0.21339745819568634, "learning_rate": 8.597069084638135e-05, "loss": 0.0064, "step": 28220 }, { "grad_norm": 0.13350501656532288, "learning_rate": 8.595920418070939e-05, "loss": 0.0058, "step": 28230 }, { "grad_norm": 0.14246851205825806, "learning_rate": 8.594771358259295e-05, "loss": 0.0064, "step": 28240 }, { "grad_norm": 0.1118985041975975, "learning_rate": 8.593621905328858e-05, "loss": 0.0072, "step": 28250 }, { "grad_norm": 0.12053028494119644, "learning_rate": 8.592472059405333e-05, "loss": 0.0077, "step": 28260 }, { "grad_norm": 0.19463908672332764, "learning_rate": 8.591321820614464e-05, "loss": 0.0081, "step": 28270 }, { "grad_norm": 0.07161613553762436, "learning_rate": 8.590171189082041e-05, "loss": 0.0055, "step": 28280 }, { "grad_norm": 0.10935773700475693, "learning_rate": 8.589020164933894e-05, "loss": 0.0054, "step": 28290 }, { "grad_norm": 0.1097244843840599, "learning_rate": 8.587868748295898e-05, "loss": 0.0052, "step": 28300 }, { "grad_norm": 0.0976918488740921, "learning_rate": 8.586716939293971e-05, "loss": 0.0066, "step": 28310 }, { "grad_norm": 0.09809000790119171, "learning_rate": 8.58556473805407e-05, "loss": 0.0054, "step": 28320 }, { "grad_norm": 0.10575724393129349, "learning_rate": 8.584412144702202e-05, "loss": 0.0047, "step": 28330 }, { "grad_norm": 0.11318891495466232, "learning_rate": 8.58325915936441e-05, "loss": 0.0062, "step": 28340 }, { "grad_norm": 0.10815466195344925, "learning_rate": 8.582105782166783e-05, "loss": 0.0055, "step": 28350 }, { "grad_norm": 0.12973693013191223, "learning_rate": 8.580952013235455e-05, "loss": 0.0069, "step": 28360 }, { "grad_norm": 0.1264285296201706, "learning_rate": 8.579797852696596e-05, "loss": 0.0058, "step": 28370 }, { "grad_norm": 0.16005465388298035, "learning_rate": 8.578643300676428e-05, "loss": 0.0066, "step": 28380 }, { "grad_norm": 0.12616252899169922, "learning_rate": 8.577488357301209e-05, "loss": 0.0064, "step": 28390 }, { "grad_norm": 0.13389329612255096, "learning_rate": 8.576333022697242e-05, "loss": 0.0075, "step": 28400 }, { "grad_norm": 0.13547246158123016, "learning_rate": 8.575177296990873e-05, "loss": 0.008, "step": 28410 }, { "grad_norm": 0.10519737750291824, "learning_rate": 8.574021180308489e-05, "loss": 0.007, "step": 28420 }, { "grad_norm": 0.12311842292547226, "learning_rate": 8.572864672776523e-05, "loss": 0.0064, "step": 28430 }, { "grad_norm": 0.10627817362546921, "learning_rate": 8.571707774521447e-05, "loss": 0.0063, "step": 28440 }, { "grad_norm": 0.12862342596054077, "learning_rate": 8.57055048566978e-05, "loss": 0.0065, "step": 28450 }, { "grad_norm": 0.10198923200368881, "learning_rate": 8.569392806348078e-05, "loss": 0.005, "step": 28460 }, { "grad_norm": 0.27762579917907715, "learning_rate": 8.568234736682947e-05, "loss": 0.0074, "step": 28470 }, { "grad_norm": 0.16344793140888214, "learning_rate": 8.567076276801029e-05, "loss": 0.0082, "step": 28480 }, { "grad_norm": 0.12101816385984421, "learning_rate": 8.565917426829013e-05, "loss": 0.0096, "step": 28490 }, { "grad_norm": 0.14104394614696503, "learning_rate": 8.564758186893628e-05, "loss": 0.0091, "step": 28500 }, { "grad_norm": 0.09676503390073776, "learning_rate": 8.563598557121649e-05, "loss": 0.0069, "step": 28510 }, { "grad_norm": 0.11429625004529953, "learning_rate": 8.562438537639888e-05, "loss": 0.0059, "step": 28520 }, { "grad_norm": 0.10232023149728775, "learning_rate": 8.561278128575206e-05, "loss": 0.0052, "step": 28530 }, { "grad_norm": 0.1213131844997406, "learning_rate": 8.5601173300545e-05, "loss": 0.0068, "step": 28540 }, { "grad_norm": 0.08292629569768906, "learning_rate": 8.558956142204717e-05, "loss": 0.005, "step": 28550 }, { "grad_norm": 0.12269514054059982, "learning_rate": 8.55779456515284e-05, "loss": 0.0056, "step": 28560 }, { "grad_norm": 0.12347190827131271, "learning_rate": 8.556632599025898e-05, "loss": 0.0059, "step": 28570 }, { "grad_norm": 0.10832846909761429, "learning_rate": 8.555470243950964e-05, "loss": 0.0085, "step": 28580 }, { "grad_norm": 0.10778695344924927, "learning_rate": 8.554307500055148e-05, "loss": 0.0053, "step": 28590 }, { "grad_norm": 0.07618127018213272, "learning_rate": 8.553144367465609e-05, "loss": 0.0073, "step": 28600 }, { "grad_norm": 0.11641543358564377, "learning_rate": 8.551980846309544e-05, "loss": 0.0061, "step": 28610 }, { "grad_norm": 0.09468582272529602, "learning_rate": 8.550816936714193e-05, "loss": 0.0061, "step": 28620 }, { "grad_norm": 0.1408054083585739, "learning_rate": 8.549652638806841e-05, "loss": 0.0064, "step": 28630 }, { "grad_norm": 0.10329747945070267, "learning_rate": 8.548487952714812e-05, "loss": 0.0056, "step": 28640 }, { "grad_norm": 0.09715306013822556, "learning_rate": 8.547322878565478e-05, "loss": 0.0053, "step": 28650 }, { "grad_norm": 0.11156398802995682, "learning_rate": 8.546157416486245e-05, "loss": 0.0058, "step": 28660 }, { "grad_norm": 0.12322898954153061, "learning_rate": 8.54499156660457e-05, "loss": 0.0067, "step": 28670 }, { "grad_norm": 0.08883221447467804, "learning_rate": 8.543825329047947e-05, "loss": 0.0051, "step": 28680 }, { "grad_norm": 0.151343435049057, "learning_rate": 8.542658703943913e-05, "loss": 0.006, "step": 28690 }, { "grad_norm": 0.10367319732904434, "learning_rate": 8.541491691420051e-05, "loss": 0.005, "step": 28700 }, { "grad_norm": 0.1807703822851181, "learning_rate": 8.54032429160398e-05, "loss": 0.005, "step": 28710 }, { "grad_norm": 0.12298797070980072, "learning_rate": 8.539156504623369e-05, "loss": 0.0059, "step": 28720 }, { "grad_norm": 0.2808924615383148, "learning_rate": 8.537988330605923e-05, "loss": 0.0063, "step": 28730 }, { "grad_norm": 0.17705678939819336, "learning_rate": 8.536819769679393e-05, "loss": 0.0077, "step": 28740 }, { "grad_norm": 0.12705835700035095, "learning_rate": 8.53565082197157e-05, "loss": 0.0091, "step": 28750 }, { "grad_norm": 0.14584821462631226, "learning_rate": 8.534481487610289e-05, "loss": 0.0085, "step": 28760 }, { "grad_norm": 0.10658682882785797, "learning_rate": 8.533311766723428e-05, "loss": 0.0059, "step": 28770 }, { "grad_norm": 0.0908418670296669, "learning_rate": 8.532141659438901e-05, "loss": 0.0061, "step": 28780 }, { "grad_norm": 0.1213829442858696, "learning_rate": 8.530971165884675e-05, "loss": 0.0079, "step": 28790 }, { "grad_norm": 0.11905372142791748, "learning_rate": 8.529800286188752e-05, "loss": 0.0076, "step": 28800 }, { "grad_norm": 0.07517540454864502, "learning_rate": 8.528629020479175e-05, "loss": 0.0066, "step": 28810 }, { "grad_norm": 0.09397099912166595, "learning_rate": 8.527457368884033e-05, "loss": 0.0069, "step": 28820 }, { "grad_norm": 0.1313159465789795, "learning_rate": 8.526285331531458e-05, "loss": 0.0054, "step": 28830 }, { "grad_norm": 0.13563942909240723, "learning_rate": 8.525112908549621e-05, "loss": 0.0057, "step": 28840 }, { "grad_norm": 0.12345194816589355, "learning_rate": 8.523940100066735e-05, "loss": 0.0067, "step": 28850 }, { "grad_norm": 0.15866480767726898, "learning_rate": 8.52276690621106e-05, "loss": 0.0079, "step": 28860 }, { "grad_norm": 0.11472085118293762, "learning_rate": 8.521593327110889e-05, "loss": 0.0089, "step": 28870 }, { "grad_norm": 0.14798903465270996, "learning_rate": 8.520419362894569e-05, "loss": 0.0092, "step": 28880 }, { "grad_norm": 0.13993865251541138, "learning_rate": 8.51924501369048e-05, "loss": 0.0087, "step": 28890 }, { "grad_norm": 0.12842024862766266, "learning_rate": 8.518070279627047e-05, "loss": 0.0086, "step": 28900 }, { "grad_norm": 0.10566802322864532, "learning_rate": 8.516895160832737e-05, "loss": 0.0087, "step": 28910 }, { "grad_norm": 0.0924956202507019, "learning_rate": 8.515719657436061e-05, "loss": 0.0061, "step": 28920 }, { "grad_norm": 0.10458288341760635, "learning_rate": 8.514543769565568e-05, "loss": 0.0061, "step": 28930 }, { "grad_norm": 0.08624589443206787, "learning_rate": 8.513367497349853e-05, "loss": 0.0051, "step": 28940 }, { "grad_norm": 0.0886104553937912, "learning_rate": 8.51219084091755e-05, "loss": 0.0059, "step": 28950 }, { "grad_norm": 0.14734241366386414, "learning_rate": 8.511013800397338e-05, "loss": 0.007, "step": 28960 }, { "grad_norm": 0.15000911056995392, "learning_rate": 8.509836375917937e-05, "loss": 0.0064, "step": 28970 }, { "grad_norm": 0.11981173604726791, "learning_rate": 8.508658567608104e-05, "loss": 0.0078, "step": 28980 }, { "grad_norm": 0.11486387252807617, "learning_rate": 8.507480375596647e-05, "loss": 0.0073, "step": 28990 }, { "grad_norm": 0.12368018925189972, "learning_rate": 8.506301800012408e-05, "loss": 0.007, "step": 29000 }, { "grad_norm": 0.12981900572776794, "learning_rate": 8.505122840984278e-05, "loss": 0.0074, "step": 29010 }, { "grad_norm": 0.10895543545484543, "learning_rate": 8.503943498641182e-05, "loss": 0.0058, "step": 29020 }, { "grad_norm": 0.08673340082168579, "learning_rate": 8.502763773112095e-05, "loss": 0.0057, "step": 29030 }, { "grad_norm": 0.13471244275569916, "learning_rate": 8.501583664526026e-05, "loss": 0.0055, "step": 29040 }, { "grad_norm": 0.10996735841035843, "learning_rate": 8.500403173012032e-05, "loss": 0.0059, "step": 29050 }, { "grad_norm": 0.10000328719615936, "learning_rate": 8.499222298699211e-05, "loss": 0.0063, "step": 29060 }, { "grad_norm": 0.1387118548154831, "learning_rate": 8.498041041716701e-05, "loss": 0.0057, "step": 29070 }, { "grad_norm": 0.08161535114049911, "learning_rate": 8.496859402193681e-05, "loss": 0.006, "step": 29080 }, { "grad_norm": 0.10385049134492874, "learning_rate": 8.495677380259374e-05, "loss": 0.005, "step": 29090 }, { "grad_norm": 0.12766844034194946, "learning_rate": 8.494494976043045e-05, "loss": 0.0051, "step": 29100 }, { "grad_norm": 0.13749675452709198, "learning_rate": 8.493312189673998e-05, "loss": 0.0057, "step": 29110 }, { "grad_norm": 0.09401759505271912, "learning_rate": 8.492129021281584e-05, "loss": 0.0053, "step": 29120 }, { "grad_norm": 0.10245157778263092, "learning_rate": 8.490945470995188e-05, "loss": 0.0055, "step": 29130 }, { "grad_norm": 0.15372046828269958, "learning_rate": 8.489761538944247e-05, "loss": 0.0078, "step": 29140 }, { "grad_norm": 0.1232684850692749, "learning_rate": 8.48857722525823e-05, "loss": 0.008, "step": 29150 }, { "grad_norm": 0.16962333023548126, "learning_rate": 8.487392530066652e-05, "loss": 0.0083, "step": 29160 }, { "grad_norm": 0.12947657704353333, "learning_rate": 8.486207453499069e-05, "loss": 0.0074, "step": 29170 }, { "grad_norm": 0.11983385682106018, "learning_rate": 8.485021995685082e-05, "loss": 0.007, "step": 29180 }, { "grad_norm": 0.14121153950691223, "learning_rate": 8.483836156754328e-05, "loss": 0.0057, "step": 29190 }, { "grad_norm": 0.07950446009635925, "learning_rate": 8.482649936836491e-05, "loss": 0.0049, "step": 29200 }, { "grad_norm": 0.14610037207603455, "learning_rate": 8.481463336061293e-05, "loss": 0.0073, "step": 29210 }, { "grad_norm": 0.1103031188249588, "learning_rate": 8.480276354558496e-05, "loss": 0.0062, "step": 29220 }, { "grad_norm": 0.10276828706264496, "learning_rate": 8.479088992457913e-05, "loss": 0.0058, "step": 29230 }, { "grad_norm": 0.08152271807193756, "learning_rate": 8.477901249889387e-05, "loss": 0.0069, "step": 29240 }, { "grad_norm": 0.1013248860836029, "learning_rate": 8.47671312698281e-05, "loss": 0.0056, "step": 29250 }, { "grad_norm": 0.10067356377840042, "learning_rate": 8.475524623868112e-05, "loss": 0.0068, "step": 29260 }, { "grad_norm": 0.13347260653972626, "learning_rate": 8.474335740675266e-05, "loss": 0.006, "step": 29270 }, { "grad_norm": 0.30795764923095703, "learning_rate": 8.473146477534289e-05, "loss": 0.0071, "step": 29280 }, { "grad_norm": 0.11411549150943756, "learning_rate": 8.471956834575232e-05, "loss": 0.0058, "step": 29290 }, { "grad_norm": 0.1574317365884781, "learning_rate": 8.470766811928197e-05, "loss": 0.0066, "step": 29300 }, { "grad_norm": 0.20583561062812805, "learning_rate": 8.469576409723323e-05, "loss": 0.0086, "step": 29310 }, { "grad_norm": 0.13804766535758972, "learning_rate": 8.468385628090788e-05, "loss": 0.0098, "step": 29320 }, { "grad_norm": 0.1606605350971222, "learning_rate": 8.467194467160815e-05, "loss": 0.0071, "step": 29330 }, { "grad_norm": 0.1183474212884903, "learning_rate": 8.466002927063667e-05, "loss": 0.0068, "step": 29340 }, { "grad_norm": 0.09993008524179459, "learning_rate": 8.464811007929651e-05, "loss": 0.0069, "step": 29350 }, { "grad_norm": 0.14410340785980225, "learning_rate": 8.463618709889114e-05, "loss": 0.006, "step": 29360 }, { "grad_norm": 0.14807087182998657, "learning_rate": 8.462426033072442e-05, "loss": 0.0068, "step": 29370 }, { "grad_norm": 0.08994691073894501, "learning_rate": 8.461232977610061e-05, "loss": 0.0068, "step": 29380 }, { "grad_norm": 0.2772112190723419, "learning_rate": 8.46003954363245e-05, "loss": 0.0077, "step": 29390 }, { "grad_norm": 0.1823030412197113, "learning_rate": 8.458845731270115e-05, "loss": 0.0082, "step": 29400 }, { "grad_norm": 0.1351601928472519, "learning_rate": 8.45765154065361e-05, "loss": 0.0085, "step": 29410 }, { "grad_norm": 0.09879688918590546, "learning_rate": 8.456456971913532e-05, "loss": 0.0072, "step": 29420 }, { "grad_norm": 0.18131248652935028, "learning_rate": 8.455262025180517e-05, "loss": 0.0089, "step": 29430 }, { "grad_norm": 0.1545669585466385, "learning_rate": 8.454066700585242e-05, "loss": 0.0068, "step": 29440 }, { "grad_norm": 0.12146555632352829, "learning_rate": 8.452870998258423e-05, "loss": 0.0084, "step": 29450 }, { "grad_norm": 0.1621330976486206, "learning_rate": 8.451674918330825e-05, "loss": 0.0104, "step": 29460 }, { "grad_norm": 0.10573705285787582, "learning_rate": 8.450478460933246e-05, "loss": 0.0083, "step": 29470 }, { "grad_norm": 0.11124976724386215, "learning_rate": 8.449281626196532e-05, "loss": 0.0074, "step": 29480 }, { "grad_norm": 0.12290698289871216, "learning_rate": 8.448084414251564e-05, "loss": 0.0073, "step": 29490 }, { "grad_norm": 0.11238030344247818, "learning_rate": 8.446886825229271e-05, "loss": 0.0069, "step": 29500 }, { "grad_norm": 0.0886429026722908, "learning_rate": 8.445688859260615e-05, "loss": 0.0056, "step": 29510 }, { "grad_norm": 0.13106393814086914, "learning_rate": 8.444490516476606e-05, "loss": 0.0066, "step": 29520 }, { "grad_norm": 0.0734676793217659, "learning_rate": 8.443291797008293e-05, "loss": 0.005, "step": 29530 }, { "grad_norm": 0.11027305573225021, "learning_rate": 8.442092700986765e-05, "loss": 0.0049, "step": 29540 }, { "grad_norm": 0.17818699777126312, "learning_rate": 8.440893228543156e-05, "loss": 0.0055, "step": 29550 }, { "grad_norm": 0.20163793861865997, "learning_rate": 8.439693379808638e-05, "loss": 0.0055, "step": 29560 }, { "grad_norm": 0.20820021629333496, "learning_rate": 8.43849315491442e-05, "loss": 0.0069, "step": 29570 }, { "grad_norm": 0.19590964913368225, "learning_rate": 8.437292553991763e-05, "loss": 0.0092, "step": 29580 }, { "grad_norm": 0.1629038155078888, "learning_rate": 8.436091577171959e-05, "loss": 0.007, "step": 29590 }, { "grad_norm": 0.13399501144886017, "learning_rate": 8.434890224586347e-05, "loss": 0.006, "step": 29600 }, { "grad_norm": 0.1234370768070221, "learning_rate": 8.433688496366303e-05, "loss": 0.0054, "step": 29610 }, { "grad_norm": 0.11104191094636917, "learning_rate": 8.432486392643248e-05, "loss": 0.0072, "step": 29620 }, { "grad_norm": 0.1226830706000328, "learning_rate": 8.431283913548643e-05, "loss": 0.0082, "step": 29630 }, { "grad_norm": 0.08836281299591064, "learning_rate": 8.430081059213985e-05, "loss": 0.007, "step": 29640 }, { "grad_norm": 0.20148882269859314, "learning_rate": 8.428877829770823e-05, "loss": 0.0053, "step": 29650 }, { "grad_norm": 0.08657548576593399, "learning_rate": 8.427674225350735e-05, "loss": 0.0055, "step": 29660 }, { "grad_norm": 0.09889352321624756, "learning_rate": 8.426470246085347e-05, "loss": 0.0042, "step": 29670 }, { "grad_norm": 0.10462427884340286, "learning_rate": 8.425265892106324e-05, "loss": 0.0046, "step": 29680 }, { "grad_norm": 0.07861513644456863, "learning_rate": 8.424061163545374e-05, "loss": 0.0054, "step": 29690 }, { "grad_norm": 0.12607666850090027, "learning_rate": 8.422856060534243e-05, "loss": 0.0058, "step": 29700 }, { "grad_norm": 0.15334928035736084, "learning_rate": 8.421650583204718e-05, "loss": 0.0058, "step": 29710 }, { "grad_norm": 0.12791480123996735, "learning_rate": 8.420444731688633e-05, "loss": 0.005, "step": 29720 }, { "grad_norm": 0.09758685529232025, "learning_rate": 8.419238506117852e-05, "loss": 0.0049, "step": 29730 }, { "grad_norm": 0.09575751423835754, "learning_rate": 8.418031906624289e-05, "loss": 0.0046, "step": 29740 }, { "grad_norm": 0.1261870265007019, "learning_rate": 8.416824933339898e-05, "loss": 0.0053, "step": 29750 }, { "grad_norm": 0.1160050630569458, "learning_rate": 8.415617586396667e-05, "loss": 0.0053, "step": 29760 }, { "grad_norm": 0.11046940088272095, "learning_rate": 8.414409865926632e-05, "loss": 0.0054, "step": 29770 }, { "grad_norm": 0.09804551303386688, "learning_rate": 8.413201772061867e-05, "loss": 0.0063, "step": 29780 }, { "grad_norm": 0.14118735492229462, "learning_rate": 8.411993304934488e-05, "loss": 0.0057, "step": 29790 }, { "grad_norm": 0.12677223980426788, "learning_rate": 8.410784464676654e-05, "loss": 0.006, "step": 29800 }, { "grad_norm": 0.10857904702425003, "learning_rate": 8.409575251420556e-05, "loss": 0.0072, "step": 29810 }, { "grad_norm": 0.13339492678642273, "learning_rate": 8.408365665298435e-05, "loss": 0.006, "step": 29820 }, { "grad_norm": 0.11234408617019653, "learning_rate": 8.40715570644257e-05, "loss": 0.0064, "step": 29830 }, { "grad_norm": 0.09493166208267212, "learning_rate": 8.40594537498528e-05, "loss": 0.0064, "step": 29840 }, { "grad_norm": 0.10171282291412354, "learning_rate": 8.404734671058924e-05, "loss": 0.0067, "step": 29850 }, { "grad_norm": 0.11844778060913086, "learning_rate": 8.403523594795902e-05, "loss": 0.0065, "step": 29860 }, { "grad_norm": 0.10507231205701828, "learning_rate": 8.402312146328659e-05, "loss": 0.0055, "step": 29870 }, { "grad_norm": 0.09448294341564178, "learning_rate": 8.401100325789675e-05, "loss": 0.0054, "step": 29880 }, { "grad_norm": 0.09593979269266129, "learning_rate": 8.399888133311472e-05, "loss": 0.0058, "step": 29890 }, { "grad_norm": 0.08097705245018005, "learning_rate": 8.398675569026613e-05, "loss": 0.0053, "step": 29900 }, { "grad_norm": 0.10148512572050095, "learning_rate": 8.397462633067705e-05, "loss": 0.0053, "step": 29910 }, { "grad_norm": 0.10838013142347336, "learning_rate": 8.396249325567392e-05, "loss": 0.0064, "step": 29920 }, { "grad_norm": 0.1426846832036972, "learning_rate": 8.395035646658357e-05, "loss": 0.0071, "step": 29930 }, { "grad_norm": 0.10062592476606369, "learning_rate": 8.39382159647333e-05, "loss": 0.0059, "step": 29940 }, { "grad_norm": 0.1049879863858223, "learning_rate": 8.392607175145075e-05, "loss": 0.0074, "step": 29950 }, { "grad_norm": 0.08533309400081635, "learning_rate": 8.3913923828064e-05, "loss": 0.0055, "step": 29960 }, { "grad_norm": 0.11425768584012985, "learning_rate": 8.390177219590152e-05, "loss": 0.0059, "step": 29970 }, { "grad_norm": 0.21483905613422394, "learning_rate": 8.388961685629222e-05, "loss": 0.007, "step": 29980 }, { "grad_norm": 0.0856069028377533, "learning_rate": 8.387745781056536e-05, "loss": 0.0058, "step": 29990 }, { "grad_norm": 0.11148544400930405, "learning_rate": 8.386529506005065e-05, "loss": 0.0063, "step": 30000 }, { "grad_norm": 0.10428725928068161, "learning_rate": 8.38531286060782e-05, "loss": 0.008, "step": 30010 }, { "grad_norm": 0.10683482140302658, "learning_rate": 8.384095844997849e-05, "loss": 0.0059, "step": 30020 }, { "grad_norm": 0.13812558352947235, "learning_rate": 8.382878459308245e-05, "loss": 0.0054, "step": 30030 }, { "grad_norm": 0.11499733477830887, "learning_rate": 8.381660703672138e-05, "loss": 0.0084, "step": 30040 }, { "grad_norm": 0.15828688442707062, "learning_rate": 8.380442578222702e-05, "loss": 0.0074, "step": 30050 }, { "grad_norm": 0.19510535895824432, "learning_rate": 8.379224083093146e-05, "loss": 0.0079, "step": 30060 }, { "grad_norm": 0.2390129566192627, "learning_rate": 8.378005218416727e-05, "loss": 0.0132, "step": 30070 }, { "grad_norm": 0.13947167992591858, "learning_rate": 8.376785984326735e-05, "loss": 0.008, "step": 30080 }, { "grad_norm": 0.18944862484931946, "learning_rate": 8.375566380956506e-05, "loss": 0.0097, "step": 30090 }, { "grad_norm": 0.12184024602174759, "learning_rate": 8.374346408439411e-05, "loss": 0.0111, "step": 30100 }, { "grad_norm": 0.10965024679899216, "learning_rate": 8.373126066908868e-05, "loss": 0.0076, "step": 30110 }, { "grad_norm": 0.14425520598888397, "learning_rate": 8.371905356498326e-05, "loss": 0.0061, "step": 30120 }, { "grad_norm": 0.1463247388601303, "learning_rate": 8.370684277341288e-05, "loss": 0.0065, "step": 30130 }, { "grad_norm": 0.1774134784936905, "learning_rate": 8.369462829571282e-05, "loss": 0.0064, "step": 30140 }, { "grad_norm": 0.18077321350574493, "learning_rate": 8.36824101332189e-05, "loss": 0.0096, "step": 30150 }, { "grad_norm": 0.07986753433942795, "learning_rate": 8.367018828726721e-05, "loss": 0.008, "step": 30160 }, { "grad_norm": 0.10904278606176376, "learning_rate": 8.365796275919438e-05, "loss": 0.0075, "step": 30170 }, { "grad_norm": 0.09965783357620239, "learning_rate": 8.364573355033734e-05, "loss": 0.0066, "step": 30180 }, { "grad_norm": 0.13283969461917877, "learning_rate": 8.363350066203346e-05, "loss": 0.0057, "step": 30190 }, { "grad_norm": 0.12579204142093658, "learning_rate": 8.362126409562053e-05, "loss": 0.0061, "step": 30200 }, { "grad_norm": 0.12750238180160522, "learning_rate": 8.360902385243667e-05, "loss": 0.0072, "step": 30210 }, { "grad_norm": 0.16047771275043488, "learning_rate": 8.359677993382052e-05, "loss": 0.0074, "step": 30220 }, { "grad_norm": 0.1294582188129425, "learning_rate": 8.358453234111103e-05, "loss": 0.0077, "step": 30230 }, { "grad_norm": 0.30884596705436707, "learning_rate": 8.357228107564756e-05, "loss": 0.0092, "step": 30240 }, { "grad_norm": 0.16641810536384583, "learning_rate": 8.356002613876993e-05, "loss": 0.0132, "step": 30250 }, { "grad_norm": 0.10969198495149612, "learning_rate": 8.35477675318183e-05, "loss": 0.0098, "step": 30260 }, { "grad_norm": 0.15227749943733215, "learning_rate": 8.353550525613323e-05, "loss": 0.0096, "step": 30270 }, { "grad_norm": 0.12661083042621613, "learning_rate": 8.352323931305572e-05, "loss": 0.0072, "step": 30280 }, { "grad_norm": 0.097429558634758, "learning_rate": 8.351096970392717e-05, "loss": 0.006, "step": 30290 }, { "grad_norm": 0.14483200013637543, "learning_rate": 8.349869643008937e-05, "loss": 0.0074, "step": 30300 }, { "grad_norm": 0.19030702114105225, "learning_rate": 8.348641949288449e-05, "loss": 0.0071, "step": 30310 }, { "grad_norm": 0.09668540954589844, "learning_rate": 8.34741388936551e-05, "loss": 0.0068, "step": 30320 }, { "grad_norm": 0.11973526328802109, "learning_rate": 8.346185463374423e-05, "loss": 0.009, "step": 30330 }, { "grad_norm": 0.15222926437854767, "learning_rate": 8.344956671449524e-05, "loss": 0.0076, "step": 30340 }, { "grad_norm": 0.12173286080360413, "learning_rate": 8.343727513725192e-05, "loss": 0.0082, "step": 30350 }, { "grad_norm": 0.09857016801834106, "learning_rate": 8.342497990335847e-05, "loss": 0.006, "step": 30360 }, { "grad_norm": 0.10778288543224335, "learning_rate": 8.341268101415946e-05, "loss": 0.006, "step": 30370 }, { "grad_norm": 0.10941637307405472, "learning_rate": 8.34003784709999e-05, "loss": 0.006, "step": 30380 }, { "grad_norm": 0.08886770159006119, "learning_rate": 8.338807227522516e-05, "loss": 0.0057, "step": 30390 }, { "grad_norm": 0.14666520059108734, "learning_rate": 8.337576242818103e-05, "loss": 0.0091, "step": 30400 }, { "grad_norm": 0.09753324091434479, "learning_rate": 8.336344893121372e-05, "loss": 0.0105, "step": 30410 }, { "grad_norm": 0.2876099646091461, "learning_rate": 8.335113178566977e-05, "loss": 0.008, "step": 30420 }, { "grad_norm": 0.09609691053628922, "learning_rate": 8.33388109928962e-05, "loss": 0.0059, "step": 30430 }, { "grad_norm": 0.10082603991031647, "learning_rate": 8.33264865542404e-05, "loss": 0.0048, "step": 30440 }, { "grad_norm": 0.1831136792898178, "learning_rate": 8.331415847105013e-05, "loss": 0.0062, "step": 30450 }, { "grad_norm": 0.09717226028442383, "learning_rate": 8.330182674467357e-05, "loss": 0.0055, "step": 30460 }, { "grad_norm": 0.11856424063444138, "learning_rate": 8.32894913764593e-05, "loss": 0.0057, "step": 30470 }, { "grad_norm": 0.12440420687198639, "learning_rate": 8.327715236775633e-05, "loss": 0.0062, "step": 30480 }, { "grad_norm": 0.10686112940311432, "learning_rate": 8.326480971991398e-05, "loss": 0.0087, "step": 30490 }, { "grad_norm": 0.11574187874794006, "learning_rate": 8.325246343428206e-05, "loss": 0.0095, "step": 30500 }, { "grad_norm": 0.14105619490146637, "learning_rate": 8.324011351221072e-05, "loss": 0.0085, "step": 30510 }, { "grad_norm": 0.0984809473156929, "learning_rate": 8.322775995505057e-05, "loss": 0.0088, "step": 30520 }, { "grad_norm": 0.11862905323505402, "learning_rate": 8.321540276415254e-05, "loss": 0.0069, "step": 30530 }, { "grad_norm": 0.09923037886619568, "learning_rate": 8.320304194086798e-05, "loss": 0.0072, "step": 30540 }, { "grad_norm": 0.14947497844696045, "learning_rate": 8.31906774865487e-05, "loss": 0.007, "step": 30550 }, { "grad_norm": 0.10697685182094574, "learning_rate": 8.317830940254682e-05, "loss": 0.0088, "step": 30560 }, { "grad_norm": 0.12505677342414856, "learning_rate": 8.316593769021491e-05, "loss": 0.0076, "step": 30570 }, { "grad_norm": 0.09570979326963425, "learning_rate": 8.315356235090592e-05, "loss": 0.0077, "step": 30580 }, { "grad_norm": 0.11917740851640701, "learning_rate": 8.314118338597319e-05, "loss": 0.0071, "step": 30590 }, { "grad_norm": 0.13892708718776703, "learning_rate": 8.312880079677048e-05, "loss": 0.0067, "step": 30600 }, { "grad_norm": 0.1164887547492981, "learning_rate": 8.311641458465191e-05, "loss": 0.006, "step": 30610 }, { "grad_norm": 0.11677806824445724, "learning_rate": 8.310402475097205e-05, "loss": 0.0061, "step": 30620 }, { "grad_norm": 0.10960231721401215, "learning_rate": 8.309163129708581e-05, "loss": 0.0057, "step": 30630 }, { "grad_norm": 0.11482342332601547, "learning_rate": 8.307923422434852e-05, "loss": 0.0051, "step": 30640 }, { "grad_norm": 0.09373831003904343, "learning_rate": 8.30668335341159e-05, "loss": 0.0047, "step": 30650 }, { "grad_norm": 0.11050964146852493, "learning_rate": 8.305442922774408e-05, "loss": 0.006, "step": 30660 }, { "grad_norm": 0.1179034486413002, "learning_rate": 8.304202130658959e-05, "loss": 0.0073, "step": 30670 }, { "grad_norm": 0.09854702651500702, "learning_rate": 8.302960977200931e-05, "loss": 0.0068, "step": 30680 }, { "grad_norm": 0.13775528967380524, "learning_rate": 8.301719462536058e-05, "loss": 0.006, "step": 30690 }, { "grad_norm": 0.12474276125431061, "learning_rate": 8.300477586800108e-05, "loss": 0.0052, "step": 30700 }, { "grad_norm": 0.08825714886188507, "learning_rate": 8.299235350128892e-05, "loss": 0.0056, "step": 30710 }, { "grad_norm": 0.08350599557161331, "learning_rate": 8.297992752658258e-05, "loss": 0.0045, "step": 30720 }, { "grad_norm": 0.09224376827478409, "learning_rate": 8.296749794524094e-05, "loss": 0.0073, "step": 30730 }, { "grad_norm": 0.1435175985097885, "learning_rate": 8.295506475862332e-05, "loss": 0.0059, "step": 30740 }, { "grad_norm": 0.12204396724700928, "learning_rate": 8.294262796808933e-05, "loss": 0.0065, "step": 30750 }, { "grad_norm": 0.1108747199177742, "learning_rate": 8.293018757499909e-05, "loss": 0.0072, "step": 30760 }, { "grad_norm": 0.11265142261981964, "learning_rate": 8.291774358071305e-05, "loss": 0.0055, "step": 30770 }, { "grad_norm": 0.1245725080370903, "learning_rate": 8.290529598659205e-05, "loss": 0.0066, "step": 30780 }, { "grad_norm": 0.1095229983329773, "learning_rate": 8.289284479399738e-05, "loss": 0.007, "step": 30790 }, { "grad_norm": 0.1433759182691574, "learning_rate": 8.288039000429064e-05, "loss": 0.0062, "step": 30800 }, { "grad_norm": 0.08961545675992966, "learning_rate": 8.286793161883388e-05, "loss": 0.0063, "step": 30810 }, { "grad_norm": 0.11645283550024033, "learning_rate": 8.285546963898954e-05, "loss": 0.006, "step": 30820 }, { "grad_norm": 0.18975700438022614, "learning_rate": 8.284300406612044e-05, "loss": 0.0056, "step": 30830 }, { "grad_norm": 0.10066787153482437, "learning_rate": 8.283053490158978e-05, "loss": 0.0049, "step": 30840 }, { "grad_norm": 0.4037944972515106, "learning_rate": 8.28180621467612e-05, "loss": 0.0056, "step": 30850 }, { "grad_norm": 0.2186235785484314, "learning_rate": 8.280558580299867e-05, "loss": 0.0108, "step": 30860 }, { "grad_norm": 0.18912562727928162, "learning_rate": 8.279310587166661e-05, "loss": 0.0101, "step": 30870 }, { "grad_norm": 0.14236047863960266, "learning_rate": 8.278062235412978e-05, "loss": 0.0089, "step": 30880 }, { "grad_norm": 0.1672884076833725, "learning_rate": 8.276813525175339e-05, "loss": 0.0115, "step": 30890 }, { "grad_norm": 0.16318824887275696, "learning_rate": 8.2755644565903e-05, "loss": 0.0102, "step": 30900 }, { "grad_norm": 0.10094167292118073, "learning_rate": 8.274315029794454e-05, "loss": 0.0065, "step": 30910 }, { "grad_norm": 0.1051848903298378, "learning_rate": 8.273065244924443e-05, "loss": 0.0059, "step": 30920 }, { "grad_norm": 0.11982036381959915, "learning_rate": 8.271815102116936e-05, "loss": 0.0073, "step": 30930 }, { "grad_norm": 0.1664225310087204, "learning_rate": 8.270564601508648e-05, "loss": 0.0063, "step": 30940 }, { "grad_norm": 0.1284128874540329, "learning_rate": 8.269313743236333e-05, "loss": 0.0079, "step": 30950 }, { "grad_norm": 0.13212665915489197, "learning_rate": 8.268062527436783e-05, "loss": 0.0088, "step": 30960 }, { "grad_norm": 0.11394623667001724, "learning_rate": 8.266810954246828e-05, "loss": 0.0091, "step": 30970 }, { "grad_norm": 0.11553109437227249, "learning_rate": 8.265559023803338e-05, "loss": 0.0082, "step": 30980 }, { "grad_norm": 0.1415831446647644, "learning_rate": 8.264306736243223e-05, "loss": 0.0083, "step": 30990 }, { "grad_norm": 0.07689402252435684, "learning_rate": 8.263054091703432e-05, "loss": 0.0068, "step": 31000 }, { "grad_norm": 0.1390269696712494, "learning_rate": 8.26180109032095e-05, "loss": 0.0077, "step": 31010 }, { "grad_norm": 0.14944195747375488, "learning_rate": 8.260547732232807e-05, "loss": 0.0063, "step": 31020 }, { "grad_norm": 0.09269152581691742, "learning_rate": 8.259294017576063e-05, "loss": 0.0068, "step": 31030 }, { "grad_norm": 0.12194294482469559, "learning_rate": 8.258039946487828e-05, "loss": 0.0064, "step": 31040 }, { "grad_norm": 0.1033538356423378, "learning_rate": 8.256785519105241e-05, "loss": 0.0072, "step": 31050 }, { "grad_norm": 0.09565317630767822, "learning_rate": 8.255530735565488e-05, "loss": 0.0075, "step": 31060 }, { "grad_norm": 0.13547874987125397, "learning_rate": 8.254275596005787e-05, "loss": 0.0062, "step": 31070 }, { "grad_norm": 0.14828170835971832, "learning_rate": 8.2530201005634e-05, "loss": 0.0096, "step": 31080 }, { "grad_norm": 0.12198140472173691, "learning_rate": 8.251764249375626e-05, "loss": 0.0069, "step": 31090 }, { "grad_norm": 0.10637401789426804, "learning_rate": 8.250508042579803e-05, "loss": 0.0082, "step": 31100 }, { "grad_norm": 0.08675647526979446, "learning_rate": 8.249251480313307e-05, "loss": 0.0061, "step": 31110 }, { "grad_norm": 0.12439614534378052, "learning_rate": 8.247994562713555e-05, "loss": 0.007, "step": 31120 }, { "grad_norm": 0.125233992934227, "learning_rate": 8.246737289917998e-05, "loss": 0.0052, "step": 31130 }, { "grad_norm": 0.15643012523651123, "learning_rate": 8.245479662064135e-05, "loss": 0.0053, "step": 31140 }, { "grad_norm": 0.09180709719657898, "learning_rate": 8.244221679289496e-05, "loss": 0.0054, "step": 31150 }, { "grad_norm": 0.097713403403759, "learning_rate": 8.242963341731652e-05, "loss": 0.0053, "step": 31160 }, { "grad_norm": 0.09386361390352249, "learning_rate": 8.24170464952821e-05, "loss": 0.0047, "step": 31170 }, { "grad_norm": 0.09435176104307175, "learning_rate": 8.240445602816824e-05, "loss": 0.0054, "step": 31180 }, { "grad_norm": 0.09769034385681152, "learning_rate": 8.239186201735179e-05, "loss": 0.0061, "step": 31190 }, { "grad_norm": 0.0957634225487709, "learning_rate": 8.237926446420998e-05, "loss": 0.005, "step": 31200 }, { "grad_norm": 0.07205114513635635, "learning_rate": 8.236666337012052e-05, "loss": 0.0052, "step": 31210 }, { "grad_norm": 0.08261866122484207, "learning_rate": 8.23540587364614e-05, "loss": 0.0053, "step": 31220 }, { "grad_norm": 0.10767904669046402, "learning_rate": 8.234145056461107e-05, "loss": 0.0042, "step": 31230 }, { "grad_norm": 0.08964364975690842, "learning_rate": 8.232883885594832e-05, "loss": 0.004, "step": 31240 }, { "grad_norm": 0.09504608064889908, "learning_rate": 8.231622361185236e-05, "loss": 0.0056, "step": 31250 }, { "grad_norm": 0.09195134043693542, "learning_rate": 8.230360483370278e-05, "loss": 0.0044, "step": 31260 }, { "grad_norm": 0.09898301213979721, "learning_rate": 8.229098252287953e-05, "loss": 0.0052, "step": 31270 }, { "grad_norm": 0.08368878811597824, "learning_rate": 8.2278356680763e-05, "loss": 0.0048, "step": 31280 }, { "grad_norm": 0.1491839438676834, "learning_rate": 8.22657273087339e-05, "loss": 0.0061, "step": 31290 }, { "grad_norm": 0.11245984584093094, "learning_rate": 8.225309440817336e-05, "loss": 0.0086, "step": 31300 }, { "grad_norm": 0.1268722116947174, "learning_rate": 8.224045798046293e-05, "loss": 0.0077, "step": 31310 }, { "grad_norm": 0.10910651087760925, "learning_rate": 8.22278180269845e-05, "loss": 0.0063, "step": 31320 }, { "grad_norm": 0.14498066902160645, "learning_rate": 8.221517454912031e-05, "loss": 0.0074, "step": 31330 }, { "grad_norm": 0.10608743876218796, "learning_rate": 8.220252754825308e-05, "loss": 0.0067, "step": 31340 }, { "grad_norm": 0.10048876702785492, "learning_rate": 8.218987702576586e-05, "loss": 0.0053, "step": 31350 }, { "grad_norm": 0.12765657901763916, "learning_rate": 8.217722298304207e-05, "loss": 0.0049, "step": 31360 }, { "grad_norm": 0.09783033281564713, "learning_rate": 8.216456542146557e-05, "loss": 0.006, "step": 31370 }, { "grad_norm": 0.17405377328395844, "learning_rate": 8.215190434242055e-05, "loss": 0.0065, "step": 31380 }, { "grad_norm": 0.20241287350654602, "learning_rate": 8.213923974729161e-05, "loss": 0.0088, "step": 31390 }, { "grad_norm": 0.15437524020671844, "learning_rate": 8.212657163746373e-05, "loss": 0.0069, "step": 31400 }, { "grad_norm": 0.12518776953220367, "learning_rate": 8.211390001432227e-05, "loss": 0.0076, "step": 31410 }, { "grad_norm": 0.4558642506599426, "learning_rate": 8.210122487925297e-05, "loss": 0.0071, "step": 31420 }, { "grad_norm": 0.1255113184452057, "learning_rate": 8.208854623364202e-05, "loss": 0.0119, "step": 31430 }, { "grad_norm": 0.15404941141605377, "learning_rate": 8.207586407887589e-05, "loss": 0.0077, "step": 31440 }, { "grad_norm": 0.16999487578868866, "learning_rate": 8.206317841634148e-05, "loss": 0.0072, "step": 31450 }, { "grad_norm": 0.0935153216123581, "learning_rate": 8.205048924742609e-05, "loss": 0.006, "step": 31460 }, { "grad_norm": 0.1001201942563057, "learning_rate": 8.203779657351738e-05, "loss": 0.0061, "step": 31470 }, { "grad_norm": 0.07979182153940201, "learning_rate": 8.20251003960034e-05, "loss": 0.0056, "step": 31480 }, { "grad_norm": 0.1195097267627716, "learning_rate": 8.201240071627258e-05, "loss": 0.0055, "step": 31490 }, { "grad_norm": 0.11388833820819855, "learning_rate": 8.199969753571377e-05, "loss": 0.0057, "step": 31500 }, { "grad_norm": 0.13319352269172668, "learning_rate": 8.198699085571615e-05, "loss": 0.0055, "step": 31510 }, { "grad_norm": 0.21122880280017853, "learning_rate": 8.197428067766928e-05, "loss": 0.0054, "step": 31520 }, { "grad_norm": 0.1365898698568344, "learning_rate": 8.196156700296316e-05, "loss": 0.006, "step": 31530 }, { "grad_norm": 0.13047121465206146, "learning_rate": 8.194884983298814e-05, "loss": 0.0058, "step": 31540 }, { "grad_norm": 0.11046545952558517, "learning_rate": 8.193612916913491e-05, "loss": 0.006, "step": 31550 }, { "grad_norm": 0.11795777827501297, "learning_rate": 8.192340501279463e-05, "loss": 0.0052, "step": 31560 }, { "grad_norm": 0.11837919056415558, "learning_rate": 8.191067736535876e-05, "loss": 0.0054, "step": 31570 }, { "grad_norm": 0.10034732520580292, "learning_rate": 8.18979462282192e-05, "loss": 0.0046, "step": 31580 }, { "grad_norm": 0.0930132195353508, "learning_rate": 8.188521160276819e-05, "loss": 0.0052, "step": 31590 }, { "grad_norm": 0.0853157714009285, "learning_rate": 8.187247349039837e-05, "loss": 0.005, "step": 31600 }, { "grad_norm": 0.09204021841287613, "learning_rate": 8.185973189250278e-05, "loss": 0.0052, "step": 31610 }, { "grad_norm": 0.11464277654886246, "learning_rate": 8.184698681047482e-05, "loss": 0.0051, "step": 31620 }, { "grad_norm": 0.1005404144525528, "learning_rate": 8.183423824570827e-05, "loss": 0.0057, "step": 31630 }, { "grad_norm": 0.10840971022844315, "learning_rate": 8.182148619959725e-05, "loss": 0.004, "step": 31640 }, { "grad_norm": 0.10811559855937958, "learning_rate": 8.180873067353636e-05, "loss": 0.0061, "step": 31650 }, { "grad_norm": 0.12526369094848633, "learning_rate": 8.179597166892052e-05, "loss": 0.0056, "step": 31660 }, { "grad_norm": 0.11249781399965286, "learning_rate": 8.178320918714501e-05, "loss": 0.0057, "step": 31670 }, { "grad_norm": 0.1244233027100563, "learning_rate": 8.177044322960554e-05, "loss": 0.0071, "step": 31680 }, { "grad_norm": 0.1691032201051712, "learning_rate": 8.175767379769816e-05, "loss": 0.006, "step": 31690 }, { "grad_norm": 0.1797431856393814, "learning_rate": 8.174490089281932e-05, "loss": 0.0063, "step": 31700 }, { "grad_norm": 0.13630492985248566, "learning_rate": 8.173212451636584e-05, "loss": 0.0083, "step": 31710 }, { "grad_norm": 0.10184421390295029, "learning_rate": 8.171934466973493e-05, "loss": 0.0066, "step": 31720 }, { "grad_norm": 0.07482516765594482, "learning_rate": 8.170656135432418e-05, "loss": 0.0059, "step": 31730 }, { "grad_norm": 0.1299857497215271, "learning_rate": 8.169377457153155e-05, "loss": 0.0084, "step": 31740 }, { "grad_norm": 0.10807350277900696, "learning_rate": 8.168098432275539e-05, "loss": 0.0077, "step": 31750 }, { "grad_norm": 0.1781490445137024, "learning_rate": 8.166819060939442e-05, "loss": 0.0055, "step": 31760 }, { "grad_norm": 0.19735394418239594, "learning_rate": 8.165539343284772e-05, "loss": 0.0069, "step": 31770 }, { "grad_norm": 0.10254362970590591, "learning_rate": 8.16425927945148e-05, "loss": 0.0075, "step": 31780 }, { "grad_norm": 0.10772862285375595, "learning_rate": 8.162978869579551e-05, "loss": 0.0075, "step": 31790 }, { "grad_norm": 0.0990593433380127, "learning_rate": 8.161698113809007e-05, "loss": 0.0064, "step": 31800 }, { "grad_norm": 0.13141469657421112, "learning_rate": 8.160417012279911e-05, "loss": 0.0066, "step": 31810 }, { "grad_norm": 0.1428675651550293, "learning_rate": 8.159135565132363e-05, "loss": 0.0083, "step": 31820 }, { "grad_norm": 0.12536835670471191, "learning_rate": 8.157853772506498e-05, "loss": 0.0083, "step": 31830 }, { "grad_norm": 0.1204228475689888, "learning_rate": 8.156571634542494e-05, "loss": 0.0071, "step": 31840 }, { "grad_norm": 0.13364624977111816, "learning_rate": 8.15528915138056e-05, "loss": 0.0067, "step": 31850 }, { "grad_norm": 0.16443590819835663, "learning_rate": 8.154006323160949e-05, "loss": 0.0073, "step": 31860 }, { "grad_norm": 0.10780013352632523, "learning_rate": 8.152723150023949e-05, "loss": 0.0067, "step": 31870 }, { "grad_norm": 0.16210432350635529, "learning_rate": 8.151439632109886e-05, "loss": 0.0065, "step": 31880 }, { "grad_norm": 0.10554782301187515, "learning_rate": 8.150155769559122e-05, "loss": 0.0069, "step": 31890 }, { "grad_norm": 0.09659197181463242, "learning_rate": 8.148871562512058e-05, "loss": 0.0055, "step": 31900 }, { "grad_norm": 0.11751417815685272, "learning_rate": 8.147587011109136e-05, "loss": 0.005, "step": 31910 }, { "grad_norm": 0.09926442801952362, "learning_rate": 8.14630211549083e-05, "loss": 0.0055, "step": 31920 }, { "grad_norm": 0.09186629205942154, "learning_rate": 8.145016875797655e-05, "loss": 0.0048, "step": 31930 }, { "grad_norm": 0.12378194183111191, "learning_rate": 8.143731292170164e-05, "loss": 0.0062, "step": 31940 }, { "grad_norm": 0.14700178802013397, "learning_rate": 8.142445364748944e-05, "loss": 0.0087, "step": 31950 }, { "grad_norm": 0.11554937809705734, "learning_rate": 8.141159093674624e-05, "loss": 0.0054, "step": 31960 }, { "grad_norm": 0.09697619825601578, "learning_rate": 8.139872479087869e-05, "loss": 0.0068, "step": 31970 }, { "grad_norm": 0.09061460196971893, "learning_rate": 8.13858552112938e-05, "loss": 0.006, "step": 31980 }, { "grad_norm": 0.12851710617542267, "learning_rate": 8.137298219939895e-05, "loss": 0.0094, "step": 31990 }, { "grad_norm": 0.1661318838596344, "learning_rate": 8.136010575660196e-05, "loss": 0.0085, "step": 32000 }, { "grad_norm": 0.3206389844417572, "learning_rate": 8.134722588431095e-05, "loss": 0.0112, "step": 32010 }, { "grad_norm": 0.112599678337574, "learning_rate": 8.133434258393444e-05, "loss": 0.0098, "step": 32020 }, { "grad_norm": 0.16411882638931274, "learning_rate": 8.132145585688134e-05, "loss": 0.0091, "step": 32030 }, { "grad_norm": 0.0933689996600151, "learning_rate": 8.130856570456093e-05, "loss": 0.0078, "step": 32040 }, { "grad_norm": 0.1322897970676422, "learning_rate": 8.129567212838283e-05, "loss": 0.0063, "step": 32050 }, { "grad_norm": 0.10593534260988235, "learning_rate": 8.128277512975708e-05, "loss": 0.0057, "step": 32060 }, { "grad_norm": 0.10651609301567078, "learning_rate": 8.126987471009408e-05, "loss": 0.0055, "step": 32070 }, { "grad_norm": 0.10384207963943481, "learning_rate": 8.125697087080459e-05, "loss": 0.0053, "step": 32080 }, { "grad_norm": 0.08658433705568314, "learning_rate": 8.124406361329976e-05, "loss": 0.0056, "step": 32090 }, { "grad_norm": 0.12300349771976471, "learning_rate": 8.123115293899111e-05, "loss": 0.0056, "step": 32100 }, { "grad_norm": 0.10512680560350418, "learning_rate": 8.121823884929055e-05, "loss": 0.0058, "step": 32110 }, { "grad_norm": 0.08455516397953033, "learning_rate": 8.12053213456103e-05, "loss": 0.0053, "step": 32120 }, { "grad_norm": 0.09092355519533157, "learning_rate": 8.119240042936303e-05, "loss": 0.0068, "step": 32130 }, { "grad_norm": 0.08023274689912796, "learning_rate": 8.117947610196175e-05, "loss": 0.0052, "step": 32140 }, { "grad_norm": 0.10712256282567978, "learning_rate": 8.116654836481982e-05, "loss": 0.0069, "step": 32150 }, { "grad_norm": 0.0939260721206665, "learning_rate": 8.115361721935106e-05, "loss": 0.0061, "step": 32160 }, { "grad_norm": 0.10832847654819489, "learning_rate": 8.114068266696953e-05, "loss": 0.005, "step": 32170 }, { "grad_norm": 0.19433100521564484, "learning_rate": 8.112774470908978e-05, "loss": 0.0071, "step": 32180 }, { "grad_norm": 0.13459551334381104, "learning_rate": 8.111480334712665e-05, "loss": 0.0077, "step": 32190 }, { "grad_norm": 0.09962224215269089, "learning_rate": 8.110185858249542e-05, "loss": 0.0065, "step": 32200 }, { "grad_norm": 0.12003561854362488, "learning_rate": 8.108891041661168e-05, "loss": 0.0057, "step": 32210 }, { "grad_norm": 0.16029709577560425, "learning_rate": 8.107595885089146e-05, "loss": 0.0055, "step": 32220 }, { "grad_norm": 0.13887643814086914, "learning_rate": 8.106300388675108e-05, "loss": 0.0059, "step": 32230 }, { "grad_norm": 0.09235529601573944, "learning_rate": 8.105004552560731e-05, "loss": 0.0049, "step": 32240 }, { "grad_norm": 0.11365728080272675, "learning_rate": 8.103708376887724e-05, "loss": 0.005, "step": 32250 }, { "grad_norm": 0.11368045210838318, "learning_rate": 8.102411861797836e-05, "loss": 0.0055, "step": 32260 }, { "grad_norm": 0.0870947614312172, "learning_rate": 8.101115007432851e-05, "loss": 0.0046, "step": 32270 }, { "grad_norm": 0.18098074197769165, "learning_rate": 8.09981781393459e-05, "loss": 0.0055, "step": 32280 }, { "grad_norm": 0.1094072237610817, "learning_rate": 8.098520281444915e-05, "loss": 0.0058, "step": 32290 }, { "grad_norm": 0.09557155519723892, "learning_rate": 8.09722241010572e-05, "loss": 0.0063, "step": 32300 }, { "grad_norm": 0.10055480897426605, "learning_rate": 8.095924200058939e-05, "loss": 0.0059, "step": 32310 }, { "grad_norm": 0.18081623315811157, "learning_rate": 8.094625651446541e-05, "loss": 0.006, "step": 32320 }, { "grad_norm": 0.0820540189743042, "learning_rate": 8.093326764410536e-05, "loss": 0.0076, "step": 32330 }, { "grad_norm": 0.13868562877178192, "learning_rate": 8.092027539092966e-05, "loss": 0.0085, "step": 32340 }, { "grad_norm": 0.12502321600914001, "learning_rate": 8.090727975635913e-05, "loss": 0.0079, "step": 32350 }, { "grad_norm": 0.24453771114349365, "learning_rate": 8.089428074181497e-05, "loss": 0.0072, "step": 32360 }, { "grad_norm": 0.07065845280885696, "learning_rate": 8.088127834871871e-05, "loss": 0.0062, "step": 32370 }, { "grad_norm": 0.1190551221370697, "learning_rate": 8.086827257849226e-05, "loss": 0.007, "step": 32380 }, { "grad_norm": 0.10522288829088211, "learning_rate": 8.085526343255795e-05, "loss": 0.0061, "step": 32390 }, { "grad_norm": 0.12789414823055267, "learning_rate": 8.084225091233842e-05, "loss": 0.0062, "step": 32400 }, { "grad_norm": 0.109963558614254, "learning_rate": 8.082923501925668e-05, "loss": 0.0069, "step": 32410 }, { "grad_norm": 0.1618293672800064, "learning_rate": 8.081621575473617e-05, "loss": 0.0101, "step": 32420 }, { "grad_norm": 0.13086818158626556, "learning_rate": 8.080319312020064e-05, "loss": 0.0079, "step": 32430 }, { "grad_norm": 0.09622684866189957, "learning_rate": 8.079016711707421e-05, "loss": 0.0065, "step": 32440 }, { "grad_norm": 0.08426666259765625, "learning_rate": 8.077713774678139e-05, "loss": 0.0049, "step": 32450 }, { "grad_norm": 0.23628002405166626, "learning_rate": 8.076410501074707e-05, "loss": 0.0053, "step": 32460 }, { "grad_norm": 0.21942776441574097, "learning_rate": 8.075106891039647e-05, "loss": 0.0087, "step": 32470 }, { "grad_norm": 0.13300397992134094, "learning_rate": 8.073802944715523e-05, "loss": 0.0078, "step": 32480 }, { "grad_norm": 0.12514211237430573, "learning_rate": 8.072498662244929e-05, "loss": 0.0085, "step": 32490 }, { "grad_norm": 0.16717593371868134, "learning_rate": 8.0711940437705e-05, "loss": 0.0076, "step": 32500 }, { "grad_norm": 0.12183545529842377, "learning_rate": 8.06988908943491e-05, "loss": 0.006, "step": 32510 }, { "grad_norm": 0.07266164571046829, "learning_rate": 8.068583799380863e-05, "loss": 0.0066, "step": 32520 }, { "grad_norm": 0.12323911488056183, "learning_rate": 8.067278173751104e-05, "loss": 0.0066, "step": 32530 }, { "grad_norm": 0.10923191159963608, "learning_rate": 8.065972212688417e-05, "loss": 0.0062, "step": 32540 }, { "grad_norm": 0.10138490051031113, "learning_rate": 8.064665916335618e-05, "loss": 0.0051, "step": 32550 }, { "grad_norm": 0.08350130915641785, "learning_rate": 8.063359284835564e-05, "loss": 0.0069, "step": 32560 }, { "grad_norm": 0.10369834303855896, "learning_rate": 8.062052318331142e-05, "loss": 0.0071, "step": 32570 }, { "grad_norm": 0.10093967616558075, "learning_rate": 8.060745016965283e-05, "loss": 0.0056, "step": 32580 }, { "grad_norm": 0.08962036669254303, "learning_rate": 8.059437380880952e-05, "loss": 0.0045, "step": 32590 }, { "grad_norm": 0.09262016415596008, "learning_rate": 8.058129410221146e-05, "loss": 0.0049, "step": 32600 }, { "grad_norm": 0.09644994139671326, "learning_rate": 8.056821105128908e-05, "loss": 0.0064, "step": 32610 }, { "grad_norm": 0.12050803005695343, "learning_rate": 8.05551246574731e-05, "loss": 0.0055, "step": 32620 }, { "grad_norm": 0.10452121496200562, "learning_rate": 8.05420349221946e-05, "loss": 0.0066, "step": 32630 }, { "grad_norm": 0.18693310022354126, "learning_rate": 8.05289418468851e-05, "loss": 0.0062, "step": 32640 }, { "grad_norm": 0.1386536806821823, "learning_rate": 8.051584543297642e-05, "loss": 0.0056, "step": 32650 }, { "grad_norm": 0.10557857900857925, "learning_rate": 8.050274568190074e-05, "loss": 0.0049, "step": 32660 }, { "grad_norm": 0.12207242101430893, "learning_rate": 8.048964259509067e-05, "loss": 0.0056, "step": 32670 }, { "grad_norm": 0.1066114753484726, "learning_rate": 8.047653617397914e-05, "loss": 0.006, "step": 32680 }, { "grad_norm": 0.1086309626698494, "learning_rate": 8.046342641999941e-05, "loss": 0.0056, "step": 32690 }, { "grad_norm": 0.12985427677631378, "learning_rate": 8.045031333458517e-05, "loss": 0.0057, "step": 32700 }, { "grad_norm": 0.08778627961874008, "learning_rate": 8.043719691917047e-05, "loss": 0.006, "step": 32710 }, { "grad_norm": 0.22687369585037231, "learning_rate": 8.042407717518966e-05, "loss": 0.0056, "step": 32720 }, { "grad_norm": 0.1347539722919464, "learning_rate": 8.041095410407751e-05, "loss": 0.007, "step": 32730 }, { "grad_norm": 0.09947123378515244, "learning_rate": 8.039782770726913e-05, "loss": 0.0066, "step": 32740 }, { "grad_norm": 0.1783929169178009, "learning_rate": 8.038469798620004e-05, "loss": 0.0066, "step": 32750 }, { "grad_norm": 0.1016668975353241, "learning_rate": 8.037156494230604e-05, "loss": 0.0058, "step": 32760 }, { "grad_norm": 0.10456985980272293, "learning_rate": 8.035842857702338e-05, "loss": 0.0084, "step": 32770 }, { "grad_norm": 0.14737963676452637, "learning_rate": 8.03452888917886e-05, "loss": 0.0081, "step": 32780 }, { "grad_norm": 0.09754195064306259, "learning_rate": 8.033214588803866e-05, "loss": 0.0092, "step": 32790 }, { "grad_norm": 0.07607829570770264, "learning_rate": 8.031899956721083e-05, "loss": 0.0059, "step": 32800 }, { "grad_norm": 0.08616168797016144, "learning_rate": 8.030584993074282e-05, "loss": 0.0077, "step": 32810 }, { "grad_norm": 0.1329333633184433, "learning_rate": 8.02926969800726e-05, "loss": 0.0064, "step": 32820 }, { "grad_norm": 0.10756612569093704, "learning_rate": 8.027954071663859e-05, "loss": 0.0064, "step": 32830 }, { "grad_norm": 0.1414932757616043, "learning_rate": 8.026638114187954e-05, "loss": 0.0079, "step": 32840 }, { "grad_norm": 0.11558912694454193, "learning_rate": 8.025321825723456e-05, "loss": 0.0074, "step": 32850 }, { "grad_norm": 0.16801150143146515, "learning_rate": 8.02400520641431e-05, "loss": 0.0088, "step": 32860 }, { "grad_norm": 0.14316485822200775, "learning_rate": 8.022688256404501e-05, "loss": 0.0072, "step": 32870 }, { "grad_norm": 0.11934028565883636, "learning_rate": 8.02137097583805e-05, "loss": 0.0074, "step": 32880 }, { "grad_norm": 0.12104198336601257, "learning_rate": 8.02005336485901e-05, "loss": 0.0061, "step": 32890 }, { "grad_norm": 0.10000339895486832, "learning_rate": 8.018735423611476e-05, "loss": 0.0063, "step": 32900 }, { "grad_norm": 0.0928390622138977, "learning_rate": 8.017417152239574e-05, "loss": 0.0058, "step": 32910 }, { "grad_norm": 0.10162268579006195, "learning_rate": 8.01609855088747e-05, "loss": 0.0043, "step": 32920 }, { "grad_norm": 0.09179174900054932, "learning_rate": 8.014779619699362e-05, "loss": 0.0059, "step": 32930 }, { "grad_norm": 0.10088489949703217, "learning_rate": 8.013460358819489e-05, "loss": 0.0041, "step": 32940 }, { "grad_norm": 0.07476793229579926, "learning_rate": 8.01214076839212e-05, "loss": 0.0049, "step": 32950 }, { "grad_norm": 0.0803547203540802, "learning_rate": 8.010820848561565e-05, "loss": 0.0063, "step": 32960 }, { "grad_norm": 0.09942159801721573, "learning_rate": 8.009500599472171e-05, "loss": 0.0052, "step": 32970 }, { "grad_norm": 0.0939583033323288, "learning_rate": 8.008180021268314e-05, "loss": 0.0047, "step": 32980 }, { "grad_norm": 0.08257751911878586, "learning_rate": 8.006859114094414e-05, "loss": 0.0058, "step": 32990 }, { "grad_norm": 0.16276173293590546, "learning_rate": 8.005537878094921e-05, "loss": 0.0052, "step": 33000 }, { "grad_norm": 0.06878495961427689, "learning_rate": 8.004216313414323e-05, "loss": 0.0048, "step": 33010 }, { "grad_norm": 0.09279246628284454, "learning_rate": 8.002894420197149e-05, "loss": 0.0047, "step": 33020 }, { "grad_norm": 0.09146548807621002, "learning_rate": 8.001572198587954e-05, "loss": 0.0051, "step": 33030 }, { "grad_norm": 0.13235707581043243, "learning_rate": 8.000249648731338e-05, "loss": 0.0052, "step": 33040 }, { "grad_norm": 0.21724697947502136, "learning_rate": 7.998926770771928e-05, "loss": 0.0058, "step": 33050 }, { "grad_norm": 0.1140199527144432, "learning_rate": 7.997603564854397e-05, "loss": 0.0053, "step": 33060 }, { "grad_norm": 0.07875782251358032, "learning_rate": 7.996280031123448e-05, "loss": 0.0048, "step": 33070 }, { "grad_norm": 0.10811254382133484, "learning_rate": 7.994956169723818e-05, "loss": 0.0055, "step": 33080 }, { "grad_norm": 0.12857480347156525, "learning_rate": 7.993631980800285e-05, "loss": 0.0063, "step": 33090 }, { "grad_norm": 0.11831674724817276, "learning_rate": 7.992307464497659e-05, "loss": 0.0075, "step": 33100 }, { "grad_norm": 0.12157875299453735, "learning_rate": 7.990982620960787e-05, "loss": 0.0053, "step": 33110 }, { "grad_norm": 0.1195179671049118, "learning_rate": 7.989657450334554e-05, "loss": 0.0049, "step": 33120 }, { "grad_norm": 0.10937897115945816, "learning_rate": 7.988331952763877e-05, "loss": 0.0066, "step": 33130 }, { "grad_norm": 0.07525473833084106, "learning_rate": 7.987006128393709e-05, "loss": 0.0046, "step": 33140 }, { "grad_norm": 0.1254822015762329, "learning_rate": 7.985679977369043e-05, "loss": 0.0068, "step": 33150 }, { "grad_norm": 0.11062629520893097, "learning_rate": 7.984353499834902e-05, "loss": 0.0073, "step": 33160 }, { "grad_norm": 0.15299417078495026, "learning_rate": 7.983026695936351e-05, "loss": 0.0076, "step": 33170 }, { "grad_norm": 0.09854132682085037, "learning_rate": 7.981699565818486e-05, "loss": 0.0056, "step": 33180 }, { "grad_norm": 0.10292752832174301, "learning_rate": 7.980372109626437e-05, "loss": 0.0052, "step": 33190 }, { "grad_norm": 0.07251686602830887, "learning_rate": 7.979044327505375e-05, "loss": 0.0051, "step": 33200 }, { "grad_norm": 0.1516837775707245, "learning_rate": 7.977716219600506e-05, "loss": 0.007, "step": 33210 }, { "grad_norm": 0.11179443448781967, "learning_rate": 7.97638778605707e-05, "loss": 0.0059, "step": 33220 }, { "grad_norm": 0.09247998148202896, "learning_rate": 7.975059027020338e-05, "loss": 0.0055, "step": 33230 }, { "grad_norm": 0.11064906418323517, "learning_rate": 7.973729942635623e-05, "loss": 0.005, "step": 33240 }, { "grad_norm": 0.08441868424415588, "learning_rate": 7.972400533048273e-05, "loss": 0.005, "step": 33250 }, { "grad_norm": 0.07671601325273514, "learning_rate": 7.97107079840367e-05, "loss": 0.0054, "step": 33260 }, { "grad_norm": 0.08256629854440689, "learning_rate": 7.969740738847231e-05, "loss": 0.0056, "step": 33270 }, { "grad_norm": 0.14910811185836792, "learning_rate": 7.968410354524411e-05, "loss": 0.006, "step": 33280 }, { "grad_norm": 0.1075935885310173, "learning_rate": 7.967079645580697e-05, "loss": 0.0068, "step": 33290 }, { "grad_norm": 0.1194399744272232, "learning_rate": 7.965748612161612e-05, "loss": 0.0048, "step": 33300 }, { "grad_norm": 0.1145920380949974, "learning_rate": 7.96441725441272e-05, "loss": 0.0057, "step": 33310 }, { "grad_norm": 0.08155815303325653, "learning_rate": 7.963085572479614e-05, "loss": 0.0046, "step": 33320 }, { "grad_norm": 0.07384727895259857, "learning_rate": 7.961753566507924e-05, "loss": 0.0047, "step": 33330 }, { "grad_norm": 0.14131559431552887, "learning_rate": 7.960421236643316e-05, "loss": 0.0056, "step": 33340 }, { "grad_norm": 0.14319977164268494, "learning_rate": 7.959088583031496e-05, "loss": 0.0071, "step": 33350 }, { "grad_norm": 0.11610124260187149, "learning_rate": 7.957755605818194e-05, "loss": 0.0059, "step": 33360 }, { "grad_norm": 0.13705123960971832, "learning_rate": 7.956422305149185e-05, "loss": 0.0047, "step": 33370 }, { "grad_norm": 0.1453244835138321, "learning_rate": 7.95508868117028e-05, "loss": 0.0052, "step": 33380 }, { "grad_norm": 0.10294263809919357, "learning_rate": 7.953754734027318e-05, "loss": 0.0062, "step": 33390 }, { "grad_norm": 0.20147424936294556, "learning_rate": 7.952420463866182e-05, "loss": 0.0057, "step": 33400 }, { "grad_norm": 0.12452374398708344, "learning_rate": 7.951085870832782e-05, "loss": 0.0061, "step": 33410 }, { "grad_norm": 0.12088190764188766, "learning_rate": 7.949750955073067e-05, "loss": 0.0061, "step": 33420 }, { "grad_norm": 0.11234032362699509, "learning_rate": 7.948415716733022e-05, "loss": 0.0055, "step": 33430 }, { "grad_norm": 0.14270535111427307, "learning_rate": 7.947080155958669e-05, "loss": 0.0068, "step": 33440 }, { "grad_norm": 0.09869702160358429, "learning_rate": 7.94574427289606e-05, "loss": 0.0052, "step": 33450 }, { "grad_norm": 0.1305810958147049, "learning_rate": 7.944408067691284e-05, "loss": 0.0055, "step": 33460 }, { "grad_norm": 0.11239513009786606, "learning_rate": 7.943071540490473e-05, "loss": 0.0059, "step": 33470 }, { "grad_norm": 0.10815078765153885, "learning_rate": 7.94173469143978e-05, "loss": 0.0052, "step": 33480 }, { "grad_norm": 0.11710706353187561, "learning_rate": 7.940397520685406e-05, "loss": 0.0061, "step": 33490 }, { "grad_norm": 0.13619080185890198, "learning_rate": 7.939060028373577e-05, "loss": 0.0054, "step": 33500 }, { "grad_norm": 0.11684869229793549, "learning_rate": 7.937722214650565e-05, "loss": 0.0077, "step": 33510 }, { "grad_norm": 0.09015581756830215, "learning_rate": 7.936384079662666e-05, "loss": 0.0051, "step": 33520 }, { "grad_norm": 0.1120588406920433, "learning_rate": 7.93504562355622e-05, "loss": 0.0053, "step": 33530 }, { "grad_norm": 0.08342215418815613, "learning_rate": 7.933706846477599e-05, "loss": 0.0047, "step": 33540 }, { "grad_norm": 0.09587882459163666, "learning_rate": 7.932367748573206e-05, "loss": 0.0048, "step": 33550 }, { "grad_norm": 0.12230517715215683, "learning_rate": 7.931028329989485e-05, "loss": 0.0047, "step": 33560 }, { "grad_norm": 0.08940041065216064, "learning_rate": 7.929688590872913e-05, "loss": 0.0047, "step": 33570 }, { "grad_norm": 0.09687183797359467, "learning_rate": 7.928348531370003e-05, "loss": 0.0054, "step": 33580 }, { "grad_norm": 0.10748649388551712, "learning_rate": 7.927008151627297e-05, "loss": 0.0045, "step": 33590 }, { "grad_norm": 0.08355730772018433, "learning_rate": 7.925667451791383e-05, "loss": 0.0041, "step": 33600 }, { "grad_norm": 0.09272479265928268, "learning_rate": 7.924326432008874e-05, "loss": 0.0044, "step": 33610 }, { "grad_norm": 0.1146857738494873, "learning_rate": 7.922985092426422e-05, "loss": 0.0062, "step": 33620 }, { "grad_norm": 0.06929384171962738, "learning_rate": 7.921643433190717e-05, "loss": 0.0061, "step": 33630 }, { "grad_norm": 0.09056559950113297, "learning_rate": 7.920301454448478e-05, "loss": 0.0052, "step": 33640 }, { "grad_norm": 0.1423387974500656, "learning_rate": 7.918959156346461e-05, "loss": 0.0065, "step": 33650 }, { "grad_norm": 0.08367783576250076, "learning_rate": 7.91761653903146e-05, "loss": 0.0058, "step": 33660 }, { "grad_norm": 0.09298636764287949, "learning_rate": 7.916273602650302e-05, "loss": 0.0043, "step": 33670 }, { "grad_norm": 0.06241494044661522, "learning_rate": 7.914930347349847e-05, "loss": 0.0038, "step": 33680 }, { "grad_norm": 0.061755821108818054, "learning_rate": 7.913586773276992e-05, "loss": 0.0039, "step": 33690 }, { "grad_norm": 0.0923982784152031, "learning_rate": 7.912242880578667e-05, "loss": 0.0066, "step": 33700 }, { "grad_norm": 0.10871495306491852, "learning_rate": 7.910898669401839e-05, "loss": 0.0044, "step": 33710 }, { "grad_norm": 0.09064047038555145, "learning_rate": 7.909554139893511e-05, "loss": 0.0046, "step": 33720 }, { "grad_norm": 0.09705790132284164, "learning_rate": 7.908209292200715e-05, "loss": 0.0042, "step": 33730 }, { "grad_norm": 0.08705279976129532, "learning_rate": 7.906864126470523e-05, "loss": 0.0055, "step": 33740 }, { "grad_norm": 0.08964665979146957, "learning_rate": 7.905518642850041e-05, "loss": 0.0049, "step": 33750 }, { "grad_norm": 0.10802163928747177, "learning_rate": 7.904172841486409e-05, "loss": 0.0052, "step": 33760 }, { "grad_norm": 0.08421911299228668, "learning_rate": 7.902826722526801e-05, "loss": 0.0041, "step": 33770 }, { "grad_norm": 0.13561603426933289, "learning_rate": 7.901480286118427e-05, "loss": 0.0054, "step": 33780 }, { "grad_norm": 0.08873169124126434, "learning_rate": 7.900133532408531e-05, "loss": 0.0052, "step": 33790 }, { "grad_norm": 0.09110824763774872, "learning_rate": 7.898786461544395e-05, "loss": 0.0055, "step": 33800 }, { "grad_norm": 0.08578597009181976, "learning_rate": 7.897439073673325e-05, "loss": 0.0047, "step": 33810 }, { "grad_norm": 0.08709271252155304, "learning_rate": 7.896091368942677e-05, "loss": 0.0047, "step": 33820 }, { "grad_norm": 0.09780310094356537, "learning_rate": 7.894743347499832e-05, "loss": 0.0052, "step": 33830 }, { "grad_norm": 0.08790504187345505, "learning_rate": 7.893395009492203e-05, "loss": 0.0048, "step": 33840 }, { "grad_norm": 0.10831759124994278, "learning_rate": 7.892046355067248e-05, "loss": 0.0043, "step": 33850 }, { "grad_norm": 0.08651257306337357, "learning_rate": 7.890697384372451e-05, "loss": 0.0051, "step": 33860 }, { "grad_norm": 0.11805271357297897, "learning_rate": 7.889348097555336e-05, "loss": 0.0057, "step": 33870 }, { "grad_norm": 0.07612749934196472, "learning_rate": 7.887998494763455e-05, "loss": 0.005, "step": 33880 }, { "grad_norm": 0.13963328301906586, "learning_rate": 7.886648576144404e-05, "loss": 0.0069, "step": 33890 }, { "grad_norm": 0.09160185605287552, "learning_rate": 7.885298341845802e-05, "loss": 0.0065, "step": 33900 }, { "grad_norm": 0.08496172726154327, "learning_rate": 7.883947792015311e-05, "loss": 0.0048, "step": 33910 }, { "grad_norm": 0.13331763446331024, "learning_rate": 7.882596926800628e-05, "loss": 0.0042, "step": 33920 }, { "grad_norm": 0.07159213721752167, "learning_rate": 7.881245746349477e-05, "loss": 0.0044, "step": 33930 }, { "grad_norm": 0.1261298954486847, "learning_rate": 7.879894250809623e-05, "loss": 0.006, "step": 33940 }, { "grad_norm": 0.17791984975337982, "learning_rate": 7.878542440328865e-05, "loss": 0.0063, "step": 33950 }, { "grad_norm": 0.13010789453983307, "learning_rate": 7.877190315055031e-05, "loss": 0.0074, "step": 33960 }, { "grad_norm": 0.10747207701206207, "learning_rate": 7.875837875135991e-05, "loss": 0.0066, "step": 33970 }, { "grad_norm": 0.08600252866744995, "learning_rate": 7.874485120719646e-05, "loss": 0.0067, "step": 33980 }, { "grad_norm": 0.11041729897260666, "learning_rate": 7.873132051953928e-05, "loss": 0.0048, "step": 33990 }, { "grad_norm": 0.08391238003969193, "learning_rate": 7.87177866898681e-05, "loss": 0.0058, "step": 34000 }, { "grad_norm": 0.09225599467754364, "learning_rate": 7.870424971966294e-05, "loss": 0.0049, "step": 34010 }, { "grad_norm": 0.08326974511146545, "learning_rate": 7.869070961040419e-05, "loss": 0.0049, "step": 34020 }, { "grad_norm": 0.11620515584945679, "learning_rate": 7.867716636357257e-05, "loss": 0.0063, "step": 34030 }, { "grad_norm": 0.12181723862886429, "learning_rate": 7.866361998064915e-05, "loss": 0.005, "step": 34040 }, { "grad_norm": 0.1217319667339325, "learning_rate": 7.865007046311534e-05, "loss": 0.0067, "step": 34050 }, { "grad_norm": 0.13127608597278595, "learning_rate": 7.86365178124529e-05, "loss": 0.0045, "step": 34060 }, { "grad_norm": 0.10030031204223633, "learning_rate": 7.862296203014394e-05, "loss": 0.0038, "step": 34070 }, { "grad_norm": 0.10573714226484299, "learning_rate": 7.860940311767088e-05, "loss": 0.0042, "step": 34080 }, { "grad_norm": 0.09620344638824463, "learning_rate": 7.85958410765165e-05, "loss": 0.0054, "step": 34090 }, { "grad_norm": 0.12243486940860748, "learning_rate": 7.858227590816394e-05, "loss": 0.0044, "step": 34100 }, { "grad_norm": 0.08339839428663254, "learning_rate": 7.856870761409664e-05, "loss": 0.0051, "step": 34110 }, { "grad_norm": 0.13646598160266876, "learning_rate": 7.855513619579846e-05, "loss": 0.0054, "step": 34120 }, { "grad_norm": 0.20625564455986023, "learning_rate": 7.85415616547535e-05, "loss": 0.0057, "step": 34130 }, { "grad_norm": 0.15376999974250793, "learning_rate": 7.852798399244627e-05, "loss": 0.0071, "step": 34140 }, { "grad_norm": 0.12867558002471924, "learning_rate": 7.851440321036161e-05, "loss": 0.0059, "step": 34150 }, { "grad_norm": 0.1344795972108841, "learning_rate": 7.850081930998468e-05, "loss": 0.0054, "step": 34160 }, { "grad_norm": 0.10683870315551758, "learning_rate": 7.8487232292801e-05, "loss": 0.005, "step": 34170 }, { "grad_norm": 0.08706803619861603, "learning_rate": 7.847364216029642e-05, "loss": 0.0055, "step": 34180 }, { "grad_norm": 0.11002759635448456, "learning_rate": 7.846004891395716e-05, "loss": 0.0051, "step": 34190 }, { "grad_norm": 0.08666086941957474, "learning_rate": 7.844645255526972e-05, "loss": 0.0063, "step": 34200 }, { "grad_norm": 0.12408973276615143, "learning_rate": 7.843285308572101e-05, "loss": 0.0056, "step": 34210 }, { "grad_norm": 0.3793765902519226, "learning_rate": 7.841925050679823e-05, "loss": 0.0051, "step": 34220 }, { "grad_norm": 0.14813114702701569, "learning_rate": 7.840564481998895e-05, "loss": 0.0067, "step": 34230 }, { "grad_norm": 0.1099475845694542, "learning_rate": 7.839203602678105e-05, "loss": 0.0065, "step": 34240 }, { "grad_norm": 0.12448938935995102, "learning_rate": 7.837842412866279e-05, "loss": 0.006, "step": 34250 }, { "grad_norm": 0.10915815085172653, "learning_rate": 7.836480912712272e-05, "loss": 0.0059, "step": 34260 }, { "grad_norm": 0.11589314788579941, "learning_rate": 7.835119102364976e-05, "loss": 0.0056, "step": 34270 }, { "grad_norm": 0.10049755126237869, "learning_rate": 7.83375698197332e-05, "loss": 0.0053, "step": 34280 }, { "grad_norm": 0.0945233702659607, "learning_rate": 7.83239455168626e-05, "loss": 0.0051, "step": 34290 }, { "grad_norm": 0.08626949042081833, "learning_rate": 7.83103181165279e-05, "loss": 0.0052, "step": 34300 }, { "grad_norm": 0.12364507466554642, "learning_rate": 7.829668762021937e-05, "loss": 0.0066, "step": 34310 }, { "grad_norm": 0.10492698848247528, "learning_rate": 7.828305402942764e-05, "loss": 0.0053, "step": 34320 }, { "grad_norm": 0.10755756497383118, "learning_rate": 7.826941734564363e-05, "loss": 0.0062, "step": 34330 }, { "grad_norm": 0.10115876793861389, "learning_rate": 7.825577757035865e-05, "loss": 0.0057, "step": 34340 }, { "grad_norm": 0.15425342321395874, "learning_rate": 7.824213470506431e-05, "loss": 0.007, "step": 34350 }, { "grad_norm": 0.06932787597179413, "learning_rate": 7.822848875125257e-05, "loss": 0.0055, "step": 34360 }, { "grad_norm": 0.13738025724887848, "learning_rate": 7.821483971041576e-05, "loss": 0.0062, "step": 34370 }, { "grad_norm": 0.10564988106489182, "learning_rate": 7.820118758404649e-05, "loss": 0.006, "step": 34380 }, { "grad_norm": 0.11125874519348145, "learning_rate": 7.818753237363776e-05, "loss": 0.0055, "step": 34390 }, { "grad_norm": 0.09240160882472992, "learning_rate": 7.817387408068286e-05, "loss": 0.0083, "step": 34400 }, { "grad_norm": 0.12653705477714539, "learning_rate": 7.816021270667544e-05, "loss": 0.0065, "step": 34410 }, { "grad_norm": 0.10102898627519608, "learning_rate": 7.81465482531095e-05, "loss": 0.0054, "step": 34420 }, { "grad_norm": 0.09438609331846237, "learning_rate": 7.813288072147938e-05, "loss": 0.0045, "step": 34430 }, { "grad_norm": 0.13677769899368286, "learning_rate": 7.811921011327972e-05, "loss": 0.0063, "step": 34440 }, { "grad_norm": 0.12601710855960846, "learning_rate": 7.810553643000549e-05, "loss": 0.0049, "step": 34450 }, { "grad_norm": 0.12582527101039886, "learning_rate": 7.809185967315206e-05, "loss": 0.004, "step": 34460 }, { "grad_norm": 0.09345990419387817, "learning_rate": 7.80781798442151e-05, "loss": 0.0041, "step": 34470 }, { "grad_norm": 0.057842839509248734, "learning_rate": 7.806449694469061e-05, "loss": 0.0041, "step": 34480 }, { "grad_norm": 0.13062623143196106, "learning_rate": 7.805081097607492e-05, "loss": 0.0051, "step": 34490 }, { "grad_norm": 0.1359357237815857, "learning_rate": 7.803712193986474e-05, "loss": 0.0071, "step": 34500 }, { "grad_norm": 0.12982484698295593, "learning_rate": 7.802342983755702e-05, "loss": 0.0072, "step": 34510 }, { "grad_norm": 0.11990265548229218, "learning_rate": 7.800973467064918e-05, "loss": 0.0052, "step": 34520 }, { "grad_norm": 0.10556477308273315, "learning_rate": 7.799603644063884e-05, "loss": 0.0061, "step": 34530 }, { "grad_norm": 0.10320451110601425, "learning_rate": 7.798233514902405e-05, "loss": 0.0063, "step": 34540 }, { "grad_norm": 0.09469668567180634, "learning_rate": 7.796863079730318e-05, "loss": 0.0064, "step": 34550 }, { "grad_norm": 0.09485118091106415, "learning_rate": 7.795492338697488e-05, "loss": 0.0057, "step": 34560 }, { "grad_norm": 0.08668598532676697, "learning_rate": 7.794121291953819e-05, "loss": 0.0046, "step": 34570 }, { "grad_norm": 0.07821452617645264, "learning_rate": 7.792749939649246e-05, "loss": 0.005, "step": 34580 }, { "grad_norm": 0.1254504770040512, "learning_rate": 7.79137828193374e-05, "loss": 0.0056, "step": 34590 }, { "grad_norm": 0.12815608084201813, "learning_rate": 7.790006318957301e-05, "loss": 0.0054, "step": 34600 }, { "grad_norm": 0.07060649245977402, "learning_rate": 7.788634050869965e-05, "loss": 0.0055, "step": 34610 }, { "grad_norm": 0.08604001998901367, "learning_rate": 7.787261477821803e-05, "loss": 0.005, "step": 34620 }, { "grad_norm": 0.10533910244703293, "learning_rate": 7.785888599962916e-05, "loss": 0.0046, "step": 34630 }, { "grad_norm": 0.08768974244594574, "learning_rate": 7.784515417443439e-05, "loss": 0.0049, "step": 34640 }, { "grad_norm": 0.09583622217178345, "learning_rate": 7.783141930413545e-05, "loss": 0.0055, "step": 34650 }, { "grad_norm": 0.12233845889568329, "learning_rate": 7.78176813902343e-05, "loss": 0.0055, "step": 34660 }, { "grad_norm": 0.12358878552913666, "learning_rate": 7.780394043423336e-05, "loss": 0.007, "step": 34670 }, { "grad_norm": 0.1257266253232956, "learning_rate": 7.77901964376353e-05, "loss": 0.0057, "step": 34680 }, { "grad_norm": 0.12369371205568314, "learning_rate": 7.777644940194316e-05, "loss": 0.0059, "step": 34690 }, { "grad_norm": 0.11627274751663208, "learning_rate": 7.776269932866023e-05, "loss": 0.0062, "step": 34700 }, { "grad_norm": 0.0844770073890686, "learning_rate": 7.774894621929026e-05, "loss": 0.005, "step": 34710 }, { "grad_norm": 0.13630490005016327, "learning_rate": 7.773519007533725e-05, "loss": 0.0071, "step": 34720 }, { "grad_norm": 0.12360721826553345, "learning_rate": 7.772143089830556e-05, "loss": 0.006, "step": 34730 }, { "grad_norm": 0.09882106631994247, "learning_rate": 7.770766868969985e-05, "loss": 0.0059, "step": 34740 }, { "grad_norm": 0.11648308485746384, "learning_rate": 7.769390345102518e-05, "loss": 0.0056, "step": 34750 }, { "grad_norm": 0.1188681423664093, "learning_rate": 7.768013518378683e-05, "loss": 0.006, "step": 34760 }, { "grad_norm": 0.10657629370689392, "learning_rate": 7.766636388949053e-05, "loss": 0.0046, "step": 34770 }, { "grad_norm": 0.0984049066901207, "learning_rate": 7.765258956964229e-05, "loss": 0.005, "step": 34780 }, { "grad_norm": 0.08175291121006012, "learning_rate": 7.76388122257484e-05, "loss": 0.0049, "step": 34790 }, { "grad_norm": 0.09524627774953842, "learning_rate": 7.762503185931558e-05, "loss": 0.0041, "step": 34800 }, { "grad_norm": 0.08502291887998581, "learning_rate": 7.76112484718508e-05, "loss": 0.0054, "step": 34810 }, { "grad_norm": 0.12589848041534424, "learning_rate": 7.75974620648614e-05, "loss": 0.0049, "step": 34820 }, { "grad_norm": 0.10087554156780243, "learning_rate": 7.758367263985503e-05, "loss": 0.0052, "step": 34830 }, { "grad_norm": 0.06007647141814232, "learning_rate": 7.75698801983397e-05, "loss": 0.0042, "step": 34840 }, { "grad_norm": 0.07149161398410797, "learning_rate": 7.755608474182372e-05, "loss": 0.0059, "step": 34850 }, { "grad_norm": 0.13883347809314728, "learning_rate": 7.754228627181574e-05, "loss": 0.0068, "step": 34860 }, { "grad_norm": 0.15728211402893066, "learning_rate": 7.752848478982476e-05, "loss": 0.006, "step": 34870 }, { "grad_norm": 0.09910345822572708, "learning_rate": 7.751468029736006e-05, "loss": 0.0048, "step": 34880 }, { "grad_norm": 0.10226169228553772, "learning_rate": 7.750087279593129e-05, "loss": 0.0049, "step": 34890 }, { "grad_norm": 0.11297678202390671, "learning_rate": 7.748706228704843e-05, "loss": 0.0049, "step": 34900 }, { "grad_norm": 0.07062573730945587, "learning_rate": 7.747324877222176e-05, "loss": 0.0047, "step": 34910 }, { "grad_norm": 0.1091548353433609, "learning_rate": 7.745943225296188e-05, "loss": 0.0049, "step": 34920 }, { "grad_norm": 0.09643694758415222, "learning_rate": 7.744561273077981e-05, "loss": 0.0046, "step": 34930 }, { "grad_norm": 0.09701508283615112, "learning_rate": 7.743179020718678e-05, "loss": 0.0043, "step": 34940 }, { "grad_norm": 0.09848588705062866, "learning_rate": 7.741796468369443e-05, "loss": 0.0053, "step": 34950 }, { "grad_norm": 0.0744103193283081, "learning_rate": 7.740413616181466e-05, "loss": 0.0045, "step": 34960 }, { "grad_norm": 0.10629835724830627, "learning_rate": 7.739030464305978e-05, "loss": 0.0059, "step": 34970 }, { "grad_norm": 0.05998681113123894, "learning_rate": 7.737647012894235e-05, "loss": 0.0042, "step": 34980 }, { "grad_norm": 0.09421083331108093, "learning_rate": 7.736263262097532e-05, "loss": 0.0049, "step": 34990 }, { "grad_norm": 0.10341139137744904, "learning_rate": 7.734879212067192e-05, "loss": 0.0057, "step": 35000 }, { "grad_norm": 0.08469630032777786, "learning_rate": 7.733494862954573e-05, "loss": 0.0046, "step": 35010 }, { "grad_norm": 0.11542098969221115, "learning_rate": 7.732110214911066e-05, "loss": 0.0052, "step": 35020 }, { "grad_norm": 0.14710505306720734, "learning_rate": 7.730725268088094e-05, "loss": 0.0058, "step": 35030 }, { "grad_norm": 0.09112498164176941, "learning_rate": 7.729340022637112e-05, "loss": 0.0047, "step": 35040 }, { "grad_norm": 0.09464767575263977, "learning_rate": 7.727954478709607e-05, "loss": 0.0058, "step": 35050 }, { "grad_norm": 0.06605538725852966, "learning_rate": 7.726568636457103e-05, "loss": 0.0049, "step": 35060 }, { "grad_norm": 0.1045849621295929, "learning_rate": 7.725182496031153e-05, "loss": 0.0049, "step": 35070 }, { "grad_norm": 0.08545050024986267, "learning_rate": 7.723796057583341e-05, "loss": 0.0046, "step": 35080 }, { "grad_norm": 0.10621339827775955, "learning_rate": 7.722409321265287e-05, "loss": 0.0042, "step": 35090 }, { "grad_norm": 0.10514959692955017, "learning_rate": 7.721022287228645e-05, "loss": 0.005, "step": 35100 }, { "grad_norm": 0.10331451892852783, "learning_rate": 7.719634955625094e-05, "loss": 0.0045, "step": 35110 }, { "grad_norm": 0.11755692958831787, "learning_rate": 7.718247326606355e-05, "loss": 0.005, "step": 35120 }, { "grad_norm": 0.08967312425374985, "learning_rate": 7.716859400324175e-05, "loss": 0.0041, "step": 35130 }, { "grad_norm": 0.1079578548669815, "learning_rate": 7.715471176930335e-05, "loss": 0.0042, "step": 35140 }, { "grad_norm": 0.19678038358688354, "learning_rate": 7.714082656576651e-05, "loss": 0.0046, "step": 35150 }, { "grad_norm": 0.08201887458562851, "learning_rate": 7.712693839414968e-05, "loss": 0.0051, "step": 35160 }, { "grad_norm": 0.09694919735193253, "learning_rate": 7.711304725597164e-05, "loss": 0.005, "step": 35170 }, { "grad_norm": 0.09950127452611923, "learning_rate": 7.709915315275151e-05, "loss": 0.0051, "step": 35180 }, { "grad_norm": 0.11636051535606384, "learning_rate": 7.708525608600876e-05, "loss": 0.0067, "step": 35190 }, { "grad_norm": 0.0981614962220192, "learning_rate": 7.707135605726311e-05, "loss": 0.0051, "step": 35200 }, { "grad_norm": 0.10202880203723907, "learning_rate": 7.705745306803466e-05, "loss": 0.0066, "step": 35210 }, { "grad_norm": 0.1672060638666153, "learning_rate": 7.704354711984383e-05, "loss": 0.0056, "step": 35220 }, { "grad_norm": 0.10561536997556686, "learning_rate": 7.702963821421133e-05, "loss": 0.0054, "step": 35230 }, { "grad_norm": 0.11337202042341232, "learning_rate": 7.701572635265826e-05, "loss": 0.0058, "step": 35240 }, { "grad_norm": 0.07879692316055298, "learning_rate": 7.700181153670596e-05, "loss": 0.004, "step": 35250 }, { "grad_norm": 0.07715544104576111, "learning_rate": 7.698789376787616e-05, "loss": 0.0048, "step": 35260 }, { "grad_norm": 0.11331816017627716, "learning_rate": 7.697397304769085e-05, "loss": 0.005, "step": 35270 }, { "grad_norm": 0.12553825974464417, "learning_rate": 7.696004937767241e-05, "loss": 0.0071, "step": 35280 }, { "grad_norm": 0.09471054375171661, "learning_rate": 7.694612275934352e-05, "loss": 0.005, "step": 35290 }, { "grad_norm": 0.12761808931827545, "learning_rate": 7.693219319422714e-05, "loss": 0.0059, "step": 35300 }, { "grad_norm": 0.12573248147964478, "learning_rate": 7.69182606838466e-05, "loss": 0.0041, "step": 35310 }, { "grad_norm": 0.08232981711626053, "learning_rate": 7.690432522972558e-05, "loss": 0.0048, "step": 35320 }, { "grad_norm": 0.09222664684057236, "learning_rate": 7.689038683338796e-05, "loss": 0.0049, "step": 35330 }, { "grad_norm": 0.09112349152565002, "learning_rate": 7.687644549635808e-05, "loss": 0.0049, "step": 35340 }, { "grad_norm": 0.09431610256433487, "learning_rate": 7.686250122016053e-05, "loss": 0.0043, "step": 35350 }, { "grad_norm": 0.1365090161561966, "learning_rate": 7.684855400632022e-05, "loss": 0.0053, "step": 35360 }, { "grad_norm": 0.10469028353691101, "learning_rate": 7.683460385636243e-05, "loss": 0.0072, "step": 35370 }, { "grad_norm": 0.10247493535280228, "learning_rate": 7.68206507718127e-05, "loss": 0.0061, "step": 35380 }, { "grad_norm": 0.1146962121129036, "learning_rate": 7.680669475419693e-05, "loss": 0.005, "step": 35390 }, { "grad_norm": 0.11922801285982132, "learning_rate": 7.679273580504132e-05, "loss": 0.0069, "step": 35400 }, { "grad_norm": 0.09494592994451523, "learning_rate": 7.67787739258724e-05, "loss": 0.0059, "step": 35410 }, { "grad_norm": 0.11497675627470016, "learning_rate": 7.676480911821705e-05, "loss": 0.0059, "step": 35420 }, { "grad_norm": 0.11035238206386566, "learning_rate": 7.675084138360238e-05, "loss": 0.0054, "step": 35430 }, { "grad_norm": 0.11180105805397034, "learning_rate": 7.673687072355592e-05, "loss": 0.0068, "step": 35440 }, { "grad_norm": 0.11386837065219879, "learning_rate": 7.67228971396055e-05, "loss": 0.0063, "step": 35450 }, { "grad_norm": 0.13049253821372986, "learning_rate": 7.670892063327922e-05, "loss": 0.0055, "step": 35460 }, { "grad_norm": 0.0841047391295433, "learning_rate": 7.669494120610552e-05, "loss": 0.006, "step": 35470 }, { "grad_norm": 0.0716504454612732, "learning_rate": 7.66809588596132e-05, "loss": 0.0045, "step": 35480 }, { "grad_norm": 0.08869212865829468, "learning_rate": 7.666697359533132e-05, "loss": 0.0053, "step": 35490 }, { "grad_norm": 0.13608579337596893, "learning_rate": 7.665298541478932e-05, "loss": 0.005, "step": 35500 }, { "grad_norm": 0.08973261713981628, "learning_rate": 7.66389943195169e-05, "loss": 0.0044, "step": 35510 }, { "grad_norm": 0.09728194773197174, "learning_rate": 7.66250003110441e-05, "loss": 0.0055, "step": 35520 }, { "grad_norm": 0.09553764015436172, "learning_rate": 7.66110033909013e-05, "loss": 0.0041, "step": 35530 }, { "grad_norm": 0.10437005013227463, "learning_rate": 7.659700356061918e-05, "loss": 0.005, "step": 35540 }, { "grad_norm": 0.11223003268241882, "learning_rate": 7.658300082172875e-05, "loss": 0.006, "step": 35550 }, { "grad_norm": 0.10961678624153137, "learning_rate": 7.65689951757613e-05, "loss": 0.0052, "step": 35560 }, { "grad_norm": 0.11717677116394043, "learning_rate": 7.655498662424849e-05, "loss": 0.0045, "step": 35570 }, { "grad_norm": 0.07145336270332336, "learning_rate": 7.654097516872227e-05, "loss": 0.0059, "step": 35580 }, { "grad_norm": 0.08089242875576019, "learning_rate": 7.65269608107149e-05, "loss": 0.0047, "step": 35590 }, { "grad_norm": 0.08171938359737396, "learning_rate": 7.6512943551759e-05, "loss": 0.0055, "step": 35600 }, { "grad_norm": 0.11546256393194199, "learning_rate": 7.649892339338743e-05, "loss": 0.0061, "step": 35610 }, { "grad_norm": 0.11538615822792053, "learning_rate": 7.648490033713344e-05, "loss": 0.0049, "step": 35620 }, { "grad_norm": 0.10893691331148148, "learning_rate": 7.647087438453058e-05, "loss": 0.0056, "step": 35630 }, { "grad_norm": 0.09972716122865677, "learning_rate": 7.64568455371127e-05, "loss": 0.0048, "step": 35640 }, { "grad_norm": 0.22304880619049072, "learning_rate": 7.644281379641396e-05, "loss": 0.0049, "step": 35650 }, { "grad_norm": 0.11292757093906403, "learning_rate": 7.642877916396887e-05, "loss": 0.0049, "step": 35660 }, { "grad_norm": 0.12430600821971893, "learning_rate": 7.641474164131221e-05, "loss": 0.0048, "step": 35670 }, { "grad_norm": 0.10488355904817581, "learning_rate": 7.640070122997913e-05, "loss": 0.0053, "step": 35680 }, { "grad_norm": 0.11199899762868881, "learning_rate": 7.638665793150505e-05, "loss": 0.0056, "step": 35690 }, { "grad_norm": 0.0963953286409378, "learning_rate": 7.637261174742574e-05, "loss": 0.0041, "step": 35700 }, { "grad_norm": 0.1398579627275467, "learning_rate": 7.635856267927725e-05, "loss": 0.0066, "step": 35710 }, { "grad_norm": 0.09671038389205933, "learning_rate": 7.634451072859597e-05, "loss": 0.0058, "step": 35720 }, { "grad_norm": 0.1047857329249382, "learning_rate": 7.633045589691863e-05, "loss": 0.0053, "step": 35730 }, { "grad_norm": 0.14266374707221985, "learning_rate": 7.63163981857822e-05, "loss": 0.0054, "step": 35740 }, { "grad_norm": 0.11391990631818771, "learning_rate": 7.630233759672403e-05, "loss": 0.0052, "step": 35750 }, { "grad_norm": 0.12618832290172577, "learning_rate": 7.628827413128177e-05, "loss": 0.0045, "step": 35760 }, { "grad_norm": 0.13309194147586823, "learning_rate": 7.627420779099338e-05, "loss": 0.0053, "step": 35770 }, { "grad_norm": 0.08736526966094971, "learning_rate": 7.626013857739711e-05, "loss": 0.0057, "step": 35780 }, { "grad_norm": 0.12095960974693298, "learning_rate": 7.624606649203158e-05, "loss": 0.0051, "step": 35790 }, { "grad_norm": 0.1281832754611969, "learning_rate": 7.623199153643569e-05, "loss": 0.0052, "step": 35800 }, { "grad_norm": 0.11573582142591476, "learning_rate": 7.621791371214863e-05, "loss": 0.0065, "step": 35810 }, { "grad_norm": 0.12228931486606598, "learning_rate": 7.620383302070995e-05, "loss": 0.0058, "step": 35820 }, { "grad_norm": 0.09728898853063583, "learning_rate": 7.61897494636595e-05, "loss": 0.0047, "step": 35830 }, { "grad_norm": 0.07476309686899185, "learning_rate": 7.617566304253739e-05, "loss": 0.0053, "step": 35840 }, { "grad_norm": 0.11349017173051834, "learning_rate": 7.616157375888416e-05, "loss": 0.0074, "step": 35850 }, { "grad_norm": 0.13441511988639832, "learning_rate": 7.614748161424053e-05, "loss": 0.005, "step": 35860 }, { "grad_norm": 0.1059776023030281, "learning_rate": 7.613338661014763e-05, "loss": 0.006, "step": 35870 }, { "grad_norm": 0.09975942224264145, "learning_rate": 7.611928874814686e-05, "loss": 0.0041, "step": 35880 }, { "grad_norm": 0.12200814485549927, "learning_rate": 7.610518802977996e-05, "loss": 0.0047, "step": 35890 }, { "grad_norm": 0.09865628927946091, "learning_rate": 7.609108445658893e-05, "loss": 0.0048, "step": 35900 }, { "grad_norm": 0.09786941856145859, "learning_rate": 7.607697803011612e-05, "loss": 0.0062, "step": 35910 }, { "grad_norm": 0.20139718055725098, "learning_rate": 7.606286875190421e-05, "loss": 0.0054, "step": 35920 }, { "grad_norm": 0.16330064833164215, "learning_rate": 7.604875662349617e-05, "loss": 0.0058, "step": 35930 }, { "grad_norm": 0.12592442333698273, "learning_rate": 7.603464164643525e-05, "loss": 0.0059, "step": 35940 }, { "grad_norm": 0.12912553548812866, "learning_rate": 7.602052382226507e-05, "loss": 0.0082, "step": 35950 }, { "grad_norm": 0.13053633272647858, "learning_rate": 7.600640315252954e-05, "loss": 0.0057, "step": 35960 }, { "grad_norm": 0.1154513955116272, "learning_rate": 7.599227963877284e-05, "loss": 0.0062, "step": 35970 }, { "grad_norm": 0.112571582198143, "learning_rate": 7.597815328253954e-05, "loss": 0.0074, "step": 35980 }, { "grad_norm": 0.09559289366006851, "learning_rate": 7.596402408537443e-05, "loss": 0.0059, "step": 35990 }, { "grad_norm": 0.11912091821432114, "learning_rate": 7.59498920488227e-05, "loss": 0.0077, "step": 36000 }, { "grad_norm": 0.10582098364830017, "learning_rate": 7.593575717442979e-05, "loss": 0.0052, "step": 36010 }, { "grad_norm": 0.09590081125497818, "learning_rate": 7.592161946374147e-05, "loss": 0.0054, "step": 36020 }, { "grad_norm": 0.07436007261276245, "learning_rate": 7.590747891830381e-05, "loss": 0.0049, "step": 36030 }, { "grad_norm": 0.11864540725946426, "learning_rate": 7.58933355396632e-05, "loss": 0.0046, "step": 36040 }, { "grad_norm": 0.06931822001934052, "learning_rate": 7.587918932936636e-05, "loss": 0.0043, "step": 36050 }, { "grad_norm": 0.05944841355085373, "learning_rate": 7.586504028896028e-05, "loss": 0.0038, "step": 36060 }, { "grad_norm": 0.07462581247091293, "learning_rate": 7.585088841999228e-05, "loss": 0.0045, "step": 36070 }, { "grad_norm": 0.04961618408560753, "learning_rate": 7.583673372400999e-05, "loss": 0.0051, "step": 36080 }, { "grad_norm": 0.08022892475128174, "learning_rate": 7.582257620256134e-05, "loss": 0.0045, "step": 36090 }, { "grad_norm": 0.08618222177028656, "learning_rate": 7.580841585719458e-05, "loss": 0.0041, "step": 36100 }, { "grad_norm": 0.08898710459470749, "learning_rate": 7.579425268945825e-05, "loss": 0.0042, "step": 36110 }, { "grad_norm": 0.1145598292350769, "learning_rate": 7.578008670090127e-05, "loss": 0.0042, "step": 36120 }, { "grad_norm": 0.1227957159280777, "learning_rate": 7.576591789307272e-05, "loss": 0.005, "step": 36130 }, { "grad_norm": 0.08214563131332397, "learning_rate": 7.575174626752216e-05, "loss": 0.004, "step": 36140 }, { "grad_norm": 0.08206033706665039, "learning_rate": 7.573757182579934e-05, "loss": 0.0043, "step": 36150 }, { "grad_norm": 0.07835553586483002, "learning_rate": 7.572339456945435e-05, "loss": 0.0039, "step": 36160 }, { "grad_norm": 0.11212866753339767, "learning_rate": 7.570921450003762e-05, "loss": 0.0051, "step": 36170 }, { "grad_norm": 0.12138693779706955, "learning_rate": 7.569503161909983e-05, "loss": 0.0051, "step": 36180 }, { "grad_norm": 0.05600355938076973, "learning_rate": 7.568084592819202e-05, "loss": 0.0044, "step": 36190 }, { "grad_norm": 0.18624451756477356, "learning_rate": 7.566665742886551e-05, "loss": 0.0051, "step": 36200 }, { "grad_norm": 0.16393782198429108, "learning_rate": 7.565246612267194e-05, "loss": 0.0058, "step": 36210 }, { "grad_norm": 0.11566793918609619, "learning_rate": 7.563827201116325e-05, "loss": 0.0056, "step": 36220 }, { "grad_norm": 0.17156380414962769, "learning_rate": 7.562407509589164e-05, "loss": 0.0067, "step": 36230 }, { "grad_norm": 0.17180082201957703, "learning_rate": 7.560987537840973e-05, "loss": 0.0085, "step": 36240 }, { "grad_norm": 0.1354171335697174, "learning_rate": 7.559567286027036e-05, "loss": 0.0078, "step": 36250 }, { "grad_norm": 0.09970258921384811, "learning_rate": 7.558146754302668e-05, "loss": 0.008, "step": 36260 }, { "grad_norm": 0.19898900389671326, "learning_rate": 7.556725942823217e-05, "loss": 0.0095, "step": 36270 }, { "grad_norm": 0.14842204749584198, "learning_rate": 7.555304851744061e-05, "loss": 0.0096, "step": 36280 }, { "grad_norm": 0.14961832761764526, "learning_rate": 7.553883481220608e-05, "loss": 0.0095, "step": 36290 }, { "grad_norm": 0.13289716839790344, "learning_rate": 7.552461831408298e-05, "loss": 0.0069, "step": 36300 }, { "grad_norm": 0.13525570929050446, "learning_rate": 7.551039902462599e-05, "loss": 0.0066, "step": 36310 }, { "grad_norm": 0.09459882229566574, "learning_rate": 7.549617694539011e-05, "loss": 0.0054, "step": 36320 }, { "grad_norm": 0.11249185353517532, "learning_rate": 7.548195207793066e-05, "loss": 0.0065, "step": 36330 }, { "grad_norm": 0.10073508322238922, "learning_rate": 7.546772442380323e-05, "loss": 0.0051, "step": 36340 }, { "grad_norm": 0.11760455369949341, "learning_rate": 7.545349398456376e-05, "loss": 0.0048, "step": 36350 }, { "grad_norm": 0.20865006744861603, "learning_rate": 7.543926076176845e-05, "loss": 0.0046, "step": 36360 }, { "grad_norm": 0.16931045055389404, "learning_rate": 7.542502475697385e-05, "loss": 0.0075, "step": 36370 }, { "grad_norm": 0.09483250975608826, "learning_rate": 7.541078597173675e-05, "loss": 0.0064, "step": 36380 }, { "grad_norm": 0.11600087583065033, "learning_rate": 7.539654440761431e-05, "loss": 0.0077, "step": 36390 }, { "grad_norm": 0.1125088781118393, "learning_rate": 7.538230006616395e-05, "loss": 0.0088, "step": 36400 }, { "grad_norm": 0.10311995446681976, "learning_rate": 7.536805294894342e-05, "loss": 0.0074, "step": 36410 }, { "grad_norm": 0.12852826714515686, "learning_rate": 7.535380305751076e-05, "loss": 0.0066, "step": 36420 }, { "grad_norm": 0.079328753054142, "learning_rate": 7.533955039342431e-05, "loss": 0.0057, "step": 36430 }, { "grad_norm": 0.11760909855365753, "learning_rate": 7.532529495824274e-05, "loss": 0.0056, "step": 36440 }, { "grad_norm": 0.10781034082174301, "learning_rate": 7.531103675352497e-05, "loss": 0.0046, "step": 36450 }, { "grad_norm": 0.10564000904560089, "learning_rate": 7.529677578083029e-05, "loss": 0.0056, "step": 36460 }, { "grad_norm": 0.0651494562625885, "learning_rate": 7.528251204171823e-05, "loss": 0.0037, "step": 36470 }, { "grad_norm": 0.09571744501590729, "learning_rate": 7.526824553774867e-05, "loss": 0.0059, "step": 36480 }, { "grad_norm": 0.10168573260307312, "learning_rate": 7.525397627048178e-05, "loss": 0.0084, "step": 36490 }, { "grad_norm": 0.1097121313214302, "learning_rate": 7.523970424147802e-05, "loss": 0.0073, "step": 36500 }, { "grad_norm": 0.11912877857685089, "learning_rate": 7.522542945229813e-05, "loss": 0.0065, "step": 36510 }, { "grad_norm": 0.10111401975154877, "learning_rate": 7.521115190450321e-05, "loss": 0.0066, "step": 36520 }, { "grad_norm": 0.12064261734485626, "learning_rate": 7.519687159965462e-05, "loss": 0.007, "step": 36530 }, { "grad_norm": 0.1226578950881958, "learning_rate": 7.518258853931403e-05, "loss": 0.0047, "step": 36540 }, { "grad_norm": 0.12544631958007812, "learning_rate": 7.516830272504342e-05, "loss": 0.0055, "step": 36550 }, { "grad_norm": 0.13756465911865234, "learning_rate": 7.515401415840509e-05, "loss": 0.0045, "step": 36560 }, { "grad_norm": 0.0986713245511055, "learning_rate": 7.513972284096155e-05, "loss": 0.0041, "step": 36570 }, { "grad_norm": 0.07171263545751572, "learning_rate": 7.512542877427574e-05, "loss": 0.0042, "step": 36580 }, { "grad_norm": 0.0754050686955452, "learning_rate": 7.51111319599108e-05, "loss": 0.0044, "step": 36590 }, { "grad_norm": 0.06550586223602295, "learning_rate": 7.509683239943024e-05, "loss": 0.0038, "step": 36600 }, { "grad_norm": 0.05805697292089462, "learning_rate": 7.50825300943978e-05, "loss": 0.0035, "step": 36610 }, { "grad_norm": 0.06238376721739769, "learning_rate": 7.50682250463776e-05, "loss": 0.0049, "step": 36620 }, { "grad_norm": 0.07591508328914642, "learning_rate": 7.5053917256934e-05, "loss": 0.0054, "step": 36630 }, { "grad_norm": 0.09505780786275864, "learning_rate": 7.503960672763166e-05, "loss": 0.006, "step": 36640 }, { "grad_norm": 0.12493111193180084, "learning_rate": 7.502529346003559e-05, "loss": 0.0049, "step": 36650 }, { "grad_norm": 0.11141765862703323, "learning_rate": 7.501097745571106e-05, "loss": 0.0052, "step": 36660 }, { "grad_norm": 0.1083204373717308, "learning_rate": 7.499665871622361e-05, "loss": 0.0049, "step": 36670 }, { "grad_norm": 0.09282667934894562, "learning_rate": 7.498233724313917e-05, "loss": 0.0046, "step": 36680 }, { "grad_norm": 0.10959257185459137, "learning_rate": 7.496801303802387e-05, "loss": 0.0043, "step": 36690 }, { "grad_norm": 0.15302278101444244, "learning_rate": 7.495368610244423e-05, "loss": 0.0051, "step": 36700 }, { "grad_norm": 0.09271342307329178, "learning_rate": 7.493935643796697e-05, "loss": 0.0071, "step": 36710 }, { "grad_norm": 0.10802727937698364, "learning_rate": 7.49250240461592e-05, "loss": 0.0044, "step": 36720 }, { "grad_norm": 0.07557275146245956, "learning_rate": 7.491068892858826e-05, "loss": 0.0049, "step": 36730 }, { "grad_norm": 0.07043886929750443, "learning_rate": 7.489635108682184e-05, "loss": 0.0043, "step": 36740 }, { "grad_norm": 0.07414502650499344, "learning_rate": 7.48820105224279e-05, "loss": 0.0044, "step": 36750 }, { "grad_norm": 0.08251111209392548, "learning_rate": 7.48676672369747e-05, "loss": 0.0042, "step": 36760 }, { "grad_norm": 0.13539491593837738, "learning_rate": 7.485332123203079e-05, "loss": 0.0063, "step": 36770 }, { "grad_norm": 0.10504865646362305, "learning_rate": 7.483897250916504e-05, "loss": 0.0049, "step": 36780 }, { "grad_norm": 0.09925857186317444, "learning_rate": 7.48246210699466e-05, "loss": 0.006, "step": 36790 }, { "grad_norm": 0.11834926903247833, "learning_rate": 7.481026691594492e-05, "loss": 0.0051, "step": 36800 }, { "grad_norm": 0.12867382168769836, "learning_rate": 7.479591004872975e-05, "loss": 0.0078, "step": 36810 }, { "grad_norm": 0.1125229224562645, "learning_rate": 7.478155046987114e-05, "loss": 0.0047, "step": 36820 }, { "grad_norm": 0.11952172964811325, "learning_rate": 7.476718818093944e-05, "loss": 0.0043, "step": 36830 }, { "grad_norm": 0.0760870948433876, "learning_rate": 7.475282318350527e-05, "loss": 0.0051, "step": 36840 }, { "grad_norm": 0.09057965874671936, "learning_rate": 7.473845547913957e-05, "loss": 0.0046, "step": 36850 }, { "grad_norm": 0.1357295960187912, "learning_rate": 7.472408506941357e-05, "loss": 0.0061, "step": 36860 }, { "grad_norm": 0.12240903824567795, "learning_rate": 7.470971195589881e-05, "loss": 0.0061, "step": 36870 }, { "grad_norm": 0.08636162430047989, "learning_rate": 7.46953361401671e-05, "loss": 0.0056, "step": 36880 }, { "grad_norm": 0.2153339982032776, "learning_rate": 7.468095762379055e-05, "loss": 0.0068, "step": 36890 }, { "grad_norm": 0.1375708132982254, "learning_rate": 7.466657640834158e-05, "loss": 0.0084, "step": 36900 }, { "grad_norm": 0.08799046277999878, "learning_rate": 7.46521924953929e-05, "loss": 0.0072, "step": 36910 }, { "grad_norm": 0.14815647900104523, "learning_rate": 7.463780588651752e-05, "loss": 0.0065, "step": 36920 }, { "grad_norm": 0.10668115317821503, "learning_rate": 7.462341658328873e-05, "loss": 0.0057, "step": 36930 }, { "grad_norm": 0.09313502907752991, "learning_rate": 7.460902458728012e-05, "loss": 0.0066, "step": 36940 }, { "grad_norm": 0.1223784014582634, "learning_rate": 7.459462990006558e-05, "loss": 0.0057, "step": 36950 }, { "grad_norm": 0.09831686317920685, "learning_rate": 7.458023252321929e-05, "loss": 0.0056, "step": 36960 }, { "grad_norm": 0.11423353105783463, "learning_rate": 7.456583245831573e-05, "loss": 0.0051, "step": 36970 }, { "grad_norm": 0.13177718222141266, "learning_rate": 7.455142970692968e-05, "loss": 0.0054, "step": 36980 }, { "grad_norm": 0.09363314509391785, "learning_rate": 7.453702427063615e-05, "loss": 0.0044, "step": 36990 }, { "grad_norm": 0.08395678550004959, "learning_rate": 7.452261615101057e-05, "loss": 0.004, "step": 37000 }, { "grad_norm": 0.0855245515704155, "learning_rate": 7.450820534962856e-05, "loss": 0.0046, "step": 37010 }, { "grad_norm": 0.22986863553524017, "learning_rate": 7.449379186806604e-05, "loss": 0.0045, "step": 37020 }, { "grad_norm": 0.0980730727314949, "learning_rate": 7.44793757078993e-05, "loss": 0.0046, "step": 37030 }, { "grad_norm": 0.09350984543561935, "learning_rate": 7.44649568707048e-05, "loss": 0.0052, "step": 37040 }, { "grad_norm": 0.08571234345436096, "learning_rate": 7.445053535805942e-05, "loss": 0.0046, "step": 37050 }, { "grad_norm": 0.13508710265159607, "learning_rate": 7.443611117154024e-05, "loss": 0.007, "step": 37060 }, { "grad_norm": 0.1183989942073822, "learning_rate": 7.442168431272469e-05, "loss": 0.0068, "step": 37070 }, { "grad_norm": 0.0785166323184967, "learning_rate": 7.440725478319044e-05, "loss": 0.0058, "step": 37080 }, { "grad_norm": 0.11689376831054688, "learning_rate": 7.439282258451553e-05, "loss": 0.006, "step": 37090 }, { "grad_norm": 0.12463273108005524, "learning_rate": 7.43783877182782e-05, "loss": 0.0064, "step": 37100 }, { "grad_norm": 0.17137353122234344, "learning_rate": 7.436395018605707e-05, "loss": 0.0062, "step": 37110 }, { "grad_norm": 0.311290442943573, "learning_rate": 7.434950998943094e-05, "loss": 0.0103, "step": 37120 }, { "grad_norm": 0.15966887772083282, "learning_rate": 7.433506712997902e-05, "loss": 0.0074, "step": 37130 }, { "grad_norm": 0.1179732158780098, "learning_rate": 7.432062160928077e-05, "loss": 0.0052, "step": 37140 }, { "grad_norm": 0.09630043804645538, "learning_rate": 7.430617342891588e-05, "loss": 0.005, "step": 37150 }, { "grad_norm": 0.11632882803678513, "learning_rate": 7.429172259046442e-05, "loss": 0.0061, "step": 37160 }, { "grad_norm": 0.22886213660240173, "learning_rate": 7.42772690955067e-05, "loss": 0.0067, "step": 37170 }, { "grad_norm": 0.09699055552482605, "learning_rate": 7.426281294562332e-05, "loss": 0.0059, "step": 37180 }, { "grad_norm": 0.11948683857917786, "learning_rate": 7.424835414239519e-05, "loss": 0.0075, "step": 37190 }, { "grad_norm": 0.0926937684416771, "learning_rate": 7.423389268740352e-05, "loss": 0.0075, "step": 37200 }, { "grad_norm": 0.1002437174320221, "learning_rate": 7.42194285822298e-05, "loss": 0.0058, "step": 37210 }, { "grad_norm": 0.14526686072349548, "learning_rate": 7.420496182845576e-05, "loss": 0.0053, "step": 37220 }, { "grad_norm": 0.12417765706777573, "learning_rate": 7.419049242766348e-05, "loss": 0.0056, "step": 37230 }, { "grad_norm": 0.10447856783866882, "learning_rate": 7.417602038143533e-05, "loss": 0.0051, "step": 37240 }, { "grad_norm": 0.094309501349926, "learning_rate": 7.416154569135393e-05, "loss": 0.0054, "step": 37250 }, { "grad_norm": 0.0775911957025528, "learning_rate": 7.414706835900222e-05, "loss": 0.0056, "step": 37260 }, { "grad_norm": 0.09395918995141983, "learning_rate": 7.413258838596345e-05, "loss": 0.006, "step": 37270 }, { "grad_norm": 0.11794045567512512, "learning_rate": 7.411810577382106e-05, "loss": 0.0052, "step": 37280 }, { "grad_norm": 0.12113088369369507, "learning_rate": 7.410362052415889e-05, "loss": 0.0068, "step": 37290 }, { "grad_norm": 0.10773523151874542, "learning_rate": 7.408913263856102e-05, "loss": 0.0058, "step": 37300 }, { "grad_norm": 0.07559609413146973, "learning_rate": 7.407464211861183e-05, "loss": 0.0045, "step": 37310 }, { "grad_norm": 0.0613529197871685, "learning_rate": 7.406014896589597e-05, "loss": 0.0044, "step": 37320 }, { "grad_norm": 0.0778612494468689, "learning_rate": 7.404565318199838e-05, "loss": 0.0048, "step": 37330 }, { "grad_norm": 0.10989590734243393, "learning_rate": 7.403115476850432e-05, "loss": 0.0041, "step": 37340 }, { "grad_norm": 0.07487493753433228, "learning_rate": 7.401665372699932e-05, "loss": 0.0049, "step": 37350 }, { "grad_norm": 0.09185470640659332, "learning_rate": 7.400215005906916e-05, "loss": 0.0046, "step": 37360 }, { "grad_norm": 0.1434205174446106, "learning_rate": 7.398764376629996e-05, "loss": 0.0053, "step": 37370 }, { "grad_norm": 0.12767907977104187, "learning_rate": 7.39731348502781e-05, "loss": 0.0046, "step": 37380 }, { "grad_norm": 0.12185750901699066, "learning_rate": 7.395862331259028e-05, "loss": 0.0051, "step": 37390 }, { "grad_norm": 0.08455272018909454, "learning_rate": 7.394410915482341e-05, "loss": 0.0049, "step": 37400 }, { "grad_norm": 0.13516341149806976, "learning_rate": 7.392959237856476e-05, "loss": 0.007, "step": 37410 }, { "grad_norm": 0.08721170574426651, "learning_rate": 7.391507298540187e-05, "loss": 0.0055, "step": 37420 }, { "grad_norm": 0.08251463621854782, "learning_rate": 7.390055097692257e-05, "loss": 0.0045, "step": 37430 }, { "grad_norm": 0.1260903924703598, "learning_rate": 7.388602635471494e-05, "loss": 0.0075, "step": 37440 }, { "grad_norm": 0.11637125164270401, "learning_rate": 7.387149912036738e-05, "loss": 0.0084, "step": 37450 }, { "grad_norm": 0.094977967441082, "learning_rate": 7.385696927546854e-05, "loss": 0.0058, "step": 37460 }, { "grad_norm": 0.0975615605711937, "learning_rate": 7.384243682160743e-05, "loss": 0.0045, "step": 37470 }, { "grad_norm": 0.08785359561443329, "learning_rate": 7.382790176037329e-05, "loss": 0.0048, "step": 37480 }, { "grad_norm": 0.08682481944561005, "learning_rate": 7.381336409335562e-05, "loss": 0.0043, "step": 37490 }, { "grad_norm": 0.09612098336219788, "learning_rate": 7.379882382214426e-05, "loss": 0.0047, "step": 37500 }, { "grad_norm": 0.07751433551311493, "learning_rate": 7.378428094832931e-05, "loss": 0.0065, "step": 37510 }, { "grad_norm": 0.09585877507925034, "learning_rate": 7.376973547350114e-05, "loss": 0.005, "step": 37520 }, { "grad_norm": 0.10235542058944702, "learning_rate": 7.375518739925046e-05, "loss": 0.0044, "step": 37530 }, { "grad_norm": 0.11336753517389297, "learning_rate": 7.374063672716818e-05, "loss": 0.0045, "step": 37540 }, { "grad_norm": 0.19219845533370972, "learning_rate": 7.372608345884558e-05, "loss": 0.0041, "step": 37550 }, { "grad_norm": 0.10410257428884506, "learning_rate": 7.371152759587415e-05, "loss": 0.0045, "step": 37560 }, { "grad_norm": 0.0842386782169342, "learning_rate": 7.369696913984572e-05, "loss": 0.0057, "step": 37570 }, { "grad_norm": 0.09004264324903488, "learning_rate": 7.368240809235237e-05, "loss": 0.0045, "step": 37580 }, { "grad_norm": 0.06138245388865471, "learning_rate": 7.366784445498649e-05, "loss": 0.0056, "step": 37590 }, { "grad_norm": 0.12820883095264435, "learning_rate": 7.36532782293407e-05, "loss": 0.0074, "step": 37600 }, { "grad_norm": 0.13972650468349457, "learning_rate": 7.363870941700797e-05, "loss": 0.005, "step": 37610 }, { "grad_norm": 0.11619475483894348, "learning_rate": 7.362413801958152e-05, "loss": 0.0059, "step": 37620 }, { "grad_norm": 0.10668714344501495, "learning_rate": 7.360956403865486e-05, "loss": 0.0067, "step": 37630 }, { "grad_norm": 0.1370115727186203, "learning_rate": 7.359498747582177e-05, "loss": 0.0054, "step": 37640 }, { "grad_norm": 0.11405016481876373, "learning_rate": 7.358040833267634e-05, "loss": 0.0052, "step": 37650 }, { "grad_norm": 0.10832444578409195, "learning_rate": 7.356582661081288e-05, "loss": 0.0056, "step": 37660 }, { "grad_norm": 0.09252731502056122, "learning_rate": 7.355124231182607e-05, "loss": 0.0041, "step": 37670 }, { "grad_norm": 0.1073535680770874, "learning_rate": 7.353665543731083e-05, "loss": 0.005, "step": 37680 }, { "grad_norm": 0.15792056918144226, "learning_rate": 7.352206598886229e-05, "loss": 0.0071, "step": 37690 }, { "grad_norm": 0.189960777759552, "learning_rate": 7.350747396807601e-05, "loss": 0.0073, "step": 37700 }, { "grad_norm": 0.13155430555343628, "learning_rate": 7.349287937654772e-05, "loss": 0.0073, "step": 37710 }, { "grad_norm": 0.12868458032608032, "learning_rate": 7.347828221587345e-05, "loss": 0.0048, "step": 37720 }, { "grad_norm": 0.08728650957345963, "learning_rate": 7.346368248764955e-05, "loss": 0.0053, "step": 37730 }, { "grad_norm": 0.10835711658000946, "learning_rate": 7.34490801934726e-05, "loss": 0.0054, "step": 37740 }, { "grad_norm": 0.1336643546819687, "learning_rate": 7.343447533493947e-05, "loss": 0.0068, "step": 37750 }, { "grad_norm": 0.14709676802158356, "learning_rate": 7.341986791364739e-05, "loss": 0.0063, "step": 37760 }, { "grad_norm": 0.17516520619392395, "learning_rate": 7.340525793119374e-05, "loss": 0.0076, "step": 37770 }, { "grad_norm": 0.08168023079633713, "learning_rate": 7.33906453891763e-05, "loss": 0.0068, "step": 37780 }, { "grad_norm": 0.08096280694007874, "learning_rate": 7.337603028919303e-05, "loss": 0.0048, "step": 37790 }, { "grad_norm": 0.09375683963298798, "learning_rate": 7.336141263284225e-05, "loss": 0.0043, "step": 37800 }, { "grad_norm": 0.13460522890090942, "learning_rate": 7.334679242172249e-05, "loss": 0.0055, "step": 37810 }, { "grad_norm": 0.09118868410587311, "learning_rate": 7.33321696574326e-05, "loss": 0.0061, "step": 37820 }, { "grad_norm": 0.10974306613206863, "learning_rate": 7.331754434157174e-05, "loss": 0.006, "step": 37830 }, { "grad_norm": 0.09293379634618759, "learning_rate": 7.330291647573928e-05, "loss": 0.0047, "step": 37840 }, { "grad_norm": 0.059371139854192734, "learning_rate": 7.32882860615349e-05, "loss": 0.0053, "step": 37850 }, { "grad_norm": 0.1002967581152916, "learning_rate": 7.327365310055858e-05, "loss": 0.0051, "step": 37860 }, { "grad_norm": 0.08765405416488647, "learning_rate": 7.325901759441055e-05, "loss": 0.0055, "step": 37870 }, { "grad_norm": 0.16164301335811615, "learning_rate": 7.324437954469132e-05, "loss": 0.0057, "step": 37880 }, { "grad_norm": 0.08271472901105881, "learning_rate": 7.32297389530017e-05, "loss": 0.0042, "step": 37890 }, { "grad_norm": 0.09084325283765793, "learning_rate": 7.321509582094274e-05, "loss": 0.0043, "step": 37900 }, { "grad_norm": 0.10894922912120819, "learning_rate": 7.320045015011582e-05, "loss": 0.0042, "step": 37910 }, { "grad_norm": 0.08490745723247528, "learning_rate": 7.318580194212253e-05, "loss": 0.0049, "step": 37920 }, { "grad_norm": 0.1162656620144844, "learning_rate": 7.317115119856483e-05, "loss": 0.0049, "step": 37930 }, { "grad_norm": 0.08820950984954834, "learning_rate": 7.315649792104485e-05, "loss": 0.004, "step": 37940 }, { "grad_norm": 0.11206929385662079, "learning_rate": 7.314184211116507e-05, "loss": 0.0045, "step": 37950 }, { "grad_norm": 0.07901794463396072, "learning_rate": 7.312718377052823e-05, "loss": 0.0045, "step": 37960 }, { "grad_norm": 0.08468946069478989, "learning_rate": 7.311252290073737e-05, "loss": 0.0053, "step": 37970 }, { "grad_norm": 0.0903230756521225, "learning_rate": 7.309785950339572e-05, "loss": 0.0058, "step": 37980 }, { "grad_norm": 0.11183317005634308, "learning_rate": 7.308319358010692e-05, "loss": 0.0036, "step": 37990 }, { "grad_norm": 0.10236501693725586, "learning_rate": 7.306852513247474e-05, "loss": 0.0052, "step": 38000 }, { "grad_norm": 0.13119827210903168, "learning_rate": 7.305385416210336e-05, "loss": 0.0046, "step": 38010 }, { "grad_norm": 0.09080056846141815, "learning_rate": 7.303918067059711e-05, "loss": 0.0058, "step": 38020 }, { "grad_norm": 0.0852225199341774, "learning_rate": 7.302450465956075e-05, "loss": 0.0055, "step": 38030 }, { "grad_norm": 0.11722573637962341, "learning_rate": 7.300982613059914e-05, "loss": 0.0073, "step": 38040 }, { "grad_norm": 0.1962345540523529, "learning_rate": 7.299514508531757e-05, "loss": 0.0065, "step": 38050 }, { "grad_norm": 0.10535432398319244, "learning_rate": 7.298046152532148e-05, "loss": 0.0069, "step": 38060 }, { "grad_norm": 0.1084538921713829, "learning_rate": 7.296577545221668e-05, "loss": 0.0059, "step": 38070 }, { "grad_norm": 0.12271111458539963, "learning_rate": 7.29510868676092e-05, "loss": 0.0061, "step": 38080 }, { "grad_norm": 0.0871964618563652, "learning_rate": 7.293639577310538e-05, "loss": 0.0056, "step": 38090 }, { "grad_norm": 0.07573843002319336, "learning_rate": 7.292170217031179e-05, "loss": 0.0054, "step": 38100 }, { "grad_norm": 0.07032602280378342, "learning_rate": 7.290700606083532e-05, "loss": 0.0048, "step": 38110 }, { "grad_norm": 0.10769537836313248, "learning_rate": 7.28923074462831e-05, "loss": 0.0053, "step": 38120 }, { "grad_norm": 0.07652284950017929, "learning_rate": 7.287760632826255e-05, "loss": 0.0054, "step": 38130 }, { "grad_norm": 0.0846104696393013, "learning_rate": 7.286290270838139e-05, "loss": 0.0044, "step": 38140 }, { "grad_norm": 0.07449191063642502, "learning_rate": 7.284819658824756e-05, "loss": 0.005, "step": 38150 }, { "grad_norm": 0.11342623829841614, "learning_rate": 7.283348796946929e-05, "loss": 0.0054, "step": 38160 }, { "grad_norm": 0.11389026045799255, "learning_rate": 7.28187768536551e-05, "loss": 0.0046, "step": 38170 }, { "grad_norm": 0.07586803287267685, "learning_rate": 7.280406324241379e-05, "loss": 0.0046, "step": 38180 }, { "grad_norm": 0.0993025004863739, "learning_rate": 7.278934713735441e-05, "loss": 0.0057, "step": 38190 }, { "grad_norm": 0.1267949491739273, "learning_rate": 7.277462854008629e-05, "loss": 0.005, "step": 38200 }, { "grad_norm": 0.0952286571264267, "learning_rate": 7.275990745221904e-05, "loss": 0.0046, "step": 38210 }, { "grad_norm": 0.10441350936889648, "learning_rate": 7.274518387536252e-05, "loss": 0.0044, "step": 38220 }, { "grad_norm": 0.07244034111499786, "learning_rate": 7.273045781112689e-05, "loss": 0.0039, "step": 38230 }, { "grad_norm": 0.08650526404380798, "learning_rate": 7.271572926112257e-05, "loss": 0.0035, "step": 38240 }, { "grad_norm": 0.09379247575998306, "learning_rate": 7.270099822696024e-05, "loss": 0.0041, "step": 38250 }, { "grad_norm": 0.09365924447774887, "learning_rate": 7.26862647102509e-05, "loss": 0.0041, "step": 38260 }, { "grad_norm": 0.12586921453475952, "learning_rate": 7.267152871260573e-05, "loss": 0.0049, "step": 38270 }, { "grad_norm": 0.11922571808099747, "learning_rate": 7.265679023563629e-05, "loss": 0.0044, "step": 38280 }, { "grad_norm": 0.10771964490413666, "learning_rate": 7.264204928095433e-05, "loss": 0.0055, "step": 38290 }, { "grad_norm": 0.1117173358798027, "learning_rate": 7.262730585017188e-05, "loss": 0.0056, "step": 38300 }, { "grad_norm": 0.10917726159095764, "learning_rate": 7.261255994490131e-05, "loss": 0.0056, "step": 38310 }, { "grad_norm": 0.08607357740402222, "learning_rate": 7.259781156675517e-05, "loss": 0.0048, "step": 38320 }, { "grad_norm": 0.08081947267055511, "learning_rate": 7.258306071734633e-05, "loss": 0.0043, "step": 38330 }, { "grad_norm": 0.09841153770685196, "learning_rate": 7.256830739828792e-05, "loss": 0.0047, "step": 38340 }, { "grad_norm": 0.09428350627422333, "learning_rate": 7.255355161119336e-05, "loss": 0.0041, "step": 38350 }, { "grad_norm": 0.10616182535886765, "learning_rate": 7.253879335767628e-05, "loss": 0.0035, "step": 38360 }, { "grad_norm": 0.05834103748202324, "learning_rate": 7.252403263935067e-05, "loss": 0.0034, "step": 38370 }, { "grad_norm": 0.05952080711722374, "learning_rate": 7.25092694578307e-05, "loss": 0.0032, "step": 38380 }, { "grad_norm": 0.11161050945520401, "learning_rate": 7.249450381473087e-05, "loss": 0.0042, "step": 38390 }, { "grad_norm": 0.09565921127796173, "learning_rate": 7.247973571166593e-05, "loss": 0.0042, "step": 38400 }, { "grad_norm": 0.10461454093456268, "learning_rate": 7.246496515025089e-05, "loss": 0.0036, "step": 38410 }, { "grad_norm": 0.08633881062269211, "learning_rate": 7.245019213210101e-05, "loss": 0.0036, "step": 38420 }, { "grad_norm": 0.13611234724521637, "learning_rate": 7.24354166588319e-05, "loss": 0.0045, "step": 38430 }, { "grad_norm": 0.08432678133249283, "learning_rate": 7.242063873205935e-05, "loss": 0.004, "step": 38440 }, { "grad_norm": 0.10889036953449249, "learning_rate": 7.240585835339946e-05, "loss": 0.0057, "step": 38450 }, { "grad_norm": 0.1383892148733139, "learning_rate": 7.239107552446857e-05, "loss": 0.0051, "step": 38460 }, { "grad_norm": 0.08435249328613281, "learning_rate": 7.237629024688334e-05, "loss": 0.0044, "step": 38470 }, { "grad_norm": 0.09347932040691376, "learning_rate": 7.236150252226064e-05, "loss": 0.0046, "step": 38480 }, { "grad_norm": 0.0825258269906044, "learning_rate": 7.234671235221765e-05, "loss": 0.0044, "step": 38490 }, { "grad_norm": 0.062382910400629044, "learning_rate": 7.233191973837179e-05, "loss": 0.0035, "step": 38500 }, { "grad_norm": 0.08303957432508469, "learning_rate": 7.231712468234078e-05, "loss": 0.0042, "step": 38510 }, { "grad_norm": 0.13667790591716766, "learning_rate": 7.230232718574254e-05, "loss": 0.0038, "step": 38520 }, { "grad_norm": 0.08884484320878983, "learning_rate": 7.228752725019535e-05, "loss": 0.0042, "step": 38530 }, { "grad_norm": 0.07711353152990341, "learning_rate": 7.227272487731769e-05, "loss": 0.0041, "step": 38540 }, { "grad_norm": 0.09324522316455841, "learning_rate": 7.225792006872831e-05, "loss": 0.0044, "step": 38550 }, { "grad_norm": 0.12943483889102936, "learning_rate": 7.224311282604628e-05, "loss": 0.0057, "step": 38560 }, { "grad_norm": 0.09534357488155365, "learning_rate": 7.222830315089085e-05, "loss": 0.0053, "step": 38570 }, { "grad_norm": 0.0834885984659195, "learning_rate": 7.22134910448816e-05, "loss": 0.0042, "step": 38580 }, { "grad_norm": 0.1929347664117813, "learning_rate": 7.219867650963839e-05, "loss": 0.005, "step": 38590 }, { "grad_norm": 0.11691384762525558, "learning_rate": 7.21838595467813e-05, "loss": 0.005, "step": 38600 }, { "grad_norm": 0.1223309114575386, "learning_rate": 7.216904015793067e-05, "loss": 0.0054, "step": 38610 }, { "grad_norm": 0.10824853181838989, "learning_rate": 7.215421834470713e-05, "loss": 0.0045, "step": 38620 }, { "grad_norm": 0.10229938477277756, "learning_rate": 7.213939410873158e-05, "loss": 0.0037, "step": 38630 }, { "grad_norm": 0.11408650130033493, "learning_rate": 7.212456745162519e-05, "loss": 0.0046, "step": 38640 }, { "grad_norm": 0.07873186469078064, "learning_rate": 7.210973837500937e-05, "loss": 0.0049, "step": 38650 }, { "grad_norm": 0.07384326308965683, "learning_rate": 7.209490688050578e-05, "loss": 0.0047, "step": 38660 }, { "grad_norm": 0.08768939226865768, "learning_rate": 7.208007296973641e-05, "loss": 0.005, "step": 38670 }, { "grad_norm": 0.10341943800449371, "learning_rate": 7.206523664432345e-05, "loss": 0.0053, "step": 38680 }, { "grad_norm": 0.1280244141817093, "learning_rate": 7.205039790588939e-05, "loss": 0.004, "step": 38690 }, { "grad_norm": 0.11019106954336166, "learning_rate": 7.203555675605697e-05, "loss": 0.0047, "step": 38700 }, { "grad_norm": 0.10064315050840378, "learning_rate": 7.202071319644917e-05, "loss": 0.0034, "step": 38710 }, { "grad_norm": 0.08654078841209412, "learning_rate": 7.200586722868932e-05, "loss": 0.0041, "step": 38720 }, { "grad_norm": 0.08034288883209229, "learning_rate": 7.199101885440088e-05, "loss": 0.0047, "step": 38730 }, { "grad_norm": 0.08596089482307434, "learning_rate": 7.197616807520771e-05, "loss": 0.0043, "step": 38740 }, { "grad_norm": 0.12037262320518494, "learning_rate": 7.196131489273381e-05, "loss": 0.0053, "step": 38750 }, { "grad_norm": 0.30717653036117554, "learning_rate": 7.194645930860355e-05, "loss": 0.0058, "step": 38760 }, { "grad_norm": 0.09560498595237732, "learning_rate": 7.19316013244415e-05, "loss": 0.0045, "step": 38770 }, { "grad_norm": 0.17438989877700806, "learning_rate": 7.191674094187248e-05, "loss": 0.0052, "step": 38780 }, { "grad_norm": 0.11700906604528427, "learning_rate": 7.190187816252165e-05, "loss": 0.008, "step": 38790 }, { "grad_norm": 0.132406547665596, "learning_rate": 7.188701298801435e-05, "loss": 0.0054, "step": 38800 }, { "grad_norm": 0.10721051692962646, "learning_rate": 7.18721454199762e-05, "loss": 0.0052, "step": 38810 }, { "grad_norm": 0.09485648572444916, "learning_rate": 7.185727546003312e-05, "loss": 0.0054, "step": 38820 }, { "grad_norm": 0.13525763154029846, "learning_rate": 7.184240310981126e-05, "loss": 0.0073, "step": 38830 }, { "grad_norm": 0.09940583258867264, "learning_rate": 7.182752837093704e-05, "loss": 0.0049, "step": 38840 }, { "grad_norm": 0.09267833083868027, "learning_rate": 7.181265124503711e-05, "loss": 0.0068, "step": 38850 }, { "grad_norm": 0.08992839604616165, "learning_rate": 7.179777173373847e-05, "loss": 0.0045, "step": 38860 }, { "grad_norm": 0.12804627418518066, "learning_rate": 7.178288983866826e-05, "loss": 0.0083, "step": 38870 }, { "grad_norm": 0.10606151819229126, "learning_rate": 7.176800556145397e-05, "loss": 0.0067, "step": 38880 }, { "grad_norm": 0.12180512398481369, "learning_rate": 7.175311890372334e-05, "loss": 0.006, "step": 38890 }, { "grad_norm": 0.06891562789678574, "learning_rate": 7.17382298671043e-05, "loss": 0.0057, "step": 38900 }, { "grad_norm": 0.08976025879383087, "learning_rate": 7.172333845322515e-05, "loss": 0.0054, "step": 38910 }, { "grad_norm": 0.10277153551578522, "learning_rate": 7.170844466371436e-05, "loss": 0.0054, "step": 38920 }, { "grad_norm": 0.07605423778295517, "learning_rate": 7.16935485002007e-05, "loss": 0.0046, "step": 38930 }, { "grad_norm": 0.09432259947061539, "learning_rate": 7.167864996431319e-05, "loss": 0.0042, "step": 38940 }, { "grad_norm": 0.09168468415737152, "learning_rate": 7.166374905768111e-05, "loss": 0.0039, "step": 38950 }, { "grad_norm": 0.0840178057551384, "learning_rate": 7.164884578193401e-05, "loss": 0.0038, "step": 38960 }, { "grad_norm": 0.07987125962972641, "learning_rate": 7.163394013870168e-05, "loss": 0.0032, "step": 38970 }, { "grad_norm": 0.06425914168357849, "learning_rate": 7.161903212961419e-05, "loss": 0.0032, "step": 38980 }, { "grad_norm": 0.0680558830499649, "learning_rate": 7.160412175630183e-05, "loss": 0.0044, "step": 38990 }, { "grad_norm": 0.09405320882797241, "learning_rate": 7.158920902039521e-05, "loss": 0.0044, "step": 39000 }, { "grad_norm": 0.09814127534627914, "learning_rate": 7.157429392352514e-05, "loss": 0.0036, "step": 39010 }, { "grad_norm": 0.08832818269729614, "learning_rate": 7.155937646732275e-05, "loss": 0.0036, "step": 39020 }, { "grad_norm": 0.09004504233598709, "learning_rate": 7.154445665341933e-05, "loss": 0.0044, "step": 39030 }, { "grad_norm": 0.10359065234661102, "learning_rate": 7.152953448344654e-05, "loss": 0.0036, "step": 39040 }, { "grad_norm": 0.08904048055410385, "learning_rate": 7.151460995903624e-05, "loss": 0.0036, "step": 39050 }, { "grad_norm": 0.06235584244132042, "learning_rate": 7.149968308182052e-05, "loss": 0.0038, "step": 39060 }, { "grad_norm": 0.07776930183172226, "learning_rate": 7.14847538534318e-05, "loss": 0.0038, "step": 39070 }, { "grad_norm": 0.07595628499984741, "learning_rate": 7.14698222755027e-05, "loss": 0.0039, "step": 39080 }, { "grad_norm": 0.08773185312747955, "learning_rate": 7.14548883496661e-05, "loss": 0.0043, "step": 39090 }, { "grad_norm": 0.07792533189058304, "learning_rate": 7.143995207755517e-05, "loss": 0.0035, "step": 39100 }, { "grad_norm": 0.08730807155370712, "learning_rate": 7.142501346080333e-05, "loss": 0.004, "step": 39110 }, { "grad_norm": 0.07439957559108734, "learning_rate": 7.141007250104421e-05, "loss": 0.0035, "step": 39120 }, { "grad_norm": 0.07093986868858337, "learning_rate": 7.139512919991176e-05, "loss": 0.0038, "step": 39130 }, { "grad_norm": 0.07137317210435867, "learning_rate": 7.138018355904014e-05, "loss": 0.0044, "step": 39140 }, { "grad_norm": 0.08751524239778519, "learning_rate": 7.13652355800638e-05, "loss": 0.0048, "step": 39150 }, { "grad_norm": 0.08382352441549301, "learning_rate": 7.135028526461743e-05, "loss": 0.0038, "step": 39160 }, { "grad_norm": 0.0839848741889, "learning_rate": 7.133533261433594e-05, "loss": 0.0041, "step": 39170 }, { "grad_norm": 0.08153531700372696, "learning_rate": 7.132037763085457e-05, "loss": 0.0056, "step": 39180 }, { "grad_norm": 0.09464866667985916, "learning_rate": 7.130542031580875e-05, "loss": 0.0043, "step": 39190 }, { "grad_norm": 0.060171179473400116, "learning_rate": 7.12904606708342e-05, "loss": 0.0034, "step": 39200 }, { "grad_norm": 0.0656198039650917, "learning_rate": 7.127549869756687e-05, "loss": 0.0045, "step": 39210 }, { "grad_norm": 0.09557630121707916, "learning_rate": 7.126053439764299e-05, "loss": 0.0048, "step": 39220 }, { "grad_norm": 0.09481224417686462, "learning_rate": 7.124556777269904e-05, "loss": 0.0041, "step": 39230 }, { "grad_norm": 0.08355320990085602, "learning_rate": 7.123059882437174e-05, "loss": 0.0036, "step": 39240 }, { "grad_norm": 0.07980458438396454, "learning_rate": 7.121562755429807e-05, "loss": 0.0036, "step": 39250 }, { "grad_norm": 0.09785211086273193, "learning_rate": 7.120065396411528e-05, "loss": 0.004, "step": 39260 }, { "grad_norm": 0.11218225955963135, "learning_rate": 7.118567805546084e-05, "loss": 0.0047, "step": 39270 }, { "grad_norm": 0.1948263794183731, "learning_rate": 7.117069982997248e-05, "loss": 0.0043, "step": 39280 }, { "grad_norm": 0.07407087087631226, "learning_rate": 7.115571928928825e-05, "loss": 0.0035, "step": 39290 }, { "grad_norm": 0.13144662976264954, "learning_rate": 7.114073643504635e-05, "loss": 0.005, "step": 39300 }, { "grad_norm": 0.0914919301867485, "learning_rate": 7.11257512688853e-05, "loss": 0.0072, "step": 39310 }, { "grad_norm": 0.08797460049390793, "learning_rate": 7.111076379244384e-05, "loss": 0.0053, "step": 39320 }, { "grad_norm": 0.15905168652534485, "learning_rate": 7.109577400736101e-05, "loss": 0.0048, "step": 39330 }, { "grad_norm": 0.07142625004053116, "learning_rate": 7.108078191527605e-05, "loss": 0.0053, "step": 39340 }, { "grad_norm": 0.10373803973197937, "learning_rate": 7.106578751782847e-05, "loss": 0.0052, "step": 39350 }, { "grad_norm": 0.1415674090385437, "learning_rate": 7.105079081665803e-05, "loss": 0.005, "step": 39360 }, { "grad_norm": 0.09007208794355392, "learning_rate": 7.103579181340476e-05, "loss": 0.006, "step": 39370 }, { "grad_norm": 0.07640925794839859, "learning_rate": 7.102079050970893e-05, "loss": 0.005, "step": 39380 }, { "grad_norm": 0.08194774389266968, "learning_rate": 7.100578690721105e-05, "loss": 0.0044, "step": 39390 }, { "grad_norm": 0.07558959722518921, "learning_rate": 7.09907810075519e-05, "loss": 0.005, "step": 39400 }, { "grad_norm": 0.10891687124967575, "learning_rate": 7.097577281237249e-05, "loss": 0.0052, "step": 39410 }, { "grad_norm": 0.1098145917057991, "learning_rate": 7.09607623233141e-05, "loss": 0.0045, "step": 39420 }, { "grad_norm": 0.10095969587564468, "learning_rate": 7.094574954201827e-05, "loss": 0.004, "step": 39430 }, { "grad_norm": 0.09076232463121414, "learning_rate": 7.093073447012675e-05, "loss": 0.0043, "step": 39440 }, { "grad_norm": 0.07503221929073334, "learning_rate": 7.09157171092816e-05, "loss": 0.0042, "step": 39450 }, { "grad_norm": 0.11088468134403229, "learning_rate": 7.090069746112504e-05, "loss": 0.0044, "step": 39460 }, { "grad_norm": 0.10331030189990997, "learning_rate": 7.088567552729965e-05, "loss": 0.0036, "step": 39470 }, { "grad_norm": 0.06576306372880936, "learning_rate": 7.087065130944818e-05, "loss": 0.0039, "step": 39480 }, { "grad_norm": 0.08372274786233902, "learning_rate": 7.085562480921366e-05, "loss": 0.0032, "step": 39490 }, { "grad_norm": 0.06285632401704788, "learning_rate": 7.084059602823937e-05, "loss": 0.0029, "step": 39500 }, { "grad_norm": 0.07381129264831543, "learning_rate": 7.082556496816882e-05, "loss": 0.0037, "step": 39510 }, { "grad_norm": 0.07003135234117508, "learning_rate": 7.081053163064582e-05, "loss": 0.0027, "step": 39520 }, { "grad_norm": 0.09794057905673981, "learning_rate": 7.079549601731434e-05, "loss": 0.0038, "step": 39530 }, { "grad_norm": 0.12076273560523987, "learning_rate": 7.07804581298187e-05, "loss": 0.0051, "step": 39540 }, { "grad_norm": 0.08464652299880981, "learning_rate": 7.07654179698034e-05, "loss": 0.0031, "step": 39550 }, { "grad_norm": 0.08146578818559647, "learning_rate": 7.075037553891321e-05, "loss": 0.0045, "step": 39560 }, { "grad_norm": 0.07190293818712234, "learning_rate": 7.073533083879315e-05, "loss": 0.0054, "step": 39570 }, { "grad_norm": 0.1317155510187149, "learning_rate": 7.072028387108849e-05, "loss": 0.005, "step": 39580 }, { "grad_norm": 0.12339256703853607, "learning_rate": 7.070523463744472e-05, "loss": 0.0067, "step": 39590 }, { "grad_norm": 0.17694276571273804, "learning_rate": 7.069018313950763e-05, "loss": 0.0074, "step": 39600 }, { "grad_norm": 0.15403740108013153, "learning_rate": 7.067512937892322e-05, "loss": 0.0062, "step": 39610 }, { "grad_norm": 0.0880391076207161, "learning_rate": 7.066007335733775e-05, "loss": 0.0057, "step": 39620 }, { "grad_norm": 0.07812756299972534, "learning_rate": 7.064501507639772e-05, "loss": 0.0049, "step": 39630 }, { "grad_norm": 0.09422854334115982, "learning_rate": 7.062995453774987e-05, "loss": 0.0044, "step": 39640 }, { "grad_norm": 0.15159247815608978, "learning_rate": 7.061489174304121e-05, "loss": 0.0049, "step": 39650 }, { "grad_norm": 0.10837015509605408, "learning_rate": 7.0599826693919e-05, "loss": 0.0047, "step": 39660 }, { "grad_norm": 0.07572350651025772, "learning_rate": 7.05847593920307e-05, "loss": 0.0042, "step": 39670 }, { "grad_norm": 0.10906185954809189, "learning_rate": 7.056968983902406e-05, "loss": 0.0051, "step": 39680 }, { "grad_norm": 0.10853605717420578, "learning_rate": 7.055461803654706e-05, "loss": 0.0062, "step": 39690 }, { "grad_norm": 0.09648405760526657, "learning_rate": 7.053954398624794e-05, "loss": 0.006, "step": 39700 }, { "grad_norm": 0.1116512194275856, "learning_rate": 7.052446768977518e-05, "loss": 0.0054, "step": 39710 }, { "grad_norm": 0.0852370411157608, "learning_rate": 7.050938914877748e-05, "loss": 0.006, "step": 39720 }, { "grad_norm": 0.06999856233596802, "learning_rate": 7.049430836490382e-05, "loss": 0.0068, "step": 39730 }, { "grad_norm": 0.10176799446344376, "learning_rate": 7.047922533980342e-05, "loss": 0.0061, "step": 39740 }, { "grad_norm": 0.08730439841747284, "learning_rate": 7.046414007512571e-05, "loss": 0.0063, "step": 39750 }, { "grad_norm": 0.09727261960506439, "learning_rate": 7.044905257252042e-05, "loss": 0.0049, "step": 39760 }, { "grad_norm": 0.08553140610456467, "learning_rate": 7.04339628336375e-05, "loss": 0.0051, "step": 39770 }, { "grad_norm": 0.09493555128574371, "learning_rate": 7.041887086012711e-05, "loss": 0.0047, "step": 39780 }, { "grad_norm": 0.09413812309503555, "learning_rate": 7.040377665363969e-05, "loss": 0.0055, "step": 39790 }, { "grad_norm": 0.08786562830209732, "learning_rate": 7.038868021582594e-05, "loss": 0.0053, "step": 39800 }, { "grad_norm": 0.09216390550136566, "learning_rate": 7.037358154833679e-05, "loss": 0.004, "step": 39810 }, { "grad_norm": 0.10281329602003098, "learning_rate": 7.035848065282339e-05, "loss": 0.0069, "step": 39820 }, { "grad_norm": 0.1281358152627945, "learning_rate": 7.034337753093714e-05, "loss": 0.0043, "step": 39830 }, { "grad_norm": 0.11022710055112839, "learning_rate": 7.032827218432972e-05, "loss": 0.0038, "step": 39840 }, { "grad_norm": 0.09671728312969208, "learning_rate": 7.031316461465302e-05, "loss": 0.0046, "step": 39850 }, { "grad_norm": 0.0891757681965828, "learning_rate": 7.029805482355915e-05, "loss": 0.0039, "step": 39860 }, { "grad_norm": 0.09032357484102249, "learning_rate": 7.028294281270055e-05, "loss": 0.0053, "step": 39870 }, { "grad_norm": 0.0880143791437149, "learning_rate": 7.026782858372982e-05, "loss": 0.0048, "step": 39880 }, { "grad_norm": 0.08561398833990097, "learning_rate": 7.025271213829982e-05, "loss": 0.0048, "step": 39890 }, { "grad_norm": 0.09884612262248993, "learning_rate": 7.023759347806366e-05, "loss": 0.0048, "step": 39900 }, { "grad_norm": 0.13423894345760345, "learning_rate": 7.022247260467472e-05, "loss": 0.005, "step": 39910 }, { "grad_norm": 0.09025257080793381, "learning_rate": 7.020734951978658e-05, "loss": 0.004, "step": 39920 }, { "grad_norm": 0.08680690079927444, "learning_rate": 7.019222422505307e-05, "loss": 0.0037, "step": 39930 }, { "grad_norm": 0.08245466649532318, "learning_rate": 7.017709672212827e-05, "loss": 0.0041, "step": 39940 }, { "grad_norm": 0.11214621365070343, "learning_rate": 7.016196701266652e-05, "loss": 0.0036, "step": 39950 }, { "grad_norm": 0.11671045422554016, "learning_rate": 7.014683509832238e-05, "loss": 0.0043, "step": 39960 }, { "grad_norm": 0.10480491071939468, "learning_rate": 7.013170098075063e-05, "loss": 0.0042, "step": 39970 }, { "grad_norm": 0.09290150552988052, "learning_rate": 7.011656466160632e-05, "loss": 0.0049, "step": 39980 }, { "grad_norm": 0.12327057868242264, "learning_rate": 7.010142614254475e-05, "loss": 0.0045, "step": 39990 }, { "grad_norm": 0.16614685952663422, "learning_rate": 7.008628542522147e-05, "loss": 0.0057, "step": 40000 }, { "grad_norm": 0.11211668699979782, "learning_rate": 7.007114251129218e-05, "loss": 0.0045, "step": 40010 }, { "grad_norm": 0.09962151944637299, "learning_rate": 7.005599740241293e-05, "loss": 0.004, "step": 40020 }, { "grad_norm": 0.10640081018209457, "learning_rate": 7.004085010023996e-05, "loss": 0.0048, "step": 40030 }, { "grad_norm": 0.09468626230955124, "learning_rate": 7.002570060642976e-05, "loss": 0.0037, "step": 40040 }, { "grad_norm": 0.07236252725124359, "learning_rate": 7.001054892263903e-05, "loss": 0.0047, "step": 40050 }, { "grad_norm": 0.07237526774406433, "learning_rate": 6.999539505052477e-05, "loss": 0.0037, "step": 40060 }, { "grad_norm": 0.07962881028652191, "learning_rate": 6.998023899174418e-05, "loss": 0.0039, "step": 40070 }, { "grad_norm": 0.08250793814659119, "learning_rate": 6.996508074795467e-05, "loss": 0.0053, "step": 40080 }, { "grad_norm": 0.07684281468391418, "learning_rate": 6.994992032081396e-05, "loss": 0.0037, "step": 40090 }, { "grad_norm": 0.10922935605049133, "learning_rate": 6.993475771197995e-05, "loss": 0.0057, "step": 40100 }, { "grad_norm": 0.09890209883451462, "learning_rate": 6.991959292311082e-05, "loss": 0.0067, "step": 40110 }, { "grad_norm": 0.11406426876783371, "learning_rate": 6.990442595586495e-05, "loss": 0.0065, "step": 40120 }, { "grad_norm": 0.09312892705202103, "learning_rate": 6.988925681190098e-05, "loss": 0.0057, "step": 40130 }, { "grad_norm": 0.12037984281778336, "learning_rate": 6.987408549287778e-05, "loss": 0.0063, "step": 40140 }, { "grad_norm": 0.09068118035793304, "learning_rate": 6.985891200045449e-05, "loss": 0.0049, "step": 40150 }, { "grad_norm": 0.17739112675189972, "learning_rate": 6.984373633629045e-05, "loss": 0.0054, "step": 40160 }, { "grad_norm": 0.09897762537002563, "learning_rate": 6.98285585020452e-05, "loss": 0.005, "step": 40170 }, { "grad_norm": 0.06931602209806442, "learning_rate": 6.981337849937864e-05, "loss": 0.0036, "step": 40180 }, { "grad_norm": 0.08972156792879105, "learning_rate": 6.979819632995078e-05, "loss": 0.0054, "step": 40190 }, { "grad_norm": 0.11337658762931824, "learning_rate": 6.978301199542193e-05, "loss": 0.0033, "step": 40200 }, { "grad_norm": 0.1304224729537964, "learning_rate": 6.976782549745263e-05, "loss": 0.0053, "step": 40210 }, { "grad_norm": 0.10416479408740997, "learning_rate": 6.975263683770367e-05, "loss": 0.0049, "step": 40220 }, { "grad_norm": 0.11042589694261551, "learning_rate": 6.973744601783604e-05, "loss": 0.0051, "step": 40230 }, { "grad_norm": 0.08549481630325317, "learning_rate": 6.972225303951097e-05, "loss": 0.0068, "step": 40240 }, { "grad_norm": 0.08529867231845856, "learning_rate": 6.970705790438998e-05, "loss": 0.0057, "step": 40250 }, { "grad_norm": 0.10236170887947083, "learning_rate": 6.969186061413477e-05, "loss": 0.0054, "step": 40260 }, { "grad_norm": 0.09875041246414185, "learning_rate": 6.967666117040727e-05, "loss": 0.0081, "step": 40270 }, { "grad_norm": 0.10734518617391586, "learning_rate": 6.966145957486972e-05, "loss": 0.0048, "step": 40280 }, { "grad_norm": 0.08467457443475723, "learning_rate": 6.964625582918449e-05, "loss": 0.0063, "step": 40290 }, { "grad_norm": 0.1158471405506134, "learning_rate": 6.963104993501425e-05, "loss": 0.0073, "step": 40300 }, { "grad_norm": 0.1513037383556366, "learning_rate": 6.961584189402192e-05, "loss": 0.007, "step": 40310 }, { "grad_norm": 0.16120830178260803, "learning_rate": 6.96006317078706e-05, "loss": 0.0061, "step": 40320 }, { "grad_norm": 0.08428547531366348, "learning_rate": 6.958541937822367e-05, "loss": 0.0056, "step": 40330 }, { "grad_norm": 0.09079120308160782, "learning_rate": 6.957020490674473e-05, "loss": 0.0045, "step": 40340 }, { "grad_norm": 0.1285264790058136, "learning_rate": 6.95549882950976e-05, "loss": 0.0059, "step": 40350 }, { "grad_norm": 0.09222692996263504, "learning_rate": 6.953976954494635e-05, "loss": 0.0053, "step": 40360 }, { "grad_norm": 0.10773787647485733, "learning_rate": 6.952454865795528e-05, "loss": 0.0052, "step": 40370 }, { "grad_norm": 0.08551697432994843, "learning_rate": 6.950932563578892e-05, "loss": 0.0047, "step": 40380 }, { "grad_norm": 0.06456514447927475, "learning_rate": 6.949410048011206e-05, "loss": 0.0047, "step": 40390 }, { "grad_norm": 0.13703656196594238, "learning_rate": 6.947887319258966e-05, "loss": 0.0054, "step": 40400 }, { "grad_norm": 0.07738706469535828, "learning_rate": 6.946364377488696e-05, "loss": 0.0052, "step": 40410 }, { "grad_norm": 0.13018976151943207, "learning_rate": 6.944841222866947e-05, "loss": 0.0053, "step": 40420 }, { "grad_norm": 0.1287519633769989, "learning_rate": 6.943317855560284e-05, "loss": 0.0061, "step": 40430 }, { "grad_norm": 0.0835876613855362, "learning_rate": 6.941794275735302e-05, "loss": 0.0055, "step": 40440 }, { "grad_norm": 0.10612813383340836, "learning_rate": 6.94027048355862e-05, "loss": 0.0057, "step": 40450 }, { "grad_norm": 0.11327695846557617, "learning_rate": 6.938746479196871e-05, "loss": 0.005, "step": 40460 }, { "grad_norm": 0.14114144444465637, "learning_rate": 6.937222262816724e-05, "loss": 0.0071, "step": 40470 }, { "grad_norm": 0.11408951133489609, "learning_rate": 6.935697834584865e-05, "loss": 0.0066, "step": 40480 }, { "grad_norm": 0.11811915040016174, "learning_rate": 6.934173194667998e-05, "loss": 0.0067, "step": 40490 }, { "grad_norm": 0.14740292727947235, "learning_rate": 6.93264834323286e-05, "loss": 0.0061, "step": 40500 }, { "grad_norm": 0.07697780430316925, "learning_rate": 6.931123280446206e-05, "loss": 0.0063, "step": 40510 }, { "grad_norm": 0.11604183912277222, "learning_rate": 6.929598006474811e-05, "loss": 0.0049, "step": 40520 }, { "grad_norm": 0.17266234755516052, "learning_rate": 6.92807252148548e-05, "loss": 0.0046, "step": 40530 }, { "grad_norm": 0.12150051444768906, "learning_rate": 6.926546825645037e-05, "loss": 0.0051, "step": 40540 }, { "grad_norm": 0.18976551294326782, "learning_rate": 6.92502091912033e-05, "loss": 0.004, "step": 40550 }, { "grad_norm": 0.1025877594947815, "learning_rate": 6.923494802078229e-05, "loss": 0.0047, "step": 40560 }, { "grad_norm": 0.10545092821121216, "learning_rate": 6.921968474685628e-05, "loss": 0.005, "step": 40570 }, { "grad_norm": 0.1227809265255928, "learning_rate": 6.920441937109445e-05, "loss": 0.0058, "step": 40580 }, { "grad_norm": 0.09530524164438248, "learning_rate": 6.91891518951662e-05, "loss": 0.004, "step": 40590 }, { "grad_norm": 0.15078304708003998, "learning_rate": 6.917388232074114e-05, "loss": 0.0048, "step": 40600 }, { "grad_norm": 0.1108582466840744, "learning_rate": 6.915861064948914e-05, "loss": 0.0052, "step": 40610 }, { "grad_norm": 0.09547597169876099, "learning_rate": 6.914333688308025e-05, "loss": 0.005, "step": 40620 }, { "grad_norm": 0.08823152631521225, "learning_rate": 6.912806102318487e-05, "loss": 0.004, "step": 40630 }, { "grad_norm": 0.06967179477214813, "learning_rate": 6.911278307147347e-05, "loss": 0.0051, "step": 40640 }, { "grad_norm": 0.10125426948070526, "learning_rate": 6.909750302961684e-05, "loss": 0.0036, "step": 40650 }, { "grad_norm": 0.10094611346721649, "learning_rate": 6.9082220899286e-05, "loss": 0.005, "step": 40660 }, { "grad_norm": 0.09151238203048706, "learning_rate": 6.906693668215219e-05, "loss": 0.0041, "step": 40670 }, { "grad_norm": 0.08960478752851486, "learning_rate": 6.905165037988683e-05, "loss": 0.0049, "step": 40680 }, { "grad_norm": 0.11954150348901749, "learning_rate": 6.903636199416164e-05, "loss": 0.0046, "step": 40690 }, { "grad_norm": 0.12083358317613602, "learning_rate": 6.902107152664851e-05, "loss": 0.0066, "step": 40700 }, { "grad_norm": 0.13246679306030273, "learning_rate": 6.90057789790196e-05, "loss": 0.0059, "step": 40710 }, { "grad_norm": 0.11836410313844681, "learning_rate": 6.899048435294728e-05, "loss": 0.0049, "step": 40720 }, { "grad_norm": 0.07395736873149872, "learning_rate": 6.897518765010415e-05, "loss": 0.0058, "step": 40730 }, { "grad_norm": 0.0861722081899643, "learning_rate": 6.895988887216302e-05, "loss": 0.0051, "step": 40740 }, { "grad_norm": 0.07144927233457565, "learning_rate": 6.894458802079694e-05, "loss": 0.0035, "step": 40750 }, { "grad_norm": 0.09670236706733704, "learning_rate": 6.892928509767922e-05, "loss": 0.005, "step": 40760 }, { "grad_norm": 0.10702469199895859, "learning_rate": 6.891398010448333e-05, "loss": 0.0048, "step": 40770 }, { "grad_norm": 0.14095258712768555, "learning_rate": 6.8898673042883e-05, "loss": 0.0039, "step": 40780 }, { "grad_norm": 0.07781775295734406, "learning_rate": 6.888336391455222e-05, "loss": 0.004, "step": 40790 }, { "grad_norm": 0.051904983818531036, "learning_rate": 6.886805272116513e-05, "loss": 0.0043, "step": 40800 }, { "grad_norm": 0.06902889907360077, "learning_rate": 6.885273946439617e-05, "loss": 0.0044, "step": 40810 }, { "grad_norm": 0.12963759899139404, "learning_rate": 6.883742414591998e-05, "loss": 0.0062, "step": 40820 }, { "grad_norm": 0.10396280139684677, "learning_rate": 6.88221067674114e-05, "loss": 0.0061, "step": 40830 }, { "grad_norm": 0.1180562824010849, "learning_rate": 6.88067873305455e-05, "loss": 0.0063, "step": 40840 }, { "grad_norm": 0.1073959469795227, "learning_rate": 6.879146583699765e-05, "loss": 0.0075, "step": 40850 }, { "grad_norm": 0.11158730834722519, "learning_rate": 6.877614228844334e-05, "loss": 0.0103, "step": 40860 }, { "grad_norm": 0.15066289901733398, "learning_rate": 6.876081668655832e-05, "loss": 0.008, "step": 40870 }, { "grad_norm": 0.08151714503765106, "learning_rate": 6.874548903301861e-05, "loss": 0.0072, "step": 40880 }, { "grad_norm": 0.1117854043841362, "learning_rate": 6.873015932950039e-05, "loss": 0.0061, "step": 40890 }, { "grad_norm": 0.10864989459514618, "learning_rate": 6.871482757768012e-05, "loss": 0.0045, "step": 40900 }, { "grad_norm": 0.07860003411769867, "learning_rate": 6.869949377923442e-05, "loss": 0.0045, "step": 40910 }, { "grad_norm": 0.09907631576061249, "learning_rate": 6.868415793584022e-05, "loss": 0.0041, "step": 40920 }, { "grad_norm": 0.13883987069129944, "learning_rate": 6.86688200491746e-05, "loss": 0.0048, "step": 40930 }, { "grad_norm": 0.11099723726511002, "learning_rate": 6.865348012091488e-05, "loss": 0.0063, "step": 40940 }, { "grad_norm": 0.11522219330072403, "learning_rate": 6.86381381527386e-05, "loss": 0.0073, "step": 40950 }, { "grad_norm": 0.19884295761585236, "learning_rate": 6.862279414632359e-05, "loss": 0.005, "step": 40960 }, { "grad_norm": 0.08301534503698349, "learning_rate": 6.860744810334778e-05, "loss": 0.0057, "step": 40970 }, { "grad_norm": 0.09389153122901917, "learning_rate": 6.859210002548943e-05, "loss": 0.0058, "step": 40980 }, { "grad_norm": 0.08159977197647095, "learning_rate": 6.8576749914427e-05, "loss": 0.0038, "step": 40990 }, { "grad_norm": 0.09942826628684998, "learning_rate": 6.85613977718391e-05, "loss": 0.0052, "step": 41000 }, { "grad_norm": 0.09625107049942017, "learning_rate": 6.854604359940464e-05, "loss": 0.0047, "step": 41010 }, { "grad_norm": 0.10080952197313309, "learning_rate": 6.853068739880276e-05, "loss": 0.0044, "step": 41020 }, { "grad_norm": 0.0850275531411171, "learning_rate": 6.851532917171276e-05, "loss": 0.0046, "step": 41030 }, { "grad_norm": 0.15125542879104614, "learning_rate": 6.849996891981418e-05, "loss": 0.0059, "step": 41040 }, { "grad_norm": 0.0947791263461113, "learning_rate": 6.848460664478684e-05, "loss": 0.0057, "step": 41050 }, { "grad_norm": 0.0787186399102211, "learning_rate": 6.846924234831069e-05, "loss": 0.0044, "step": 41060 }, { "grad_norm": 0.08239023387432098, "learning_rate": 6.845387603206597e-05, "loss": 0.0052, "step": 41070 }, { "grad_norm": 0.10451879352331161, "learning_rate": 6.843850769773312e-05, "loss": 0.0036, "step": 41080 }, { "grad_norm": 0.07012593746185303, "learning_rate": 6.842313734699278e-05, "loss": 0.004, "step": 41090 }, { "grad_norm": 0.058413516730070114, "learning_rate": 6.840776498152584e-05, "loss": 0.0037, "step": 41100 }, { "grad_norm": 0.06325267255306244, "learning_rate": 6.83923906030134e-05, "loss": 0.0051, "step": 41110 }, { "grad_norm": 0.0994156077504158, "learning_rate": 6.837701421313677e-05, "loss": 0.0048, "step": 41120 }, { "grad_norm": 0.09521555155515671, "learning_rate": 6.83616358135775e-05, "loss": 0.0035, "step": 41130 }, { "grad_norm": 0.08141029626131058, "learning_rate": 6.834625540601734e-05, "loss": 0.0036, "step": 41140 }, { "grad_norm": 0.07618988305330276, "learning_rate": 6.833087299213829e-05, "loss": 0.0038, "step": 41150 }, { "grad_norm": 0.07209070026874542, "learning_rate": 6.83154885736225e-05, "loss": 0.0036, "step": 41160 }, { "grad_norm": 0.08205151557922363, "learning_rate": 6.830010215215244e-05, "loss": 0.0035, "step": 41170 }, { "grad_norm": 0.09697937220335007, "learning_rate": 6.828471372941072e-05, "loss": 0.0047, "step": 41180 }, { "grad_norm": 0.10553675144910812, "learning_rate": 6.826932330708018e-05, "loss": 0.0052, "step": 41190 }, { "grad_norm": 0.16174174845218658, "learning_rate": 6.825393088684393e-05, "loss": 0.0039, "step": 41200 }, { "grad_norm": 0.08969586342573166, "learning_rate": 6.823853647038524e-05, "loss": 0.0034, "step": 41210 }, { "grad_norm": 0.13820336759090424, "learning_rate": 6.822314005938762e-05, "loss": 0.0048, "step": 41220 }, { "grad_norm": 0.07783254981040955, "learning_rate": 6.820774165553479e-05, "loss": 0.0041, "step": 41230 }, { "grad_norm": 0.06751640141010284, "learning_rate": 6.819234126051072e-05, "loss": 0.004, "step": 41240 }, { "grad_norm": 0.06757872551679611, "learning_rate": 6.817693887599956e-05, "loss": 0.0037, "step": 41250 }, { "grad_norm": 0.08683044463396072, "learning_rate": 6.816153450368568e-05, "loss": 0.005, "step": 41260 }, { "grad_norm": 0.07854272425174713, "learning_rate": 6.814612814525372e-05, "loss": 0.0049, "step": 41270 }, { "grad_norm": 0.10290129482746124, "learning_rate": 6.813071980238846e-05, "loss": 0.0047, "step": 41280 }, { "grad_norm": 0.1476607471704483, "learning_rate": 6.811530947677492e-05, "loss": 0.006, "step": 41290 }, { "grad_norm": 0.1193963885307312, "learning_rate": 6.809989717009839e-05, "loss": 0.0069, "step": 41300 }, { "grad_norm": 0.12484145164489746, "learning_rate": 6.808448288404431e-05, "loss": 0.0065, "step": 41310 }, { "grad_norm": 0.12001243978738785, "learning_rate": 6.806906662029838e-05, "loss": 0.0049, "step": 41320 }, { "grad_norm": 0.10382763296365738, "learning_rate": 6.805364838054647e-05, "loss": 0.0051, "step": 41330 }, { "grad_norm": 0.11026397347450256, "learning_rate": 6.803822816647474e-05, "loss": 0.0045, "step": 41340 }, { "grad_norm": 0.0821840912103653, "learning_rate": 6.802280597976949e-05, "loss": 0.0043, "step": 41350 }, { "grad_norm": 0.09762857854366302, "learning_rate": 6.800738182211727e-05, "loss": 0.0042, "step": 41360 }, { "grad_norm": 0.09486772119998932, "learning_rate": 6.799195569520485e-05, "loss": 0.005, "step": 41370 }, { "grad_norm": 0.08296597748994827, "learning_rate": 6.79765276007192e-05, "loss": 0.0049, "step": 41380 }, { "grad_norm": 0.09834809601306915, "learning_rate": 6.796109754034753e-05, "loss": 0.0048, "step": 41390 }, { "grad_norm": 0.14152371883392334, "learning_rate": 6.794566551577724e-05, "loss": 0.0061, "step": 41400 }, { "grad_norm": 0.13364194333553314, "learning_rate": 6.793023152869594e-05, "loss": 0.0045, "step": 41410 }, { "grad_norm": 0.07908660173416138, "learning_rate": 6.791479558079147e-05, "loss": 0.0034, "step": 41420 }, { "grad_norm": 0.09479524940252304, "learning_rate": 6.789935767375191e-05, "loss": 0.0043, "step": 41430 }, { "grad_norm": 0.08589717745780945, "learning_rate": 6.78839178092655e-05, "loss": 0.0044, "step": 41440 }, { "grad_norm": 0.1401423066854477, "learning_rate": 6.786847598902072e-05, "loss": 0.0058, "step": 41450 }, { "grad_norm": 0.11809468269348145, "learning_rate": 6.785303221470627e-05, "loss": 0.0047, "step": 41460 }, { "grad_norm": 0.13227961957454681, "learning_rate": 6.783758648801108e-05, "loss": 0.0056, "step": 41470 }, { "grad_norm": 0.05864667519927025, "learning_rate": 6.782213881062422e-05, "loss": 0.0053, "step": 41480 }, { "grad_norm": 0.10879392921924591, "learning_rate": 6.780668918423508e-05, "loss": 0.005, "step": 41490 }, { "grad_norm": 0.10198533535003662, "learning_rate": 6.779123761053317e-05, "loss": 0.0044, "step": 41500 }, { "grad_norm": 0.10401713848114014, "learning_rate": 6.777578409120826e-05, "loss": 0.0041, "step": 41510 }, { "grad_norm": 0.10042103379964828, "learning_rate": 6.776032862795033e-05, "loss": 0.0045, "step": 41520 }, { "grad_norm": 0.13434137403964996, "learning_rate": 6.774487122244956e-05, "loss": 0.0042, "step": 41530 }, { "grad_norm": 0.10056130588054657, "learning_rate": 6.772941187639637e-05, "loss": 0.0044, "step": 41540 }, { "grad_norm": 0.14778797328472137, "learning_rate": 6.771395059148134e-05, "loss": 0.0054, "step": 41550 }, { "grad_norm": 0.07113508135080338, "learning_rate": 6.76984873693953e-05, "loss": 0.0056, "step": 41560 }, { "grad_norm": 0.07392779737710953, "learning_rate": 6.768302221182931e-05, "loss": 0.005, "step": 41570 }, { "grad_norm": 0.10918468981981277, "learning_rate": 6.766755512047457e-05, "loss": 0.0048, "step": 41580 }, { "grad_norm": 0.12839660048484802, "learning_rate": 6.765208609702259e-05, "loss": 0.0044, "step": 41590 }, { "grad_norm": 0.06567282974720001, "learning_rate": 6.763661514316499e-05, "loss": 0.004, "step": 41600 }, { "grad_norm": 0.08856330066919327, "learning_rate": 6.76211422605937e-05, "loss": 0.0037, "step": 41610 }, { "grad_norm": 0.06874994933605194, "learning_rate": 6.760566745100076e-05, "loss": 0.0038, "step": 41620 }, { "grad_norm": 0.147721529006958, "learning_rate": 6.75901907160785e-05, "loss": 0.0043, "step": 41630 }, { "grad_norm": 0.10948876291513443, "learning_rate": 6.757471205751943e-05, "loss": 0.0056, "step": 41640 }, { "grad_norm": 0.10768619179725647, "learning_rate": 6.75592314770163e-05, "loss": 0.0046, "step": 41650 }, { "grad_norm": 0.06753487139940262, "learning_rate": 6.7543748976262e-05, "loss": 0.005, "step": 41660 }, { "grad_norm": 0.09910068660974503, "learning_rate": 6.752826455694968e-05, "loss": 0.0041, "step": 41670 }, { "grad_norm": 0.09861475229263306, "learning_rate": 6.751277822077271e-05, "loss": 0.0046, "step": 41680 }, { "grad_norm": 0.07712475955486298, "learning_rate": 6.749728996942465e-05, "loss": 0.0054, "step": 41690 }, { "grad_norm": 0.10437864810228348, "learning_rate": 6.748179980459924e-05, "loss": 0.0045, "step": 41700 }, { "grad_norm": 0.07273504137992859, "learning_rate": 6.746630772799052e-05, "loss": 0.0036, "step": 41710 }, { "grad_norm": 0.07669747620820999, "learning_rate": 6.745081374129262e-05, "loss": 0.0039, "step": 41720 }, { "grad_norm": 0.0839758887887001, "learning_rate": 6.743531784619998e-05, "loss": 0.0044, "step": 41730 }, { "grad_norm": 0.07900340110063553, "learning_rate": 6.741982004440719e-05, "loss": 0.0047, "step": 41740 }, { "grad_norm": 0.05182048678398132, "learning_rate": 6.740432033760907e-05, "loss": 0.0039, "step": 41750 }, { "grad_norm": 0.06971769779920578, "learning_rate": 6.738881872750066e-05, "loss": 0.004, "step": 41760 }, { "grad_norm": 0.08150149881839752, "learning_rate": 6.737331521577715e-05, "loss": 0.0049, "step": 41770 }, { "grad_norm": 0.0999046340584755, "learning_rate": 6.735780980413403e-05, "loss": 0.0045, "step": 41780 }, { "grad_norm": 0.10179443657398224, "learning_rate": 6.734230249426692e-05, "loss": 0.0052, "step": 41790 }, { "grad_norm": 0.10491029173135757, "learning_rate": 6.732679328787168e-05, "loss": 0.0039, "step": 41800 }, { "grad_norm": 0.0985504612326622, "learning_rate": 6.731128218664438e-05, "loss": 0.0034, "step": 41810 }, { "grad_norm": 0.07941185683012009, "learning_rate": 6.729576919228129e-05, "loss": 0.0034, "step": 41820 }, { "grad_norm": 0.12257120758295059, "learning_rate": 6.728025430647888e-05, "loss": 0.0041, "step": 41830 }, { "grad_norm": 0.09327907860279083, "learning_rate": 6.726473753093383e-05, "loss": 0.0035, "step": 41840 }, { "grad_norm": 0.34293079376220703, "learning_rate": 6.724921886734305e-05, "loss": 0.0057, "step": 41850 }, { "grad_norm": 0.10889997333288193, "learning_rate": 6.723369831740363e-05, "loss": 0.0043, "step": 41860 }, { "grad_norm": 0.10005957633256912, "learning_rate": 6.721817588281288e-05, "loss": 0.0062, "step": 41870 }, { "grad_norm": 0.11588864773511887, "learning_rate": 6.720265156526828e-05, "loss": 0.0049, "step": 41880 }, { "grad_norm": 0.19025322794914246, "learning_rate": 6.718712536646758e-05, "loss": 0.0052, "step": 41890 }, { "grad_norm": 0.11169176548719406, "learning_rate": 6.71715972881087e-05, "loss": 0.0052, "step": 41900 }, { "grad_norm": 0.11954380571842194, "learning_rate": 6.715606733188974e-05, "loss": 0.0045, "step": 41910 }, { "grad_norm": 0.08827391266822815, "learning_rate": 6.714053549950909e-05, "loss": 0.0059, "step": 41920 }, { "grad_norm": 0.12187203019857407, "learning_rate": 6.71250017926652e-05, "loss": 0.0058, "step": 41930 }, { "grad_norm": 0.11421797424554825, "learning_rate": 6.71094662130569e-05, "loss": 0.0049, "step": 41940 }, { "grad_norm": 0.08545517921447754, "learning_rate": 6.709392876238307e-05, "loss": 0.0046, "step": 41950 }, { "grad_norm": 0.09169618040323257, "learning_rate": 6.707838944234289e-05, "loss": 0.0042, "step": 41960 }, { "grad_norm": 0.06777074187994003, "learning_rate": 6.706284825463574e-05, "loss": 0.0045, "step": 41970 }, { "grad_norm": 0.062331974506378174, "learning_rate": 6.704730520096114e-05, "loss": 0.0034, "step": 41980 }, { "grad_norm": 0.06507875770330429, "learning_rate": 6.703176028301888e-05, "loss": 0.003, "step": 41990 }, { "grad_norm": 0.05090352147817612, "learning_rate": 6.701621350250892e-05, "loss": 0.003, "step": 42000 }, { "grad_norm": 0.06247740238904953, "learning_rate": 6.700066486113144e-05, "loss": 0.0035, "step": 42010 }, { "grad_norm": 0.06234964355826378, "learning_rate": 6.69851143605868e-05, "loss": 0.0035, "step": 42020 }, { "grad_norm": 0.06838791817426682, "learning_rate": 6.696956200257561e-05, "loss": 0.0033, "step": 42030 }, { "grad_norm": 0.0772450789809227, "learning_rate": 6.695400778879863e-05, "loss": 0.0038, "step": 42040 }, { "grad_norm": 0.11675380915403366, "learning_rate": 6.693845172095683e-05, "loss": 0.0045, "step": 42050 }, { "grad_norm": 0.10018576681613922, "learning_rate": 6.692289380075142e-05, "loss": 0.0048, "step": 42060 }, { "grad_norm": 0.14698228240013123, "learning_rate": 6.690733402988379e-05, "loss": 0.0064, "step": 42070 }, { "grad_norm": 0.09893801063299179, "learning_rate": 6.689177241005553e-05, "loss": 0.0046, "step": 42080 }, { "grad_norm": 0.08449488878250122, "learning_rate": 6.687620894296844e-05, "loss": 0.0035, "step": 42090 }, { "grad_norm": 0.07470931857824326, "learning_rate": 6.686064363032451e-05, "loss": 0.0028, "step": 42100 }, { "grad_norm": 0.07708176970481873, "learning_rate": 6.684507647382596e-05, "loss": 0.0039, "step": 42110 }, { "grad_norm": 0.08161807805299759, "learning_rate": 6.682950747517513e-05, "loss": 0.003, "step": 42120 }, { "grad_norm": 0.0756722018122673, "learning_rate": 6.68139366360747e-05, "loss": 0.0034, "step": 42130 }, { "grad_norm": 0.09584114700555801, "learning_rate": 6.679836395822744e-05, "loss": 0.0036, "step": 42140 }, { "grad_norm": 0.10572609305381775, "learning_rate": 6.678278944333633e-05, "loss": 0.0044, "step": 42150 }, { "grad_norm": 0.08080144226551056, "learning_rate": 6.676721309310462e-05, "loss": 0.004, "step": 42160 }, { "grad_norm": 0.11154317110776901, "learning_rate": 6.67516349092357e-05, "loss": 0.0053, "step": 42170 }, { "grad_norm": 0.07770144939422607, "learning_rate": 6.673605489343315e-05, "loss": 0.0038, "step": 42180 }, { "grad_norm": 0.07854588329792023, "learning_rate": 6.672047304740081e-05, "loss": 0.0036, "step": 42190 }, { "grad_norm": 0.08487482368946075, "learning_rate": 6.670488937284268e-05, "loss": 0.0037, "step": 42200 }, { "grad_norm": 0.07379508763551712, "learning_rate": 6.668930387146294e-05, "loss": 0.0037, "step": 42210 }, { "grad_norm": 0.06445527821779251, "learning_rate": 6.667371654496605e-05, "loss": 0.0038, "step": 42220 }, { "grad_norm": 0.06138818711042404, "learning_rate": 6.665812739505659e-05, "loss": 0.0034, "step": 42230 }, { "grad_norm": 0.06233540549874306, "learning_rate": 6.664253642343935e-05, "loss": 0.0036, "step": 42240 }, { "grad_norm": 0.1764325350522995, "learning_rate": 6.662694363181935e-05, "loss": 0.0035, "step": 42250 }, { "grad_norm": 0.10497259348630905, "learning_rate": 6.66113490219018e-05, "loss": 0.0049, "step": 42260 }, { "grad_norm": 0.12830716371536255, "learning_rate": 6.659575259539207e-05, "loss": 0.005, "step": 42270 }, { "grad_norm": 0.14203576743602753, "learning_rate": 6.65801543539958e-05, "loss": 0.0042, "step": 42280 }, { "grad_norm": 0.18705737590789795, "learning_rate": 6.65645542994188e-05, "loss": 0.0097, "step": 42290 }, { "grad_norm": 0.13900358974933624, "learning_rate": 6.654895243336702e-05, "loss": 0.0093, "step": 42300 }, { "grad_norm": 0.14951172471046448, "learning_rate": 6.65333487575467e-05, "loss": 0.0092, "step": 42310 }, { "grad_norm": 0.1455516815185547, "learning_rate": 6.651774327366422e-05, "loss": 0.0073, "step": 42320 }, { "grad_norm": 0.138942152261734, "learning_rate": 6.650213598342613e-05, "loss": 0.0093, "step": 42330 }, { "grad_norm": 0.14870727062225342, "learning_rate": 6.648652688853931e-05, "loss": 0.0058, "step": 42340 }, { "grad_norm": 0.10400309413671494, "learning_rate": 6.647091599071066e-05, "loss": 0.0062, "step": 42350 }, { "grad_norm": 0.09673573821783066, "learning_rate": 6.645530329164743e-05, "loss": 0.0049, "step": 42360 }, { "grad_norm": 0.09549874067306519, "learning_rate": 6.643968879305694e-05, "loss": 0.007, "step": 42370 }, { "grad_norm": 0.11944150924682617, "learning_rate": 6.642407249664683e-05, "loss": 0.0055, "step": 42380 }, { "grad_norm": 0.12202620506286621, "learning_rate": 6.640845440412483e-05, "loss": 0.0046, "step": 42390 }, { "grad_norm": 0.0654386356472969, "learning_rate": 6.639283451719893e-05, "loss": 0.0044, "step": 42400 }, { "grad_norm": 0.09265010803937912, "learning_rate": 6.637721283757729e-05, "loss": 0.0042, "step": 42410 }, { "grad_norm": 0.10739142447710037, "learning_rate": 6.636158936696828e-05, "loss": 0.0048, "step": 42420 }, { "grad_norm": 0.0664442777633667, "learning_rate": 6.634596410708047e-05, "loss": 0.005, "step": 42430 }, { "grad_norm": 0.07833166420459747, "learning_rate": 6.633033705962257e-05, "loss": 0.0047, "step": 42440 }, { "grad_norm": 0.07059276849031448, "learning_rate": 6.631470822630359e-05, "loss": 0.0047, "step": 42450 }, { "grad_norm": 0.08805079013109207, "learning_rate": 6.629907760883263e-05, "loss": 0.0037, "step": 42460 }, { "grad_norm": 0.05997317656874657, "learning_rate": 6.628344520891905e-05, "loss": 0.0042, "step": 42470 }, { "grad_norm": 0.07482054829597473, "learning_rate": 6.62678110282724e-05, "loss": 0.0035, "step": 42480 }, { "grad_norm": 0.08194457739591599, "learning_rate": 6.625217506860237e-05, "loss": 0.0042, "step": 42490 }, { "grad_norm": 0.0711677297949791, "learning_rate": 6.623653733161892e-05, "loss": 0.0041, "step": 42500 }, { "grad_norm": 0.07136737555265427, "learning_rate": 6.622089781903217e-05, "loss": 0.0032, "step": 42510 }, { "grad_norm": 0.15133874118328094, "learning_rate": 6.620525653255241e-05, "loss": 0.0053, "step": 42520 }, { "grad_norm": 0.11661218851804733, "learning_rate": 6.618961347389015e-05, "loss": 0.0045, "step": 42530 }, { "grad_norm": 0.10028599202632904, "learning_rate": 6.617396864475613e-05, "loss": 0.005, "step": 42540 }, { "grad_norm": 0.098261259496212, "learning_rate": 6.61583220468612e-05, "loss": 0.0047, "step": 42550 }, { "grad_norm": 0.1061980277299881, "learning_rate": 6.614267368191645e-05, "loss": 0.0049, "step": 42560 }, { "grad_norm": 0.08098049461841583, "learning_rate": 6.61270235516332e-05, "loss": 0.004, "step": 42570 }, { "grad_norm": 0.06812366098165512, "learning_rate": 6.611137165772287e-05, "loss": 0.0038, "step": 42580 }, { "grad_norm": 0.0960550531744957, "learning_rate": 6.60957180018972e-05, "loss": 0.0047, "step": 42590 }, { "grad_norm": 0.1045348197221756, "learning_rate": 6.608006258586797e-05, "loss": 0.0039, "step": 42600 }, { "grad_norm": 0.1252179592847824, "learning_rate": 6.606440541134731e-05, "loss": 0.0048, "step": 42610 }, { "grad_norm": 0.2610074579715729, "learning_rate": 6.60487464800474e-05, "loss": 0.0049, "step": 42620 }, { "grad_norm": 0.13153503835201263, "learning_rate": 6.603308579368071e-05, "loss": 0.006, "step": 42630 }, { "grad_norm": 0.0803581178188324, "learning_rate": 6.601742335395987e-05, "loss": 0.0056, "step": 42640 }, { "grad_norm": 0.0774894505739212, "learning_rate": 6.600175916259769e-05, "loss": 0.0039, "step": 42650 }, { "grad_norm": 0.1044439822435379, "learning_rate": 6.598609322130718e-05, "loss": 0.0061, "step": 42660 }, { "grad_norm": 0.09196683764457703, "learning_rate": 6.597042553180155e-05, "loss": 0.0048, "step": 42670 }, { "grad_norm": 0.06923588365316391, "learning_rate": 6.59547560957942e-05, "loss": 0.0041, "step": 42680 }, { "grad_norm": 0.11575276404619217, "learning_rate": 6.59390849149987e-05, "loss": 0.0052, "step": 42690 }, { "grad_norm": 0.07248378545045853, "learning_rate": 6.592341199112886e-05, "loss": 0.0045, "step": 42700 }, { "grad_norm": 0.09013968706130981, "learning_rate": 6.590773732589861e-05, "loss": 0.0037, "step": 42710 }, { "grad_norm": 0.06832326948642731, "learning_rate": 6.589206092102213e-05, "loss": 0.0041, "step": 42720 }, { "grad_norm": 0.13904744386672974, "learning_rate": 6.587638277821375e-05, "loss": 0.0043, "step": 42730 }, { "grad_norm": 0.13332316279411316, "learning_rate": 6.586070289918803e-05, "loss": 0.004, "step": 42740 }, { "grad_norm": 0.10046526044607162, "learning_rate": 6.584502128565968e-05, "loss": 0.0049, "step": 42750 }, { "grad_norm": 0.07531996816396713, "learning_rate": 6.582933793934363e-05, "loss": 0.0033, "step": 42760 }, { "grad_norm": 0.08637070655822754, "learning_rate": 6.581365286195498e-05, "loss": 0.0039, "step": 42770 }, { "grad_norm": 0.12291591614484787, "learning_rate": 6.579796605520903e-05, "loss": 0.0055, "step": 42780 }, { "grad_norm": 0.0974564403295517, "learning_rate": 6.578227752082127e-05, "loss": 0.0038, "step": 42790 }, { "grad_norm": 0.07831864058971405, "learning_rate": 6.576658726050735e-05, "loss": 0.004, "step": 42800 }, { "grad_norm": 0.09137833118438721, "learning_rate": 6.575089527598316e-05, "loss": 0.0046, "step": 42810 }, { "grad_norm": 0.08810349553823471, "learning_rate": 6.573520156896476e-05, "loss": 0.0047, "step": 42820 }, { "grad_norm": 0.0886538177728653, "learning_rate": 6.571950614116835e-05, "loss": 0.0049, "step": 42830 }, { "grad_norm": 0.0700581967830658, "learning_rate": 6.570380899431039e-05, "loss": 0.004, "step": 42840 }, { "grad_norm": 0.12614110112190247, "learning_rate": 6.568811013010749e-05, "loss": 0.0059, "step": 42850 }, { "grad_norm": 0.10624603927135468, "learning_rate": 6.567240955027645e-05, "loss": 0.0042, "step": 42860 }, { "grad_norm": 0.0914531797170639, "learning_rate": 6.565670725653427e-05, "loss": 0.0037, "step": 42870 }, { "grad_norm": 0.054491862654685974, "learning_rate": 6.564100325059812e-05, "loss": 0.0033, "step": 42880 }, { "grad_norm": 0.062002960592508316, "learning_rate": 6.562529753418538e-05, "loss": 0.0039, "step": 42890 }, { "grad_norm": 0.0871419906616211, "learning_rate": 6.56095901090136e-05, "loss": 0.0038, "step": 42900 }, { "grad_norm": 0.08323561400175095, "learning_rate": 6.55938809768005e-05, "loss": 0.0031, "step": 42910 }, { "grad_norm": 0.07418379187583923, "learning_rate": 6.557817013926402e-05, "loss": 0.0056, "step": 42920 }, { "grad_norm": 0.09753990173339844, "learning_rate": 6.556245759812227e-05, "loss": 0.0044, "step": 42930 }, { "grad_norm": 0.0781136229634285, "learning_rate": 6.554674335509357e-05, "loss": 0.004, "step": 42940 }, { "grad_norm": 0.07777382433414459, "learning_rate": 6.553102741189638e-05, "loss": 0.0036, "step": 42950 }, { "grad_norm": 0.09858659654855728, "learning_rate": 6.55153097702494e-05, "loss": 0.0037, "step": 42960 }, { "grad_norm": 0.06722725927829742, "learning_rate": 6.549959043187146e-05, "loss": 0.004, "step": 42970 }, { "grad_norm": 0.08293172717094421, "learning_rate": 6.54838693984816e-05, "loss": 0.0039, "step": 42980 }, { "grad_norm": 0.08476860076189041, "learning_rate": 6.546814667179909e-05, "loss": 0.0037, "step": 42990 }, { "grad_norm": 0.06990116089582443, "learning_rate": 6.545242225354328e-05, "loss": 0.0045, "step": 43000 }, { "grad_norm": 0.15557435154914856, "learning_rate": 6.543669614543383e-05, "loss": 0.0071, "step": 43010 }, { "grad_norm": 0.08222787082195282, "learning_rate": 6.542096834919049e-05, "loss": 0.0068, "step": 43020 }, { "grad_norm": 0.09043978154659271, "learning_rate": 6.540523886653324e-05, "loss": 0.0048, "step": 43030 }, { "grad_norm": 0.05585411190986633, "learning_rate": 6.538950769918222e-05, "loss": 0.0052, "step": 43040 }, { "grad_norm": 0.06923103332519531, "learning_rate": 6.537377484885779e-05, "loss": 0.0043, "step": 43050 }, { "grad_norm": 0.07708602398633957, "learning_rate": 6.535804031728045e-05, "loss": 0.0035, "step": 43060 }, { "grad_norm": 0.09545394033193588, "learning_rate": 6.534230410617092e-05, "loss": 0.0039, "step": 43070 }, { "grad_norm": 0.10863785445690155, "learning_rate": 6.532656621725007e-05, "loss": 0.0037, "step": 43080 }, { "grad_norm": 0.06540754437446594, "learning_rate": 6.531082665223898e-05, "loss": 0.0044, "step": 43090 }, { "grad_norm": 0.1088062971830368, "learning_rate": 6.529508541285889e-05, "loss": 0.0035, "step": 43100 }, { "grad_norm": 0.06953567266464233, "learning_rate": 6.527934250083128e-05, "loss": 0.0028, "step": 43110 }, { "grad_norm": 0.0733075961470604, "learning_rate": 6.526359791787775e-05, "loss": 0.0031, "step": 43120 }, { "grad_norm": 0.0787891298532486, "learning_rate": 6.524785166572009e-05, "loss": 0.0033, "step": 43130 }, { "grad_norm": 0.08675359189510345, "learning_rate": 6.523210374608029e-05, "loss": 0.0038, "step": 43140 }, { "grad_norm": 0.07516886293888092, "learning_rate": 6.521635416068054e-05, "loss": 0.0039, "step": 43150 }, { "grad_norm": 0.07856201380491257, "learning_rate": 6.520060291124317e-05, "loss": 0.0035, "step": 43160 }, { "grad_norm": 0.07637880742549896, "learning_rate": 6.518484999949072e-05, "loss": 0.0056, "step": 43170 }, { "grad_norm": 0.08300889283418655, "learning_rate": 6.516909542714591e-05, "loss": 0.0038, "step": 43180 }, { "grad_norm": 0.06435579806566238, "learning_rate": 6.515333919593161e-05, "loss": 0.0035, "step": 43190 }, { "grad_norm": 0.10018093883991241, "learning_rate": 6.513758130757094e-05, "loss": 0.0045, "step": 43200 }, { "grad_norm": 0.08311552554368973, "learning_rate": 6.512182176378713e-05, "loss": 0.0037, "step": 43210 }, { "grad_norm": 0.11006749421358109, "learning_rate": 6.510606056630362e-05, "loss": 0.0049, "step": 43220 }, { "grad_norm": 0.10143797844648361, "learning_rate": 6.509029771684405e-05, "loss": 0.005, "step": 43230 }, { "grad_norm": 0.09226579964160919, "learning_rate": 6.50745332171322e-05, "loss": 0.0047, "step": 43240 }, { "grad_norm": 0.14367790520191193, "learning_rate": 6.505876706889207e-05, "loss": 0.005, "step": 43250 }, { "grad_norm": 0.08519449084997177, "learning_rate": 6.504299927384781e-05, "loss": 0.0039, "step": 43260 }, { "grad_norm": 0.08411501348018646, "learning_rate": 6.502722983372378e-05, "loss": 0.0046, "step": 43270 }, { "grad_norm": 0.09858304262161255, "learning_rate": 6.501145875024446e-05, "loss": 0.0057, "step": 43280 }, { "grad_norm": 0.0682872012257576, "learning_rate": 6.499568602513459e-05, "loss": 0.0039, "step": 43290 }, { "grad_norm": 0.0891435369849205, "learning_rate": 6.497991166011903e-05, "loss": 0.0042, "step": 43300 }, { "grad_norm": 0.08213905990123749, "learning_rate": 6.49641356569229e-05, "loss": 0.0037, "step": 43310 }, { "grad_norm": 0.113523930311203, "learning_rate": 6.494835801727135e-05, "loss": 0.0051, "step": 43320 }, { "grad_norm": 0.10271348804235458, "learning_rate": 6.493257874288986e-05, "loss": 0.004, "step": 43330 }, { "grad_norm": 0.11224208772182465, "learning_rate": 6.491679783550399e-05, "loss": 0.0044, "step": 43340 }, { "grad_norm": 0.07697148621082306, "learning_rate": 6.490101529683957e-05, "loss": 0.0042, "step": 43350 }, { "grad_norm": 0.19941598176956177, "learning_rate": 6.488523112862248e-05, "loss": 0.0038, "step": 43360 }, { "grad_norm": 0.08573053777217865, "learning_rate": 6.486944533257891e-05, "loss": 0.004, "step": 43370 }, { "grad_norm": 0.1189369335770607, "learning_rate": 6.485365791043517e-05, "loss": 0.0052, "step": 43380 }, { "grad_norm": 0.09889945387840271, "learning_rate": 6.483786886391772e-05, "loss": 0.0035, "step": 43390 }, { "grad_norm": 0.05639270320534706, "learning_rate": 6.482207819475323e-05, "loss": 0.0032, "step": 43400 }, { "grad_norm": 0.06885334849357605, "learning_rate": 6.480628590466857e-05, "loss": 0.0038, "step": 43410 }, { "grad_norm": 0.10021413862705231, "learning_rate": 6.479049199539074e-05, "loss": 0.0034, "step": 43420 }, { "grad_norm": 0.08782395720481873, "learning_rate": 6.477469646864693e-05, "loss": 0.0032, "step": 43430 }, { "grad_norm": 0.07479480654001236, "learning_rate": 6.475889932616454e-05, "loss": 0.0033, "step": 43440 }, { "grad_norm": 0.09503573924303055, "learning_rate": 6.474310056967111e-05, "loss": 0.0037, "step": 43450 }, { "grad_norm": 0.07757823914289474, "learning_rate": 6.472730020089437e-05, "loss": 0.004, "step": 43460 }, { "grad_norm": 0.10486495494842529, "learning_rate": 6.471149822156222e-05, "loss": 0.0041, "step": 43470 }, { "grad_norm": 0.0748593732714653, "learning_rate": 6.469569463340275e-05, "loss": 0.0041, "step": 43480 }, { "grad_norm": 0.11706521362066269, "learning_rate": 6.467988943814418e-05, "loss": 0.0046, "step": 43490 }, { "grad_norm": 0.06067875400185585, "learning_rate": 6.4664082637515e-05, "loss": 0.0036, "step": 43500 }, { "grad_norm": 0.11416208744049072, "learning_rate": 6.464827423324377e-05, "loss": 0.004, "step": 43510 }, { "grad_norm": 0.08325780928134918, "learning_rate": 6.463246422705931e-05, "loss": 0.0036, "step": 43520 }, { "grad_norm": 0.09031511843204498, "learning_rate": 6.461665262069055e-05, "loss": 0.004, "step": 43530 }, { "grad_norm": 0.0927891656756401, "learning_rate": 6.460083941586665e-05, "loss": 0.0037, "step": 43540 }, { "grad_norm": 0.09592081606388092, "learning_rate": 6.45850246143169e-05, "loss": 0.0043, "step": 43550 }, { "grad_norm": 0.09685412049293518, "learning_rate": 6.456920821777077e-05, "loss": 0.0053, "step": 43560 }, { "grad_norm": 0.08797494322061539, "learning_rate": 6.455339022795796e-05, "loss": 0.0038, "step": 43570 }, { "grad_norm": 0.06696321070194244, "learning_rate": 6.453757064660825e-05, "loss": 0.0032, "step": 43580 }, { "grad_norm": 0.07293689250946045, "learning_rate": 6.452174947545169e-05, "loss": 0.0039, "step": 43590 }, { "grad_norm": 0.07279811054468155, "learning_rate": 6.450592671621842e-05, "loss": 0.0046, "step": 43600 }, { "grad_norm": 0.07684590667486191, "learning_rate": 6.449010237063882e-05, "loss": 0.004, "step": 43610 }, { "grad_norm": 0.05360235646367073, "learning_rate": 6.447427644044343e-05, "loss": 0.0037, "step": 43620 }, { "grad_norm": 0.10697616636753082, "learning_rate": 6.44584489273629e-05, "loss": 0.0037, "step": 43630 }, { "grad_norm": 0.08509675413370132, "learning_rate": 6.444261983312817e-05, "loss": 0.0032, "step": 43640 }, { "grad_norm": 0.0902261883020401, "learning_rate": 6.442678915947023e-05, "loss": 0.0037, "step": 43650 }, { "grad_norm": 0.05908331647515297, "learning_rate": 6.441095690812032e-05, "loss": 0.0038, "step": 43660 }, { "grad_norm": 0.0625418871641159, "learning_rate": 6.439512308080982e-05, "loss": 0.0033, "step": 43670 }, { "grad_norm": 0.1005292683839798, "learning_rate": 6.43792876792703e-05, "loss": 0.004, "step": 43680 }, { "grad_norm": 0.08566047251224518, "learning_rate": 6.43634507052335e-05, "loss": 0.0036, "step": 43690 }, { "grad_norm": 0.11387030780315399, "learning_rate": 6.434761216043135e-05, "loss": 0.0046, "step": 43700 }, { "grad_norm": 0.10170263051986694, "learning_rate": 6.433177204659589e-05, "loss": 0.0046, "step": 43710 }, { "grad_norm": 0.0807969719171524, "learning_rate": 6.431593036545938e-05, "loss": 0.0039, "step": 43720 }, { "grad_norm": 0.07122869789600372, "learning_rate": 6.430008711875426e-05, "loss": 0.0039, "step": 43730 }, { "grad_norm": 0.06997791677713394, "learning_rate": 6.42842423082131e-05, "loss": 0.003, "step": 43740 }, { "grad_norm": 0.06168210506439209, "learning_rate": 6.426839593556868e-05, "loss": 0.0034, "step": 43750 }, { "grad_norm": 0.11996796727180481, "learning_rate": 6.425254800255392e-05, "loss": 0.0044, "step": 43760 }, { "grad_norm": 0.06760716438293457, "learning_rate": 6.423669851090196e-05, "loss": 0.0033, "step": 43770 }, { "grad_norm": 0.07369088381528854, "learning_rate": 6.422084746234604e-05, "loss": 0.0037, "step": 43780 }, { "grad_norm": 0.0600915364921093, "learning_rate": 6.420499485861965e-05, "loss": 0.0028, "step": 43790 }, { "grad_norm": 0.08708053827285767, "learning_rate": 6.418914070145636e-05, "loss": 0.0046, "step": 43800 }, { "grad_norm": 0.07121000438928604, "learning_rate": 6.417328499258996e-05, "loss": 0.0045, "step": 43810 }, { "grad_norm": 0.07079613208770752, "learning_rate": 6.415742773375446e-05, "loss": 0.0035, "step": 43820 }, { "grad_norm": 0.22937944531440735, "learning_rate": 6.414156892668393e-05, "loss": 0.007, "step": 43830 }, { "grad_norm": 0.08112230151891708, "learning_rate": 6.412570857311267e-05, "loss": 0.0044, "step": 43840 }, { "grad_norm": 0.12354040890932083, "learning_rate": 6.410984667477518e-05, "loss": 0.0055, "step": 43850 }, { "grad_norm": 0.0912032425403595, "learning_rate": 6.409398323340607e-05, "loss": 0.004, "step": 43860 }, { "grad_norm": 0.10533691197633743, "learning_rate": 6.407811825074012e-05, "loss": 0.0042, "step": 43870 }, { "grad_norm": 0.06500043720006943, "learning_rate": 6.406225172851234e-05, "loss": 0.0036, "step": 43880 }, { "grad_norm": 0.0902094617486, "learning_rate": 6.404638366845786e-05, "loss": 0.0046, "step": 43890 }, { "grad_norm": 0.10472533106803894, "learning_rate": 6.403051407231196e-05, "loss": 0.0039, "step": 43900 }, { "grad_norm": 0.09717769920825958, "learning_rate": 6.401464294181016e-05, "loss": 0.0047, "step": 43910 }, { "grad_norm": 0.06452937424182892, "learning_rate": 6.399877027868808e-05, "loss": 0.0038, "step": 43920 }, { "grad_norm": 0.07786591351032257, "learning_rate": 6.39828960846815e-05, "loss": 0.0038, "step": 43930 }, { "grad_norm": 0.07237919420003891, "learning_rate": 6.396702036152644e-05, "loss": 0.0038, "step": 43940 }, { "grad_norm": 0.0710383802652359, "learning_rate": 6.395114311095905e-05, "loss": 0.0036, "step": 43950 }, { "grad_norm": 0.0759420171380043, "learning_rate": 6.393526433471559e-05, "loss": 0.0035, "step": 43960 }, { "grad_norm": 0.0723726898431778, "learning_rate": 6.391938403453262e-05, "loss": 0.0045, "step": 43970 }, { "grad_norm": 0.10852058976888657, "learning_rate": 6.390350221214671e-05, "loss": 0.0048, "step": 43980 }, { "grad_norm": 0.11887418478727341, "learning_rate": 6.38876188692947e-05, "loss": 0.0044, "step": 43990 }, { "grad_norm": 0.06817133724689484, "learning_rate": 6.387173400771358e-05, "loss": 0.0037, "step": 44000 }, { "grad_norm": 0.05934031680226326, "learning_rate": 6.385584762914047e-05, "loss": 0.0024, "step": 44010 }, { "grad_norm": 0.06996594369411469, "learning_rate": 6.38399597353127e-05, "loss": 0.0027, "step": 44020 }, { "grad_norm": 0.06130514666438103, "learning_rate": 6.382407032796775e-05, "loss": 0.003, "step": 44030 }, { "grad_norm": 0.06029128283262253, "learning_rate": 6.380817940884325e-05, "loss": 0.0031, "step": 44040 }, { "grad_norm": 0.13361404836177826, "learning_rate": 6.379228697967702e-05, "loss": 0.0043, "step": 44050 }, { "grad_norm": 0.09766858071088791, "learning_rate": 6.3776393042207e-05, "loss": 0.0035, "step": 44060 }, { "grad_norm": 0.07929905503988266, "learning_rate": 6.376049759817137e-05, "loss": 0.0036, "step": 44070 }, { "grad_norm": 0.0694841593503952, "learning_rate": 6.37446006493084e-05, "loss": 0.0052, "step": 44080 }, { "grad_norm": 0.10911276936531067, "learning_rate": 6.372870219735656e-05, "loss": 0.005, "step": 44090 }, { "grad_norm": 0.0862378180027008, "learning_rate": 6.371280224405452e-05, "loss": 0.0049, "step": 44100 }, { "grad_norm": 0.09076273441314697, "learning_rate": 6.369690079114101e-05, "loss": 0.0051, "step": 44110 }, { "grad_norm": 0.11501624435186386, "learning_rate": 6.368099784035504e-05, "loss": 0.0055, "step": 44120 }, { "grad_norm": 0.06385966390371323, "learning_rate": 6.366509339343572e-05, "loss": 0.0041, "step": 44130 }, { "grad_norm": 0.15113505721092224, "learning_rate": 6.364918745212232e-05, "loss": 0.0046, "step": 44140 }, { "grad_norm": 0.10578490793704987, "learning_rate": 6.363328001815431e-05, "loss": 0.0054, "step": 44150 }, { "grad_norm": 0.09376710653305054, "learning_rate": 6.361737109327128e-05, "loss": 0.0051, "step": 44160 }, { "grad_norm": 0.06934380531311035, "learning_rate": 6.360146067921301e-05, "loss": 0.0044, "step": 44170 }, { "grad_norm": 0.09577465057373047, "learning_rate": 6.35855487777195e-05, "loss": 0.0039, "step": 44180 }, { "grad_norm": 0.08461959660053253, "learning_rate": 6.356963539053076e-05, "loss": 0.0042, "step": 44190 }, { "grad_norm": 0.06492093205451965, "learning_rate": 6.355372051938711e-05, "loss": 0.0037, "step": 44200 }, { "grad_norm": 0.08712461590766907, "learning_rate": 6.353780416602894e-05, "loss": 0.0046, "step": 44210 }, { "grad_norm": 0.11078468710184097, "learning_rate": 6.352188633219689e-05, "loss": 0.0041, "step": 44220 }, { "grad_norm": 0.07135825604200363, "learning_rate": 6.350596701963166e-05, "loss": 0.0039, "step": 44230 }, { "grad_norm": 0.12741747498512268, "learning_rate": 6.349004623007419e-05, "loss": 0.006, "step": 44240 }, { "grad_norm": 0.09981782734394073, "learning_rate": 6.347412396526555e-05, "loss": 0.0058, "step": 44250 }, { "grad_norm": 0.0791914239525795, "learning_rate": 6.345820022694696e-05, "loss": 0.0047, "step": 44260 }, { "grad_norm": 0.07807297259569168, "learning_rate": 6.344227501685984e-05, "loss": 0.0051, "step": 44270 }, { "grad_norm": 0.11187047511339188, "learning_rate": 6.342634833674572e-05, "loss": 0.0049, "step": 44280 }, { "grad_norm": 0.06923051923513412, "learning_rate": 6.341042018834635e-05, "loss": 0.0037, "step": 44290 }, { "grad_norm": 0.09047011286020279, "learning_rate": 6.339449057340359e-05, "loss": 0.0032, "step": 44300 }, { "grad_norm": 0.09251344949007034, "learning_rate": 6.337855949365945e-05, "loss": 0.0046, "step": 44310 }, { "grad_norm": 0.08251911401748657, "learning_rate": 6.336262695085619e-05, "loss": 0.0043, "step": 44320 }, { "grad_norm": 0.09741627424955368, "learning_rate": 6.334669294673612e-05, "loss": 0.0049, "step": 44330 }, { "grad_norm": 0.09961596876382828, "learning_rate": 6.333075748304179e-05, "loss": 0.0049, "step": 44340 }, { "grad_norm": 0.09164240211248398, "learning_rate": 6.331482056151585e-05, "loss": 0.0045, "step": 44350 }, { "grad_norm": 0.0864352136850357, "learning_rate": 6.329888218390117e-05, "loss": 0.005, "step": 44360 }, { "grad_norm": 0.0778060257434845, "learning_rate": 6.328294235194072e-05, "loss": 0.0042, "step": 44370 }, { "grad_norm": 0.09009253978729248, "learning_rate": 6.326700106737765e-05, "loss": 0.004, "step": 44380 }, { "grad_norm": 0.09293077141046524, "learning_rate": 6.325105833195531e-05, "loss": 0.0039, "step": 44390 }, { "grad_norm": 0.0868525579571724, "learning_rate": 6.323511414741715e-05, "loss": 0.0036, "step": 44400 }, { "grad_norm": 0.08135561645030975, "learning_rate": 6.321916851550678e-05, "loss": 0.0032, "step": 44410 }, { "grad_norm": 0.06928332895040512, "learning_rate": 6.320322143796806e-05, "loss": 0.004, "step": 44420 }, { "grad_norm": 0.0739288330078125, "learning_rate": 6.318727291654488e-05, "loss": 0.0032, "step": 44430 }, { "grad_norm": 0.06197074428200722, "learning_rate": 6.317132295298134e-05, "loss": 0.0027, "step": 44440 }, { "grad_norm": 0.08652407675981522, "learning_rate": 6.315537154902173e-05, "loss": 0.0035, "step": 44450 }, { "grad_norm": 0.07110317051410675, "learning_rate": 6.313941870641048e-05, "loss": 0.0038, "step": 44460 }, { "grad_norm": 0.0719558447599411, "learning_rate": 6.312346442689214e-05, "loss": 0.0039, "step": 44470 }, { "grad_norm": 0.08370382338762283, "learning_rate": 6.310750871221147e-05, "loss": 0.0053, "step": 44480 }, { "grad_norm": 0.06900815665721893, "learning_rate": 6.309155156411335e-05, "loss": 0.0036, "step": 44490 }, { "grad_norm": 0.1122586652636528, "learning_rate": 6.307559298434284e-05, "loss": 0.0064, "step": 44500 }, { "grad_norm": 0.10027293860912323, "learning_rate": 6.305963297464512e-05, "loss": 0.0039, "step": 44510 }, { "grad_norm": 0.06909660249948502, "learning_rate": 6.304367153676561e-05, "loss": 0.0041, "step": 44520 }, { "grad_norm": 0.06036994978785515, "learning_rate": 6.302770867244978e-05, "loss": 0.0055, "step": 44530 }, { "grad_norm": 0.07807517796754837, "learning_rate": 6.301174438344328e-05, "loss": 0.0044, "step": 44540 }, { "grad_norm": 0.10380709171295166, "learning_rate": 6.299577867149202e-05, "loss": 0.0055, "step": 44550 }, { "grad_norm": 0.133942648768425, "learning_rate": 6.297981153834192e-05, "loss": 0.0044, "step": 44560 }, { "grad_norm": 0.11213048547506332, "learning_rate": 6.296384298573916e-05, "loss": 0.0059, "step": 44570 }, { "grad_norm": 0.12352249026298523, "learning_rate": 6.294787301543001e-05, "loss": 0.0049, "step": 44580 }, { "grad_norm": 0.08226925879716873, "learning_rate": 6.293190162916095e-05, "loss": 0.0042, "step": 44590 }, { "grad_norm": 0.10343819856643677, "learning_rate": 6.291592882867855e-05, "loss": 0.0032, "step": 44600 }, { "grad_norm": 0.1128518283367157, "learning_rate": 6.28999546157296e-05, "loss": 0.0048, "step": 44610 }, { "grad_norm": 0.08154880255460739, "learning_rate": 6.288397899206102e-05, "loss": 0.0041, "step": 44620 }, { "grad_norm": 0.09272228181362152, "learning_rate": 6.286800195941984e-05, "loss": 0.0038, "step": 44630 }, { "grad_norm": 0.07603482902050018, "learning_rate": 6.285202351955334e-05, "loss": 0.0031, "step": 44640 }, { "grad_norm": 0.10056179761886597, "learning_rate": 6.283604367420887e-05, "loss": 0.0035, "step": 44650 }, { "grad_norm": 0.08578193932771683, "learning_rate": 6.282006242513394e-05, "loss": 0.0036, "step": 44660 }, { "grad_norm": 0.08171916007995605, "learning_rate": 6.280407977407628e-05, "loss": 0.0035, "step": 44670 }, { "grad_norm": 0.14684666693210602, "learning_rate": 6.27880957227837e-05, "loss": 0.0054, "step": 44680 }, { "grad_norm": 0.0937536433339119, "learning_rate": 6.27721102730042e-05, "loss": 0.0053, "step": 44690 }, { "grad_norm": 0.08773371577262878, "learning_rate": 6.275612342648592e-05, "loss": 0.0039, "step": 44700 }, { "grad_norm": 0.07818055152893066, "learning_rate": 6.274013518497716e-05, "loss": 0.0035, "step": 44710 }, { "grad_norm": 0.06835142523050308, "learning_rate": 6.272414555022636e-05, "loss": 0.0052, "step": 44720 }, { "grad_norm": 0.09111090749502182, "learning_rate": 6.270815452398215e-05, "loss": 0.0041, "step": 44730 }, { "grad_norm": 0.12051951885223389, "learning_rate": 6.269216210799326e-05, "loss": 0.0048, "step": 44740 }, { "grad_norm": 0.08365719020366669, "learning_rate": 6.26761683040086e-05, "loss": 0.0046, "step": 44750 }, { "grad_norm": 0.07047062367200851, "learning_rate": 6.266017311377723e-05, "loss": 0.0038, "step": 44760 }, { "grad_norm": 0.08197721838951111, "learning_rate": 6.264417653904839e-05, "loss": 0.0036, "step": 44770 }, { "grad_norm": 0.06288139522075653, "learning_rate": 6.262817858157139e-05, "loss": 0.0031, "step": 44780 }, { "grad_norm": 0.08771330863237381, "learning_rate": 6.261217924309576e-05, "loss": 0.0035, "step": 44790 }, { "grad_norm": 0.09371877461671829, "learning_rate": 6.259617852537118e-05, "loss": 0.005, "step": 44800 }, { "grad_norm": 0.11431269347667694, "learning_rate": 6.258017643014747e-05, "loss": 0.0062, "step": 44810 }, { "grad_norm": 0.07276929914951324, "learning_rate": 6.256417295917456e-05, "loss": 0.004, "step": 44820 }, { "grad_norm": 0.0973648875951767, "learning_rate": 6.254816811420258e-05, "loss": 0.0049, "step": 44830 }, { "grad_norm": 0.0965314731001854, "learning_rate": 6.253216189698183e-05, "loss": 0.004, "step": 44840 }, { "grad_norm": 0.12299703061580658, "learning_rate": 6.251615430926267e-05, "loss": 0.0039, "step": 44850 }, { "grad_norm": 0.07580216974020004, "learning_rate": 6.25001453527957e-05, "loss": 0.0039, "step": 44860 }, { "grad_norm": 0.09740325808525085, "learning_rate": 6.248413502933164e-05, "loss": 0.0042, "step": 44870 }, { "grad_norm": 0.4213807284832001, "learning_rate": 6.246812334062133e-05, "loss": 0.0038, "step": 44880 }, { "grad_norm": 0.11168172955513, "learning_rate": 6.245211028841579e-05, "loss": 0.0061, "step": 44890 }, { "grad_norm": 0.0901595875620842, "learning_rate": 6.24360958744662e-05, "loss": 0.0052, "step": 44900 }, { "grad_norm": 0.07057884335517883, "learning_rate": 6.242008010052387e-05, "loss": 0.0042, "step": 44910 }, { "grad_norm": 0.108303964138031, "learning_rate": 6.240406296834024e-05, "loss": 0.005, "step": 44920 }, { "grad_norm": 0.07287253439426422, "learning_rate": 6.238804447966694e-05, "loss": 0.004, "step": 44930 }, { "grad_norm": 0.07406919449567795, "learning_rate": 6.237202463625573e-05, "loss": 0.0052, "step": 44940 }, { "grad_norm": 0.03698865324258804, "learning_rate": 6.235600343985848e-05, "loss": 0.0036, "step": 44950 }, { "grad_norm": 0.0732077807188034, "learning_rate": 6.233998089222729e-05, "loss": 0.0034, "step": 44960 }, { "grad_norm": 0.07580117136240005, "learning_rate": 6.232395699511433e-05, "loss": 0.0029, "step": 44970 }, { "grad_norm": 0.058666568249464035, "learning_rate": 6.230793175027199e-05, "loss": 0.0037, "step": 44980 }, { "grad_norm": 0.08944573253393173, "learning_rate": 6.22919051594527e-05, "loss": 0.0044, "step": 44990 }, { "grad_norm": 0.1199757382273674, "learning_rate": 6.227587722440917e-05, "loss": 0.0043, "step": 45000 }, { "grad_norm": 0.1154218539595604, "learning_rate": 6.225984794689414e-05, "loss": 0.0047, "step": 45010 }, { "grad_norm": 0.11654717475175858, "learning_rate": 6.22438173286606e-05, "loss": 0.0053, "step": 45020 }, { "grad_norm": 0.08468528836965561, "learning_rate": 6.222778537146159e-05, "loss": 0.0042, "step": 45030 }, { "grad_norm": 0.11236162483692169, "learning_rate": 6.221175207705037e-05, "loss": 0.0057, "step": 45040 }, { "grad_norm": 0.09456054866313934, "learning_rate": 6.219571744718027e-05, "loss": 0.0043, "step": 45050 }, { "grad_norm": 0.056393515318632126, "learning_rate": 6.217968148360487e-05, "loss": 0.0038, "step": 45060 }, { "grad_norm": 0.09226938337087631, "learning_rate": 6.216364418807782e-05, "loss": 0.0048, "step": 45070 }, { "grad_norm": 0.07852716743946075, "learning_rate": 6.214760556235292e-05, "loss": 0.0035, "step": 45080 }, { "grad_norm": 0.17241136729717255, "learning_rate": 6.213156560818414e-05, "loss": 0.0034, "step": 45090 }, { "grad_norm": 0.1066536158323288, "learning_rate": 6.211552432732559e-05, "loss": 0.0042, "step": 45100 }, { "grad_norm": 0.08376219123601913, "learning_rate": 6.20994817215315e-05, "loss": 0.0044, "step": 45110 }, { "grad_norm": 0.10075085610151291, "learning_rate": 6.208343779255628e-05, "loss": 0.0039, "step": 45120 }, { "grad_norm": 0.07602208852767944, "learning_rate": 6.206739254215449e-05, "loss": 0.0037, "step": 45130 }, { "grad_norm": 0.12535026669502258, "learning_rate": 6.205134597208077e-05, "loss": 0.0047, "step": 45140 }, { "grad_norm": 0.06795518845319748, "learning_rate": 6.203529808409e-05, "loss": 0.0041, "step": 45150 }, { "grad_norm": 0.08926834166049957, "learning_rate": 6.201924887993712e-05, "loss": 0.004, "step": 45160 }, { "grad_norm": 0.11313740909099579, "learning_rate": 6.200319836137725e-05, "loss": 0.0042, "step": 45170 }, { "grad_norm": 0.08098924160003662, "learning_rate": 6.198714653016565e-05, "loss": 0.0041, "step": 45180 }, { "grad_norm": 0.14666995406150818, "learning_rate": 6.197109338805774e-05, "loss": 0.0043, "step": 45190 }, { "grad_norm": 0.06630445271730423, "learning_rate": 6.195503893680903e-05, "loss": 0.0045, "step": 45200 }, { "grad_norm": 0.07933533936738968, "learning_rate": 6.193898317817524e-05, "loss": 0.0043, "step": 45210 }, { "grad_norm": 0.06720510870218277, "learning_rate": 6.192292611391221e-05, "loss": 0.0036, "step": 45220 }, { "grad_norm": 0.0696425586938858, "learning_rate": 6.190686774577591e-05, "loss": 0.003, "step": 45230 }, { "grad_norm": 0.04971780255436897, "learning_rate": 6.189080807552245e-05, "loss": 0.0034, "step": 45240 }, { "grad_norm": 0.07391475141048431, "learning_rate": 6.187474710490809e-05, "loss": 0.0042, "step": 45250 }, { "grad_norm": 0.06703752279281616, "learning_rate": 6.185868483568926e-05, "loss": 0.0037, "step": 45260 }, { "grad_norm": 0.10171529650688171, "learning_rate": 6.184262126962245e-05, "loss": 0.0041, "step": 45270 }, { "grad_norm": 0.08804751187562943, "learning_rate": 6.182655640846442e-05, "loss": 0.0035, "step": 45280 }, { "grad_norm": 0.07004863023757935, "learning_rate": 6.181049025397196e-05, "loss": 0.004, "step": 45290 }, { "grad_norm": 0.06330125778913498, "learning_rate": 6.179442280790202e-05, "loss": 0.0049, "step": 45300 }, { "grad_norm": 0.0935438722372055, "learning_rate": 6.177835407201174e-05, "loss": 0.0048, "step": 45310 }, { "grad_norm": 0.08238158375024796, "learning_rate": 6.176228404805839e-05, "loss": 0.0031, "step": 45320 }, { "grad_norm": 0.08186393231153488, "learning_rate": 6.174621273779932e-05, "loss": 0.0029, "step": 45330 }, { "grad_norm": 0.0940820649266243, "learning_rate": 6.17301401429921e-05, "loss": 0.0034, "step": 45340 }, { "grad_norm": 0.06899750977754593, "learning_rate": 6.171406626539436e-05, "loss": 0.0043, "step": 45350 }, { "grad_norm": 0.10865581035614014, "learning_rate": 6.169799110676398e-05, "loss": 0.0033, "step": 45360 }, { "grad_norm": 0.1517678201198578, "learning_rate": 6.168191466885885e-05, "loss": 0.0045, "step": 45370 }, { "grad_norm": 0.05672434717416763, "learning_rate": 6.166583695343711e-05, "loss": 0.0042, "step": 45380 }, { "grad_norm": 0.08202294260263443, "learning_rate": 6.164975796225698e-05, "loss": 0.0047, "step": 45390 }, { "grad_norm": 0.0756060853600502, "learning_rate": 6.163367769707683e-05, "loss": 0.0052, "step": 45400 }, { "grad_norm": 0.08340366184711456, "learning_rate": 6.161759615965519e-05, "loss": 0.0042, "step": 45410 }, { "grad_norm": 0.10845906287431717, "learning_rate": 6.16015133517507e-05, "loss": 0.0039, "step": 45420 }, { "grad_norm": 0.29251760244369507, "learning_rate": 6.158542927512214e-05, "loss": 0.0036, "step": 45430 }, { "grad_norm": 0.18289685249328613, "learning_rate": 6.156934393152846e-05, "loss": 0.0047, "step": 45440 }, { "grad_norm": 0.13870683312416077, "learning_rate": 6.15532573227287e-05, "loss": 0.0048, "step": 45450 }, { "grad_norm": 0.08129812777042389, "learning_rate": 6.153716945048212e-05, "loss": 0.0051, "step": 45460 }, { "grad_norm": 0.15514232218265533, "learning_rate": 6.152108031654802e-05, "loss": 0.0053, "step": 45470 }, { "grad_norm": 0.07615087926387787, "learning_rate": 6.15049899226859e-05, "loss": 0.0041, "step": 45480 }, { "grad_norm": 0.07515482604503632, "learning_rate": 6.148889827065537e-05, "loss": 0.0041, "step": 45490 }, { "grad_norm": 0.05614819750189781, "learning_rate": 6.147280536221622e-05, "loss": 0.0034, "step": 45500 }, { "grad_norm": 0.07551544904708862, "learning_rate": 6.145671119912832e-05, "loss": 0.0033, "step": 45510 }, { "grad_norm": 0.09511536359786987, "learning_rate": 6.144061578315169e-05, "loss": 0.005, "step": 45520 }, { "grad_norm": 0.0825226753950119, "learning_rate": 6.142451911604654e-05, "loss": 0.0045, "step": 45530 }, { "grad_norm": 0.10399632155895233, "learning_rate": 6.140842119957315e-05, "loss": 0.0039, "step": 45540 }, { "grad_norm": 0.10618321597576141, "learning_rate": 6.139232203549197e-05, "loss": 0.0057, "step": 45550 }, { "grad_norm": 0.10892964154481888, "learning_rate": 6.137622162556357e-05, "loss": 0.0044, "step": 45560 }, { "grad_norm": 0.12258482724428177, "learning_rate": 6.13601199715487e-05, "loss": 0.005, "step": 45570 }, { "grad_norm": 0.07556363940238953, "learning_rate": 6.134401707520816e-05, "loss": 0.0042, "step": 45580 }, { "grad_norm": 0.07933824509382248, "learning_rate": 6.132791293830298e-05, "loss": 0.0042, "step": 45590 }, { "grad_norm": 0.07199857383966446, "learning_rate": 6.131180756259428e-05, "loss": 0.0043, "step": 45600 }, { "grad_norm": 0.06608004122972488, "learning_rate": 6.129570094984331e-05, "loss": 0.0037, "step": 45610 }, { "grad_norm": 0.07286589592695236, "learning_rate": 6.127959310181145e-05, "loss": 0.0038, "step": 45620 }, { "grad_norm": 0.08318153768777847, "learning_rate": 6.126348402026026e-05, "loss": 0.0036, "step": 45630 }, { "grad_norm": 0.09982669353485107, "learning_rate": 6.12473737069514e-05, "loss": 0.0044, "step": 45640 }, { "grad_norm": 0.08103223890066147, "learning_rate": 6.123126216364665e-05, "loss": 0.003, "step": 45650 }, { "grad_norm": 0.08087613433599472, "learning_rate": 6.121514939210797e-05, "loss": 0.004, "step": 45660 }, { "grad_norm": 0.09519381821155548, "learning_rate": 6.119903539409741e-05, "loss": 0.0053, "step": 45670 }, { "grad_norm": 0.06851623952388763, "learning_rate": 6.118292017137716e-05, "loss": 0.0058, "step": 45680 }, { "grad_norm": 0.08809620887041092, "learning_rate": 6.116680372570959e-05, "loss": 0.0035, "step": 45690 }, { "grad_norm": 0.11609477549791336, "learning_rate": 6.115068605885713e-05, "loss": 0.0033, "step": 45700 }, { "grad_norm": 0.08297321200370789, "learning_rate": 6.113456717258243e-05, "loss": 0.0039, "step": 45710 }, { "grad_norm": 0.07547511905431747, "learning_rate": 6.11184470686482e-05, "loss": 0.0031, "step": 45720 }, { "grad_norm": 0.20221032202243805, "learning_rate": 6.110232574881731e-05, "loss": 0.0031, "step": 45730 }, { "grad_norm": 0.06311823427677155, "learning_rate": 6.108620321485277e-05, "loss": 0.0032, "step": 45740 }, { "grad_norm": 0.23696891963481903, "learning_rate": 6.107007946851773e-05, "loss": 0.0037, "step": 45750 }, { "grad_norm": 0.09168621152639389, "learning_rate": 6.105395451157542e-05, "loss": 0.0068, "step": 45760 }, { "grad_norm": 0.1978285163640976, "learning_rate": 6.103782834578928e-05, "loss": 0.0062, "step": 45770 }, { "grad_norm": 0.10534492135047913, "learning_rate": 6.102170097292281e-05, "loss": 0.0065, "step": 45780 }, { "grad_norm": 0.11988793313503265, "learning_rate": 6.1005572394739716e-05, "loss": 0.0055, "step": 45790 }, { "grad_norm": 0.06082630529999733, "learning_rate": 6.0989442613003765e-05, "loss": 0.0043, "step": 45800 }, { "grad_norm": 0.08853461593389511, "learning_rate": 6.097331162947888e-05, "loss": 0.004, "step": 45810 }, { "grad_norm": 0.08051014691591263, "learning_rate": 6.095717944592914e-05, "loss": 0.0043, "step": 45820 }, { "grad_norm": 0.07862603664398193, "learning_rate": 6.094104606411873e-05, "loss": 0.003, "step": 45830 }, { "grad_norm": 0.08399216830730438, "learning_rate": 6.0924911485811966e-05, "loss": 0.0036, "step": 45840 }, { "grad_norm": 0.12154195457696915, "learning_rate": 6.090877571277331e-05, "loss": 0.0033, "step": 45850 }, { "grad_norm": 0.08724801987409592, "learning_rate": 6.089263874676736e-05, "loss": 0.0032, "step": 45860 }, { "grad_norm": 0.12010657042264938, "learning_rate": 6.0876500589558796e-05, "loss": 0.0037, "step": 45870 }, { "grad_norm": 0.06728079915046692, "learning_rate": 6.086036124291248e-05, "loss": 0.0036, "step": 45880 }, { "grad_norm": 0.11202571541070938, "learning_rate": 6.084422070859339e-05, "loss": 0.0066, "step": 45890 }, { "grad_norm": 0.22761695086956024, "learning_rate": 6.082807898836663e-05, "loss": 0.0061, "step": 45900 }, { "grad_norm": 0.12137458473443985, "learning_rate": 6.081193608399742e-05, "loss": 0.0054, "step": 45910 }, { "grad_norm": 0.08612369745969772, "learning_rate": 6.0795791997251164e-05, "loss": 0.0045, "step": 45920 }, { "grad_norm": 0.07842282950878143, "learning_rate": 6.0779646729893294e-05, "loss": 0.004, "step": 45930 }, { "grad_norm": 0.07562674582004547, "learning_rate": 6.0763500283689476e-05, "loss": 0.0041, "step": 45940 }, { "grad_norm": 0.1986623853445053, "learning_rate": 6.0747352660405455e-05, "loss": 0.0041, "step": 45950 }, { "grad_norm": 0.12509533762931824, "learning_rate": 6.073120386180709e-05, "loss": 0.0038, "step": 45960 }, { "grad_norm": 0.08611192554235458, "learning_rate": 6.0715053889660425e-05, "loss": 0.0051, "step": 45970 }, { "grad_norm": 0.08004720509052277, "learning_rate": 6.069890274573157e-05, "loss": 0.0053, "step": 45980 }, { "grad_norm": 0.13036863505840302, "learning_rate": 6.068275043178679e-05, "loss": 0.0069, "step": 45990 }, { "grad_norm": 0.1407131850719452, "learning_rate": 6.066659694959248e-05, "loss": 0.0067, "step": 46000 }, { "grad_norm": 0.10022412985563278, "learning_rate": 6.0650442300915176e-05, "loss": 0.0068, "step": 46010 }, { "grad_norm": 0.12257987260818481, "learning_rate": 6.063428648752152e-05, "loss": 0.0061, "step": 46020 }, { "grad_norm": 0.10324115306138992, "learning_rate": 6.0618129511178266e-05, "loss": 0.0044, "step": 46030 }, { "grad_norm": 0.236689031124115, "learning_rate": 6.060197137365234e-05, "loss": 0.0048, "step": 46040 }, { "grad_norm": 0.12427639216184616, "learning_rate": 6.058581207671077e-05, "loss": 0.0056, "step": 46050 }, { "grad_norm": 0.07053434103727341, "learning_rate": 6.056965162212072e-05, "loss": 0.0043, "step": 46060 }, { "grad_norm": 0.09933178871870041, "learning_rate": 6.055349001164943e-05, "loss": 0.004, "step": 46070 }, { "grad_norm": 0.07897479832172394, "learning_rate": 6.0537327247064347e-05, "loss": 0.0039, "step": 46080 }, { "grad_norm": 0.06046318635344505, "learning_rate": 6.0521163330133014e-05, "loss": 0.0031, "step": 46090 }, { "grad_norm": 0.07658398151397705, "learning_rate": 6.050499826262306e-05, "loss": 0.003, "step": 46100 }, { "grad_norm": 0.08920491486787796, "learning_rate": 6.0488832046302294e-05, "loss": 0.0041, "step": 46110 }, { "grad_norm": 0.106556236743927, "learning_rate": 6.0472664682938626e-05, "loss": 0.0045, "step": 46120 }, { "grad_norm": 0.2719743251800537, "learning_rate": 6.045649617430009e-05, "loss": 0.0041, "step": 46130 }, { "grad_norm": 0.10118387639522552, "learning_rate": 6.0440326522154866e-05, "loss": 0.0038, "step": 46140 }, { "grad_norm": 0.05726264417171478, "learning_rate": 6.0424155728271224e-05, "loss": 0.0046, "step": 46150 }, { "grad_norm": 0.07746601849794388, "learning_rate": 6.040798379441758e-05, "loss": 0.0047, "step": 46160 }, { "grad_norm": 0.08952849358320236, "learning_rate": 6.0391810722362485e-05, "loss": 0.0052, "step": 46170 }, { "grad_norm": 0.09434954077005386, "learning_rate": 6.037563651387458e-05, "loss": 0.004, "step": 46180 }, { "grad_norm": 0.05733263120055199, "learning_rate": 6.0359461170722666e-05, "loss": 0.0033, "step": 46190 }, { "grad_norm": 0.1254538595676422, "learning_rate": 6.034328469467566e-05, "loss": 0.0044, "step": 46200 }, { "grad_norm": 0.07488616555929184, "learning_rate": 6.0327107087502596e-05, "loss": 0.0044, "step": 46210 }, { "grad_norm": 0.09553037583827972, "learning_rate": 6.031092835097262e-05, "loss": 0.005, "step": 46220 }, { "grad_norm": 0.10588240623474121, "learning_rate": 6.0294748486855024e-05, "loss": 0.0052, "step": 46230 }, { "grad_norm": 0.09087708592414856, "learning_rate": 6.0278567496919216e-05, "loss": 0.0044, "step": 46240 }, { "grad_norm": 0.06566695123910904, "learning_rate": 6.026238538293472e-05, "loss": 0.0035, "step": 46250 }, { "grad_norm": 0.06901651620864868, "learning_rate": 6.024620214667118e-05, "loss": 0.0035, "step": 46260 }, { "grad_norm": 0.06180638447403908, "learning_rate": 6.0230017789898384e-05, "loss": 0.0036, "step": 46270 }, { "grad_norm": 0.14657776057720184, "learning_rate": 6.021383231438622e-05, "loss": 0.005, "step": 46280 }, { "grad_norm": 0.19954220950603485, "learning_rate": 6.0197645721904704e-05, "loss": 0.0048, "step": 46290 }, { "grad_norm": 0.12110946327447891, "learning_rate": 6.0181458014224e-05, "loss": 0.0055, "step": 46300 }, { "grad_norm": 0.07836192101240158, "learning_rate": 6.016526919311435e-05, "loss": 0.0036, "step": 46310 }, { "grad_norm": 0.31662464141845703, "learning_rate": 6.014907926034613e-05, "loss": 0.0042, "step": 46320 }, { "grad_norm": 0.07260707765817642, "learning_rate": 6.0132888217689875e-05, "loss": 0.0035, "step": 46330 }, { "grad_norm": 0.10373737663030624, "learning_rate": 6.01166960669162e-05, "loss": 0.0037, "step": 46340 }, { "grad_norm": 0.056214869022369385, "learning_rate": 6.0100502809795845e-05, "loss": 0.0031, "step": 46350 }, { "grad_norm": 0.08422140032052994, "learning_rate": 6.00843084480997e-05, "loss": 0.0046, "step": 46360 }, { "grad_norm": 0.0943451076745987, "learning_rate": 6.0068112983598736e-05, "loss": 0.0042, "step": 46370 }, { "grad_norm": 0.09329117089509964, "learning_rate": 6.005191641806407e-05, "loss": 0.0042, "step": 46380 }, { "grad_norm": 0.07161585241556168, "learning_rate": 6.003571875326694e-05, "loss": 0.0055, "step": 46390 }, { "grad_norm": 0.1155005693435669, "learning_rate": 6.00195199909787e-05, "loss": 0.0054, "step": 46400 }, { "grad_norm": 0.07504066079854965, "learning_rate": 6.000332013297082e-05, "loss": 0.0043, "step": 46410 }, { "grad_norm": 0.1269698590040207, "learning_rate": 5.998711918101487e-05, "loss": 0.0061, "step": 46420 }, { "grad_norm": 0.11081513017416, "learning_rate": 5.997091713688261e-05, "loss": 0.0051, "step": 46430 }, { "grad_norm": 0.10752743482589722, "learning_rate": 5.9954714002345836e-05, "loss": 0.0042, "step": 46440 }, { "grad_norm": 0.06720344722270966, "learning_rate": 5.993850977917649e-05, "loss": 0.0043, "step": 46450 }, { "grad_norm": 0.0953952744603157, "learning_rate": 5.992230446914667e-05, "loss": 0.0036, "step": 46460 }, { "grad_norm": 0.11725173890590668, "learning_rate": 5.990609807402855e-05, "loss": 0.0062, "step": 46470 }, { "grad_norm": 0.1197165995836258, "learning_rate": 5.988989059559443e-05, "loss": 0.0049, "step": 46480 }, { "grad_norm": 0.07670070976018906, "learning_rate": 5.987368203561675e-05, "loss": 0.0032, "step": 46490 }, { "grad_norm": 0.09256835281848907, "learning_rate": 5.9857472395868055e-05, "loss": 0.0036, "step": 46500 }, { "grad_norm": 0.10026693344116211, "learning_rate": 5.9841261678120983e-05, "loss": 0.0035, "step": 46510 }, { "grad_norm": 0.08662591129541397, "learning_rate": 5.982504988414834e-05, "loss": 0.0034, "step": 46520 }, { "grad_norm": 0.0877104252576828, "learning_rate": 5.9808837015723015e-05, "loss": 0.0035, "step": 46530 }, { "grad_norm": 0.09227906912565231, "learning_rate": 5.9792623074618016e-05, "loss": 0.0044, "step": 46540 }, { "grad_norm": 0.18565186858177185, "learning_rate": 5.977640806260648e-05, "loss": 0.0046, "step": 46550 }, { "grad_norm": 0.08920863270759583, "learning_rate": 5.976019198146166e-05, "loss": 0.004, "step": 46560 }, { "grad_norm": 0.09362363815307617, "learning_rate": 5.974397483295692e-05, "loss": 0.0034, "step": 46570 }, { "grad_norm": 0.11965528130531311, "learning_rate": 5.972775661886575e-05, "loss": 0.0042, "step": 46580 }, { "grad_norm": 0.05672524496912956, "learning_rate": 5.971153734096173e-05, "loss": 0.0032, "step": 46590 }, { "grad_norm": 0.08899004757404327, "learning_rate": 5.969531700101859e-05, "loss": 0.0041, "step": 46600 }, { "grad_norm": 0.05874719098210335, "learning_rate": 5.9679095600810155e-05, "loss": 0.0029, "step": 46610 }, { "grad_norm": 0.09511498361825943, "learning_rate": 5.9662873142110384e-05, "loss": 0.0037, "step": 46620 }, { "grad_norm": 0.08963128179311752, "learning_rate": 5.964664962669333e-05, "loss": 0.0038, "step": 46630 }, { "grad_norm": 0.07384523004293442, "learning_rate": 5.9630425056333186e-05, "loss": 0.004, "step": 46640 }, { "grad_norm": 0.09327306598424911, "learning_rate": 5.961419943280422e-05, "loss": 0.0043, "step": 46650 }, { "grad_norm": 0.08313500881195068, "learning_rate": 5.959797275788087e-05, "loss": 0.0049, "step": 46660 }, { "grad_norm": 0.07317472994327545, "learning_rate": 5.958174503333765e-05, "loss": 0.0033, "step": 46670 }, { "grad_norm": 0.08812014013528824, "learning_rate": 5.9565516260949195e-05, "loss": 0.0044, "step": 46680 }, { "grad_norm": 0.10301654040813446, "learning_rate": 5.9549286442490273e-05, "loss": 0.0045, "step": 46690 }, { "grad_norm": 0.07601505517959595, "learning_rate": 5.953305557973572e-05, "loss": 0.0041, "step": 46700 }, { "grad_norm": 0.08671147376298904, "learning_rate": 5.9516823674460564e-05, "loss": 0.0036, "step": 46710 }, { "grad_norm": 0.08983220905065536, "learning_rate": 5.9500590728439875e-05, "loss": 0.0053, "step": 46720 }, { "grad_norm": 0.07937533408403397, "learning_rate": 5.948435674344886e-05, "loss": 0.0045, "step": 46730 }, { "grad_norm": 0.062496840953826904, "learning_rate": 5.946812172126285e-05, "loss": 0.0032, "step": 46740 }, { "grad_norm": 0.08635279536247253, "learning_rate": 5.945188566365729e-05, "loss": 0.005, "step": 46750 }, { "grad_norm": 0.058153484016656876, "learning_rate": 5.943564857240773e-05, "loss": 0.0049, "step": 46760 }, { "grad_norm": 0.08209875971078873, "learning_rate": 5.941941044928981e-05, "loss": 0.0043, "step": 46770 }, { "grad_norm": 0.09760336577892303, "learning_rate": 5.940317129607935e-05, "loss": 0.0062, "step": 46780 }, { "grad_norm": 0.060428764671087265, "learning_rate": 5.9386931114552204e-05, "loss": 0.0035, "step": 46790 }, { "grad_norm": 0.06944669038057327, "learning_rate": 5.937068990648438e-05, "loss": 0.0048, "step": 46800 }, { "grad_norm": 0.0669902116060257, "learning_rate": 5.935444767365199e-05, "loss": 0.0044, "step": 46810 }, { "grad_norm": 0.10172561556100845, "learning_rate": 5.933820441783129e-05, "loss": 0.004, "step": 46820 }, { "grad_norm": 0.07683773338794708, "learning_rate": 5.932196014079857e-05, "loss": 0.0037, "step": 46830 }, { "grad_norm": 0.07998686283826828, "learning_rate": 5.930571484433032e-05, "loss": 0.0033, "step": 46840 }, { "grad_norm": 0.07528772205114365, "learning_rate": 5.928946853020309e-05, "loss": 0.0037, "step": 46850 }, { "grad_norm": 0.07507059723138809, "learning_rate": 5.927322120019352e-05, "loss": 0.004, "step": 46860 }, { "grad_norm": 0.10490697622299194, "learning_rate": 5.925697285607844e-05, "loss": 0.0036, "step": 46870 }, { "grad_norm": 0.07918145507574081, "learning_rate": 5.924072349963472e-05, "loss": 0.0039, "step": 46880 }, { "grad_norm": 0.07653272897005081, "learning_rate": 5.922447313263938e-05, "loss": 0.0028, "step": 46890 }, { "grad_norm": 0.08508153259754181, "learning_rate": 5.920822175686952e-05, "loss": 0.005, "step": 46900 }, { "grad_norm": 0.08258869498968124, "learning_rate": 5.919196937410239e-05, "loss": 0.0039, "step": 46910 }, { "grad_norm": 0.06313127279281616, "learning_rate": 5.917571598611529e-05, "loss": 0.0041, "step": 46920 }, { "grad_norm": 0.05556059256196022, "learning_rate": 5.9159461594685686e-05, "loss": 0.0037, "step": 46930 }, { "grad_norm": 0.08902191370725632, "learning_rate": 5.914320620159114e-05, "loss": 0.0041, "step": 46940 }, { "grad_norm": 0.07673894613981247, "learning_rate": 5.912694980860931e-05, "loss": 0.0035, "step": 46950 }, { "grad_norm": 0.08695141971111298, "learning_rate": 5.9110692417517964e-05, "loss": 0.0033, "step": 46960 }, { "grad_norm": 0.07709107547998428, "learning_rate": 5.9094434030095e-05, "loss": 0.0031, "step": 46970 }, { "grad_norm": 0.07135570049285889, "learning_rate": 5.907817464811841e-05, "loss": 0.0026, "step": 46980 }, { "grad_norm": 0.12867498397827148, "learning_rate": 5.906191427336627e-05, "loss": 0.004, "step": 46990 }, { "grad_norm": 0.10086789727210999, "learning_rate": 5.9045652907616835e-05, "loss": 0.0038, "step": 47000 }, { "grad_norm": 0.10916807502508163, "learning_rate": 5.902939055264838e-05, "loss": 0.0041, "step": 47010 }, { "grad_norm": 0.09256412833929062, "learning_rate": 5.901312721023935e-05, "loss": 0.0033, "step": 47020 }, { "grad_norm": 0.05480325594544411, "learning_rate": 5.8996862882168294e-05, "loss": 0.0029, "step": 47030 }, { "grad_norm": 0.07691085338592529, "learning_rate": 5.8980597570213826e-05, "loss": 0.0031, "step": 47040 }, { "grad_norm": 0.06244584172964096, "learning_rate": 5.896433127615471e-05, "loss": 0.0038, "step": 47050 }, { "grad_norm": 0.07198924571275711, "learning_rate": 5.894806400176981e-05, "loss": 0.0039, "step": 47060 }, { "grad_norm": 0.09861142188310623, "learning_rate": 5.893179574883808e-05, "loss": 0.0036, "step": 47070 }, { "grad_norm": 0.09172313660383224, "learning_rate": 5.8915526519138585e-05, "loss": 0.0043, "step": 47080 }, { "grad_norm": 0.12700803577899933, "learning_rate": 5.889925631445053e-05, "loss": 0.0041, "step": 47090 }, { "grad_norm": 0.07742566615343094, "learning_rate": 5.888298513655318e-05, "loss": 0.0036, "step": 47100 }, { "grad_norm": 0.06847143173217773, "learning_rate": 5.886671298722592e-05, "loss": 0.0038, "step": 47110 }, { "grad_norm": 0.07706289738416672, "learning_rate": 5.885043986824828e-05, "loss": 0.0042, "step": 47120 }, { "grad_norm": 0.07908019423484802, "learning_rate": 5.883416578139982e-05, "loss": 0.0031, "step": 47130 }, { "grad_norm": 0.06225522980093956, "learning_rate": 5.881789072846029e-05, "loss": 0.0043, "step": 47140 }, { "grad_norm": 0.087717704474926, "learning_rate": 5.8801614711209474e-05, "loss": 0.0034, "step": 47150 }, { "grad_norm": 0.09156187623739243, "learning_rate": 5.878533773142732e-05, "loss": 0.0032, "step": 47160 }, { "grad_norm": 0.0704856738448143, "learning_rate": 5.8769059790893834e-05, "loss": 0.0035, "step": 47170 }, { "grad_norm": 0.08793658763170242, "learning_rate": 5.8752780891389145e-05, "loss": 0.0043, "step": 47180 }, { "grad_norm": 0.0717587098479271, "learning_rate": 5.873650103469351e-05, "loss": 0.0032, "step": 47190 }, { "grad_norm": 0.06877795606851578, "learning_rate": 5.872022022258725e-05, "loss": 0.0029, "step": 47200 }, { "grad_norm": 0.08778207004070282, "learning_rate": 5.8703938456850814e-05, "loss": 0.0045, "step": 47210 }, { "grad_norm": 0.12402137368917465, "learning_rate": 5.868765573926477e-05, "loss": 0.0035, "step": 47220 }, { "grad_norm": 0.09398212283849716, "learning_rate": 5.867137207160974e-05, "loss": 0.0034, "step": 47230 }, { "grad_norm": 0.09262124449014664, "learning_rate": 5.86550874556665e-05, "loss": 0.0034, "step": 47240 }, { "grad_norm": 0.1049213856458664, "learning_rate": 5.863880189321592e-05, "loss": 0.0035, "step": 47250 }, { "grad_norm": 0.07055896520614624, "learning_rate": 5.862251538603894e-05, "loss": 0.0029, "step": 47260 }, { "grad_norm": 0.05745413526892662, "learning_rate": 5.8606227935916656e-05, "loss": 0.0038, "step": 47270 }, { "grad_norm": 0.08705846220254898, "learning_rate": 5.858993954463021e-05, "loss": 0.0032, "step": 47280 }, { "grad_norm": 0.06076744943857193, "learning_rate": 5.8573650213960896e-05, "loss": 0.0032, "step": 47290 }, { "grad_norm": 0.07909481972455978, "learning_rate": 5.8557359945690084e-05, "loss": 0.0033, "step": 47300 }, { "grad_norm": 0.06054552644491196, "learning_rate": 5.8541068741599246e-05, "loss": 0.0029, "step": 47310 }, { "grad_norm": 0.07449042797088623, "learning_rate": 5.8524776603469985e-05, "loss": 0.0029, "step": 47320 }, { "grad_norm": 0.08198460936546326, "learning_rate": 5.850848353308397e-05, "loss": 0.0034, "step": 47330 }, { "grad_norm": 0.07430458068847656, "learning_rate": 5.849218953222297e-05, "loss": 0.0029, "step": 47340 }, { "grad_norm": 0.07307735830545425, "learning_rate": 5.847589460266891e-05, "loss": 0.0028, "step": 47350 }, { "grad_norm": 0.06544483453035355, "learning_rate": 5.845959874620375e-05, "loss": 0.004, "step": 47360 }, { "grad_norm": 0.08034854382276535, "learning_rate": 5.84433019646096e-05, "loss": 0.0031, "step": 47370 }, { "grad_norm": 0.07946810126304626, "learning_rate": 5.842700425966863e-05, "loss": 0.0033, "step": 47380 }, { "grad_norm": 0.10837064683437347, "learning_rate": 5.841070563316315e-05, "loss": 0.0037, "step": 47390 }, { "grad_norm": 0.08964309096336365, "learning_rate": 5.839440608687554e-05, "loss": 0.0034, "step": 47400 }, { "grad_norm": 0.08427327871322632, "learning_rate": 5.837810562258831e-05, "loss": 0.0036, "step": 47410 }, { "grad_norm": 0.12992781400680542, "learning_rate": 5.836180424208405e-05, "loss": 0.0027, "step": 47420 }, { "grad_norm": 0.09200184792280197, "learning_rate": 5.834550194714543e-05, "loss": 0.0045, "step": 47430 }, { "grad_norm": 0.06511174142360687, "learning_rate": 5.832919873955528e-05, "loss": 0.004, "step": 47440 }, { "grad_norm": 0.08142392337322235, "learning_rate": 5.831289462109648e-05, "loss": 0.0043, "step": 47450 }, { "grad_norm": 0.06628294289112091, "learning_rate": 5.829658959355202e-05, "loss": 0.003, "step": 47460 }, { "grad_norm": 0.07514069229364395, "learning_rate": 5.8280283658704995e-05, "loss": 0.0038, "step": 47470 }, { "grad_norm": 0.07571639120578766, "learning_rate": 5.826397681833859e-05, "loss": 0.0028, "step": 47480 }, { "grad_norm": 0.0745522528886795, "learning_rate": 5.824766907423612e-05, "loss": 0.0044, "step": 47490 }, { "grad_norm": 0.08473149687051773, "learning_rate": 5.8231360428180945e-05, "loss": 0.004, "step": 47500 }, { "grad_norm": 0.07886895537376404, "learning_rate": 5.821505088195658e-05, "loss": 0.0042, "step": 47510 }, { "grad_norm": 0.05483851581811905, "learning_rate": 5.819874043734661e-05, "loss": 0.0035, "step": 47520 }, { "grad_norm": 0.07599373906850815, "learning_rate": 5.8182429096134695e-05, "loss": 0.0035, "step": 47530 }, { "grad_norm": 0.0907200276851654, "learning_rate": 5.816611686010465e-05, "loss": 0.0045, "step": 47540 }, { "grad_norm": 0.11849507689476013, "learning_rate": 5.814980373104033e-05, "loss": 0.0053, "step": 47550 }, { "grad_norm": 0.07857631891965866, "learning_rate": 5.813348971072572e-05, "loss": 0.0033, "step": 47560 }, { "grad_norm": 0.08459281176328659, "learning_rate": 5.811717480094492e-05, "loss": 0.0035, "step": 47570 }, { "grad_norm": 0.09532944858074188, "learning_rate": 5.810085900348209e-05, "loss": 0.0044, "step": 47580 }, { "grad_norm": 0.06460914015769958, "learning_rate": 5.8084542320121483e-05, "loss": 0.0036, "step": 47590 }, { "grad_norm": 0.08329907059669495, "learning_rate": 5.8068224752647497e-05, "loss": 0.0039, "step": 47600 }, { "grad_norm": 0.09797742962837219, "learning_rate": 5.805190630284457e-05, "loss": 0.0036, "step": 47610 }, { "grad_norm": 0.07134060561656952, "learning_rate": 5.803558697249729e-05, "loss": 0.0035, "step": 47620 }, { "grad_norm": 0.06034095212817192, "learning_rate": 5.8019266763390276e-05, "loss": 0.004, "step": 47630 }, { "grad_norm": 0.18379555642604828, "learning_rate": 5.800294567730833e-05, "loss": 0.0036, "step": 47640 }, { "grad_norm": 0.06582550704479218, "learning_rate": 5.798662371603628e-05, "loss": 0.0037, "step": 47650 }, { "grad_norm": 0.11550786346197128, "learning_rate": 5.797030088135904e-05, "loss": 0.0041, "step": 47660 }, { "grad_norm": 0.09515003859996796, "learning_rate": 5.7953977175061705e-05, "loss": 0.0043, "step": 47670 }, { "grad_norm": 0.08622969686985016, "learning_rate": 5.793765259892938e-05, "loss": 0.0049, "step": 47680 }, { "grad_norm": 0.15366142988204956, "learning_rate": 5.792132715474729e-05, "loss": 0.0054, "step": 47690 }, { "grad_norm": 0.11047866940498352, "learning_rate": 5.790500084430078e-05, "loss": 0.0042, "step": 47700 }, { "grad_norm": 0.09113182127475739, "learning_rate": 5.7888673669375274e-05, "loss": 0.0032, "step": 47710 }, { "grad_norm": 0.07248402386903763, "learning_rate": 5.787234563175625e-05, "loss": 0.0032, "step": 47720 }, { "grad_norm": 0.06419762223958969, "learning_rate": 5.7856016733229355e-05, "loss": 0.0041, "step": 47730 }, { "grad_norm": 0.09331068396568298, "learning_rate": 5.7839686975580297e-05, "loss": 0.0033, "step": 47740 }, { "grad_norm": 0.09022149443626404, "learning_rate": 5.782335636059484e-05, "loss": 0.0032, "step": 47750 }, { "grad_norm": 0.06426529586315155, "learning_rate": 5.780702489005889e-05, "loss": 0.0032, "step": 47760 }, { "grad_norm": 0.1056896448135376, "learning_rate": 5.779069256575845e-05, "loss": 0.0049, "step": 47770 }, { "grad_norm": 0.06553520262241364, "learning_rate": 5.7774359389479574e-05, "loss": 0.0046, "step": 47780 }, { "grad_norm": 0.1064959168434143, "learning_rate": 5.775802536300845e-05, "loss": 0.0032, "step": 47790 }, { "grad_norm": 0.0865098387002945, "learning_rate": 5.774169048813134e-05, "loss": 0.0039, "step": 47800 }, { "grad_norm": 0.2135966718196869, "learning_rate": 5.77253547666346e-05, "loss": 0.0049, "step": 47810 }, { "grad_norm": 0.2663615643978119, "learning_rate": 5.770901820030465e-05, "loss": 0.0061, "step": 47820 }, { "grad_norm": 0.13378380239009857, "learning_rate": 5.769268079092809e-05, "loss": 0.0064, "step": 47830 }, { "grad_norm": 0.09553011506795883, "learning_rate": 5.767634254029151e-05, "loss": 0.0047, "step": 47840 }, { "grad_norm": 0.13607379794120789, "learning_rate": 5.7660003450181655e-05, "loss": 0.0048, "step": 47850 }, { "grad_norm": 0.10232865810394287, "learning_rate": 5.764366352238534e-05, "loss": 0.0047, "step": 47860 }, { "grad_norm": 0.11482667177915573, "learning_rate": 5.7627322758689474e-05, "loss": 0.0046, "step": 47870 }, { "grad_norm": 0.08943463861942291, "learning_rate": 5.761098116088105e-05, "loss": 0.0033, "step": 47880 }, { "grad_norm": 0.06738237291574478, "learning_rate": 5.759463873074717e-05, "loss": 0.0042, "step": 47890 }, { "grad_norm": 0.06170113757252693, "learning_rate": 5.757829547007504e-05, "loss": 0.0049, "step": 47900 }, { "grad_norm": 0.07523016631603241, "learning_rate": 5.756195138065189e-05, "loss": 0.0033, "step": 47910 }, { "grad_norm": 0.10748061537742615, "learning_rate": 5.754560646426511e-05, "loss": 0.0034, "step": 47920 }, { "grad_norm": 0.07481628656387329, "learning_rate": 5.752926072270216e-05, "loss": 0.0041, "step": 47930 }, { "grad_norm": 0.07912256568670273, "learning_rate": 5.7512914157750563e-05, "loss": 0.0036, "step": 47940 }, { "grad_norm": 0.08560676127672195, "learning_rate": 5.749656677119798e-05, "loss": 0.0037, "step": 47950 }, { "grad_norm": 0.07376586645841599, "learning_rate": 5.7480218564832125e-05, "loss": 0.0031, "step": 47960 }, { "grad_norm": 0.06392541527748108, "learning_rate": 5.746386954044082e-05, "loss": 0.0029, "step": 47970 }, { "grad_norm": 0.05965283513069153, "learning_rate": 5.744751969981195e-05, "loss": 0.0031, "step": 47980 }, { "grad_norm": 0.06531985104084015, "learning_rate": 5.7431169044733526e-05, "loss": 0.0035, "step": 47990 }, { "grad_norm": 0.09622220695018768, "learning_rate": 5.741481757699364e-05, "loss": 0.0033, "step": 48000 }, { "grad_norm": 0.06131618842482567, "learning_rate": 5.7398465298380434e-05, "loss": 0.0029, "step": 48010 }, { "grad_norm": 0.050724469125270844, "learning_rate": 5.7382112210682193e-05, "loss": 0.0034, "step": 48020 }, { "grad_norm": 0.06709104031324387, "learning_rate": 5.7365758315687266e-05, "loss": 0.0023, "step": 48030 }, { "grad_norm": 0.3479957580566406, "learning_rate": 5.734940361518407e-05, "loss": 0.0038, "step": 48040 }, { "grad_norm": 0.06266792118549347, "learning_rate": 5.733304811096116e-05, "loss": 0.003, "step": 48050 }, { "grad_norm": 0.09334055334329605, "learning_rate": 5.731669180480713e-05, "loss": 0.0042, "step": 48060 }, { "grad_norm": 0.07761288434267044, "learning_rate": 5.730033469851067e-05, "loss": 0.004, "step": 48070 }, { "grad_norm": 0.07371708750724792, "learning_rate": 5.7283976793860607e-05, "loss": 0.0036, "step": 48080 }, { "grad_norm": 0.12272322177886963, "learning_rate": 5.726761809264577e-05, "loss": 0.0034, "step": 48090 }, { "grad_norm": 0.055357251316308975, "learning_rate": 5.7251258596655155e-05, "loss": 0.0033, "step": 48100 }, { "grad_norm": 0.10831006616353989, "learning_rate": 5.72348983076778e-05, "loss": 0.0039, "step": 48110 }, { "grad_norm": 0.10501453280448914, "learning_rate": 5.7218537227502854e-05, "loss": 0.0038, "step": 48120 }, { "grad_norm": 0.06302320212125778, "learning_rate": 5.720217535791951e-05, "loss": 0.0032, "step": 48130 }, { "grad_norm": 0.10168673098087311, "learning_rate": 5.718581270071711e-05, "loss": 0.006, "step": 48140 }, { "grad_norm": 0.08414867520332336, "learning_rate": 5.716944925768505e-05, "loss": 0.0051, "step": 48150 }, { "grad_norm": 0.08541394770145416, "learning_rate": 5.7153085030612786e-05, "loss": 0.0066, "step": 48160 }, { "grad_norm": 0.1047859936952591, "learning_rate": 5.713672002128989e-05, "loss": 0.0069, "step": 48170 }, { "grad_norm": 0.11274402588605881, "learning_rate": 5.712035423150602e-05, "loss": 0.0107, "step": 48180 }, { "grad_norm": 0.11018536984920502, "learning_rate": 5.710398766305094e-05, "loss": 0.0057, "step": 48190 }, { "grad_norm": 0.08214882761240005, "learning_rate": 5.708762031771442e-05, "loss": 0.0035, "step": 48200 }, { "grad_norm": 0.07021453976631165, "learning_rate": 5.707125219728642e-05, "loss": 0.0045, "step": 48210 }, { "grad_norm": 0.07106579840183258, "learning_rate": 5.7054883303556905e-05, "loss": 0.0038, "step": 48220 }, { "grad_norm": 0.0626407191157341, "learning_rate": 5.703851363831595e-05, "loss": 0.0042, "step": 48230 }, { "grad_norm": 0.08617711067199707, "learning_rate": 5.702214320335374e-05, "loss": 0.0034, "step": 48240 }, { "grad_norm": 0.09812811017036438, "learning_rate": 5.7005772000460514e-05, "loss": 0.004, "step": 48250 }, { "grad_norm": 0.09334176778793335, "learning_rate": 5.698940003142656e-05, "loss": 0.0032, "step": 48260 }, { "grad_norm": 0.12235451489686966, "learning_rate": 5.697302729804236e-05, "loss": 0.004, "step": 48270 }, { "grad_norm": 0.05534962937235832, "learning_rate": 5.695665380209837e-05, "loss": 0.0037, "step": 48280 }, { "grad_norm": 0.06488760560750961, "learning_rate": 5.6940279545385165e-05, "loss": 0.0045, "step": 48290 }, { "grad_norm": 0.08279679715633392, "learning_rate": 5.692390452969344e-05, "loss": 0.0036, "step": 48300 }, { "grad_norm": 0.08090221881866455, "learning_rate": 5.690752875681392e-05, "loss": 0.0034, "step": 48310 }, { "grad_norm": 0.11614003777503967, "learning_rate": 5.6891152228537435e-05, "loss": 0.0043, "step": 48320 }, { "grad_norm": 0.07643395662307739, "learning_rate": 5.687477494665492e-05, "loss": 0.0034, "step": 48330 }, { "grad_norm": 0.08989343792200089, "learning_rate": 5.685839691295733e-05, "loss": 0.0034, "step": 48340 }, { "grad_norm": 0.07882153987884521, "learning_rate": 5.6842018129235786e-05, "loss": 0.0032, "step": 48350 }, { "grad_norm": 0.11808833479881287, "learning_rate": 5.6825638597281404e-05, "loss": 0.0034, "step": 48360 }, { "grad_norm": 0.08138665556907654, "learning_rate": 5.680925831888546e-05, "loss": 0.0031, "step": 48370 }, { "grad_norm": 0.06235438585281372, "learning_rate": 5.6792877295839274e-05, "loss": 0.0028, "step": 48380 }, { "grad_norm": 0.09771546721458435, "learning_rate": 5.6776495529934224e-05, "loss": 0.003, "step": 48390 }, { "grad_norm": 0.08086904138326645, "learning_rate": 5.6760113022961824e-05, "loss": 0.0031, "step": 48400 }, { "grad_norm": 0.08691875636577606, "learning_rate": 5.6743729776713617e-05, "loss": 0.0037, "step": 48410 }, { "grad_norm": 0.0886790081858635, "learning_rate": 5.672734579298126e-05, "loss": 0.0032, "step": 48420 }, { "grad_norm": 0.08019509166479111, "learning_rate": 5.671096107355649e-05, "loss": 0.0039, "step": 48430 }, { "grad_norm": 0.07011198997497559, "learning_rate": 5.66945756202311e-05, "loss": 0.0038, "step": 48440 }, { "grad_norm": 0.09705504029989243, "learning_rate": 5.667818943479699e-05, "loss": 0.004, "step": 48450 }, { "grad_norm": 0.06881513446569443, "learning_rate": 5.666180251904612e-05, "loss": 0.0031, "step": 48460 }, { "grad_norm": 0.07833107560873032, "learning_rate": 5.6645414874770555e-05, "loss": 0.0033, "step": 48470 }, { "grad_norm": 0.06674269586801529, "learning_rate": 5.66290265037624e-05, "loss": 0.0033, "step": 48480 }, { "grad_norm": 0.0684233158826828, "learning_rate": 5.661263740781386e-05, "loss": 0.0031, "step": 48490 }, { "grad_norm": 0.06786816567182541, "learning_rate": 5.6596247588717254e-05, "loss": 0.0043, "step": 48500 }, { "grad_norm": 0.1355637013912201, "learning_rate": 5.6579857048264926e-05, "loss": 0.0041, "step": 48510 }, { "grad_norm": 0.10163883119821548, "learning_rate": 5.6563465788249314e-05, "loss": 0.0048, "step": 48520 }, { "grad_norm": 0.0923035740852356, "learning_rate": 5.6547073810462956e-05, "loss": 0.004, "step": 48530 }, { "grad_norm": 0.07101649045944214, "learning_rate": 5.653068111669846e-05, "loss": 0.0033, "step": 48540 }, { "grad_norm": 0.12385807186365128, "learning_rate": 5.651428770874848e-05, "loss": 0.0051, "step": 48550 }, { "grad_norm": 0.10292641818523407, "learning_rate": 5.64978935884058e-05, "loss": 0.0052, "step": 48560 }, { "grad_norm": 0.12090523540973663, "learning_rate": 5.6481498757463244e-05, "loss": 0.0056, "step": 48570 }, { "grad_norm": 0.05349336937069893, "learning_rate": 5.646510321771373e-05, "loss": 0.0042, "step": 48580 }, { "grad_norm": 0.0760534331202507, "learning_rate": 5.644870697095024e-05, "loss": 0.0035, "step": 48590 }, { "grad_norm": 0.0738903060555458, "learning_rate": 5.643231001896586e-05, "loss": 0.0043, "step": 48600 }, { "grad_norm": 0.0791177824139595, "learning_rate": 5.6415912363553726e-05, "loss": 0.0038, "step": 48610 }, { "grad_norm": 0.07364539057016373, "learning_rate": 5.639951400650706e-05, "loss": 0.0042, "step": 48620 }, { "grad_norm": 0.10825517028570175, "learning_rate": 5.6383114949619165e-05, "loss": 0.0035, "step": 48630 }, { "grad_norm": 0.08716508001089096, "learning_rate": 5.636671519468342e-05, "loss": 0.0044, "step": 48640 }, { "grad_norm": 0.10866730660200119, "learning_rate": 5.635031474349327e-05, "loss": 0.0061, "step": 48650 }, { "grad_norm": 0.11844734847545624, "learning_rate": 5.6333913597842246e-05, "loss": 0.0058, "step": 48660 }, { "grad_norm": 0.10583063215017319, "learning_rate": 5.6317511759523955e-05, "loss": 0.0055, "step": 48670 }, { "grad_norm": 0.09283234924077988, "learning_rate": 5.630110923033207e-05, "loss": 0.0061, "step": 48680 }, { "grad_norm": 0.09263809770345688, "learning_rate": 5.628470601206036e-05, "loss": 0.0049, "step": 48690 }, { "grad_norm": 0.07020828872919083, "learning_rate": 5.626830210650266e-05, "loss": 0.0044, "step": 48700 }, { "grad_norm": 0.07945916801691055, "learning_rate": 5.625189751545285e-05, "loss": 0.0035, "step": 48710 }, { "grad_norm": 0.06648362427949905, "learning_rate": 5.6235492240704936e-05, "loss": 0.0041, "step": 48720 }, { "grad_norm": 0.060071639716625214, "learning_rate": 5.621908628405296e-05, "loss": 0.0039, "step": 48730 }, { "grad_norm": 0.07317759096622467, "learning_rate": 5.620267964729106e-05, "loss": 0.0031, "step": 48740 }, { "grad_norm": 0.09948890656232834, "learning_rate": 5.618627233221344e-05, "loss": 0.0034, "step": 48750 }, { "grad_norm": 0.08056309819221497, "learning_rate": 5.616986434061438e-05, "loss": 0.0036, "step": 48760 }, { "grad_norm": 0.08291026949882507, "learning_rate": 5.615345567428822e-05, "loss": 0.004, "step": 48770 }, { "grad_norm": 0.12109571695327759, "learning_rate": 5.613704633502941e-05, "loss": 0.0041, "step": 48780 }, { "grad_norm": 0.07942087203264236, "learning_rate": 5.612063632463245e-05, "loss": 0.0043, "step": 48790 }, { "grad_norm": 0.06259773671627045, "learning_rate": 5.610422564489188e-05, "loss": 0.0042, "step": 48800 }, { "grad_norm": 0.07277899980545044, "learning_rate": 5.608781429760239e-05, "loss": 0.0036, "step": 48810 }, { "grad_norm": 0.05332743376493454, "learning_rate": 5.607140228455866e-05, "loss": 0.003, "step": 48820 }, { "grad_norm": 0.08380783349275589, "learning_rate": 5.605498960755553e-05, "loss": 0.0037, "step": 48830 }, { "grad_norm": 0.07186544686555862, "learning_rate": 5.603857626838782e-05, "loss": 0.0035, "step": 48840 }, { "grad_norm": 0.07082279771566391, "learning_rate": 5.60221622688505e-05, "loss": 0.0034, "step": 48850 }, { "grad_norm": 0.0832602009177208, "learning_rate": 5.6005747610738565e-05, "loss": 0.0028, "step": 48860 }, { "grad_norm": 0.08561019599437714, "learning_rate": 5.59893322958471e-05, "loss": 0.0027, "step": 48870 }, { "grad_norm": 0.07574943453073502, "learning_rate": 5.5972916325971256e-05, "loss": 0.0035, "step": 48880 }, { "grad_norm": 0.06765647977590561, "learning_rate": 5.595649970290628e-05, "loss": 0.0027, "step": 48890 }, { "grad_norm": 0.07195210456848145, "learning_rate": 5.5940082428447426e-05, "loss": 0.0026, "step": 48900 }, { "grad_norm": 0.06304165720939636, "learning_rate": 5.592366450439012e-05, "loss": 0.0027, "step": 48910 }, { "grad_norm": 0.07596246898174286, "learning_rate": 5.590724593252975e-05, "loss": 0.0036, "step": 48920 }, { "grad_norm": 0.06838509440422058, "learning_rate": 5.589082671466184e-05, "loss": 0.0043, "step": 48930 }, { "grad_norm": 0.07852939516305923, "learning_rate": 5.587440685258199e-05, "loss": 0.0028, "step": 48940 }, { "grad_norm": 0.09530319273471832, "learning_rate": 5.585798634808583e-05, "loss": 0.0028, "step": 48950 }, { "grad_norm": 0.07683438807725906, "learning_rate": 5.584156520296909e-05, "loss": 0.003, "step": 48960 }, { "grad_norm": 0.06634138524532318, "learning_rate": 5.582514341902757e-05, "loss": 0.0029, "step": 48970 }, { "grad_norm": 0.09103960543870926, "learning_rate": 5.580872099805713e-05, "loss": 0.0036, "step": 48980 }, { "grad_norm": 0.07899553328752518, "learning_rate": 5.5792297941853674e-05, "loss": 0.0034, "step": 48990 }, { "grad_norm": 0.06857962906360626, "learning_rate": 5.5775874252213247e-05, "loss": 0.0031, "step": 49000 }, { "grad_norm": 0.06927546858787537, "learning_rate": 5.575944993093189e-05, "loss": 0.0038, "step": 49010 }, { "grad_norm": 0.08637315034866333, "learning_rate": 5.574302497980574e-05, "loss": 0.0054, "step": 49020 }, { "grad_norm": 0.09841200709342957, "learning_rate": 5.5726599400631005e-05, "loss": 0.0035, "step": 49030 }, { "grad_norm": 0.08524400740861893, "learning_rate": 5.571017319520399e-05, "loss": 0.0034, "step": 49040 }, { "grad_norm": 0.09941041469573975, "learning_rate": 5.5693746365320984e-05, "loss": 0.0039, "step": 49050 }, { "grad_norm": 0.06990012526512146, "learning_rate": 5.567731891277848e-05, "loss": 0.0042, "step": 49060 }, { "grad_norm": 0.09878788888454437, "learning_rate": 5.566089083937288e-05, "loss": 0.0039, "step": 49070 }, { "grad_norm": 0.08270907402038574, "learning_rate": 5.564446214690079e-05, "loss": 0.0034, "step": 49080 }, { "grad_norm": 0.08552916347980499, "learning_rate": 5.56280328371588e-05, "loss": 0.0034, "step": 49090 }, { "grad_norm": 0.0735815241932869, "learning_rate": 5.5611602911943596e-05, "loss": 0.003, "step": 49100 }, { "grad_norm": 0.09335426241159439, "learning_rate": 5.559517237305194e-05, "loss": 0.0033, "step": 49110 }, { "grad_norm": 0.07397956401109695, "learning_rate": 5.557874122228064e-05, "loss": 0.003, "step": 49120 }, { "grad_norm": 0.06539388000965118, "learning_rate": 5.55623094614266e-05, "loss": 0.0029, "step": 49130 }, { "grad_norm": 0.06454309076070786, "learning_rate": 5.5545877092286743e-05, "loss": 0.0025, "step": 49140 }, { "grad_norm": 0.07149957865476608, "learning_rate": 5.552944411665811e-05, "loss": 0.0033, "step": 49150 }, { "grad_norm": 0.09707542508840561, "learning_rate": 5.5513010536337795e-05, "loss": 0.0032, "step": 49160 }, { "grad_norm": 0.08670364320278168, "learning_rate": 5.5496576353122934e-05, "loss": 0.0036, "step": 49170 }, { "grad_norm": 0.09428153932094574, "learning_rate": 5.548014156881074e-05, "loss": 0.0039, "step": 49180 }, { "grad_norm": 0.0889870896935463, "learning_rate": 5.5463706185198494e-05, "loss": 0.0031, "step": 49190 }, { "grad_norm": 0.07025988399982452, "learning_rate": 5.544727020408358e-05, "loss": 0.0043, "step": 49200 }, { "grad_norm": 0.07252366095781326, "learning_rate": 5.5430833627263366e-05, "loss": 0.0048, "step": 49210 }, { "grad_norm": 0.06615881621837616, "learning_rate": 5.5414396456535354e-05, "loss": 0.0025, "step": 49220 }, { "grad_norm": 0.051436059176921844, "learning_rate": 5.539795869369709e-05, "loss": 0.0029, "step": 49230 }, { "grad_norm": 0.10055489093065262, "learning_rate": 5.538152034054618e-05, "loss": 0.0036, "step": 49240 }, { "grad_norm": 0.06720937043428421, "learning_rate": 5.536508139888028e-05, "loss": 0.0034, "step": 49250 }, { "grad_norm": 0.07383395731449127, "learning_rate": 5.534864187049716e-05, "loss": 0.0031, "step": 49260 }, { "grad_norm": 0.08021394908428192, "learning_rate": 5.5332201757194615e-05, "loss": 0.0047, "step": 49270 }, { "grad_norm": 0.11950221657752991, "learning_rate": 5.531576106077049e-05, "loss": 0.0033, "step": 49280 }, { "grad_norm": 0.08322852104902267, "learning_rate": 5.529931978302272e-05, "loss": 0.0043, "step": 49290 }, { "grad_norm": 0.09370879083871841, "learning_rate": 5.528287792574932e-05, "loss": 0.0041, "step": 49300 }, { "grad_norm": 0.06051967665553093, "learning_rate": 5.5266435490748324e-05, "loss": 0.0027, "step": 49310 }, { "grad_norm": 0.07057762891054153, "learning_rate": 5.524999247981787e-05, "loss": 0.0028, "step": 49320 }, { "grad_norm": 0.07057575136423111, "learning_rate": 5.523354889475613e-05, "loss": 0.0033, "step": 49330 }, { "grad_norm": 0.0625339075922966, "learning_rate": 5.521710473736134e-05, "loss": 0.003, "step": 49340 }, { "grad_norm": 0.05620357394218445, "learning_rate": 5.5200660009431826e-05, "loss": 0.0027, "step": 49350 }, { "grad_norm": 0.11069219559431076, "learning_rate": 5.518421471276596e-05, "loss": 0.0037, "step": 49360 }, { "grad_norm": 0.07155990600585938, "learning_rate": 5.5167768849162174e-05, "loss": 0.0039, "step": 49370 }, { "grad_norm": 0.07195066660642624, "learning_rate": 5.515132242041893e-05, "loss": 0.0033, "step": 49380 }, { "grad_norm": 0.0701843723654747, "learning_rate": 5.513487542833483e-05, "loss": 0.0028, "step": 49390 }, { "grad_norm": 0.06499996781349182, "learning_rate": 5.5118427874708466e-05, "loss": 0.0031, "step": 49400 }, { "grad_norm": 0.09087783843278885, "learning_rate": 5.510197976133853e-05, "loss": 0.0039, "step": 49410 }, { "grad_norm": 0.08878985047340393, "learning_rate": 5.508553109002376e-05, "loss": 0.0034, "step": 49420 }, { "grad_norm": 0.08571244776248932, "learning_rate": 5.5069081862562957e-05, "loss": 0.0032, "step": 49430 }, { "grad_norm": 0.0801301971077919, "learning_rate": 5.5052632080754965e-05, "loss": 0.0034, "step": 49440 }, { "grad_norm": 0.08410409837961197, "learning_rate": 5.503618174639874e-05, "loss": 0.0034, "step": 49450 }, { "grad_norm": 0.08382611721754074, "learning_rate": 5.501973086129325e-05, "loss": 0.0035, "step": 49460 }, { "grad_norm": 0.13157042860984802, "learning_rate": 5.500327942723753e-05, "loss": 0.0038, "step": 49470 }, { "grad_norm": 0.07045148313045502, "learning_rate": 5.498682744603071e-05, "loss": 0.004, "step": 49480 }, { "grad_norm": 0.11737145483493805, "learning_rate": 5.497037491947194e-05, "loss": 0.0052, "step": 49490 }, { "grad_norm": 0.07665074616670609, "learning_rate": 5.4953921849360424e-05, "loss": 0.0034, "step": 49500 }, { "grad_norm": 0.08974684029817581, "learning_rate": 5.493746823749547e-05, "loss": 0.0042, "step": 49510 }, { "grad_norm": 0.07939028739929199, "learning_rate": 5.49210140856764e-05, "loss": 0.0051, "step": 49520 }, { "grad_norm": 0.07883036136627197, "learning_rate": 5.4904559395702647e-05, "loss": 0.0042, "step": 49530 }, { "grad_norm": 0.0829998105764389, "learning_rate": 5.488810416937364e-05, "loss": 0.0045, "step": 49540 }, { "grad_norm": 0.10141976922750473, "learning_rate": 5.4871648408488926e-05, "loss": 0.005, "step": 49550 }, { "grad_norm": 0.10850876569747925, "learning_rate": 5.485519211484807e-05, "loss": 0.0036, "step": 49560 }, { "grad_norm": 0.07268897444009781, "learning_rate": 5.4838735290250676e-05, "loss": 0.004, "step": 49570 }, { "grad_norm": 0.09668050706386566, "learning_rate": 5.482227793649649e-05, "loss": 0.0035, "step": 49580 }, { "grad_norm": 0.10825567692518234, "learning_rate": 5.480582005538524e-05, "loss": 0.0058, "step": 49590 }, { "grad_norm": 0.06021616980433464, "learning_rate": 5.478936164871671e-05, "loss": 0.0036, "step": 49600 }, { "grad_norm": 0.0750868171453476, "learning_rate": 5.477290271829081e-05, "loss": 0.0038, "step": 49610 }, { "grad_norm": 0.07545974105596542, "learning_rate": 5.475644326590744e-05, "loss": 0.003, "step": 49620 }, { "grad_norm": 0.07912004739046097, "learning_rate": 5.473998329336658e-05, "loss": 0.003, "step": 49630 }, { "grad_norm": 0.08825639635324478, "learning_rate": 5.4723522802468286e-05, "loss": 0.0043, "step": 49640 }, { "grad_norm": 0.08489182591438293, "learning_rate": 5.470706179501264e-05, "loss": 0.004, "step": 49650 }, { "grad_norm": 0.11328259110450745, "learning_rate": 5.4690600272799776e-05, "loss": 0.0034, "step": 49660 }, { "grad_norm": 0.09465458989143372, "learning_rate": 5.467413823762993e-05, "loss": 0.0031, "step": 49670 }, { "grad_norm": 0.06649202108383179, "learning_rate": 5.465767569130335e-05, "loss": 0.0031, "step": 49680 }, { "grad_norm": 0.0919322818517685, "learning_rate": 5.464121263562036e-05, "loss": 0.0036, "step": 49690 }, { "grad_norm": 0.10096749663352966, "learning_rate": 5.4624749072381343e-05, "loss": 0.0035, "step": 49700 }, { "grad_norm": 0.09024706482887268, "learning_rate": 5.460828500338672e-05, "loss": 0.0035, "step": 49710 }, { "grad_norm": 0.10998499393463135, "learning_rate": 5.459182043043698e-05, "loss": 0.0053, "step": 49720 }, { "grad_norm": 0.12413681298494339, "learning_rate": 5.457535535533265e-05, "loss": 0.0062, "step": 49730 }, { "grad_norm": 0.09105376154184341, "learning_rate": 5.4558889779874334e-05, "loss": 0.0039, "step": 49740 }, { "grad_norm": 0.0953880324959755, "learning_rate": 5.454242370586269e-05, "loss": 0.0061, "step": 49750 }, { "grad_norm": 0.08280759304761887, "learning_rate": 5.452595713509843e-05, "loss": 0.0039, "step": 49760 }, { "grad_norm": 0.07489580661058426, "learning_rate": 5.4509490069382274e-05, "loss": 0.0039, "step": 49770 }, { "grad_norm": 0.09243878722190857, "learning_rate": 5.449302251051509e-05, "loss": 0.004, "step": 49780 }, { "grad_norm": 0.05576569586992264, "learning_rate": 5.44765544602977e-05, "loss": 0.0031, "step": 49790 }, { "grad_norm": 0.07891155779361725, "learning_rate": 5.446008592053107e-05, "loss": 0.0046, "step": 49800 }, { "grad_norm": 0.08913829177618027, "learning_rate": 5.4443616893016135e-05, "loss": 0.0038, "step": 49810 }, { "grad_norm": 0.08780127763748169, "learning_rate": 5.4427147379553934e-05, "loss": 0.0041, "step": 49820 }, { "grad_norm": 0.08316973596811295, "learning_rate": 5.441067738194556e-05, "loss": 0.0048, "step": 49830 }, { "grad_norm": 0.06972505152225494, "learning_rate": 5.439420690199214e-05, "loss": 0.0034, "step": 49840 }, { "grad_norm": 0.08144547045230865, "learning_rate": 5.437773594149484e-05, "loss": 0.004, "step": 49850 }, { "grad_norm": 0.09008192270994186, "learning_rate": 5.436126450225495e-05, "loss": 0.0049, "step": 49860 }, { "grad_norm": 0.11948342621326447, "learning_rate": 5.434479258607371e-05, "loss": 0.0041, "step": 49870 }, { "grad_norm": 0.06407797336578369, "learning_rate": 5.43283201947525e-05, "loss": 0.0033, "step": 49880 }, { "grad_norm": 0.080284483730793, "learning_rate": 5.431184733009268e-05, "loss": 0.0033, "step": 49890 }, { "grad_norm": 0.07194071263074875, "learning_rate": 5.4295373993895736e-05, "loss": 0.0027, "step": 49900 }, { "grad_norm": 0.07485229521989822, "learning_rate": 5.4278900187963157e-05, "loss": 0.003, "step": 49910 }, { "grad_norm": 0.07562863081693649, "learning_rate": 5.426242591409646e-05, "loss": 0.0033, "step": 49920 }, { "grad_norm": 0.09798067808151245, "learning_rate": 5.42459511740973e-05, "loss": 0.0025, "step": 49930 }, { "grad_norm": 0.11198906600475311, "learning_rate": 5.422947596976729e-05, "loss": 0.0041, "step": 49940 }, { "grad_norm": 0.08406133204698563, "learning_rate": 5.4213000302908134e-05, "loss": 0.0038, "step": 49950 }, { "grad_norm": 0.05441479757428169, "learning_rate": 5.419652417532162e-05, "loss": 0.0028, "step": 49960 }, { "grad_norm": 0.07254557311534882, "learning_rate": 5.4180047588809534e-05, "loss": 0.003, "step": 49970 }, { "grad_norm": 0.05902407318353653, "learning_rate": 5.4163570545173704e-05, "loss": 0.0029, "step": 49980 }, { "grad_norm": 0.04987410083413124, "learning_rate": 5.414709304621608e-05, "loss": 0.0035, "step": 49990 }, { "grad_norm": 0.06058702990412712, "learning_rate": 5.413061509373858e-05, "loss": 0.0029, "step": 50000 }, { "grad_norm": 0.0758795291185379, "learning_rate": 5.411413668954324e-05, "loss": 0.0036, "step": 50010 }, { "grad_norm": 0.19270426034927368, "learning_rate": 5.4097657835432083e-05, "loss": 0.0034, "step": 50020 }, { "grad_norm": 0.057997994124889374, "learning_rate": 5.408117853320723e-05, "loss": 0.0025, "step": 50030 }, { "grad_norm": 0.07886508852243423, "learning_rate": 5.406469878467084e-05, "loss": 0.0036, "step": 50040 }, { "grad_norm": 0.09914764761924744, "learning_rate": 5.404821859162509e-05, "loss": 0.0033, "step": 50050 }, { "grad_norm": 0.16469107568264008, "learning_rate": 5.403173795587225e-05, "loss": 0.0043, "step": 50060 }, { "grad_norm": 0.07634548097848892, "learning_rate": 5.4015256879214606e-05, "loss": 0.0038, "step": 50070 }, { "grad_norm": 0.09133089333772659, "learning_rate": 5.39987753634545e-05, "loss": 0.004, "step": 50080 }, { "grad_norm": 0.09709157049655914, "learning_rate": 5.3982293410394335e-05, "loss": 0.0047, "step": 50090 }, { "grad_norm": 0.0855531170964241, "learning_rate": 5.396581102183655e-05, "loss": 0.005, "step": 50100 }, { "grad_norm": 0.10826057195663452, "learning_rate": 5.3949328199583634e-05, "loss": 0.0052, "step": 50110 }, { "grad_norm": 0.09460718184709549, "learning_rate": 5.393284494543813e-05, "loss": 0.0038, "step": 50120 }, { "grad_norm": 0.14131087064743042, "learning_rate": 5.391636126120262e-05, "loss": 0.0045, "step": 50130 }, { "grad_norm": 0.3097897171974182, "learning_rate": 5.389987714867971e-05, "loss": 0.0042, "step": 50140 }, { "grad_norm": 0.1352442502975464, "learning_rate": 5.3883392609672114e-05, "loss": 0.0055, "step": 50150 }, { "grad_norm": 0.09929666668176651, "learning_rate": 5.3866907645982545e-05, "loss": 0.0036, "step": 50160 }, { "grad_norm": 0.08764434605836868, "learning_rate": 5.385042225941375e-05, "loss": 0.0038, "step": 50170 }, { "grad_norm": 0.08584430068731308, "learning_rate": 5.3833936451768585e-05, "loss": 0.0039, "step": 50180 }, { "grad_norm": 0.07543960213661194, "learning_rate": 5.3817450224849884e-05, "loss": 0.0052, "step": 50190 }, { "grad_norm": 0.11392431706190109, "learning_rate": 5.380096358046056e-05, "loss": 0.0048, "step": 50200 }, { "grad_norm": 0.10522527247667313, "learning_rate": 5.378447652040359e-05, "loss": 0.0033, "step": 50210 }, { "grad_norm": 0.07573343068361282, "learning_rate": 5.376798904648195e-05, "loss": 0.0037, "step": 50220 }, { "grad_norm": 0.06277265399694443, "learning_rate": 5.375150116049869e-05, "loss": 0.0032, "step": 50230 }, { "grad_norm": 0.09032009541988373, "learning_rate": 5.373501286425691e-05, "loss": 0.0027, "step": 50240 }, { "grad_norm": 0.1316193789243698, "learning_rate": 5.3718524159559726e-05, "loss": 0.0056, "step": 50250 }, { "grad_norm": 0.0774427279829979, "learning_rate": 5.370203504821034e-05, "loss": 0.0042, "step": 50260 }, { "grad_norm": 0.08366815745830536, "learning_rate": 5.368554553201196e-05, "loss": 0.0044, "step": 50270 }, { "grad_norm": 0.10219860076904297, "learning_rate": 5.3669055612767874e-05, "loss": 0.004, "step": 50280 }, { "grad_norm": 0.1131516844034195, "learning_rate": 5.3652565292281376e-05, "loss": 0.0036, "step": 50290 }, { "grad_norm": 0.07692917436361313, "learning_rate": 5.3636074572355824e-05, "loss": 0.0038, "step": 50300 }, { "grad_norm": 0.09658560901880264, "learning_rate": 5.361958345479463e-05, "loss": 0.0038, "step": 50310 }, { "grad_norm": 0.06833642721176147, "learning_rate": 5.360309194140123e-05, "loss": 0.0043, "step": 50320 }, { "grad_norm": 0.08775712549686432, "learning_rate": 5.3586600033979106e-05, "loss": 0.004, "step": 50330 }, { "grad_norm": 0.10998331755399704, "learning_rate": 5.357010773433181e-05, "loss": 0.0044, "step": 50340 }, { "grad_norm": 0.08526743948459625, "learning_rate": 5.3553615044262895e-05, "loss": 0.0029, "step": 50350 }, { "grad_norm": 0.06737583130598068, "learning_rate": 5.3537121965575964e-05, "loss": 0.0029, "step": 50360 }, { "grad_norm": 0.07260514795780182, "learning_rate": 5.352062850007471e-05, "loss": 0.0037, "step": 50370 }, { "grad_norm": 0.09063711762428284, "learning_rate": 5.350413464956282e-05, "loss": 0.0039, "step": 50380 }, { "grad_norm": 0.09261628240346909, "learning_rate": 5.348764041584403e-05, "loss": 0.0036, "step": 50390 }, { "grad_norm": 0.08366412669420242, "learning_rate": 5.347114580072212e-05, "loss": 0.0047, "step": 50400 }, { "grad_norm": 0.130976140499115, "learning_rate": 5.345465080600093e-05, "loss": 0.0054, "step": 50410 }, { "grad_norm": 0.06510507315397263, "learning_rate": 5.3438155433484326e-05, "loss": 0.0039, "step": 50420 }, { "grad_norm": 0.07527697831392288, "learning_rate": 5.3421659684976197e-05, "loss": 0.0033, "step": 50430 }, { "grad_norm": 0.06559600681066513, "learning_rate": 5.340516356228052e-05, "loss": 0.0035, "step": 50440 }, { "grad_norm": 0.09376395493745804, "learning_rate": 5.338866706720128e-05, "loss": 0.0034, "step": 50450 }, { "grad_norm": 0.07481250911951065, "learning_rate": 5.337217020154249e-05, "loss": 0.0042, "step": 50460 }, { "grad_norm": 0.06695582717657089, "learning_rate": 5.335567296710825e-05, "loss": 0.0048, "step": 50470 }, { "grad_norm": 0.0727362185716629, "learning_rate": 5.333917536570265e-05, "loss": 0.0041, "step": 50480 }, { "grad_norm": 0.049110982567071915, "learning_rate": 5.332267739912986e-05, "loss": 0.0027, "step": 50490 }, { "grad_norm": 0.05766148865222931, "learning_rate": 5.330617906919405e-05, "loss": 0.0026, "step": 50500 }, { "grad_norm": 0.053811538964509964, "learning_rate": 5.328968037769949e-05, "loss": 0.0035, "step": 50510 }, { "grad_norm": 0.052262529730796814, "learning_rate": 5.327318132645042e-05, "loss": 0.0027, "step": 50520 }, { "grad_norm": 0.06364022195339203, "learning_rate": 5.325668191725116e-05, "loss": 0.0029, "step": 50530 }, { "grad_norm": 0.06306412816047668, "learning_rate": 5.324018215190606e-05, "loss": 0.0031, "step": 50540 }, { "grad_norm": 0.08122286200523376, "learning_rate": 5.3223682032219515e-05, "loss": 0.0033, "step": 50550 }, { "grad_norm": 0.10302852839231491, "learning_rate": 5.320718155999595e-05, "loss": 0.0033, "step": 50560 }, { "grad_norm": 0.07085723429918289, "learning_rate": 5.3190680737039835e-05, "loss": 0.0031, "step": 50570 }, { "grad_norm": 0.07180625200271606, "learning_rate": 5.317417956515567e-05, "loss": 0.0032, "step": 50580 }, { "grad_norm": 0.08905138075351715, "learning_rate": 5.315767804614798e-05, "loss": 0.003, "step": 50590 }, { "grad_norm": 0.09010287374258041, "learning_rate": 5.3141176181821395e-05, "loss": 0.0035, "step": 50600 }, { "grad_norm": 0.12752234935760498, "learning_rate": 5.31246739739805e-05, "loss": 0.0043, "step": 50610 }, { "grad_norm": 0.08345440030097961, "learning_rate": 5.310817142442994e-05, "loss": 0.0037, "step": 50620 }, { "grad_norm": 0.07962110638618469, "learning_rate": 5.309166853497445e-05, "loss": 0.0032, "step": 50630 }, { "grad_norm": 0.07661158591508865, "learning_rate": 5.307516530741873e-05, "loss": 0.0033, "step": 50640 }, { "grad_norm": 0.07973705977201462, "learning_rate": 5.305866174356754e-05, "loss": 0.004, "step": 50650 }, { "grad_norm": 0.057360079139471054, "learning_rate": 5.304215784522571e-05, "loss": 0.0023, "step": 50660 }, { "grad_norm": 0.05534398928284645, "learning_rate": 5.302565361419808e-05, "loss": 0.0027, "step": 50670 }, { "grad_norm": 0.07477809488773346, "learning_rate": 5.3009149052289507e-05, "loss": 0.0031, "step": 50680 }, { "grad_norm": 0.07014487683773041, "learning_rate": 5.299264416130493e-05, "loss": 0.0023, "step": 50690 }, { "grad_norm": 0.07602822035551071, "learning_rate": 5.297613894304928e-05, "loss": 0.0028, "step": 50700 }, { "grad_norm": 0.077012799680233, "learning_rate": 5.2959633399327534e-05, "loss": 0.0029, "step": 50710 }, { "grad_norm": 0.06358057260513306, "learning_rate": 5.294312753194476e-05, "loss": 0.0037, "step": 50720 }, { "grad_norm": 0.45843949913978577, "learning_rate": 5.292662134270596e-05, "loss": 0.0028, "step": 50730 }, { "grad_norm": 0.10400725156068802, "learning_rate": 5.291011483341626e-05, "loss": 0.0036, "step": 50740 }, { "grad_norm": 0.06936703622341156, "learning_rate": 5.2893608005880767e-05, "loss": 0.0043, "step": 50750 }, { "grad_norm": 0.0720098689198494, "learning_rate": 5.287710086190467e-05, "loss": 0.0043, "step": 50760 }, { "grad_norm": 0.0630376935005188, "learning_rate": 5.286059340329314e-05, "loss": 0.0036, "step": 50770 }, { "grad_norm": 0.07549319416284561, "learning_rate": 5.284408563185141e-05, "loss": 0.0045, "step": 50780 }, { "grad_norm": 0.07856278121471405, "learning_rate": 5.282757754938476e-05, "loss": 0.0034, "step": 50790 }, { "grad_norm": 0.09289611130952835, "learning_rate": 5.281106915769849e-05, "loss": 0.0035, "step": 50800 }, { "grad_norm": 0.08401118218898773, "learning_rate": 5.2794560458597897e-05, "loss": 0.0031, "step": 50810 }, { "grad_norm": 0.09539610147476196, "learning_rate": 5.27780514538884e-05, "loss": 0.0047, "step": 50820 }, { "grad_norm": 0.07564505934715271, "learning_rate": 5.2761542145375365e-05, "loss": 0.0035, "step": 50830 }, { "grad_norm": 0.11031465232372284, "learning_rate": 5.274503253486421e-05, "loss": 0.0035, "step": 50840 }, { "grad_norm": 0.04887663945555687, "learning_rate": 5.272852262416046e-05, "loss": 0.003, "step": 50850 }, { "grad_norm": 0.0692378431558609, "learning_rate": 5.2712012415069555e-05, "loss": 0.004, "step": 50860 }, { "grad_norm": 0.0585598424077034, "learning_rate": 5.269550190939705e-05, "loss": 0.0031, "step": 50870 }, { "grad_norm": 0.09513884037733078, "learning_rate": 5.267899110894852e-05, "loss": 0.0037, "step": 50880 }, { "grad_norm": 0.06788837164640427, "learning_rate": 5.266248001552955e-05, "loss": 0.0043, "step": 50890 }, { "grad_norm": 0.08557208627462387, "learning_rate": 5.264596863094575e-05, "loss": 0.0028, "step": 50900 }, { "grad_norm": 0.09008323401212692, "learning_rate": 5.262945695700282e-05, "loss": 0.0031, "step": 50910 }, { "grad_norm": 0.11305879056453705, "learning_rate": 5.261294499550643e-05, "loss": 0.0045, "step": 50920 }, { "grad_norm": 0.08770125359296799, "learning_rate": 5.25964327482623e-05, "loss": 0.0044, "step": 50930 }, { "grad_norm": 0.09400758892297745, "learning_rate": 5.257992021707617e-05, "loss": 0.004, "step": 50940 }, { "grad_norm": 0.08967126160860062, "learning_rate": 5.256340740375387e-05, "loss": 0.0038, "step": 50950 }, { "grad_norm": 0.05802978202700615, "learning_rate": 5.254689431010117e-05, "loss": 0.0032, "step": 50960 }, { "grad_norm": 0.08752787858247757, "learning_rate": 5.253038093792395e-05, "loss": 0.0028, "step": 50970 }, { "grad_norm": 0.07657483965158463, "learning_rate": 5.251386728902806e-05, "loss": 0.0043, "step": 50980 }, { "grad_norm": 0.11449577659368515, "learning_rate": 5.2497353365219446e-05, "loss": 0.0045, "step": 50990 }, { "grad_norm": 0.08548170328140259, "learning_rate": 5.2480839168304e-05, "loss": 0.0044, "step": 51000 }, { "grad_norm": 0.06438098102807999, "learning_rate": 5.246432470008773e-05, "loss": 0.0027, "step": 51010 }, { "grad_norm": 0.06888292729854584, "learning_rate": 5.2447809962376616e-05, "loss": 0.0029, "step": 51020 }, { "grad_norm": 0.09775900095701218, "learning_rate": 5.243129495697666e-05, "loss": 0.0035, "step": 51030 }, { "grad_norm": 0.08891928941011429, "learning_rate": 5.241477968569397e-05, "loss": 0.0032, "step": 51040 }, { "grad_norm": 0.17184144258499146, "learning_rate": 5.23982641503346e-05, "loss": 0.0035, "step": 51050 }, { "grad_norm": 0.09620114415884018, "learning_rate": 5.238174835270465e-05, "loss": 0.0036, "step": 51060 }, { "grad_norm": 0.05770092085003853, "learning_rate": 5.23652322946103e-05, "loss": 0.0026, "step": 51070 }, { "grad_norm": 0.06585173308849335, "learning_rate": 5.234871597785771e-05, "loss": 0.0029, "step": 51080 }, { "grad_norm": 0.06265328824520111, "learning_rate": 5.233219940425306e-05, "loss": 0.0024, "step": 51090 }, { "grad_norm": 0.08119308948516846, "learning_rate": 5.2315682575602576e-05, "loss": 0.0033, "step": 51100 }, { "grad_norm": 0.08643331378698349, "learning_rate": 5.229916549371254e-05, "loss": 0.003, "step": 51110 }, { "grad_norm": 0.07635696232318878, "learning_rate": 5.228264816038924e-05, "loss": 0.0027, "step": 51120 }, { "grad_norm": 0.06623110920190811, "learning_rate": 5.2266130577438935e-05, "loss": 0.0031, "step": 51130 }, { "grad_norm": 0.04251008480787277, "learning_rate": 5.224961274666802e-05, "loss": 0.0026, "step": 51140 }, { "grad_norm": 0.07122895866632462, "learning_rate": 5.223309466988283e-05, "loss": 0.0043, "step": 51150 }, { "grad_norm": 0.06917261332273483, "learning_rate": 5.221657634888976e-05, "loss": 0.0041, "step": 51160 }, { "grad_norm": 0.10395494848489761, "learning_rate": 5.2200057785495235e-05, "loss": 0.0048, "step": 51170 }, { "grad_norm": 0.08947083353996277, "learning_rate": 5.2183538981505706e-05, "loss": 0.0037, "step": 51180 }, { "grad_norm": 0.06300738453865051, "learning_rate": 5.216701993872762e-05, "loss": 0.0037, "step": 51190 }, { "grad_norm": 0.12769579887390137, "learning_rate": 5.215050065896752e-05, "loss": 0.0027, "step": 51200 }, { "grad_norm": 0.09969175606966019, "learning_rate": 5.2133981144031875e-05, "loss": 0.0033, "step": 51210 }, { "grad_norm": 0.08611899614334106, "learning_rate": 5.211746139572727e-05, "loss": 0.0039, "step": 51220 }, { "grad_norm": 0.108866386115551, "learning_rate": 5.210094141586026e-05, "loss": 0.0039, "step": 51230 }, { "grad_norm": 0.11427655071020126, "learning_rate": 5.2084421206237476e-05, "loss": 0.0041, "step": 51240 }, { "grad_norm": 0.07842601835727692, "learning_rate": 5.20679007686655e-05, "loss": 0.0037, "step": 51250 }, { "grad_norm": 0.07583863288164139, "learning_rate": 5.205138010495102e-05, "loss": 0.0046, "step": 51260 }, { "grad_norm": 0.0853390023112297, "learning_rate": 5.20348592169007e-05, "loss": 0.0031, "step": 51270 }, { "grad_norm": 0.08143056929111481, "learning_rate": 5.201833810632123e-05, "loss": 0.0027, "step": 51280 }, { "grad_norm": 0.05315444618463516, "learning_rate": 5.200181677501933e-05, "loss": 0.0026, "step": 51290 }, { "grad_norm": 0.07732497155666351, "learning_rate": 5.1985295224801777e-05, "loss": 0.0031, "step": 51300 }, { "grad_norm": 0.0560186542570591, "learning_rate": 5.196877345747531e-05, "loss": 0.0028, "step": 51310 }, { "grad_norm": 0.06326545774936676, "learning_rate": 5.195225147484674e-05, "loss": 0.0021, "step": 51320 }, { "grad_norm": 0.06600672006607056, "learning_rate": 5.193572927872288e-05, "loss": 0.0027, "step": 51330 }, { "grad_norm": 0.06658772379159927, "learning_rate": 5.191920687091059e-05, "loss": 0.0023, "step": 51340 }, { "grad_norm": 0.0631287693977356, "learning_rate": 5.1902684253216695e-05, "loss": 0.0022, "step": 51350 }, { "grad_norm": 0.05594281852245331, "learning_rate": 5.1886161427448134e-05, "loss": 0.0028, "step": 51360 }, { "grad_norm": 0.0541783943772316, "learning_rate": 5.186963839541179e-05, "loss": 0.0036, "step": 51370 }, { "grad_norm": 0.09544892609119415, "learning_rate": 5.185311515891459e-05, "loss": 0.0041, "step": 51380 }, { "grad_norm": 0.08498341590166092, "learning_rate": 5.1836591719763516e-05, "loss": 0.0028, "step": 51390 }, { "grad_norm": 0.045691344887018204, "learning_rate": 5.182006807976554e-05, "loss": 0.0024, "step": 51400 }, { "grad_norm": 0.06250789761543274, "learning_rate": 5.180354424072763e-05, "loss": 0.0027, "step": 51410 }, { "grad_norm": 0.06328622251749039, "learning_rate": 5.178702020445686e-05, "loss": 0.0028, "step": 51420 }, { "grad_norm": 0.10903923213481903, "learning_rate": 5.177049597276024e-05, "loss": 0.0037, "step": 51430 }, { "grad_norm": 0.13983143866062164, "learning_rate": 5.175397154744482e-05, "loss": 0.0037, "step": 51440 }, { "grad_norm": 0.09091688692569733, "learning_rate": 5.1737446930317736e-05, "loss": 0.0044, "step": 51450 }, { "grad_norm": 0.0786956325173378, "learning_rate": 5.172092212318604e-05, "loss": 0.0033, "step": 51460 }, { "grad_norm": 0.08447922766208649, "learning_rate": 5.170439712785692e-05, "loss": 0.0036, "step": 51470 }, { "grad_norm": 0.12614527344703674, "learning_rate": 5.1687871946137467e-05, "loss": 0.0045, "step": 51480 }, { "grad_norm": 0.08452905714511871, "learning_rate": 5.16713465798349e-05, "loss": 0.0038, "step": 51490 }, { "grad_norm": 0.16418571770191193, "learning_rate": 5.165482103075637e-05, "loss": 0.0048, "step": 51500 }, { "grad_norm": 0.09559009224176407, "learning_rate": 5.163829530070909e-05, "loss": 0.0035, "step": 51510 }, { "grad_norm": 0.09960806369781494, "learning_rate": 5.162176939150032e-05, "loss": 0.0027, "step": 51520 }, { "grad_norm": 0.06110407039523125, "learning_rate": 5.160524330493728e-05, "loss": 0.0031, "step": 51530 }, { "grad_norm": 0.07185009866952896, "learning_rate": 5.158871704282724e-05, "loss": 0.0048, "step": 51540 }, { "grad_norm": 0.07385186105966568, "learning_rate": 5.15721906069775e-05, "loss": 0.0027, "step": 51550 }, { "grad_norm": 0.0904737040400505, "learning_rate": 5.155566399919536e-05, "loss": 0.0031, "step": 51560 }, { "grad_norm": 0.0697886198759079, "learning_rate": 5.153913722128813e-05, "loss": 0.0033, "step": 51570 }, { "grad_norm": 0.1129012256860733, "learning_rate": 5.1522610275063176e-05, "loss": 0.0037, "step": 51580 }, { "grad_norm": 0.07321067899465561, "learning_rate": 5.150608316232786e-05, "loss": 0.0029, "step": 51590 }, { "grad_norm": 0.07110471278429031, "learning_rate": 5.148955588488955e-05, "loss": 0.0032, "step": 51600 }, { "grad_norm": 0.08476985991001129, "learning_rate": 5.147302844455564e-05, "loss": 0.0043, "step": 51610 }, { "grad_norm": 0.0852542370557785, "learning_rate": 5.145650084313356e-05, "loss": 0.0033, "step": 51620 }, { "grad_norm": 0.06358764320611954, "learning_rate": 5.1439973082430734e-05, "loss": 0.0036, "step": 51630 }, { "grad_norm": 0.09895424544811249, "learning_rate": 5.142344516425461e-05, "loss": 0.0033, "step": 51640 }, { "grad_norm": 0.07646879553794861, "learning_rate": 5.1406917090412673e-05, "loss": 0.0036, "step": 51650 }, { "grad_norm": 0.07373416423797607, "learning_rate": 5.1390388862712404e-05, "loss": 0.0031, "step": 51660 }, { "grad_norm": 0.10129114985466003, "learning_rate": 5.137386048296129e-05, "loss": 0.0029, "step": 51670 }, { "grad_norm": 0.0792209729552269, "learning_rate": 5.135733195296686e-05, "loss": 0.0028, "step": 51680 }, { "grad_norm": 0.08104248344898224, "learning_rate": 5.134080327453664e-05, "loss": 0.0029, "step": 51690 }, { "grad_norm": 0.07021685689687729, "learning_rate": 5.1324274449478204e-05, "loss": 0.0029, "step": 51700 }, { "grad_norm": 0.06013081595301628, "learning_rate": 5.130774547959909e-05, "loss": 0.0032, "step": 51710 }, { "grad_norm": 0.08661625534296036, "learning_rate": 5.129121636670692e-05, "loss": 0.0039, "step": 51720 }, { "grad_norm": 0.05492696166038513, "learning_rate": 5.1274687112609246e-05, "loss": 0.0029, "step": 51730 }, { "grad_norm": 0.05605752393603325, "learning_rate": 5.125815771911372e-05, "loss": 0.0029, "step": 51740 }, { "grad_norm": 0.08274860680103302, "learning_rate": 5.1241628188027956e-05, "loss": 0.0032, "step": 51750 }, { "grad_norm": 0.06216210871934891, "learning_rate": 5.1225098521159585e-05, "loss": 0.0026, "step": 51760 }, { "grad_norm": 0.05936560407280922, "learning_rate": 5.12085687203163e-05, "loss": 0.0036, "step": 51770 }, { "grad_norm": 0.06400670856237411, "learning_rate": 5.119203878730575e-05, "loss": 0.0027, "step": 51780 }, { "grad_norm": 0.0862298533320427, "learning_rate": 5.117550872393564e-05, "loss": 0.004, "step": 51790 }, { "grad_norm": 0.07171246409416199, "learning_rate": 5.1158978532013645e-05, "loss": 0.0026, "step": 51800 }, { "grad_norm": 0.0917634665966034, "learning_rate": 5.114244821334751e-05, "loss": 0.0031, "step": 51810 }, { "grad_norm": 0.08472491055727005, "learning_rate": 5.112591776974497e-05, "loss": 0.0033, "step": 51820 }, { "grad_norm": 0.12778516113758087, "learning_rate": 5.110938720301374e-05, "loss": 0.0052, "step": 51830 }, { "grad_norm": 0.10228238999843597, "learning_rate": 5.1092856514961595e-05, "loss": 0.0036, "step": 51840 }, { "grad_norm": 0.13653038442134857, "learning_rate": 5.107632570739631e-05, "loss": 0.0037, "step": 51850 }, { "grad_norm": 0.07121738791465759, "learning_rate": 5.1059794782125645e-05, "loss": 0.0036, "step": 51860 }, { "grad_norm": 0.06719717383384705, "learning_rate": 5.104326374095744e-05, "loss": 0.0041, "step": 51870 }, { "grad_norm": 0.08658257126808167, "learning_rate": 5.102673258569947e-05, "loss": 0.0031, "step": 51880 }, { "grad_norm": 0.12580199539661407, "learning_rate": 5.101020131815956e-05, "loss": 0.0028, "step": 51890 }, { "grad_norm": 0.058537520468235016, "learning_rate": 5.099366994014556e-05, "loss": 0.0026, "step": 51900 }, { "grad_norm": 0.07026577740907669, "learning_rate": 5.097713845346532e-05, "loss": 0.0026, "step": 51910 }, { "grad_norm": 0.07900775969028473, "learning_rate": 5.096060685992668e-05, "loss": 0.0027, "step": 51920 }, { "grad_norm": 0.05949823558330536, "learning_rate": 5.0944075161337514e-05, "loss": 0.0029, "step": 51930 }, { "grad_norm": 0.07020438462495804, "learning_rate": 5.09275433595057e-05, "loss": 0.0029, "step": 51940 }, { "grad_norm": 0.29270973801612854, "learning_rate": 5.0911011456239157e-05, "loss": 0.0026, "step": 51950 }, { "grad_norm": 0.06286739557981491, "learning_rate": 5.089447945334574e-05, "loss": 0.0038, "step": 51960 }, { "grad_norm": 0.06730588525533676, "learning_rate": 5.087794735263343e-05, "loss": 0.0033, "step": 51970 }, { "grad_norm": 0.0629471093416214, "learning_rate": 5.086141515591011e-05, "loss": 0.0042, "step": 51980 }, { "grad_norm": 0.07727865874767303, "learning_rate": 5.084488286498371e-05, "loss": 0.0031, "step": 51990 }, { "grad_norm": 0.09959389269351959, "learning_rate": 5.08283504816622e-05, "loss": 0.0047, "step": 52000 }, { "grad_norm": 0.09867830574512482, "learning_rate": 5.081181800775353e-05, "loss": 0.004, "step": 52010 }, { "grad_norm": 0.07508641481399536, "learning_rate": 5.079528544506565e-05, "loss": 0.0029, "step": 52020 }, { "grad_norm": 0.08544453233480453, "learning_rate": 5.077875279540657e-05, "loss": 0.0041, "step": 52030 }, { "grad_norm": 0.08405979722738266, "learning_rate": 5.076222006058424e-05, "loss": 0.0049, "step": 52040 }, { "grad_norm": 0.1751604825258255, "learning_rate": 5.074568724240667e-05, "loss": 0.0032, "step": 52050 }, { "grad_norm": 0.08169400691986084, "learning_rate": 5.072915434268187e-05, "loss": 0.0036, "step": 52060 }, { "grad_norm": 0.22909466922283173, "learning_rate": 5.071262136321785e-05, "loss": 0.0037, "step": 52070 }, { "grad_norm": 0.15982015430927277, "learning_rate": 5.069608830582262e-05, "loss": 0.0026, "step": 52080 }, { "grad_norm": 0.0805196538567543, "learning_rate": 5.067955517230423e-05, "loss": 0.0042, "step": 52090 }, { "grad_norm": 0.06507791578769684, "learning_rate": 5.06630219644707e-05, "loss": 0.0026, "step": 52100 }, { "grad_norm": 0.08724188804626465, "learning_rate": 5.064648868413008e-05, "loss": 0.0027, "step": 52110 }, { "grad_norm": 0.05753135308623314, "learning_rate": 5.062995533309044e-05, "loss": 0.003, "step": 52120 }, { "grad_norm": 0.0773613452911377, "learning_rate": 5.061342191315983e-05, "loss": 0.0033, "step": 52130 }, { "grad_norm": 0.08474941551685333, "learning_rate": 5.0596888426146325e-05, "loss": 0.0027, "step": 52140 }, { "grad_norm": 0.07244031131267548, "learning_rate": 5.058035487385798e-05, "loss": 0.0028, "step": 52150 }, { "grad_norm": 0.04801322519779205, "learning_rate": 5.0563821258102915e-05, "loss": 0.0022, "step": 52160 }, { "grad_norm": 0.06064548343420029, "learning_rate": 5.05472875806892e-05, "loss": 0.0028, "step": 52170 }, { "grad_norm": 0.0949726402759552, "learning_rate": 5.053075384342494e-05, "loss": 0.0031, "step": 52180 }, { "grad_norm": 0.076385997235775, "learning_rate": 5.0514220048118234e-05, "loss": 0.0024, "step": 52190 }, { "grad_norm": 0.05517774820327759, "learning_rate": 5.049768619657721e-05, "loss": 0.0037, "step": 52200 }, { "grad_norm": 0.08717665821313858, "learning_rate": 5.048115229060996e-05, "loss": 0.003, "step": 52210 }, { "grad_norm": 0.0663432627916336, "learning_rate": 5.046461833202462e-05, "loss": 0.0035, "step": 52220 }, { "grad_norm": 0.12229431420564651, "learning_rate": 5.044808432262932e-05, "loss": 0.0038, "step": 52230 }, { "grad_norm": 0.07116436213254929, "learning_rate": 5.043155026423219e-05, "loss": 0.0036, "step": 52240 }, { "grad_norm": 0.055470436811447144, "learning_rate": 5.041501615864137e-05, "loss": 0.0025, "step": 52250 }, { "grad_norm": 0.0693853422999382, "learning_rate": 5.039848200766502e-05, "loss": 0.0035, "step": 52260 }, { "grad_norm": 0.06932435929775238, "learning_rate": 5.038194781311125e-05, "loss": 0.0029, "step": 52270 }, { "grad_norm": 0.0815742239356041, "learning_rate": 5.0365413576788264e-05, "loss": 0.0026, "step": 52280 }, { "grad_norm": 0.09321536868810654, "learning_rate": 5.0348879300504194e-05, "loss": 0.0034, "step": 52290 }, { "grad_norm": 0.05462487041950226, "learning_rate": 5.033234498606721e-05, "loss": 0.0034, "step": 52300 }, { "grad_norm": 0.0568239800632, "learning_rate": 5.0315810635285476e-05, "loss": 0.0022, "step": 52310 }, { "grad_norm": 0.07414178550243378, "learning_rate": 5.0299276249967164e-05, "loss": 0.003, "step": 52320 }, { "grad_norm": 0.09098265320062637, "learning_rate": 5.0282741831920454e-05, "loss": 0.0033, "step": 52330 }, { "grad_norm": 0.06853790581226349, "learning_rate": 5.026620738295351e-05, "loss": 0.0025, "step": 52340 }, { "grad_norm": 0.07480665296316147, "learning_rate": 5.024967290487455e-05, "loss": 0.0033, "step": 52350 }, { "grad_norm": 0.06486759334802628, "learning_rate": 5.023313839949172e-05, "loss": 0.0025, "step": 52360 }, { "grad_norm": 0.0709741935133934, "learning_rate": 5.0216603868613224e-05, "loss": 0.0033, "step": 52370 }, { "grad_norm": 0.06299114972352982, "learning_rate": 5.020006931404726e-05, "loss": 0.0032, "step": 52380 }, { "grad_norm": 0.07005372643470764, "learning_rate": 5.0183534737602014e-05, "loss": 0.0036, "step": 52390 }, { "grad_norm": 0.05671985447406769, "learning_rate": 5.0167000141085686e-05, "loss": 0.0036, "step": 52400 }, { "grad_norm": 0.07553752511739731, "learning_rate": 5.015046552630647e-05, "loss": 0.003, "step": 52410 }, { "grad_norm": 0.06666941940784454, "learning_rate": 5.013393089507258e-05, "loss": 0.0033, "step": 52420 }, { "grad_norm": 0.06807969510555267, "learning_rate": 5.011739624919219e-05, "loss": 0.0032, "step": 52430 }, { "grad_norm": 0.06376001983880997, "learning_rate": 5.0100861590473545e-05, "loss": 0.0031, "step": 52440 }, { "grad_norm": 0.08064531534910202, "learning_rate": 5.008432692072482e-05, "loss": 0.0037, "step": 52450 }, { "grad_norm": 0.08170661330223083, "learning_rate": 5.006779224175423e-05, "loss": 0.0033, "step": 52460 }, { "grad_norm": 0.0697636753320694, "learning_rate": 5.005125755536998e-05, "loss": 0.0034, "step": 52470 }, { "grad_norm": 0.09583116322755814, "learning_rate": 5.003472286338029e-05, "loss": 0.0029, "step": 52480 }, { "grad_norm": 0.06822584569454193, "learning_rate": 5.001818816759335e-05, "loss": 0.003, "step": 52490 }, { "grad_norm": 0.09716296195983887, "learning_rate": 5.000165346981738e-05, "loss": 0.0031, "step": 52500 }, { "grad_norm": 0.08225242048501968, "learning_rate": 4.9985118771860595e-05, "loss": 0.0024, "step": 52510 }, { "grad_norm": 0.08250170201063156, "learning_rate": 4.996858407553119e-05, "loss": 0.0033, "step": 52520 }, { "grad_norm": 0.059553805738687515, "learning_rate": 4.99520493826374e-05, "loss": 0.0029, "step": 52530 }, { "grad_norm": 0.06881263852119446, "learning_rate": 4.9935514694987384e-05, "loss": 0.0028, "step": 52540 }, { "grad_norm": 0.05665326863527298, "learning_rate": 4.991898001438941e-05, "loss": 0.002, "step": 52550 }, { "grad_norm": 0.04428067058324814, "learning_rate": 4.990244534265166e-05, "loss": 0.0027, "step": 52560 }, { "grad_norm": 0.07634680718183517, "learning_rate": 4.988591068158234e-05, "loss": 0.0031, "step": 52570 }, { "grad_norm": 0.06515146791934967, "learning_rate": 4.986937603298965e-05, "loss": 0.0031, "step": 52580 }, { "grad_norm": 0.0846315547823906, "learning_rate": 4.985284139868181e-05, "loss": 0.0031, "step": 52590 }, { "grad_norm": 0.05612417310476303, "learning_rate": 4.9836306780467e-05, "loss": 0.0028, "step": 52600 }, { "grad_norm": 0.06954672187566757, "learning_rate": 4.981977218015343e-05, "loss": 0.0028, "step": 52610 }, { "grad_norm": 0.05436793342232704, "learning_rate": 4.980323759954932e-05, "loss": 0.0034, "step": 52620 }, { "grad_norm": 0.09494209289550781, "learning_rate": 4.978670304046284e-05, "loss": 0.003, "step": 52630 }, { "grad_norm": 0.0807601660490036, "learning_rate": 4.9770168504702195e-05, "loss": 0.0037, "step": 52640 }, { "grad_norm": 0.08556270599365234, "learning_rate": 4.975363399407559e-05, "loss": 0.0031, "step": 52650 }, { "grad_norm": 0.08932390064001083, "learning_rate": 4.9737099510391194e-05, "loss": 0.0035, "step": 52660 }, { "grad_norm": 0.07665979862213135, "learning_rate": 4.972056505545719e-05, "loss": 0.0036, "step": 52670 }, { "grad_norm": 0.061617642641067505, "learning_rate": 4.9704030631081775e-05, "loss": 0.0053, "step": 52680 }, { "grad_norm": 0.06543080508708954, "learning_rate": 4.968749623907313e-05, "loss": 0.0033, "step": 52690 }, { "grad_norm": 0.06608076393604279, "learning_rate": 4.967096188123943e-05, "loss": 0.0035, "step": 52700 }, { "grad_norm": 0.09445137530565262, "learning_rate": 4.965442755938884e-05, "loss": 0.0033, "step": 52710 }, { "grad_norm": 0.09915712475776672, "learning_rate": 4.963789327532954e-05, "loss": 0.0046, "step": 52720 }, { "grad_norm": 0.0521569587290287, "learning_rate": 4.962135903086968e-05, "loss": 0.0028, "step": 52730 }, { "grad_norm": 0.06172827631235123, "learning_rate": 4.960482482781741e-05, "loss": 0.0029, "step": 52740 }, { "grad_norm": 0.12069524824619293, "learning_rate": 4.958829066798092e-05, "loss": 0.003, "step": 52750 }, { "grad_norm": 0.25881698727607727, "learning_rate": 4.957175655316835e-05, "loss": 0.0036, "step": 52760 }, { "grad_norm": 0.1014808714389801, "learning_rate": 4.9555222485187834e-05, "loss": 0.0037, "step": 52770 }, { "grad_norm": 0.09346790611743927, "learning_rate": 4.953868846584752e-05, "loss": 0.0037, "step": 52780 }, { "grad_norm": 0.067520871758461, "learning_rate": 4.9522154496955544e-05, "loss": 0.0027, "step": 52790 }, { "grad_norm": 0.12236765027046204, "learning_rate": 4.950562058032001e-05, "loss": 0.0034, "step": 52800 }, { "grad_norm": 0.06897223740816116, "learning_rate": 4.948908671774909e-05, "loss": 0.0033, "step": 52810 }, { "grad_norm": 0.2261897623538971, "learning_rate": 4.947255291105087e-05, "loss": 0.004, "step": 52820 }, { "grad_norm": 0.0789112076163292, "learning_rate": 4.945601916203348e-05, "loss": 0.0026, "step": 52830 }, { "grad_norm": 0.08735651522874832, "learning_rate": 4.9439485472505016e-05, "loss": 0.0033, "step": 52840 }, { "grad_norm": 0.11232338100671768, "learning_rate": 4.942295184427356e-05, "loss": 0.0031, "step": 52850 }, { "grad_norm": 0.09081760048866272, "learning_rate": 4.9406418279147234e-05, "loss": 0.0028, "step": 52860 }, { "grad_norm": 0.13081452250480652, "learning_rate": 4.9389884778934116e-05, "loss": 0.0033, "step": 52870 }, { "grad_norm": 0.07046466320753098, "learning_rate": 4.937335134544228e-05, "loss": 0.0029, "step": 52880 }, { "grad_norm": 0.076787568628788, "learning_rate": 4.935681798047979e-05, "loss": 0.0029, "step": 52890 }, { "grad_norm": 0.09190913289785385, "learning_rate": 4.934028468585473e-05, "loss": 0.0045, "step": 52900 }, { "grad_norm": 0.10099121183156967, "learning_rate": 4.932375146337514e-05, "loss": 0.0028, "step": 52910 }, { "grad_norm": 0.07269048690795898, "learning_rate": 4.930721831484908e-05, "loss": 0.0034, "step": 52920 }, { "grad_norm": 0.05745505914092064, "learning_rate": 4.929068524208456e-05, "loss": 0.0032, "step": 52930 }, { "grad_norm": 0.10454456508159637, "learning_rate": 4.927415224688965e-05, "loss": 0.0048, "step": 52940 }, { "grad_norm": 0.11802806705236435, "learning_rate": 4.9257619331072375e-05, "loss": 0.0032, "step": 52950 }, { "grad_norm": 0.05151665210723877, "learning_rate": 4.924108649644072e-05, "loss": 0.0027, "step": 52960 }, { "grad_norm": 0.0828666016459465, "learning_rate": 4.922455374480271e-05, "loss": 0.0036, "step": 52970 }, { "grad_norm": 0.07419668138027191, "learning_rate": 4.920802107796634e-05, "loss": 0.0028, "step": 52980 }, { "grad_norm": 0.06935534626245499, "learning_rate": 4.9191488497739564e-05, "loss": 0.0026, "step": 52990 }, { "grad_norm": 0.07254811376333237, "learning_rate": 4.917495600593042e-05, "loss": 0.0025, "step": 53000 }, { "grad_norm": 0.06696958094835281, "learning_rate": 4.915842360434684e-05, "loss": 0.0023, "step": 53010 }, { "grad_norm": 0.07516641914844513, "learning_rate": 4.914189129479679e-05, "loss": 0.0026, "step": 53020 }, { "grad_norm": 0.06482943147420883, "learning_rate": 4.912535907908822e-05, "loss": 0.0032, "step": 53030 }, { "grad_norm": 0.052931495010852814, "learning_rate": 4.910882695902907e-05, "loss": 0.0037, "step": 53040 }, { "grad_norm": 0.0773751512169838, "learning_rate": 4.9092294936427234e-05, "loss": 0.0029, "step": 53050 }, { "grad_norm": 0.05921147018671036, "learning_rate": 4.9075763013090685e-05, "loss": 0.0034, "step": 53060 }, { "grad_norm": 0.07403049618005753, "learning_rate": 4.9059231190827305e-05, "loss": 0.0034, "step": 53070 }, { "grad_norm": 0.07757978141307831, "learning_rate": 4.904269947144499e-05, "loss": 0.0031, "step": 53080 }, { "grad_norm": 0.06928006559610367, "learning_rate": 4.902616785675161e-05, "loss": 0.0028, "step": 53090 }, { "grad_norm": 0.07921089977025986, "learning_rate": 4.900963634855505e-05, "loss": 0.0026, "step": 53100 }, { "grad_norm": 0.09242470562458038, "learning_rate": 4.8993104948663173e-05, "loss": 0.003, "step": 53110 }, { "grad_norm": 0.0840369388461113, "learning_rate": 4.8976573658883823e-05, "loss": 0.004, "step": 53120 }, { "grad_norm": 0.07899650186300278, "learning_rate": 4.896004248102484e-05, "loss": 0.0036, "step": 53130 }, { "grad_norm": 0.09872165322303772, "learning_rate": 4.8943511416894045e-05, "loss": 0.0032, "step": 53140 }, { "grad_norm": 0.0607881136238575, "learning_rate": 4.892698046829926e-05, "loss": 0.0023, "step": 53150 }, { "grad_norm": 0.07007057964801788, "learning_rate": 4.8910449637048275e-05, "loss": 0.003, "step": 53160 }, { "grad_norm": 0.06309864670038223, "learning_rate": 4.8893918924948884e-05, "loss": 0.0025, "step": 53170 }, { "grad_norm": 0.06073552742600441, "learning_rate": 4.8877388333808835e-05, "loss": 0.0029, "step": 53180 }, { "grad_norm": 0.07438739389181137, "learning_rate": 4.8860857865435925e-05, "loss": 0.0023, "step": 53190 }, { "grad_norm": 0.09896988421678543, "learning_rate": 4.884432752163789e-05, "loss": 0.005, "step": 53200 }, { "grad_norm": 0.11459863185882568, "learning_rate": 4.8827797304222466e-05, "loss": 0.0041, "step": 53210 }, { "grad_norm": 0.11779090762138367, "learning_rate": 4.8811267214997366e-05, "loss": 0.0028, "step": 53220 }, { "grad_norm": 0.07307810336351395, "learning_rate": 4.879473725577029e-05, "loss": 0.0029, "step": 53230 }, { "grad_norm": 0.3620987832546234, "learning_rate": 4.877820742834891e-05, "loss": 0.0074, "step": 53240 }, { "grad_norm": 0.12594042718410492, "learning_rate": 4.8761677734540956e-05, "loss": 0.005, "step": 53250 }, { "grad_norm": 0.08614569902420044, "learning_rate": 4.874514817615407e-05, "loss": 0.0041, "step": 53260 }, { "grad_norm": 0.10379932075738907, "learning_rate": 4.872861875499588e-05, "loss": 0.0049, "step": 53270 }, { "grad_norm": 0.10011274367570877, "learning_rate": 4.871208947287404e-05, "loss": 0.0034, "step": 53280 }, { "grad_norm": 0.08394595235586166, "learning_rate": 4.869556033159615e-05, "loss": 0.0033, "step": 53290 }, { "grad_norm": 0.10372868925333023, "learning_rate": 4.8679031332969835e-05, "loss": 0.0041, "step": 53300 }, { "grad_norm": 0.11560031026601791, "learning_rate": 4.8662502478802635e-05, "loss": 0.0065, "step": 53310 }, { "grad_norm": 0.09964508563280106, "learning_rate": 4.8645973770902176e-05, "loss": 0.006, "step": 53320 }, { "grad_norm": 0.06455229967832565, "learning_rate": 4.862944521107599e-05, "loss": 0.0035, "step": 53330 }, { "grad_norm": 0.0892808809876442, "learning_rate": 4.861291680113162e-05, "loss": 0.0042, "step": 53340 }, { "grad_norm": 0.09503065794706345, "learning_rate": 4.8596388542876563e-05, "loss": 0.0037, "step": 53350 }, { "grad_norm": 0.09493212401866913, "learning_rate": 4.857986043811836e-05, "loss": 0.0045, "step": 53360 }, { "grad_norm": 0.0915474072098732, "learning_rate": 4.856333248866448e-05, "loss": 0.0038, "step": 53370 }, { "grad_norm": 0.08914093673229218, "learning_rate": 4.85468046963224e-05, "loss": 0.0028, "step": 53380 }, { "grad_norm": 0.10748354345560074, "learning_rate": 4.853027706289957e-05, "loss": 0.0035, "step": 53390 }, { "grad_norm": 0.08698085695505142, "learning_rate": 4.851374959020344e-05, "loss": 0.0034, "step": 53400 }, { "grad_norm": 0.07237076759338379, "learning_rate": 4.849722228004141e-05, "loss": 0.003, "step": 53410 }, { "grad_norm": 0.0744117721915245, "learning_rate": 4.848069513422091e-05, "loss": 0.0032, "step": 53420 }, { "grad_norm": 0.05208548158407211, "learning_rate": 4.8464168154549294e-05, "loss": 0.0027, "step": 53430 }, { "grad_norm": 0.07044117152690887, "learning_rate": 4.844764134283392e-05, "loss": 0.0031, "step": 53440 }, { "grad_norm": 0.06292032450437546, "learning_rate": 4.843111470088218e-05, "loss": 0.0028, "step": 53450 }, { "grad_norm": 0.0646747276186943, "learning_rate": 4.841458823050137e-05, "loss": 0.0024, "step": 53460 }, { "grad_norm": 0.06078604236245155, "learning_rate": 4.8398061933498816e-05, "loss": 0.0023, "step": 53470 }, { "grad_norm": 0.0611426942050457, "learning_rate": 4.83815358116818e-05, "loss": 0.0025, "step": 53480 }, { "grad_norm": 0.044266700744628906, "learning_rate": 4.8365009866857584e-05, "loss": 0.0021, "step": 53490 }, { "grad_norm": 0.06024504452943802, "learning_rate": 4.834848410083342e-05, "loss": 0.0027, "step": 53500 }, { "grad_norm": 0.09101380407810211, "learning_rate": 4.833195851541657e-05, "loss": 0.0027, "step": 53510 }, { "grad_norm": 0.12891283631324768, "learning_rate": 4.831543311241423e-05, "loss": 0.0038, "step": 53520 }, { "grad_norm": 0.09645186364650726, "learning_rate": 4.829890789363359e-05, "loss": 0.0027, "step": 53530 }, { "grad_norm": 0.0870882123708725, "learning_rate": 4.828238286088182e-05, "loss": 0.0037, "step": 53540 }, { "grad_norm": 0.08688965439796448, "learning_rate": 4.826585801596609e-05, "loss": 0.0037, "step": 53550 }, { "grad_norm": 0.07229097187519073, "learning_rate": 4.8249333360693486e-05, "loss": 0.003, "step": 53560 }, { "grad_norm": 0.08036959171295166, "learning_rate": 4.8232808896871185e-05, "loss": 0.003, "step": 53570 }, { "grad_norm": 0.07428815960884094, "learning_rate": 4.821628462630624e-05, "loss": 0.0027, "step": 53580 }, { "grad_norm": 0.05783337354660034, "learning_rate": 4.8199760550805735e-05, "loss": 0.0028, "step": 53590 }, { "grad_norm": 0.10059075802564621, "learning_rate": 4.818323667217669e-05, "loss": 0.0031, "step": 53600 }, { "grad_norm": 0.06660525500774384, "learning_rate": 4.8166712992226165e-05, "loss": 0.0033, "step": 53610 }, { "grad_norm": 0.07282088696956635, "learning_rate": 4.8150189512761154e-05, "loss": 0.0028, "step": 53620 }, { "grad_norm": 0.06339338421821594, "learning_rate": 4.813366623558862e-05, "loss": 0.0038, "step": 53630 }, { "grad_norm": 0.07977008819580078, "learning_rate": 4.811714316251554e-05, "loss": 0.003, "step": 53640 }, { "grad_norm": 0.05759918689727783, "learning_rate": 4.810062029534886e-05, "loss": 0.0023, "step": 53650 }, { "grad_norm": 0.0519016869366169, "learning_rate": 4.8084097635895505e-05, "loss": 0.0034, "step": 53660 }, { "grad_norm": 0.0570373609662056, "learning_rate": 4.8067575185962335e-05, "loss": 0.0028, "step": 53670 }, { "grad_norm": 0.08715953677892685, "learning_rate": 4.805105294735625e-05, "loss": 0.004, "step": 53680 }, { "grad_norm": 0.09916187077760696, "learning_rate": 4.803453092188406e-05, "loss": 0.0034, "step": 53690 }, { "grad_norm": 0.07167455554008484, "learning_rate": 4.801800911135263e-05, "loss": 0.0025, "step": 53700 }, { "grad_norm": 0.07965677231550217, "learning_rate": 4.8001487517568744e-05, "loss": 0.0036, "step": 53710 }, { "grad_norm": 0.08381063491106033, "learning_rate": 4.798496614233918e-05, "loss": 0.0029, "step": 53720 }, { "grad_norm": 0.06421691924333572, "learning_rate": 4.796844498747069e-05, "loss": 0.0028, "step": 53730 }, { "grad_norm": 0.06548886001110077, "learning_rate": 4.795192405477e-05, "loss": 0.003, "step": 53740 }, { "grad_norm": 0.09315869212150574, "learning_rate": 4.79354033460438e-05, "loss": 0.0026, "step": 53750 }, { "grad_norm": 0.05962192639708519, "learning_rate": 4.79188828630988e-05, "loss": 0.0027, "step": 53760 }, { "grad_norm": 0.12039980292320251, "learning_rate": 4.790236260774165e-05, "loss": 0.0032, "step": 53770 }, { "grad_norm": 0.06994666904211044, "learning_rate": 4.788584258177896e-05, "loss": 0.0025, "step": 53780 }, { "grad_norm": 0.07314051687717438, "learning_rate": 4.7869322787017356e-05, "loss": 0.0026, "step": 53790 }, { "grad_norm": 0.10294179618358612, "learning_rate": 4.78528032252634e-05, "loss": 0.0034, "step": 53800 }, { "grad_norm": 0.10232222825288773, "learning_rate": 4.783628389832366e-05, "loss": 0.003, "step": 53810 }, { "grad_norm": 0.05890953913331032, "learning_rate": 4.7819764808004644e-05, "loss": 0.0024, "step": 53820 }, { "grad_norm": 0.07166987657546997, "learning_rate": 4.780324595611289e-05, "loss": 0.0032, "step": 53830 }, { "grad_norm": 0.05441073328256607, "learning_rate": 4.7786727344454854e-05, "loss": 0.0027, "step": 53840 }, { "grad_norm": 0.2227925807237625, "learning_rate": 4.777020897483697e-05, "loss": 0.0036, "step": 53850 }, { "grad_norm": 0.06557342410087585, "learning_rate": 4.775369084906569e-05, "loss": 0.0027, "step": 53860 }, { "grad_norm": 0.1021660640835762, "learning_rate": 4.773717296894742e-05, "loss": 0.0042, "step": 53870 }, { "grad_norm": 0.14317691326141357, "learning_rate": 4.772065533628848e-05, "loss": 0.0033, "step": 53880 }, { "grad_norm": 0.0910939872264862, "learning_rate": 4.7704137952895265e-05, "loss": 0.0031, "step": 53890 }, { "grad_norm": 0.07111383974552155, "learning_rate": 4.768762082057407e-05, "loss": 0.0036, "step": 53900 }, { "grad_norm": 0.07335057854652405, "learning_rate": 4.7671103941131186e-05, "loss": 0.0029, "step": 53910 }, { "grad_norm": 0.06899211555719376, "learning_rate": 4.765458731637287e-05, "loss": 0.0041, "step": 53920 }, { "grad_norm": 0.06324039399623871, "learning_rate": 4.763807094810537e-05, "loss": 0.0032, "step": 53930 }, { "grad_norm": 0.059958316385746, "learning_rate": 4.762155483813487e-05, "loss": 0.0028, "step": 53940 }, { "grad_norm": 0.07966820895671844, "learning_rate": 4.760503898826754e-05, "loss": 0.0028, "step": 53950 }, { "grad_norm": 0.07474663853645325, "learning_rate": 4.758852340030957e-05, "loss": 0.0032, "step": 53960 }, { "grad_norm": 0.09963833540678024, "learning_rate": 4.757200807606705e-05, "loss": 0.0039, "step": 53970 }, { "grad_norm": 0.046395041048526764, "learning_rate": 4.755549301734607e-05, "loss": 0.003, "step": 53980 }, { "grad_norm": 0.08297092467546463, "learning_rate": 4.753897822595271e-05, "loss": 0.0032, "step": 53990 }, { "grad_norm": 0.07471703737974167, "learning_rate": 4.7522463703692994e-05, "loss": 0.0028, "step": 54000 }, { "grad_norm": 0.13210467994213104, "learning_rate": 4.75059494523729e-05, "loss": 0.0035, "step": 54010 }, { "grad_norm": 0.07460907846689224, "learning_rate": 4.748943547379843e-05, "loss": 0.0032, "step": 54020 }, { "grad_norm": 0.08048216998577118, "learning_rate": 4.7472921769775535e-05, "loss": 0.0031, "step": 54030 }, { "grad_norm": 0.07331380248069763, "learning_rate": 4.745640834211011e-05, "loss": 0.0031, "step": 54040 }, { "grad_norm": 0.05868779495358467, "learning_rate": 4.7439895192608054e-05, "loss": 0.0023, "step": 54050 }, { "grad_norm": 0.1004970371723175, "learning_rate": 4.7423382323075186e-05, "loss": 0.0031, "step": 54060 }, { "grad_norm": 0.0735560953617096, "learning_rate": 4.740686973531737e-05, "loss": 0.0026, "step": 54070 }, { "grad_norm": 0.07685847580432892, "learning_rate": 4.7390357431140375e-05, "loss": 0.0026, "step": 54080 }, { "grad_norm": 0.07267553359270096, "learning_rate": 4.737384541234998e-05, "loss": 0.0036, "step": 54090 }, { "grad_norm": 0.07873711735010147, "learning_rate": 4.735733368075189e-05, "loss": 0.003, "step": 54100 }, { "grad_norm": 0.07242922484874725, "learning_rate": 4.734082223815182e-05, "loss": 0.0031, "step": 54110 }, { "grad_norm": 0.06753179430961609, "learning_rate": 4.7324311086355436e-05, "loss": 0.0029, "step": 54120 }, { "grad_norm": 0.06849268823862076, "learning_rate": 4.730780022716836e-05, "loss": 0.0031, "step": 54130 }, { "grad_norm": 0.097519151866436, "learning_rate": 4.729128966239619e-05, "loss": 0.0027, "step": 54140 }, { "grad_norm": 0.07019013166427612, "learning_rate": 4.7274779393844536e-05, "loss": 0.0031, "step": 54150 }, { "grad_norm": 0.07878334820270538, "learning_rate": 4.725826942331891e-05, "loss": 0.0028, "step": 54160 }, { "grad_norm": 0.05153675749897957, "learning_rate": 4.724175975262481e-05, "loss": 0.0029, "step": 54170 }, { "grad_norm": 0.0765775665640831, "learning_rate": 4.722525038356772e-05, "loss": 0.0022, "step": 54180 }, { "grad_norm": 0.05652947723865509, "learning_rate": 4.7208741317953075e-05, "loss": 0.0024, "step": 54190 }, { "grad_norm": 0.05959858000278473, "learning_rate": 4.7192232557586264e-05, "loss": 0.0033, "step": 54200 }, { "grad_norm": 0.05594545975327492, "learning_rate": 4.717572410427269e-05, "loss": 0.0035, "step": 54210 }, { "grad_norm": 0.07010294497013092, "learning_rate": 4.71592159598177e-05, "loss": 0.0036, "step": 54220 }, { "grad_norm": 0.09676217287778854, "learning_rate": 4.714270812602657e-05, "loss": 0.0025, "step": 54230 }, { "grad_norm": 0.0684170201420784, "learning_rate": 4.712620060470458e-05, "loss": 0.0036, "step": 54240 }, { "grad_norm": 0.05806944519281387, "learning_rate": 4.7109693397656986e-05, "loss": 0.0033, "step": 54250 }, { "grad_norm": 0.0670706257224083, "learning_rate": 4.709318650668894e-05, "loss": 0.0027, "step": 54260 }, { "grad_norm": 0.05759730190038681, "learning_rate": 4.7076679933605676e-05, "loss": 0.0023, "step": 54270 }, { "grad_norm": 0.07589448988437653, "learning_rate": 4.70601736802123e-05, "loss": 0.0031, "step": 54280 }, { "grad_norm": 0.06885047256946564, "learning_rate": 4.704366774831391e-05, "loss": 0.003, "step": 54290 }, { "grad_norm": 0.060359206050634384, "learning_rate": 4.702716213971557e-05, "loss": 0.004, "step": 54300 }, { "grad_norm": 0.055003076791763306, "learning_rate": 4.70106568562223e-05, "loss": 0.0023, "step": 54310 }, { "grad_norm": 0.0660797581076622, "learning_rate": 4.6994151899639105e-05, "loss": 0.0035, "step": 54320 }, { "grad_norm": 0.1697942614555359, "learning_rate": 4.697764727177093e-05, "loss": 0.0031, "step": 54330 }, { "grad_norm": 0.07114126533269882, "learning_rate": 4.696114297442272e-05, "loss": 0.0041, "step": 54340 }, { "grad_norm": 0.1276724487543106, "learning_rate": 4.694463900939933e-05, "loss": 0.0042, "step": 54350 }, { "grad_norm": 0.07985126227140427, "learning_rate": 4.692813537850564e-05, "loss": 0.0039, "step": 54360 }, { "grad_norm": 0.07434176653623581, "learning_rate": 4.6911632083546445e-05, "loss": 0.0048, "step": 54370 }, { "grad_norm": 0.06258555501699448, "learning_rate": 4.6895129126326525e-05, "loss": 0.0027, "step": 54380 }, { "grad_norm": 0.051562387496232986, "learning_rate": 4.687862650865059e-05, "loss": 0.0031, "step": 54390 }, { "grad_norm": 0.0789204090833664, "learning_rate": 4.686212423232339e-05, "loss": 0.0033, "step": 54400 }, { "grad_norm": 0.0700882077217102, "learning_rate": 4.6845622299149565e-05, "loss": 0.0026, "step": 54410 }, { "grad_norm": 0.05034729838371277, "learning_rate": 4.682912071093374e-05, "loss": 0.0032, "step": 54420 }, { "grad_norm": 0.07498861849308014, "learning_rate": 4.681261946948052e-05, "loss": 0.0026, "step": 54430 }, { "grad_norm": 0.08274982124567032, "learning_rate": 4.679611857659443e-05, "loss": 0.0031, "step": 54440 }, { "grad_norm": 0.05035588517785072, "learning_rate": 4.6779618034079975e-05, "loss": 0.0028, "step": 54450 }, { "grad_norm": 0.061398159712553024, "learning_rate": 4.6763117843741675e-05, "loss": 0.0036, "step": 54460 }, { "grad_norm": 0.06728263944387436, "learning_rate": 4.674661800738394e-05, "loss": 0.0032, "step": 54470 }, { "grad_norm": 0.04964457079768181, "learning_rate": 4.673011852681117e-05, "loss": 0.0037, "step": 54480 }, { "grad_norm": 0.06072668358683586, "learning_rate": 4.671361940382772e-05, "loss": 0.0025, "step": 54490 }, { "grad_norm": 0.05830087512731552, "learning_rate": 4.669712064023791e-05, "loss": 0.0022, "step": 54500 }, { "grad_norm": 0.06094582751393318, "learning_rate": 4.668062223784602e-05, "loss": 0.0026, "step": 54510 }, { "grad_norm": 0.056572575122117996, "learning_rate": 4.6664124198456274e-05, "loss": 0.0027, "step": 54520 }, { "grad_norm": 0.06403122842311859, "learning_rate": 4.6647626523872905e-05, "loss": 0.0025, "step": 54530 }, { "grad_norm": 0.055845484137535095, "learning_rate": 4.6631129215900066e-05, "loss": 0.0021, "step": 54540 }, { "grad_norm": 0.08879842609167099, "learning_rate": 4.661463227634187e-05, "loss": 0.0026, "step": 54550 }, { "grad_norm": 0.05493651703000069, "learning_rate": 4.659813570700238e-05, "loss": 0.0026, "step": 54560 }, { "grad_norm": 0.07924731075763702, "learning_rate": 4.658163950968568e-05, "loss": 0.0023, "step": 54570 }, { "grad_norm": 0.08091669529676437, "learning_rate": 4.6565143686195726e-05, "loss": 0.0022, "step": 54580 }, { "grad_norm": 0.07314179837703705, "learning_rate": 4.654864823833649e-05, "loss": 0.0023, "step": 54590 }, { "grad_norm": 0.06189890205860138, "learning_rate": 4.65321531679119e-05, "loss": 0.0026, "step": 54600 }, { "grad_norm": 0.0638677328824997, "learning_rate": 4.6515658476725834e-05, "loss": 0.0023, "step": 54610 }, { "grad_norm": 0.07211340218782425, "learning_rate": 4.649916416658212e-05, "loss": 0.0037, "step": 54620 }, { "grad_norm": 0.0617234893143177, "learning_rate": 4.648267023928455e-05, "loss": 0.0027, "step": 54630 }, { "grad_norm": 0.058843016624450684, "learning_rate": 4.646617669663687e-05, "loss": 0.0031, "step": 54640 }, { "grad_norm": 0.05929896607995033, "learning_rate": 4.644968354044277e-05, "loss": 0.0023, "step": 54650 }, { "grad_norm": 0.05366167426109314, "learning_rate": 4.643319077250596e-05, "loss": 0.0022, "step": 54660 }, { "grad_norm": 0.06190834194421768, "learning_rate": 4.6416698394630055e-05, "loss": 0.0019, "step": 54670 }, { "grad_norm": 0.09892474114894867, "learning_rate": 4.640020640861862e-05, "loss": 0.0025, "step": 54680 }, { "grad_norm": 0.10757997632026672, "learning_rate": 4.6383714816275195e-05, "loss": 0.0023, "step": 54690 }, { "grad_norm": 0.0769570916891098, "learning_rate": 4.636722361940329e-05, "loss": 0.0024, "step": 54700 }, { "grad_norm": 0.06730765104293823, "learning_rate": 4.635073281980632e-05, "loss": 0.0035, "step": 54710 }, { "grad_norm": 0.05395396798849106, "learning_rate": 4.633424241928775e-05, "loss": 0.002, "step": 54720 }, { "grad_norm": 0.07088840752840042, "learning_rate": 4.6317752419650914e-05, "loss": 0.0033, "step": 54730 }, { "grad_norm": 0.07155153155326843, "learning_rate": 4.630126282269914e-05, "loss": 0.0022, "step": 54740 }, { "grad_norm": 0.066811703145504, "learning_rate": 4.628477363023569e-05, "loss": 0.004, "step": 54750 }, { "grad_norm": 0.08156538754701614, "learning_rate": 4.626828484406383e-05, "loss": 0.0028, "step": 54760 }, { "grad_norm": 0.06066711246967316, "learning_rate": 4.625179646598669e-05, "loss": 0.0031, "step": 54770 }, { "grad_norm": 0.07448559999465942, "learning_rate": 4.6235308497807487e-05, "loss": 0.0025, "step": 54780 }, { "grad_norm": 0.05508005619049072, "learning_rate": 4.6218820941329276e-05, "loss": 0.0022, "step": 54790 }, { "grad_norm": 0.05207338556647301, "learning_rate": 4.620233379835513e-05, "loss": 0.0023, "step": 54800 }, { "grad_norm": 0.053718067705631256, "learning_rate": 4.618584707068804e-05, "loss": 0.0029, "step": 54810 }, { "grad_norm": 0.0755675733089447, "learning_rate": 4.616936076013099e-05, "loss": 0.0022, "step": 54820 }, { "grad_norm": 0.07053493708372116, "learning_rate": 4.615287486848688e-05, "loss": 0.0029, "step": 54830 }, { "grad_norm": 0.1279430240392685, "learning_rate": 4.6136389397558584e-05, "loss": 0.0044, "step": 54840 }, { "grad_norm": 0.06868185102939606, "learning_rate": 4.6119904349148935e-05, "loss": 0.0028, "step": 54850 }, { "grad_norm": 0.06320211291313171, "learning_rate": 4.6103419725060726e-05, "loss": 0.0027, "step": 54860 }, { "grad_norm": 0.05717796832323074, "learning_rate": 4.608693552709667e-05, "loss": 0.0028, "step": 54870 }, { "grad_norm": 0.053960032761096954, "learning_rate": 4.607045175705947e-05, "loss": 0.0035, "step": 54880 }, { "grad_norm": 0.07499374449253082, "learning_rate": 4.605396841675176e-05, "loss": 0.0039, "step": 54890 }, { "grad_norm": 0.08053741604089737, "learning_rate": 4.6037485507976094e-05, "loss": 0.0048, "step": 54900 }, { "grad_norm": 0.061444852501153946, "learning_rate": 4.602100303253509e-05, "loss": 0.0026, "step": 54910 }, { "grad_norm": 0.06940062344074249, "learning_rate": 4.600452099223122e-05, "loss": 0.0042, "step": 54920 }, { "grad_norm": 0.07949309051036835, "learning_rate": 4.598803938886691e-05, "loss": 0.0025, "step": 54930 }, { "grad_norm": 0.052048634737730026, "learning_rate": 4.597155822424459e-05, "loss": 0.0027, "step": 54940 }, { "grad_norm": 0.10299469530582428, "learning_rate": 4.59550775001666e-05, "loss": 0.0036, "step": 54950 }, { "grad_norm": 0.06857293844223022, "learning_rate": 4.5938597218435234e-05, "loss": 0.0028, "step": 54960 }, { "grad_norm": 0.07530813664197922, "learning_rate": 4.592211738085279e-05, "loss": 0.0024, "step": 54970 }, { "grad_norm": 0.05126016587018967, "learning_rate": 4.590563798922146e-05, "loss": 0.0028, "step": 54980 }, { "grad_norm": 0.06932180374860764, "learning_rate": 4.5889159045343404e-05, "loss": 0.0021, "step": 54990 }, { "grad_norm": 0.08249469101428986, "learning_rate": 4.5872680551020734e-05, "loss": 0.0031, "step": 55000 }, { "grad_norm": 0.038603395223617554, "learning_rate": 4.585620250805552e-05, "loss": 0.0031, "step": 55010 }, { "grad_norm": 0.08200179785490036, "learning_rate": 4.583972491824975e-05, "loss": 0.0033, "step": 55020 }, { "grad_norm": 0.07875894010066986, "learning_rate": 4.582324778340541e-05, "loss": 0.0031, "step": 55030 }, { "grad_norm": 0.06585756689310074, "learning_rate": 4.5806771105324424e-05, "loss": 0.0031, "step": 55040 }, { "grad_norm": 0.05682782828807831, "learning_rate": 4.579029488580865e-05, "loss": 0.0022, "step": 55050 }, { "grad_norm": 0.09107225388288498, "learning_rate": 4.5773819126659876e-05, "loss": 0.0027, "step": 55060 }, { "grad_norm": 0.06812741607427597, "learning_rate": 4.575734382967991e-05, "loss": 0.0025, "step": 55070 }, { "grad_norm": 0.047361601144075394, "learning_rate": 4.574086899667044e-05, "loss": 0.0028, "step": 55080 }, { "grad_norm": 0.17327076196670532, "learning_rate": 4.5724394629433106e-05, "loss": 0.0033, "step": 55090 }, { "grad_norm": 0.0776846706867218, "learning_rate": 4.570792072976958e-05, "loss": 0.0023, "step": 55100 }, { "grad_norm": 0.06097107008099556, "learning_rate": 4.569144729948138e-05, "loss": 0.0021, "step": 55110 }, { "grad_norm": 0.06813345849514008, "learning_rate": 4.5674974340370033e-05, "loss": 0.003, "step": 55120 }, { "grad_norm": 0.06584915518760681, "learning_rate": 4.565850185423699e-05, "loss": 0.0028, "step": 55130 }, { "grad_norm": 0.07534454762935638, "learning_rate": 4.564202984288365e-05, "loss": 0.0026, "step": 55140 }, { "grad_norm": 0.08282314985990524, "learning_rate": 4.5625558308111356e-05, "loss": 0.0043, "step": 55150 }, { "grad_norm": 0.08720893412828445, "learning_rate": 4.560908725172145e-05, "loss": 0.0033, "step": 55160 }, { "grad_norm": 0.10209605097770691, "learning_rate": 4.559261667551516e-05, "loss": 0.0032, "step": 55170 }, { "grad_norm": 0.08631248027086258, "learning_rate": 4.5576146581293685e-05, "loss": 0.0031, "step": 55180 }, { "grad_norm": 0.3314840495586395, "learning_rate": 4.5559676970858164e-05, "loss": 0.0041, "step": 55190 }, { "grad_norm": 0.10300199687480927, "learning_rate": 4.5543207846009705e-05, "loss": 0.0034, "step": 55200 }, { "grad_norm": 0.0759178102016449, "learning_rate": 4.552673920854933e-05, "loss": 0.0038, "step": 55210 }, { "grad_norm": 0.06961552053689957, "learning_rate": 4.5510271060278006e-05, "loss": 0.0045, "step": 55220 }, { "grad_norm": 0.08105465024709702, "learning_rate": 4.5493803402996704e-05, "loss": 0.0026, "step": 55230 }, { "grad_norm": 0.08186173439025879, "learning_rate": 4.5477336238506295e-05, "loss": 0.003, "step": 55240 }, { "grad_norm": 0.07513102144002914, "learning_rate": 4.546086956860759e-05, "loss": 0.0035, "step": 55250 }, { "grad_norm": 0.14645259082317352, "learning_rate": 4.5444403395101374e-05, "loss": 0.0031, "step": 55260 }, { "grad_norm": 0.08519914746284485, "learning_rate": 4.542793771978834e-05, "loss": 0.0031, "step": 55270 }, { "grad_norm": 0.07432830333709717, "learning_rate": 4.541147254446916e-05, "loss": 0.004, "step": 55280 }, { "grad_norm": 0.09347512573003769, "learning_rate": 4.539500787094445e-05, "loss": 0.0029, "step": 55290 }, { "grad_norm": 0.08075574785470963, "learning_rate": 4.537854370101475e-05, "loss": 0.0028, "step": 55300 }, { "grad_norm": 0.08533170819282532, "learning_rate": 4.5362080036480566e-05, "loss": 0.0031, "step": 55310 }, { "grad_norm": 0.04628005251288414, "learning_rate": 4.534561687914235e-05, "loss": 0.0024, "step": 55320 }, { "grad_norm": 0.12390338629484177, "learning_rate": 4.5329154230800466e-05, "loss": 0.0043, "step": 55330 }, { "grad_norm": 0.10599128156900406, "learning_rate": 4.531269209325525e-05, "loss": 0.0029, "step": 55340 }, { "grad_norm": 0.06882426887750626, "learning_rate": 4.529623046830696e-05, "loss": 0.0041, "step": 55350 }, { "grad_norm": 0.08478546142578125, "learning_rate": 4.5279769357755866e-05, "loss": 0.0033, "step": 55360 }, { "grad_norm": 0.08649618178606033, "learning_rate": 4.5263308763402084e-05, "loss": 0.0038, "step": 55370 }, { "grad_norm": 0.09157413244247437, "learning_rate": 4.524684868704574e-05, "loss": 0.0029, "step": 55380 }, { "grad_norm": 0.08843014389276505, "learning_rate": 4.523038913048688e-05, "loss": 0.0038, "step": 55390 }, { "grad_norm": 0.08758548647165298, "learning_rate": 4.521393009552549e-05, "loss": 0.0037, "step": 55400 }, { "grad_norm": 0.061497028917074203, "learning_rate": 4.5197471583961484e-05, "loss": 0.0033, "step": 55410 }, { "grad_norm": 0.07118411362171173, "learning_rate": 4.51810135975948e-05, "loss": 0.0037, "step": 55420 }, { "grad_norm": 0.058318689465522766, "learning_rate": 4.516455613822521e-05, "loss": 0.0027, "step": 55430 }, { "grad_norm": 0.051244039088487625, "learning_rate": 4.514809920765249e-05, "loss": 0.0021, "step": 55440 }, { "grad_norm": 0.07423142343759537, "learning_rate": 4.5131642807676356e-05, "loss": 0.0023, "step": 55450 }, { "grad_norm": 0.06798680126667023, "learning_rate": 4.511518694009644e-05, "loss": 0.0026, "step": 55460 }, { "grad_norm": 0.1634690761566162, "learning_rate": 4.509873160671231e-05, "loss": 0.0033, "step": 55470 }, { "grad_norm": 0.07180041819810867, "learning_rate": 4.508227680932356e-05, "loss": 0.0026, "step": 55480 }, { "grad_norm": 0.059911567717790604, "learning_rate": 4.5065822549729605e-05, "loss": 0.0029, "step": 55490 }, { "grad_norm": 0.07496245950460434, "learning_rate": 4.504936882972989e-05, "loss": 0.0033, "step": 55500 }, { "grad_norm": 0.06159280985593796, "learning_rate": 4.503291565112374e-05, "loss": 0.0026, "step": 55510 }, { "grad_norm": 0.061585862189531326, "learning_rate": 4.501646301571048e-05, "loss": 0.0026, "step": 55520 }, { "grad_norm": 0.07536688446998596, "learning_rate": 4.500001092528933e-05, "loss": 0.0027, "step": 55530 }, { "grad_norm": 0.12807795405387878, "learning_rate": 4.4983559381659455e-05, "loss": 0.0035, "step": 55540 }, { "grad_norm": 0.06580293923616409, "learning_rate": 4.496710838661999e-05, "loss": 0.0027, "step": 55550 }, { "grad_norm": 0.08215323090553284, "learning_rate": 4.495065794196999e-05, "loss": 0.0029, "step": 55560 }, { "grad_norm": 0.13402262330055237, "learning_rate": 4.4934208049508455e-05, "loss": 0.0036, "step": 55570 }, { "grad_norm": 0.09113701432943344, "learning_rate": 4.4917758711034303e-05, "loss": 0.0038, "step": 55580 }, { "grad_norm": 0.06072688847780228, "learning_rate": 4.490130992834642e-05, "loss": 0.005, "step": 55590 }, { "grad_norm": 0.06886756420135498, "learning_rate": 4.488486170324359e-05, "loss": 0.0039, "step": 55600 }, { "grad_norm": 0.14989544451236725, "learning_rate": 4.486841403752463e-05, "loss": 0.0052, "step": 55610 }, { "grad_norm": 0.0730823278427124, "learning_rate": 4.4851966932988174e-05, "loss": 0.0035, "step": 55620 }, { "grad_norm": 0.05680279806256294, "learning_rate": 4.4835520391432894e-05, "loss": 0.0033, "step": 55630 }, { "grad_norm": 0.06075727939605713, "learning_rate": 4.481907441465733e-05, "loss": 0.0026, "step": 55640 }, { "grad_norm": 0.0665695071220398, "learning_rate": 4.480262900446e-05, "loss": 0.0026, "step": 55650 }, { "grad_norm": 0.07403063774108887, "learning_rate": 4.478618416263933e-05, "loss": 0.0027, "step": 55660 }, { "grad_norm": 0.08140137046575546, "learning_rate": 4.476973989099373e-05, "loss": 0.0031, "step": 55670 }, { "grad_norm": 0.06172248348593712, "learning_rate": 4.475329619132153e-05, "loss": 0.0028, "step": 55680 }, { "grad_norm": 0.057416435331106186, "learning_rate": 4.4736853065420956e-05, "loss": 0.0027, "step": 55690 }, { "grad_norm": 0.06255793571472168, "learning_rate": 4.472041051509023e-05, "loss": 0.0031, "step": 55700 }, { "grad_norm": 0.0907863974571228, "learning_rate": 4.470396854212746e-05, "loss": 0.0026, "step": 55710 }, { "grad_norm": 0.07293717563152313, "learning_rate": 4.468752714833072e-05, "loss": 0.0023, "step": 55720 }, { "grad_norm": 0.06689997017383575, "learning_rate": 4.467108633549803e-05, "loss": 0.0031, "step": 55730 }, { "grad_norm": 0.07466975599527359, "learning_rate": 4.4654646105427335e-05, "loss": 0.0031, "step": 55740 }, { "grad_norm": 0.0682736486196518, "learning_rate": 4.463820645991651e-05, "loss": 0.0034, "step": 55750 }, { "grad_norm": 0.06653264909982681, "learning_rate": 4.462176740076334e-05, "loss": 0.0031, "step": 55760 }, { "grad_norm": 0.0624079629778862, "learning_rate": 4.4605328929765624e-05, "loss": 0.0029, "step": 55770 }, { "grad_norm": 0.09471254795789719, "learning_rate": 4.458889104872102e-05, "loss": 0.0026, "step": 55780 }, { "grad_norm": 0.07090573012828827, "learning_rate": 4.457245375942715e-05, "loss": 0.0024, "step": 55790 }, { "grad_norm": 0.07384113222360611, "learning_rate": 4.455601706368157e-05, "loss": 0.0026, "step": 55800 }, { "grad_norm": 0.09417910873889923, "learning_rate": 4.453958096328179e-05, "loss": 0.0021, "step": 55810 }, { "grad_norm": 0.08100197464227676, "learning_rate": 4.452314546002523e-05, "loss": 0.0024, "step": 55820 }, { "grad_norm": 0.06584969162940979, "learning_rate": 4.4506710555709246e-05, "loss": 0.0028, "step": 55830 }, { "grad_norm": 0.048569340258836746, "learning_rate": 4.449027625213114e-05, "loss": 0.0025, "step": 55840 }, { "grad_norm": 0.05623948946595192, "learning_rate": 4.4473842551088135e-05, "loss": 0.0021, "step": 55850 }, { "grad_norm": 0.06885084509849548, "learning_rate": 4.445740945437738e-05, "loss": 0.0031, "step": 55860 }, { "grad_norm": 0.05537700280547142, "learning_rate": 4.444097696379602e-05, "loss": 0.0034, "step": 55870 }, { "grad_norm": 0.06222613528370857, "learning_rate": 4.442454508114106e-05, "loss": 0.0027, "step": 55880 }, { "grad_norm": 0.07508378475904465, "learning_rate": 4.4408113808209465e-05, "loss": 0.003, "step": 55890 }, { "grad_norm": 0.05167922377586365, "learning_rate": 4.439168314679813e-05, "loss": 0.0024, "step": 55900 }, { "grad_norm": 0.06205074116587639, "learning_rate": 4.43752530987039e-05, "loss": 0.0028, "step": 55910 }, { "grad_norm": 0.05832629278302193, "learning_rate": 4.4358823665723515e-05, "loss": 0.0025, "step": 55920 }, { "grad_norm": 0.039784085005521774, "learning_rate": 4.434239484965371e-05, "loss": 0.0024, "step": 55930 }, { "grad_norm": 0.05488711595535278, "learning_rate": 4.4325966652291103e-05, "loss": 0.0022, "step": 55940 }, { "grad_norm": 0.07313796132802963, "learning_rate": 4.430953907543225e-05, "loss": 0.0028, "step": 55950 }, { "grad_norm": 0.05284300073981285, "learning_rate": 4.4293112120873645e-05, "loss": 0.003, "step": 55960 }, { "grad_norm": 0.05148620903491974, "learning_rate": 4.427668579041172e-05, "loss": 0.002, "step": 55970 }, { "grad_norm": 0.05364363640546799, "learning_rate": 4.4260260085842816e-05, "loss": 0.003, "step": 55980 }, { "grad_norm": 0.059432197362184525, "learning_rate": 4.4243835008963256e-05, "loss": 0.0026, "step": 55990 }, { "grad_norm": 0.0734643042087555, "learning_rate": 4.422741056156925e-05, "loss": 0.0026, "step": 56000 }, { "grad_norm": 0.06464904546737671, "learning_rate": 4.421098674545693e-05, "loss": 0.0029, "step": 56010 }, { "grad_norm": 0.07935579866170883, "learning_rate": 4.419456356242241e-05, "loss": 0.0028, "step": 56020 }, { "grad_norm": 0.0530337356030941, "learning_rate": 4.4178141014261685e-05, "loss": 0.0028, "step": 56030 }, { "grad_norm": 0.07466598600149155, "learning_rate": 4.416171910277071e-05, "loss": 0.0037, "step": 56040 }, { "grad_norm": 0.07845421880483627, "learning_rate": 4.414529782974533e-05, "loss": 0.0033, "step": 56050 }, { "grad_norm": 0.06718330085277557, "learning_rate": 4.41288771969814e-05, "loss": 0.0028, "step": 56060 }, { "grad_norm": 0.07697447389364243, "learning_rate": 4.411245720627464e-05, "loss": 0.004, "step": 56070 }, { "grad_norm": 0.06156957894563675, "learning_rate": 4.40960378594207e-05, "loss": 0.0022, "step": 56080 }, { "grad_norm": 0.07600629329681396, "learning_rate": 4.407961915821519e-05, "loss": 0.0041, "step": 56090 }, { "grad_norm": 0.07790613174438477, "learning_rate": 4.406320110445362e-05, "loss": 0.0031, "step": 56100 }, { "grad_norm": 0.12037357687950134, "learning_rate": 4.404678369993144e-05, "loss": 0.0035, "step": 56110 }, { "grad_norm": 0.09412963688373566, "learning_rate": 4.403036694644406e-05, "loss": 0.0036, "step": 56120 }, { "grad_norm": 0.10386180877685547, "learning_rate": 4.4013950845786764e-05, "loss": 0.0037, "step": 56130 }, { "grad_norm": 0.08129146695137024, "learning_rate": 4.399753539975482e-05, "loss": 0.0031, "step": 56140 }, { "grad_norm": 0.10588812083005905, "learning_rate": 4.398112061014337e-05, "loss": 0.0037, "step": 56150 }, { "grad_norm": 0.07489017397165298, "learning_rate": 4.396470647874753e-05, "loss": 0.0024, "step": 56160 }, { "grad_norm": 0.14188233017921448, "learning_rate": 4.394829300736229e-05, "loss": 0.0027, "step": 56170 }, { "grad_norm": 0.08799871057271957, "learning_rate": 4.393188019778265e-05, "loss": 0.0028, "step": 56180 }, { "grad_norm": 0.0942840725183487, "learning_rate": 4.391546805180347e-05, "loss": 0.0031, "step": 56190 }, { "grad_norm": 0.06648151576519012, "learning_rate": 4.389905657121955e-05, "loss": 0.0031, "step": 56200 }, { "grad_norm": 0.07495210319757462, "learning_rate": 4.3882645757825635e-05, "loss": 0.0032, "step": 56210 }, { "grad_norm": 0.07454205304384232, "learning_rate": 4.386623561341637e-05, "loss": 0.0037, "step": 56220 }, { "grad_norm": 0.08158718794584274, "learning_rate": 4.384982613978637e-05, "loss": 0.0026, "step": 56230 }, { "grad_norm": 0.052753616124391556, "learning_rate": 4.383341733873012e-05, "loss": 0.0025, "step": 56240 }, { "grad_norm": 0.06295919418334961, "learning_rate": 4.381700921204209e-05, "loss": 0.0023, "step": 56250 }, { "grad_norm": 0.0611623115837574, "learning_rate": 4.380060176151663e-05, "loss": 0.0033, "step": 56260 }, { "grad_norm": 0.08511371910572052, "learning_rate": 4.378419498894805e-05, "loss": 0.003, "step": 56270 }, { "grad_norm": 0.09565237909555435, "learning_rate": 4.376778889613056e-05, "loss": 0.0022, "step": 56280 }, { "grad_norm": 0.060163550078868866, "learning_rate": 4.375138348485831e-05, "loss": 0.0024, "step": 56290 }, { "grad_norm": 0.0632355809211731, "learning_rate": 4.3734978756925346e-05, "loss": 0.0037, "step": 56300 }, { "grad_norm": 0.058164164423942566, "learning_rate": 4.3718574714125706e-05, "loss": 0.0033, "step": 56310 }, { "grad_norm": 0.06326286494731903, "learning_rate": 4.370217135825329e-05, "loss": 0.0037, "step": 56320 }, { "grad_norm": 0.06307101994752884, "learning_rate": 4.368576869110194e-05, "loss": 0.0032, "step": 56330 }, { "grad_norm": 0.07013888657093048, "learning_rate": 4.366936671446544e-05, "loss": 0.0043, "step": 56340 }, { "grad_norm": 0.11606686562299728, "learning_rate": 4.3652965430137474e-05, "loss": 0.0033, "step": 56350 }, { "grad_norm": 0.08480223268270493, "learning_rate": 4.3636564839911646e-05, "loss": 0.0036, "step": 56360 }, { "grad_norm": 0.084404855966568, "learning_rate": 4.3620164945581545e-05, "loss": 0.0034, "step": 56370 }, { "grad_norm": 0.05953215807676315, "learning_rate": 4.360376574894061e-05, "loss": 0.0031, "step": 56380 }, { "grad_norm": 0.07591700553894043, "learning_rate": 4.358736725178224e-05, "loss": 0.0029, "step": 56390 }, { "grad_norm": 0.06410995870828629, "learning_rate": 4.357096945589974e-05, "loss": 0.0036, "step": 56400 }, { "grad_norm": 0.08114075660705566, "learning_rate": 4.355457236308636e-05, "loss": 0.0027, "step": 56410 }, { "grad_norm": 0.06902904808521271, "learning_rate": 4.353817597513526e-05, "loss": 0.0034, "step": 56420 }, { "grad_norm": 0.0741959810256958, "learning_rate": 4.352178029383948e-05, "loss": 0.0033, "step": 56430 }, { "grad_norm": 0.07724590599536896, "learning_rate": 4.3505385320992105e-05, "loss": 0.0033, "step": 56440 }, { "grad_norm": 0.06403248012065887, "learning_rate": 4.348899105838602e-05, "loss": 0.0027, "step": 56450 }, { "grad_norm": 0.06858321279287338, "learning_rate": 4.3472597507814087e-05, "loss": 0.0025, "step": 56460 }, { "grad_norm": 0.0645555779337883, "learning_rate": 4.3456204671069066e-05, "loss": 0.0026, "step": 56470 }, { "grad_norm": 0.04131188616156578, "learning_rate": 4.343981254994367e-05, "loss": 0.002, "step": 56480 }, { "grad_norm": 0.0565531887114048, "learning_rate": 4.34234211462305e-05, "loss": 0.0023, "step": 56490 }, { "grad_norm": 0.07630365341901779, "learning_rate": 4.340703046172213e-05, "loss": 0.0028, "step": 56500 }, { "grad_norm": 0.09403761476278305, "learning_rate": 4.339064049821097e-05, "loss": 0.0025, "step": 56510 }, { "grad_norm": 0.07081124931573868, "learning_rate": 4.3374251257489446e-05, "loss": 0.0026, "step": 56520 }, { "grad_norm": 0.06151386722922325, "learning_rate": 4.335786274134984e-05, "loss": 0.0023, "step": 56530 }, { "grad_norm": 0.041461143642663956, "learning_rate": 4.334147495158438e-05, "loss": 0.0027, "step": 56540 }, { "grad_norm": 0.04373079538345337, "learning_rate": 4.33250878899852e-05, "loss": 0.0023, "step": 56550 }, { "grad_norm": 0.06084985285997391, "learning_rate": 4.330870155834437e-05, "loss": 0.0022, "step": 56560 }, { "grad_norm": 0.08457693457603455, "learning_rate": 4.3292315958453884e-05, "loss": 0.0025, "step": 56570 }, { "grad_norm": 0.06871353834867477, "learning_rate": 4.3275931092105646e-05, "loss": 0.0023, "step": 56580 }, { "grad_norm": 0.07299335300922394, "learning_rate": 4.325954696109147e-05, "loss": 0.0028, "step": 56590 }, { "grad_norm": 0.059326719492673874, "learning_rate": 4.32431635672031e-05, "loss": 0.002, "step": 56600 }, { "grad_norm": 0.0644371286034584, "learning_rate": 4.3226780912232214e-05, "loss": 0.0024, "step": 56610 }, { "grad_norm": 0.09912730008363724, "learning_rate": 4.321039899797036e-05, "loss": 0.0034, "step": 56620 }, { "grad_norm": 0.06534526497125626, "learning_rate": 4.319401782620908e-05, "loss": 0.0025, "step": 56630 }, { "grad_norm": 0.06160586699843407, "learning_rate": 4.317763739873978e-05, "loss": 0.0029, "step": 56640 }, { "grad_norm": 0.06415468454360962, "learning_rate": 4.31612577173538e-05, "loss": 0.0032, "step": 56650 }, { "grad_norm": 0.08760783821344376, "learning_rate": 4.314487878384239e-05, "loss": 0.0034, "step": 56660 }, { "grad_norm": 0.11556316167116165, "learning_rate": 4.3128500599996736e-05, "loss": 0.0031, "step": 56670 }, { "grad_norm": 0.12930074334144592, "learning_rate": 4.31121231676079e-05, "loss": 0.0045, "step": 56680 }, { "grad_norm": 0.10179697722196579, "learning_rate": 4.309574648846694e-05, "loss": 0.0057, "step": 56690 }, { "grad_norm": 0.09733936190605164, "learning_rate": 4.3079370564364755e-05, "loss": 0.0046, "step": 56700 }, { "grad_norm": 0.08321452140808105, "learning_rate": 4.306299539709221e-05, "loss": 0.0036, "step": 56710 }, { "grad_norm": 0.104085274040699, "learning_rate": 4.3046620988440046e-05, "loss": 0.0031, "step": 56720 }, { "grad_norm": 0.07589141279459, "learning_rate": 4.303024734019897e-05, "loss": 0.0028, "step": 56730 }, { "grad_norm": 0.06012149900197983, "learning_rate": 4.3013874454159565e-05, "loss": 0.0027, "step": 56740 }, { "grad_norm": 0.07960695028305054, "learning_rate": 4.299750233211233e-05, "loss": 0.0029, "step": 56750 }, { "grad_norm": 0.07071243226528168, "learning_rate": 4.2981130975847715e-05, "loss": 0.0027, "step": 56760 }, { "grad_norm": 0.045872628688812256, "learning_rate": 4.2964760387156076e-05, "loss": 0.0026, "step": 56770 }, { "grad_norm": 0.04568279907107353, "learning_rate": 4.294839056782767e-05, "loss": 0.0025, "step": 56780 }, { "grad_norm": 0.043068770319223404, "learning_rate": 4.293202151965266e-05, "loss": 0.0025, "step": 56790 }, { "grad_norm": 0.05028417706489563, "learning_rate": 4.291565324442115e-05, "loss": 0.0024, "step": 56800 }, { "grad_norm": 0.05811735615134239, "learning_rate": 4.2899285743923135e-05, "loss": 0.0022, "step": 56810 }, { "grad_norm": 0.06095191463828087, "learning_rate": 4.288291901994857e-05, "loss": 0.002, "step": 56820 }, { "grad_norm": 0.056267932057380676, "learning_rate": 4.286655307428729e-05, "loss": 0.0022, "step": 56830 }, { "grad_norm": 0.07492539286613464, "learning_rate": 4.285018790872903e-05, "loss": 0.0034, "step": 56840 }, { "grad_norm": 0.05701431632041931, "learning_rate": 4.283382352506347e-05, "loss": 0.0024, "step": 56850 }, { "grad_norm": 0.05504627153277397, "learning_rate": 4.2817459925080205e-05, "loss": 0.0023, "step": 56860 }, { "grad_norm": 0.06555266678333282, "learning_rate": 4.28010971105687e-05, "loss": 0.0025, "step": 56870 }, { "grad_norm": 0.068882517516613, "learning_rate": 4.278473508331841e-05, "loss": 0.0032, "step": 56880 }, { "grad_norm": 0.091828353703022, "learning_rate": 4.276837384511864e-05, "loss": 0.0025, "step": 56890 }, { "grad_norm": 0.05864270031452179, "learning_rate": 4.275201339775864e-05, "loss": 0.0024, "step": 56900 }, { "grad_norm": 0.06483113765716553, "learning_rate": 4.273565374302756e-05, "loss": 0.0025, "step": 56910 }, { "grad_norm": 0.07963812351226807, "learning_rate": 4.2719294882714454e-05, "loss": 0.0035, "step": 56920 }, { "grad_norm": 0.15392933785915375, "learning_rate": 4.270293681860831e-05, "loss": 0.0028, "step": 56930 }, { "grad_norm": 0.08350064605474472, "learning_rate": 4.268657955249802e-05, "loss": 0.0029, "step": 56940 }, { "grad_norm": 0.08218944817781448, "learning_rate": 4.267022308617241e-05, "loss": 0.0028, "step": 56950 }, { "grad_norm": 0.05864536389708519, "learning_rate": 4.265386742142018e-05, "loss": 0.0022, "step": 56960 }, { "grad_norm": 0.0774255320429802, "learning_rate": 4.263751256002996e-05, "loss": 0.003, "step": 56970 }, { "grad_norm": 0.06026284396648407, "learning_rate": 4.26211585037903e-05, "loss": 0.0024, "step": 56980 }, { "grad_norm": 0.064162977039814, "learning_rate": 4.2604805254489656e-05, "loss": 0.0027, "step": 56990 }, { "grad_norm": 0.044398508965969086, "learning_rate": 4.258845281391638e-05, "loss": 0.0026, "step": 57000 }, { "grad_norm": 0.06101352721452713, "learning_rate": 4.257210118385876e-05, "loss": 0.0023, "step": 57010 }, { "grad_norm": 0.07374837249517441, "learning_rate": 4.2555750366104994e-05, "loss": 0.0022, "step": 57020 }, { "grad_norm": 0.05447318032383919, "learning_rate": 4.2539400362443184e-05, "loss": 0.002, "step": 57030 }, { "grad_norm": 0.0708872452378273, "learning_rate": 4.252305117466133e-05, "loss": 0.0032, "step": 57040 }, { "grad_norm": 0.06109749153256416, "learning_rate": 4.2506702804547356e-05, "loss": 0.0021, "step": 57050 }, { "grad_norm": 0.05136625096201897, "learning_rate": 4.249035525388907e-05, "loss": 0.0021, "step": 57060 }, { "grad_norm": 0.10910692811012268, "learning_rate": 4.247400852447427e-05, "loss": 0.0032, "step": 57070 }, { "grad_norm": 0.08521313220262527, "learning_rate": 4.245766261809059e-05, "loss": 0.0028, "step": 57080 }, { "grad_norm": 0.07884158939123154, "learning_rate": 4.244131753652558e-05, "loss": 0.0029, "step": 57090 }, { "grad_norm": 0.11164786666631699, "learning_rate": 4.242497328156672e-05, "loss": 0.0028, "step": 57100 }, { "grad_norm": 0.07345603406429291, "learning_rate": 4.240862985500139e-05, "loss": 0.0027, "step": 57110 }, { "grad_norm": 0.0803685337305069, "learning_rate": 4.2392287258616894e-05, "loss": 0.0037, "step": 57120 }, { "grad_norm": 0.050544481724500656, "learning_rate": 4.2375945494200396e-05, "loss": 0.0029, "step": 57130 }, { "grad_norm": 0.05879935249686241, "learning_rate": 4.2359604563539054e-05, "loss": 0.0026, "step": 57140 }, { "grad_norm": 0.06647833436727524, "learning_rate": 4.2343264468419885e-05, "loss": 0.0027, "step": 57150 }, { "grad_norm": 0.10514767467975616, "learning_rate": 4.232692521062979e-05, "loss": 0.0039, "step": 57160 }, { "grad_norm": 0.103117935359478, "learning_rate": 4.231058679195562e-05, "loss": 0.0037, "step": 57170 }, { "grad_norm": 0.08439509570598602, "learning_rate": 4.229424921418411e-05, "loss": 0.0026, "step": 57180 }, { "grad_norm": 0.07820860296487808, "learning_rate": 4.2277912479101916e-05, "loss": 0.0026, "step": 57190 }, { "grad_norm": 0.09204493463039398, "learning_rate": 4.226157658849562e-05, "loss": 0.0029, "step": 57200 }, { "grad_norm": 0.12101633846759796, "learning_rate": 4.2245241544151664e-05, "loss": 0.0027, "step": 57210 }, { "grad_norm": 0.08350533246994019, "learning_rate": 4.2228907347856425e-05, "loss": 0.0024, "step": 57220 }, { "grad_norm": 0.07774857431650162, "learning_rate": 4.2212574001396214e-05, "loss": 0.0039, "step": 57230 }, { "grad_norm": 0.06989547610282898, "learning_rate": 4.21962415065572e-05, "loss": 0.0027, "step": 57240 }, { "grad_norm": 0.052518248558044434, "learning_rate": 4.217990986512548e-05, "loss": 0.0028, "step": 57250 }, { "grad_norm": 0.0699070394039154, "learning_rate": 4.216357907888704e-05, "loss": 0.0031, "step": 57260 }, { "grad_norm": 0.06601528078317642, "learning_rate": 4.2147249149627824e-05, "loss": 0.0028, "step": 57270 }, { "grad_norm": 0.09501606225967407, "learning_rate": 4.213092007913364e-05, "loss": 0.0032, "step": 57280 }, { "grad_norm": 0.0659269392490387, "learning_rate": 4.2114591869190205e-05, "loss": 0.0026, "step": 57290 }, { "grad_norm": 0.04720544442534447, "learning_rate": 4.209826452158315e-05, "loss": 0.0024, "step": 57300 }, { "grad_norm": 0.05066750943660736, "learning_rate": 4.2081938038098e-05, "loss": 0.003, "step": 57310 }, { "grad_norm": 0.07257035374641418, "learning_rate": 4.2065612420520185e-05, "loss": 0.0024, "step": 57320 }, { "grad_norm": 0.06107319891452789, "learning_rate": 4.204928767063509e-05, "loss": 0.0023, "step": 57330 }, { "grad_norm": 0.07492437213659286, "learning_rate": 4.2032963790227944e-05, "loss": 0.0025, "step": 57340 }, { "grad_norm": 0.07153021544218063, "learning_rate": 4.2016640781083887e-05, "loss": 0.0024, "step": 57350 }, { "grad_norm": 0.07761240005493164, "learning_rate": 4.2000318644988004e-05, "loss": 0.0026, "step": 57360 }, { "grad_norm": 0.13711443543434143, "learning_rate": 4.198399738372524e-05, "loss": 0.0037, "step": 57370 }, { "grad_norm": 0.0734860822558403, "learning_rate": 4.196767699908045e-05, "loss": 0.0028, "step": 57380 }, { "grad_norm": 0.05180356651544571, "learning_rate": 4.195135749283845e-05, "loss": 0.0026, "step": 57390 }, { "grad_norm": 0.06342000514268875, "learning_rate": 4.1935038866783894e-05, "loss": 0.0027, "step": 57400 }, { "grad_norm": 0.09468791633844376, "learning_rate": 4.1918721122701355e-05, "loss": 0.0027, "step": 57410 }, { "grad_norm": 0.05851779133081436, "learning_rate": 4.190240426237533e-05, "loss": 0.0025, "step": 57420 }, { "grad_norm": 0.04088125005364418, "learning_rate": 4.188608828759019e-05, "loss": 0.003, "step": 57430 }, { "grad_norm": 0.07214067876338959, "learning_rate": 4.1869773200130246e-05, "loss": 0.0025, "step": 57440 }, { "grad_norm": 0.046156253665685654, "learning_rate": 4.185345900177967e-05, "loss": 0.0036, "step": 57450 }, { "grad_norm": 0.10466007143259048, "learning_rate": 4.183714569432258e-05, "loss": 0.0026, "step": 57460 }, { "grad_norm": 0.13662303984165192, "learning_rate": 4.182083327954295e-05, "loss": 0.0029, "step": 57470 }, { "grad_norm": 0.18243445456027985, "learning_rate": 4.1804521759224714e-05, "loss": 0.0043, "step": 57480 }, { "grad_norm": 0.06386145204305649, "learning_rate": 4.178821113515165e-05, "loss": 0.0038, "step": 57490 }, { "grad_norm": 0.07166944444179535, "learning_rate": 4.1771901409107476e-05, "loss": 0.0026, "step": 57500 }, { "grad_norm": 0.09202004969120026, "learning_rate": 4.175559258287577e-05, "loss": 0.0026, "step": 57510 }, { "grad_norm": 0.04745349660515785, "learning_rate": 4.173928465824008e-05, "loss": 0.0022, "step": 57520 }, { "grad_norm": 0.07744729518890381, "learning_rate": 4.17229776369838e-05, "loss": 0.0025, "step": 57530 }, { "grad_norm": 0.05178070813417435, "learning_rate": 4.170667152089025e-05, "loss": 0.0025, "step": 57540 }, { "grad_norm": 0.06889332830905914, "learning_rate": 4.1690366311742626e-05, "loss": 0.0023, "step": 57550 }, { "grad_norm": 0.07121540606021881, "learning_rate": 4.1674062011324055e-05, "loss": 0.0027, "step": 57560 }, { "grad_norm": 0.06264859437942505, "learning_rate": 4.165775862141752e-05, "loss": 0.0025, "step": 57570 }, { "grad_norm": 0.0658370703458786, "learning_rate": 4.164145614380598e-05, "loss": 0.0022, "step": 57580 }, { "grad_norm": 0.06670516729354858, "learning_rate": 4.162515458027223e-05, "loss": 0.0029, "step": 57590 }, { "grad_norm": 0.06010546162724495, "learning_rate": 4.160885393259899e-05, "loss": 0.0021, "step": 57600 }, { "grad_norm": 0.049763135612010956, "learning_rate": 4.159255420256886e-05, "loss": 0.0023, "step": 57610 }, { "grad_norm": 0.08956656605005264, "learning_rate": 4.157625539196436e-05, "loss": 0.0026, "step": 57620 }, { "grad_norm": 0.06569770723581314, "learning_rate": 4.155995750256791e-05, "loss": 0.0022, "step": 57630 }, { "grad_norm": 0.06104138121008873, "learning_rate": 4.15436605361618e-05, "loss": 0.0021, "step": 57640 }, { "grad_norm": 0.05875221639871597, "learning_rate": 4.152736449452827e-05, "loss": 0.0021, "step": 57650 }, { "grad_norm": 0.05642416700720787, "learning_rate": 4.151106937944943e-05, "loss": 0.0021, "step": 57660 }, { "grad_norm": 0.058787960559129715, "learning_rate": 4.149477519270727e-05, "loss": 0.0029, "step": 57670 }, { "grad_norm": 0.062300905585289, "learning_rate": 4.147848193608371e-05, "loss": 0.0027, "step": 57680 }, { "grad_norm": 0.060083888471126556, "learning_rate": 4.146218961136056e-05, "loss": 0.0018, "step": 57690 }, { "grad_norm": 0.08152464777231216, "learning_rate": 4.14458982203195e-05, "loss": 0.0035, "step": 57700 }, { "grad_norm": 0.07382933795452118, "learning_rate": 4.1429607764742165e-05, "loss": 0.0023, "step": 57710 }, { "grad_norm": 0.060589488595724106, "learning_rate": 4.141331824641003e-05, "loss": 0.0021, "step": 57720 }, { "grad_norm": 0.04012792930006981, "learning_rate": 4.139702966710452e-05, "loss": 0.0024, "step": 57730 }, { "grad_norm": 0.05938807502388954, "learning_rate": 4.13807420286069e-05, "loss": 0.0023, "step": 57740 }, { "grad_norm": 0.05922430008649826, "learning_rate": 4.136445533269837e-05, "loss": 0.0023, "step": 57750 }, { "grad_norm": 0.08309783041477203, "learning_rate": 4.134816958116003e-05, "loss": 0.0024, "step": 57760 }, { "grad_norm": 0.0612071231007576, "learning_rate": 4.133188477577283e-05, "loss": 0.0022, "step": 57770 }, { "grad_norm": 0.07118908315896988, "learning_rate": 4.131560091831771e-05, "loss": 0.0026, "step": 57780 }, { "grad_norm": 0.053318679332733154, "learning_rate": 4.1299318010575405e-05, "loss": 0.0024, "step": 57790 }, { "grad_norm": 0.08486675471067429, "learning_rate": 4.128303605432661e-05, "loss": 0.0028, "step": 57800 }, { "grad_norm": 0.05498352646827698, "learning_rate": 4.126675505135188e-05, "loss": 0.0024, "step": 57810 }, { "grad_norm": 0.042426325380802155, "learning_rate": 4.125047500343169e-05, "loss": 0.0024, "step": 57820 }, { "grad_norm": 0.07314690202474594, "learning_rate": 4.1234195912346376e-05, "loss": 0.0021, "step": 57830 }, { "grad_norm": 0.07611899822950363, "learning_rate": 4.1217917779876235e-05, "loss": 0.0023, "step": 57840 }, { "grad_norm": 0.07563576847314835, "learning_rate": 4.120164060780141e-05, "loss": 0.0037, "step": 57850 }, { "grad_norm": 0.060297634452581406, "learning_rate": 4.1185364397901935e-05, "loss": 0.0023, "step": 57860 }, { "grad_norm": 0.05671457201242447, "learning_rate": 4.116908915195776e-05, "loss": 0.002, "step": 57870 }, { "grad_norm": 0.053592752665281296, "learning_rate": 4.115281487174872e-05, "loss": 0.0019, "step": 57880 }, { "grad_norm": 0.06511799246072769, "learning_rate": 4.113654155905452e-05, "loss": 0.0025, "step": 57890 }, { "grad_norm": 0.08458591252565384, "learning_rate": 4.112026921565483e-05, "loss": 0.0023, "step": 57900 }, { "grad_norm": 0.04915659874677658, "learning_rate": 4.1103997843329147e-05, "loss": 0.0022, "step": 57910 }, { "grad_norm": 0.1162513867020607, "learning_rate": 4.108772744385689e-05, "loss": 0.0024, "step": 57920 }, { "grad_norm": 0.08549916744232178, "learning_rate": 4.1071458019017346e-05, "loss": 0.0027, "step": 57930 }, { "grad_norm": 0.10743214935064316, "learning_rate": 4.1055189570589734e-05, "loss": 0.0037, "step": 57940 }, { "grad_norm": 0.08245055377483368, "learning_rate": 4.103892210035315e-05, "loss": 0.0033, "step": 57950 }, { "grad_norm": 0.0668279230594635, "learning_rate": 4.102265561008656e-05, "loss": 0.0024, "step": 57960 }, { "grad_norm": 0.10108613967895508, "learning_rate": 4.100639010156885e-05, "loss": 0.0027, "step": 57970 }, { "grad_norm": 0.07365906238555908, "learning_rate": 4.099012557657882e-05, "loss": 0.0027, "step": 57980 }, { "grad_norm": 0.06359998881816864, "learning_rate": 4.09738620368951e-05, "loss": 0.0027, "step": 57990 }, { "grad_norm": 0.10543425381183624, "learning_rate": 4.095759948429626e-05, "loss": 0.0024, "step": 58000 }, { "grad_norm": 0.05247393250465393, "learning_rate": 4.094133792056075e-05, "loss": 0.0022, "step": 58010 }, { "grad_norm": 0.05208270996809006, "learning_rate": 4.092507734746688e-05, "loss": 0.0021, "step": 58020 }, { "grad_norm": 0.05074447765946388, "learning_rate": 4.090881776679293e-05, "loss": 0.002, "step": 58030 }, { "grad_norm": 0.08566594868898392, "learning_rate": 4.0892559180316994e-05, "loss": 0.0023, "step": 58040 }, { "grad_norm": 0.06553100049495697, "learning_rate": 4.08763015898171e-05, "loss": 0.0027, "step": 58050 }, { "grad_norm": 0.07153583317995071, "learning_rate": 4.0860044997071145e-05, "loss": 0.0024, "step": 58060 }, { "grad_norm": 0.058925382792949677, "learning_rate": 4.084378940385693e-05, "loss": 0.0028, "step": 58070 }, { "grad_norm": 0.0666363462805748, "learning_rate": 4.082753481195211e-05, "loss": 0.0026, "step": 58080 }, { "grad_norm": 0.05818450450897217, "learning_rate": 4.0811281223134324e-05, "loss": 0.0027, "step": 58090 }, { "grad_norm": 0.06120657920837402, "learning_rate": 4.0795028639181e-05, "loss": 0.0042, "step": 58100 }, { "grad_norm": 0.08873745799064636, "learning_rate": 4.077877706186951e-05, "loss": 0.0022, "step": 58110 }, { "grad_norm": 0.07793871313333511, "learning_rate": 4.076252649297709e-05, "loss": 0.0026, "step": 58120 }, { "grad_norm": 0.06729321926832199, "learning_rate": 4.07462769342809e-05, "loss": 0.0024, "step": 58130 }, { "grad_norm": 0.08286190778017044, "learning_rate": 4.073002838755793e-05, "loss": 0.0029, "step": 58140 }, { "grad_norm": 0.054976578801870346, "learning_rate": 4.0713780854585115e-05, "loss": 0.0023, "step": 58150 }, { "grad_norm": 0.06928917020559311, "learning_rate": 4.069753433713927e-05, "loss": 0.0037, "step": 58160 }, { "grad_norm": 0.09609492868185043, "learning_rate": 4.068128883699709e-05, "loss": 0.0033, "step": 58170 }, { "grad_norm": 0.08774612843990326, "learning_rate": 4.0665044355935155e-05, "loss": 0.0033, "step": 58180 }, { "grad_norm": 0.08346191793680191, "learning_rate": 4.064880089572993e-05, "loss": 0.0035, "step": 58190 }, { "grad_norm": 0.0845891535282135, "learning_rate": 4.063255845815779e-05, "loss": 0.0039, "step": 58200 }, { "grad_norm": 0.08930613100528717, "learning_rate": 4.0616317044994945e-05, "loss": 0.0034, "step": 58210 }, { "grad_norm": 0.0894082561135292, "learning_rate": 4.0600076658017585e-05, "loss": 0.003, "step": 58220 }, { "grad_norm": 0.08871397376060486, "learning_rate": 4.0583837299001706e-05, "loss": 0.0026, "step": 58230 }, { "grad_norm": 0.07002442330121994, "learning_rate": 4.056759896972322e-05, "loss": 0.0026, "step": 58240 }, { "grad_norm": 0.06880168616771698, "learning_rate": 4.055136167195793e-05, "loss": 0.0036, "step": 58250 }, { "grad_norm": 0.08255381882190704, "learning_rate": 4.053512540748152e-05, "loss": 0.0022, "step": 58260 }, { "grad_norm": 0.06209350377321243, "learning_rate": 4.051889017806955e-05, "loss": 0.002, "step": 58270 }, { "grad_norm": 0.04771944135427475, "learning_rate": 4.0502655985497516e-05, "loss": 0.0024, "step": 58280 }, { "grad_norm": 0.06839237362146378, "learning_rate": 4.048642283154074e-05, "loss": 0.0026, "step": 58290 }, { "grad_norm": 0.09134473651647568, "learning_rate": 4.0470190717974464e-05, "loss": 0.0035, "step": 58300 }, { "grad_norm": 0.064432792365551, "learning_rate": 4.045395964657381e-05, "loss": 0.0027, "step": 58310 }, { "grad_norm": 0.05446677282452583, "learning_rate": 4.043772961911376e-05, "loss": 0.0022, "step": 58320 }, { "grad_norm": 0.07441777735948563, "learning_rate": 4.042150063736924e-05, "loss": 0.0024, "step": 58330 }, { "grad_norm": 0.03994143754243851, "learning_rate": 4.0405272703114976e-05, "loss": 0.0021, "step": 58340 }, { "grad_norm": 0.03922301158308983, "learning_rate": 4.038904581812569e-05, "loss": 0.0024, "step": 58350 }, { "grad_norm": 0.038477323949337006, "learning_rate": 4.03728199841759e-05, "loss": 0.0022, "step": 58360 }, { "grad_norm": 0.0445093996822834, "learning_rate": 4.0356595203040044e-05, "loss": 0.0019, "step": 58370 }, { "grad_norm": 0.060349442064762115, "learning_rate": 4.0340371476492426e-05, "loss": 0.0019, "step": 58380 }, { "grad_norm": 0.058919016271829605, "learning_rate": 4.032414880630727e-05, "loss": 0.0019, "step": 58390 }, { "grad_norm": 0.08108863979578018, "learning_rate": 4.0307927194258646e-05, "loss": 0.0028, "step": 58400 }, { "grad_norm": 0.09969508647918701, "learning_rate": 4.029170664212054e-05, "loss": 0.0027, "step": 58410 }, { "grad_norm": 0.061643894761800766, "learning_rate": 4.0275487151666794e-05, "loss": 0.0022, "step": 58420 }, { "grad_norm": 0.09470382332801819, "learning_rate": 4.025926872467116e-05, "loss": 0.0026, "step": 58430 }, { "grad_norm": 0.10746823251247406, "learning_rate": 4.024305136290725e-05, "loss": 0.0029, "step": 58440 }, { "grad_norm": 0.05864568427205086, "learning_rate": 4.022683506814858e-05, "loss": 0.0026, "step": 58450 }, { "grad_norm": 0.0772448480129242, "learning_rate": 4.021061984216853e-05, "loss": 0.0025, "step": 58460 }, { "grad_norm": 0.04908899590373039, "learning_rate": 4.019440568674035e-05, "loss": 0.0023, "step": 58470 }, { "grad_norm": 0.05152135714888573, "learning_rate": 4.017819260363724e-05, "loss": 0.0036, "step": 58480 }, { "grad_norm": 0.0545620322227478, "learning_rate": 4.016198059463223e-05, "loss": 0.0021, "step": 58490 }, { "grad_norm": 0.054430726915597916, "learning_rate": 4.0145769661498233e-05, "loss": 0.0022, "step": 58500 }, { "grad_norm": 0.06981852650642395, "learning_rate": 4.012955980600804e-05, "loss": 0.0022, "step": 58510 }, { "grad_norm": 0.061604633927345276, "learning_rate": 4.0113351029934355e-05, "loss": 0.0025, "step": 58520 }, { "grad_norm": 0.07997238636016846, "learning_rate": 4.0097143335049714e-05, "loss": 0.0026, "step": 58530 }, { "grad_norm": 0.05535992980003357, "learning_rate": 4.0080936723126606e-05, "loss": 0.0039, "step": 58540 }, { "grad_norm": 0.07407084852457047, "learning_rate": 4.006473119593735e-05, "loss": 0.003, "step": 58550 }, { "grad_norm": 0.048545319586992264, "learning_rate": 4.004852675525415e-05, "loss": 0.0025, "step": 58560 }, { "grad_norm": 0.051814667880535126, "learning_rate": 4.00323234028491e-05, "loss": 0.0034, "step": 58570 }, { "grad_norm": 0.060714609920978546, "learning_rate": 4.0016121140494186e-05, "loss": 0.0025, "step": 58580 }, { "grad_norm": 0.09275231510400772, "learning_rate": 3.999991996996122e-05, "loss": 0.0024, "step": 58590 }, { "grad_norm": 0.08398756384849548, "learning_rate": 3.9983719893022e-05, "loss": 0.0025, "step": 58600 }, { "grad_norm": 0.060640547424554825, "learning_rate": 3.996752091144812e-05, "loss": 0.0026, "step": 58610 }, { "grad_norm": 0.05999346449971199, "learning_rate": 3.9951323027011055e-05, "loss": 0.0026, "step": 58620 }, { "grad_norm": 0.06696518510580063, "learning_rate": 3.993512624148219e-05, "loss": 0.0025, "step": 58630 }, { "grad_norm": 0.08178574591875076, "learning_rate": 3.99189305566328e-05, "loss": 0.0019, "step": 58640 }, { "grad_norm": 0.06730809062719345, "learning_rate": 3.990273597423401e-05, "loss": 0.002, "step": 58650 }, { "grad_norm": 0.06283655762672424, "learning_rate": 3.988654249605681e-05, "loss": 0.0023, "step": 58660 }, { "grad_norm": 0.058380115777254105, "learning_rate": 3.987035012387213e-05, "loss": 0.0025, "step": 58670 }, { "grad_norm": 0.08642380684614182, "learning_rate": 3.9854158859450735e-05, "loss": 0.0028, "step": 58680 }, { "grad_norm": 0.08697709441184998, "learning_rate": 3.983796870456328e-05, "loss": 0.0025, "step": 58690 }, { "grad_norm": 0.054107893258333206, "learning_rate": 3.982177966098028e-05, "loss": 0.0025, "step": 58700 }, { "grad_norm": 0.08285339176654816, "learning_rate": 3.980559173047216e-05, "loss": 0.0025, "step": 58710 }, { "grad_norm": 0.0829588770866394, "learning_rate": 3.9789404914809176e-05, "loss": 0.0029, "step": 58720 }, { "grad_norm": 0.05135613679885864, "learning_rate": 3.9773219215761544e-05, "loss": 0.0021, "step": 58730 }, { "grad_norm": 0.08342769742012024, "learning_rate": 3.975703463509928e-05, "loss": 0.0027, "step": 58740 }, { "grad_norm": 0.060240697115659714, "learning_rate": 3.974085117459231e-05, "loss": 0.0022, "step": 58750 }, { "grad_norm": 0.07689762115478516, "learning_rate": 3.972466883601043e-05, "loss": 0.0027, "step": 58760 }, { "grad_norm": 0.0577738992869854, "learning_rate": 3.9708487621123316e-05, "loss": 0.0037, "step": 58770 }, { "grad_norm": 0.07113245874643326, "learning_rate": 3.969230753170051e-05, "loss": 0.003, "step": 58780 }, { "grad_norm": 0.077623650431633, "learning_rate": 3.967612856951146e-05, "loss": 0.0032, "step": 58790 }, { "grad_norm": 0.0711410716176033, "learning_rate": 3.9659950736325476e-05, "loss": 0.0033, "step": 58800 }, { "grad_norm": 0.0900111049413681, "learning_rate": 3.964377403391174e-05, "loss": 0.0035, "step": 58810 }, { "grad_norm": 0.060756292194128036, "learning_rate": 3.9627598464039303e-05, "loss": 0.0023, "step": 58820 }, { "grad_norm": 0.0641852393746376, "learning_rate": 3.9611424028477096e-05, "loss": 0.0025, "step": 58830 }, { "grad_norm": 0.06369753181934357, "learning_rate": 3.959525072899394e-05, "loss": 0.0029, "step": 58840 }, { "grad_norm": 0.10651610791683197, "learning_rate": 3.9579078567358506e-05, "loss": 0.0028, "step": 58850 }, { "grad_norm": 0.05955846980214119, "learning_rate": 3.956290754533939e-05, "loss": 0.002, "step": 58860 }, { "grad_norm": 0.07722210139036179, "learning_rate": 3.954673766470501e-05, "loss": 0.0021, "step": 58870 }, { "grad_norm": 0.07982918620109558, "learning_rate": 3.9530568927223674e-05, "loss": 0.0033, "step": 58880 }, { "grad_norm": 0.06948242336511612, "learning_rate": 3.95144013346636e-05, "loss": 0.0027, "step": 58890 }, { "grad_norm": 0.06178635358810425, "learning_rate": 3.949823488879282e-05, "loss": 0.0021, "step": 58900 }, { "grad_norm": 0.0676526427268982, "learning_rate": 3.9482069591379276e-05, "loss": 0.0026, "step": 58910 }, { "grad_norm": 0.05633757263422012, "learning_rate": 3.9465905444190784e-05, "loss": 0.0024, "step": 58920 }, { "grad_norm": 0.0710851177573204, "learning_rate": 3.944974244899505e-05, "loss": 0.0031, "step": 58930 }, { "grad_norm": 0.08235789090394974, "learning_rate": 3.9433580607559616e-05, "loss": 0.0027, "step": 58940 }, { "grad_norm": 0.07192534953355789, "learning_rate": 3.941741992165193e-05, "loss": 0.0027, "step": 58950 }, { "grad_norm": 0.06313076615333557, "learning_rate": 3.940126039303928e-05, "loss": 0.002, "step": 58960 }, { "grad_norm": 0.07577812671661377, "learning_rate": 3.938510202348886e-05, "loss": 0.0024, "step": 58970 }, { "grad_norm": 0.06674184650182724, "learning_rate": 3.93689448147677e-05, "loss": 0.0023, "step": 58980 }, { "grad_norm": 0.04603883996605873, "learning_rate": 3.935278876864278e-05, "loss": 0.0022, "step": 58990 }, { "grad_norm": 0.047467537224292755, "learning_rate": 3.933663388688087e-05, "loss": 0.0029, "step": 59000 }, { "grad_norm": 0.054178789258003235, "learning_rate": 3.9320480171248645e-05, "loss": 0.0032, "step": 59010 }, { "grad_norm": 0.07923250645399094, "learning_rate": 3.930432762351265e-05, "loss": 0.0024, "step": 59020 }, { "grad_norm": 0.051514580845832825, "learning_rate": 3.928817624543931e-05, "loss": 0.0025, "step": 59030 }, { "grad_norm": 0.0635070949792862, "learning_rate": 3.9272026038794885e-05, "loss": 0.003, "step": 59040 }, { "grad_norm": 0.06146632134914398, "learning_rate": 3.9255877005345586e-05, "loss": 0.0024, "step": 59050 }, { "grad_norm": 0.059932783246040344, "learning_rate": 3.923972914685743e-05, "loss": 0.003, "step": 59060 }, { "grad_norm": 0.05856781825423241, "learning_rate": 3.922358246509631e-05, "loss": 0.0022, "step": 59070 }, { "grad_norm": 0.07846762239933014, "learning_rate": 3.920743696182801e-05, "loss": 0.003, "step": 59080 }, { "grad_norm": 0.051517385989427567, "learning_rate": 3.9191292638818164e-05, "loss": 0.0021, "step": 59090 }, { "grad_norm": 0.0502135343849659, "learning_rate": 3.9175149497832316e-05, "loss": 0.0015, "step": 59100 }, { "grad_norm": 0.07095769792795181, "learning_rate": 3.915900754063584e-05, "loss": 0.0024, "step": 59110 }, { "grad_norm": 0.06807652860879898, "learning_rate": 3.9142866768994014e-05, "loss": 0.0018, "step": 59120 }, { "grad_norm": 0.06593821942806244, "learning_rate": 3.912672718467193e-05, "loss": 0.0023, "step": 59130 }, { "grad_norm": 0.05917998403310776, "learning_rate": 3.911058878943463e-05, "loss": 0.0024, "step": 59140 }, { "grad_norm": 0.0439717173576355, "learning_rate": 3.9094451585046974e-05, "loss": 0.0018, "step": 59150 }, { "grad_norm": 0.05289597436785698, "learning_rate": 3.90783155732737e-05, "loss": 0.002, "step": 59160 }, { "grad_norm": 0.06710442155599594, "learning_rate": 3.9062180755879373e-05, "loss": 0.0025, "step": 59170 }, { "grad_norm": 0.06876819580793381, "learning_rate": 3.904604713462855e-05, "loss": 0.002, "step": 59180 }, { "grad_norm": 0.055334530770778656, "learning_rate": 3.902991471128554e-05, "loss": 0.0022, "step": 59190 }, { "grad_norm": 0.09341456741094589, "learning_rate": 3.901378348761456e-05, "loss": 0.0044, "step": 59200 }, { "grad_norm": 0.07820086926221848, "learning_rate": 3.899765346537971e-05, "loss": 0.0026, "step": 59210 }, { "grad_norm": 0.04337037727236748, "learning_rate": 3.898152464634493e-05, "loss": 0.0022, "step": 59220 }, { "grad_norm": 0.057387348264455795, "learning_rate": 3.896539703227402e-05, "loss": 0.0019, "step": 59230 }, { "grad_norm": 0.04467136412858963, "learning_rate": 3.894927062493073e-05, "loss": 0.0019, "step": 59240 }, { "grad_norm": 0.058652110397815704, "learning_rate": 3.893314542607859e-05, "loss": 0.0025, "step": 59250 }, { "grad_norm": 0.08669039607048035, "learning_rate": 3.891702143748103e-05, "loss": 0.0022, "step": 59260 }, { "grad_norm": 0.07199177145957947, "learning_rate": 3.890089866090133e-05, "loss": 0.0036, "step": 59270 }, { "grad_norm": 0.0871616005897522, "learning_rate": 3.888477709810267e-05, "loss": 0.003, "step": 59280 }, { "grad_norm": 0.07574988901615143, "learning_rate": 3.886865675084806e-05, "loss": 0.0035, "step": 59290 }, { "grad_norm": 0.05294535681605339, "learning_rate": 3.885253762090043e-05, "loss": 0.0022, "step": 59300 }, { "grad_norm": 0.07521191239356995, "learning_rate": 3.8836419710022534e-05, "loss": 0.0028, "step": 59310 }, { "grad_norm": 0.06072647124528885, "learning_rate": 3.882030301997698e-05, "loss": 0.0021, "step": 59320 }, { "grad_norm": 0.09018409252166748, "learning_rate": 3.880418755252629e-05, "loss": 0.0028, "step": 59330 }, { "grad_norm": 0.042571015655994415, "learning_rate": 3.87880733094328e-05, "loss": 0.0033, "step": 59340 }, { "grad_norm": 0.14198224246501923, "learning_rate": 3.877196029245877e-05, "loss": 0.0025, "step": 59350 }, { "grad_norm": 0.07994364947080612, "learning_rate": 3.875584850336627e-05, "loss": 0.0029, "step": 59360 }, { "grad_norm": 0.0537368468940258, "learning_rate": 3.873973794391728e-05, "loss": 0.002, "step": 59370 }, { "grad_norm": 0.055197373032569885, "learning_rate": 3.872362861587361e-05, "loss": 0.0017, "step": 59380 }, { "grad_norm": 0.04498417675495148, "learning_rate": 3.870752052099698e-05, "loss": 0.0024, "step": 59390 }, { "grad_norm": 0.055525004863739014, "learning_rate": 3.8691413661048916e-05, "loss": 0.0028, "step": 59400 }, { "grad_norm": 0.05372828617691994, "learning_rate": 3.867530803779085e-05, "loss": 0.0028, "step": 59410 }, { "grad_norm": 0.08622337877750397, "learning_rate": 3.8659203652984054e-05, "loss": 0.0032, "step": 59420 }, { "grad_norm": 0.06185273081064224, "learning_rate": 3.864310050838972e-05, "loss": 0.0022, "step": 59430 }, { "grad_norm": 0.059355102479457855, "learning_rate": 3.8626998605768835e-05, "loss": 0.0031, "step": 59440 }, { "grad_norm": 0.07452896982431412, "learning_rate": 3.8610897946882294e-05, "loss": 0.0025, "step": 59450 }, { "grad_norm": 0.13005948066711426, "learning_rate": 3.859479853349082e-05, "loss": 0.0028, "step": 59460 }, { "grad_norm": 0.08762004971504211, "learning_rate": 3.857870036735503e-05, "loss": 0.0027, "step": 59470 }, { "grad_norm": 0.07708226144313812, "learning_rate": 3.856260345023538e-05, "loss": 0.0036, "step": 59480 }, { "grad_norm": 0.07920950651168823, "learning_rate": 3.854650778389224e-05, "loss": 0.0026, "step": 59490 }, { "grad_norm": 0.08834870159626007, "learning_rate": 3.8530413370085786e-05, "loss": 0.0024, "step": 59500 }, { "grad_norm": 0.06811227649450302, "learning_rate": 3.851432021057608e-05, "loss": 0.0022, "step": 59510 }, { "grad_norm": 0.07513132691383362, "learning_rate": 3.849822830712306e-05, "loss": 0.0036, "step": 59520 }, { "grad_norm": 0.08021728694438934, "learning_rate": 3.848213766148649e-05, "loss": 0.0029, "step": 59530 }, { "grad_norm": 0.05099665746092796, "learning_rate": 3.846604827542602e-05, "loss": 0.0025, "step": 59540 }, { "grad_norm": 0.07696180045604706, "learning_rate": 3.844996015070117e-05, "loss": 0.0032, "step": 59550 }, { "grad_norm": 0.04894891008734703, "learning_rate": 3.843387328907132e-05, "loss": 0.0027, "step": 59560 }, { "grad_norm": 0.07913735508918762, "learning_rate": 3.8417787692295696e-05, "loss": 0.0027, "step": 59570 }, { "grad_norm": 0.054421938955783844, "learning_rate": 3.84017033621334e-05, "loss": 0.0037, "step": 59580 }, { "grad_norm": 0.07404084503650665, "learning_rate": 3.838562030034338e-05, "loss": 0.0022, "step": 59590 }, { "grad_norm": 0.08785959333181381, "learning_rate": 3.836953850868447e-05, "loss": 0.0029, "step": 59600 }, { "grad_norm": 0.07154370844364166, "learning_rate": 3.835345798891533e-05, "loss": 0.0026, "step": 59610 }, { "grad_norm": 0.06462058424949646, "learning_rate": 3.833737874279452e-05, "loss": 0.0029, "step": 59620 }, { "grad_norm": 0.0549028217792511, "learning_rate": 3.8321300772080433e-05, "loss": 0.0022, "step": 59630 }, { "grad_norm": 0.050255950540304184, "learning_rate": 3.8305224078531345e-05, "loss": 0.002, "step": 59640 }, { "grad_norm": 0.11119520664215088, "learning_rate": 3.828914866390537e-05, "loss": 0.0027, "step": 59650 }, { "grad_norm": 0.0634584054350853, "learning_rate": 3.827307452996048e-05, "loss": 0.0025, "step": 59660 }, { "grad_norm": 0.06174875795841217, "learning_rate": 3.825700167845454e-05, "loss": 0.0024, "step": 59670 }, { "grad_norm": 0.05501591041684151, "learning_rate": 3.8240930111145214e-05, "loss": 0.0023, "step": 59680 }, { "grad_norm": 0.06336790323257446, "learning_rate": 3.8224859829790116e-05, "loss": 0.002, "step": 59690 }, { "grad_norm": 0.05523402988910675, "learning_rate": 3.820879083614664e-05, "loss": 0.0022, "step": 59700 }, { "grad_norm": 0.11479905992746353, "learning_rate": 3.819272313197208e-05, "loss": 0.0026, "step": 59710 }, { "grad_norm": 0.06687748432159424, "learning_rate": 3.817665671902355e-05, "loss": 0.002, "step": 59720 }, { "grad_norm": 0.055573564022779465, "learning_rate": 3.816059159905808e-05, "loss": 0.0026, "step": 59730 }, { "grad_norm": 0.08921847492456436, "learning_rate": 3.8144527773832476e-05, "loss": 0.0029, "step": 59740 }, { "grad_norm": 0.06242932751774788, "learning_rate": 3.812846524510352e-05, "loss": 0.002, "step": 59750 }, { "grad_norm": 0.07273570448160172, "learning_rate": 3.811240401462775e-05, "loss": 0.0024, "step": 59760 }, { "grad_norm": 0.07466933131217957, "learning_rate": 3.809634408416162e-05, "loss": 0.0035, "step": 59770 }, { "grad_norm": 0.08922406286001205, "learning_rate": 3.8080285455461394e-05, "loss": 0.0022, "step": 59780 }, { "grad_norm": 0.04302661865949631, "learning_rate": 3.806422813028323e-05, "loss": 0.0024, "step": 59790 }, { "grad_norm": 0.10536123067140579, "learning_rate": 3.8048172110383114e-05, "loss": 0.0025, "step": 59800 }, { "grad_norm": 0.08094122260808945, "learning_rate": 3.8032117397516944e-05, "loss": 0.0028, "step": 59810 }, { "grad_norm": 0.07399773597717285, "learning_rate": 3.801606399344042e-05, "loss": 0.0023, "step": 59820 }, { "grad_norm": 0.06872184574604034, "learning_rate": 3.800001189990913e-05, "loss": 0.0028, "step": 59830 }, { "grad_norm": 0.0569329559803009, "learning_rate": 3.798396111867847e-05, "loss": 0.0021, "step": 59840 }, { "grad_norm": 0.04594867676496506, "learning_rate": 3.796791165150378e-05, "loss": 0.0018, "step": 59850 }, { "grad_norm": 0.05587432533502579, "learning_rate": 3.795186350014018e-05, "loss": 0.0023, "step": 59860 }, { "grad_norm": 0.06940976530313492, "learning_rate": 3.793581666634266e-05, "loss": 0.0024, "step": 59870 }, { "grad_norm": 0.06690783053636551, "learning_rate": 3.791977115186609e-05, "loss": 0.0019, "step": 59880 }, { "grad_norm": 0.08483845740556717, "learning_rate": 3.79037269584652e-05, "loss": 0.0021, "step": 59890 }, { "grad_norm": 0.06475558131933212, "learning_rate": 3.7887684087894545e-05, "loss": 0.0023, "step": 59900 }, { "grad_norm": 0.06048021838068962, "learning_rate": 3.787164254190854e-05, "loss": 0.0025, "step": 59910 }, { "grad_norm": 0.07519466429948807, "learning_rate": 3.7855602322261473e-05, "loss": 0.0026, "step": 59920 }, { "grad_norm": 0.08347359299659729, "learning_rate": 3.783956343070746e-05, "loss": 0.002, "step": 59930 }, { "grad_norm": 0.07612063735723495, "learning_rate": 3.7823525869000526e-05, "loss": 0.0026, "step": 59940 }, { "grad_norm": 0.08730566501617432, "learning_rate": 3.78074896388945e-05, "loss": 0.0026, "step": 59950 }, { "grad_norm": 0.05350586026906967, "learning_rate": 3.7791454742143075e-05, "loss": 0.0021, "step": 59960 }, { "grad_norm": 0.07463400810956955, "learning_rate": 3.7775421180499795e-05, "loss": 0.003, "step": 59970 }, { "grad_norm": 0.07172086834907532, "learning_rate": 3.7759388955718086e-05, "loss": 0.0026, "step": 59980 }, { "grad_norm": 0.12326430529356003, "learning_rate": 3.774335806955117e-05, "loss": 0.0035, "step": 59990 }, { "grad_norm": 0.09958473592996597, "learning_rate": 3.7727328523752215e-05, "loss": 0.0031, "step": 60000 }, { "grad_norm": 0.09842871129512787, "learning_rate": 3.771130032007415e-05, "loss": 0.0039, "step": 60010 }, { "grad_norm": 0.05283636227250099, "learning_rate": 3.769527346026982e-05, "loss": 0.0021, "step": 60020 }, { "grad_norm": 0.050799816846847534, "learning_rate": 3.767924794609188e-05, "loss": 0.0026, "step": 60030 }, { "grad_norm": 0.045588426291942596, "learning_rate": 3.7663223779292854e-05, "loss": 0.002, "step": 60040 }, { "grad_norm": 0.03927299752831459, "learning_rate": 3.764720096162512e-05, "loss": 0.0021, "step": 60050 }, { "grad_norm": 0.10360720008611679, "learning_rate": 3.763117949484092e-05, "loss": 0.0023, "step": 60060 }, { "grad_norm": 0.06299979239702225, "learning_rate": 3.7615159380692344e-05, "loss": 0.0032, "step": 60070 }, { "grad_norm": 0.07057135552167892, "learning_rate": 3.759914062093132e-05, "loss": 0.0028, "step": 60080 }, { "grad_norm": 0.059071771800518036, "learning_rate": 3.7583123217309615e-05, "loss": 0.0026, "step": 60090 }, { "grad_norm": 0.061271924525499344, "learning_rate": 3.7567107171578904e-05, "loss": 0.0029, "step": 60100 }, { "grad_norm": 0.0758221372961998, "learning_rate": 3.755109248549066e-05, "loss": 0.0032, "step": 60110 }, { "grad_norm": 0.06225285306572914, "learning_rate": 3.75350791607962e-05, "loss": 0.0025, "step": 60120 }, { "grad_norm": 0.08950982242822647, "learning_rate": 3.751906719924676e-05, "loss": 0.0027, "step": 60130 }, { "grad_norm": 0.10324566811323166, "learning_rate": 3.750305660259337e-05, "loss": 0.0031, "step": 60140 }, { "grad_norm": 0.0782947763800621, "learning_rate": 3.748704737258693e-05, "loss": 0.0023, "step": 60150 }, { "grad_norm": 0.05595439672470093, "learning_rate": 3.747103951097816e-05, "loss": 0.0023, "step": 60160 }, { "grad_norm": 0.06211629509925842, "learning_rate": 3.745503301951768e-05, "loss": 0.0032, "step": 60170 }, { "grad_norm": 0.25053659081459045, "learning_rate": 3.743902789995592e-05, "loss": 0.0028, "step": 60180 }, { "grad_norm": 0.057022713124752045, "learning_rate": 3.7423024154043183e-05, "loss": 0.0024, "step": 60190 }, { "grad_norm": 0.16251331567764282, "learning_rate": 3.740702178352963e-05, "loss": 0.0036, "step": 60200 }, { "grad_norm": 0.1053074449300766, "learning_rate": 3.739102079016523e-05, "loss": 0.0026, "step": 60210 }, { "grad_norm": 0.06560418009757996, "learning_rate": 3.7375021175699846e-05, "loss": 0.003, "step": 60220 }, { "grad_norm": 0.04652802646160126, "learning_rate": 3.735902294188317e-05, "loss": 0.0023, "step": 60230 }, { "grad_norm": 0.07325227558612823, "learning_rate": 3.734302609046473e-05, "loss": 0.0024, "step": 60240 }, { "grad_norm": 0.06082860007882118, "learning_rate": 3.732703062319391e-05, "loss": 0.0024, "step": 60250 }, { "grad_norm": 0.07009538263082504, "learning_rate": 3.731103654181999e-05, "loss": 0.0027, "step": 60260 }, { "grad_norm": 0.05107397958636284, "learning_rate": 3.729504384809203e-05, "loss": 0.0021, "step": 60270 }, { "grad_norm": 0.0637149065732956, "learning_rate": 3.727905254375898e-05, "loss": 0.0021, "step": 60280 }, { "grad_norm": 0.04967513307929039, "learning_rate": 3.7263062630569614e-05, "loss": 0.0017, "step": 60290 }, { "grad_norm": 0.06665831059217453, "learning_rate": 3.724707411027256e-05, "loss": 0.002, "step": 60300 }, { "grad_norm": 0.04877157881855965, "learning_rate": 3.723108698461631e-05, "loss": 0.002, "step": 60310 }, { "grad_norm": 0.06417476385831833, "learning_rate": 3.72151012553492e-05, "loss": 0.0026, "step": 60320 }, { "grad_norm": 0.06379779428243637, "learning_rate": 3.719911692421939e-05, "loss": 0.0025, "step": 60330 }, { "grad_norm": 0.0452013723552227, "learning_rate": 3.71831339929749e-05, "loss": 0.0018, "step": 60340 }, { "grad_norm": 0.04047112539410591, "learning_rate": 3.716715246336362e-05, "loss": 0.002, "step": 60350 }, { "grad_norm": 0.05758155882358551, "learning_rate": 3.715117233713324e-05, "loss": 0.0026, "step": 60360 }, { "grad_norm": 0.062400732189416885, "learning_rate": 3.713519361603135e-05, "loss": 0.003, "step": 60370 }, { "grad_norm": 0.05066187307238579, "learning_rate": 3.711921630180532e-05, "loss": 0.0024, "step": 60380 }, { "grad_norm": 0.060245826840400696, "learning_rate": 3.710324039620245e-05, "loss": 0.0032, "step": 60390 }, { "grad_norm": 0.0868871659040451, "learning_rate": 3.708726590096982e-05, "loss": 0.0027, "step": 60400 }, { "grad_norm": 0.0644182339310646, "learning_rate": 3.707129281785437e-05, "loss": 0.003, "step": 60410 }, { "grad_norm": 0.0665387511253357, "learning_rate": 3.705532114860291e-05, "loss": 0.0024, "step": 60420 }, { "grad_norm": 0.12376782298088074, "learning_rate": 3.703935089496207e-05, "loss": 0.0038, "step": 60430 }, { "grad_norm": 0.06618554145097733, "learning_rate": 3.702338205867829e-05, "loss": 0.004, "step": 60440 }, { "grad_norm": 0.07371210306882858, "learning_rate": 3.7007414641497976e-05, "loss": 0.0028, "step": 60450 }, { "grad_norm": 0.09422434866428375, "learning_rate": 3.699144864516726e-05, "loss": 0.0025, "step": 60460 }, { "grad_norm": 0.19045160710811615, "learning_rate": 3.6975484071432153e-05, "loss": 0.0036, "step": 60470 }, { "grad_norm": 0.07780320942401886, "learning_rate": 3.695952092203854e-05, "loss": 0.0029, "step": 60480 }, { "grad_norm": 0.07566429674625397, "learning_rate": 3.69435591987321e-05, "loss": 0.0035, "step": 60490 }, { "grad_norm": 0.07318558543920517, "learning_rate": 3.6927598903258374e-05, "loss": 0.0023, "step": 60500 }, { "grad_norm": 0.08135019242763519, "learning_rate": 3.691164003736281e-05, "loss": 0.0032, "step": 60510 }, { "grad_norm": 0.07826323807239532, "learning_rate": 3.689568260279061e-05, "loss": 0.0025, "step": 60520 }, { "grad_norm": 0.0809582993388176, "learning_rate": 3.687972660128686e-05, "loss": 0.0025, "step": 60530 }, { "grad_norm": 0.07372023910284042, "learning_rate": 3.686377203459648e-05, "loss": 0.0026, "step": 60540 }, { "grad_norm": 0.0464334636926651, "learning_rate": 3.6847818904464226e-05, "loss": 0.0017, "step": 60550 }, { "grad_norm": 0.05757390707731247, "learning_rate": 3.683186721263474e-05, "loss": 0.0025, "step": 60560 }, { "grad_norm": 0.09129685908555984, "learning_rate": 3.6815916960852435e-05, "loss": 0.0031, "step": 60570 }, { "grad_norm": 0.08741920441389084, "learning_rate": 3.679996815086165e-05, "loss": 0.0025, "step": 60580 }, { "grad_norm": 0.062945157289505, "learning_rate": 3.678402078440649e-05, "loss": 0.0021, "step": 60590 }, { "grad_norm": 0.05793749913573265, "learning_rate": 3.676807486323096e-05, "loss": 0.0019, "step": 60600 }, { "grad_norm": 0.04641110077500343, "learning_rate": 3.6752130389078864e-05, "loss": 0.0023, "step": 60610 }, { "grad_norm": 0.05367431044578552, "learning_rate": 3.673618736369388e-05, "loss": 0.0026, "step": 60620 }, { "grad_norm": 0.060739994049072266, "learning_rate": 3.6720245788819474e-05, "loss": 0.0029, "step": 60630 }, { "grad_norm": 0.0863199532032013, "learning_rate": 3.6704305666199044e-05, "loss": 0.0025, "step": 60640 }, { "grad_norm": 0.08574321120977402, "learning_rate": 3.668836699757576e-05, "loss": 0.0035, "step": 60650 }, { "grad_norm": 0.06838196516036987, "learning_rate": 3.667242978469265e-05, "loss": 0.002, "step": 60660 }, { "grad_norm": 0.07507354766130447, "learning_rate": 3.6656494029292575e-05, "loss": 0.0027, "step": 60670 }, { "grad_norm": 0.06355579942464828, "learning_rate": 3.664055973311825e-05, "loss": 0.0022, "step": 60680 }, { "grad_norm": 0.051804110407829285, "learning_rate": 3.662462689791221e-05, "loss": 0.0024, "step": 60690 }, { "grad_norm": 0.06936577707529068, "learning_rate": 3.660869552541689e-05, "loss": 0.0024, "step": 60700 }, { "grad_norm": 0.0632789209485054, "learning_rate": 3.659276561737448e-05, "loss": 0.0023, "step": 60710 }, { "grad_norm": 0.045347660779953, "learning_rate": 3.657683717552708e-05, "loss": 0.0029, "step": 60720 }, { "grad_norm": 0.06451152265071869, "learning_rate": 3.6560910201616574e-05, "loss": 0.0025, "step": 60730 }, { "grad_norm": 0.05453112721443176, "learning_rate": 3.654498469738472e-05, "loss": 0.0025, "step": 60740 }, { "grad_norm": 0.05384374409914017, "learning_rate": 3.6529060664573126e-05, "loss": 0.0031, "step": 60750 }, { "grad_norm": 0.06252624094486237, "learning_rate": 3.6513138104923176e-05, "loss": 0.0029, "step": 60760 }, { "grad_norm": 0.06851896643638611, "learning_rate": 3.6497217020176176e-05, "loss": 0.0026, "step": 60770 }, { "grad_norm": 0.06045937165617943, "learning_rate": 3.648129741207323e-05, "loss": 0.0024, "step": 60780 }, { "grad_norm": 0.08854204416275024, "learning_rate": 3.646537928235527e-05, "loss": 0.0036, "step": 60790 }, { "grad_norm": 0.07043900340795517, "learning_rate": 3.6449462632763075e-05, "loss": 0.0024, "step": 60800 }, { "grad_norm": 0.05588576942682266, "learning_rate": 3.643354746503729e-05, "loss": 0.0026, "step": 60810 }, { "grad_norm": 0.0586027093231678, "learning_rate": 3.6417633780918345e-05, "loss": 0.0023, "step": 60820 }, { "grad_norm": 0.05160659924149513, "learning_rate": 3.640172158214655e-05, "loss": 0.0032, "step": 60830 }, { "grad_norm": 0.06912123411893845, "learning_rate": 3.6385810870462035e-05, "loss": 0.0019, "step": 60840 }, { "grad_norm": 0.08248595148324966, "learning_rate": 3.636990164760479e-05, "loss": 0.0021, "step": 60850 }, { "grad_norm": 0.058720074594020844, "learning_rate": 3.6353993915314596e-05, "loss": 0.0025, "step": 60860 }, { "grad_norm": 0.045092809945344925, "learning_rate": 3.6338087675331124e-05, "loss": 0.0019, "step": 60870 }, { "grad_norm": 0.07279480993747711, "learning_rate": 3.632218292939383e-05, "loss": 0.0025, "step": 60880 }, { "grad_norm": 0.0511033870279789, "learning_rate": 3.6306279679242033e-05, "loss": 0.0027, "step": 60890 }, { "grad_norm": 0.06358032673597336, "learning_rate": 3.629037792661492e-05, "loss": 0.0026, "step": 60900 }, { "grad_norm": 0.0632525160908699, "learning_rate": 3.627447767325147e-05, "loss": 0.002, "step": 60910 }, { "grad_norm": 0.06289834529161453, "learning_rate": 3.62585789208905e-05, "loss": 0.002, "step": 60920 }, { "grad_norm": 0.060760945081710815, "learning_rate": 3.624268167127068e-05, "loss": 0.0021, "step": 60930 }, { "grad_norm": 0.04055705666542053, "learning_rate": 3.622678592613051e-05, "loss": 0.0022, "step": 60940 }, { "grad_norm": 0.08794164657592773, "learning_rate": 3.6210891687208304e-05, "loss": 0.0034, "step": 60950 }, { "grad_norm": 0.09176647663116455, "learning_rate": 3.619499895624228e-05, "loss": 0.0025, "step": 60960 }, { "grad_norm": 0.09760189801454544, "learning_rate": 3.61791077349704e-05, "loss": 0.0033, "step": 60970 }, { "grad_norm": 0.04791257157921791, "learning_rate": 3.616321802513053e-05, "loss": 0.0018, "step": 60980 }, { "grad_norm": 0.050402041524648666, "learning_rate": 3.6147329828460325e-05, "loss": 0.0017, "step": 60990 }, { "grad_norm": 0.05779816210269928, "learning_rate": 3.6131443146697306e-05, "loss": 0.0023, "step": 61000 }, { "grad_norm": 0.08087149262428284, "learning_rate": 3.611555798157879e-05, "loss": 0.0027, "step": 61010 }, { "grad_norm": 0.06399279832839966, "learning_rate": 3.6099674334842e-05, "loss": 0.0023, "step": 61020 }, { "grad_norm": 0.057506147772073746, "learning_rate": 3.6083792208223934e-05, "loss": 0.0026, "step": 61030 }, { "grad_norm": 0.0629502609372139, "learning_rate": 3.606791160346142e-05, "loss": 0.0022, "step": 61040 }, { "grad_norm": 0.08203628659248352, "learning_rate": 3.6052032522291134e-05, "loss": 0.0028, "step": 61050 }, { "grad_norm": 0.06298638880252838, "learning_rate": 3.603615496644962e-05, "loss": 0.0022, "step": 61060 }, { "grad_norm": 0.071625255048275, "learning_rate": 3.60202789376732e-05, "loss": 0.0019, "step": 61070 }, { "grad_norm": 0.04866486415266991, "learning_rate": 3.6004404437698036e-05, "loss": 0.0025, "step": 61080 }, { "grad_norm": 0.12244396656751633, "learning_rate": 3.5988531468260184e-05, "loss": 0.0026, "step": 61090 }, { "grad_norm": 0.047842659056186676, "learning_rate": 3.597266003109546e-05, "loss": 0.0022, "step": 61100 }, { "grad_norm": 0.056206557899713516, "learning_rate": 3.5956790127939557e-05, "loss": 0.002, "step": 61110 }, { "grad_norm": 0.3516215682029724, "learning_rate": 3.594092176052796e-05, "loss": 0.0033, "step": 61120 }, { "grad_norm": 0.07691410183906555, "learning_rate": 3.592505493059603e-05, "loss": 0.0023, "step": 61130 }, { "grad_norm": 0.059601448476314545, "learning_rate": 3.5909189639878905e-05, "loss": 0.0025, "step": 61140 }, { "grad_norm": 0.1541709452867508, "learning_rate": 3.5893325890111626e-05, "loss": 0.0043, "step": 61150 }, { "grad_norm": 0.079445019364357, "learning_rate": 3.5877463683029034e-05, "loss": 0.0026, "step": 61160 }, { "grad_norm": 0.05889008939266205, "learning_rate": 3.586160302036578e-05, "loss": 0.0021, "step": 61170 }, { "grad_norm": 0.09573853760957718, "learning_rate": 3.584574390385636e-05, "loss": 0.0043, "step": 61180 }, { "grad_norm": 0.0642416700720787, "learning_rate": 3.582988633523511e-05, "loss": 0.0018, "step": 61190 }, { "grad_norm": 0.052098438143730164, "learning_rate": 3.581403031623616e-05, "loss": 0.0027, "step": 61200 }, { "grad_norm": 0.07696392387151718, "learning_rate": 3.579817584859354e-05, "loss": 0.0031, "step": 61210 }, { "grad_norm": 0.07447999715805054, "learning_rate": 3.578232293404107e-05, "loss": 0.0025, "step": 61220 }, { "grad_norm": 0.044857561588287354, "learning_rate": 3.5766471574312385e-05, "loss": 0.0025, "step": 61230 }, { "grad_norm": 0.05765646696090698, "learning_rate": 3.5750621771140964e-05, "loss": 0.0025, "step": 61240 }, { "grad_norm": 0.06378220021724701, "learning_rate": 3.5734773526260115e-05, "loss": 0.0021, "step": 61250 }, { "grad_norm": 0.0729542225599289, "learning_rate": 3.5718926841402996e-05, "loss": 0.0036, "step": 61260 }, { "grad_norm": 0.05178680270910263, "learning_rate": 3.5703081718302555e-05, "loss": 0.0027, "step": 61270 }, { "grad_norm": 0.04291559010744095, "learning_rate": 3.568723815869162e-05, "loss": 0.002, "step": 61280 }, { "grad_norm": 0.04802162945270538, "learning_rate": 3.5671396164302777e-05, "loss": 0.0021, "step": 61290 }, { "grad_norm": 0.05949685350060463, "learning_rate": 3.565555573686853e-05, "loss": 0.0031, "step": 61300 }, { "grad_norm": 0.06294847279787064, "learning_rate": 3.563971687812113e-05, "loss": 0.0018, "step": 61310 }, { "grad_norm": 0.0656142607331276, "learning_rate": 3.5623879589792706e-05, "loss": 0.0024, "step": 61320 }, { "grad_norm": 0.09765163064002991, "learning_rate": 3.5608043873615185e-05, "loss": 0.0028, "step": 61330 }, { "grad_norm": 0.10123012214899063, "learning_rate": 3.559220973132035e-05, "loss": 0.0027, "step": 61340 }, { "grad_norm": 0.08105991780757904, "learning_rate": 3.557637716463981e-05, "loss": 0.0034, "step": 61350 }, { "grad_norm": 0.05809653177857399, "learning_rate": 3.556054617530497e-05, "loss": 0.0026, "step": 61360 }, { "grad_norm": 0.06673775613307953, "learning_rate": 3.5544716765047106e-05, "loss": 0.0021, "step": 61370 }, { "grad_norm": 0.0643359050154686, "learning_rate": 3.5528888935597274e-05, "loss": 0.0021, "step": 61380 }, { "grad_norm": 0.05164853483438492, "learning_rate": 3.551306268868636e-05, "loss": 0.0019, "step": 61390 }, { "grad_norm": 0.05744212493300438, "learning_rate": 3.5497238026045166e-05, "loss": 0.0032, "step": 61400 }, { "grad_norm": 0.06542707979679108, "learning_rate": 3.548141494940421e-05, "loss": 0.0025, "step": 61410 }, { "grad_norm": 0.05622105300426483, "learning_rate": 3.546559346049388e-05, "loss": 0.0025, "step": 61420 }, { "grad_norm": 0.06673672795295715, "learning_rate": 3.544977356104441e-05, "loss": 0.0025, "step": 61430 }, { "grad_norm": 0.07128991186618805, "learning_rate": 3.543395525278582e-05, "loss": 0.0022, "step": 61440 }, { "grad_norm": 0.08639304339885712, "learning_rate": 3.5418138537447985e-05, "loss": 0.0023, "step": 61450 }, { "grad_norm": 0.15737052261829376, "learning_rate": 3.5402323416760574e-05, "loss": 0.0042, "step": 61460 }, { "grad_norm": 0.08726401627063751, "learning_rate": 3.538650989245316e-05, "loss": 0.0038, "step": 61470 }, { "grad_norm": 0.07721107453107834, "learning_rate": 3.537069796625504e-05, "loss": 0.0028, "step": 61480 }, { "grad_norm": 0.09632062166929245, "learning_rate": 3.53548876398954e-05, "loss": 0.0027, "step": 61490 }, { "grad_norm": 0.06743987649679184, "learning_rate": 3.5339078915103214e-05, "loss": 0.0022, "step": 61500 }, { "grad_norm": 0.058172840625047684, "learning_rate": 3.532327179360733e-05, "loss": 0.002, "step": 61510 }, { "grad_norm": 0.06906135380268097, "learning_rate": 3.530746627713636e-05, "loss": 0.0019, "step": 61520 }, { "grad_norm": 0.07459285110235214, "learning_rate": 3.5291662367418805e-05, "loss": 0.0027, "step": 61530 }, { "grad_norm": 0.06040971726179123, "learning_rate": 3.527586006618293e-05, "loss": 0.0022, "step": 61540 }, { "grad_norm": 0.04693603515625, "learning_rate": 3.526005937515687e-05, "loss": 0.0019, "step": 61550 }, { "grad_norm": 0.055962808430194855, "learning_rate": 3.524426029606856e-05, "loss": 0.0018, "step": 61560 }, { "grad_norm": 0.07639940083026886, "learning_rate": 3.5228462830645746e-05, "loss": 0.0021, "step": 61570 }, { "grad_norm": 0.07872083783149719, "learning_rate": 3.5212666980616036e-05, "loss": 0.0027, "step": 61580 }, { "grad_norm": 0.05305122584104538, "learning_rate": 3.519687274770681e-05, "loss": 0.002, "step": 61590 }, { "grad_norm": 0.04443752020597458, "learning_rate": 3.518108013364535e-05, "loss": 0.0021, "step": 61600 }, { "grad_norm": 0.05508482828736305, "learning_rate": 3.516528914015868e-05, "loss": 0.0021, "step": 61610 }, { "grad_norm": 0.05763622000813484, "learning_rate": 3.5149499768973694e-05, "loss": 0.0019, "step": 61620 }, { "grad_norm": 0.06705738604068756, "learning_rate": 3.513371202181708e-05, "loss": 0.0022, "step": 61630 }, { "grad_norm": 0.05352674424648285, "learning_rate": 3.511792590041537e-05, "loss": 0.0021, "step": 61640 }, { "grad_norm": 0.06573609262704849, "learning_rate": 3.510214140649488e-05, "loss": 0.0025, "step": 61650 }, { "grad_norm": 0.08212669938802719, "learning_rate": 3.508635854178184e-05, "loss": 0.0021, "step": 61660 }, { "grad_norm": 0.05709180608391762, "learning_rate": 3.50705773080022e-05, "loss": 0.0022, "step": 61670 }, { "grad_norm": 0.05721622332930565, "learning_rate": 3.505479770688178e-05, "loss": 0.0036, "step": 61680 }, { "grad_norm": 0.06317795813083649, "learning_rate": 3.503901974014622e-05, "loss": 0.0019, "step": 61690 }, { "grad_norm": 0.05969005823135376, "learning_rate": 3.502324340952095e-05, "loss": 0.0022, "step": 61700 }, { "grad_norm": 0.05090857297182083, "learning_rate": 3.500746871673126e-05, "loss": 0.0023, "step": 61710 }, { "grad_norm": 0.06971710175275803, "learning_rate": 3.499169566350227e-05, "loss": 0.0018, "step": 61720 }, { "grad_norm": 0.05898590013384819, "learning_rate": 3.4975924251558876e-05, "loss": 0.0027, "step": 61730 }, { "grad_norm": 0.06904743611812592, "learning_rate": 3.496015448262582e-05, "loss": 0.0025, "step": 61740 }, { "grad_norm": 0.06691984087228775, "learning_rate": 3.494438635842764e-05, "loss": 0.0021, "step": 61750 }, { "grad_norm": 0.07250145822763443, "learning_rate": 3.492861988068874e-05, "loss": 0.0022, "step": 61760 }, { "grad_norm": 0.05072847381234169, "learning_rate": 3.491285505113332e-05, "loss": 0.0027, "step": 61770 }, { "grad_norm": 0.06844713538885117, "learning_rate": 3.4897091871485376e-05, "loss": 0.0017, "step": 61780 }, { "grad_norm": 0.06290709227323532, "learning_rate": 3.488133034346875e-05, "loss": 0.0023, "step": 61790 }, { "grad_norm": 0.05686964467167854, "learning_rate": 3.486557046880713e-05, "loss": 0.0026, "step": 61800 }, { "grad_norm": 0.06435991823673248, "learning_rate": 3.484981224922395e-05, "loss": 0.0025, "step": 61810 }, { "grad_norm": 0.08277329802513123, "learning_rate": 3.4834055686442544e-05, "loss": 0.0033, "step": 61820 }, { "grad_norm": 0.06915535032749176, "learning_rate": 3.4818300782186e-05, "loss": 0.0043, "step": 61830 }, { "grad_norm": 0.07563350349664688, "learning_rate": 3.4802547538177236e-05, "loss": 0.0021, "step": 61840 }, { "grad_norm": 0.08368942141532898, "learning_rate": 3.478679595613905e-05, "loss": 0.0023, "step": 61850 }, { "grad_norm": 0.19006295502185822, "learning_rate": 3.4771046037793975e-05, "loss": 0.0027, "step": 61860 }, { "grad_norm": 0.08589693158864975, "learning_rate": 3.4755297784864416e-05, "loss": 0.004, "step": 61870 }, { "grad_norm": 0.13815909624099731, "learning_rate": 3.473955119907257e-05, "loss": 0.0042, "step": 61880 }, { "grad_norm": 0.0779080018401146, "learning_rate": 3.472380628214046e-05, "loss": 0.0026, "step": 61890 }, { "grad_norm": 0.0775030180811882, "learning_rate": 3.4708063035789906e-05, "loss": 0.0025, "step": 61900 }, { "grad_norm": 0.060488346964120865, "learning_rate": 3.4692321461742614e-05, "loss": 0.002, "step": 61910 }, { "grad_norm": 0.05271916463971138, "learning_rate": 3.4676581561720023e-05, "loss": 0.0016, "step": 61920 }, { "grad_norm": 0.0775662288069725, "learning_rate": 3.466084333744344e-05, "loss": 0.0022, "step": 61930 }, { "grad_norm": 0.07947751134634018, "learning_rate": 3.4645106790633964e-05, "loss": 0.0027, "step": 61940 }, { "grad_norm": 0.08395987749099731, "learning_rate": 3.4629371923012526e-05, "loss": 0.0024, "step": 61950 }, { "grad_norm": 0.06421974301338196, "learning_rate": 3.4613638736299856e-05, "loss": 0.0023, "step": 61960 }, { "grad_norm": 0.06169263646006584, "learning_rate": 3.459790723221652e-05, "loss": 0.0017, "step": 61970 }, { "grad_norm": 0.06426038593053818, "learning_rate": 3.45821774124829e-05, "loss": 0.0023, "step": 61980 }, { "grad_norm": 0.07112405449151993, "learning_rate": 3.456644927881918e-05, "loss": 0.0021, "step": 61990 }, { "grad_norm": 0.05300851911306381, "learning_rate": 3.4550722832945354e-05, "loss": 0.0018, "step": 62000 }, { "grad_norm": 0.06032325327396393, "learning_rate": 3.453499807658126e-05, "loss": 0.002, "step": 62010 }, { "grad_norm": 0.03607027977705002, "learning_rate": 3.451927501144653e-05, "loss": 0.0021, "step": 62020 }, { "grad_norm": 0.04539237916469574, "learning_rate": 3.450355363926059e-05, "loss": 0.0018, "step": 62030 }, { "grad_norm": 0.043292850255966187, "learning_rate": 3.448783396174273e-05, "loss": 0.0022, "step": 62040 }, { "grad_norm": 0.06063906475901604, "learning_rate": 3.4472115980612034e-05, "loss": 0.003, "step": 62050 }, { "grad_norm": 0.054097577929496765, "learning_rate": 3.4456399697587396e-05, "loss": 0.0017, "step": 62060 }, { "grad_norm": 0.09229674935340881, "learning_rate": 3.4440685114387506e-05, "loss": 0.0029, "step": 62070 }, { "grad_norm": 0.046515993773937225, "learning_rate": 3.442497223273091e-05, "loss": 0.0024, "step": 62080 }, { "grad_norm": 0.04134286195039749, "learning_rate": 3.4409261054335926e-05, "loss": 0.0022, "step": 62090 }, { "grad_norm": 0.07219839841127396, "learning_rate": 3.439355158092069e-05, "loss": 0.002, "step": 62100 }, { "grad_norm": 0.05464715510606766, "learning_rate": 3.437784381420322e-05, "loss": 0.003, "step": 62110 }, { "grad_norm": 0.0601375438272953, "learning_rate": 3.436213775590126e-05, "loss": 0.0022, "step": 62120 }, { "grad_norm": 0.05381615459918976, "learning_rate": 3.4346433407732405e-05, "loss": 0.0021, "step": 62130 }, { "grad_norm": 0.06139087677001953, "learning_rate": 3.433073077141405e-05, "loss": 0.0024, "step": 62140 }, { "grad_norm": 0.0813223198056221, "learning_rate": 3.431502984866343e-05, "loss": 0.0025, "step": 62150 }, { "grad_norm": 0.06021345406770706, "learning_rate": 3.429933064119753e-05, "loss": 0.0024, "step": 62160 }, { "grad_norm": 0.05052224546670914, "learning_rate": 3.4283633150733255e-05, "loss": 0.0019, "step": 62170 }, { "grad_norm": 0.0787314847111702, "learning_rate": 3.4267937378987225e-05, "loss": 0.0023, "step": 62180 }, { "grad_norm": 0.07124961167573929, "learning_rate": 3.425224332767592e-05, "loss": 0.0026, "step": 62190 }, { "grad_norm": 0.06534340977668762, "learning_rate": 3.4236550998515595e-05, "loss": 0.0026, "step": 62200 }, { "grad_norm": 0.043682824820280075, "learning_rate": 3.422086039322235e-05, "loss": 0.0019, "step": 62210 }, { "grad_norm": 0.06990636140108109, "learning_rate": 3.420517151351209e-05, "loss": 0.002, "step": 62220 }, { "grad_norm": 0.13551214337348938, "learning_rate": 3.418948436110053e-05, "loss": 0.0036, "step": 62230 }, { "grad_norm": 0.12189000844955444, "learning_rate": 3.417379893770319e-05, "loss": 0.0033, "step": 62240 }, { "grad_norm": 0.0989951565861702, "learning_rate": 3.41581152450354e-05, "loss": 0.0037, "step": 62250 }, { "grad_norm": 0.07064877450466156, "learning_rate": 3.414243328481231e-05, "loss": 0.0023, "step": 62260 }, { "grad_norm": 0.06842079013586044, "learning_rate": 3.412675305874888e-05, "loss": 0.0023, "step": 62270 }, { "grad_norm": 0.10411625355482101, "learning_rate": 3.4111074568559876e-05, "loss": 0.003, "step": 62280 }, { "grad_norm": 0.0866522341966629, "learning_rate": 3.4095397815959846e-05, "loss": 0.003, "step": 62290 }, { "grad_norm": 0.08356092125177383, "learning_rate": 3.407972280266322e-05, "loss": 0.0023, "step": 62300 }, { "grad_norm": 0.061942022293806076, "learning_rate": 3.4064049530384166e-05, "loss": 0.0018, "step": 62310 }, { "grad_norm": 0.040933337062597275, "learning_rate": 3.40483780008367e-05, "loss": 0.0018, "step": 62320 }, { "grad_norm": 0.08588054031133652, "learning_rate": 3.4032708215734635e-05, "loss": 0.0027, "step": 62330 }, { "grad_norm": 0.058311376720666885, "learning_rate": 3.4017040176791584e-05, "loss": 0.0029, "step": 62340 }, { "grad_norm": 0.044997598975896835, "learning_rate": 3.4001373885720976e-05, "loss": 0.0021, "step": 62350 }, { "grad_norm": 0.08228236436843872, "learning_rate": 3.398570934423608e-05, "loss": 0.0021, "step": 62360 }, { "grad_norm": 0.08336199820041656, "learning_rate": 3.397004655404994e-05, "loss": 0.0018, "step": 62370 }, { "grad_norm": 0.06800642609596252, "learning_rate": 3.395438551687541e-05, "loss": 0.003, "step": 62380 }, { "grad_norm": 0.06961110234260559, "learning_rate": 3.393872623442515e-05, "loss": 0.0031, "step": 62390 }, { "grad_norm": 0.051264919340610504, "learning_rate": 3.392306870841164e-05, "loss": 0.0019, "step": 62400 }, { "grad_norm": 0.06680747121572495, "learning_rate": 3.390741294054715e-05, "loss": 0.002, "step": 62410 }, { "grad_norm": 0.05257392302155495, "learning_rate": 3.38917589325438e-05, "loss": 0.0019, "step": 62420 }, { "grad_norm": 0.10865988582372665, "learning_rate": 3.387610668611349e-05, "loss": 0.0036, "step": 62430 }, { "grad_norm": 0.09608844667673111, "learning_rate": 3.38604562029679e-05, "loss": 0.0026, "step": 62440 }, { "grad_norm": 0.05805007740855217, "learning_rate": 3.384480748481855e-05, "loss": 0.0042, "step": 62450 }, { "grad_norm": 0.06729801744222641, "learning_rate": 3.382916053337676e-05, "loss": 0.0031, "step": 62460 }, { "grad_norm": 0.061414800584316254, "learning_rate": 3.381351535035366e-05, "loss": 0.0021, "step": 62470 }, { "grad_norm": 0.05091143026947975, "learning_rate": 3.379787193746019e-05, "loss": 0.0019, "step": 62480 }, { "grad_norm": 0.047866687178611755, "learning_rate": 3.378223029640709e-05, "loss": 0.0034, "step": 62490 }, { "grad_norm": 0.055098701268434525, "learning_rate": 3.376659042890489e-05, "loss": 0.0022, "step": 62500 }, { "grad_norm": 0.07540459930896759, "learning_rate": 3.375095233666397e-05, "loss": 0.0028, "step": 62510 }, { "grad_norm": 0.07988777756690979, "learning_rate": 3.373531602139447e-05, "loss": 0.0023, "step": 62520 }, { "grad_norm": 0.048183340579271317, "learning_rate": 3.371968148480635e-05, "loss": 0.0021, "step": 62530 }, { "grad_norm": 0.07243374735116959, "learning_rate": 3.370404872860937e-05, "loss": 0.0022, "step": 62540 }, { "grad_norm": 0.06525357812643051, "learning_rate": 3.3688417754513145e-05, "loss": 0.0022, "step": 62550 }, { "grad_norm": 0.05612272396683693, "learning_rate": 3.367278856422703e-05, "loss": 0.0027, "step": 62560 }, { "grad_norm": 0.06021187826991081, "learning_rate": 3.365716115946021e-05, "loss": 0.0021, "step": 62570 }, { "grad_norm": 0.0647396668791771, "learning_rate": 3.3641535541921666e-05, "loss": 0.0023, "step": 62580 }, { "grad_norm": 0.07974334806203842, "learning_rate": 3.36259117133202e-05, "loss": 0.0022, "step": 62590 }, { "grad_norm": 0.07022974640130997, "learning_rate": 3.36102896753644e-05, "loss": 0.0026, "step": 62600 }, { "grad_norm": 0.09906529635190964, "learning_rate": 3.35946694297627e-05, "loss": 0.0024, "step": 62610 }, { "grad_norm": 0.04676714539527893, "learning_rate": 3.357905097822328e-05, "loss": 0.0023, "step": 62620 }, { "grad_norm": 0.06952499598264694, "learning_rate": 3.356343432245416e-05, "loss": 0.0022, "step": 62630 }, { "grad_norm": 0.11086846888065338, "learning_rate": 3.354781946416315e-05, "loss": 0.0024, "step": 62640 }, { "grad_norm": 0.056640174239873886, "learning_rate": 3.3532206405057864e-05, "loss": 0.0021, "step": 62650 }, { "grad_norm": 0.05224509537220001, "learning_rate": 3.351659514684573e-05, "loss": 0.0016, "step": 62660 }, { "grad_norm": 0.07370539009571075, "learning_rate": 3.350098569123395e-05, "loss": 0.0034, "step": 62670 }, { "grad_norm": 0.04493332281708717, "learning_rate": 3.348537803992958e-05, "loss": 0.0018, "step": 62680 }, { "grad_norm": 0.06297976523637772, "learning_rate": 3.346977219463945e-05, "loss": 0.0021, "step": 62690 }, { "grad_norm": 0.056034788489341736, "learning_rate": 3.345416815707017e-05, "loss": 0.0021, "step": 62700 }, { "grad_norm": 0.09801996499300003, "learning_rate": 3.3438565928928185e-05, "loss": 0.0041, "step": 62710 }, { "grad_norm": 0.061632342636585236, "learning_rate": 3.3422965511919736e-05, "loss": 0.0026, "step": 62720 }, { "grad_norm": 0.05242561176419258, "learning_rate": 3.3407366907750846e-05, "loss": 0.002, "step": 62730 }, { "grad_norm": 0.06773938983678818, "learning_rate": 3.339177011812737e-05, "loss": 0.0023, "step": 62740 }, { "grad_norm": 0.06597254425287247, "learning_rate": 3.3376175144754935e-05, "loss": 0.0043, "step": 62750 }, { "grad_norm": 0.04336698353290558, "learning_rate": 3.3360581989339004e-05, "loss": 0.003, "step": 62760 }, { "grad_norm": 0.07515629380941391, "learning_rate": 3.3344990653584804e-05, "loss": 0.0028, "step": 62770 }, { "grad_norm": 0.0663030818104744, "learning_rate": 3.332940113919739e-05, "loss": 0.0017, "step": 62780 }, { "grad_norm": 0.05500059574842453, "learning_rate": 3.331381344788161e-05, "loss": 0.0028, "step": 62790 }, { "grad_norm": 0.04803456738591194, "learning_rate": 3.3298227581342076e-05, "loss": 0.0022, "step": 62800 }, { "grad_norm": 0.05206569284200668, "learning_rate": 3.328264354128329e-05, "loss": 0.0018, "step": 62810 }, { "grad_norm": 0.10111478716135025, "learning_rate": 3.3267061329409465e-05, "loss": 0.002, "step": 62820 }, { "grad_norm": 0.06861209124326706, "learning_rate": 3.325148094742466e-05, "loss": 0.0015, "step": 62830 }, { "grad_norm": 0.06497448682785034, "learning_rate": 3.323590239703271e-05, "loss": 0.0019, "step": 62840 }, { "grad_norm": 0.06912407279014587, "learning_rate": 3.322032567993728e-05, "loss": 0.002, "step": 62850 }, { "grad_norm": 0.04284288361668587, "learning_rate": 3.3204750797841774e-05, "loss": 0.0015, "step": 62860 }, { "grad_norm": 0.1002330482006073, "learning_rate": 3.31891777524495e-05, "loss": 0.0022, "step": 62870 }, { "grad_norm": 0.07279891520738602, "learning_rate": 3.317360654546347e-05, "loss": 0.0025, "step": 62880 }, { "grad_norm": 0.05397685989737511, "learning_rate": 3.3158037178586534e-05, "loss": 0.0021, "step": 62890 }, { "grad_norm": 0.055306822061538696, "learning_rate": 3.3142469653521316e-05, "loss": 0.0018, "step": 62900 }, { "grad_norm": 0.05339360982179642, "learning_rate": 3.3126903971970284e-05, "loss": 0.0019, "step": 62910 }, { "grad_norm": 0.061833497136831284, "learning_rate": 3.311134013563564e-05, "loss": 0.0021, "step": 62920 }, { "grad_norm": 0.049932487308979034, "learning_rate": 3.309577814621947e-05, "loss": 0.0018, "step": 62930 }, { "grad_norm": 0.05528418347239494, "learning_rate": 3.308021800542359e-05, "loss": 0.0019, "step": 62940 }, { "grad_norm": 0.06167106330394745, "learning_rate": 3.306465971494963e-05, "loss": 0.0026, "step": 62950 }, { "grad_norm": 0.08276695013046265, "learning_rate": 3.304910327649901e-05, "loss": 0.0024, "step": 62960 }, { "grad_norm": 0.08404503762722015, "learning_rate": 3.303354869177297e-05, "loss": 0.0019, "step": 62970 }, { "grad_norm": 0.043541330844163895, "learning_rate": 3.301799596247255e-05, "loss": 0.0022, "step": 62980 }, { "grad_norm": 0.060176167637109756, "learning_rate": 3.300244509029854e-05, "loss": 0.0017, "step": 62990 }, { "grad_norm": 0.08132289350032806, "learning_rate": 3.2986896076951585e-05, "loss": 0.0039, "step": 63000 }, { "grad_norm": 0.04041029512882233, "learning_rate": 3.297134892413211e-05, "loss": 0.0016, "step": 63010 }, { "grad_norm": 0.059711094945669174, "learning_rate": 3.295580363354031e-05, "loss": 0.0035, "step": 63020 }, { "grad_norm": 0.06021890416741371, "learning_rate": 3.2940260206876194e-05, "loss": 0.0028, "step": 63030 }, { "grad_norm": 0.07379216700792313, "learning_rate": 3.292471864583958e-05, "loss": 0.0026, "step": 63040 }, { "grad_norm": 0.06238831207156181, "learning_rate": 3.290917895213004e-05, "loss": 0.0026, "step": 63050 }, { "grad_norm": 0.050357893109321594, "learning_rate": 3.289364112744701e-05, "loss": 0.002, "step": 63060 }, { "grad_norm": 0.0722811371088028, "learning_rate": 3.287810517348967e-05, "loss": 0.0018, "step": 63070 }, { "grad_norm": 0.05166875943541527, "learning_rate": 3.2862571091957014e-05, "loss": 0.0016, "step": 63080 }, { "grad_norm": 0.07273610681295395, "learning_rate": 3.284703888454781e-05, "loss": 0.0025, "step": 63090 }, { "grad_norm": 0.06314269453287125, "learning_rate": 3.2831508552960646e-05, "loss": 0.0023, "step": 63100 }, { "grad_norm": 0.05952324718236923, "learning_rate": 3.281598009889387e-05, "loss": 0.0019, "step": 63110 }, { "grad_norm": 0.06662046164274216, "learning_rate": 3.28004535240457e-05, "loss": 0.0016, "step": 63120 }, { "grad_norm": 0.06704623252153397, "learning_rate": 3.278492883011407e-05, "loss": 0.0021, "step": 63130 }, { "grad_norm": 0.048290327191352844, "learning_rate": 3.276940601879675e-05, "loss": 0.0027, "step": 63140 }, { "grad_norm": 0.07828082144260406, "learning_rate": 3.275388509179127e-05, "loss": 0.0026, "step": 63150 }, { "grad_norm": 0.04980441555380821, "learning_rate": 3.2738366050794985e-05, "loss": 0.0023, "step": 63160 }, { "grad_norm": 0.05579211190342903, "learning_rate": 3.2722848897505034e-05, "loss": 0.0016, "step": 63170 }, { "grad_norm": 0.06966686248779297, "learning_rate": 3.2707333633618356e-05, "loss": 0.0024, "step": 63180 }, { "grad_norm": 0.053516972810029984, "learning_rate": 3.269182026083167e-05, "loss": 0.002, "step": 63190 }, { "grad_norm": 0.05885401740670204, "learning_rate": 3.2676308780841516e-05, "loss": 0.0016, "step": 63200 }, { "grad_norm": 0.05055946856737137, "learning_rate": 3.266079919534416e-05, "loss": 0.0018, "step": 63210 }, { "grad_norm": 0.09031789749860764, "learning_rate": 3.264529150603575e-05, "loss": 0.0019, "step": 63220 }, { "grad_norm": 0.05580509081482887, "learning_rate": 3.262978571461217e-05, "loss": 0.0022, "step": 63230 }, { "grad_norm": 0.07239146530628204, "learning_rate": 3.26142818227691e-05, "loss": 0.0019, "step": 63240 }, { "grad_norm": 0.051735032349824905, "learning_rate": 3.259877983220203e-05, "loss": 0.0016, "step": 63250 }, { "grad_norm": 0.07002037018537521, "learning_rate": 3.2583279744606246e-05, "loss": 0.0026, "step": 63260 }, { "grad_norm": 0.062287382781505585, "learning_rate": 3.25677815616768e-05, "loss": 0.0017, "step": 63270 }, { "grad_norm": 0.06487096846103668, "learning_rate": 3.255228528510856e-05, "loss": 0.0018, "step": 63280 }, { "grad_norm": 0.054248642176389694, "learning_rate": 3.253679091659617e-05, "loss": 0.002, "step": 63290 }, { "grad_norm": 0.06766059994697571, "learning_rate": 3.252129845783404e-05, "loss": 0.0017, "step": 63300 }, { "grad_norm": 0.05848347768187523, "learning_rate": 3.2505807910516456e-05, "loss": 0.0019, "step": 63310 }, { "grad_norm": 0.04074794799089432, "learning_rate": 3.249031927633742e-05, "loss": 0.0014, "step": 63320 }, { "grad_norm": 0.060542281717061996, "learning_rate": 3.2474832556990737e-05, "loss": 0.0023, "step": 63330 }, { "grad_norm": 0.09036260843276978, "learning_rate": 3.245934775417002e-05, "loss": 0.0036, "step": 63340 }, { "grad_norm": 0.053906913846731186, "learning_rate": 3.244386486956866e-05, "loss": 0.0031, "step": 63350 }, { "grad_norm": 0.08026347309350967, "learning_rate": 3.2428383904879846e-05, "loss": 0.0026, "step": 63360 }, { "grad_norm": 0.15321755409240723, "learning_rate": 3.241290486179652e-05, "loss": 0.0037, "step": 63370 }, { "grad_norm": 0.07981694489717484, "learning_rate": 3.239742774201151e-05, "loss": 0.003, "step": 63380 }, { "grad_norm": 0.0945572704076767, "learning_rate": 3.238195254721733e-05, "loss": 0.0038, "step": 63390 }, { "grad_norm": 0.08652956038713455, "learning_rate": 3.236647927910633e-05, "loss": 0.0024, "step": 63400 }, { "grad_norm": 0.06593635678291321, "learning_rate": 3.2351007939370646e-05, "loss": 0.0023, "step": 63410 }, { "grad_norm": 0.056002888828516006, "learning_rate": 3.2335538529702184e-05, "loss": 0.0024, "step": 63420 }, { "grad_norm": 0.044267769902944565, "learning_rate": 3.232007105179268e-05, "loss": 0.0018, "step": 63430 }, { "grad_norm": 0.05450603365898132, "learning_rate": 3.230460550733363e-05, "loss": 0.0026, "step": 63440 }, { "grad_norm": 0.06312394142150879, "learning_rate": 3.2289141898016315e-05, "loss": 0.002, "step": 63450 }, { "grad_norm": 0.06229901313781738, "learning_rate": 3.227368022553181e-05, "loss": 0.0023, "step": 63460 }, { "grad_norm": 0.05989474803209305, "learning_rate": 3.225822049157099e-05, "loss": 0.0018, "step": 63470 }, { "grad_norm": 0.05258253589272499, "learning_rate": 3.2242762697824494e-05, "loss": 0.002, "step": 63480 }, { "grad_norm": 0.07753193378448486, "learning_rate": 3.222730684598279e-05, "loss": 0.0022, "step": 63490 }, { "grad_norm": 0.06287211179733276, "learning_rate": 3.221185293773605e-05, "loss": 0.0017, "step": 63500 }, { "grad_norm": 0.06567802280187607, "learning_rate": 3.2196400974774354e-05, "loss": 0.0018, "step": 63510 }, { "grad_norm": 0.07857029885053635, "learning_rate": 3.2180950958787485e-05, "loss": 0.0019, "step": 63520 }, { "grad_norm": 0.0722523033618927, "learning_rate": 3.2165502891465024e-05, "loss": 0.0021, "step": 63530 }, { "grad_norm": 0.09612327069044113, "learning_rate": 3.215005677449636e-05, "loss": 0.0022, "step": 63540 }, { "grad_norm": 0.05656920745968819, "learning_rate": 3.2134612609570644e-05, "loss": 0.0022, "step": 63550 }, { "grad_norm": 0.09030279517173767, "learning_rate": 3.211917039837682e-05, "loss": 0.0023, "step": 63560 }, { "grad_norm": 0.0501125194132328, "learning_rate": 3.210373014260365e-05, "loss": 0.002, "step": 63570 }, { "grad_norm": 0.06488023698329926, "learning_rate": 3.208829184393965e-05, "loss": 0.0026, "step": 63580 }, { "grad_norm": 0.04393882304430008, "learning_rate": 3.207285550407312e-05, "loss": 0.0023, "step": 63590 }, { "grad_norm": 0.06264559179544449, "learning_rate": 3.205742112469217e-05, "loss": 0.0024, "step": 63600 }, { "grad_norm": 0.055100537836551666, "learning_rate": 3.204198870748465e-05, "loss": 0.0022, "step": 63610 }, { "grad_norm": 0.06190182641148567, "learning_rate": 3.202655825413824e-05, "loss": 0.0017, "step": 63620 }, { "grad_norm": 0.0725850909948349, "learning_rate": 3.2011129766340416e-05, "loss": 0.0022, "step": 63630 }, { "grad_norm": 0.075466588139534, "learning_rate": 3.1995703245778384e-05, "loss": 0.0018, "step": 63640 }, { "grad_norm": 0.06668899953365326, "learning_rate": 3.1980278694139186e-05, "loss": 0.003, "step": 63650 }, { "grad_norm": 0.05279335007071495, "learning_rate": 3.196485611310962e-05, "loss": 0.0023, "step": 63660 }, { "grad_norm": 0.05777968838810921, "learning_rate": 3.1949435504376257e-05, "loss": 0.0019, "step": 63670 }, { "grad_norm": 0.07034258544445038, "learning_rate": 3.1934016869625505e-05, "loss": 0.0023, "step": 63680 }, { "grad_norm": 0.06150926649570465, "learning_rate": 3.191860021054349e-05, "loss": 0.0024, "step": 63690 }, { "grad_norm": 0.0801444724202156, "learning_rate": 3.190318552881618e-05, "loss": 0.002, "step": 63700 }, { "grad_norm": 0.08363673090934753, "learning_rate": 3.1887772826129285e-05, "loss": 0.0023, "step": 63710 }, { "grad_norm": 0.06430500000715256, "learning_rate": 3.1872362104168335e-05, "loss": 0.0023, "step": 63720 }, { "grad_norm": 0.07340425997972488, "learning_rate": 3.185695336461861e-05, "loss": 0.0019, "step": 63730 }, { "grad_norm": 0.08681352436542511, "learning_rate": 3.184154660916519e-05, "loss": 0.0021, "step": 63740 }, { "grad_norm": 0.09962629526853561, "learning_rate": 3.1826141839492906e-05, "loss": 0.0022, "step": 63750 }, { "grad_norm": 0.04553021863102913, "learning_rate": 3.181073905728644e-05, "loss": 0.0019, "step": 63760 }, { "grad_norm": 0.07850637286901474, "learning_rate": 3.1795338264230214e-05, "loss": 0.0024, "step": 63770 }, { "grad_norm": 0.06792458146810532, "learning_rate": 3.1779939462008426e-05, "loss": 0.0018, "step": 63780 }, { "grad_norm": 0.059238433837890625, "learning_rate": 3.1764542652305066e-05, "loss": 0.0019, "step": 63790 }, { "grad_norm": 0.05324440822005272, "learning_rate": 3.17491478368039e-05, "loss": 0.0016, "step": 63800 }, { "grad_norm": 0.049606405198574066, "learning_rate": 3.173375501718847e-05, "loss": 0.0021, "step": 63810 }, { "grad_norm": 0.07242485880851746, "learning_rate": 3.171836419514215e-05, "loss": 0.0025, "step": 63820 }, { "grad_norm": 0.049417782574892044, "learning_rate": 3.170297537234803e-05, "loss": 0.0019, "step": 63830 }, { "grad_norm": 0.05060649663209915, "learning_rate": 3.168758855048901e-05, "loss": 0.0016, "step": 63840 }, { "grad_norm": 0.06761704385280609, "learning_rate": 3.1672203731247785e-05, "loss": 0.0027, "step": 63850 }, { "grad_norm": 0.04235260561108589, "learning_rate": 3.16568209163068e-05, "loss": 0.0017, "step": 63860 }, { "grad_norm": 0.04759012907743454, "learning_rate": 3.1641440107348295e-05, "loss": 0.0019, "step": 63870 }, { "grad_norm": 0.05232953652739525, "learning_rate": 3.162606130605429e-05, "loss": 0.002, "step": 63880 }, { "grad_norm": 0.044734708964824677, "learning_rate": 3.16106845141066e-05, "loss": 0.0017, "step": 63890 }, { "grad_norm": 0.05040018633008003, "learning_rate": 3.159530973318682e-05, "loss": 0.0019, "step": 63900 }, { "grad_norm": 0.07009606808423996, "learning_rate": 3.1579936964976286e-05, "loss": 0.0022, "step": 63910 }, { "grad_norm": 0.06898820400238037, "learning_rate": 3.156456621115615e-05, "loss": 0.0025, "step": 63920 }, { "grad_norm": 0.05679282173514366, "learning_rate": 3.154919747340734e-05, "loss": 0.0024, "step": 63930 }, { "grad_norm": 0.055778082460165024, "learning_rate": 3.153383075341054e-05, "loss": 0.0015, "step": 63940 }, { "grad_norm": 0.0434613898396492, "learning_rate": 3.1518466052846264e-05, "loss": 0.0015, "step": 63950 }, { "grad_norm": 0.04896984249353409, "learning_rate": 3.150310337339474e-05, "loss": 0.0014, "step": 63960 }, { "grad_norm": 0.057112473994493484, "learning_rate": 3.148774271673603e-05, "loss": 0.0025, "step": 63970 }, { "grad_norm": 0.07780402153730392, "learning_rate": 3.147238408454995e-05, "loss": 0.002, "step": 63980 }, { "grad_norm": 0.07621745020151138, "learning_rate": 3.1457027478516086e-05, "loss": 0.002, "step": 63990 }, { "grad_norm": 0.07439146190881729, "learning_rate": 3.144167290031382e-05, "loss": 0.0019, "step": 64000 }, { "grad_norm": 0.062112957239151, "learning_rate": 3.142632035162227e-05, "loss": 0.0022, "step": 64010 }, { "grad_norm": 0.05320781096816063, "learning_rate": 3.141096983412044e-05, "loss": 0.0015, "step": 64020 }, { "grad_norm": 0.05669097229838371, "learning_rate": 3.139562134948699e-05, "loss": 0.0023, "step": 64030 }, { "grad_norm": 0.11651826649904251, "learning_rate": 3.1380274899400424e-05, "loss": 0.0019, "step": 64040 }, { "grad_norm": 0.05492878332734108, "learning_rate": 3.1364930485539e-05, "loss": 0.0024, "step": 64050 }, { "grad_norm": 0.05483618751168251, "learning_rate": 3.134958810958076e-05, "loss": 0.0021, "step": 64060 }, { "grad_norm": 0.06646032631397247, "learning_rate": 3.13342477732035e-05, "loss": 0.0036, "step": 64070 }, { "grad_norm": 0.07543105632066727, "learning_rate": 3.131890947808487e-05, "loss": 0.0026, "step": 64080 }, { "grad_norm": 0.07693223655223846, "learning_rate": 3.13035732259022e-05, "loss": 0.0027, "step": 64090 }, { "grad_norm": 0.09015093743801117, "learning_rate": 3.1288239018332656e-05, "loss": 0.0022, "step": 64100 }, { "grad_norm": 0.06441512703895569, "learning_rate": 3.1272906857053164e-05, "loss": 0.0022, "step": 64110 }, { "grad_norm": 0.060820721089839935, "learning_rate": 3.125757674374042e-05, "loss": 0.0022, "step": 64120 }, { "grad_norm": 0.06258022785186768, "learning_rate": 3.124224868007088e-05, "loss": 0.002, "step": 64130 }, { "grad_norm": 0.053792692720890045, "learning_rate": 3.1226922667720854e-05, "loss": 0.002, "step": 64140 }, { "grad_norm": 0.05272713303565979, "learning_rate": 3.121159870836633e-05, "loss": 0.0026, "step": 64150 }, { "grad_norm": 0.06556154042482376, "learning_rate": 3.119627680368312e-05, "loss": 0.0034, "step": 64160 }, { "grad_norm": 0.07113680243492126, "learning_rate": 3.1180956955346816e-05, "loss": 0.0032, "step": 64170 }, { "grad_norm": 0.07575714588165283, "learning_rate": 3.116563916503276e-05, "loss": 0.0038, "step": 64180 }, { "grad_norm": 0.07871463894844055, "learning_rate": 3.1150323434416095e-05, "loss": 0.0019, "step": 64190 }, { "grad_norm": 0.06451825052499771, "learning_rate": 3.1135009765171696e-05, "loss": 0.0021, "step": 64200 }, { "grad_norm": 0.05534999445080757, "learning_rate": 3.111969815897429e-05, "loss": 0.0022, "step": 64210 }, { "grad_norm": 0.05493004992604256, "learning_rate": 3.1104388617498295e-05, "loss": 0.0016, "step": 64220 }, { "grad_norm": 0.05591977387666702, "learning_rate": 3.1089081142417966e-05, "loss": 0.0017, "step": 64230 }, { "grad_norm": 0.057522114366292953, "learning_rate": 3.1073775735407285e-05, "loss": 0.0023, "step": 64240 }, { "grad_norm": 0.05344970151782036, "learning_rate": 3.1058472398140035e-05, "loss": 0.0021, "step": 64250 }, { "grad_norm": 0.0741843730211258, "learning_rate": 3.1043171132289736e-05, "loss": 0.002, "step": 64260 }, { "grad_norm": 0.05172545462846756, "learning_rate": 3.1027871939529774e-05, "loss": 0.0018, "step": 64270 }, { "grad_norm": 0.06199568510055542, "learning_rate": 3.1012574821533214e-05, "loss": 0.0019, "step": 64280 }, { "grad_norm": 0.05116036534309387, "learning_rate": 3.099727977997291e-05, "loss": 0.002, "step": 64290 }, { "grad_norm": 0.07569751888513565, "learning_rate": 3.098198681652154e-05, "loss": 0.0018, "step": 64300 }, { "grad_norm": 0.06210274621844292, "learning_rate": 3.0966695932851475e-05, "loss": 0.0019, "step": 64310 }, { "grad_norm": 0.051632605493068695, "learning_rate": 3.0951407130634926e-05, "loss": 0.0018, "step": 64320 }, { "grad_norm": 0.07174569368362427, "learning_rate": 3.0936120411543864e-05, "loss": 0.0024, "step": 64330 }, { "grad_norm": 0.0607309564948082, "learning_rate": 3.0920835777250015e-05, "loss": 0.0022, "step": 64340 }, { "grad_norm": 0.03773137181997299, "learning_rate": 3.090555322942488e-05, "loss": 0.0018, "step": 64350 }, { "grad_norm": 0.06823192536830902, "learning_rate": 3.089027276973973e-05, "loss": 0.0022, "step": 64360 }, { "grad_norm": 0.05770331248641014, "learning_rate": 3.0874994399865615e-05, "loss": 0.0031, "step": 64370 }, { "grad_norm": 0.06432782113552094, "learning_rate": 3.085971812147337e-05, "loss": 0.0025, "step": 64380 }, { "grad_norm": 0.05748927965760231, "learning_rate": 3.0844443936233557e-05, "loss": 0.0019, "step": 64390 }, { "grad_norm": 0.07095129042863846, "learning_rate": 3.082917184581657e-05, "loss": 0.0022, "step": 64400 }, { "grad_norm": 0.03998145833611488, "learning_rate": 3.0813901851892513e-05, "loss": 0.0019, "step": 64410 }, { "grad_norm": 0.10224220156669617, "learning_rate": 3.0798633956131316e-05, "loss": 0.0022, "step": 64420 }, { "grad_norm": 0.0589648075401783, "learning_rate": 3.078336816020264e-05, "loss": 0.002, "step": 64430 }, { "grad_norm": 0.04775869846343994, "learning_rate": 3.0768104465775935e-05, "loss": 0.0023, "step": 64440 }, { "grad_norm": 0.05930785462260246, "learning_rate": 3.0752842874520396e-05, "loss": 0.0023, "step": 64450 }, { "grad_norm": 0.05006998032331467, "learning_rate": 3.073758338810503e-05, "loss": 0.0017, "step": 64460 }, { "grad_norm": 0.05523066967725754, "learning_rate": 3.072232600819859e-05, "loss": 0.0027, "step": 64470 }, { "grad_norm": 0.06400556117296219, "learning_rate": 3.070707073646959e-05, "loss": 0.002, "step": 64480 }, { "grad_norm": 0.04339083656668663, "learning_rate": 3.069181757458633e-05, "loss": 0.002, "step": 64490 }, { "grad_norm": 0.05039794743061066, "learning_rate": 3.067656652421687e-05, "loss": 0.0017, "step": 64500 }, { "grad_norm": 0.043654587119817734, "learning_rate": 3.0661317587029025e-05, "loss": 0.0013, "step": 64510 }, { "grad_norm": 0.05729510635137558, "learning_rate": 3.064607076469043e-05, "loss": 0.002, "step": 64520 }, { "grad_norm": 0.042087122797966, "learning_rate": 3.063082605886845e-05, "loss": 0.0015, "step": 64530 }, { "grad_norm": 0.06870099157094955, "learning_rate": 3.061558347123019e-05, "loss": 0.0019, "step": 64540 }, { "grad_norm": 0.06665327399969101, "learning_rate": 3.060034300344259e-05, "loss": 0.0024, "step": 64550 }, { "grad_norm": 0.04183217138051987, "learning_rate": 3.058510465717231e-05, "loss": 0.0014, "step": 64560 }, { "grad_norm": 0.04839923977851868, "learning_rate": 3.0569868434085804e-05, "loss": 0.0016, "step": 64570 }, { "grad_norm": 0.057847678661346436, "learning_rate": 3.055463433584924e-05, "loss": 0.0019, "step": 64580 }, { "grad_norm": 0.056252170354127884, "learning_rate": 3.0539402364128656e-05, "loss": 0.0022, "step": 64590 }, { "grad_norm": 0.043025076389312744, "learning_rate": 3.052417252058977e-05, "loss": 0.0018, "step": 64600 }, { "grad_norm": 0.04342074319720268, "learning_rate": 3.0508944806898092e-05, "loss": 0.0017, "step": 64610 }, { "grad_norm": 0.05593905225396156, "learning_rate": 3.04937192247189e-05, "loss": 0.0018, "step": 64620 }, { "grad_norm": 0.04857053980231285, "learning_rate": 3.047849577571725e-05, "loss": 0.0014, "step": 64630 }, { "grad_norm": 0.05810972675681114, "learning_rate": 3.046327446155795e-05, "loss": 0.0021, "step": 64640 }, { "grad_norm": 0.06147903576493263, "learning_rate": 3.0448055283905587e-05, "loss": 0.0021, "step": 64650 }, { "grad_norm": 0.07403156906366348, "learning_rate": 3.0432838244424494e-05, "loss": 0.0021, "step": 64660 }, { "grad_norm": 0.06779778748750687, "learning_rate": 3.04176233447788e-05, "loss": 0.0025, "step": 64670 }, { "grad_norm": 0.05807585269212723, "learning_rate": 3.040241058663238e-05, "loss": 0.0022, "step": 64680 }, { "grad_norm": 0.07711220532655716, "learning_rate": 3.038719997164887e-05, "loss": 0.0022, "step": 64690 }, { "grad_norm": 0.05357784405350685, "learning_rate": 3.0371991501491683e-05, "loss": 0.0017, "step": 64700 }, { "grad_norm": 0.05284224450588226, "learning_rate": 3.0356785177823975e-05, "loss": 0.0022, "step": 64710 }, { "grad_norm": 0.045348528772592545, "learning_rate": 3.0341581002308726e-05, "loss": 0.0015, "step": 64720 }, { "grad_norm": 0.06668628007173538, "learning_rate": 3.0326378976608627e-05, "loss": 0.002, "step": 64730 }, { "grad_norm": 0.04348672181367874, "learning_rate": 3.0311179102386146e-05, "loss": 0.0015, "step": 64740 }, { "grad_norm": 0.07187147438526154, "learning_rate": 3.0295981381303513e-05, "loss": 0.0019, "step": 64750 }, { "grad_norm": 0.04988522082567215, "learning_rate": 3.0280785815022727e-05, "loss": 0.0014, "step": 64760 }, { "grad_norm": 0.05538449063897133, "learning_rate": 3.0265592405205544e-05, "loss": 0.0017, "step": 64770 }, { "grad_norm": 0.049072276800870895, "learning_rate": 3.0250401153513517e-05, "loss": 0.0015, "step": 64780 }, { "grad_norm": 0.06232462078332901, "learning_rate": 3.0235212061607936e-05, "loss": 0.0016, "step": 64790 }, { "grad_norm": 0.050123315304517746, "learning_rate": 3.0220025131149842e-05, "loss": 0.0025, "step": 64800 }, { "grad_norm": 0.10255541652441025, "learning_rate": 3.020484036380007e-05, "loss": 0.0021, "step": 64810 }, { "grad_norm": 0.07145002484321594, "learning_rate": 3.0189657761219188e-05, "loss": 0.0019, "step": 64820 }, { "grad_norm": 0.05556591972708702, "learning_rate": 3.017447732506753e-05, "loss": 0.0028, "step": 64830 }, { "grad_norm": 0.07857600599527359, "learning_rate": 3.015929905700524e-05, "loss": 0.0022, "step": 64840 }, { "grad_norm": 0.06588900089263916, "learning_rate": 3.014412295869218e-05, "loss": 0.0026, "step": 64850 }, { "grad_norm": 0.07546882331371307, "learning_rate": 3.0128949031787977e-05, "loss": 0.0023, "step": 64860 }, { "grad_norm": 0.05459445342421532, "learning_rate": 3.0113777277952022e-05, "loss": 0.0018, "step": 64870 }, { "grad_norm": 0.04126264527440071, "learning_rate": 3.0098607698843496e-05, "loss": 0.0031, "step": 64880 }, { "grad_norm": 0.040971849113702774, "learning_rate": 3.0083440296121312e-05, "loss": 0.0015, "step": 64890 }, { "grad_norm": 0.07109639793634415, "learning_rate": 3.0068275071444142e-05, "loss": 0.0025, "step": 64900 }, { "grad_norm": 0.07066909223794937, "learning_rate": 3.0053112026470442e-05, "loss": 0.0025, "step": 64910 }, { "grad_norm": 0.06697839498519897, "learning_rate": 3.0037951162858434e-05, "loss": 0.0021, "step": 64920 }, { "grad_norm": 0.053791873157024384, "learning_rate": 3.002279248226607e-05, "loss": 0.0021, "step": 64930 }, { "grad_norm": 0.06623941659927368, "learning_rate": 3.000763598635109e-05, "loss": 0.0021, "step": 64940 }, { "grad_norm": 0.03784627839922905, "learning_rate": 2.999248167677098e-05, "loss": 0.002, "step": 64950 }, { "grad_norm": 0.06700526177883148, "learning_rate": 2.9977329555182976e-05, "loss": 0.0019, "step": 64960 }, { "grad_norm": 0.04825190454721451, "learning_rate": 2.9962179623244124e-05, "loss": 0.0016, "step": 64970 }, { "grad_norm": 0.09058961272239685, "learning_rate": 2.9947031882611175e-05, "loss": 0.0033, "step": 64980 }, { "grad_norm": 0.060177817940711975, "learning_rate": 2.9931886334940685e-05, "loss": 0.0029, "step": 64990 }, { "grad_norm": 0.05150635540485382, "learning_rate": 2.9916742981888923e-05, "loss": 0.0023, "step": 65000 }, { "grad_norm": 0.06982320547103882, "learning_rate": 2.990160182511196e-05, "loss": 0.0019, "step": 65010 }, { "grad_norm": 0.0555846244096756, "learning_rate": 2.9886462866265574e-05, "loss": 0.0025, "step": 65020 }, { "grad_norm": 0.08458809554576874, "learning_rate": 2.9871326107005392e-05, "loss": 0.0022, "step": 65030 }, { "grad_norm": 0.05517042428255081, "learning_rate": 2.9856191548986735e-05, "loss": 0.002, "step": 65040 }, { "grad_norm": 0.09201088547706604, "learning_rate": 2.9841059193864672e-05, "loss": 0.0019, "step": 65050 }, { "grad_norm": 0.06342591345310211, "learning_rate": 2.982592904329407e-05, "loss": 0.002, "step": 65060 }, { "grad_norm": 0.06169149652123451, "learning_rate": 2.981080109892954e-05, "loss": 0.0022, "step": 65070 }, { "grad_norm": 0.06997790187597275, "learning_rate": 2.9795675362425434e-05, "loss": 0.0024, "step": 65080 }, { "grad_norm": 0.05609486997127533, "learning_rate": 2.97805518354359e-05, "loss": 0.0021, "step": 65090 }, { "grad_norm": 0.0581878237426281, "learning_rate": 2.9765430519614825e-05, "loss": 0.0018, "step": 65100 }, { "grad_norm": 0.053027115762233734, "learning_rate": 2.975031141661584e-05, "loss": 0.0018, "step": 65110 }, { "grad_norm": 0.05169885233044624, "learning_rate": 2.9735194528092347e-05, "loss": 0.0013, "step": 65120 }, { "grad_norm": 0.05550776422023773, "learning_rate": 2.9720079855697525e-05, "loss": 0.002, "step": 65130 }, { "grad_norm": 0.08336304128170013, "learning_rate": 2.970496740108427e-05, "loss": 0.002, "step": 65140 }, { "grad_norm": 0.030148949474096298, "learning_rate": 2.968985716590527e-05, "loss": 0.0018, "step": 65150 }, { "grad_norm": 0.035686004906892776, "learning_rate": 2.9674749151812953e-05, "loss": 0.0019, "step": 65160 }, { "grad_norm": 0.04469071701169014, "learning_rate": 2.965964336045951e-05, "loss": 0.0017, "step": 65170 }, { "grad_norm": 0.06460228562355042, "learning_rate": 2.9644539793496894e-05, "loss": 0.0019, "step": 65180 }, { "grad_norm": 0.0638979971408844, "learning_rate": 2.9629438452576807e-05, "loss": 0.0019, "step": 65190 }, { "grad_norm": 0.04722944274544716, "learning_rate": 2.9614339339350698e-05, "loss": 0.002, "step": 65200 }, { "grad_norm": 0.06555979698896408, "learning_rate": 2.959924245546979e-05, "loss": 0.002, "step": 65210 }, { "grad_norm": 0.04755803570151329, "learning_rate": 2.9584147802585038e-05, "loss": 0.0019, "step": 65220 }, { "grad_norm": 0.05774260684847832, "learning_rate": 2.956905538234721e-05, "loss": 0.0021, "step": 65230 }, { "grad_norm": 0.056755971163511276, "learning_rate": 2.9553965196406764e-05, "loss": 0.0018, "step": 65240 }, { "grad_norm": 0.04819377511739731, "learning_rate": 2.9538877246413943e-05, "loss": 0.0026, "step": 65250 }, { "grad_norm": 0.0627211257815361, "learning_rate": 2.952379153401874e-05, "loss": 0.0022, "step": 65260 }, { "grad_norm": 0.06057009473443031, "learning_rate": 2.9508708060870914e-05, "loss": 0.0017, "step": 65270 }, { "grad_norm": 0.051126688718795776, "learning_rate": 2.9493626828619948e-05, "loss": 0.0018, "step": 65280 }, { "grad_norm": 0.0648510754108429, "learning_rate": 2.9478547838915137e-05, "loss": 0.003, "step": 65290 }, { "grad_norm": 0.07158311456441879, "learning_rate": 2.946347109340548e-05, "loss": 0.0022, "step": 65300 }, { "grad_norm": 0.0718952864408493, "learning_rate": 2.944839659373976e-05, "loss": 0.002, "step": 65310 }, { "grad_norm": 0.061100609600543976, "learning_rate": 2.943332434156647e-05, "loss": 0.0032, "step": 65320 }, { "grad_norm": 0.04347573220729828, "learning_rate": 2.9418254338533912e-05, "loss": 0.0019, "step": 65330 }, { "grad_norm": 0.04906696453690529, "learning_rate": 2.940318658629011e-05, "loss": 0.0022, "step": 65340 }, { "grad_norm": 0.05197790637612343, "learning_rate": 2.938812108648287e-05, "loss": 0.0021, "step": 65350 }, { "grad_norm": 0.04925227537751198, "learning_rate": 2.937305784075972e-05, "loss": 0.0017, "step": 65360 }, { "grad_norm": 0.04081159830093384, "learning_rate": 2.9357996850767937e-05, "loss": 0.0022, "step": 65370 }, { "grad_norm": 0.055034954100847244, "learning_rate": 2.9342938118154595e-05, "loss": 0.0015, "step": 65380 }, { "grad_norm": 0.04965847358107567, "learning_rate": 2.9327881644566485e-05, "loss": 0.0016, "step": 65390 }, { "grad_norm": 0.05297944322228432, "learning_rate": 2.9312827431650157e-05, "loss": 0.0019, "step": 65400 }, { "grad_norm": 0.05803609639406204, "learning_rate": 2.92977754810519e-05, "loss": 0.0022, "step": 65410 }, { "grad_norm": 0.1428464949131012, "learning_rate": 2.928272579441781e-05, "loss": 0.0024, "step": 65420 }, { "grad_norm": 0.07108637690544128, "learning_rate": 2.9267678373393676e-05, "loss": 0.0025, "step": 65430 }, { "grad_norm": 0.04693666845560074, "learning_rate": 2.925263321962507e-05, "loss": 0.0021, "step": 65440 }, { "grad_norm": 0.07850196957588196, "learning_rate": 2.9237590334757292e-05, "loss": 0.0026, "step": 65450 }, { "grad_norm": 0.07584811747074127, "learning_rate": 2.922254972043543e-05, "loss": 0.0021, "step": 65460 }, { "grad_norm": 0.04530392959713936, "learning_rate": 2.9207511378304265e-05, "loss": 0.0023, "step": 65470 }, { "grad_norm": 0.041970428079366684, "learning_rate": 2.9192475310008428e-05, "loss": 0.0022, "step": 65480 }, { "grad_norm": 0.13739065825939178, "learning_rate": 2.917744151719219e-05, "loss": 0.002, "step": 65490 }, { "grad_norm": 0.10791271924972534, "learning_rate": 2.916241000149964e-05, "loss": 0.002, "step": 65500 }, { "grad_norm": 0.14707709848880768, "learning_rate": 2.9147380764574622e-05, "loss": 0.0027, "step": 65510 }, { "grad_norm": 0.05544043704867363, "learning_rate": 2.9132353808060674e-05, "loss": 0.0016, "step": 65520 }, { "grad_norm": 0.06123910844326019, "learning_rate": 2.9117329133601134e-05, "loss": 0.0016, "step": 65530 }, { "grad_norm": 0.05851234123110771, "learning_rate": 2.9102306742839093e-05, "loss": 0.0023, "step": 65540 }, { "grad_norm": 0.054240334779024124, "learning_rate": 2.9087286637417375e-05, "loss": 0.0017, "step": 65550 }, { "grad_norm": 0.05308850109577179, "learning_rate": 2.9072268818978537e-05, "loss": 0.003, "step": 65560 }, { "grad_norm": 0.19846339523792267, "learning_rate": 2.905725328916493e-05, "loss": 0.002, "step": 65570 }, { "grad_norm": 0.03224950656294823, "learning_rate": 2.9042240049618608e-05, "loss": 0.0018, "step": 65580 }, { "grad_norm": 0.04708007350564003, "learning_rate": 2.9027229101981413e-05, "loss": 0.0018, "step": 65590 }, { "grad_norm": 0.11100712418556213, "learning_rate": 2.901222044789488e-05, "loss": 0.0018, "step": 65600 }, { "grad_norm": 0.07436097413301468, "learning_rate": 2.89972140890004e-05, "loss": 0.002, "step": 65610 }, { "grad_norm": 0.08248503506183624, "learning_rate": 2.8982210026938988e-05, "loss": 0.0022, "step": 65620 }, { "grad_norm": 0.05467642471194267, "learning_rate": 2.896720826335151e-05, "loss": 0.0036, "step": 65630 }, { "grad_norm": 0.0921575054526329, "learning_rate": 2.8952208799878493e-05, "loss": 0.0025, "step": 65640 }, { "grad_norm": 0.06819546967744827, "learning_rate": 2.89372116381603e-05, "loss": 0.0017, "step": 65650 }, { "grad_norm": 0.042410194873809814, "learning_rate": 2.8922216779836932e-05, "loss": 0.0023, "step": 65660 }, { "grad_norm": 0.05651569366455078, "learning_rate": 2.8907224226548285e-05, "loss": 0.0037, "step": 65670 }, { "grad_norm": 0.10278582572937012, "learning_rate": 2.8892233979933868e-05, "loss": 0.0029, "step": 65680 }, { "grad_norm": 0.06964126974344254, "learning_rate": 2.8877246041633026e-05, "loss": 0.0023, "step": 65690 }, { "grad_norm": 0.07565665990114212, "learning_rate": 2.886226041328478e-05, "loss": 0.0026, "step": 65700 }, { "grad_norm": 0.05907013639807701, "learning_rate": 2.8847277096527957e-05, "loss": 0.002, "step": 65710 }, { "grad_norm": 0.0638597384095192, "learning_rate": 2.8832296093001105e-05, "loss": 0.003, "step": 65720 }, { "grad_norm": 0.06688190251588821, "learning_rate": 2.8817317404342547e-05, "loss": 0.0019, "step": 65730 }, { "grad_norm": 0.0540696382522583, "learning_rate": 2.880234103219029e-05, "loss": 0.0017, "step": 65740 }, { "grad_norm": 0.057796310633420944, "learning_rate": 2.878736697818215e-05, "loss": 0.0018, "step": 65750 }, { "grad_norm": 0.06617115437984467, "learning_rate": 2.877239524395568e-05, "loss": 0.0022, "step": 65760 }, { "grad_norm": 0.05036496743559837, "learning_rate": 2.8757425831148134e-05, "loss": 0.0021, "step": 65770 }, { "grad_norm": 0.05255606397986412, "learning_rate": 2.874245874139658e-05, "loss": 0.0024, "step": 65780 }, { "grad_norm": 0.05862957239151001, "learning_rate": 2.8727493976337737e-05, "loss": 0.0019, "step": 65790 }, { "grad_norm": 0.047660693526268005, "learning_rate": 2.8712531537608218e-05, "loss": 0.0019, "step": 65800 }, { "grad_norm": 0.04906298220157623, "learning_rate": 2.8697571426844217e-05, "loss": 0.0018, "step": 65810 }, { "grad_norm": 0.06928658485412598, "learning_rate": 2.86826136456818e-05, "loss": 0.0018, "step": 65820 }, { "grad_norm": 0.042259663343429565, "learning_rate": 2.8667658195756685e-05, "loss": 0.0017, "step": 65830 }, { "grad_norm": 0.05441482737660408, "learning_rate": 2.8652705078704412e-05, "loss": 0.0019, "step": 65840 }, { "grad_norm": 0.040062323212623596, "learning_rate": 2.863775429616018e-05, "loss": 0.0015, "step": 65850 }, { "grad_norm": 0.06536075472831726, "learning_rate": 2.862280584975906e-05, "loss": 0.0016, "step": 65860 }, { "grad_norm": 0.039556149393320084, "learning_rate": 2.8607859741135735e-05, "loss": 0.0018, "step": 65870 }, { "grad_norm": 0.03637095168232918, "learning_rate": 2.8592915971924723e-05, "loss": 0.0026, "step": 65880 }, { "grad_norm": 0.06412703543901443, "learning_rate": 2.8577974543760212e-05, "loss": 0.0025, "step": 65890 }, { "grad_norm": 0.0684737041592598, "learning_rate": 2.8563035458276222e-05, "loss": 0.0029, "step": 65900 }, { "grad_norm": 0.05013043433427811, "learning_rate": 2.8548098717106424e-05, "loss": 0.0019, "step": 65910 }, { "grad_norm": 0.04509204626083374, "learning_rate": 2.85331643218843e-05, "loss": 0.0026, "step": 65920 }, { "grad_norm": 0.06366253644227982, "learning_rate": 2.8518232274243046e-05, "loss": 0.0022, "step": 65930 }, { "grad_norm": 0.04895211011171341, "learning_rate": 2.8503302575815638e-05, "loss": 0.0017, "step": 65940 }, { "grad_norm": 0.039565928280353546, "learning_rate": 2.8488375228234716e-05, "loss": 0.0016, "step": 65950 }, { "grad_norm": 0.05622095614671707, "learning_rate": 2.8473450233132738e-05, "loss": 0.0014, "step": 65960 }, { "grad_norm": 0.0965440645813942, "learning_rate": 2.8458527592141898e-05, "loss": 0.002, "step": 65970 }, { "grad_norm": 0.06023783981800079, "learning_rate": 2.8443607306894073e-05, "loss": 0.0021, "step": 65980 }, { "grad_norm": 0.06241302564740181, "learning_rate": 2.8428689379020946e-05, "loss": 0.002, "step": 65990 }, { "grad_norm": 0.06532598286867142, "learning_rate": 2.841377381015391e-05, "loss": 0.0029, "step": 66000 }, { "grad_norm": 0.04569448530673981, "learning_rate": 2.8398860601924144e-05, "loss": 0.0015, "step": 66010 }, { "grad_norm": 0.042035944759845734, "learning_rate": 2.838394975596248e-05, "loss": 0.0018, "step": 66020 }, { "grad_norm": 0.04312746226787567, "learning_rate": 2.83690412738996e-05, "loss": 0.0016, "step": 66030 }, { "grad_norm": 0.06943462789058685, "learning_rate": 2.835413515736579e-05, "loss": 0.0025, "step": 66040 }, { "grad_norm": 0.06260114163160324, "learning_rate": 2.8339231407991267e-05, "loss": 0.002, "step": 66050 }, { "grad_norm": 0.09586107730865479, "learning_rate": 2.832433002740581e-05, "loss": 0.0021, "step": 66060 }, { "grad_norm": 0.06536206603050232, "learning_rate": 2.8309431017239047e-05, "loss": 0.0021, "step": 66070 }, { "grad_norm": 0.07092438638210297, "learning_rate": 2.829453437912028e-05, "loss": 0.0019, "step": 66080 }, { "grad_norm": 0.07025209814310074, "learning_rate": 2.8279640114678617e-05, "loss": 0.0027, "step": 66090 }, { "grad_norm": 0.04869865998625755, "learning_rate": 2.8264748225542836e-05, "loss": 0.0019, "step": 66100 }, { "grad_norm": 0.06136192008852959, "learning_rate": 2.8249858713341516e-05, "loss": 0.002, "step": 66110 }, { "grad_norm": 0.04972183704376221, "learning_rate": 2.823497157970294e-05, "loss": 0.0019, "step": 66120 }, { "grad_norm": 0.05884940177202225, "learning_rate": 2.8220086826255164e-05, "loss": 0.0018, "step": 66130 }, { "grad_norm": 0.09720799326896667, "learning_rate": 2.8205204454625934e-05, "loss": 0.0022, "step": 66140 }, { "grad_norm": 0.06551020592451096, "learning_rate": 2.8190324466442798e-05, "loss": 0.0027, "step": 66150 }, { "grad_norm": 0.0665360540151596, "learning_rate": 2.8175446863332965e-05, "loss": 0.0021, "step": 66160 }, { "grad_norm": 0.05123293399810791, "learning_rate": 2.816057164692345e-05, "loss": 0.0023, "step": 66170 }, { "grad_norm": 0.06118084862828255, "learning_rate": 2.8145698818840992e-05, "loss": 0.002, "step": 66180 }, { "grad_norm": 0.030673174187541008, "learning_rate": 2.8130828380712065e-05, "loss": 0.0024, "step": 66190 }, { "grad_norm": 0.04579118266701698, "learning_rate": 2.811596033416285e-05, "loss": 0.0026, "step": 66200 }, { "grad_norm": 0.042297035455703735, "learning_rate": 2.8101094680819307e-05, "loss": 0.0019, "step": 66210 }, { "grad_norm": 0.05770869553089142, "learning_rate": 2.808623142230715e-05, "loss": 0.0018, "step": 66220 }, { "grad_norm": 0.04617978259921074, "learning_rate": 2.8071370560251757e-05, "loss": 0.0014, "step": 66230 }, { "grad_norm": 0.037597399204969406, "learning_rate": 2.805651209627831e-05, "loss": 0.0017, "step": 66240 }, { "grad_norm": 0.05545184016227722, "learning_rate": 2.804165603201171e-05, "loss": 0.0019, "step": 66250 }, { "grad_norm": 0.06017608940601349, "learning_rate": 2.802680236907661e-05, "loss": 0.0023, "step": 66260 }, { "grad_norm": 0.06325783580541611, "learning_rate": 2.8011951109097344e-05, "loss": 0.0022, "step": 66270 }, { "grad_norm": 0.0790218636393547, "learning_rate": 2.7997102253698072e-05, "loss": 0.002, "step": 66280 }, { "grad_norm": 0.04390878602862358, "learning_rate": 2.7982255804502588e-05, "loss": 0.0021, "step": 66290 }, { "grad_norm": 0.04406167194247246, "learning_rate": 2.7967411763134516e-05, "loss": 0.0024, "step": 66300 }, { "grad_norm": 0.06107530742883682, "learning_rate": 2.7952570131217166e-05, "loss": 0.0019, "step": 66310 }, { "grad_norm": 0.046114347875118256, "learning_rate": 2.7937730910373618e-05, "loss": 0.0019, "step": 66320 }, { "grad_norm": 0.06099343299865723, "learning_rate": 2.7922894102226628e-05, "loss": 0.0018, "step": 66330 }, { "grad_norm": 0.0648796558380127, "learning_rate": 2.790805970839876e-05, "loss": 0.0022, "step": 66340 }, { "grad_norm": 0.04648321866989136, "learning_rate": 2.789322773051225e-05, "loss": 0.0017, "step": 66350 }, { "grad_norm": 0.0517583005130291, "learning_rate": 2.7878398170189124e-05, "loss": 0.0016, "step": 66360 }, { "grad_norm": 0.04258599132299423, "learning_rate": 2.7863571029051112e-05, "loss": 0.0014, "step": 66370 }, { "grad_norm": 0.06495900452136993, "learning_rate": 2.784874630871971e-05, "loss": 0.0016, "step": 66380 }, { "grad_norm": 0.049215178936719894, "learning_rate": 2.7833924010816086e-05, "loss": 0.0016, "step": 66390 }, { "grad_norm": 0.04387679696083069, "learning_rate": 2.7819104136961226e-05, "loss": 0.0028, "step": 66400 }, { "grad_norm": 0.053021274507045746, "learning_rate": 2.780428668877577e-05, "loss": 0.0016, "step": 66410 }, { "grad_norm": 0.06471845507621765, "learning_rate": 2.7789471667880152e-05, "loss": 0.0018, "step": 66420 }, { "grad_norm": 0.060005586594343185, "learning_rate": 2.7774659075894523e-05, "loss": 0.0022, "step": 66430 }, { "grad_norm": 0.04937873035669327, "learning_rate": 2.7759848914438778e-05, "loss": 0.0014, "step": 66440 }, { "grad_norm": 0.04432276263833046, "learning_rate": 2.7745041185132496e-05, "loss": 0.0014, "step": 66450 }, { "grad_norm": 0.061024006456136703, "learning_rate": 2.773023588959506e-05, "loss": 0.0018, "step": 66460 }, { "grad_norm": 0.09362856298685074, "learning_rate": 2.7715433029445555e-05, "loss": 0.0022, "step": 66470 }, { "grad_norm": 0.06485117971897125, "learning_rate": 2.7700632606302766e-05, "loss": 0.0023, "step": 66480 }, { "grad_norm": 0.1582833081483841, "learning_rate": 2.768583462178528e-05, "loss": 0.0021, "step": 66490 }, { "grad_norm": 0.0633242279291153, "learning_rate": 2.767103907751136e-05, "loss": 0.0018, "step": 66500 }, { "grad_norm": 0.05454694479703903, "learning_rate": 2.7656245975099065e-05, "loss": 0.002, "step": 66510 }, { "grad_norm": 0.06034671887755394, "learning_rate": 2.7641455316166093e-05, "loss": 0.0026, "step": 66520 }, { "grad_norm": 0.08167394250631332, "learning_rate": 2.7626667102329973e-05, "loss": 0.0026, "step": 66530 }, { "grad_norm": 0.049546126276254654, "learning_rate": 2.761188133520788e-05, "loss": 0.0019, "step": 66540 }, { "grad_norm": 0.06288273632526398, "learning_rate": 2.7597098016416784e-05, "loss": 0.0017, "step": 66550 }, { "grad_norm": 0.04787367582321167, "learning_rate": 2.7582317147573368e-05, "loss": 0.0021, "step": 66560 }, { "grad_norm": 0.04455254599452019, "learning_rate": 2.7567538730294064e-05, "loss": 0.0019, "step": 66570 }, { "grad_norm": 0.09019068628549576, "learning_rate": 2.7552762766194973e-05, "loss": 0.0021, "step": 66580 }, { "grad_norm": 0.07018487900495529, "learning_rate": 2.753798925689202e-05, "loss": 0.0023, "step": 66590 }, { "grad_norm": 0.05286628007888794, "learning_rate": 2.7523218204000774e-05, "loss": 0.0016, "step": 66600 }, { "grad_norm": 0.053467217832803726, "learning_rate": 2.7508449609136612e-05, "loss": 0.0029, "step": 66610 }, { "grad_norm": 0.0503842756152153, "learning_rate": 2.749368347391454e-05, "loss": 0.002, "step": 66620 }, { "grad_norm": 0.06795688718557358, "learning_rate": 2.7478919799949443e-05, "loss": 0.0028, "step": 66630 }, { "grad_norm": 0.0555541031062603, "learning_rate": 2.7464158588855803e-05, "loss": 0.0022, "step": 66640 }, { "grad_norm": 0.059689007699489594, "learning_rate": 2.7449399842247918e-05, "loss": 0.0021, "step": 66650 }, { "grad_norm": 0.05486374720931053, "learning_rate": 2.7434643561739736e-05, "loss": 0.0018, "step": 66660 }, { "grad_norm": 0.0600571371614933, "learning_rate": 2.7419889748945016e-05, "loss": 0.0021, "step": 66670 }, { "grad_norm": 0.041483987122774124, "learning_rate": 2.7405138405477203e-05, "loss": 0.0016, "step": 66680 }, { "grad_norm": 0.058346498757600784, "learning_rate": 2.73903895329495e-05, "loss": 0.0021, "step": 66690 }, { "grad_norm": 0.055588435381650925, "learning_rate": 2.737564313297478e-05, "loss": 0.0019, "step": 66700 }, { "grad_norm": 0.05319878086447716, "learning_rate": 2.736089920716572e-05, "loss": 0.0016, "step": 66710 }, { "grad_norm": 0.05305474251508713, "learning_rate": 2.7346157757134706e-05, "loss": 0.0019, "step": 66720 }, { "grad_norm": 0.05139468237757683, "learning_rate": 2.733141878449379e-05, "loss": 0.0018, "step": 66730 }, { "grad_norm": 0.04600853472948074, "learning_rate": 2.7316682290854844e-05, "loss": 0.0018, "step": 66740 }, { "grad_norm": 0.037779100239276886, "learning_rate": 2.7301948277829414e-05, "loss": 0.0023, "step": 66750 }, { "grad_norm": 0.041686780750751495, "learning_rate": 2.7287216747028814e-05, "loss": 0.0017, "step": 66760 }, { "grad_norm": 0.036184344440698624, "learning_rate": 2.7272487700064025e-05, "loss": 0.0014, "step": 66770 }, { "grad_norm": 0.07089050114154816, "learning_rate": 2.725776113854583e-05, "loss": 0.0017, "step": 66780 }, { "grad_norm": 0.06204596161842346, "learning_rate": 2.7243037064084664e-05, "loss": 0.0016, "step": 66790 }, { "grad_norm": 0.08043587952852249, "learning_rate": 2.7228315478290766e-05, "loss": 0.0017, "step": 66800 }, { "grad_norm": 0.049006037414073944, "learning_rate": 2.721359638277402e-05, "loss": 0.0019, "step": 66810 }, { "grad_norm": 0.05901753157377243, "learning_rate": 2.7198879779144148e-05, "loss": 0.0018, "step": 66820 }, { "grad_norm": 0.05794376879930496, "learning_rate": 2.7184165669010485e-05, "loss": 0.002, "step": 66830 }, { "grad_norm": 0.04510270059108734, "learning_rate": 2.716945405398218e-05, "loss": 0.0021, "step": 66840 }, { "grad_norm": 0.04358339309692383, "learning_rate": 2.715474493566803e-05, "loss": 0.0015, "step": 66850 }, { "grad_norm": 0.06301245838403702, "learning_rate": 2.7140038315676652e-05, "loss": 0.0016, "step": 66860 }, { "grad_norm": 0.03733198717236519, "learning_rate": 2.7125334195616275e-05, "loss": 0.0016, "step": 66870 }, { "grad_norm": 0.046408168971538544, "learning_rate": 2.7110632577094997e-05, "loss": 0.0018, "step": 66880 }, { "grad_norm": 0.06945376098155975, "learning_rate": 2.709593346172051e-05, "loss": 0.0019, "step": 66890 }, { "grad_norm": 0.05898163840174675, "learning_rate": 2.708123685110032e-05, "loss": 0.0016, "step": 66900 }, { "grad_norm": 0.07309342920780182, "learning_rate": 2.7066542746841588e-05, "loss": 0.0019, "step": 66910 }, { "grad_norm": 0.042020149528980255, "learning_rate": 2.7051851150551267e-05, "loss": 0.002, "step": 66920 }, { "grad_norm": 0.05390989035367966, "learning_rate": 2.7037162063835996e-05, "loss": 0.0024, "step": 66930 }, { "grad_norm": 0.0853402242064476, "learning_rate": 2.702247548830218e-05, "loss": 0.0024, "step": 66940 }, { "grad_norm": 0.08371204882860184, "learning_rate": 2.7007791425555883e-05, "loss": 0.002, "step": 66950 }, { "grad_norm": 0.07017886638641357, "learning_rate": 2.6993109877202945e-05, "loss": 0.0019, "step": 66960 }, { "grad_norm": 0.06396963447332382, "learning_rate": 2.6978430844848935e-05, "loss": 0.0017, "step": 66970 }, { "grad_norm": 0.042347341775894165, "learning_rate": 2.69637543300991e-05, "loss": 0.0019, "step": 66980 }, { "grad_norm": 0.07245160639286041, "learning_rate": 2.694908033455848e-05, "loss": 0.0019, "step": 66990 }, { "grad_norm": 0.0651409700512886, "learning_rate": 2.6934408859831733e-05, "loss": 0.0021, "step": 67000 }, { "grad_norm": 0.04891689494252205, "learning_rate": 2.69197399075234e-05, "loss": 0.0015, "step": 67010 }, { "grad_norm": 0.05286368727684021, "learning_rate": 2.6905073479237584e-05, "loss": 0.0015, "step": 67020 }, { "grad_norm": 0.05026590824127197, "learning_rate": 2.6890409576578235e-05, "loss": 0.0016, "step": 67030 }, { "grad_norm": 0.041448257863521576, "learning_rate": 2.687574820114892e-05, "loss": 0.0017, "step": 67040 }, { "grad_norm": 0.061509277671575546, "learning_rate": 2.686108935455305e-05, "loss": 0.0022, "step": 67050 }, { "grad_norm": 0.08112379908561707, "learning_rate": 2.6846433038393603e-05, "loss": 0.0021, "step": 67060 }, { "grad_norm": 0.0653320923447609, "learning_rate": 2.6831779254273482e-05, "loss": 0.0022, "step": 67070 }, { "grad_norm": 0.040473226457834244, "learning_rate": 2.6817128003795122e-05, "loss": 0.0015, "step": 67080 }, { "grad_norm": 0.04987616464495659, "learning_rate": 2.680247928856081e-05, "loss": 0.0017, "step": 67090 }, { "grad_norm": 0.073650062084198, "learning_rate": 2.6787833110172466e-05, "loss": 0.0019, "step": 67100 }, { "grad_norm": 0.07653188705444336, "learning_rate": 2.6773189470231807e-05, "loss": 0.0035, "step": 67110 }, { "grad_norm": 0.06569796800613403, "learning_rate": 2.675854837034022e-05, "loss": 0.0017, "step": 67120 }, { "grad_norm": 0.0692516341805458, "learning_rate": 2.6743909812098833e-05, "loss": 0.0019, "step": 67130 }, { "grad_norm": 0.06996948271989822, "learning_rate": 2.67292737971085e-05, "loss": 0.002, "step": 67140 }, { "grad_norm": 0.05747554823756218, "learning_rate": 2.671464032696982e-05, "loss": 0.0017, "step": 67150 }, { "grad_norm": 0.04682212695479393, "learning_rate": 2.6700009403283055e-05, "loss": 0.0023, "step": 67160 }, { "grad_norm": 0.05710683390498161, "learning_rate": 2.6685381027648214e-05, "loss": 0.0035, "step": 67170 }, { "grad_norm": 0.04582592099905014, "learning_rate": 2.6670755201665086e-05, "loss": 0.0027, "step": 67180 }, { "grad_norm": 0.06741289794445038, "learning_rate": 2.6656131926933066e-05, "loss": 0.0025, "step": 67190 }, { "grad_norm": 0.06890217959880829, "learning_rate": 2.664151120505136e-05, "loss": 0.0021, "step": 67200 }, { "grad_norm": 0.054568272083997726, "learning_rate": 2.6626893037618873e-05, "loss": 0.0024, "step": 67210 }, { "grad_norm": 0.04855673387646675, "learning_rate": 2.6612277426234243e-05, "loss": 0.0019, "step": 67220 }, { "grad_norm": 0.06634856015443802, "learning_rate": 2.659766437249577e-05, "loss": 0.0027, "step": 67230 }, { "grad_norm": 0.05105385556817055, "learning_rate": 2.6583053878001557e-05, "loss": 0.0019, "step": 67240 }, { "grad_norm": 0.07101278007030487, "learning_rate": 2.6568445944349323e-05, "loss": 0.0022, "step": 67250 }, { "grad_norm": 0.0484524667263031, "learning_rate": 2.6553840573136657e-05, "loss": 0.0017, "step": 67260 }, { "grad_norm": 0.038310207426548004, "learning_rate": 2.6539237765960706e-05, "loss": 0.0016, "step": 67270 }, { "grad_norm": 0.06484746932983398, "learning_rate": 2.6524637524418465e-05, "loss": 0.0024, "step": 67280 }, { "grad_norm": 0.0671195313334465, "learning_rate": 2.651003985010655e-05, "loss": 0.0023, "step": 67290 }, { "grad_norm": 0.07517372071743011, "learning_rate": 2.649544474462138e-05, "loss": 0.002, "step": 67300 }, { "grad_norm": 0.07357975840568542, "learning_rate": 2.6480852209559015e-05, "loss": 0.0024, "step": 67310 }, { "grad_norm": 0.0950268805027008, "learning_rate": 2.6466262246515296e-05, "loss": 0.0024, "step": 67320 }, { "grad_norm": 0.05551545321941376, "learning_rate": 2.6451674857085746e-05, "loss": 0.0021, "step": 67330 }, { "grad_norm": 0.062324944883584976, "learning_rate": 2.6437090042865655e-05, "loss": 0.0022, "step": 67340 }, { "grad_norm": 0.042209964245557785, "learning_rate": 2.6422507805449947e-05, "loss": 0.0026, "step": 67350 }, { "grad_norm": 0.058180514723062515, "learning_rate": 2.6407928146433357e-05, "loss": 0.0021, "step": 67360 }, { "grad_norm": 0.034232813864946365, "learning_rate": 2.639335106741026e-05, "loss": 0.0015, "step": 67370 }, { "grad_norm": 0.05135868862271309, "learning_rate": 2.6378776569974796e-05, "loss": 0.002, "step": 67380 }, { "grad_norm": 0.04247448965907097, "learning_rate": 2.6364204655720814e-05, "loss": 0.0016, "step": 67390 }, { "grad_norm": 0.06181855499744415, "learning_rate": 2.6349635326241894e-05, "loss": 0.0018, "step": 67400 }, { "grad_norm": 0.06840045750141144, "learning_rate": 2.633506858313127e-05, "loss": 0.002, "step": 67410 }, { "grad_norm": 0.0682496652007103, "learning_rate": 2.6320504427981975e-05, "loss": 0.0022, "step": 67420 }, { "grad_norm": 0.07869452238082886, "learning_rate": 2.6305942862386734e-05, "loss": 0.0022, "step": 67430 }, { "grad_norm": 0.060560766607522964, "learning_rate": 2.629138388793793e-05, "loss": 0.0021, "step": 67440 }, { "grad_norm": 0.04642724618315697, "learning_rate": 2.6276827506227743e-05, "loss": 0.0019, "step": 67450 }, { "grad_norm": 0.057409558445215225, "learning_rate": 2.6262273718848037e-05, "loss": 0.0019, "step": 67460 }, { "grad_norm": 0.0603427030146122, "learning_rate": 2.6247722527390407e-05, "loss": 0.0016, "step": 67470 }, { "grad_norm": 0.04608604684472084, "learning_rate": 2.6233173933446104e-05, "loss": 0.0013, "step": 67480 }, { "grad_norm": 0.041627541184425354, "learning_rate": 2.6218627938606188e-05, "loss": 0.0016, "step": 67490 }, { "grad_norm": 0.060069240629673004, "learning_rate": 2.620408454446135e-05, "loss": 0.0018, "step": 67500 }, { "grad_norm": 0.06027461588382721, "learning_rate": 2.6189543752602046e-05, "loss": 0.0021, "step": 67510 }, { "grad_norm": 0.05031098425388336, "learning_rate": 2.6175005564618444e-05, "loss": 0.0025, "step": 67520 }, { "grad_norm": 0.08806183934211731, "learning_rate": 2.6160469982100428e-05, "loss": 0.0017, "step": 67530 }, { "grad_norm": 0.05340227484703064, "learning_rate": 2.6145937006637555e-05, "loss": 0.0019, "step": 67540 }, { "grad_norm": 0.08111587166786194, "learning_rate": 2.6131406639819166e-05, "loss": 0.0018, "step": 67550 }, { "grad_norm": 0.06778489798307419, "learning_rate": 2.611687888323424e-05, "loss": 0.0017, "step": 67560 }, { "grad_norm": 0.055842574685811996, "learning_rate": 2.610235373847153e-05, "loss": 0.0013, "step": 67570 }, { "grad_norm": 0.05116450786590576, "learning_rate": 2.6087831207119494e-05, "loss": 0.0019, "step": 67580 }, { "grad_norm": 0.04000948369503021, "learning_rate": 2.60733112907663e-05, "loss": 0.002, "step": 67590 }, { "grad_norm": 0.036660145968198776, "learning_rate": 2.6058793990999786e-05, "loss": 0.0018, "step": 67600 }, { "grad_norm": 0.048301029950380325, "learning_rate": 2.6044279309407593e-05, "loss": 0.002, "step": 67610 }, { "grad_norm": 0.05173024907708168, "learning_rate": 2.6029767247576976e-05, "loss": 0.002, "step": 67620 }, { "grad_norm": 0.08976869285106659, "learning_rate": 2.601525780709497e-05, "loss": 0.0014, "step": 67630 }, { "grad_norm": 0.0853143110871315, "learning_rate": 2.6000750989548318e-05, "loss": 0.0018, "step": 67640 }, { "grad_norm": 0.06181728094816208, "learning_rate": 2.5986246796523462e-05, "loss": 0.0017, "step": 67650 }, { "grad_norm": 0.0973532572388649, "learning_rate": 2.5971745229606546e-05, "loss": 0.0018, "step": 67660 }, { "grad_norm": 0.04203835502266884, "learning_rate": 2.5957246290383442e-05, "loss": 0.0015, "step": 67670 }, { "grad_norm": 0.06497140973806381, "learning_rate": 2.594274998043975e-05, "loss": 0.0023, "step": 67680 }, { "grad_norm": 0.03889366239309311, "learning_rate": 2.5928256301360733e-05, "loss": 0.0015, "step": 67690 }, { "grad_norm": 0.04272356256842613, "learning_rate": 2.591376525473142e-05, "loss": 0.0016, "step": 67700 }, { "grad_norm": 0.05388830602169037, "learning_rate": 2.5899276842136523e-05, "loss": 0.0022, "step": 67710 }, { "grad_norm": 0.04808368906378746, "learning_rate": 2.5884791065160495e-05, "loss": 0.0017, "step": 67720 }, { "grad_norm": 0.06775792688131332, "learning_rate": 2.5870307925387436e-05, "loss": 0.0017, "step": 67730 }, { "grad_norm": 0.08344108611345291, "learning_rate": 2.585582742440125e-05, "loss": 0.0022, "step": 67740 }, { "grad_norm": 0.06775316596031189, "learning_rate": 2.5841349563785455e-05, "loss": 0.002, "step": 67750 }, { "grad_norm": 0.04736749827861786, "learning_rate": 2.5826874345123355e-05, "loss": 0.0018, "step": 67760 }, { "grad_norm": 0.053434718400239944, "learning_rate": 2.581240176999792e-05, "loss": 0.0015, "step": 67770 }, { "grad_norm": 0.05120701715350151, "learning_rate": 2.579793183999189e-05, "loss": 0.0014, "step": 67780 }, { "grad_norm": 0.045157741755247116, "learning_rate": 2.5783464556687627e-05, "loss": 0.0019, "step": 67790 }, { "grad_norm": 0.05660662800073624, "learning_rate": 2.576899992166729e-05, "loss": 0.0019, "step": 67800 }, { "grad_norm": 0.04639868065714836, "learning_rate": 2.5754537936512668e-05, "loss": 0.0017, "step": 67810 }, { "grad_norm": 0.060024987906217575, "learning_rate": 2.574007860280535e-05, "loss": 0.0022, "step": 67820 }, { "grad_norm": 0.046221233904361725, "learning_rate": 2.5725621922126518e-05, "loss": 0.0014, "step": 67830 }, { "grad_norm": 0.060961462557315826, "learning_rate": 2.571116789605722e-05, "loss": 0.002, "step": 67840 }, { "grad_norm": 0.4630526602268219, "learning_rate": 2.5696716526178054e-05, "loss": 0.0018, "step": 67850 }, { "grad_norm": 0.06514602154493332, "learning_rate": 2.568226781406946e-05, "loss": 0.0021, "step": 67860 }, { "grad_norm": 0.04642432555556297, "learning_rate": 2.5667821761311474e-05, "loss": 0.0017, "step": 67870 }, { "grad_norm": 0.04967782273888588, "learning_rate": 2.5653378369483917e-05, "loss": 0.0017, "step": 67880 }, { "grad_norm": 0.10645262151956558, "learning_rate": 2.5638937640166292e-05, "loss": 0.0023, "step": 67890 }, { "grad_norm": 0.05124703422188759, "learning_rate": 2.5624499574937842e-05, "loss": 0.0023, "step": 67900 }, { "grad_norm": 0.04153243452310562, "learning_rate": 2.5610064175377456e-05, "loss": 0.0016, "step": 67910 }, { "grad_norm": 0.066195048391819, "learning_rate": 2.5595631443063782e-05, "loss": 0.0024, "step": 67920 }, { "grad_norm": 0.05594716593623161, "learning_rate": 2.5581201379575177e-05, "loss": 0.0026, "step": 67930 }, { "grad_norm": 0.07258529961109161, "learning_rate": 2.5566773986489655e-05, "loss": 0.0033, "step": 67940 }, { "grad_norm": 0.05090542510151863, "learning_rate": 2.5552349265385e-05, "loss": 0.002, "step": 67950 }, { "grad_norm": 0.06216766685247421, "learning_rate": 2.5537927217838675e-05, "loss": 0.0021, "step": 67960 }, { "grad_norm": 0.054907359182834625, "learning_rate": 2.5523507845427864e-05, "loss": 0.0015, "step": 67970 }, { "grad_norm": 0.0550679974257946, "learning_rate": 2.5509091149729413e-05, "loss": 0.0017, "step": 67980 }, { "grad_norm": 0.08296290785074234, "learning_rate": 2.5494677132319962e-05, "loss": 0.0025, "step": 67990 }, { "grad_norm": 0.06973034143447876, "learning_rate": 2.548026579477575e-05, "loss": 0.0021, "step": 68000 }, { "grad_norm": 0.06105402484536171, "learning_rate": 2.5465857138672834e-05, "loss": 0.002, "step": 68010 }, { "grad_norm": 0.05308759957551956, "learning_rate": 2.5451451165586848e-05, "loss": 0.0017, "step": 68020 }, { "grad_norm": 0.03325944021344185, "learning_rate": 2.543704787709329e-05, "loss": 0.0013, "step": 68030 }, { "grad_norm": 0.07538730651140213, "learning_rate": 2.542264727476723e-05, "loss": 0.0016, "step": 68040 }, { "grad_norm": 0.046411558985710144, "learning_rate": 2.5408249360183535e-05, "loss": 0.0017, "step": 68050 }, { "grad_norm": 0.06446634232997894, "learning_rate": 2.5393854134916694e-05, "loss": 0.0028, "step": 68060 }, { "grad_norm": 0.08986447751522064, "learning_rate": 2.5379461600540993e-05, "loss": 0.002, "step": 68070 }, { "grad_norm": 0.05374826118350029, "learning_rate": 2.5365071758630317e-05, "loss": 0.0017, "step": 68080 }, { "grad_norm": 0.047799576073884964, "learning_rate": 2.535068461075839e-05, "loss": 0.0014, "step": 68090 }, { "grad_norm": 0.05559144541621208, "learning_rate": 2.5336300158498516e-05, "loss": 0.0031, "step": 68100 }, { "grad_norm": 0.04396636784076691, "learning_rate": 2.5321918403423793e-05, "loss": 0.0019, "step": 68110 }, { "grad_norm": 0.051627472043037415, "learning_rate": 2.5307539347106957e-05, "loss": 0.0018, "step": 68120 }, { "grad_norm": 0.056230444461107254, "learning_rate": 2.529316299112049e-05, "loss": 0.0018, "step": 68130 }, { "grad_norm": 0.04769518971443176, "learning_rate": 2.5278789337036567e-05, "loss": 0.0014, "step": 68140 }, { "grad_norm": 0.04354875162243843, "learning_rate": 2.52644183864271e-05, "loss": 0.0015, "step": 68150 }, { "grad_norm": 0.05001814663410187, "learning_rate": 2.525005014086363e-05, "loss": 0.0018, "step": 68160 }, { "grad_norm": 0.06975600868463516, "learning_rate": 2.5235684601917465e-05, "loss": 0.0021, "step": 68170 }, { "grad_norm": 0.06629149615764618, "learning_rate": 2.5221321771159618e-05, "loss": 0.0025, "step": 68180 }, { "grad_norm": 0.06361204385757446, "learning_rate": 2.5206961650160743e-05, "loss": 0.0021, "step": 68190 }, { "grad_norm": 0.06606578081846237, "learning_rate": 2.519260424049129e-05, "loss": 0.0024, "step": 68200 }, { "grad_norm": 0.057190027087926865, "learning_rate": 2.5178249543721304e-05, "loss": 0.0018, "step": 68210 }, { "grad_norm": 0.06572483479976654, "learning_rate": 2.5163897561420668e-05, "loss": 0.0016, "step": 68220 }, { "grad_norm": 0.036092765629291534, "learning_rate": 2.514954829515883e-05, "loss": 0.0017, "step": 68230 }, { "grad_norm": 0.049772635102272034, "learning_rate": 2.513520174650505e-05, "loss": 0.0016, "step": 68240 }, { "grad_norm": 0.037123870104551315, "learning_rate": 2.5120857917028207e-05, "loss": 0.0014, "step": 68250 }, { "grad_norm": 0.05303535237908363, "learning_rate": 2.5106516808296953e-05, "loss": 0.0015, "step": 68260 }, { "grad_norm": 0.062239937484264374, "learning_rate": 2.5092178421879552e-05, "loss": 0.0019, "step": 68270 }, { "grad_norm": 0.17043161392211914, "learning_rate": 2.5077842759344116e-05, "loss": 0.0019, "step": 68280 }, { "grad_norm": 0.05714944005012512, "learning_rate": 2.5063509822258304e-05, "loss": 0.0016, "step": 68290 }, { "grad_norm": 0.07728936523199081, "learning_rate": 2.5049179612189576e-05, "loss": 0.0022, "step": 68300 }, { "grad_norm": 0.08021201193332672, "learning_rate": 2.5034852130705043e-05, "loss": 0.0023, "step": 68310 }, { "grad_norm": 0.06303548812866211, "learning_rate": 2.5020527379371572e-05, "loss": 0.0018, "step": 68320 }, { "grad_norm": 0.05580068752169609, "learning_rate": 2.500620535975564e-05, "loss": 0.0024, "step": 68330 }, { "grad_norm": 0.04853653162717819, "learning_rate": 2.4991886073423515e-05, "loss": 0.0014, "step": 68340 }, { "grad_norm": 0.08723629266023636, "learning_rate": 2.4977569521941134e-05, "loss": 0.0029, "step": 68350 }, { "grad_norm": 0.0393582321703434, "learning_rate": 2.4963255706874143e-05, "loss": 0.0021, "step": 68360 }, { "grad_norm": 0.03626960888504982, "learning_rate": 2.4948944629787853e-05, "loss": 0.0015, "step": 68370 }, { "grad_norm": 0.04611946642398834, "learning_rate": 2.493463629224731e-05, "loss": 0.0019, "step": 68380 }, { "grad_norm": 0.03818727657198906, "learning_rate": 2.492033069581728e-05, "loss": 0.0014, "step": 68390 }, { "grad_norm": 0.05149427428841591, "learning_rate": 2.4906027842062137e-05, "loss": 0.0015, "step": 68400 }, { "grad_norm": 0.06567765027284622, "learning_rate": 2.48917277325461e-05, "loss": 0.0015, "step": 68410 }, { "grad_norm": 0.05212872847914696, "learning_rate": 2.487743036883295e-05, "loss": 0.002, "step": 68420 }, { "grad_norm": 0.06362663954496384, "learning_rate": 2.486313575248626e-05, "loss": 0.0021, "step": 68430 }, { "grad_norm": 0.0423160083591938, "learning_rate": 2.4848843885069235e-05, "loss": 0.002, "step": 68440 }, { "grad_norm": 0.04520334303379059, "learning_rate": 2.483455476814484e-05, "loss": 0.0017, "step": 68450 }, { "grad_norm": 0.05785130336880684, "learning_rate": 2.4820268403275672e-05, "loss": 0.0016, "step": 68460 }, { "grad_norm": 0.04979558289051056, "learning_rate": 2.480598479202413e-05, "loss": 0.0016, "step": 68470 }, { "grad_norm": 0.10643282532691956, "learning_rate": 2.4791703935952193e-05, "loss": 0.0024, "step": 68480 }, { "grad_norm": 0.06122899055480957, "learning_rate": 2.477742583662163e-05, "loss": 0.0019, "step": 68490 }, { "grad_norm": 0.06027674302458763, "learning_rate": 2.476315049559384e-05, "loss": 0.0015, "step": 68500 }, { "grad_norm": 0.03977940231561661, "learning_rate": 2.4748877914429987e-05, "loss": 0.0017, "step": 68510 }, { "grad_norm": 0.049387458711862564, "learning_rate": 2.4734608094690857e-05, "loss": 0.0017, "step": 68520 }, { "grad_norm": 0.04756324738264084, "learning_rate": 2.4720341037937006e-05, "loss": 0.002, "step": 68530 }, { "grad_norm": 0.06349322199821472, "learning_rate": 2.4706076745728645e-05, "loss": 0.0015, "step": 68540 }, { "grad_norm": 0.07061212509870529, "learning_rate": 2.469181521962573e-05, "loss": 0.0014, "step": 68550 }, { "grad_norm": 0.07402443140745163, "learning_rate": 2.467755646118783e-05, "loss": 0.0015, "step": 68560 }, { "grad_norm": 0.03490496799349785, "learning_rate": 2.46633004719743e-05, "loss": 0.0012, "step": 68570 }, { "grad_norm": 0.1160372942686081, "learning_rate": 2.4649047253544123e-05, "loss": 0.0017, "step": 68580 }, { "grad_norm": 0.09260217845439911, "learning_rate": 2.4634796807456016e-05, "loss": 0.0031, "step": 68590 }, { "grad_norm": 0.0759110078215599, "learning_rate": 2.46205491352684e-05, "loss": 0.0023, "step": 68600 }, { "grad_norm": 0.056071843951940536, "learning_rate": 2.4606304238539375e-05, "loss": 0.0029, "step": 68610 }, { "grad_norm": 0.06619758158922195, "learning_rate": 2.4592062118826753e-05, "loss": 0.002, "step": 68620 }, { "grad_norm": 0.04261254891753197, "learning_rate": 2.4577822777688004e-05, "loss": 0.0013, "step": 68630 }, { "grad_norm": 0.055225882679224014, "learning_rate": 2.4563586216680347e-05, "loss": 0.0016, "step": 68640 }, { "grad_norm": 0.07740595936775208, "learning_rate": 2.4549352437360622e-05, "loss": 0.0034, "step": 68650 }, { "grad_norm": 0.08171948045492172, "learning_rate": 2.4535121441285493e-05, "loss": 0.0021, "step": 68660 }, { "grad_norm": 0.052522964775562286, "learning_rate": 2.4520893230011172e-05, "loss": 0.0023, "step": 68670 }, { "grad_norm": 0.049007147550582886, "learning_rate": 2.450666780509368e-05, "loss": 0.0016, "step": 68680 }, { "grad_norm": 0.04733724892139435, "learning_rate": 2.4492445168088656e-05, "loss": 0.0017, "step": 68690 }, { "grad_norm": 0.041536878794431686, "learning_rate": 2.447822532055149e-05, "loss": 0.0015, "step": 68700 }, { "grad_norm": 0.0394691601395607, "learning_rate": 2.4464008264037212e-05, "loss": 0.0013, "step": 68710 }, { "grad_norm": 0.05916132032871246, "learning_rate": 2.4449794000100605e-05, "loss": 0.0022, "step": 68720 }, { "grad_norm": 0.043348971754312515, "learning_rate": 2.4435582530296108e-05, "loss": 0.0023, "step": 68730 }, { "grad_norm": 0.0587773434817791, "learning_rate": 2.4421373856177887e-05, "loss": 0.0015, "step": 68740 }, { "grad_norm": 0.045615218579769135, "learning_rate": 2.440716797929975e-05, "loss": 0.0018, "step": 68750 }, { "grad_norm": 0.06457816064357758, "learning_rate": 2.439296490121526e-05, "loss": 0.0019, "step": 68760 }, { "grad_norm": 0.06371389329433441, "learning_rate": 2.4378764623477608e-05, "loss": 0.0022, "step": 68770 }, { "grad_norm": 0.06636390835046768, "learning_rate": 2.436456714763974e-05, "loss": 0.0026, "step": 68780 }, { "grad_norm": 0.046072233468294144, "learning_rate": 2.4350372475254262e-05, "loss": 0.0021, "step": 68790 }, { "grad_norm": 0.0585949681699276, "learning_rate": 2.433618060787351e-05, "loss": 0.0023, "step": 68800 }, { "grad_norm": 0.039164990186691284, "learning_rate": 2.432199154704944e-05, "loss": 0.002, "step": 68810 }, { "grad_norm": 0.06292573362588882, "learning_rate": 2.430780529433377e-05, "loss": 0.0018, "step": 68820 }, { "grad_norm": 0.050848543643951416, "learning_rate": 2.42936218512779e-05, "loss": 0.002, "step": 68830 }, { "grad_norm": 0.0654374212026596, "learning_rate": 2.427944121943288e-05, "loss": 0.0016, "step": 68840 }, { "grad_norm": 0.05874871462583542, "learning_rate": 2.4265263400349497e-05, "loss": 0.0025, "step": 68850 }, { "grad_norm": 0.04311610385775566, "learning_rate": 2.4251088395578213e-05, "loss": 0.002, "step": 68860 }, { "grad_norm": 0.04718789458274841, "learning_rate": 2.4236916206669214e-05, "loss": 0.0017, "step": 68870 }, { "grad_norm": 0.05509135127067566, "learning_rate": 2.4222746835172304e-05, "loss": 0.0022, "step": 68880 }, { "grad_norm": 0.03300910443067551, "learning_rate": 2.4208580282637065e-05, "loss": 0.0013, "step": 68890 }, { "grad_norm": 0.044064562767744064, "learning_rate": 2.4194416550612687e-05, "loss": 0.0019, "step": 68900 }, { "grad_norm": 0.05200808495283127, "learning_rate": 2.418025564064812e-05, "loss": 0.0017, "step": 68910 }, { "grad_norm": 0.03370828926563263, "learning_rate": 2.4166097554291978e-05, "loss": 0.0013, "step": 68920 }, { "grad_norm": 0.04766426235437393, "learning_rate": 2.4151942293092583e-05, "loss": 0.0018, "step": 68930 }, { "grad_norm": 0.06113520637154579, "learning_rate": 2.41377898585979e-05, "loss": 0.0017, "step": 68940 }, { "grad_norm": 0.06147703528404236, "learning_rate": 2.4123640252355652e-05, "loss": 0.0014, "step": 68950 }, { "grad_norm": 0.2705737054347992, "learning_rate": 2.410949347591318e-05, "loss": 0.0022, "step": 68960 }, { "grad_norm": 0.03852560371160507, "learning_rate": 2.4095349530817578e-05, "loss": 0.0019, "step": 68970 }, { "grad_norm": 0.0627339705824852, "learning_rate": 2.4081208418615603e-05, "loss": 0.0016, "step": 68980 }, { "grad_norm": 0.07048255205154419, "learning_rate": 2.4067070140853725e-05, "loss": 0.0027, "step": 68990 }, { "grad_norm": 0.06341904401779175, "learning_rate": 2.405293469907805e-05, "loss": 0.0017, "step": 69000 }, { "grad_norm": 0.07490704953670502, "learning_rate": 2.4038802094834446e-05, "loss": 0.0021, "step": 69010 }, { "grad_norm": 0.09427382051944733, "learning_rate": 2.4024672329668395e-05, "loss": 0.0024, "step": 69020 }, { "grad_norm": 0.050812628120183945, "learning_rate": 2.4010545405125125e-05, "loss": 0.002, "step": 69030 }, { "grad_norm": 0.051677554845809937, "learning_rate": 2.3996421322749528e-05, "loss": 0.0025, "step": 69040 }, { "grad_norm": 0.051150958985090256, "learning_rate": 2.3982300084086224e-05, "loss": 0.0026, "step": 69050 }, { "grad_norm": 0.06353823840618134, "learning_rate": 2.396818169067944e-05, "loss": 0.0027, "step": 69060 }, { "grad_norm": 0.047943126410245895, "learning_rate": 2.3954066144073178e-05, "loss": 0.0016, "step": 69070 }, { "grad_norm": 0.13384881615638733, "learning_rate": 2.39399534458111e-05, "loss": 0.0016, "step": 69080 }, { "grad_norm": 0.036607153713703156, "learning_rate": 2.392584359743651e-05, "loss": 0.0013, "step": 69090 }, { "grad_norm": 0.06248117610812187, "learning_rate": 2.3911736600492463e-05, "loss": 0.0021, "step": 69100 }, { "grad_norm": 0.056974709033966064, "learning_rate": 2.3897632456521672e-05, "loss": 0.0014, "step": 69110 }, { "grad_norm": 0.08884472399950027, "learning_rate": 2.388353116706658e-05, "loss": 0.0018, "step": 69120 }, { "grad_norm": 0.04477006569504738, "learning_rate": 2.3869432733669232e-05, "loss": 0.0024, "step": 69130 }, { "grad_norm": 0.06212559714913368, "learning_rate": 2.385533715787145e-05, "loss": 0.0017, "step": 69140 }, { "grad_norm": 0.04940534010529518, "learning_rate": 2.3841244441214673e-05, "loss": 0.0021, "step": 69150 }, { "grad_norm": 0.15750956535339355, "learning_rate": 2.3827154585240073e-05, "loss": 0.0022, "step": 69160 }, { "grad_norm": 0.06163807958364487, "learning_rate": 2.3813067591488498e-05, "loss": 0.0026, "step": 69170 }, { "grad_norm": 0.08645747601985931, "learning_rate": 2.3798983461500494e-05, "loss": 0.0026, "step": 69180 }, { "grad_norm": 0.06807516515254974, "learning_rate": 2.3784902196816256e-05, "loss": 0.002, "step": 69190 }, { "grad_norm": 0.042702920734882355, "learning_rate": 2.3770823798975716e-05, "loss": 0.0016, "step": 69200 }, { "grad_norm": 0.03434502333402634, "learning_rate": 2.3756748269518424e-05, "loss": 0.0018, "step": 69210 }, { "grad_norm": 0.07322487980127335, "learning_rate": 2.374267560998371e-05, "loss": 0.0021, "step": 69220 }, { "grad_norm": 0.05245095491409302, "learning_rate": 2.3728605821910477e-05, "loss": 0.002, "step": 69230 }, { "grad_norm": 0.05249006673693657, "learning_rate": 2.3714538906837453e-05, "loss": 0.0016, "step": 69240 }, { "grad_norm": 0.04736952483654022, "learning_rate": 2.3700474866302913e-05, "loss": 0.0019, "step": 69250 }, { "grad_norm": 0.07614554464817047, "learning_rate": 2.3686413701844917e-05, "loss": 0.0014, "step": 69260 }, { "grad_norm": 0.052310243248939514, "learning_rate": 2.367235541500114e-05, "loss": 0.0017, "step": 69270 }, { "grad_norm": 0.10351830720901489, "learning_rate": 2.3658300007308993e-05, "loss": 0.002, "step": 69280 }, { "grad_norm": 0.09698987007141113, "learning_rate": 2.364424748030555e-05, "loss": 0.0023, "step": 69290 }, { "grad_norm": 0.04547742009162903, "learning_rate": 2.3630197835527595e-05, "loss": 0.0016, "step": 69300 }, { "grad_norm": 0.04924830049276352, "learning_rate": 2.3616151074511546e-05, "loss": 0.0018, "step": 69310 }, { "grad_norm": 0.05701926350593567, "learning_rate": 2.3602107198793545e-05, "loss": 0.0017, "step": 69320 }, { "grad_norm": 0.0288352333009243, "learning_rate": 2.3588066209909427e-05, "loss": 0.0016, "step": 69330 }, { "grad_norm": 0.04148486256599426, "learning_rate": 2.3574028109394662e-05, "loss": 0.003, "step": 69340 }, { "grad_norm": 0.03999824449419975, "learning_rate": 2.3559992898784454e-05, "loss": 0.002, "step": 69350 }, { "grad_norm": 0.05843086168169975, "learning_rate": 2.354596057961367e-05, "loss": 0.002, "step": 69360 }, { "grad_norm": 0.06479492038488388, "learning_rate": 2.353193115341688e-05, "loss": 0.0019, "step": 69370 }, { "grad_norm": 0.05941884592175484, "learning_rate": 2.3517904621728294e-05, "loss": 0.0019, "step": 69380 }, { "grad_norm": 0.03755369782447815, "learning_rate": 2.3503880986081855e-05, "loss": 0.0017, "step": 69390 }, { "grad_norm": 0.05241704732179642, "learning_rate": 2.3489860248011136e-05, "loss": 0.0021, "step": 69400 }, { "grad_norm": 0.044087380170822144, "learning_rate": 2.3475842409049464e-05, "loss": 0.0013, "step": 69410 }, { "grad_norm": 0.05812712013721466, "learning_rate": 2.346182747072975e-05, "loss": 0.0019, "step": 69420 }, { "grad_norm": 0.07030978798866272, "learning_rate": 2.3447815434584718e-05, "loss": 0.0021, "step": 69430 }, { "grad_norm": 0.056092917919158936, "learning_rate": 2.3433806302146656e-05, "loss": 0.0016, "step": 69440 }, { "grad_norm": 0.050796546041965485, "learning_rate": 2.3419800074947612e-05, "loss": 0.002, "step": 69450 }, { "grad_norm": 0.062421053647994995, "learning_rate": 2.3405796754519248e-05, "loss": 0.0017, "step": 69460 }, { "grad_norm": 0.03513652831315994, "learning_rate": 2.3391796342392992e-05, "loss": 0.0016, "step": 69470 }, { "grad_norm": 0.04611043632030487, "learning_rate": 2.337779884009984e-05, "loss": 0.0017, "step": 69480 }, { "grad_norm": 0.06848066300153732, "learning_rate": 2.3363804249170624e-05, "loss": 0.0017, "step": 69490 }, { "grad_norm": 0.0712333545088768, "learning_rate": 2.3349812571135703e-05, "loss": 0.0015, "step": 69500 }, { "grad_norm": 0.06444458663463593, "learning_rate": 2.333582380752523e-05, "loss": 0.0022, "step": 69510 }, { "grad_norm": 0.0454753078520298, "learning_rate": 2.3321837959868963e-05, "loss": 0.0016, "step": 69520 }, { "grad_norm": 0.09570349752902985, "learning_rate": 2.3307855029696383e-05, "loss": 0.0018, "step": 69530 }, { "grad_norm": 0.04417028650641441, "learning_rate": 2.3293875018536648e-05, "loss": 0.0015, "step": 69540 }, { "grad_norm": 0.06501241028308868, "learning_rate": 2.327989792791861e-05, "loss": 0.0018, "step": 69550 }, { "grad_norm": 0.06970930844545364, "learning_rate": 2.326592375937074e-05, "loss": 0.0017, "step": 69560 }, { "grad_norm": 0.055881280452013016, "learning_rate": 2.3251952514421248e-05, "loss": 0.0014, "step": 69570 }, { "grad_norm": 0.05232127755880356, "learning_rate": 2.323798419459804e-05, "loss": 0.0014, "step": 69580 }, { "grad_norm": 0.047048840671777725, "learning_rate": 2.3224018801428625e-05, "loss": 0.0016, "step": 69590 }, { "grad_norm": 0.06045026704668999, "learning_rate": 2.3210056336440273e-05, "loss": 0.0017, "step": 69600 }, { "grad_norm": 0.0948423370718956, "learning_rate": 2.3196096801159844e-05, "loss": 0.0025, "step": 69610 }, { "grad_norm": 0.053903114050626755, "learning_rate": 2.3182140197114012e-05, "loss": 0.0017, "step": 69620 }, { "grad_norm": 0.05170869082212448, "learning_rate": 2.3168186525828985e-05, "loss": 0.0016, "step": 69630 }, { "grad_norm": 0.047776974737644196, "learning_rate": 2.3154235788830757e-05, "loss": 0.0016, "step": 69640 }, { "grad_norm": 0.049769267439842224, "learning_rate": 2.3140287987644927e-05, "loss": 0.0019, "step": 69650 }, { "grad_norm": 0.040820345282554626, "learning_rate": 2.3126343123796833e-05, "loss": 0.0022, "step": 69660 }, { "grad_norm": 0.04450872167944908, "learning_rate": 2.3112401198811418e-05, "loss": 0.0014, "step": 69670 }, { "grad_norm": 0.03692622110247612, "learning_rate": 2.3098462214213418e-05, "loss": 0.0013, "step": 69680 }, { "grad_norm": 0.04119633510708809, "learning_rate": 2.308452617152712e-05, "loss": 0.0015, "step": 69690 }, { "grad_norm": 0.061649058014154434, "learning_rate": 2.3070593072276603e-05, "loss": 0.0019, "step": 69700 }, { "grad_norm": 0.06080181524157524, "learning_rate": 2.3056662917985518e-05, "loss": 0.0013, "step": 69710 }, { "grad_norm": 0.047014202922582626, "learning_rate": 2.3042735710177283e-05, "loss": 0.0021, "step": 69720 }, { "grad_norm": 0.04691410809755325, "learning_rate": 2.302881145037492e-05, "loss": 0.0015, "step": 69730 }, { "grad_norm": 0.03848068416118622, "learning_rate": 2.3014890140101192e-05, "loss": 0.0015, "step": 69740 }, { "grad_norm": 0.04383663833141327, "learning_rate": 2.30009717808785e-05, "loss": 0.0018, "step": 69750 }, { "grad_norm": 0.059894707053899765, "learning_rate": 2.2987056374228967e-05, "loss": 0.0016, "step": 69760 }, { "grad_norm": 0.0428369902074337, "learning_rate": 2.297314392167431e-05, "loss": 0.0015, "step": 69770 }, { "grad_norm": 0.03728166222572327, "learning_rate": 2.295923442473601e-05, "loss": 0.0011, "step": 69780 }, { "grad_norm": 0.06630302965641022, "learning_rate": 2.294532788493519e-05, "loss": 0.0021, "step": 69790 }, { "grad_norm": 0.06677418202161789, "learning_rate": 2.293142430379262e-05, "loss": 0.0019, "step": 69800 }, { "grad_norm": 0.06304289400577545, "learning_rate": 2.291752368282879e-05, "loss": 0.0015, "step": 69810 }, { "grad_norm": 0.2054770141839981, "learning_rate": 2.290362602356385e-05, "loss": 0.0017, "step": 69820 }, { "grad_norm": 0.08367603272199631, "learning_rate": 2.288973132751765e-05, "loss": 0.0023, "step": 69830 }, { "grad_norm": 0.07186152786016464, "learning_rate": 2.287583959620965e-05, "loss": 0.0019, "step": 69840 }, { "grad_norm": 0.08566226065158844, "learning_rate": 2.2861950831159074e-05, "loss": 0.0022, "step": 69850 }, { "grad_norm": 0.09295152127742767, "learning_rate": 2.284806503388472e-05, "loss": 0.0022, "step": 69860 }, { "grad_norm": 0.06989585608243942, "learning_rate": 2.283418220590518e-05, "loss": 0.0018, "step": 69870 }, { "grad_norm": 0.05197058618068695, "learning_rate": 2.282030234873862e-05, "loss": 0.0015, "step": 69880 }, { "grad_norm": 0.041027750819921494, "learning_rate": 2.2806425463902948e-05, "loss": 0.0019, "step": 69890 }, { "grad_norm": 0.04435499384999275, "learning_rate": 2.279255155291568e-05, "loss": 0.0017, "step": 69900 }, { "grad_norm": 0.04506370797753334, "learning_rate": 2.277868061729409e-05, "loss": 0.0015, "step": 69910 }, { "grad_norm": 0.05235988274216652, "learning_rate": 2.276481265855504e-05, "loss": 0.002, "step": 69920 }, { "grad_norm": 0.04901866987347603, "learning_rate": 2.2750947678215128e-05, "loss": 0.0019, "step": 69930 }, { "grad_norm": 0.04636506736278534, "learning_rate": 2.2737085677790616e-05, "loss": 0.0016, "step": 69940 }, { "grad_norm": 0.06541602313518524, "learning_rate": 2.2723226658797443e-05, "loss": 0.0016, "step": 69950 }, { "grad_norm": 0.06628298759460449, "learning_rate": 2.270937062275117e-05, "loss": 0.0016, "step": 69960 }, { "grad_norm": 0.04432531073689461, "learning_rate": 2.2695517571167113e-05, "loss": 0.0013, "step": 69970 }, { "grad_norm": 0.04579286649823189, "learning_rate": 2.268166750556019e-05, "loss": 0.0021, "step": 69980 }, { "grad_norm": 0.06673501431941986, "learning_rate": 2.266782042744503e-05, "loss": 0.0017, "step": 69990 }, { "grad_norm": 0.054483987390995026, "learning_rate": 2.2653976338335935e-05, "loss": 0.0023, "step": 70000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 5000, "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": 64, "trial_name": null, "trial_params": null }