{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.9975224733352661, "learning_rate": 1.8e-07, "loss": 1.551, "step": 10 }, { "grad_norm": 0.9788910150527954, "learning_rate": 3.8e-07, "loss": 1.553, "step": 20 }, { "grad_norm": 0.9372209310531616, "learning_rate": 5.8e-07, "loss": 1.5502, "step": 30 }, { "grad_norm": 1.0687583684921265, "learning_rate": 7.8e-07, "loss": 1.5408, "step": 40 }, { "grad_norm": 0.8944749236106873, "learning_rate": 9.8e-07, "loss": 1.5196, "step": 50 }, { "grad_norm": 1.307900071144104, "learning_rate": 1.18e-06, "loss": 1.4578, "step": 60 }, { "grad_norm": 0.9971070289611816, "learning_rate": 1.3800000000000001e-06, "loss": 1.358, "step": 70 }, { "grad_norm": 0.9231767058372498, "learning_rate": 1.5800000000000003e-06, "loss": 1.3447, "step": 80 }, { "grad_norm": 0.9857848882675171, "learning_rate": 1.7800000000000001e-06, "loss": 1.3011, "step": 90 }, { "grad_norm": 1.4238364696502686, "learning_rate": 1.98e-06, "loss": 1.2511, "step": 100 }, { "grad_norm": 1.0690946578979492, "learning_rate": 2.1800000000000003e-06, "loss": 1.1912, "step": 110 }, { "grad_norm": 0.8021936416625977, "learning_rate": 2.38e-06, "loss": 1.1609, "step": 120 }, { "grad_norm": 0.8536638617515564, "learning_rate": 2.5800000000000003e-06, "loss": 1.1479, "step": 130 }, { "grad_norm": 0.7306159138679504, "learning_rate": 2.78e-06, "loss": 1.1436, "step": 140 }, { "grad_norm": 0.8626981973648071, "learning_rate": 2.9800000000000003e-06, "loss": 1.1382, "step": 150 }, { "grad_norm": 0.6184470653533936, "learning_rate": 3.1800000000000005e-06, "loss": 1.1372, "step": 160 }, { "grad_norm": 0.7347499132156372, "learning_rate": 3.38e-06, "loss": 1.1192, "step": 170 }, { "grad_norm": 0.5373278260231018, "learning_rate": 3.58e-06, "loss": 1.1077, "step": 180 }, { "grad_norm": 0.7212187647819519, "learning_rate": 3.7800000000000002e-06, "loss": 1.1092, "step": 190 }, { "grad_norm": 0.6823059320449829, "learning_rate": 3.98e-06, "loss": 1.104, "step": 200 }, { "grad_norm": 0.8753969669342041, "learning_rate": 4.18e-06, "loss": 1.0945, "step": 210 }, { "grad_norm": 0.9287396669387817, "learning_rate": 4.38e-06, "loss": 1.0929, "step": 220 }, { "grad_norm": 0.7131221294403076, "learning_rate": 4.58e-06, "loss": 1.0754, "step": 230 }, { "grad_norm": 0.8192726373672485, "learning_rate": 4.780000000000001e-06, "loss": 1.0635, "step": 240 }, { "grad_norm": 0.8605013489723206, "learning_rate": 4.98e-06, "loss": 1.0579, "step": 250 }, { "grad_norm": 1.1971544027328491, "learning_rate": 5.18e-06, "loss": 1.051, "step": 260 }, { "grad_norm": 0.7031174302101135, "learning_rate": 5.38e-06, "loss": 1.0269, "step": 270 }, { "grad_norm": 1.1460418701171875, "learning_rate": 5.580000000000001e-06, "loss": 1.0069, "step": 280 }, { "grad_norm": 1.2078816890716553, "learning_rate": 5.78e-06, "loss": 0.9711, "step": 290 }, { "grad_norm": 1.0267271995544434, "learning_rate": 5.98e-06, "loss": 0.9391, "step": 300 }, { "grad_norm": 1.1650363206863403, "learning_rate": 6.18e-06, "loss": 0.9041, "step": 310 }, { "grad_norm": 1.0979350805282593, "learning_rate": 6.38e-06, "loss": 0.8486, "step": 320 }, { "grad_norm": 1.5374418497085571, "learning_rate": 6.58e-06, "loss": 0.8207, "step": 330 }, { "grad_norm": 1.3747873306274414, "learning_rate": 6.78e-06, "loss": 0.7759, "step": 340 }, { "grad_norm": 1.3674312829971313, "learning_rate": 6.98e-06, "loss": 0.7209, "step": 350 }, { "grad_norm": 1.5208721160888672, "learning_rate": 7.180000000000001e-06, "loss": 0.6804, "step": 360 }, { "grad_norm": 1.4818483591079712, "learning_rate": 7.3800000000000005e-06, "loss": 0.6376, "step": 370 }, { "grad_norm": 1.6344751119613647, "learning_rate": 7.580000000000001e-06, "loss": 0.6046, "step": 380 }, { "grad_norm": 1.778784990310669, "learning_rate": 7.78e-06, "loss": 0.5584, "step": 390 }, { "grad_norm": 1.5689412355422974, "learning_rate": 7.98e-06, "loss": 0.5267, "step": 400 }, { "grad_norm": 1.7120646238327026, "learning_rate": 8.18e-06, "loss": 0.5041, "step": 410 }, { "grad_norm": 2.220107316970825, "learning_rate": 8.380000000000001e-06, "loss": 0.4603, "step": 420 }, { "grad_norm": 2.0523900985717773, "learning_rate": 8.580000000000001e-06, "loss": 0.4426, "step": 430 }, { "grad_norm": 1.807788610458374, "learning_rate": 8.78e-06, "loss": 0.3994, "step": 440 }, { "grad_norm": 2.18395733833313, "learning_rate": 8.98e-06, "loss": 0.383, "step": 450 }, { "grad_norm": 2.104902505874634, "learning_rate": 9.180000000000002e-06, "loss": 0.3617, "step": 460 }, { "grad_norm": 1.9469685554504395, "learning_rate": 9.38e-06, "loss": 0.3281, "step": 470 }, { "grad_norm": 2.6318626403808594, "learning_rate": 9.58e-06, "loss": 0.3206, "step": 480 }, { "grad_norm": 1.907653570175171, "learning_rate": 9.78e-06, "loss": 0.2887, "step": 490 }, { "grad_norm": 2.093496799468994, "learning_rate": 9.980000000000001e-06, "loss": 0.2692, "step": 500 }, { "grad_norm": 2.2928149700164795, "learning_rate": 1.018e-05, "loss": 0.2439, "step": 510 }, { "grad_norm": 1.6951100826263428, "learning_rate": 1.038e-05, "loss": 0.2302, "step": 520 }, { "grad_norm": 1.7244486808776855, "learning_rate": 1.058e-05, "loss": 0.2203, "step": 530 }, { "grad_norm": 1.7386807203292847, "learning_rate": 1.0780000000000002e-05, "loss": 0.2069, "step": 540 }, { "grad_norm": 2.1259958744049072, "learning_rate": 1.098e-05, "loss": 0.1802, "step": 550 }, { "grad_norm": 1.8718994855880737, "learning_rate": 1.118e-05, "loss": 0.1913, "step": 560 }, { "grad_norm": 1.9048571586608887, "learning_rate": 1.1380000000000001e-05, "loss": 0.195, "step": 570 }, { "grad_norm": 2.14871883392334, "learning_rate": 1.1580000000000001e-05, "loss": 0.1904, "step": 580 }, { "grad_norm": 1.874372959136963, "learning_rate": 1.178e-05, "loss": 0.1765, "step": 590 }, { "grad_norm": 2.1462464332580566, "learning_rate": 1.198e-05, "loss": 0.1776, "step": 600 }, { "grad_norm": 1.8166368007659912, "learning_rate": 1.2180000000000002e-05, "loss": 0.1688, "step": 610 }, { "grad_norm": 1.97954261302948, "learning_rate": 1.238e-05, "loss": 0.1672, "step": 620 }, { "grad_norm": 1.9214550256729126, "learning_rate": 1.258e-05, "loss": 0.1629, "step": 630 }, { "grad_norm": 2.16848087310791, "learning_rate": 1.278e-05, "loss": 0.1642, "step": 640 }, { "grad_norm": 1.9148555994033813, "learning_rate": 1.2980000000000001e-05, "loss": 0.1596, "step": 650 }, { "grad_norm": 2.2031590938568115, "learning_rate": 1.3180000000000001e-05, "loss": 0.1588, "step": 660 }, { "grad_norm": 1.8737298250198364, "learning_rate": 1.338e-05, "loss": 0.1415, "step": 670 }, { "grad_norm": 2.4444494247436523, "learning_rate": 1.358e-05, "loss": 0.1453, "step": 680 }, { "grad_norm": 1.9248008728027344, "learning_rate": 1.3780000000000002e-05, "loss": 0.1352, "step": 690 }, { "grad_norm": 2.6030020713806152, "learning_rate": 1.3980000000000002e-05, "loss": 0.1439, "step": 700 }, { "grad_norm": 2.1256344318389893, "learning_rate": 1.4180000000000001e-05, "loss": 0.1372, "step": 710 }, { "grad_norm": 2.0386714935302734, "learning_rate": 1.4380000000000001e-05, "loss": 0.1392, "step": 720 }, { "grad_norm": 2.0249977111816406, "learning_rate": 1.4580000000000003e-05, "loss": 0.1359, "step": 730 }, { "grad_norm": 1.6407544612884521, "learning_rate": 1.4779999999999999e-05, "loss": 0.1364, "step": 740 }, { "grad_norm": 1.7784180641174316, "learning_rate": 1.4979999999999999e-05, "loss": 0.1348, "step": 750 }, { "grad_norm": 2.3298146724700928, "learning_rate": 1.518e-05, "loss": 0.1306, "step": 760 }, { "grad_norm": 1.7861031293869019, "learning_rate": 1.538e-05, "loss": 0.1363, "step": 770 }, { "grad_norm": 1.388842225074768, "learning_rate": 1.558e-05, "loss": 0.1266, "step": 780 }, { "grad_norm": 2.239680290222168, "learning_rate": 1.578e-05, "loss": 0.1151, "step": 790 }, { "grad_norm": 2.578012704849243, "learning_rate": 1.598e-05, "loss": 0.1212, "step": 800 }, { "grad_norm": 1.6944953203201294, "learning_rate": 1.618e-05, "loss": 0.113, "step": 810 }, { "grad_norm": 2.2617812156677246, "learning_rate": 1.6380000000000002e-05, "loss": 0.1181, "step": 820 }, { "grad_norm": 1.8611632585525513, "learning_rate": 1.658e-05, "loss": 0.1148, "step": 830 }, { "grad_norm": 1.6048526763916016, "learning_rate": 1.6780000000000002e-05, "loss": 0.1194, "step": 840 }, { "grad_norm": 1.7985590696334839, "learning_rate": 1.698e-05, "loss": 0.1163, "step": 850 }, { "grad_norm": 1.9091578722000122, "learning_rate": 1.718e-05, "loss": 0.1171, "step": 860 }, { "grad_norm": 1.6670420169830322, "learning_rate": 1.7380000000000003e-05, "loss": 0.1117, "step": 870 }, { "grad_norm": 1.9098756313323975, "learning_rate": 1.758e-05, "loss": 0.1151, "step": 880 }, { "grad_norm": 1.4564955234527588, "learning_rate": 1.7780000000000003e-05, "loss": 0.1088, "step": 890 }, { "grad_norm": 1.6268330812454224, "learning_rate": 1.798e-05, "loss": 0.1119, "step": 900 }, { "grad_norm": 1.66782808303833, "learning_rate": 1.818e-05, "loss": 0.1028, "step": 910 }, { "grad_norm": 1.5663467645645142, "learning_rate": 1.838e-05, "loss": 0.0991, "step": 920 }, { "grad_norm": 1.5252668857574463, "learning_rate": 1.858e-05, "loss": 0.1069, "step": 930 }, { "grad_norm": 1.4857734441757202, "learning_rate": 1.878e-05, "loss": 0.1014, "step": 940 }, { "grad_norm": 2.183403253555298, "learning_rate": 1.898e-05, "loss": 0.1077, "step": 950 }, { "grad_norm": 1.6803869009017944, "learning_rate": 1.918e-05, "loss": 0.102, "step": 960 }, { "grad_norm": 1.8586409091949463, "learning_rate": 1.938e-05, "loss": 0.1037, "step": 970 }, { "grad_norm": 1.5640246868133545, "learning_rate": 1.9580000000000002e-05, "loss": 0.0991, "step": 980 }, { "grad_norm": 1.6739190816879272, "learning_rate": 1.978e-05, "loss": 0.1089, "step": 990 }, { "grad_norm": 1.3955771923065186, "learning_rate": 1.9980000000000002e-05, "loss": 0.0942, "step": 1000 }, { "grad_norm": 2.086839199066162, "learning_rate": 2.0180000000000003e-05, "loss": 0.0975, "step": 1010 }, { "grad_norm": 1.4920032024383545, "learning_rate": 2.038e-05, "loss": 0.1059, "step": 1020 }, { "grad_norm": 1.6632213592529297, "learning_rate": 2.0580000000000003e-05, "loss": 0.0928, "step": 1030 }, { "grad_norm": 1.4797157049179077, "learning_rate": 2.078e-05, "loss": 0.0874, "step": 1040 }, { "grad_norm": 1.599825143814087, "learning_rate": 2.098e-05, "loss": 0.0898, "step": 1050 }, { "grad_norm": 1.5796217918395996, "learning_rate": 2.118e-05, "loss": 0.0838, "step": 1060 }, { "grad_norm": 1.797655463218689, "learning_rate": 2.138e-05, "loss": 0.0928, "step": 1070 }, { "grad_norm": 1.6076314449310303, "learning_rate": 2.158e-05, "loss": 0.0972, "step": 1080 }, { "grad_norm": 1.9435464143753052, "learning_rate": 2.178e-05, "loss": 0.0895, "step": 1090 }, { "grad_norm": 1.661603331565857, "learning_rate": 2.198e-05, "loss": 0.0843, "step": 1100 }, { "grad_norm": 1.8126975297927856, "learning_rate": 2.218e-05, "loss": 0.0829, "step": 1110 }, { "grad_norm": 1.2139863967895508, "learning_rate": 2.2380000000000003e-05, "loss": 0.0857, "step": 1120 }, { "grad_norm": 1.2630420923233032, "learning_rate": 2.258e-05, "loss": 0.0917, "step": 1130 }, { "grad_norm": 1.7044087648391724, "learning_rate": 2.2780000000000002e-05, "loss": 0.0858, "step": 1140 }, { "grad_norm": 1.7858256101608276, "learning_rate": 2.298e-05, "loss": 0.0843, "step": 1150 }, { "grad_norm": 1.9050724506378174, "learning_rate": 2.318e-05, "loss": 0.0882, "step": 1160 }, { "grad_norm": 1.6129579544067383, "learning_rate": 2.3380000000000003e-05, "loss": 0.0877, "step": 1170 }, { "grad_norm": 1.6313070058822632, "learning_rate": 2.358e-05, "loss": 0.083, "step": 1180 }, { "grad_norm": 1.5644216537475586, "learning_rate": 2.3780000000000003e-05, "loss": 0.0774, "step": 1190 }, { "grad_norm": 1.488529086112976, "learning_rate": 2.398e-05, "loss": 0.0867, "step": 1200 }, { "grad_norm": 1.6354787349700928, "learning_rate": 2.418e-05, "loss": 0.0799, "step": 1210 }, { "grad_norm": 1.5297985076904297, "learning_rate": 2.438e-05, "loss": 0.0797, "step": 1220 }, { "grad_norm": 1.7572240829467773, "learning_rate": 2.4580000000000002e-05, "loss": 0.0838, "step": 1230 }, { "grad_norm": 1.6243019104003906, "learning_rate": 2.478e-05, "loss": 0.0844, "step": 1240 }, { "grad_norm": 1.5983842611312866, "learning_rate": 2.498e-05, "loss": 0.0795, "step": 1250 }, { "grad_norm": 1.4330590963363647, "learning_rate": 2.5180000000000003e-05, "loss": 0.0832, "step": 1260 }, { "grad_norm": 1.8184115886688232, "learning_rate": 2.5380000000000004e-05, "loss": 0.0812, "step": 1270 }, { "grad_norm": 1.615639090538025, "learning_rate": 2.5580000000000002e-05, "loss": 0.0882, "step": 1280 }, { "grad_norm": 1.2239577770233154, "learning_rate": 2.5779999999999997e-05, "loss": 0.0808, "step": 1290 }, { "grad_norm": 1.6016101837158203, "learning_rate": 2.598e-05, "loss": 0.0774, "step": 1300 }, { "grad_norm": 1.376372218132019, "learning_rate": 2.618e-05, "loss": 0.0746, "step": 1310 }, { "grad_norm": 1.3961836099624634, "learning_rate": 2.6379999999999998e-05, "loss": 0.0736, "step": 1320 }, { "grad_norm": 1.6977592706680298, "learning_rate": 2.658e-05, "loss": 0.0852, "step": 1330 }, { "grad_norm": 1.2673205137252808, "learning_rate": 2.678e-05, "loss": 0.0803, "step": 1340 }, { "grad_norm": 1.7499539852142334, "learning_rate": 2.698e-05, "loss": 0.0787, "step": 1350 }, { "grad_norm": 1.2209681272506714, "learning_rate": 2.718e-05, "loss": 0.0776, "step": 1360 }, { "grad_norm": 1.5990358591079712, "learning_rate": 2.738e-05, "loss": 0.0752, "step": 1370 }, { "grad_norm": 1.2981661558151245, "learning_rate": 2.758e-05, "loss": 0.0713, "step": 1380 }, { "grad_norm": 1.527499794960022, "learning_rate": 2.778e-05, "loss": 0.0751, "step": 1390 }, { "grad_norm": 2.1717453002929688, "learning_rate": 2.798e-05, "loss": 0.0738, "step": 1400 }, { "grad_norm": 1.5965702533721924, "learning_rate": 2.818e-05, "loss": 0.0724, "step": 1410 }, { "grad_norm": 1.573650598526001, "learning_rate": 2.8380000000000003e-05, "loss": 0.0728, "step": 1420 }, { "grad_norm": 1.2911020517349243, "learning_rate": 2.858e-05, "loss": 0.0733, "step": 1430 }, { "grad_norm": 1.4007647037506104, "learning_rate": 2.8780000000000002e-05, "loss": 0.0703, "step": 1440 }, { "grad_norm": 1.2874647378921509, "learning_rate": 2.898e-05, "loss": 0.0715, "step": 1450 }, { "grad_norm": 1.298115611076355, "learning_rate": 2.9180000000000002e-05, "loss": 0.0847, "step": 1460 }, { "grad_norm": 1.294065237045288, "learning_rate": 2.9380000000000003e-05, "loss": 0.0739, "step": 1470 }, { "grad_norm": 1.3410983085632324, "learning_rate": 2.958e-05, "loss": 0.0768, "step": 1480 }, { "grad_norm": 1.6264344453811646, "learning_rate": 2.9780000000000003e-05, "loss": 0.0771, "step": 1490 }, { "grad_norm": 1.4962483644485474, "learning_rate": 2.998e-05, "loss": 0.0768, "step": 1500 }, { "grad_norm": 1.1761730909347534, "learning_rate": 3.0180000000000002e-05, "loss": 0.0806, "step": 1510 }, { "grad_norm": 1.3541561365127563, "learning_rate": 3.0380000000000004e-05, "loss": 0.0713, "step": 1520 }, { "grad_norm": 1.2907086610794067, "learning_rate": 3.058e-05, "loss": 0.0741, "step": 1530 }, { "grad_norm": 1.4249893426895142, "learning_rate": 3.078e-05, "loss": 0.0804, "step": 1540 }, { "grad_norm": 1.677174687385559, "learning_rate": 3.0980000000000005e-05, "loss": 0.0778, "step": 1550 }, { "grad_norm": 1.1475248336791992, "learning_rate": 3.118e-05, "loss": 0.0777, "step": 1560 }, { "grad_norm": 1.6261640787124634, "learning_rate": 3.138e-05, "loss": 0.077, "step": 1570 }, { "grad_norm": 1.6344152688980103, "learning_rate": 3.1580000000000006e-05, "loss": 0.0727, "step": 1580 }, { "grad_norm": 1.5872993469238281, "learning_rate": 3.1780000000000004e-05, "loss": 0.0739, "step": 1590 }, { "grad_norm": 1.0340197086334229, "learning_rate": 3.198e-05, "loss": 0.0782, "step": 1600 }, { "grad_norm": 1.1817866563796997, "learning_rate": 3.218e-05, "loss": 0.0702, "step": 1610 }, { "grad_norm": 1.2296122312545776, "learning_rate": 3.238e-05, "loss": 0.0661, "step": 1620 }, { "grad_norm": 1.1662694215774536, "learning_rate": 3.2579999999999996e-05, "loss": 0.0735, "step": 1630 }, { "grad_norm": 1.4646083116531372, "learning_rate": 3.278e-05, "loss": 0.0719, "step": 1640 }, { "grad_norm": 1.095267653465271, "learning_rate": 3.298e-05, "loss": 0.0701, "step": 1650 }, { "grad_norm": 1.3375276327133179, "learning_rate": 3.318e-05, "loss": 0.0668, "step": 1660 }, { "grad_norm": 1.1763993501663208, "learning_rate": 3.338e-05, "loss": 0.0672, "step": 1670 }, { "grad_norm": 1.0491218566894531, "learning_rate": 3.358e-05, "loss": 0.0674, "step": 1680 }, { "grad_norm": 1.6378374099731445, "learning_rate": 3.378e-05, "loss": 0.0737, "step": 1690 }, { "grad_norm": 1.339936375617981, "learning_rate": 3.398e-05, "loss": 0.0768, "step": 1700 }, { "grad_norm": 1.313605785369873, "learning_rate": 3.418e-05, "loss": 0.0716, "step": 1710 }, { "grad_norm": 1.3290271759033203, "learning_rate": 3.438e-05, "loss": 0.069, "step": 1720 }, { "grad_norm": 1.5068126916885376, "learning_rate": 3.4580000000000004e-05, "loss": 0.0694, "step": 1730 }, { "grad_norm": 1.2394928932189941, "learning_rate": 3.478e-05, "loss": 0.0674, "step": 1740 }, { "grad_norm": 1.4200352430343628, "learning_rate": 3.498e-05, "loss": 0.0722, "step": 1750 }, { "grad_norm": 1.1422483921051025, "learning_rate": 3.518e-05, "loss": 0.0724, "step": 1760 }, { "grad_norm": 1.1815210580825806, "learning_rate": 3.5380000000000003e-05, "loss": 0.0734, "step": 1770 }, { "grad_norm": 0.9596393704414368, "learning_rate": 3.558e-05, "loss": 0.0698, "step": 1780 }, { "grad_norm": 1.0392882823944092, "learning_rate": 3.578e-05, "loss": 0.0722, "step": 1790 }, { "grad_norm": 1.4709360599517822, "learning_rate": 3.5980000000000004e-05, "loss": 0.0677, "step": 1800 }, { "grad_norm": 1.3237780332565308, "learning_rate": 3.618e-05, "loss": 0.0654, "step": 1810 }, { "grad_norm": 1.4277013540267944, "learning_rate": 3.638e-05, "loss": 0.0698, "step": 1820 }, { "grad_norm": 1.054428219795227, "learning_rate": 3.6580000000000006e-05, "loss": 0.0614, "step": 1830 }, { "grad_norm": 1.2031663656234741, "learning_rate": 3.6780000000000004e-05, "loss": 0.0662, "step": 1840 }, { "grad_norm": 1.3193025588989258, "learning_rate": 3.698e-05, "loss": 0.0687, "step": 1850 }, { "grad_norm": 1.0653842687606812, "learning_rate": 3.7180000000000007e-05, "loss": 0.0681, "step": 1860 }, { "grad_norm": 0.8878337144851685, "learning_rate": 3.7380000000000005e-05, "loss": 0.0674, "step": 1870 }, { "grad_norm": 0.9198649525642395, "learning_rate": 3.758e-05, "loss": 0.0661, "step": 1880 }, { "grad_norm": 1.2348819971084595, "learning_rate": 3.778000000000001e-05, "loss": 0.0686, "step": 1890 }, { "grad_norm": 0.938966691493988, "learning_rate": 3.7980000000000006e-05, "loss": 0.0654, "step": 1900 }, { "grad_norm": 1.0537033081054688, "learning_rate": 3.818e-05, "loss": 0.0652, "step": 1910 }, { "grad_norm": 0.966640830039978, "learning_rate": 3.838e-05, "loss": 0.065, "step": 1920 }, { "grad_norm": 1.0263370275497437, "learning_rate": 3.858e-05, "loss": 0.0685, "step": 1930 }, { "grad_norm": 0.9395450353622437, "learning_rate": 3.878e-05, "loss": 0.0689, "step": 1940 }, { "grad_norm": 0.8812814354896545, "learning_rate": 3.898e-05, "loss": 0.0649, "step": 1950 }, { "grad_norm": 1.1552670001983643, "learning_rate": 3.918e-05, "loss": 0.063, "step": 1960 }, { "grad_norm": 0.980377197265625, "learning_rate": 3.938e-05, "loss": 0.0626, "step": 1970 }, { "grad_norm": 1.1364667415618896, "learning_rate": 3.958e-05, "loss": 0.0658, "step": 1980 }, { "grad_norm": 1.187811017036438, "learning_rate": 3.978e-05, "loss": 0.0641, "step": 1990 }, { "grad_norm": 0.8644867539405823, "learning_rate": 3.998e-05, "loss": 0.0638, "step": 2000 }, { "grad_norm": 1.211003065109253, "learning_rate": 4.018e-05, "loss": 0.0602, "step": 2010 }, { "grad_norm": 1.215433120727539, "learning_rate": 4.038e-05, "loss": 0.0667, "step": 2020 }, { "grad_norm": 1.234449028968811, "learning_rate": 4.058e-05, "loss": 0.0639, "step": 2030 }, { "grad_norm": 1.0969312191009521, "learning_rate": 4.078e-05, "loss": 0.0615, "step": 2040 }, { "grad_norm": 1.2620173692703247, "learning_rate": 4.0980000000000004e-05, "loss": 0.0704, "step": 2050 }, { "grad_norm": 1.5413395166397095, "learning_rate": 4.118e-05, "loss": 0.0646, "step": 2060 }, { "grad_norm": 1.205174446105957, "learning_rate": 4.138e-05, "loss": 0.0678, "step": 2070 }, { "grad_norm": 0.8233559131622314, "learning_rate": 4.1580000000000005e-05, "loss": 0.0589, "step": 2080 }, { "grad_norm": 0.9579275250434875, "learning_rate": 4.178e-05, "loss": 0.0679, "step": 2090 }, { "grad_norm": 0.7717534303665161, "learning_rate": 4.198e-05, "loss": 0.0644, "step": 2100 }, { "grad_norm": 1.2402584552764893, "learning_rate": 4.2180000000000006e-05, "loss": 0.0595, "step": 2110 }, { "grad_norm": 0.9612022042274475, "learning_rate": 4.2380000000000004e-05, "loss": 0.0605, "step": 2120 }, { "grad_norm": 0.8668773770332336, "learning_rate": 4.258e-05, "loss": 0.0631, "step": 2130 }, { "grad_norm": 1.2786411046981812, "learning_rate": 4.278e-05, "loss": 0.0586, "step": 2140 }, { "grad_norm": 1.1637139320373535, "learning_rate": 4.2980000000000005e-05, "loss": 0.0614, "step": 2150 }, { "grad_norm": 1.0468307733535767, "learning_rate": 4.318e-05, "loss": 0.0555, "step": 2160 }, { "grad_norm": 0.9679360389709473, "learning_rate": 4.338e-05, "loss": 0.059, "step": 2170 }, { "grad_norm": 1.1773749589920044, "learning_rate": 4.3580000000000006e-05, "loss": 0.0619, "step": 2180 }, { "grad_norm": 1.17970871925354, "learning_rate": 4.3780000000000004e-05, "loss": 0.0594, "step": 2190 }, { "grad_norm": 1.1370575428009033, "learning_rate": 4.398e-05, "loss": 0.0679, "step": 2200 }, { "grad_norm": 0.9872722625732422, "learning_rate": 4.418000000000001e-05, "loss": 0.0637, "step": 2210 }, { "grad_norm": 1.1032280921936035, "learning_rate": 4.438e-05, "loss": 0.0694, "step": 2220 }, { "grad_norm": 0.8907199501991272, "learning_rate": 4.458e-05, "loss": 0.0584, "step": 2230 }, { "grad_norm": 1.1562308073043823, "learning_rate": 4.478e-05, "loss": 0.0597, "step": 2240 }, { "grad_norm": 1.0145926475524902, "learning_rate": 4.498e-05, "loss": 0.0639, "step": 2250 }, { "grad_norm": 1.120290994644165, "learning_rate": 4.518e-05, "loss": 0.0647, "step": 2260 }, { "grad_norm": 1.0963430404663086, "learning_rate": 4.538e-05, "loss": 0.0598, "step": 2270 }, { "grad_norm": 0.9054792523384094, "learning_rate": 4.558e-05, "loss": 0.0547, "step": 2280 }, { "grad_norm": 0.937497615814209, "learning_rate": 4.578e-05, "loss": 0.0613, "step": 2290 }, { "grad_norm": 1.0640075206756592, "learning_rate": 4.5980000000000004e-05, "loss": 0.0573, "step": 2300 }, { "grad_norm": 1.000028133392334, "learning_rate": 4.618e-05, "loss": 0.0639, "step": 2310 }, { "grad_norm": 0.8749353289604187, "learning_rate": 4.638e-05, "loss": 0.0621, "step": 2320 }, { "grad_norm": 0.8822588920593262, "learning_rate": 4.6580000000000005e-05, "loss": 0.0624, "step": 2330 }, { "grad_norm": 1.0748307704925537, "learning_rate": 4.678e-05, "loss": 0.0562, "step": 2340 }, { "grad_norm": 1.0502281188964844, "learning_rate": 4.698e-05, "loss": 0.0636, "step": 2350 }, { "grad_norm": 0.8359048962593079, "learning_rate": 4.718e-05, "loss": 0.0564, "step": 2360 }, { "grad_norm": 0.8407859802246094, "learning_rate": 4.7380000000000004e-05, "loss": 0.0558, "step": 2370 }, { "grad_norm": 0.9285060167312622, "learning_rate": 4.758e-05, "loss": 0.058, "step": 2380 }, { "grad_norm": 0.9562541246414185, "learning_rate": 4.778e-05, "loss": 0.0599, "step": 2390 }, { "grad_norm": 1.0191346406936646, "learning_rate": 4.7980000000000005e-05, "loss": 0.0598, "step": 2400 }, { "grad_norm": 0.9836097359657288, "learning_rate": 4.818e-05, "loss": 0.0546, "step": 2410 }, { "grad_norm": 0.7053503394126892, "learning_rate": 4.838e-05, "loss": 0.064, "step": 2420 }, { "grad_norm": 0.9777510762214661, "learning_rate": 4.8580000000000006e-05, "loss": 0.0557, "step": 2430 }, { "grad_norm": 0.8878880739212036, "learning_rate": 4.8780000000000004e-05, "loss": 0.0575, "step": 2440 }, { "grad_norm": 1.2205071449279785, "learning_rate": 4.898e-05, "loss": 0.0576, "step": 2450 }, { "grad_norm": 1.2274940013885498, "learning_rate": 4.918000000000001e-05, "loss": 0.0571, "step": 2460 }, { "grad_norm": 1.0375521183013916, "learning_rate": 4.9380000000000005e-05, "loss": 0.0571, "step": 2470 }, { "grad_norm": 0.7612653970718384, "learning_rate": 4.958e-05, "loss": 0.0548, "step": 2480 }, { "grad_norm": 0.793198823928833, "learning_rate": 4.978e-05, "loss": 0.0583, "step": 2490 }, { "grad_norm": 0.9814950823783875, "learning_rate": 4.9980000000000006e-05, "loss": 0.0546, "step": 2500 }, { "grad_norm": 1.1109364032745361, "learning_rate": 5.0180000000000004e-05, "loss": 0.0573, "step": 2510 }, { "grad_norm": 0.8188556432723999, "learning_rate": 5.038e-05, "loss": 0.0559, "step": 2520 }, { "grad_norm": 0.8688899278640747, "learning_rate": 5.058000000000001e-05, "loss": 0.0575, "step": 2530 }, { "grad_norm": 0.9791549444198608, "learning_rate": 5.0780000000000005e-05, "loss": 0.0522, "step": 2540 }, { "grad_norm": 0.749591052532196, "learning_rate": 5.098e-05, "loss": 0.0577, "step": 2550 }, { "grad_norm": 0.8270071148872375, "learning_rate": 5.118000000000001e-05, "loss": 0.0591, "step": 2560 }, { "grad_norm": 0.7590640783309937, "learning_rate": 5.1380000000000006e-05, "loss": 0.054, "step": 2570 }, { "grad_norm": 0.9992347955703735, "learning_rate": 5.1580000000000004e-05, "loss": 0.0582, "step": 2580 }, { "grad_norm": 0.8885387778282166, "learning_rate": 5.178000000000001e-05, "loss": 0.0561, "step": 2590 }, { "grad_norm": 0.8538034558296204, "learning_rate": 5.198000000000001e-05, "loss": 0.0619, "step": 2600 }, { "grad_norm": 0.681715726852417, "learning_rate": 5.2180000000000005e-05, "loss": 0.0608, "step": 2610 }, { "grad_norm": 0.9976449012756348, "learning_rate": 5.238000000000001e-05, "loss": 0.0601, "step": 2620 }, { "grad_norm": 0.7447198033332825, "learning_rate": 5.258000000000001e-05, "loss": 0.0576, "step": 2630 }, { "grad_norm": 0.8613282442092896, "learning_rate": 5.2780000000000006e-05, "loss": 0.0594, "step": 2640 }, { "grad_norm": 0.8713308572769165, "learning_rate": 5.2980000000000004e-05, "loss": 0.0577, "step": 2650 }, { "grad_norm": 0.6700264811515808, "learning_rate": 5.318000000000001e-05, "loss": 0.0565, "step": 2660 }, { "grad_norm": 0.7628989219665527, "learning_rate": 5.338000000000001e-05, "loss": 0.0584, "step": 2670 }, { "grad_norm": 0.8815659284591675, "learning_rate": 5.3580000000000005e-05, "loss": 0.0569, "step": 2680 }, { "grad_norm": 0.6319178938865662, "learning_rate": 5.378e-05, "loss": 0.0556, "step": 2690 }, { "grad_norm": 0.8138101696968079, "learning_rate": 5.3979999999999995e-05, "loss": 0.0531, "step": 2700 }, { "grad_norm": 0.993640661239624, "learning_rate": 5.418e-05, "loss": 0.0533, "step": 2710 }, { "grad_norm": 0.9643045663833618, "learning_rate": 5.438e-05, "loss": 0.0525, "step": 2720 }, { "grad_norm": 0.8881089091300964, "learning_rate": 5.4579999999999996e-05, "loss": 0.0556, "step": 2730 }, { "grad_norm": 0.8946599364280701, "learning_rate": 5.478e-05, "loss": 0.055, "step": 2740 }, { "grad_norm": 0.5772756338119507, "learning_rate": 5.498e-05, "loss": 0.0499, "step": 2750 }, { "grad_norm": 0.646065890789032, "learning_rate": 5.518e-05, "loss": 0.0506, "step": 2760 }, { "grad_norm": 0.5758311152458191, "learning_rate": 5.538e-05, "loss": 0.053, "step": 2770 }, { "grad_norm": 0.7722363471984863, "learning_rate": 5.558e-05, "loss": 0.0489, "step": 2780 }, { "grad_norm": 1.0460387468338013, "learning_rate": 5.578e-05, "loss": 0.049, "step": 2790 }, { "grad_norm": 0.9481039047241211, "learning_rate": 5.5979999999999996e-05, "loss": 0.0523, "step": 2800 }, { "grad_norm": 0.7996304631233215, "learning_rate": 5.618e-05, "loss": 0.0532, "step": 2810 }, { "grad_norm": 0.6484392881393433, "learning_rate": 5.638e-05, "loss": 0.05, "step": 2820 }, { "grad_norm": 0.7919263243675232, "learning_rate": 5.658e-05, "loss": 0.0497, "step": 2830 }, { "grad_norm": 0.6285794973373413, "learning_rate": 5.678e-05, "loss": 0.0521, "step": 2840 }, { "grad_norm": 0.9985947012901306, "learning_rate": 5.698e-05, "loss": 0.0538, "step": 2850 }, { "grad_norm": 0.7377331256866455, "learning_rate": 5.718e-05, "loss": 0.0582, "step": 2860 }, { "grad_norm": 0.9662452340126038, "learning_rate": 5.738e-05, "loss": 0.057, "step": 2870 }, { "grad_norm": 0.8617846369743347, "learning_rate": 5.758e-05, "loss": 0.0558, "step": 2880 }, { "grad_norm": 0.7702245712280273, "learning_rate": 5.778e-05, "loss": 0.0547, "step": 2890 }, { "grad_norm": 0.5738171339035034, "learning_rate": 5.7980000000000004e-05, "loss": 0.049, "step": 2900 }, { "grad_norm": 0.9326099753379822, "learning_rate": 5.818e-05, "loss": 0.0536, "step": 2910 }, { "grad_norm": 0.8017061948776245, "learning_rate": 5.838e-05, "loss": 0.0553, "step": 2920 }, { "grad_norm": 0.7413862347602844, "learning_rate": 5.858e-05, "loss": 0.0603, "step": 2930 }, { "grad_norm": 1.0474483966827393, "learning_rate": 5.878e-05, "loss": 0.0605, "step": 2940 }, { "grad_norm": 0.7569717764854431, "learning_rate": 5.898e-05, "loss": 0.0517, "step": 2950 }, { "grad_norm": 0.879721462726593, "learning_rate": 5.918e-05, "loss": 0.056, "step": 2960 }, { "grad_norm": 0.9167556166648865, "learning_rate": 5.9380000000000004e-05, "loss": 0.054, "step": 2970 }, { "grad_norm": 0.8646329641342163, "learning_rate": 5.958e-05, "loss": 0.0513, "step": 2980 }, { "grad_norm": 0.7131762504577637, "learning_rate": 5.978e-05, "loss": 0.0521, "step": 2990 }, { "grad_norm": 0.5936396718025208, "learning_rate": 5.9980000000000005e-05, "loss": 0.0522, "step": 3000 }, { "grad_norm": 0.7116773128509521, "learning_rate": 6.018e-05, "loss": 0.054, "step": 3010 }, { "grad_norm": 0.9421036839485168, "learning_rate": 6.038e-05, "loss": 0.0533, "step": 3020 }, { "grad_norm": 0.8187962174415588, "learning_rate": 6.0580000000000006e-05, "loss": 0.0519, "step": 3030 }, { "grad_norm": 0.7167229652404785, "learning_rate": 6.0780000000000004e-05, "loss": 0.0541, "step": 3040 }, { "grad_norm": 0.6832262873649597, "learning_rate": 6.098e-05, "loss": 0.0485, "step": 3050 }, { "grad_norm": 0.8122378587722778, "learning_rate": 6.118000000000001e-05, "loss": 0.0497, "step": 3060 }, { "grad_norm": 0.7508593201637268, "learning_rate": 6.138e-05, "loss": 0.0508, "step": 3070 }, { "grad_norm": 0.8974077105522156, "learning_rate": 6.158e-05, "loss": 0.0522, "step": 3080 }, { "grad_norm": 0.664301335811615, "learning_rate": 6.178000000000001e-05, "loss": 0.0521, "step": 3090 }, { "grad_norm": 0.6709455251693726, "learning_rate": 6.198e-05, "loss": 0.0513, "step": 3100 }, { "grad_norm": 0.726954460144043, "learning_rate": 6.218e-05, "loss": 0.0511, "step": 3110 }, { "grad_norm": 0.5517457723617554, "learning_rate": 6.238000000000001e-05, "loss": 0.0494, "step": 3120 }, { "grad_norm": 0.966600239276886, "learning_rate": 6.258e-05, "loss": 0.0537, "step": 3130 }, { "grad_norm": 0.6393757462501526, "learning_rate": 6.278e-05, "loss": 0.05, "step": 3140 }, { "grad_norm": 0.5176573395729065, "learning_rate": 6.298000000000001e-05, "loss": 0.0523, "step": 3150 }, { "grad_norm": 0.7349201440811157, "learning_rate": 6.318e-05, "loss": 0.0474, "step": 3160 }, { "grad_norm": 0.6503294706344604, "learning_rate": 6.338e-05, "loss": 0.049, "step": 3170 }, { "grad_norm": 0.7154936194419861, "learning_rate": 6.358000000000001e-05, "loss": 0.0477, "step": 3180 }, { "grad_norm": 0.5571959018707275, "learning_rate": 6.378e-05, "loss": 0.0483, "step": 3190 }, { "grad_norm": 0.7404274344444275, "learning_rate": 6.398000000000001e-05, "loss": 0.0545, "step": 3200 }, { "grad_norm": 0.8104550838470459, "learning_rate": 6.418000000000001e-05, "loss": 0.0528, "step": 3210 }, { "grad_norm": 0.7237273454666138, "learning_rate": 6.438e-05, "loss": 0.0502, "step": 3220 }, { "grad_norm": 0.7358949184417725, "learning_rate": 6.458000000000001e-05, "loss": 0.0494, "step": 3230 }, { "grad_norm": 0.9967756271362305, "learning_rate": 6.478000000000001e-05, "loss": 0.0543, "step": 3240 }, { "grad_norm": 0.751807451248169, "learning_rate": 6.498e-05, "loss": 0.048, "step": 3250 }, { "grad_norm": 0.6971145868301392, "learning_rate": 6.518000000000001e-05, "loss": 0.0528, "step": 3260 }, { "grad_norm": 0.585674524307251, "learning_rate": 6.538000000000001e-05, "loss": 0.0487, "step": 3270 }, { "grad_norm": 0.6536195874214172, "learning_rate": 6.558e-05, "loss": 0.046, "step": 3280 }, { "grad_norm": 0.5735354423522949, "learning_rate": 6.578000000000001e-05, "loss": 0.0576, "step": 3290 }, { "grad_norm": 0.7424114346504211, "learning_rate": 6.598e-05, "loss": 0.0498, "step": 3300 }, { "grad_norm": 0.6457546949386597, "learning_rate": 6.618e-05, "loss": 0.0466, "step": 3310 }, { "grad_norm": 0.9071348309516907, "learning_rate": 6.638e-05, "loss": 0.0498, "step": 3320 }, { "grad_norm": 0.5607154965400696, "learning_rate": 6.658e-05, "loss": 0.047, "step": 3330 }, { "grad_norm": 0.7025681734085083, "learning_rate": 6.678e-05, "loss": 0.0496, "step": 3340 }, { "grad_norm": 0.7961622476577759, "learning_rate": 6.698e-05, "loss": 0.0515, "step": 3350 }, { "grad_norm": 0.5584286451339722, "learning_rate": 6.718e-05, "loss": 0.0494, "step": 3360 }, { "grad_norm": 0.8126237392425537, "learning_rate": 6.738e-05, "loss": 0.053, "step": 3370 }, { "grad_norm": 0.5751201510429382, "learning_rate": 6.758e-05, "loss": 0.0457, "step": 3380 }, { "grad_norm": 0.753041684627533, "learning_rate": 6.778e-05, "loss": 0.0428, "step": 3390 }, { "grad_norm": 0.6321828365325928, "learning_rate": 6.798e-05, "loss": 0.0512, "step": 3400 }, { "grad_norm": 0.6862341165542603, "learning_rate": 6.818e-05, "loss": 0.0452, "step": 3410 }, { "grad_norm": 0.5841202139854431, "learning_rate": 6.838e-05, "loss": 0.0491, "step": 3420 }, { "grad_norm": 0.6049500107765198, "learning_rate": 6.858e-05, "loss": 0.049, "step": 3430 }, { "grad_norm": 0.6386799812316895, "learning_rate": 6.878e-05, "loss": 0.0575, "step": 3440 }, { "grad_norm": 0.7754188179969788, "learning_rate": 6.898e-05, "loss": 0.0517, "step": 3450 }, { "grad_norm": 0.5438117384910583, "learning_rate": 6.918e-05, "loss": 0.0545, "step": 3460 }, { "grad_norm": 0.5988237261772156, "learning_rate": 6.938e-05, "loss": 0.0432, "step": 3470 }, { "grad_norm": 0.9784885048866272, "learning_rate": 6.958e-05, "loss": 0.0524, "step": 3480 }, { "grad_norm": 0.5516456365585327, "learning_rate": 6.978e-05, "loss": 0.0472, "step": 3490 }, { "grad_norm": 0.7843689918518066, "learning_rate": 6.998e-05, "loss": 0.0506, "step": 3500 }, { "grad_norm": 0.5072706341743469, "learning_rate": 7.018e-05, "loss": 0.0447, "step": 3510 }, { "grad_norm": 0.5779539942741394, "learning_rate": 7.038e-05, "loss": 0.0475, "step": 3520 }, { "grad_norm": 0.7591243982315063, "learning_rate": 7.058e-05, "loss": 0.0503, "step": 3530 }, { "grad_norm": 0.5841754674911499, "learning_rate": 7.078e-05, "loss": 0.0466, "step": 3540 }, { "grad_norm": 0.7736421823501587, "learning_rate": 7.098e-05, "loss": 0.0473, "step": 3550 }, { "grad_norm": 0.7235401272773743, "learning_rate": 7.118e-05, "loss": 0.0474, "step": 3560 }, { "grad_norm": 0.5745367407798767, "learning_rate": 7.138e-05, "loss": 0.0457, "step": 3570 }, { "grad_norm": 0.600956916809082, "learning_rate": 7.158e-05, "loss": 0.0507, "step": 3580 }, { "grad_norm": 0.7090939879417419, "learning_rate": 7.178000000000001e-05, "loss": 0.0468, "step": 3590 }, { "grad_norm": 0.5784317851066589, "learning_rate": 7.198e-05, "loss": 0.0495, "step": 3600 }, { "grad_norm": 0.5205085277557373, "learning_rate": 7.218e-05, "loss": 0.0464, "step": 3610 }, { "grad_norm": 0.5818207263946533, "learning_rate": 7.238000000000001e-05, "loss": 0.048, "step": 3620 }, { "grad_norm": 0.7329183220863342, "learning_rate": 7.258e-05, "loss": 0.0456, "step": 3630 }, { "grad_norm": 0.4861656427383423, "learning_rate": 7.278e-05, "loss": 0.0485, "step": 3640 }, { "grad_norm": 0.8041912317276001, "learning_rate": 7.298000000000001e-05, "loss": 0.046, "step": 3650 }, { "grad_norm": 0.539353609085083, "learning_rate": 7.318e-05, "loss": 0.0479, "step": 3660 }, { "grad_norm": 0.8287153840065002, "learning_rate": 7.338e-05, "loss": 0.0489, "step": 3670 }, { "grad_norm": 0.5185717940330505, "learning_rate": 7.358000000000001e-05, "loss": 0.04, "step": 3680 }, { "grad_norm": 0.560111939907074, "learning_rate": 7.378e-05, "loss": 0.0459, "step": 3690 }, { "grad_norm": 0.6353963613510132, "learning_rate": 7.398e-05, "loss": 0.0494, "step": 3700 }, { "grad_norm": 0.5790730118751526, "learning_rate": 7.418000000000001e-05, "loss": 0.0496, "step": 3710 }, { "grad_norm": 0.7015461325645447, "learning_rate": 7.438e-05, "loss": 0.0512, "step": 3720 }, { "grad_norm": 0.5258327722549438, "learning_rate": 7.458000000000001e-05, "loss": 0.0487, "step": 3730 }, { "grad_norm": 0.5163690447807312, "learning_rate": 7.478e-05, "loss": 0.0484, "step": 3740 }, { "grad_norm": 0.5047724843025208, "learning_rate": 7.498e-05, "loss": 0.0529, "step": 3750 }, { "grad_norm": 0.809824526309967, "learning_rate": 7.518000000000001e-05, "loss": 0.0473, "step": 3760 }, { "grad_norm": 0.4843330383300781, "learning_rate": 7.538e-05, "loss": 0.046, "step": 3770 }, { "grad_norm": 0.5816749930381775, "learning_rate": 7.558e-05, "loss": 0.0432, "step": 3780 }, { "grad_norm": 0.5709459781646729, "learning_rate": 7.578000000000001e-05, "loss": 0.0522, "step": 3790 }, { "grad_norm": 0.5581595301628113, "learning_rate": 7.598e-05, "loss": 0.0488, "step": 3800 }, { "grad_norm": 0.6346007585525513, "learning_rate": 7.618e-05, "loss": 0.0437, "step": 3810 }, { "grad_norm": 0.6537196040153503, "learning_rate": 7.638000000000001e-05, "loss": 0.0444, "step": 3820 }, { "grad_norm": 0.6706584095954895, "learning_rate": 7.658e-05, "loss": 0.0474, "step": 3830 }, { "grad_norm": 0.5351465344429016, "learning_rate": 7.678000000000001e-05, "loss": 0.0453, "step": 3840 }, { "grad_norm": 0.5414372086524963, "learning_rate": 7.698000000000001e-05, "loss": 0.0462, "step": 3850 }, { "grad_norm": 0.5260214805603027, "learning_rate": 7.718e-05, "loss": 0.0447, "step": 3860 }, { "grad_norm": 0.45789405703544617, "learning_rate": 7.738000000000001e-05, "loss": 0.0464, "step": 3870 }, { "grad_norm": 0.5879605412483215, "learning_rate": 7.758000000000001e-05, "loss": 0.0448, "step": 3880 }, { "grad_norm": 0.723677933216095, "learning_rate": 7.778e-05, "loss": 0.0438, "step": 3890 }, { "grad_norm": 0.5441688299179077, "learning_rate": 7.798000000000001e-05, "loss": 0.0449, "step": 3900 }, { "grad_norm": 0.5159058570861816, "learning_rate": 7.818000000000001e-05, "loss": 0.048, "step": 3910 }, { "grad_norm": 0.585374116897583, "learning_rate": 7.838e-05, "loss": 0.052, "step": 3920 }, { "grad_norm": 0.5419237017631531, "learning_rate": 7.858000000000001e-05, "loss": 0.0414, "step": 3930 }, { "grad_norm": 0.5026785135269165, "learning_rate": 7.878e-05, "loss": 0.0448, "step": 3940 }, { "grad_norm": 0.5343493223190308, "learning_rate": 7.897999999999999e-05, "loss": 0.0445, "step": 3950 }, { "grad_norm": 0.4866940379142761, "learning_rate": 7.918e-05, "loss": 0.0457, "step": 3960 }, { "grad_norm": 0.5426163673400879, "learning_rate": 7.938e-05, "loss": 0.0452, "step": 3970 }, { "grad_norm": 0.6618219614028931, "learning_rate": 7.958e-05, "loss": 0.0536, "step": 3980 }, { "grad_norm": 0.5372304320335388, "learning_rate": 7.978e-05, "loss": 0.0506, "step": 3990 }, { "grad_norm": 0.6457937359809875, "learning_rate": 7.998e-05, "loss": 0.0439, "step": 4000 }, { "grad_norm": 0.6827357411384583, "learning_rate": 8.018e-05, "loss": 0.0497, "step": 4010 }, { "grad_norm": 0.5433101654052734, "learning_rate": 8.038e-05, "loss": 0.0479, "step": 4020 }, { "grad_norm": 0.6147820353507996, "learning_rate": 8.058e-05, "loss": 0.0477, "step": 4030 }, { "grad_norm": 0.6205549240112305, "learning_rate": 8.078e-05, "loss": 0.0461, "step": 4040 }, { "grad_norm": 0.5457535982131958, "learning_rate": 8.098e-05, "loss": 0.0449, "step": 4050 }, { "grad_norm": 0.4675854444503784, "learning_rate": 8.118e-05, "loss": 0.0456, "step": 4060 }, { "grad_norm": 0.6504355669021606, "learning_rate": 8.138e-05, "loss": 0.047, "step": 4070 }, { "grad_norm": 0.6084593534469604, "learning_rate": 8.158e-05, "loss": 0.0437, "step": 4080 }, { "grad_norm": 0.6045984029769897, "learning_rate": 8.178e-05, "loss": 0.0435, "step": 4090 }, { "grad_norm": 0.48167523741722107, "learning_rate": 8.198e-05, "loss": 0.0434, "step": 4100 }, { "grad_norm": 0.6700831651687622, "learning_rate": 8.218e-05, "loss": 0.0445, "step": 4110 }, { "grad_norm": 0.5450020432472229, "learning_rate": 8.238000000000001e-05, "loss": 0.0465, "step": 4120 }, { "grad_norm": 0.4632725715637207, "learning_rate": 8.258e-05, "loss": 0.0405, "step": 4130 }, { "grad_norm": 0.3497019112110138, "learning_rate": 8.278e-05, "loss": 0.0455, "step": 4140 }, { "grad_norm": 0.4897456765174866, "learning_rate": 8.298000000000001e-05, "loss": 0.0387, "step": 4150 }, { "grad_norm": 0.49974730610847473, "learning_rate": 8.318e-05, "loss": 0.0436, "step": 4160 }, { "grad_norm": 0.5713818073272705, "learning_rate": 8.338e-05, "loss": 0.0418, "step": 4170 }, { "grad_norm": 0.44841891527175903, "learning_rate": 8.358e-05, "loss": 0.0412, "step": 4180 }, { "grad_norm": 0.48230791091918945, "learning_rate": 8.378e-05, "loss": 0.0373, "step": 4190 }, { "grad_norm": 0.4889425039291382, "learning_rate": 8.398e-05, "loss": 0.0398, "step": 4200 }, { "grad_norm": 0.7523491978645325, "learning_rate": 8.418e-05, "loss": 0.0425, "step": 4210 }, { "grad_norm": 0.525765061378479, "learning_rate": 8.438e-05, "loss": 0.0409, "step": 4220 }, { "grad_norm": 0.5017731189727783, "learning_rate": 8.458e-05, "loss": 0.0385, "step": 4230 }, { "grad_norm": 0.4968859553337097, "learning_rate": 8.478e-05, "loss": 0.043, "step": 4240 }, { "grad_norm": 0.6163761019706726, "learning_rate": 8.498e-05, "loss": 0.0399, "step": 4250 }, { "grad_norm": 0.61353999376297, "learning_rate": 8.518000000000001e-05, "loss": 0.0416, "step": 4260 }, { "grad_norm": 0.45355451107025146, "learning_rate": 8.538e-05, "loss": 0.0373, "step": 4270 }, { "grad_norm": 0.5551420450210571, "learning_rate": 8.558e-05, "loss": 0.0399, "step": 4280 }, { "grad_norm": 0.42952898144721985, "learning_rate": 8.578000000000001e-05, "loss": 0.0408, "step": 4290 }, { "grad_norm": 0.3732469975948334, "learning_rate": 8.598e-05, "loss": 0.0404, "step": 4300 }, { "grad_norm": 0.41509848833084106, "learning_rate": 8.618e-05, "loss": 0.0419, "step": 4310 }, { "grad_norm": 0.5405118465423584, "learning_rate": 8.638000000000001e-05, "loss": 0.04, "step": 4320 }, { "grad_norm": 0.5031703114509583, "learning_rate": 8.658e-05, "loss": 0.0411, "step": 4330 }, { "grad_norm": 0.5535834431648254, "learning_rate": 8.678e-05, "loss": 0.0428, "step": 4340 }, { "grad_norm": 0.4803304970264435, "learning_rate": 8.698000000000001e-05, "loss": 0.0404, "step": 4350 }, { "grad_norm": 0.5785989165306091, "learning_rate": 8.718e-05, "loss": 0.0412, "step": 4360 }, { "grad_norm": 0.39817187190055847, "learning_rate": 8.738000000000001e-05, "loss": 0.0419, "step": 4370 }, { "grad_norm": 0.4228054881095886, "learning_rate": 8.758000000000001e-05, "loss": 0.0404, "step": 4380 }, { "grad_norm": 0.6565912365913391, "learning_rate": 8.778e-05, "loss": 0.0421, "step": 4390 }, { "grad_norm": 0.49560627341270447, "learning_rate": 8.798000000000001e-05, "loss": 0.0451, "step": 4400 }, { "grad_norm": 0.5272500514984131, "learning_rate": 8.818000000000001e-05, "loss": 0.0462, "step": 4410 }, { "grad_norm": 0.5669817328453064, "learning_rate": 8.838e-05, "loss": 0.0434, "step": 4420 }, { "grad_norm": 0.4383881688117981, "learning_rate": 8.858000000000001e-05, "loss": 0.046, "step": 4430 }, { "grad_norm": 0.529341995716095, "learning_rate": 8.878000000000001e-05, "loss": 0.0452, "step": 4440 }, { "grad_norm": 0.681519091129303, "learning_rate": 8.898e-05, "loss": 0.0462, "step": 4450 }, { "grad_norm": 0.5899010300636292, "learning_rate": 8.918000000000001e-05, "loss": 0.0415, "step": 4460 }, { "grad_norm": 0.552308201789856, "learning_rate": 8.938e-05, "loss": 0.0421, "step": 4470 }, { "grad_norm": 0.5406711101531982, "learning_rate": 8.958e-05, "loss": 0.0444, "step": 4480 }, { "grad_norm": 0.5028188824653625, "learning_rate": 8.978000000000001e-05, "loss": 0.0414, "step": 4490 }, { "grad_norm": 0.4711349904537201, "learning_rate": 8.998e-05, "loss": 0.0435, "step": 4500 }, { "grad_norm": 0.4359803795814514, "learning_rate": 9.018000000000001e-05, "loss": 0.039, "step": 4510 }, { "grad_norm": 0.43766140937805176, "learning_rate": 9.038000000000001e-05, "loss": 0.0392, "step": 4520 }, { "grad_norm": 0.4266323447227478, "learning_rate": 9.058e-05, "loss": 0.0424, "step": 4530 }, { "grad_norm": 0.5485581159591675, "learning_rate": 9.078000000000001e-05, "loss": 0.0406, "step": 4540 }, { "grad_norm": 0.5791836380958557, "learning_rate": 9.098000000000001e-05, "loss": 0.0416, "step": 4550 }, { "grad_norm": 0.48284798860549927, "learning_rate": 9.118e-05, "loss": 0.0408, "step": 4560 }, { "grad_norm": 0.5601641535758972, "learning_rate": 9.138e-05, "loss": 0.0382, "step": 4570 }, { "grad_norm": 0.40926411747932434, "learning_rate": 9.158e-05, "loss": 0.0427, "step": 4580 }, { "grad_norm": 0.5669175386428833, "learning_rate": 9.178e-05, "loss": 0.0437, "step": 4590 }, { "grad_norm": 0.3983975350856781, "learning_rate": 9.198e-05, "loss": 0.0424, "step": 4600 }, { "grad_norm": 0.49079540371894836, "learning_rate": 9.218e-05, "loss": 0.0448, "step": 4610 }, { "grad_norm": 0.5684144496917725, "learning_rate": 9.238e-05, "loss": 0.0446, "step": 4620 }, { "grad_norm": 0.5550910830497742, "learning_rate": 9.258e-05, "loss": 0.0445, "step": 4630 }, { "grad_norm": 0.5026065111160278, "learning_rate": 9.278e-05, "loss": 0.0407, "step": 4640 }, { "grad_norm": 0.5312486290931702, "learning_rate": 9.298e-05, "loss": 0.0417, "step": 4650 }, { "grad_norm": 0.4753267765045166, "learning_rate": 9.318e-05, "loss": 0.0439, "step": 4660 }, { "grad_norm": 0.5649738311767578, "learning_rate": 9.338e-05, "loss": 0.0452, "step": 4670 }, { "grad_norm": 0.4067171812057495, "learning_rate": 9.358e-05, "loss": 0.0424, "step": 4680 }, { "grad_norm": 0.5612300038337708, "learning_rate": 9.378e-05, "loss": 0.0454, "step": 4690 }, { "grad_norm": 0.42577436566352844, "learning_rate": 9.398e-05, "loss": 0.0402, "step": 4700 }, { "grad_norm": 0.3561828136444092, "learning_rate": 9.418e-05, "loss": 0.0395, "step": 4710 }, { "grad_norm": 0.519399106502533, "learning_rate": 9.438e-05, "loss": 0.0414, "step": 4720 }, { "grad_norm": 0.6033222675323486, "learning_rate": 9.458e-05, "loss": 0.0407, "step": 4730 }, { "grad_norm": 0.49622711539268494, "learning_rate": 9.478e-05, "loss": 0.0418, "step": 4740 }, { "grad_norm": 0.4575178325176239, "learning_rate": 9.498e-05, "loss": 0.0477, "step": 4750 }, { "grad_norm": 0.49623963236808777, "learning_rate": 9.518000000000001e-05, "loss": 0.0439, "step": 4760 }, { "grad_norm": 0.37838563323020935, "learning_rate": 9.538e-05, "loss": 0.0438, "step": 4770 }, { "grad_norm": 0.49385279417037964, "learning_rate": 9.558e-05, "loss": 0.0461, "step": 4780 }, { "grad_norm": 0.4169091582298279, "learning_rate": 9.578000000000001e-05, "loss": 0.0451, "step": 4790 }, { "grad_norm": 0.37672117352485657, "learning_rate": 9.598e-05, "loss": 0.0441, "step": 4800 }, { "grad_norm": 0.3934929370880127, "learning_rate": 9.618e-05, "loss": 0.0493, "step": 4810 }, { "grad_norm": 0.4146836996078491, "learning_rate": 9.638000000000001e-05, "loss": 0.0428, "step": 4820 }, { "grad_norm": 0.45145994424819946, "learning_rate": 9.658e-05, "loss": 0.0443, "step": 4830 }, { "grad_norm": 0.41763001680374146, "learning_rate": 9.678e-05, "loss": 0.0413, "step": 4840 }, { "grad_norm": 0.46199461817741394, "learning_rate": 9.698000000000001e-05, "loss": 0.0387, "step": 4850 }, { "grad_norm": 0.39983585476875305, "learning_rate": 9.718e-05, "loss": 0.043, "step": 4860 }, { "grad_norm": 0.37119796872138977, "learning_rate": 9.738e-05, "loss": 0.0422, "step": 4870 }, { "grad_norm": 0.4067266583442688, "learning_rate": 9.758000000000001e-05, "loss": 0.0445, "step": 4880 }, { "grad_norm": 0.5364513993263245, "learning_rate": 9.778e-05, "loss": 0.045, "step": 4890 }, { "grad_norm": 0.417329877614975, "learning_rate": 9.798000000000001e-05, "loss": 0.0425, "step": 4900 }, { "grad_norm": 0.46560192108154297, "learning_rate": 9.818000000000001e-05, "loss": 0.04, "step": 4910 }, { "grad_norm": 0.5174978971481323, "learning_rate": 9.838e-05, "loss": 0.044, "step": 4920 }, { "grad_norm": 0.42216989398002625, "learning_rate": 9.858000000000001e-05, "loss": 0.0437, "step": 4930 }, { "grad_norm": 0.44136396050453186, "learning_rate": 9.878e-05, "loss": 0.0388, "step": 4940 }, { "grad_norm": 0.4786333441734314, "learning_rate": 9.898e-05, "loss": 0.0415, "step": 4950 }, { "grad_norm": 0.42887330055236816, "learning_rate": 9.918000000000001e-05, "loss": 0.0413, "step": 4960 }, { "grad_norm": 0.3763371407985687, "learning_rate": 9.938e-05, "loss": 0.0392, "step": 4970 }, { "grad_norm": 0.5108075737953186, "learning_rate": 9.958e-05, "loss": 0.0415, "step": 4980 }, { "grad_norm": 0.8270316123962402, "learning_rate": 9.978000000000001e-05, "loss": 0.0413, "step": 4990 }, { "grad_norm": 0.397341787815094, "learning_rate": 9.998e-05, "loss": 0.0497, "step": 5000 }, { "grad_norm": 0.39929959177970886, "learning_rate": 9.999999778549045e-05, "loss": 0.0439, "step": 5010 }, { "grad_norm": 0.5613687038421631, "learning_rate": 9.999999013039593e-05, "loss": 0.046, "step": 5020 }, { "grad_norm": 0.50724858045578, "learning_rate": 9.999997700737766e-05, "loss": 0.0412, "step": 5030 }, { "grad_norm": 0.47708743810653687, "learning_rate": 9.999995841643709e-05, "loss": 0.0421, "step": 5040 }, { "grad_norm": 0.5865389108657837, "learning_rate": 9.999993435757623e-05, "loss": 0.0401, "step": 5050 }, { "grad_norm": 0.41585490107536316, "learning_rate": 9.999990483079773e-05, "loss": 0.0406, "step": 5060 }, { "grad_norm": 0.5261132121086121, "learning_rate": 9.999986983610481e-05, "loss": 0.0413, "step": 5070 }, { "grad_norm": 0.49273985624313354, "learning_rate": 9.99998293735013e-05, "loss": 0.0444, "step": 5080 }, { "grad_norm": 0.4202691912651062, "learning_rate": 9.999978344299161e-05, "loss": 0.0462, "step": 5090 }, { "grad_norm": 0.5040944218635559, "learning_rate": 9.99997320445808e-05, "loss": 0.0404, "step": 5100 }, { "grad_norm": 0.4478529989719391, "learning_rate": 9.999967517827444e-05, "loss": 0.0433, "step": 5110 }, { "grad_norm": 0.44097790122032166, "learning_rate": 9.999961284407879e-05, "loss": 0.0408, "step": 5120 }, { "grad_norm": 0.4494256377220154, "learning_rate": 9.999954504200067e-05, "loss": 0.0442, "step": 5130 }, { "grad_norm": 0.4798661470413208, "learning_rate": 9.999947177204744e-05, "loss": 0.0441, "step": 5140 }, { "grad_norm": 0.399030476808548, "learning_rate": 9.999939303422718e-05, "loss": 0.0449, "step": 5150 }, { "grad_norm": 0.4240924119949341, "learning_rate": 9.999930882854847e-05, "loss": 0.041, "step": 5160 }, { "grad_norm": 0.3710813522338867, "learning_rate": 9.999921915502051e-05, "loss": 0.0376, "step": 5170 }, { "grad_norm": 0.444669634103775, "learning_rate": 9.99991240136531e-05, "loss": 0.0381, "step": 5180 }, { "grad_norm": 0.38486409187316895, "learning_rate": 9.999902340445668e-05, "loss": 0.0405, "step": 5190 }, { "grad_norm": 0.4766444265842438, "learning_rate": 9.999891732744224e-05, "loss": 0.0395, "step": 5200 }, { "grad_norm": 0.45755940675735474, "learning_rate": 9.999880578262135e-05, "loss": 0.041, "step": 5210 }, { "grad_norm": 0.43284228444099426, "learning_rate": 9.999868877000624e-05, "loss": 0.0389, "step": 5220 }, { "grad_norm": 0.4678821563720703, "learning_rate": 9.99985662896097e-05, "loss": 0.0411, "step": 5230 }, { "grad_norm": 0.49120932817459106, "learning_rate": 9.999843834144513e-05, "loss": 0.0402, "step": 5240 }, { "grad_norm": 0.5443093180656433, "learning_rate": 9.99983049255265e-05, "loss": 0.0446, "step": 5250 }, { "grad_norm": 0.5326570272445679, "learning_rate": 9.999816604186843e-05, "loss": 0.0555, "step": 5260 }, { "grad_norm": 0.4300270080566406, "learning_rate": 9.999802169048609e-05, "loss": 0.0503, "step": 5270 }, { "grad_norm": 0.41250282526016235, "learning_rate": 9.999787187139527e-05, "loss": 0.0407, "step": 5280 }, { "grad_norm": 0.4004277288913727, "learning_rate": 9.999771658461234e-05, "loss": 0.0383, "step": 5290 }, { "grad_norm": 0.5143145322799683, "learning_rate": 9.999755583015431e-05, "loss": 0.0427, "step": 5300 }, { "grad_norm": 0.3642270267009735, "learning_rate": 9.999738960803874e-05, "loss": 0.0388, "step": 5310 }, { "grad_norm": 0.3067939579486847, "learning_rate": 9.99972179182838e-05, "loss": 0.0368, "step": 5320 }, { "grad_norm": 0.4158315062522888, "learning_rate": 9.99970407609083e-05, "loss": 0.0384, "step": 5330 }, { "grad_norm": 0.3692384660243988, "learning_rate": 9.999685813593159e-05, "loss": 0.0388, "step": 5340 }, { "grad_norm": 0.30440735816955566, "learning_rate": 9.999667004337362e-05, "loss": 0.0351, "step": 5350 }, { "grad_norm": 0.3989730775356293, "learning_rate": 9.9996476483255e-05, "loss": 0.0376, "step": 5360 }, { "grad_norm": 0.4260944426059723, "learning_rate": 9.999627745559688e-05, "loss": 0.0367, "step": 5370 }, { "grad_norm": 0.427773118019104, "learning_rate": 9.999607296042101e-05, "loss": 0.0347, "step": 5380 }, { "grad_norm": 0.4268186688423157, "learning_rate": 9.99958629977498e-05, "loss": 0.0349, "step": 5390 }, { "grad_norm": 0.5151520371437073, "learning_rate": 9.999564756760615e-05, "loss": 0.0354, "step": 5400 }, { "grad_norm": 0.3963541090488434, "learning_rate": 9.999542667001366e-05, "loss": 0.0397, "step": 5410 }, { "grad_norm": 0.4334373474121094, "learning_rate": 9.999520030499647e-05, "loss": 0.0392, "step": 5420 }, { "grad_norm": 0.5068620443344116, "learning_rate": 9.999496847257936e-05, "loss": 0.0386, "step": 5430 }, { "grad_norm": 0.38695403933525085, "learning_rate": 9.999473117278764e-05, "loss": 0.0366, "step": 5440 }, { "grad_norm": 0.3627166152000427, "learning_rate": 9.999448840564731e-05, "loss": 0.0343, "step": 5450 }, { "grad_norm": 0.4354136288166046, "learning_rate": 9.999424017118488e-05, "loss": 0.0358, "step": 5460 }, { "grad_norm": 0.46292948722839355, "learning_rate": 9.999398646942751e-05, "loss": 0.0381, "step": 5470 }, { "grad_norm": 0.425296425819397, "learning_rate": 9.999372730040296e-05, "loss": 0.0388, "step": 5480 }, { "grad_norm": 0.36421820521354675, "learning_rate": 9.999346266413953e-05, "loss": 0.0356, "step": 5490 }, { "grad_norm": 0.43300268054008484, "learning_rate": 9.99931925606662e-05, "loss": 0.0406, "step": 5500 }, { "grad_norm": 0.5187178254127502, "learning_rate": 9.99929169900125e-05, "loss": 0.0363, "step": 5510 }, { "grad_norm": 0.297522634267807, "learning_rate": 9.999263595220855e-05, "loss": 0.0375, "step": 5520 }, { "grad_norm": 0.5504851937294006, "learning_rate": 9.99923494472851e-05, "loss": 0.0352, "step": 5530 }, { "grad_norm": 0.453182578086853, "learning_rate": 9.999205747527348e-05, "loss": 0.0377, "step": 5540 }, { "grad_norm": 0.5340377688407898, "learning_rate": 9.999176003620561e-05, "loss": 0.0388, "step": 5550 }, { "grad_norm": 0.386752188205719, "learning_rate": 9.999145713011405e-05, "loss": 0.0404, "step": 5560 }, { "grad_norm": 0.3518471419811249, "learning_rate": 9.999114875703186e-05, "loss": 0.037, "step": 5570 }, { "grad_norm": 0.5140269994735718, "learning_rate": 9.999083491699281e-05, "loss": 0.0409, "step": 5580 }, { "grad_norm": 0.42127418518066406, "learning_rate": 9.999051561003123e-05, "loss": 0.0361, "step": 5590 }, { "grad_norm": 0.4243949353694916, "learning_rate": 9.999019083618202e-05, "loss": 0.0421, "step": 5600 }, { "grad_norm": 0.5000735521316528, "learning_rate": 9.99898605954807e-05, "loss": 0.0396, "step": 5610 }, { "grad_norm": 0.4628535509109497, "learning_rate": 9.998952488796338e-05, "loss": 0.0411, "step": 5620 }, { "grad_norm": 0.5252373814582825, "learning_rate": 9.998918371366676e-05, "loss": 0.0438, "step": 5630 }, { "grad_norm": 0.5636419653892517, "learning_rate": 9.99888370726282e-05, "loss": 0.0408, "step": 5640 }, { "grad_norm": 0.4502127468585968, "learning_rate": 9.998848496488556e-05, "loss": 0.0406, "step": 5650 }, { "grad_norm": 0.2830300033092499, "learning_rate": 9.998812739047736e-05, "loss": 0.0365, "step": 5660 }, { "grad_norm": 0.4179527759552002, "learning_rate": 9.99877643494427e-05, "loss": 0.0397, "step": 5670 }, { "grad_norm": 0.3008710443973541, "learning_rate": 9.998739584182128e-05, "loss": 0.0331, "step": 5680 }, { "grad_norm": 0.5122311115264893, "learning_rate": 9.998702186765342e-05, "loss": 0.0375, "step": 5690 }, { "grad_norm": 0.3921786844730377, "learning_rate": 9.998664242698e-05, "loss": 0.0332, "step": 5700 }, { "grad_norm": 0.4067893922328949, "learning_rate": 9.998625751984251e-05, "loss": 0.0391, "step": 5710 }, { "grad_norm": 0.4593198895454407, "learning_rate": 9.998586714628307e-05, "loss": 0.0365, "step": 5720 }, { "grad_norm": 0.39926210045814514, "learning_rate": 9.998547130634432e-05, "loss": 0.0409, "step": 5730 }, { "grad_norm": 0.3877190947532654, "learning_rate": 9.99850700000696e-05, "loss": 0.0396, "step": 5740 }, { "grad_norm": 0.4329858720302582, "learning_rate": 9.998466322750278e-05, "loss": 0.0365, "step": 5750 }, { "grad_norm": 0.4701058864593506, "learning_rate": 9.998425098868834e-05, "loss": 0.0364, "step": 5760 }, { "grad_norm": 0.40196692943573, "learning_rate": 9.998383328367136e-05, "loss": 0.0376, "step": 5770 }, { "grad_norm": 0.3198049068450928, "learning_rate": 9.99834101124975e-05, "loss": 0.0363, "step": 5780 }, { "grad_norm": 0.365826278924942, "learning_rate": 9.998298147521309e-05, "loss": 0.0368, "step": 5790 }, { "grad_norm": 0.34877586364746094, "learning_rate": 9.998254737186496e-05, "loss": 0.0379, "step": 5800 }, { "grad_norm": 0.3323003947734833, "learning_rate": 9.99821078025006e-05, "loss": 0.0365, "step": 5810 }, { "grad_norm": 0.396135538816452, "learning_rate": 9.998166276716807e-05, "loss": 0.0349, "step": 5820 }, { "grad_norm": 0.4706515967845917, "learning_rate": 9.998121226591606e-05, "loss": 0.0337, "step": 5830 }, { "grad_norm": 0.6308317184448242, "learning_rate": 9.998075629879382e-05, "loss": 0.0346, "step": 5840 }, { "grad_norm": 0.5030755400657654, "learning_rate": 9.99802948658512e-05, "loss": 0.0376, "step": 5850 }, { "grad_norm": 0.45359840989112854, "learning_rate": 9.99798279671387e-05, "loss": 0.0382, "step": 5860 }, { "grad_norm": 0.3123444616794586, "learning_rate": 9.997935560270734e-05, "loss": 0.0353, "step": 5870 }, { "grad_norm": 0.4493919909000397, "learning_rate": 9.997887777260879e-05, "loss": 0.0396, "step": 5880 }, { "grad_norm": 0.4076724946498871, "learning_rate": 9.997839447689532e-05, "loss": 0.0325, "step": 5890 }, { "grad_norm": 0.3806590139865875, "learning_rate": 9.997790571561978e-05, "loss": 0.0358, "step": 5900 }, { "grad_norm": 0.5485077500343323, "learning_rate": 9.99774114888356e-05, "loss": 0.0372, "step": 5910 }, { "grad_norm": 0.4434952437877655, "learning_rate": 9.997691179659684e-05, "loss": 0.0372, "step": 5920 }, { "grad_norm": 0.44390982389450073, "learning_rate": 9.997640663895815e-05, "loss": 0.0388, "step": 5930 }, { "grad_norm": 0.386062890291214, "learning_rate": 9.997589601597477e-05, "loss": 0.0369, "step": 5940 }, { "grad_norm": 0.40225985646247864, "learning_rate": 9.997537992770252e-05, "loss": 0.0383, "step": 5950 }, { "grad_norm": 0.27407723665237427, "learning_rate": 9.997485837419788e-05, "loss": 0.0344, "step": 5960 }, { "grad_norm": 0.4362132251262665, "learning_rate": 9.997433135551786e-05, "loss": 0.0339, "step": 5970 }, { "grad_norm": 0.3479194641113281, "learning_rate": 9.997379887172009e-05, "loss": 0.0351, "step": 5980 }, { "grad_norm": 0.4513731002807617, "learning_rate": 9.997326092286281e-05, "loss": 0.0379, "step": 5990 }, { "grad_norm": 0.3031649887561798, "learning_rate": 9.997271750900486e-05, "loss": 0.0367, "step": 6000 }, { "grad_norm": 0.29613983631134033, "learning_rate": 9.997216863020565e-05, "loss": 0.0366, "step": 6010 }, { "grad_norm": 0.3330073654651642, "learning_rate": 9.99716142865252e-05, "loss": 0.0361, "step": 6020 }, { "grad_norm": 0.5256592631340027, "learning_rate": 9.997105447802415e-05, "loss": 0.038, "step": 6030 }, { "grad_norm": 0.4019038677215576, "learning_rate": 9.997048920476373e-05, "loss": 0.0343, "step": 6040 }, { "grad_norm": 0.427051842212677, "learning_rate": 9.996991846680572e-05, "loss": 0.0363, "step": 6050 }, { "grad_norm": 0.3440536558628082, "learning_rate": 9.996934226421257e-05, "loss": 0.0364, "step": 6060 }, { "grad_norm": 0.30843237042427063, "learning_rate": 9.996876059704726e-05, "loss": 0.0333, "step": 6070 }, { "grad_norm": 0.3050723671913147, "learning_rate": 9.996817346537343e-05, "loss": 0.0391, "step": 6080 }, { "grad_norm": 0.337971568107605, "learning_rate": 9.996758086925526e-05, "loss": 0.0348, "step": 6090 }, { "grad_norm": 0.41162270307540894, "learning_rate": 9.996698280875759e-05, "loss": 0.0358, "step": 6100 }, { "grad_norm": 0.45799800753593445, "learning_rate": 9.99663792839458e-05, "loss": 0.0384, "step": 6110 }, { "grad_norm": 0.34209704399108887, "learning_rate": 9.99657702948859e-05, "loss": 0.0378, "step": 6120 }, { "grad_norm": 0.2843440771102905, "learning_rate": 9.996515584164448e-05, "loss": 0.0333, "step": 6130 }, { "grad_norm": 0.37961632013320923, "learning_rate": 9.996453592428873e-05, "loss": 0.0364, "step": 6140 }, { "grad_norm": 0.3599735200405121, "learning_rate": 9.996391054288646e-05, "loss": 0.0311, "step": 6150 }, { "grad_norm": 0.36396563053131104, "learning_rate": 9.996327969750605e-05, "loss": 0.0349, "step": 6160 }, { "grad_norm": 0.4434325098991394, "learning_rate": 9.996264338821649e-05, "loss": 0.0334, "step": 6170 }, { "grad_norm": 0.37124744057655334, "learning_rate": 9.996200161508735e-05, "loss": 0.0363, "step": 6180 }, { "grad_norm": 0.2865388095378876, "learning_rate": 9.996135437818885e-05, "loss": 0.035, "step": 6190 }, { "grad_norm": 0.3248513340950012, "learning_rate": 9.996070167759175e-05, "loss": 0.0348, "step": 6200 }, { "grad_norm": 0.3913320302963257, "learning_rate": 9.996004351336743e-05, "loss": 0.0359, "step": 6210 }, { "grad_norm": 0.3883376717567444, "learning_rate": 9.995937988558785e-05, "loss": 0.0351, "step": 6220 }, { "grad_norm": 0.30743131041526794, "learning_rate": 9.995871079432561e-05, "loss": 0.0333, "step": 6230 }, { "grad_norm": 0.2947206497192383, "learning_rate": 9.995803623965389e-05, "loss": 0.0379, "step": 6240 }, { "grad_norm": 0.29608529806137085, "learning_rate": 9.995735622164641e-05, "loss": 0.0338, "step": 6250 }, { "grad_norm": 0.3887060582637787, "learning_rate": 9.995667074037758e-05, "loss": 0.0315, "step": 6260 }, { "grad_norm": 0.2611384391784668, "learning_rate": 9.995597979592232e-05, "loss": 0.0325, "step": 6270 }, { "grad_norm": 0.2719224989414215, "learning_rate": 9.995528338835625e-05, "loss": 0.0336, "step": 6280 }, { "grad_norm": 0.28054043650627136, "learning_rate": 9.995458151775547e-05, "loss": 0.036, "step": 6290 }, { "grad_norm": 0.45706623792648315, "learning_rate": 9.995387418419677e-05, "loss": 0.0325, "step": 6300 }, { "grad_norm": 0.4073222279548645, "learning_rate": 9.99531613877575e-05, "loss": 0.0319, "step": 6310 }, { "grad_norm": 0.3296639919281006, "learning_rate": 9.995244312851559e-05, "loss": 0.0329, "step": 6320 }, { "grad_norm": 0.43654075264930725, "learning_rate": 9.995171940654961e-05, "loss": 0.0342, "step": 6330 }, { "grad_norm": 0.47308146953582764, "learning_rate": 9.995099022193871e-05, "loss": 0.0373, "step": 6340 }, { "grad_norm": 0.4837159812450409, "learning_rate": 9.995025557476261e-05, "loss": 0.0407, "step": 6350 }, { "grad_norm": 0.3722970187664032, "learning_rate": 9.994951546510165e-05, "loss": 0.0352, "step": 6360 }, { "grad_norm": 0.41013649106025696, "learning_rate": 9.994876989303679e-05, "loss": 0.0347, "step": 6370 }, { "grad_norm": 0.3916681408882141, "learning_rate": 9.994801885864955e-05, "loss": 0.0331, "step": 6380 }, { "grad_norm": 0.4265499711036682, "learning_rate": 9.994726236202205e-05, "loss": 0.035, "step": 6390 }, { "grad_norm": 0.4204399585723877, "learning_rate": 9.994650040323704e-05, "loss": 0.0372, "step": 6400 }, { "grad_norm": 0.40114837884902954, "learning_rate": 9.994573298237784e-05, "loss": 0.0362, "step": 6410 }, { "grad_norm": 0.35254159569740295, "learning_rate": 9.994496009952837e-05, "loss": 0.0352, "step": 6420 }, { "grad_norm": 0.45781514048576355, "learning_rate": 9.994418175477316e-05, "loss": 0.0346, "step": 6430 }, { "grad_norm": 0.4171578884124756, "learning_rate": 9.994339794819733e-05, "loss": 0.035, "step": 6440 }, { "grad_norm": 0.3187430799007416, "learning_rate": 9.994260867988658e-05, "loss": 0.0318, "step": 6450 }, { "grad_norm": 0.3890190124511719, "learning_rate": 9.994181394992723e-05, "loss": 0.0334, "step": 6460 }, { "grad_norm": 0.34653133153915405, "learning_rate": 9.994101375840618e-05, "loss": 0.0286, "step": 6470 }, { "grad_norm": 0.26024141907691956, "learning_rate": 9.994020810541098e-05, "loss": 0.0318, "step": 6480 }, { "grad_norm": 0.43454793095588684, "learning_rate": 9.99393969910297e-05, "loss": 0.0333, "step": 6490 }, { "grad_norm": 0.34684625267982483, "learning_rate": 9.993858041535104e-05, "loss": 0.0344, "step": 6500 }, { "grad_norm": 0.32612162828445435, "learning_rate": 9.99377583784643e-05, "loss": 0.036, "step": 6510 }, { "grad_norm": 0.4974979758262634, "learning_rate": 9.993693088045939e-05, "loss": 0.033, "step": 6520 }, { "grad_norm": 0.31062519550323486, "learning_rate": 9.99360979214268e-05, "loss": 0.0357, "step": 6530 }, { "grad_norm": 0.34023845195770264, "learning_rate": 9.99352595014576e-05, "loss": 0.0285, "step": 6540 }, { "grad_norm": 0.42129629850387573, "learning_rate": 9.993441562064354e-05, "loss": 0.0332, "step": 6550 }, { "grad_norm": 0.5159804821014404, "learning_rate": 9.993356627907685e-05, "loss": 0.0353, "step": 6560 }, { "grad_norm": 0.3815820515155792, "learning_rate": 9.99327114768504e-05, "loss": 0.0343, "step": 6570 }, { "grad_norm": 0.3275502622127533, "learning_rate": 9.99318512140577e-05, "loss": 0.0292, "step": 6580 }, { "grad_norm": 0.4016001522541046, "learning_rate": 9.993098549079284e-05, "loss": 0.032, "step": 6590 }, { "grad_norm": 0.32022950053215027, "learning_rate": 9.993011430715047e-05, "loss": 0.0348, "step": 6600 }, { "grad_norm": 0.2868673801422119, "learning_rate": 9.992923766322586e-05, "loss": 0.0329, "step": 6610 }, { "grad_norm": 0.31561800837516785, "learning_rate": 9.99283555591149e-05, "loss": 0.035, "step": 6620 }, { "grad_norm": 0.3700001537799835, "learning_rate": 9.992746799491404e-05, "loss": 0.0368, "step": 6630 }, { "grad_norm": 0.38787204027175903, "learning_rate": 9.992657497072033e-05, "loss": 0.0344, "step": 6640 }, { "grad_norm": 0.42458412051200867, "learning_rate": 9.992567648663147e-05, "loss": 0.0362, "step": 6650 }, { "grad_norm": 0.38257884979248047, "learning_rate": 9.992477254274568e-05, "loss": 0.0306, "step": 6660 }, { "grad_norm": 0.4431650936603546, "learning_rate": 9.992386313916183e-05, "loss": 0.0366, "step": 6670 }, { "grad_norm": 0.3380863070487976, "learning_rate": 9.992294827597934e-05, "loss": 0.0352, "step": 6680 }, { "grad_norm": 0.27915555238723755, "learning_rate": 9.992202795329831e-05, "loss": 0.0329, "step": 6690 }, { "grad_norm": 0.24716027081012726, "learning_rate": 9.992110217121936e-05, "loss": 0.0305, "step": 6700 }, { "grad_norm": 0.3293974697589874, "learning_rate": 9.992017092984372e-05, "loss": 0.0318, "step": 6710 }, { "grad_norm": 0.29438358545303345, "learning_rate": 9.991923422927326e-05, "loss": 0.036, "step": 6720 }, { "grad_norm": 0.2939601242542267, "learning_rate": 9.991829206961037e-05, "loss": 0.0324, "step": 6730 }, { "grad_norm": 0.26215600967407227, "learning_rate": 9.991734445095813e-05, "loss": 0.0339, "step": 6740 }, { "grad_norm": 0.48552438616752625, "learning_rate": 9.991639137342015e-05, "loss": 0.0382, "step": 6750 }, { "grad_norm": 0.2703455686569214, "learning_rate": 9.991543283710064e-05, "loss": 0.0325, "step": 6760 }, { "grad_norm": 0.394000768661499, "learning_rate": 9.991446884210445e-05, "loss": 0.036, "step": 6770 }, { "grad_norm": 0.28995048999786377, "learning_rate": 9.9913499388537e-05, "loss": 0.0353, "step": 6780 }, { "grad_norm": 0.40333759784698486, "learning_rate": 9.99125244765043e-05, "loss": 0.0316, "step": 6790 }, { "grad_norm": 0.4176333546638489, "learning_rate": 9.991154410611296e-05, "loss": 0.0331, "step": 6800 }, { "grad_norm": 0.3138960599899292, "learning_rate": 9.99105582774702e-05, "loss": 0.0326, "step": 6810 }, { "grad_norm": 0.33803126215934753, "learning_rate": 9.990956699068384e-05, "loss": 0.0338, "step": 6820 }, { "grad_norm": 0.2949196994304657, "learning_rate": 9.990857024586224e-05, "loss": 0.0297, "step": 6830 }, { "grad_norm": 0.3500555157661438, "learning_rate": 9.990756804311446e-05, "loss": 0.0324, "step": 6840 }, { "grad_norm": 0.4491446018218994, "learning_rate": 9.990656038255006e-05, "loss": 0.0311, "step": 6850 }, { "grad_norm": 0.36266762018203735, "learning_rate": 9.990554726427926e-05, "loss": 0.0301, "step": 6860 }, { "grad_norm": 0.36502277851104736, "learning_rate": 9.990452868841284e-05, "loss": 0.0291, "step": 6870 }, { "grad_norm": 0.36486005783081055, "learning_rate": 9.99035046550622e-05, "loss": 0.0374, "step": 6880 }, { "grad_norm": 0.3429112732410431, "learning_rate": 9.99024751643393e-05, "loss": 0.0333, "step": 6890 }, { "grad_norm": 0.41208508610725403, "learning_rate": 9.990144021635677e-05, "loss": 0.0363, "step": 6900 }, { "grad_norm": 0.35970383882522583, "learning_rate": 9.990039981122775e-05, "loss": 0.0355, "step": 6910 }, { "grad_norm": 0.2844679355621338, "learning_rate": 9.989935394906602e-05, "loss": 0.0328, "step": 6920 }, { "grad_norm": 0.33791568875312805, "learning_rate": 9.989830262998598e-05, "loss": 0.0333, "step": 6930 }, { "grad_norm": 0.35294511914253235, "learning_rate": 9.989724585410259e-05, "loss": 0.0312, "step": 6940 }, { "grad_norm": 0.3217146098613739, "learning_rate": 9.989618362153139e-05, "loss": 0.034, "step": 6950 }, { "grad_norm": 0.34132376313209534, "learning_rate": 9.989511593238859e-05, "loss": 0.0338, "step": 6960 }, { "grad_norm": 0.3253345787525177, "learning_rate": 9.98940427867909e-05, "loss": 0.032, "step": 6970 }, { "grad_norm": 0.33602839708328247, "learning_rate": 9.989296418485573e-05, "loss": 0.0322, "step": 6980 }, { "grad_norm": 0.3279520571231842, "learning_rate": 9.989188012670101e-05, "loss": 0.0276, "step": 6990 }, { "grad_norm": 0.24078917503356934, "learning_rate": 9.989079061244528e-05, "loss": 0.0278, "step": 7000 }, { "grad_norm": 0.30250462889671326, "learning_rate": 9.988969564220769e-05, "loss": 0.0309, "step": 7010 }, { "grad_norm": 0.4196934401988983, "learning_rate": 9.988859521610801e-05, "loss": 0.0302, "step": 7020 }, { "grad_norm": 0.29674795269966125, "learning_rate": 9.988748933426656e-05, "loss": 0.0311, "step": 7030 }, { "grad_norm": 0.2873777747154236, "learning_rate": 9.988637799680428e-05, "loss": 0.0327, "step": 7040 }, { "grad_norm": 0.32827991247177124, "learning_rate": 9.98852612038427e-05, "loss": 0.0323, "step": 7050 }, { "grad_norm": 0.39969903230667114, "learning_rate": 9.988413895550397e-05, "loss": 0.0279, "step": 7060 }, { "grad_norm": 0.3224151134490967, "learning_rate": 9.98830112519108e-05, "loss": 0.0385, "step": 7070 }, { "grad_norm": 0.263328492641449, "learning_rate": 9.98818780931865e-05, "loss": 0.0334, "step": 7080 }, { "grad_norm": 0.27316489815711975, "learning_rate": 9.988073947945502e-05, "loss": 0.0323, "step": 7090 }, { "grad_norm": 0.30871060490608215, "learning_rate": 9.987959541084087e-05, "loss": 0.0322, "step": 7100 }, { "grad_norm": 0.33903345465660095, "learning_rate": 9.987844588746915e-05, "loss": 0.0301, "step": 7110 }, { "grad_norm": 0.2656480073928833, "learning_rate": 9.987729090946558e-05, "loss": 0.0312, "step": 7120 }, { "grad_norm": 0.4520304501056671, "learning_rate": 9.987613047695647e-05, "loss": 0.0375, "step": 7130 }, { "grad_norm": 0.37252986431121826, "learning_rate": 9.987496459006871e-05, "loss": 0.0338, "step": 7140 }, { "grad_norm": 0.28246259689331055, "learning_rate": 9.987379324892982e-05, "loss": 0.0304, "step": 7150 }, { "grad_norm": 0.3238445222377777, "learning_rate": 9.987261645366788e-05, "loss": 0.0336, "step": 7160 }, { "grad_norm": 0.4010528326034546, "learning_rate": 9.987143420441158e-05, "loss": 0.0316, "step": 7170 }, { "grad_norm": 0.3322905898094177, "learning_rate": 9.987024650129022e-05, "loss": 0.0327, "step": 7180 }, { "grad_norm": 0.3581978678703308, "learning_rate": 9.986905334443368e-05, "loss": 0.0329, "step": 7190 }, { "grad_norm": 0.32030609250068665, "learning_rate": 9.986785473397245e-05, "loss": 0.0305, "step": 7200 }, { "grad_norm": 0.29995355010032654, "learning_rate": 9.98666506700376e-05, "loss": 0.0309, "step": 7210 }, { "grad_norm": 0.3072746992111206, "learning_rate": 9.986544115276081e-05, "loss": 0.0264, "step": 7220 }, { "grad_norm": 0.2854437828063965, "learning_rate": 9.986422618227433e-05, "loss": 0.0355, "step": 7230 }, { "grad_norm": 0.2949868440628052, "learning_rate": 9.986300575871106e-05, "loss": 0.0332, "step": 7240 }, { "grad_norm": 0.34734854102134705, "learning_rate": 9.986177988220444e-05, "loss": 0.0273, "step": 7250 }, { "grad_norm": 0.40559324622154236, "learning_rate": 9.986054855288856e-05, "loss": 0.0303, "step": 7260 }, { "grad_norm": 0.2936643362045288, "learning_rate": 9.985931177089802e-05, "loss": 0.032, "step": 7270 }, { "grad_norm": 0.36499980092048645, "learning_rate": 9.985806953636814e-05, "loss": 0.0329, "step": 7280 }, { "grad_norm": 0.26685577630996704, "learning_rate": 9.985682184943471e-05, "loss": 0.0341, "step": 7290 }, { "grad_norm": 0.4045690596103668, "learning_rate": 9.98555687102342e-05, "loss": 0.0338, "step": 7300 }, { "grad_norm": 0.3400069773197174, "learning_rate": 9.985431011890367e-05, "loss": 0.0317, "step": 7310 }, { "grad_norm": 0.32755130529403687, "learning_rate": 9.985304607558075e-05, "loss": 0.0362, "step": 7320 }, { "grad_norm": 0.31668898463249207, "learning_rate": 9.985177658040364e-05, "loss": 0.0298, "step": 7330 }, { "grad_norm": 0.3414134383201599, "learning_rate": 9.985050163351119e-05, "loss": 0.0305, "step": 7340 }, { "grad_norm": 0.2794857621192932, "learning_rate": 9.984922123504286e-05, "loss": 0.0307, "step": 7350 }, { "grad_norm": 0.34399017691612244, "learning_rate": 9.984793538513862e-05, "loss": 0.0373, "step": 7360 }, { "grad_norm": 0.33952227234840393, "learning_rate": 9.984664408393912e-05, "loss": 0.0345, "step": 7370 }, { "grad_norm": 0.34355440735816956, "learning_rate": 9.984534733158556e-05, "loss": 0.0303, "step": 7380 }, { "grad_norm": 0.29568928480148315, "learning_rate": 9.984404512821977e-05, "loss": 0.0337, "step": 7390 }, { "grad_norm": 0.3394313156604767, "learning_rate": 9.984273747398411e-05, "loss": 0.0297, "step": 7400 }, { "grad_norm": 0.3535849452018738, "learning_rate": 9.984142436902165e-05, "loss": 0.0294, "step": 7410 }, { "grad_norm": 0.255767822265625, "learning_rate": 9.984010581347596e-05, "loss": 0.0302, "step": 7420 }, { "grad_norm": 0.221689835190773, "learning_rate": 9.983878180749121e-05, "loss": 0.0285, "step": 7430 }, { "grad_norm": 0.2865849733352661, "learning_rate": 9.983745235121222e-05, "loss": 0.0278, "step": 7440 }, { "grad_norm": 0.24127978086471558, "learning_rate": 9.983611744478438e-05, "loss": 0.0261, "step": 7450 }, { "grad_norm": 0.2947443723678589, "learning_rate": 9.983477708835365e-05, "loss": 0.0317, "step": 7460 }, { "grad_norm": 0.32391297817230225, "learning_rate": 9.983343128206664e-05, "loss": 0.0299, "step": 7470 }, { "grad_norm": 0.43490326404571533, "learning_rate": 9.983208002607049e-05, "loss": 0.0295, "step": 7480 }, { "grad_norm": 0.34243807196617126, "learning_rate": 9.9830723320513e-05, "loss": 0.0288, "step": 7490 }, { "grad_norm": 0.3136695921421051, "learning_rate": 9.982936116554254e-05, "loss": 0.0344, "step": 7500 }, { "grad_norm": 0.38297420740127563, "learning_rate": 9.982799356130803e-05, "loss": 0.0343, "step": 7510 }, { "grad_norm": 0.32201141119003296, "learning_rate": 9.982662050795908e-05, "loss": 0.0388, "step": 7520 }, { "grad_norm": 0.3504529297351837, "learning_rate": 9.982524200564583e-05, "loss": 0.0322, "step": 7530 }, { "grad_norm": 0.2974012494087219, "learning_rate": 9.982385805451901e-05, "loss": 0.0299, "step": 7540 }, { "grad_norm": 0.339059442281723, "learning_rate": 9.982246865472998e-05, "loss": 0.0298, "step": 7550 }, { "grad_norm": 0.5149786472320557, "learning_rate": 9.982107380643069e-05, "loss": 0.0369, "step": 7560 }, { "grad_norm": 0.36339008808135986, "learning_rate": 9.981967350977368e-05, "loss": 0.0318, "step": 7570 }, { "grad_norm": 0.32982566952705383, "learning_rate": 9.981826776491208e-05, "loss": 0.033, "step": 7580 }, { "grad_norm": 0.33873239159584045, "learning_rate": 9.98168565719996e-05, "loss": 0.0332, "step": 7590 }, { "grad_norm": 0.3394455313682556, "learning_rate": 9.98154399311906e-05, "loss": 0.0333, "step": 7600 }, { "grad_norm": 0.4005190134048462, "learning_rate": 9.981401784263997e-05, "loss": 0.0318, "step": 7610 }, { "grad_norm": 0.40064066648483276, "learning_rate": 9.981259030650326e-05, "loss": 0.0301, "step": 7620 }, { "grad_norm": 0.3094565272331238, "learning_rate": 9.981115732293655e-05, "loss": 0.0309, "step": 7630 }, { "grad_norm": 0.28395313024520874, "learning_rate": 9.980971889209659e-05, "loss": 0.0297, "step": 7640 }, { "grad_norm": 0.23985162377357483, "learning_rate": 9.980827501414064e-05, "loss": 0.0292, "step": 7650 }, { "grad_norm": 0.4320547878742218, "learning_rate": 9.980682568922663e-05, "loss": 0.0365, "step": 7660 }, { "grad_norm": 0.3326107859611511, "learning_rate": 9.980537091751304e-05, "loss": 0.033, "step": 7670 }, { "grad_norm": 0.34216248989105225, "learning_rate": 9.980391069915897e-05, "loss": 0.0289, "step": 7680 }, { "grad_norm": 0.2872937023639679, "learning_rate": 9.98024450343241e-05, "loss": 0.0267, "step": 7690 }, { "grad_norm": 0.3010108470916748, "learning_rate": 9.980097392316872e-05, "loss": 0.033, "step": 7700 }, { "grad_norm": 0.3231135606765747, "learning_rate": 9.97994973658537e-05, "loss": 0.0327, "step": 7710 }, { "grad_norm": 0.33956411480903625, "learning_rate": 9.979801536254054e-05, "loss": 0.0306, "step": 7720 }, { "grad_norm": 0.2985449731349945, "learning_rate": 9.979652791339127e-05, "loss": 0.0322, "step": 7730 }, { "grad_norm": 0.28136980533599854, "learning_rate": 9.97950350185686e-05, "loss": 0.0297, "step": 7740 }, { "grad_norm": 0.3667967617511749, "learning_rate": 9.979353667823574e-05, "loss": 0.0331, "step": 7750 }, { "grad_norm": 0.27970126271247864, "learning_rate": 9.979203289255658e-05, "loss": 0.0308, "step": 7760 }, { "grad_norm": 0.2396286278963089, "learning_rate": 9.979052366169557e-05, "loss": 0.0371, "step": 7770 }, { "grad_norm": 0.33535680174827576, "learning_rate": 9.978900898581775e-05, "loss": 0.0312, "step": 7780 }, { "grad_norm": 0.299522340297699, "learning_rate": 9.978748886508875e-05, "loss": 0.0295, "step": 7790 }, { "grad_norm": 0.3098130226135254, "learning_rate": 9.978596329967484e-05, "loss": 0.0325, "step": 7800 }, { "grad_norm": 0.27049335837364197, "learning_rate": 9.978443228974284e-05, "loss": 0.0302, "step": 7810 }, { "grad_norm": 0.32331812381744385, "learning_rate": 9.978289583546015e-05, "loss": 0.0269, "step": 7820 }, { "grad_norm": 0.23372115194797516, "learning_rate": 9.978135393699484e-05, "loss": 0.0305, "step": 7830 }, { "grad_norm": 0.3448472321033478, "learning_rate": 9.977980659451548e-05, "loss": 0.0276, "step": 7840 }, { "grad_norm": 0.3155830204486847, "learning_rate": 9.977825380819135e-05, "loss": 0.0291, "step": 7850 }, { "grad_norm": 0.3932153284549713, "learning_rate": 9.97766955781922e-05, "loss": 0.0342, "step": 7860 }, { "grad_norm": 0.3314362168312073, "learning_rate": 9.977513190468848e-05, "loss": 0.0319, "step": 7870 }, { "grad_norm": 0.39465293288230896, "learning_rate": 9.977356278785116e-05, "loss": 0.0304, "step": 7880 }, { "grad_norm": 0.421457976102829, "learning_rate": 9.977198822785184e-05, "loss": 0.0327, "step": 7890 }, { "grad_norm": 0.34070318937301636, "learning_rate": 9.977040822486273e-05, "loss": 0.0308, "step": 7900 }, { "grad_norm": 0.29190587997436523, "learning_rate": 9.97688227790566e-05, "loss": 0.0275, "step": 7910 }, { "grad_norm": 0.295036643743515, "learning_rate": 9.976723189060684e-05, "loss": 0.0292, "step": 7920 }, { "grad_norm": 0.2600986361503601, "learning_rate": 9.976563555968742e-05, "loss": 0.0281, "step": 7930 }, { "grad_norm": 0.2755002975463867, "learning_rate": 9.976403378647292e-05, "loss": 0.0306, "step": 7940 }, { "grad_norm": 0.30681806802749634, "learning_rate": 9.97624265711385e-05, "loss": 0.0295, "step": 7950 }, { "grad_norm": 0.2757038474082947, "learning_rate": 9.976081391385993e-05, "loss": 0.0339, "step": 7960 }, { "grad_norm": 0.21889804303646088, "learning_rate": 9.975919581481356e-05, "loss": 0.0351, "step": 7970 }, { "grad_norm": 0.3433792293071747, "learning_rate": 9.975757227417634e-05, "loss": 0.0321, "step": 7980 }, { "grad_norm": 0.33976513147354126, "learning_rate": 9.975594329212586e-05, "loss": 0.03, "step": 7990 }, { "grad_norm": 0.42543408274650574, "learning_rate": 9.97543088688402e-05, "loss": 0.0292, "step": 8000 }, { "grad_norm": 0.26971906423568726, "learning_rate": 9.975266900449814e-05, "loss": 0.026, "step": 8010 }, { "grad_norm": 0.3774315416812897, "learning_rate": 9.975102369927898e-05, "loss": 0.0313, "step": 8020 }, { "grad_norm": 0.2927657961845398, "learning_rate": 9.974937295336269e-05, "loss": 0.0301, "step": 8030 }, { "grad_norm": 0.26062896847724915, "learning_rate": 9.974771676692975e-05, "loss": 0.0287, "step": 8040 }, { "grad_norm": 0.36228886246681213, "learning_rate": 9.974605514016131e-05, "loss": 0.0274, "step": 8050 }, { "grad_norm": 0.32144784927368164, "learning_rate": 9.974438807323907e-05, "loss": 0.0303, "step": 8060 }, { "grad_norm": 0.3173360824584961, "learning_rate": 9.974271556634535e-05, "loss": 0.0314, "step": 8070 }, { "grad_norm": 0.3535872995853424, "learning_rate": 9.974103761966302e-05, "loss": 0.0317, "step": 8080 }, { "grad_norm": 0.24759623408317566, "learning_rate": 9.973935423337563e-05, "loss": 0.0312, "step": 8090 }, { "grad_norm": 0.37784892320632935, "learning_rate": 9.973766540766722e-05, "loss": 0.0332, "step": 8100 }, { "grad_norm": 0.3444628119468689, "learning_rate": 9.97359711427225e-05, "loss": 0.0299, "step": 8110 }, { "grad_norm": 0.4536915719509125, "learning_rate": 9.973427143872677e-05, "loss": 0.032, "step": 8120 }, { "grad_norm": 0.3261953592300415, "learning_rate": 9.973256629586589e-05, "loss": 0.031, "step": 8130 }, { "grad_norm": 0.3856222927570343, "learning_rate": 9.973085571432632e-05, "loss": 0.0288, "step": 8140 }, { "grad_norm": 0.32769086956977844, "learning_rate": 9.972913969429513e-05, "loss": 0.0314, "step": 8150 }, { "grad_norm": 0.28611695766448975, "learning_rate": 9.972741823596e-05, "loss": 0.0276, "step": 8160 }, { "grad_norm": 0.2987666428089142, "learning_rate": 9.972569133950917e-05, "loss": 0.028, "step": 8170 }, { "grad_norm": 0.3181549310684204, "learning_rate": 9.972395900513151e-05, "loss": 0.0268, "step": 8180 }, { "grad_norm": 0.3448394238948822, "learning_rate": 9.972222123301645e-05, "loss": 0.0301, "step": 8190 }, { "grad_norm": 0.30785951018333435, "learning_rate": 9.972047802335403e-05, "loss": 0.0295, "step": 8200 }, { "grad_norm": 0.35775962471961975, "learning_rate": 9.971872937633488e-05, "loss": 0.026, "step": 8210 }, { "grad_norm": 0.3293589949607849, "learning_rate": 9.971697529215024e-05, "loss": 0.0258, "step": 8220 }, { "grad_norm": 0.3130868971347809, "learning_rate": 9.971521577099192e-05, "loss": 0.0279, "step": 8230 }, { "grad_norm": 0.33256995677948, "learning_rate": 9.971345081305236e-05, "loss": 0.0273, "step": 8240 }, { "grad_norm": 0.3467409014701843, "learning_rate": 9.971168041852456e-05, "loss": 0.0288, "step": 8250 }, { "grad_norm": 0.3618285059928894, "learning_rate": 9.970990458760215e-05, "loss": 0.027, "step": 8260 }, { "grad_norm": 0.34067681431770325, "learning_rate": 9.970812332047929e-05, "loss": 0.0319, "step": 8270 }, { "grad_norm": 0.3538804054260254, "learning_rate": 9.97063366173508e-05, "loss": 0.0299, "step": 8280 }, { "grad_norm": 0.5606152415275574, "learning_rate": 9.970454447841207e-05, "loss": 0.0336, "step": 8290 }, { "grad_norm": 0.4576495587825775, "learning_rate": 9.970274690385909e-05, "loss": 0.031, "step": 8300 }, { "grad_norm": 0.3145383894443512, "learning_rate": 9.970094389388844e-05, "loss": 0.034, "step": 8310 }, { "grad_norm": 0.3158317506313324, "learning_rate": 9.969913544869728e-05, "loss": 0.028, "step": 8320 }, { "grad_norm": 0.40648558735847473, "learning_rate": 9.96973215684834e-05, "loss": 0.0331, "step": 8330 }, { "grad_norm": 0.40808799862861633, "learning_rate": 9.969550225344513e-05, "loss": 0.0305, "step": 8340 }, { "grad_norm": 0.35204076766967773, "learning_rate": 9.969367750378147e-05, "loss": 0.0332, "step": 8350 }, { "grad_norm": 0.29632872343063354, "learning_rate": 9.969184731969194e-05, "loss": 0.0319, "step": 8360 }, { "grad_norm": 0.286072701215744, "learning_rate": 9.96900117013767e-05, "loss": 0.0276, "step": 8370 }, { "grad_norm": 0.36736762523651123, "learning_rate": 9.96881706490365e-05, "loss": 0.0325, "step": 8380 }, { "grad_norm": 0.3015446960926056, "learning_rate": 9.968632416287265e-05, "loss": 0.0298, "step": 8390 }, { "grad_norm": 0.31737029552459717, "learning_rate": 9.96844722430871e-05, "loss": 0.0313, "step": 8400 }, { "grad_norm": 0.284322589635849, "learning_rate": 9.968261488988235e-05, "loss": 0.0332, "step": 8410 }, { "grad_norm": 0.3076310455799103, "learning_rate": 9.968075210346155e-05, "loss": 0.029, "step": 8420 }, { "grad_norm": 0.27338963747024536, "learning_rate": 9.967888388402839e-05, "loss": 0.0296, "step": 8430 }, { "grad_norm": 0.2553826868534088, "learning_rate": 9.967701023178717e-05, "loss": 0.0297, "step": 8440 }, { "grad_norm": 0.3608287274837494, "learning_rate": 9.967513114694282e-05, "loss": 0.0371, "step": 8450 }, { "grad_norm": 0.42506152391433716, "learning_rate": 9.967324662970079e-05, "loss": 0.0279, "step": 8460 }, { "grad_norm": 0.28671255707740784, "learning_rate": 9.96713566802672e-05, "loss": 0.0317, "step": 8470 }, { "grad_norm": 0.22029690444469452, "learning_rate": 9.966946129884873e-05, "loss": 0.0289, "step": 8480 }, { "grad_norm": 0.34959664940834045, "learning_rate": 9.966756048565265e-05, "loss": 0.0302, "step": 8490 }, { "grad_norm": 0.37993475794792175, "learning_rate": 9.966565424088681e-05, "loss": 0.0342, "step": 8500 }, { "grad_norm": 0.30413275957107544, "learning_rate": 9.96637425647597e-05, "loss": 0.0343, "step": 8510 }, { "grad_norm": 0.3540544807910919, "learning_rate": 9.966182545748038e-05, "loss": 0.0353, "step": 8520 }, { "grad_norm": 0.32841426134109497, "learning_rate": 9.96599029192585e-05, "loss": 0.0333, "step": 8530 }, { "grad_norm": 0.27233392000198364, "learning_rate": 9.965797495030428e-05, "loss": 0.0311, "step": 8540 }, { "grad_norm": 0.2651784121990204, "learning_rate": 9.96560415508286e-05, "loss": 0.0284, "step": 8550 }, { "grad_norm": 0.3175239562988281, "learning_rate": 9.965410272104286e-05, "loss": 0.0293, "step": 8560 }, { "grad_norm": 0.5718841552734375, "learning_rate": 9.96521584611591e-05, "loss": 0.0371, "step": 8570 }, { "grad_norm": 0.34320807456970215, "learning_rate": 9.965020877138994e-05, "loss": 0.0329, "step": 8580 }, { "grad_norm": 0.4012939929962158, "learning_rate": 9.964825365194861e-05, "loss": 0.0322, "step": 8590 }, { "grad_norm": 0.484631210565567, "learning_rate": 9.96462931030489e-05, "loss": 0.0322, "step": 8600 }, { "grad_norm": 0.4259701073169708, "learning_rate": 9.96443271249052e-05, "loss": 0.0344, "step": 8610 }, { "grad_norm": 0.3595598042011261, "learning_rate": 9.964235571773255e-05, "loss": 0.032, "step": 8620 }, { "grad_norm": 0.2948048412799835, "learning_rate": 9.96403788817465e-05, "loss": 0.0333, "step": 8630 }, { "grad_norm": 0.3027276396751404, "learning_rate": 9.963839661716325e-05, "loss": 0.0314, "step": 8640 }, { "grad_norm": 0.3616979718208313, "learning_rate": 9.963640892419958e-05, "loss": 0.0318, "step": 8650 }, { "grad_norm": 0.2621501684188843, "learning_rate": 9.963441580307286e-05, "loss": 0.0303, "step": 8660 }, { "grad_norm": 0.3506384789943695, "learning_rate": 9.963241725400104e-05, "loss": 0.0323, "step": 8670 }, { "grad_norm": 0.24135568737983704, "learning_rate": 9.963041327720271e-05, "loss": 0.0328, "step": 8680 }, { "grad_norm": 0.23337817192077637, "learning_rate": 9.962840387289697e-05, "loss": 0.025, "step": 8690 }, { "grad_norm": 0.27965667843818665, "learning_rate": 9.962638904130363e-05, "loss": 0.0269, "step": 8700 }, { "grad_norm": 0.26577168703079224, "learning_rate": 9.962436878264298e-05, "loss": 0.0285, "step": 8710 }, { "grad_norm": 0.32297244668006897, "learning_rate": 9.962234309713598e-05, "loss": 0.0279, "step": 8720 }, { "grad_norm": 0.3234224319458008, "learning_rate": 9.962031198500414e-05, "loss": 0.0292, "step": 8730 }, { "grad_norm": 0.28006431460380554, "learning_rate": 9.961827544646958e-05, "loss": 0.0283, "step": 8740 }, { "grad_norm": 0.41857078671455383, "learning_rate": 9.961623348175501e-05, "loss": 0.0273, "step": 8750 }, { "grad_norm": 0.40531739592552185, "learning_rate": 9.961418609108377e-05, "loss": 0.0275, "step": 8760 }, { "grad_norm": 0.3201572299003601, "learning_rate": 9.961213327467971e-05, "loss": 0.0317, "step": 8770 }, { "grad_norm": 0.2479609251022339, "learning_rate": 9.961007503276736e-05, "loss": 0.0331, "step": 8780 }, { "grad_norm": 0.3467199504375458, "learning_rate": 9.960801136557179e-05, "loss": 0.0325, "step": 8790 }, { "grad_norm": 0.338176429271698, "learning_rate": 9.960594227331866e-05, "loss": 0.0258, "step": 8800 }, { "grad_norm": 0.23905061185359955, "learning_rate": 9.960386775623429e-05, "loss": 0.0262, "step": 8810 }, { "grad_norm": 0.391560435295105, "learning_rate": 9.96017878145455e-05, "loss": 0.0321, "step": 8820 }, { "grad_norm": 0.2744453549385071, "learning_rate": 9.959970244847977e-05, "loss": 0.0291, "step": 8830 }, { "grad_norm": 0.35254722833633423, "learning_rate": 9.959761165826518e-05, "loss": 0.0354, "step": 8840 }, { "grad_norm": 0.3213765323162079, "learning_rate": 9.959551544413033e-05, "loss": 0.0326, "step": 8850 }, { "grad_norm": 0.30613622069358826, "learning_rate": 9.959341380630448e-05, "loss": 0.0284, "step": 8860 }, { "grad_norm": 0.3359206020832062, "learning_rate": 9.959130674501746e-05, "loss": 0.0306, "step": 8870 }, { "grad_norm": 0.2656834125518799, "learning_rate": 9.958919426049968e-05, "loss": 0.0284, "step": 8880 }, { "grad_norm": 0.27055129408836365, "learning_rate": 9.958707635298219e-05, "loss": 0.0293, "step": 8890 }, { "grad_norm": 0.3041641116142273, "learning_rate": 9.958495302269657e-05, "loss": 0.0294, "step": 8900 }, { "grad_norm": 0.3168145418167114, "learning_rate": 9.958282426987503e-05, "loss": 0.0297, "step": 8910 }, { "grad_norm": 0.30518296360969543, "learning_rate": 9.95806900947504e-05, "loss": 0.0285, "step": 8920 }, { "grad_norm": 0.38940054178237915, "learning_rate": 9.957855049755604e-05, "loss": 0.035, "step": 8930 }, { "grad_norm": 0.3513416051864624, "learning_rate": 9.957640547852593e-05, "loss": 0.0326, "step": 8940 }, { "grad_norm": 0.31455978751182556, "learning_rate": 9.957425503789466e-05, "loss": 0.0306, "step": 8950 }, { "grad_norm": 0.26485151052474976, "learning_rate": 9.957209917589738e-05, "loss": 0.0301, "step": 8960 }, { "grad_norm": 0.2996211349964142, "learning_rate": 9.956993789276987e-05, "loss": 0.0284, "step": 8970 }, { "grad_norm": 0.3105311095714569, "learning_rate": 9.956777118874847e-05, "loss": 0.0295, "step": 8980 }, { "grad_norm": 0.2919618785381317, "learning_rate": 9.956559906407016e-05, "loss": 0.0333, "step": 8990 }, { "grad_norm": 0.2621149718761444, "learning_rate": 9.956342151897245e-05, "loss": 0.0286, "step": 9000 }, { "grad_norm": 0.34400486946105957, "learning_rate": 9.956123855369346e-05, "loss": 0.0313, "step": 9010 }, { "grad_norm": 0.27371135354042053, "learning_rate": 9.955905016847196e-05, "loss": 0.0318, "step": 9020 }, { "grad_norm": 0.3672904074192047, "learning_rate": 9.955685636354723e-05, "loss": 0.0313, "step": 9030 }, { "grad_norm": 0.28457725048065186, "learning_rate": 9.95546571391592e-05, "loss": 0.0265, "step": 9040 }, { "grad_norm": 0.27510833740234375, "learning_rate": 9.955245249554837e-05, "loss": 0.027, "step": 9050 }, { "grad_norm": 0.2382332980632782, "learning_rate": 9.955024243295582e-05, "loss": 0.0279, "step": 9060 }, { "grad_norm": 0.3440762758255005, "learning_rate": 9.954802695162328e-05, "loss": 0.0268, "step": 9070 }, { "grad_norm": 0.3067493736743927, "learning_rate": 9.954580605179302e-05, "loss": 0.0256, "step": 9080 }, { "grad_norm": 0.3366178274154663, "learning_rate": 9.954357973370788e-05, "loss": 0.0272, "step": 9090 }, { "grad_norm": 0.3012387156486511, "learning_rate": 9.954134799761135e-05, "loss": 0.0245, "step": 9100 }, { "grad_norm": 0.2461877018213272, "learning_rate": 9.953911084374748e-05, "loss": 0.0242, "step": 9110 }, { "grad_norm": 0.27775219082832336, "learning_rate": 9.953686827236093e-05, "loss": 0.0295, "step": 9120 }, { "grad_norm": 0.3560655117034912, "learning_rate": 9.953462028369695e-05, "loss": 0.0331, "step": 9130 }, { "grad_norm": 0.3251045346260071, "learning_rate": 9.953236687800136e-05, "loss": 0.0277, "step": 9140 }, { "grad_norm": 0.29841703176498413, "learning_rate": 9.95301080555206e-05, "loss": 0.032, "step": 9150 }, { "grad_norm": 0.359862357378006, "learning_rate": 9.952784381650171e-05, "loss": 0.0277, "step": 9160 }, { "grad_norm": 0.21756939589977264, "learning_rate": 9.952557416119226e-05, "loss": 0.0303, "step": 9170 }, { "grad_norm": 0.24599778652191162, "learning_rate": 9.95232990898405e-05, "loss": 0.0283, "step": 9180 }, { "grad_norm": 0.3325737714767456, "learning_rate": 9.95210186026952e-05, "loss": 0.0294, "step": 9190 }, { "grad_norm": 0.4372248947620392, "learning_rate": 9.951873270000576e-05, "loss": 0.0312, "step": 9200 }, { "grad_norm": 0.3771456182003021, "learning_rate": 9.951644138202216e-05, "loss": 0.0331, "step": 9210 }, { "grad_norm": 0.31642353534698486, "learning_rate": 9.951414464899498e-05, "loss": 0.0326, "step": 9220 }, { "grad_norm": 0.2666427195072174, "learning_rate": 9.951184250117538e-05, "loss": 0.0295, "step": 9230 }, { "grad_norm": 0.29721173644065857, "learning_rate": 9.950953493881513e-05, "loss": 0.0262, "step": 9240 }, { "grad_norm": 0.3279067873954773, "learning_rate": 9.950722196216658e-05, "loss": 0.0295, "step": 9250 }, { "grad_norm": 0.2911849319934845, "learning_rate": 9.950490357148265e-05, "loss": 0.0312, "step": 9260 }, { "grad_norm": 0.46096503734588623, "learning_rate": 9.950257976701692e-05, "loss": 0.0332, "step": 9270 }, { "grad_norm": 0.3180677890777588, "learning_rate": 9.950025054902348e-05, "loss": 0.0329, "step": 9280 }, { "grad_norm": 0.33115828037261963, "learning_rate": 9.949791591775706e-05, "loss": 0.0388, "step": 9290 }, { "grad_norm": 0.40674325823783875, "learning_rate": 9.949557587347298e-05, "loss": 0.0354, "step": 9300 }, { "grad_norm": 0.29299718141555786, "learning_rate": 9.949323041642713e-05, "loss": 0.0301, "step": 9310 }, { "grad_norm": 0.2120637148618698, "learning_rate": 9.949087954687602e-05, "loss": 0.028, "step": 9320 }, { "grad_norm": 0.2865990698337555, "learning_rate": 9.948852326507672e-05, "loss": 0.0272, "step": 9330 }, { "grad_norm": 0.25748762488365173, "learning_rate": 9.948616157128694e-05, "loss": 0.033, "step": 9340 }, { "grad_norm": 0.2908293306827545, "learning_rate": 9.948379446576493e-05, "loss": 0.0297, "step": 9350 }, { "grad_norm": 0.3831382393836975, "learning_rate": 9.948142194876952e-05, "loss": 0.0277, "step": 9360 }, { "grad_norm": 0.3132159113883972, "learning_rate": 9.947904402056024e-05, "loss": 0.0263, "step": 9370 }, { "grad_norm": 0.2908787131309509, "learning_rate": 9.947666068139708e-05, "loss": 0.0251, "step": 9380 }, { "grad_norm": 0.26148608326911926, "learning_rate": 9.947427193154071e-05, "loss": 0.0279, "step": 9390 }, { "grad_norm": 0.2620199918746948, "learning_rate": 9.947187777125233e-05, "loss": 0.0293, "step": 9400 }, { "grad_norm": 0.308754563331604, "learning_rate": 9.946947820079377e-05, "loss": 0.0274, "step": 9410 }, { "grad_norm": 0.24417582154273987, "learning_rate": 9.946707322042747e-05, "loss": 0.027, "step": 9420 }, { "grad_norm": 0.2505521774291992, "learning_rate": 9.94646628304164e-05, "loss": 0.0293, "step": 9430 }, { "grad_norm": 0.33254384994506836, "learning_rate": 9.946224703102418e-05, "loss": 0.0274, "step": 9440 }, { "grad_norm": 0.26486751437187195, "learning_rate": 9.945982582251498e-05, "loss": 0.0276, "step": 9450 }, { "grad_norm": 0.3012474775314331, "learning_rate": 9.94573992051536e-05, "loss": 0.0274, "step": 9460 }, { "grad_norm": 0.3227504789829254, "learning_rate": 9.94549671792054e-05, "loss": 0.0285, "step": 9470 }, { "grad_norm": 0.2811976671218872, "learning_rate": 9.945252974493635e-05, "loss": 0.0278, "step": 9480 }, { "grad_norm": 0.2458571046590805, "learning_rate": 9.9450086902613e-05, "loss": 0.0266, "step": 9490 }, { "grad_norm": 0.202414870262146, "learning_rate": 9.944763865250248e-05, "loss": 0.0224, "step": 9500 }, { "grad_norm": 0.33378833532333374, "learning_rate": 9.944518499487254e-05, "loss": 0.0248, "step": 9510 }, { "grad_norm": 0.2749314308166504, "learning_rate": 9.944272592999151e-05, "loss": 0.0253, "step": 9520 }, { "grad_norm": 0.397843599319458, "learning_rate": 9.94402614581283e-05, "loss": 0.0225, "step": 9530 }, { "grad_norm": 0.33306390047073364, "learning_rate": 9.943779157955244e-05, "loss": 0.0266, "step": 9540 }, { "grad_norm": 0.25829893350601196, "learning_rate": 9.943531629453403e-05, "loss": 0.0267, "step": 9550 }, { "grad_norm": 0.2688533067703247, "learning_rate": 9.943283560334375e-05, "loss": 0.0246, "step": 9560 }, { "grad_norm": 0.24257101118564606, "learning_rate": 9.943034950625288e-05, "loss": 0.0259, "step": 9570 }, { "grad_norm": 0.28164994716644287, "learning_rate": 9.942785800353332e-05, "loss": 0.0271, "step": 9580 }, { "grad_norm": 0.24064886569976807, "learning_rate": 9.942536109545751e-05, "loss": 0.0271, "step": 9590 }, { "grad_norm": 0.36414557695388794, "learning_rate": 9.942285878229853e-05, "loss": 0.0259, "step": 9600 }, { "grad_norm": 0.3448335826396942, "learning_rate": 9.942035106433001e-05, "loss": 0.0258, "step": 9610 }, { "grad_norm": 0.2585966885089874, "learning_rate": 9.94178379418262e-05, "loss": 0.0266, "step": 9620 }, { "grad_norm": 0.2803003191947937, "learning_rate": 9.941531941506194e-05, "loss": 0.0245, "step": 9630 }, { "grad_norm": 0.26758745312690735, "learning_rate": 9.941279548431263e-05, "loss": 0.0285, "step": 9640 }, { "grad_norm": 0.2863309979438782, "learning_rate": 9.941026614985431e-05, "loss": 0.0266, "step": 9650 }, { "grad_norm": 0.2663835883140564, "learning_rate": 9.940773141196357e-05, "loss": 0.0259, "step": 9660 }, { "grad_norm": 0.29245275259017944, "learning_rate": 9.94051912709176e-05, "loss": 0.0273, "step": 9670 }, { "grad_norm": 0.22333931922912598, "learning_rate": 9.940264572699421e-05, "loss": 0.025, "step": 9680 }, { "grad_norm": 0.260135293006897, "learning_rate": 9.940009478047174e-05, "loss": 0.0248, "step": 9690 }, { "grad_norm": 0.3325978219509125, "learning_rate": 9.939753843162918e-05, "loss": 0.0255, "step": 9700 }, { "grad_norm": 0.27577823400497437, "learning_rate": 9.939497668074609e-05, "loss": 0.0271, "step": 9710 }, { "grad_norm": 0.27605006098747253, "learning_rate": 9.93924095281026e-05, "loss": 0.0293, "step": 9720 }, { "grad_norm": 0.249272882938385, "learning_rate": 9.938983697397948e-05, "loss": 0.0245, "step": 9730 }, { "grad_norm": 0.27651703357696533, "learning_rate": 9.938725901865805e-05, "loss": 0.0299, "step": 9740 }, { "grad_norm": 0.28450262546539307, "learning_rate": 9.93846756624202e-05, "loss": 0.0282, "step": 9750 }, { "grad_norm": 0.24413305521011353, "learning_rate": 9.938208690554849e-05, "loss": 0.0259, "step": 9760 }, { "grad_norm": 0.30975961685180664, "learning_rate": 9.9379492748326e-05, "loss": 0.0257, "step": 9770 }, { "grad_norm": 0.5377655029296875, "learning_rate": 9.937689319103641e-05, "loss": 0.0279, "step": 9780 }, { "grad_norm": 0.2735133469104767, "learning_rate": 9.937428823396404e-05, "loss": 0.0282, "step": 9790 }, { "grad_norm": 0.2771715819835663, "learning_rate": 9.937167787739372e-05, "loss": 0.0249, "step": 9800 }, { "grad_norm": 0.2538471221923828, "learning_rate": 9.936906212161095e-05, "loss": 0.0238, "step": 9810 }, { "grad_norm": 0.22177726030349731, "learning_rate": 9.936644096690176e-05, "loss": 0.0232, "step": 9820 }, { "grad_norm": 0.3994300067424774, "learning_rate": 9.936381441355282e-05, "loss": 0.027, "step": 9830 }, { "grad_norm": 0.35256528854370117, "learning_rate": 9.936118246185136e-05, "loss": 0.0269, "step": 9840 }, { "grad_norm": 0.2549561560153961, "learning_rate": 9.935854511208518e-05, "loss": 0.032, "step": 9850 }, { "grad_norm": 0.32011497020721436, "learning_rate": 9.935590236454272e-05, "loss": 0.0281, "step": 9860 }, { "grad_norm": 0.3262996971607208, "learning_rate": 9.935325421951298e-05, "loss": 0.0283, "step": 9870 }, { "grad_norm": 0.2824798822402954, "learning_rate": 9.935060067728557e-05, "loss": 0.0293, "step": 9880 }, { "grad_norm": 0.24285288155078888, "learning_rate": 9.934794173815067e-05, "loss": 0.0269, "step": 9890 }, { "grad_norm": 0.2961962819099426, "learning_rate": 9.934527740239906e-05, "loss": 0.0272, "step": 9900 }, { "grad_norm": 0.36524906754493713, "learning_rate": 9.934260767032209e-05, "loss": 0.0268, "step": 9910 }, { "grad_norm": 0.4639340937137604, "learning_rate": 9.933993254221172e-05, "loss": 0.0299, "step": 9920 }, { "grad_norm": 0.2717544734477997, "learning_rate": 9.933725201836053e-05, "loss": 0.0313, "step": 9930 }, { "grad_norm": 0.3427317142486572, "learning_rate": 9.933456609906162e-05, "loss": 0.0259, "step": 9940 }, { "grad_norm": 0.23999840021133423, "learning_rate": 9.933187478460875e-05, "loss": 0.0288, "step": 9950 }, { "grad_norm": 0.2087501585483551, "learning_rate": 9.93291780752962e-05, "loss": 0.0267, "step": 9960 }, { "grad_norm": 0.29869556427001953, "learning_rate": 9.932647597141893e-05, "loss": 0.0257, "step": 9970 }, { "grad_norm": 0.3413742482662201, "learning_rate": 9.932376847327239e-05, "loss": 0.029, "step": 9980 }, { "grad_norm": 0.22466927766799927, "learning_rate": 9.932105558115268e-05, "loss": 0.0258, "step": 9990 }, { "grad_norm": 0.23679664731025696, "learning_rate": 9.931833729535651e-05, "loss": 0.0256, "step": 10000 }, { "grad_norm": 0.2858388125896454, "learning_rate": 9.931561361618111e-05, "loss": 0.0247, "step": 10010 }, { "grad_norm": 0.3061981797218323, "learning_rate": 9.931288454392435e-05, "loss": 0.0269, "step": 10020 }, { "grad_norm": 0.3154492676258087, "learning_rate": 9.931015007888467e-05, "loss": 0.0245, "step": 10030 }, { "grad_norm": 0.29357174038887024, "learning_rate": 9.930741022136112e-05, "loss": 0.0251, "step": 10040 }, { "grad_norm": 0.288161039352417, "learning_rate": 9.930466497165333e-05, "loss": 0.0237, "step": 10050 }, { "grad_norm": 0.16961780190467834, "learning_rate": 9.93019143300615e-05, "loss": 0.0231, "step": 10060 }, { "grad_norm": 0.21344056725502014, "learning_rate": 9.929915829688644e-05, "loss": 0.024, "step": 10070 }, { "grad_norm": 0.2841798961162567, "learning_rate": 9.929639687242955e-05, "loss": 0.0277, "step": 10080 }, { "grad_norm": 0.17530691623687744, "learning_rate": 9.929363005699281e-05, "loss": 0.0279, "step": 10090 }, { "grad_norm": 0.31113019585609436, "learning_rate": 9.92908578508788e-05, "loss": 0.0298, "step": 10100 }, { "grad_norm": 0.3247736096382141, "learning_rate": 9.928808025439069e-05, "loss": 0.027, "step": 10110 }, { "grad_norm": 0.25910815596580505, "learning_rate": 9.928529726783223e-05, "loss": 0.0262, "step": 10120 }, { "grad_norm": 0.2599576711654663, "learning_rate": 9.928250889150774e-05, "loss": 0.0239, "step": 10130 }, { "grad_norm": 0.3370806574821472, "learning_rate": 9.92797151257222e-05, "loss": 0.0245, "step": 10140 }, { "grad_norm": 0.31346750259399414, "learning_rate": 9.927691597078108e-05, "loss": 0.023, "step": 10150 }, { "grad_norm": 0.299783319234848, "learning_rate": 9.927411142699053e-05, "loss": 0.0241, "step": 10160 }, { "grad_norm": 0.3311260938644409, "learning_rate": 9.927130149465725e-05, "loss": 0.0294, "step": 10170 }, { "grad_norm": 0.2693972587585449, "learning_rate": 9.92684861740885e-05, "loss": 0.0249, "step": 10180 }, { "grad_norm": 0.2473294734954834, "learning_rate": 9.926566546559217e-05, "loss": 0.0236, "step": 10190 }, { "grad_norm": 0.29265907406806946, "learning_rate": 9.926283936947673e-05, "loss": 0.0253, "step": 10200 }, { "grad_norm": 0.2530464828014374, "learning_rate": 9.926000788605126e-05, "loss": 0.0242, "step": 10210 }, { "grad_norm": 0.3658234179019928, "learning_rate": 9.92571710156254e-05, "loss": 0.0263, "step": 10220 }, { "grad_norm": 0.3707159161567688, "learning_rate": 9.925432875850936e-05, "loss": 0.0283, "step": 10230 }, { "grad_norm": 0.2969200015068054, "learning_rate": 9.925148111501396e-05, "loss": 0.0273, "step": 10240 }, { "grad_norm": 0.2485176920890808, "learning_rate": 9.924862808545066e-05, "loss": 0.0265, "step": 10250 }, { "grad_norm": 0.3842213749885559, "learning_rate": 9.924576967013141e-05, "loss": 0.0309, "step": 10260 }, { "grad_norm": 0.3349672555923462, "learning_rate": 9.924290586936887e-05, "loss": 0.027, "step": 10270 }, { "grad_norm": 0.28389352560043335, "learning_rate": 9.924003668347614e-05, "loss": 0.0271, "step": 10280 }, { "grad_norm": 0.3747609555721283, "learning_rate": 9.923716211276704e-05, "loss": 0.0268, "step": 10290 }, { "grad_norm": 0.265157550573349, "learning_rate": 9.923428215755594e-05, "loss": 0.0282, "step": 10300 }, { "grad_norm": 0.2304009199142456, "learning_rate": 9.923139681815775e-05, "loss": 0.0258, "step": 10310 }, { "grad_norm": 0.3500032424926758, "learning_rate": 9.922850609488801e-05, "loss": 0.0253, "step": 10320 }, { "grad_norm": 0.290428102016449, "learning_rate": 9.922560998806287e-05, "loss": 0.0288, "step": 10330 }, { "grad_norm": 0.266161173582077, "learning_rate": 9.922270849799905e-05, "loss": 0.0281, "step": 10340 }, { "grad_norm": 0.2518540322780609, "learning_rate": 9.92198016250138e-05, "loss": 0.0291, "step": 10350 }, { "grad_norm": 0.23007236421108246, "learning_rate": 9.921688936942506e-05, "loss": 0.0286, "step": 10360 }, { "grad_norm": 0.39083731174468994, "learning_rate": 9.921397173155129e-05, "loss": 0.0339, "step": 10370 }, { "grad_norm": 0.31015366315841675, "learning_rate": 9.921104871171157e-05, "loss": 0.0312, "step": 10380 }, { "grad_norm": 0.237137109041214, "learning_rate": 9.920812031022554e-05, "loss": 0.033, "step": 10390 }, { "grad_norm": 0.2908317446708679, "learning_rate": 9.920518652741348e-05, "loss": 0.0301, "step": 10400 }, { "grad_norm": 0.3058565557003021, "learning_rate": 9.920224736359618e-05, "loss": 0.0284, "step": 10410 }, { "grad_norm": 0.2627919018268585, "learning_rate": 9.91993028190951e-05, "loss": 0.0266, "step": 10420 }, { "grad_norm": 0.28024929761886597, "learning_rate": 9.919635289423222e-05, "loss": 0.0237, "step": 10430 }, { "grad_norm": 0.34907519817352295, "learning_rate": 9.919339758933015e-05, "loss": 0.0261, "step": 10440 }, { "grad_norm": 0.3016636371612549, "learning_rate": 9.919043690471209e-05, "loss": 0.0268, "step": 10450 }, { "grad_norm": 0.23380187153816223, "learning_rate": 9.91874708407018e-05, "loss": 0.0274, "step": 10460 }, { "grad_norm": 0.23098719120025635, "learning_rate": 9.918449939762367e-05, "loss": 0.0253, "step": 10470 }, { "grad_norm": 0.3730524778366089, "learning_rate": 9.91815225758026e-05, "loss": 0.028, "step": 10480 }, { "grad_norm": 0.31326353549957275, "learning_rate": 9.917854037556419e-05, "loss": 0.0338, "step": 10490 }, { "grad_norm": 0.2672078013420105, "learning_rate": 9.917555279723454e-05, "loss": 0.0343, "step": 10500 }, { "grad_norm": 0.24881400167942047, "learning_rate": 9.917255984114036e-05, "loss": 0.0267, "step": 10510 }, { "grad_norm": 0.26338282227516174, "learning_rate": 9.916956150760896e-05, "loss": 0.0252, "step": 10520 }, { "grad_norm": 0.31833338737487793, "learning_rate": 9.916655779696826e-05, "loss": 0.0272, "step": 10530 }, { "grad_norm": 0.25604888796806335, "learning_rate": 9.916354870954671e-05, "loss": 0.0276, "step": 10540 }, { "grad_norm": 0.26531046628952026, "learning_rate": 9.91605342456734e-05, "loss": 0.025, "step": 10550 }, { "grad_norm": 0.31714051961898804, "learning_rate": 9.915751440567795e-05, "loss": 0.0263, "step": 10560 }, { "grad_norm": 0.28909748792648315, "learning_rate": 9.915448918989066e-05, "loss": 0.0251, "step": 10570 }, { "grad_norm": 0.288359135389328, "learning_rate": 9.915145859864232e-05, "loss": 0.0241, "step": 10580 }, { "grad_norm": 0.25291121006011963, "learning_rate": 9.914842263226437e-05, "loss": 0.0257, "step": 10590 }, { "grad_norm": 0.3737794756889343, "learning_rate": 9.914538129108882e-05, "loss": 0.025, "step": 10600 }, { "grad_norm": 0.36805570125579834, "learning_rate": 9.914233457544825e-05, "loss": 0.0232, "step": 10610 }, { "grad_norm": 0.4486929178237915, "learning_rate": 9.913928248567586e-05, "loss": 0.0325, "step": 10620 }, { "grad_norm": 0.3591736853122711, "learning_rate": 9.913622502210542e-05, "loss": 0.0304, "step": 10630 }, { "grad_norm": 0.2686847448348999, "learning_rate": 9.913316218507128e-05, "loss": 0.0311, "step": 10640 }, { "grad_norm": 0.22276324033737183, "learning_rate": 9.91300939749084e-05, "loss": 0.029, "step": 10650 }, { "grad_norm": 0.24905911087989807, "learning_rate": 9.91270203919523e-05, "loss": 0.0264, "step": 10660 }, { "grad_norm": 0.23075315356254578, "learning_rate": 9.912394143653912e-05, "loss": 0.0265, "step": 10670 }, { "grad_norm": 0.2982424199581146, "learning_rate": 9.912085710900555e-05, "loss": 0.028, "step": 10680 }, { "grad_norm": 0.23922480642795563, "learning_rate": 9.911776740968892e-05, "loss": 0.0277, "step": 10690 }, { "grad_norm": 0.21613839268684387, "learning_rate": 9.911467233892709e-05, "loss": 0.0258, "step": 10700 }, { "grad_norm": 0.29025277495384216, "learning_rate": 9.911157189705853e-05, "loss": 0.025, "step": 10710 }, { "grad_norm": 0.325430303812027, "learning_rate": 9.910846608442229e-05, "loss": 0.0261, "step": 10720 }, { "grad_norm": 0.26771679520606995, "learning_rate": 9.910535490135805e-05, "loss": 0.0264, "step": 10730 }, { "grad_norm": 0.22466178238391876, "learning_rate": 9.910223834820603e-05, "loss": 0.0255, "step": 10740 }, { "grad_norm": 0.2358098328113556, "learning_rate": 9.909911642530703e-05, "loss": 0.026, "step": 10750 }, { "grad_norm": 0.27892348170280457, "learning_rate": 9.909598913300249e-05, "loss": 0.0261, "step": 10760 }, { "grad_norm": 0.26553574204444885, "learning_rate": 9.909285647163438e-05, "loss": 0.0271, "step": 10770 }, { "grad_norm": 0.2526853382587433, "learning_rate": 9.908971844154531e-05, "loss": 0.025, "step": 10780 }, { "grad_norm": 0.25232574343681335, "learning_rate": 9.908657504307843e-05, "loss": 0.0273, "step": 10790 }, { "grad_norm": 0.22909311950206757, "learning_rate": 9.908342627657751e-05, "loss": 0.0267, "step": 10800 }, { "grad_norm": 0.26633521914482117, "learning_rate": 9.908027214238689e-05, "loss": 0.0259, "step": 10810 }, { "grad_norm": 0.26717081665992737, "learning_rate": 9.90771126408515e-05, "loss": 0.0249, "step": 10820 }, { "grad_norm": 0.21462252736091614, "learning_rate": 9.907394777231685e-05, "loss": 0.0235, "step": 10830 }, { "grad_norm": 0.25441041588783264, "learning_rate": 9.907077753712905e-05, "loss": 0.0235, "step": 10840 }, { "grad_norm": 0.32010307908058167, "learning_rate": 9.906760193563482e-05, "loss": 0.0282, "step": 10850 }, { "grad_norm": 0.38363999128341675, "learning_rate": 9.906442096818139e-05, "loss": 0.0268, "step": 10860 }, { "grad_norm": 0.29144248366355896, "learning_rate": 9.906123463511665e-05, "loss": 0.0267, "step": 10870 }, { "grad_norm": 0.24677881598472595, "learning_rate": 9.905804293678907e-05, "loss": 0.0273, "step": 10880 }, { "grad_norm": 0.3109222650527954, "learning_rate": 9.905484587354766e-05, "loss": 0.0279, "step": 10890 }, { "grad_norm": 0.257295161485672, "learning_rate": 9.905164344574205e-05, "loss": 0.0248, "step": 10900 }, { "grad_norm": 0.2639507055282593, "learning_rate": 9.904843565372248e-05, "loss": 0.0249, "step": 10910 }, { "grad_norm": 0.3068907856941223, "learning_rate": 9.904522249783972e-05, "loss": 0.025, "step": 10920 }, { "grad_norm": 0.2707410454750061, "learning_rate": 9.904200397844517e-05, "loss": 0.0256, "step": 10930 }, { "grad_norm": 0.2632235586643219, "learning_rate": 9.903878009589078e-05, "loss": 0.0238, "step": 10940 }, { "grad_norm": 0.3169499337673187, "learning_rate": 9.903555085052915e-05, "loss": 0.0239, "step": 10950 }, { "grad_norm": 0.2757814824581146, "learning_rate": 9.903231624271338e-05, "loss": 0.0231, "step": 10960 }, { "grad_norm": 0.24850048124790192, "learning_rate": 9.902907627279724e-05, "loss": 0.0275, "step": 10970 }, { "grad_norm": 0.27288997173309326, "learning_rate": 9.902583094113504e-05, "loss": 0.0255, "step": 10980 }, { "grad_norm": 0.2894405126571655, "learning_rate": 9.902258024808168e-05, "loss": 0.0272, "step": 10990 }, { "grad_norm": 0.2427450567483902, "learning_rate": 9.901932419399264e-05, "loss": 0.0257, "step": 11000 }, { "grad_norm": 0.25057318806648254, "learning_rate": 9.9016062779224e-05, "loss": 0.0232, "step": 11010 }, { "grad_norm": 0.21203498542308807, "learning_rate": 9.901279600413242e-05, "loss": 0.025, "step": 11020 }, { "grad_norm": 0.2600727379322052, "learning_rate": 9.900952386907518e-05, "loss": 0.0231, "step": 11030 }, { "grad_norm": 0.2894074320793152, "learning_rate": 9.90062463744101e-05, "loss": 0.0248, "step": 11040 }, { "grad_norm": 0.29337266087532043, "learning_rate": 9.900296352049558e-05, "loss": 0.0242, "step": 11050 }, { "grad_norm": 0.24138453602790833, "learning_rate": 9.899967530769065e-05, "loss": 0.0274, "step": 11060 }, { "grad_norm": 0.23678302764892578, "learning_rate": 9.899638173635489e-05, "loss": 0.0241, "step": 11070 }, { "grad_norm": 0.30485719442367554, "learning_rate": 9.899308280684849e-05, "loss": 0.0263, "step": 11080 }, { "grad_norm": 0.2125970721244812, "learning_rate": 9.898977851953222e-05, "loss": 0.0217, "step": 11090 }, { "grad_norm": 0.33318737149238586, "learning_rate": 9.898646887476741e-05, "loss": 0.0235, "step": 11100 }, { "grad_norm": 0.3117749094963074, "learning_rate": 9.898315387291603e-05, "loss": 0.0301, "step": 11110 }, { "grad_norm": 0.2538500130176544, "learning_rate": 9.89798335143406e-05, "loss": 0.027, "step": 11120 }, { "grad_norm": 0.20158402621746063, "learning_rate": 9.897650779940419e-05, "loss": 0.0262, "step": 11130 }, { "grad_norm": 0.2074035257101059, "learning_rate": 9.897317672847054e-05, "loss": 0.0234, "step": 11140 }, { "grad_norm": 0.3484072983264923, "learning_rate": 9.89698403019039e-05, "loss": 0.0216, "step": 11150 }, { "grad_norm": 0.24690064787864685, "learning_rate": 9.896649852006917e-05, "loss": 0.0237, "step": 11160 }, { "grad_norm": 0.2727881968021393, "learning_rate": 9.896315138333177e-05, "loss": 0.0262, "step": 11170 }, { "grad_norm": 0.2724374234676361, "learning_rate": 9.895979889205774e-05, "loss": 0.0227, "step": 11180 }, { "grad_norm": 0.2222411334514618, "learning_rate": 9.895644104661372e-05, "loss": 0.0257, "step": 11190 }, { "grad_norm": 0.320443332195282, "learning_rate": 9.895307784736691e-05, "loss": 0.0246, "step": 11200 }, { "grad_norm": 0.2510613203048706, "learning_rate": 9.894970929468512e-05, "loss": 0.0245, "step": 11210 }, { "grad_norm": 0.22597724199295044, "learning_rate": 9.89463353889367e-05, "loss": 0.0227, "step": 11220 }, { "grad_norm": 0.25344914197921753, "learning_rate": 9.894295613049065e-05, "loss": 0.0293, "step": 11230 }, { "grad_norm": 0.24606068432331085, "learning_rate": 9.893957151971649e-05, "loss": 0.0222, "step": 11240 }, { "grad_norm": 0.2801874577999115, "learning_rate": 9.893618155698436e-05, "loss": 0.0242, "step": 11250 }, { "grad_norm": 0.23169277608394623, "learning_rate": 9.8932786242665e-05, "loss": 0.0208, "step": 11260 }, { "grad_norm": 0.2721833288669586, "learning_rate": 9.89293855771297e-05, "loss": 0.0216, "step": 11270 }, { "grad_norm": 0.23154689371585846, "learning_rate": 9.892597956075036e-05, "loss": 0.0218, "step": 11280 }, { "grad_norm": 0.26970693469047546, "learning_rate": 9.892256819389947e-05, "loss": 0.0245, "step": 11290 }, { "grad_norm": 0.35343995690345764, "learning_rate": 9.891915147695006e-05, "loss": 0.0252, "step": 11300 }, { "grad_norm": 0.29062628746032715, "learning_rate": 9.891572941027577e-05, "loss": 0.0263, "step": 11310 }, { "grad_norm": 0.2614213228225708, "learning_rate": 9.89123019942509e-05, "loss": 0.0232, "step": 11320 }, { "grad_norm": 0.2507038414478302, "learning_rate": 9.89088692292502e-05, "loss": 0.0234, "step": 11330 }, { "grad_norm": 0.27607566118240356, "learning_rate": 9.89054311156491e-05, "loss": 0.0252, "step": 11340 }, { "grad_norm": 0.3183014690876007, "learning_rate": 9.890198765382357e-05, "loss": 0.027, "step": 11350 }, { "grad_norm": 0.32612180709838867, "learning_rate": 9.889853884415021e-05, "loss": 0.0246, "step": 11360 }, { "grad_norm": 0.30379366874694824, "learning_rate": 9.889508468700614e-05, "loss": 0.0285, "step": 11370 }, { "grad_norm": 0.2940705716609955, "learning_rate": 9.889162518276915e-05, "loss": 0.0255, "step": 11380 }, { "grad_norm": 0.28352341055870056, "learning_rate": 9.888816033181752e-05, "loss": 0.0256, "step": 11390 }, { "grad_norm": 0.26330235600471497, "learning_rate": 9.888469013453018e-05, "loss": 0.0262, "step": 11400 }, { "grad_norm": 0.2894163429737091, "learning_rate": 9.888121459128663e-05, "loss": 0.027, "step": 11410 }, { "grad_norm": 0.25775644183158875, "learning_rate": 9.887773370246693e-05, "loss": 0.0246, "step": 11420 }, { "grad_norm": 0.2729383409023285, "learning_rate": 9.887424746845177e-05, "loss": 0.0262, "step": 11430 }, { "grad_norm": 0.2668342590332031, "learning_rate": 9.887075588962239e-05, "loss": 0.0261, "step": 11440 }, { "grad_norm": 0.31063392758369446, "learning_rate": 9.88672589663606e-05, "loss": 0.0294, "step": 11450 }, { "grad_norm": 0.29173895716667175, "learning_rate": 9.886375669904886e-05, "loss": 0.0283, "step": 11460 }, { "grad_norm": 0.20364466309547424, "learning_rate": 9.886024908807014e-05, "loss": 0.024, "step": 11470 }, { "grad_norm": 0.22273297607898712, "learning_rate": 9.885673613380806e-05, "loss": 0.023, "step": 11480 }, { "grad_norm": 0.2658102512359619, "learning_rate": 9.885321783664676e-05, "loss": 0.0231, "step": 11490 }, { "grad_norm": 0.2857418656349182, "learning_rate": 9.884969419697101e-05, "loss": 0.0237, "step": 11500 }, { "grad_norm": 0.25536054372787476, "learning_rate": 9.884616521516614e-05, "loss": 0.023, "step": 11510 }, { "grad_norm": 0.3006768524646759, "learning_rate": 9.88426308916181e-05, "loss": 0.024, "step": 11520 }, { "grad_norm": 0.2958314120769501, "learning_rate": 9.883909122671335e-05, "loss": 0.0278, "step": 11530 }, { "grad_norm": 0.2634080946445465, "learning_rate": 9.883554622083904e-05, "loss": 0.026, "step": 11540 }, { "grad_norm": 0.22239162027835846, "learning_rate": 9.88319958743828e-05, "loss": 0.0262, "step": 11550 }, { "grad_norm": 0.219709575176239, "learning_rate": 9.882844018773291e-05, "loss": 0.0243, "step": 11560 }, { "grad_norm": 0.27145373821258545, "learning_rate": 9.882487916127823e-05, "loss": 0.0227, "step": 11570 }, { "grad_norm": 0.3066153824329376, "learning_rate": 9.882131279540815e-05, "loss": 0.0262, "step": 11580 }, { "grad_norm": 0.29725420475006104, "learning_rate": 9.881774109051271e-05, "loss": 0.0261, "step": 11590 }, { "grad_norm": 0.20475895702838898, "learning_rate": 9.881416404698252e-05, "loss": 0.0226, "step": 11600 }, { "grad_norm": 0.27938857674598694, "learning_rate": 9.881058166520873e-05, "loss": 0.0251, "step": 11610 }, { "grad_norm": 0.2707037031650543, "learning_rate": 9.880699394558311e-05, "loss": 0.0279, "step": 11620 }, { "grad_norm": 0.2613072395324707, "learning_rate": 9.880340088849801e-05, "loss": 0.029, "step": 11630 }, { "grad_norm": 0.2675364315509796, "learning_rate": 9.879980249434637e-05, "loss": 0.0244, "step": 11640 }, { "grad_norm": 0.2868051826953888, "learning_rate": 9.879619876352168e-05, "loss": 0.0261, "step": 11650 }, { "grad_norm": 0.2461802065372467, "learning_rate": 9.879258969641809e-05, "loss": 0.0235, "step": 11660 }, { "grad_norm": 0.2799616754055023, "learning_rate": 9.878897529343023e-05, "loss": 0.0254, "step": 11670 }, { "grad_norm": 0.22588039934635162, "learning_rate": 9.878535555495338e-05, "loss": 0.0233, "step": 11680 }, { "grad_norm": 0.2236168384552002, "learning_rate": 9.87817304813834e-05, "loss": 0.0214, "step": 11690 }, { "grad_norm": 0.2842499911785126, "learning_rate": 9.877810007311671e-05, "loss": 0.0255, "step": 11700 }, { "grad_norm": 0.20329026877880096, "learning_rate": 9.877446433055035e-05, "loss": 0.0236, "step": 11710 }, { "grad_norm": 0.23506934940814972, "learning_rate": 9.877082325408191e-05, "loss": 0.0242, "step": 11720 }, { "grad_norm": 0.29283401370048523, "learning_rate": 9.876717684410954e-05, "loss": 0.0252, "step": 11730 }, { "grad_norm": 0.3710646331310272, "learning_rate": 9.876352510103204e-05, "loss": 0.024, "step": 11740 }, { "grad_norm": 0.23968513309955597, "learning_rate": 9.875986802524875e-05, "loss": 0.0256, "step": 11750 }, { "grad_norm": 0.3261071741580963, "learning_rate": 9.87562056171596e-05, "loss": 0.0259, "step": 11760 }, { "grad_norm": 0.2797462046146393, "learning_rate": 9.875253787716511e-05, "loss": 0.0244, "step": 11770 }, { "grad_norm": 0.39148303866386414, "learning_rate": 9.874886480566637e-05, "loss": 0.0245, "step": 11780 }, { "grad_norm": 0.3326127529144287, "learning_rate": 9.874518640306507e-05, "loss": 0.0284, "step": 11790 }, { "grad_norm": 0.38661372661590576, "learning_rate": 9.874150266976347e-05, "loss": 0.0282, "step": 11800 }, { "grad_norm": 0.2858535349369049, "learning_rate": 9.873781360616443e-05, "loss": 0.0276, "step": 11810 }, { "grad_norm": 0.26760348677635193, "learning_rate": 9.873411921267137e-05, "loss": 0.0275, "step": 11820 }, { "grad_norm": 0.257721871137619, "learning_rate": 9.873041948968829e-05, "loss": 0.0253, "step": 11830 }, { "grad_norm": 0.21471069753170013, "learning_rate": 9.872671443761981e-05, "loss": 0.0252, "step": 11840 }, { "grad_norm": 0.3426585793495178, "learning_rate": 9.872300405687109e-05, "loss": 0.0264, "step": 11850 }, { "grad_norm": 0.34311211109161377, "learning_rate": 9.871928834784792e-05, "loss": 0.0275, "step": 11860 }, { "grad_norm": 0.19577166438102722, "learning_rate": 9.871556731095661e-05, "loss": 0.029, "step": 11870 }, { "grad_norm": 0.24037015438079834, "learning_rate": 9.871184094660411e-05, "loss": 0.0263, "step": 11880 }, { "grad_norm": 0.20846059918403625, "learning_rate": 9.870810925519791e-05, "loss": 0.0242, "step": 11890 }, { "grad_norm": 0.24984151124954224, "learning_rate": 9.870437223714612e-05, "loss": 0.0243, "step": 11900 }, { "grad_norm": 0.18279190361499786, "learning_rate": 9.87006298928574e-05, "loss": 0.024, "step": 11910 }, { "grad_norm": 0.2844248414039612, "learning_rate": 9.869688222274103e-05, "loss": 0.0239, "step": 11920 }, { "grad_norm": 0.26622632145881653, "learning_rate": 9.869312922720681e-05, "loss": 0.0286, "step": 11930 }, { "grad_norm": 0.5897768139839172, "learning_rate": 9.868937090666521e-05, "loss": 0.0281, "step": 11940 }, { "grad_norm": 0.22957269847393036, "learning_rate": 9.86856072615272e-05, "loss": 0.0243, "step": 11950 }, { "grad_norm": 0.25426602363586426, "learning_rate": 9.868183829220438e-05, "loss": 0.0262, "step": 11960 }, { "grad_norm": 0.2068662792444229, "learning_rate": 9.867806399910893e-05, "loss": 0.0236, "step": 11970 }, { "grad_norm": 0.23722834885120392, "learning_rate": 9.867428438265356e-05, "loss": 0.0234, "step": 11980 }, { "grad_norm": 0.26232942938804626, "learning_rate": 9.867049944325165e-05, "loss": 0.0208, "step": 11990 }, { "grad_norm": 0.22845494747161865, "learning_rate": 9.86667091813171e-05, "loss": 0.0235, "step": 12000 }, { "grad_norm": 0.28112685680389404, "learning_rate": 9.866291359726438e-05, "loss": 0.0246, "step": 12010 }, { "grad_norm": 0.2979620695114136, "learning_rate": 9.865911269150861e-05, "loss": 0.0232, "step": 12020 }, { "grad_norm": 0.27232977747917175, "learning_rate": 9.865530646446544e-05, "loss": 0.0228, "step": 12030 }, { "grad_norm": 0.22101207077503204, "learning_rate": 9.86514949165511e-05, "loss": 0.022, "step": 12040 }, { "grad_norm": 0.20225952565670013, "learning_rate": 9.864767804818243e-05, "loss": 0.0226, "step": 12050 }, { "grad_norm": 0.21152575314044952, "learning_rate": 9.86438558597768e-05, "loss": 0.0228, "step": 12060 }, { "grad_norm": 0.2176004946231842, "learning_rate": 9.864002835175225e-05, "loss": 0.0203, "step": 12070 }, { "grad_norm": 0.26453983783721924, "learning_rate": 9.863619552452734e-05, "loss": 0.0215, "step": 12080 }, { "grad_norm": 0.2838873267173767, "learning_rate": 9.863235737852119e-05, "loss": 0.0213, "step": 12090 }, { "grad_norm": 0.2911619544029236, "learning_rate": 9.862851391415356e-05, "loss": 0.0229, "step": 12100 }, { "grad_norm": 0.23945991694927216, "learning_rate": 9.862466513184477e-05, "loss": 0.0228, "step": 12110 }, { "grad_norm": 0.2726638913154602, "learning_rate": 9.86208110320157e-05, "loss": 0.0245, "step": 12120 }, { "grad_norm": 0.2695308327674866, "learning_rate": 9.861695161508784e-05, "loss": 0.0246, "step": 12130 }, { "grad_norm": 0.2826770544052124, "learning_rate": 9.861308688148324e-05, "loss": 0.0252, "step": 12140 }, { "grad_norm": 0.2172670066356659, "learning_rate": 9.860921683162455e-05, "loss": 0.0232, "step": 12150 }, { "grad_norm": 0.2977447211742401, "learning_rate": 9.860534146593499e-05, "loss": 0.0234, "step": 12160 }, { "grad_norm": 0.3196479380130768, "learning_rate": 9.860146078483836e-05, "loss": 0.0227, "step": 12170 }, { "grad_norm": 0.2599658966064453, "learning_rate": 9.859757478875905e-05, "loss": 0.025, "step": 12180 }, { "grad_norm": 0.22958718240261078, "learning_rate": 9.859368347812204e-05, "loss": 0.0218, "step": 12190 }, { "grad_norm": 0.2611585557460785, "learning_rate": 9.858978685335285e-05, "loss": 0.0247, "step": 12200 }, { "grad_norm": 0.32810521125793457, "learning_rate": 9.858588491487763e-05, "loss": 0.0266, "step": 12210 }, { "grad_norm": 0.24123112857341766, "learning_rate": 9.858197766312308e-05, "loss": 0.0229, "step": 12220 }, { "grad_norm": 0.26225391030311584, "learning_rate": 9.857806509851649e-05, "loss": 0.0255, "step": 12230 }, { "grad_norm": 0.22078928351402283, "learning_rate": 9.857414722148574e-05, "loss": 0.0263, "step": 12240 }, { "grad_norm": 0.24710585176944733, "learning_rate": 9.857022403245928e-05, "loss": 0.0223, "step": 12250 }, { "grad_norm": 0.1711253970861435, "learning_rate": 9.856629553186615e-05, "loss": 0.0212, "step": 12260 }, { "grad_norm": 0.2453335076570511, "learning_rate": 9.856236172013595e-05, "loss": 0.0232, "step": 12270 }, { "grad_norm": 0.2455371767282486, "learning_rate": 9.85584225976989e-05, "loss": 0.021, "step": 12280 }, { "grad_norm": 0.2998409569263458, "learning_rate": 9.855447816498575e-05, "loss": 0.0204, "step": 12290 }, { "grad_norm": 0.28642207384109497, "learning_rate": 9.855052842242787e-05, "loss": 0.0236, "step": 12300 }, { "grad_norm": 0.26886245608329773, "learning_rate": 9.85465733704572e-05, "loss": 0.0269, "step": 12310 }, { "grad_norm": 0.1979418694972992, "learning_rate": 9.854261300950624e-05, "loss": 0.0208, "step": 12320 }, { "grad_norm": 0.3174593448638916, "learning_rate": 9.853864734000813e-05, "loss": 0.0215, "step": 12330 }, { "grad_norm": 0.22353577613830566, "learning_rate": 9.85346763623965e-05, "loss": 0.0232, "step": 12340 }, { "grad_norm": 0.29380160570144653, "learning_rate": 9.853070007710564e-05, "loss": 0.0252, "step": 12350 }, { "grad_norm": 0.29063376784324646, "learning_rate": 9.85267184845704e-05, "loss": 0.0228, "step": 12360 }, { "grad_norm": 0.3230121433734894, "learning_rate": 9.852273158522616e-05, "loss": 0.0219, "step": 12370 }, { "grad_norm": 0.2738851010799408, "learning_rate": 9.851873937950896e-05, "loss": 0.0211, "step": 12380 }, { "grad_norm": 0.24779453873634338, "learning_rate": 9.851474186785537e-05, "loss": 0.0217, "step": 12390 }, { "grad_norm": 0.34299972653388977, "learning_rate": 9.851073905070254e-05, "loss": 0.0225, "step": 12400 }, { "grad_norm": 0.22163991630077362, "learning_rate": 9.850673092848824e-05, "loss": 0.0217, "step": 12410 }, { "grad_norm": 0.24365286529064178, "learning_rate": 9.850271750165077e-05, "loss": 0.0213, "step": 12420 }, { "grad_norm": 0.3447042405605316, "learning_rate": 9.849869877062902e-05, "loss": 0.0267, "step": 12430 }, { "grad_norm": 0.297629177570343, "learning_rate": 9.849467473586252e-05, "loss": 0.0255, "step": 12440 }, { "grad_norm": 0.2589130997657776, "learning_rate": 9.849064539779127e-05, "loss": 0.0248, "step": 12450 }, { "grad_norm": 0.26408684253692627, "learning_rate": 9.848661075685594e-05, "loss": 0.023, "step": 12460 }, { "grad_norm": 0.24009177088737488, "learning_rate": 9.848257081349778e-05, "loss": 0.0274, "step": 12470 }, { "grad_norm": 0.2696816027164459, "learning_rate": 9.847852556815856e-05, "loss": 0.0246, "step": 12480 }, { "grad_norm": 0.32852017879486084, "learning_rate": 9.847447502128067e-05, "loss": 0.0283, "step": 12490 }, { "grad_norm": 0.38577884435653687, "learning_rate": 9.847041917330708e-05, "loss": 0.0216, "step": 12500 }, { "grad_norm": 0.26276832818984985, "learning_rate": 9.846635802468132e-05, "loss": 0.0266, "step": 12510 }, { "grad_norm": 0.2615661025047302, "learning_rate": 9.84622915758475e-05, "loss": 0.0234, "step": 12520 }, { "grad_norm": 0.2457466423511505, "learning_rate": 9.845821982725034e-05, "loss": 0.0253, "step": 12530 }, { "grad_norm": 0.40282583236694336, "learning_rate": 9.845414277933514e-05, "loss": 0.021, "step": 12540 }, { "grad_norm": 0.23552948236465454, "learning_rate": 9.845006043254771e-05, "loss": 0.0233, "step": 12550 }, { "grad_norm": 0.28661230206489563, "learning_rate": 9.844597278733451e-05, "loss": 0.0242, "step": 12560 }, { "grad_norm": 0.23361217975616455, "learning_rate": 9.844187984414259e-05, "loss": 0.0229, "step": 12570 }, { "grad_norm": 0.29092949628829956, "learning_rate": 9.84377816034195e-05, "loss": 0.026, "step": 12580 }, { "grad_norm": 0.2386806756258011, "learning_rate": 9.843367806561345e-05, "loss": 0.0219, "step": 12590 }, { "grad_norm": 0.2684301733970642, "learning_rate": 9.842956923117317e-05, "loss": 0.022, "step": 12600 }, { "grad_norm": 0.1827227920293808, "learning_rate": 9.842545510054802e-05, "loss": 0.019, "step": 12610 }, { "grad_norm": 0.1919942945241928, "learning_rate": 9.842133567418792e-05, "loss": 0.0197, "step": 12620 }, { "grad_norm": 0.20398063957691193, "learning_rate": 9.841721095254333e-05, "loss": 0.0228, "step": 12630 }, { "grad_norm": 0.23337127268314362, "learning_rate": 9.841308093606537e-05, "loss": 0.023, "step": 12640 }, { "grad_norm": 0.190652996301651, "learning_rate": 9.840894562520565e-05, "loss": 0.0201, "step": 12650 }, { "grad_norm": 0.18053416907787323, "learning_rate": 9.840480502041642e-05, "loss": 0.0217, "step": 12660 }, { "grad_norm": 0.22682969272136688, "learning_rate": 9.840065912215049e-05, "loss": 0.0211, "step": 12670 }, { "grad_norm": 0.26034510135650635, "learning_rate": 9.839650793086124e-05, "loss": 0.0201, "step": 12680 }, { "grad_norm": 0.3429766595363617, "learning_rate": 9.839235144700265e-05, "loss": 0.0238, "step": 12690 }, { "grad_norm": 0.2663881778717041, "learning_rate": 9.838818967102926e-05, "loss": 0.0209, "step": 12700 }, { "grad_norm": 0.41815125942230225, "learning_rate": 9.83840226033962e-05, "loss": 0.0229, "step": 12710 }, { "grad_norm": 0.3372351825237274, "learning_rate": 9.837985024455918e-05, "loss": 0.026, "step": 12720 }, { "grad_norm": 0.2736491560935974, "learning_rate": 9.837567259497447e-05, "loss": 0.0285, "step": 12730 }, { "grad_norm": 0.34039053320884705, "learning_rate": 9.837148965509894e-05, "loss": 0.0232, "step": 12740 }, { "grad_norm": 0.2735542058944702, "learning_rate": 9.836730142539001e-05, "loss": 0.0254, "step": 12750 }, { "grad_norm": 0.2561011016368866, "learning_rate": 9.836310790630574e-05, "loss": 0.0265, "step": 12760 }, { "grad_norm": 0.36333906650543213, "learning_rate": 9.83589090983047e-05, "loss": 0.0285, "step": 12770 }, { "grad_norm": 0.2150631546974182, "learning_rate": 9.835470500184605e-05, "loss": 0.0258, "step": 12780 }, { "grad_norm": 0.24524445831775665, "learning_rate": 9.835049561738957e-05, "loss": 0.0232, "step": 12790 }, { "grad_norm": 0.2469114512205124, "learning_rate": 9.834628094539558e-05, "loss": 0.0245, "step": 12800 }, { "grad_norm": 0.2697421610355377, "learning_rate": 9.834206098632499e-05, "loss": 0.0222, "step": 12810 }, { "grad_norm": 0.2124910205602646, "learning_rate": 9.833783574063931e-05, "loss": 0.0243, "step": 12820 }, { "grad_norm": 0.18200796842575073, "learning_rate": 9.833360520880058e-05, "loss": 0.0213, "step": 12830 }, { "grad_norm": 0.2152395099401474, "learning_rate": 9.832936939127144e-05, "loss": 0.0206, "step": 12840 }, { "grad_norm": 0.25390565395355225, "learning_rate": 9.832512828851515e-05, "loss": 0.0232, "step": 12850 }, { "grad_norm": 0.25002333521842957, "learning_rate": 9.832088190099546e-05, "loss": 0.0228, "step": 12860 }, { "grad_norm": 0.2372884899377823, "learning_rate": 9.831663022917679e-05, "loss": 0.0226, "step": 12870 }, { "grad_norm": 0.2611143887042999, "learning_rate": 9.831237327352407e-05, "loss": 0.0206, "step": 12880 }, { "grad_norm": 0.22561296820640564, "learning_rate": 9.830811103450286e-05, "loss": 0.026, "step": 12890 }, { "grad_norm": 0.2214168757200241, "learning_rate": 9.830384351257924e-05, "loss": 0.0234, "step": 12900 }, { "grad_norm": 0.23402069509029388, "learning_rate": 9.829957070821993e-05, "loss": 0.0213, "step": 12910 }, { "grad_norm": 0.370402455329895, "learning_rate": 9.829529262189218e-05, "loss": 0.0236, "step": 12920 }, { "grad_norm": 0.34785228967666626, "learning_rate": 9.829100925406385e-05, "loss": 0.0258, "step": 12930 }, { "grad_norm": 0.24078814685344696, "learning_rate": 9.828672060520333e-05, "loss": 0.0247, "step": 12940 }, { "grad_norm": 0.21807844936847687, "learning_rate": 9.828242667577966e-05, "loss": 0.0254, "step": 12950 }, { "grad_norm": 0.2493838518857956, "learning_rate": 9.82781274662624e-05, "loss": 0.0211, "step": 12960 }, { "grad_norm": 0.23426997661590576, "learning_rate": 9.82738229771217e-05, "loss": 0.0196, "step": 12970 }, { "grad_norm": 0.212551087141037, "learning_rate": 9.826951320882829e-05, "loss": 0.0218, "step": 12980 }, { "grad_norm": 0.2102968394756317, "learning_rate": 9.826519816185351e-05, "loss": 0.0213, "step": 12990 }, { "grad_norm": 0.27759242057800293, "learning_rate": 9.826087783666921e-05, "loss": 0.0242, "step": 13000 }, { "grad_norm": 0.22227619588375092, "learning_rate": 9.825655223374787e-05, "loss": 0.0206, "step": 13010 }, { "grad_norm": 0.24996055662631989, "learning_rate": 9.825222135356253e-05, "loss": 0.0211, "step": 13020 }, { "grad_norm": 0.26802513003349304, "learning_rate": 9.82478851965868e-05, "loss": 0.0247, "step": 13030 }, { "grad_norm": 0.2682815194129944, "learning_rate": 9.82435437632949e-05, "loss": 0.0233, "step": 13040 }, { "grad_norm": 0.3588883578777313, "learning_rate": 9.823919705416158e-05, "loss": 0.0266, "step": 13050 }, { "grad_norm": 0.26287615299224854, "learning_rate": 9.82348450696622e-05, "loss": 0.0248, "step": 13060 }, { "grad_norm": 0.2001814991235733, "learning_rate": 9.823048781027268e-05, "loss": 0.022, "step": 13070 }, { "grad_norm": 0.21999366581439972, "learning_rate": 9.822612527646953e-05, "loss": 0.0237, "step": 13080 }, { "grad_norm": 0.4011252522468567, "learning_rate": 9.822175746872984e-05, "loss": 0.0228, "step": 13090 }, { "grad_norm": 0.2714596092700958, "learning_rate": 9.821738438753123e-05, "loss": 0.0223, "step": 13100 }, { "grad_norm": 0.20697705447673798, "learning_rate": 9.821300603335196e-05, "loss": 0.0234, "step": 13110 }, { "grad_norm": 0.2043686956167221, "learning_rate": 9.820862240667085e-05, "loss": 0.0221, "step": 13120 }, { "grad_norm": 0.2878926694393158, "learning_rate": 9.820423350796726e-05, "loss": 0.0214, "step": 13130 }, { "grad_norm": 0.2818118631839752, "learning_rate": 9.819983933772118e-05, "loss": 0.0232, "step": 13140 }, { "grad_norm": 0.2166784703731537, "learning_rate": 9.819543989641314e-05, "loss": 0.0249, "step": 13150 }, { "grad_norm": 0.2994634211063385, "learning_rate": 9.819103518452423e-05, "loss": 0.0245, "step": 13160 }, { "grad_norm": 0.2636997699737549, "learning_rate": 9.818662520253618e-05, "loss": 0.0259, "step": 13170 }, { "grad_norm": 0.265764445066452, "learning_rate": 9.818220995093126e-05, "loss": 0.0246, "step": 13180 }, { "grad_norm": 0.2376195341348648, "learning_rate": 9.817778943019228e-05, "loss": 0.0253, "step": 13190 }, { "grad_norm": 0.2821318209171295, "learning_rate": 9.81733636408027e-05, "loss": 0.023, "step": 13200 }, { "grad_norm": 0.24353717267513275, "learning_rate": 9.816893258324649e-05, "loss": 0.0228, "step": 13210 }, { "grad_norm": 0.2743922472000122, "learning_rate": 9.816449625800823e-05, "loss": 0.0254, "step": 13220 }, { "grad_norm": 0.46588948369026184, "learning_rate": 9.816005466557308e-05, "loss": 0.025, "step": 13230 }, { "grad_norm": 0.21551154553890228, "learning_rate": 9.815560780642674e-05, "loss": 0.0228, "step": 13240 }, { "grad_norm": 0.18157777190208435, "learning_rate": 9.815115568105555e-05, "loss": 0.0239, "step": 13250 }, { "grad_norm": 0.27178382873535156, "learning_rate": 9.814669828994638e-05, "loss": 0.0248, "step": 13260 }, { "grad_norm": 0.4139138162136078, "learning_rate": 9.814223563358665e-05, "loss": 0.0231, "step": 13270 }, { "grad_norm": 0.196329727768898, "learning_rate": 9.813776771246443e-05, "loss": 0.024, "step": 13280 }, { "grad_norm": 0.2179359644651413, "learning_rate": 9.813329452706829e-05, "loss": 0.0266, "step": 13290 }, { "grad_norm": 0.27229246497154236, "learning_rate": 9.812881607788744e-05, "loss": 0.0266, "step": 13300 }, { "grad_norm": 0.2502240240573883, "learning_rate": 9.812433236541163e-05, "loss": 0.0246, "step": 13310 }, { "grad_norm": 0.37406209111213684, "learning_rate": 9.811984339013116e-05, "loss": 0.0339, "step": 13320 }, { "grad_norm": 0.3274857997894287, "learning_rate": 9.811534915253698e-05, "loss": 0.0305, "step": 13330 }, { "grad_norm": 0.4295433759689331, "learning_rate": 9.811084965312056e-05, "loss": 0.0283, "step": 13340 }, { "grad_norm": 0.24902690947055817, "learning_rate": 9.810634489237396e-05, "loss": 0.0272, "step": 13350 }, { "grad_norm": 0.24997840821743011, "learning_rate": 9.81018348707898e-05, "loss": 0.0283, "step": 13360 }, { "grad_norm": 0.27915894985198975, "learning_rate": 9.809731958886131e-05, "loss": 0.0232, "step": 13370 }, { "grad_norm": 0.23139624297618866, "learning_rate": 9.809279904708224e-05, "loss": 0.0237, "step": 13380 }, { "grad_norm": 0.2153775990009308, "learning_rate": 9.808827324594699e-05, "loss": 0.0234, "step": 13390 }, { "grad_norm": 0.21395272016525269, "learning_rate": 9.808374218595046e-05, "loss": 0.0236, "step": 13400 }, { "grad_norm": 0.23276302218437195, "learning_rate": 9.80792058675882e-05, "loss": 0.0223, "step": 13410 }, { "grad_norm": 0.2685870826244354, "learning_rate": 9.807466429135627e-05, "loss": 0.0238, "step": 13420 }, { "grad_norm": 0.33970391750335693, "learning_rate": 9.807011745775132e-05, "loss": 0.0265, "step": 13430 }, { "grad_norm": 0.29652148485183716, "learning_rate": 9.806556536727061e-05, "loss": 0.0255, "step": 13440 }, { "grad_norm": 0.20446482300758362, "learning_rate": 9.806100802041193e-05, "loss": 0.0226, "step": 13450 }, { "grad_norm": 0.22198674082756042, "learning_rate": 9.805644541767368e-05, "loss": 0.023, "step": 13460 }, { "grad_norm": 0.22394679486751556, "learning_rate": 9.805187755955478e-05, "loss": 0.0248, "step": 13470 }, { "grad_norm": 0.22908225655555725, "learning_rate": 9.804730444655483e-05, "loss": 0.0236, "step": 13480 }, { "grad_norm": 0.27377453446388245, "learning_rate": 9.804272607917388e-05, "loss": 0.0231, "step": 13490 }, { "grad_norm": 0.25116682052612305, "learning_rate": 9.803814245791265e-05, "loss": 0.0231, "step": 13500 }, { "grad_norm": 0.22819392383098602, "learning_rate": 9.803355358327239e-05, "loss": 0.0228, "step": 13510 }, { "grad_norm": 0.22834213078022003, "learning_rate": 9.802895945575492e-05, "loss": 0.0206, "step": 13520 }, { "grad_norm": 0.213328555226326, "learning_rate": 9.802436007586266e-05, "loss": 0.0208, "step": 13530 }, { "grad_norm": 0.17223487794399261, "learning_rate": 9.801975544409858e-05, "loss": 0.0215, "step": 13540 }, { "grad_norm": 0.4210655093193054, "learning_rate": 9.801514556096625e-05, "loss": 0.0254, "step": 13550 }, { "grad_norm": 0.31929174065589905, "learning_rate": 9.801053042696977e-05, "loss": 0.0238, "step": 13560 }, { "grad_norm": 0.22309266030788422, "learning_rate": 9.800591004261388e-05, "loss": 0.0244, "step": 13570 }, { "grad_norm": 0.23128663003444672, "learning_rate": 9.800128440840385e-05, "loss": 0.0224, "step": 13580 }, { "grad_norm": 0.25046828389167786, "learning_rate": 9.799665352484552e-05, "loss": 0.0269, "step": 13590 }, { "grad_norm": 0.24716921150684357, "learning_rate": 9.799201739244532e-05, "loss": 0.0213, "step": 13600 }, { "grad_norm": 0.2428029328584671, "learning_rate": 9.798737601171025e-05, "loss": 0.0201, "step": 13610 }, { "grad_norm": 0.23539328575134277, "learning_rate": 9.79827293831479e-05, "loss": 0.0229, "step": 13620 }, { "grad_norm": 0.19904498755931854, "learning_rate": 9.797807750726638e-05, "loss": 0.0217, "step": 13630 }, { "grad_norm": 0.24234186112880707, "learning_rate": 9.797342038457446e-05, "loss": 0.0216, "step": 13640 }, { "grad_norm": 0.25296714901924133, "learning_rate": 9.796875801558141e-05, "loss": 0.0247, "step": 13650 }, { "grad_norm": 0.19162079691886902, "learning_rate": 9.79640904007971e-05, "loss": 0.0191, "step": 13660 }, { "grad_norm": 0.19525066018104553, "learning_rate": 9.795941754073199e-05, "loss": 0.0243, "step": 13670 }, { "grad_norm": 0.2249360978603363, "learning_rate": 9.795473943589705e-05, "loss": 0.0255, "step": 13680 }, { "grad_norm": 0.22716574370861053, "learning_rate": 9.795005608680394e-05, "loss": 0.0234, "step": 13690 }, { "grad_norm": 0.17588305473327637, "learning_rate": 9.794536749396477e-05, "loss": 0.0227, "step": 13700 }, { "grad_norm": 0.16388043761253357, "learning_rate": 9.79406736578923e-05, "loss": 0.0223, "step": 13710 }, { "grad_norm": 0.1357273906469345, "learning_rate": 9.793597457909984e-05, "loss": 0.0204, "step": 13720 }, { "grad_norm": 0.19002987444400787, "learning_rate": 9.793127025810127e-05, "loss": 0.0248, "step": 13730 }, { "grad_norm": 0.24521638453006744, "learning_rate": 9.792656069541104e-05, "loss": 0.0221, "step": 13740 }, { "grad_norm": 0.23542535305023193, "learning_rate": 9.79218458915442e-05, "loss": 0.0282, "step": 13750 }, { "grad_norm": 0.2514778971672058, "learning_rate": 9.791712584701634e-05, "loss": 0.0251, "step": 13760 }, { "grad_norm": 0.30518728494644165, "learning_rate": 9.791240056234364e-05, "loss": 0.0226, "step": 13770 }, { "grad_norm": 0.25239986181259155, "learning_rate": 9.790767003804283e-05, "loss": 0.021, "step": 13780 }, { "grad_norm": 0.23398394882678986, "learning_rate": 9.790293427463126e-05, "loss": 0.023, "step": 13790 }, { "grad_norm": 0.23709328472614288, "learning_rate": 9.789819327262684e-05, "loss": 0.0208, "step": 13800 }, { "grad_norm": 0.2432878017425537, "learning_rate": 9.7893447032548e-05, "loss": 0.0201, "step": 13810 }, { "grad_norm": 0.19762757420539856, "learning_rate": 9.78886955549138e-05, "loss": 0.0216, "step": 13820 }, { "grad_norm": 0.24210871756076813, "learning_rate": 9.788393884024387e-05, "loss": 0.023, "step": 13830 }, { "grad_norm": 0.22561849653720856, "learning_rate": 9.787917688905836e-05, "loss": 0.024, "step": 13840 }, { "grad_norm": 0.2160043865442276, "learning_rate": 9.787440970187807e-05, "loss": 0.0212, "step": 13850 }, { "grad_norm": 0.2342377007007599, "learning_rate": 9.786963727922429e-05, "loss": 0.0203, "step": 13860 }, { "grad_norm": 0.20113380253314972, "learning_rate": 9.786485962161897e-05, "loss": 0.024, "step": 13870 }, { "grad_norm": 0.28562742471694946, "learning_rate": 9.786007672958455e-05, "loss": 0.0229, "step": 13880 }, { "grad_norm": 0.2826785147190094, "learning_rate": 9.78552886036441e-05, "loss": 0.0246, "step": 13890 }, { "grad_norm": 0.21678853034973145, "learning_rate": 9.785049524432124e-05, "loss": 0.0209, "step": 13900 }, { "grad_norm": 0.325340211391449, "learning_rate": 9.784569665214016e-05, "loss": 0.0273, "step": 13910 }, { "grad_norm": 0.27431970834732056, "learning_rate": 9.784089282762563e-05, "loss": 0.0253, "step": 13920 }, { "grad_norm": 0.2628304958343506, "learning_rate": 9.7836083771303e-05, "loss": 0.0223, "step": 13930 }, { "grad_norm": 0.2506124675273895, "learning_rate": 9.783126948369817e-05, "loss": 0.0225, "step": 13940 }, { "grad_norm": 0.2324661761522293, "learning_rate": 9.78264499653376e-05, "loss": 0.0253, "step": 13950 }, { "grad_norm": 0.21949255466461182, "learning_rate": 9.782162521674838e-05, "loss": 0.0264, "step": 13960 }, { "grad_norm": 0.276773065328598, "learning_rate": 9.781679523845812e-05, "loss": 0.026, "step": 13970 }, { "grad_norm": 0.2661118805408478, "learning_rate": 9.781196003099502e-05, "loss": 0.0232, "step": 13980 }, { "grad_norm": 0.28874674439430237, "learning_rate": 9.780711959488786e-05, "loss": 0.023, "step": 13990 }, { "grad_norm": 0.2975868880748749, "learning_rate": 9.780227393066599e-05, "loss": 0.0214, "step": 14000 }, { "grad_norm": 0.2608224153518677, "learning_rate": 9.77974230388593e-05, "loss": 0.0283, "step": 14010 }, { "grad_norm": 0.22275467216968536, "learning_rate": 9.779256691999829e-05, "loss": 0.0286, "step": 14020 }, { "grad_norm": 0.25721248984336853, "learning_rate": 9.778770557461403e-05, "loss": 0.0242, "step": 14030 }, { "grad_norm": 0.16992715001106262, "learning_rate": 9.778283900323812e-05, "loss": 0.022, "step": 14040 }, { "grad_norm": 0.29257291555404663, "learning_rate": 9.777796720640277e-05, "loss": 0.0254, "step": 14050 }, { "grad_norm": 0.18690627813339233, "learning_rate": 9.777309018464078e-05, "loss": 0.0205, "step": 14060 }, { "grad_norm": 0.22203201055526733, "learning_rate": 9.776820793848547e-05, "loss": 0.0175, "step": 14070 }, { "grad_norm": 0.22742100059986115, "learning_rate": 9.776332046847075e-05, "loss": 0.021, "step": 14080 }, { "grad_norm": 0.26566842198371887, "learning_rate": 9.775842777513111e-05, "loss": 0.0192, "step": 14090 }, { "grad_norm": 0.2668527066707611, "learning_rate": 9.775352985900163e-05, "loss": 0.022, "step": 14100 }, { "grad_norm": 0.26463252305984497, "learning_rate": 9.774862672061791e-05, "loss": 0.0212, "step": 14110 }, { "grad_norm": 0.2685362994670868, "learning_rate": 9.774371836051616e-05, "loss": 0.0211, "step": 14120 }, { "grad_norm": 0.2478809952735901, "learning_rate": 9.773880477923315e-05, "loss": 0.0205, "step": 14130 }, { "grad_norm": 0.20363715291023254, "learning_rate": 9.773388597730623e-05, "loss": 0.0203, "step": 14140 }, { "grad_norm": 0.3155117928981781, "learning_rate": 9.77289619552733e-05, "loss": 0.0199, "step": 14150 }, { "grad_norm": 0.23106376826763153, "learning_rate": 9.772403271367285e-05, "loss": 0.0221, "step": 14160 }, { "grad_norm": 0.22493816912174225, "learning_rate": 9.771909825304396e-05, "loss": 0.0217, "step": 14170 }, { "grad_norm": 0.20811451971530914, "learning_rate": 9.771415857392619e-05, "loss": 0.0196, "step": 14180 }, { "grad_norm": 0.17478181421756744, "learning_rate": 9.770921367685978e-05, "loss": 0.0223, "step": 14190 }, { "grad_norm": 0.23114880919456482, "learning_rate": 9.770426356238551e-05, "loss": 0.0194, "step": 14200 }, { "grad_norm": 0.2704788148403168, "learning_rate": 9.769930823104469e-05, "loss": 0.0217, "step": 14210 }, { "grad_norm": 0.22554504871368408, "learning_rate": 9.769434768337926e-05, "loss": 0.0215, "step": 14220 }, { "grad_norm": 0.19904112815856934, "learning_rate": 9.768938191993164e-05, "loss": 0.0219, "step": 14230 }, { "grad_norm": 0.22554564476013184, "learning_rate": 9.768441094124494e-05, "loss": 0.0237, "step": 14240 }, { "grad_norm": 0.2733002305030823, "learning_rate": 9.767943474786275e-05, "loss": 0.0231, "step": 14250 }, { "grad_norm": 0.2089792639017105, "learning_rate": 9.767445334032923e-05, "loss": 0.0212, "step": 14260 }, { "grad_norm": 0.2522347867488861, "learning_rate": 9.766946671918919e-05, "loss": 0.0216, "step": 14270 }, { "grad_norm": 0.2500695586204529, "learning_rate": 9.766447488498796e-05, "loss": 0.0212, "step": 14280 }, { "grad_norm": 0.2313302606344223, "learning_rate": 9.765947783827139e-05, "loss": 0.0228, "step": 14290 }, { "grad_norm": 0.29437386989593506, "learning_rate": 9.765447557958599e-05, "loss": 0.023, "step": 14300 }, { "grad_norm": 0.23964940011501312, "learning_rate": 9.764946810947879e-05, "loss": 0.0242, "step": 14310 }, { "grad_norm": 0.2250799536705017, "learning_rate": 9.764445542849738e-05, "loss": 0.0217, "step": 14320 }, { "grad_norm": 0.2724258601665497, "learning_rate": 9.763943753718998e-05, "loss": 0.019, "step": 14330 }, { "grad_norm": 0.2968827486038208, "learning_rate": 9.76344144361053e-05, "loss": 0.0259, "step": 14340 }, { "grad_norm": 0.20100516080856323, "learning_rate": 9.762938612579269e-05, "loss": 0.024, "step": 14350 }, { "grad_norm": 0.19435013830661774, "learning_rate": 9.762435260680202e-05, "loss": 0.0219, "step": 14360 }, { "grad_norm": 0.21531590819358826, "learning_rate": 9.761931387968373e-05, "loss": 0.0202, "step": 14370 }, { "grad_norm": 0.2458387315273285, "learning_rate": 9.76142699449889e-05, "loss": 0.0225, "step": 14380 }, { "grad_norm": 0.23839913308620453, "learning_rate": 9.760922080326908e-05, "loss": 0.0236, "step": 14390 }, { "grad_norm": 0.2337137758731842, "learning_rate": 9.760416645507644e-05, "loss": 0.0206, "step": 14400 }, { "grad_norm": 0.20672576129436493, "learning_rate": 9.759910690096375e-05, "loss": 0.0203, "step": 14410 }, { "grad_norm": 0.2109336256980896, "learning_rate": 9.759404214148429e-05, "loss": 0.0195, "step": 14420 }, { "grad_norm": 0.18519729375839233, "learning_rate": 9.758897217719191e-05, "loss": 0.0211, "step": 14430 }, { "grad_norm": 0.229094460606575, "learning_rate": 9.758389700864113e-05, "loss": 0.0213, "step": 14440 }, { "grad_norm": 0.27938422560691833, "learning_rate": 9.757881663638688e-05, "loss": 0.0227, "step": 14450 }, { "grad_norm": 0.2828797698020935, "learning_rate": 9.757373106098478e-05, "loss": 0.0228, "step": 14460 }, { "grad_norm": 0.27844321727752686, "learning_rate": 9.756864028299097e-05, "loss": 0.0199, "step": 14470 }, { "grad_norm": 0.24723385274410248, "learning_rate": 9.75635443029622e-05, "loss": 0.0202, "step": 14480 }, { "grad_norm": 0.31489166617393494, "learning_rate": 9.755844312145572e-05, "loss": 0.021, "step": 14490 }, { "grad_norm": 0.2015235424041748, "learning_rate": 9.755333673902941e-05, "loss": 0.0223, "step": 14500 }, { "grad_norm": 0.2168286293745041, "learning_rate": 9.75482251562417e-05, "loss": 0.0211, "step": 14510 }, { "grad_norm": 0.22615788877010345, "learning_rate": 9.754310837365155e-05, "loss": 0.0221, "step": 14520 }, { "grad_norm": 0.22407788038253784, "learning_rate": 9.753798639181856e-05, "loss": 0.0203, "step": 14530 }, { "grad_norm": 0.28892210125923157, "learning_rate": 9.753285921130286e-05, "loss": 0.0229, "step": 14540 }, { "grad_norm": 0.20592306554317474, "learning_rate": 9.752772683266512e-05, "loss": 0.0238, "step": 14550 }, { "grad_norm": 0.1521792709827423, "learning_rate": 9.752258925646665e-05, "loss": 0.0224, "step": 14560 }, { "grad_norm": 0.22855545580387115, "learning_rate": 9.751744648326926e-05, "loss": 0.0207, "step": 14570 }, { "grad_norm": 0.255960613489151, "learning_rate": 9.751229851363536e-05, "loss": 0.0207, "step": 14580 }, { "grad_norm": 0.2713901400566101, "learning_rate": 9.750714534812793e-05, "loss": 0.0221, "step": 14590 }, { "grad_norm": 0.3752094805240631, "learning_rate": 9.750198698731053e-05, "loss": 0.0201, "step": 14600 }, { "grad_norm": 0.20246663689613342, "learning_rate": 9.749682343174722e-05, "loss": 0.0201, "step": 14610 }, { "grad_norm": 0.24638578295707703, "learning_rate": 9.749165468200272e-05, "loss": 0.0189, "step": 14620 }, { "grad_norm": 0.17690208554267883, "learning_rate": 9.748648073864229e-05, "loss": 0.02, "step": 14630 }, { "grad_norm": 0.23108576238155365, "learning_rate": 9.748130160223168e-05, "loss": 0.0199, "step": 14640 }, { "grad_norm": 0.20542791485786438, "learning_rate": 9.747611727333734e-05, "loss": 0.0198, "step": 14650 }, { "grad_norm": 0.19372496008872986, "learning_rate": 9.74709277525262e-05, "loss": 0.0196, "step": 14660 }, { "grad_norm": 0.20744256675243378, "learning_rate": 9.746573304036576e-05, "loss": 0.0245, "step": 14670 }, { "grad_norm": 0.2822064161300659, "learning_rate": 9.746053313742412e-05, "loss": 0.0254, "step": 14680 }, { "grad_norm": 0.3780663311481476, "learning_rate": 9.745532804426994e-05, "loss": 0.0214, "step": 14690 }, { "grad_norm": 0.2399013489484787, "learning_rate": 9.745011776147242e-05, "loss": 0.0234, "step": 14700 }, { "grad_norm": 0.27139416337013245, "learning_rate": 9.744490228960138e-05, "loss": 0.0217, "step": 14710 }, { "grad_norm": 0.24013899266719818, "learning_rate": 9.743968162922713e-05, "loss": 0.0206, "step": 14720 }, { "grad_norm": 0.19307859241962433, "learning_rate": 9.743445578092064e-05, "loss": 0.0212, "step": 14730 }, { "grad_norm": 0.20252615213394165, "learning_rate": 9.742922474525338e-05, "loss": 0.0194, "step": 14740 }, { "grad_norm": 0.19381946325302124, "learning_rate": 9.742398852279741e-05, "loss": 0.0187, "step": 14750 }, { "grad_norm": 0.2928876280784607, "learning_rate": 9.741874711412535e-05, "loss": 0.0219, "step": 14760 }, { "grad_norm": 0.1957024484872818, "learning_rate": 9.741350051981042e-05, "loss": 0.02, "step": 14770 }, { "grad_norm": 0.22520558536052704, "learning_rate": 9.740824874042633e-05, "loss": 0.0199, "step": 14780 }, { "grad_norm": 0.25634780526161194, "learning_rate": 9.740299177654746e-05, "loss": 0.0184, "step": 14790 }, { "grad_norm": 0.17474842071533203, "learning_rate": 9.739772962874867e-05, "loss": 0.0203, "step": 14800 }, { "grad_norm": 0.23211880028247833, "learning_rate": 9.739246229760541e-05, "loss": 0.0195, "step": 14810 }, { "grad_norm": 0.2007247507572174, "learning_rate": 9.738718978369376e-05, "loss": 0.0185, "step": 14820 }, { "grad_norm": 0.2126888483762741, "learning_rate": 9.738191208759025e-05, "loss": 0.0208, "step": 14830 }, { "grad_norm": 0.2169387936592102, "learning_rate": 9.73766292098721e-05, "loss": 0.0206, "step": 14840 }, { "grad_norm": 0.1725919544696808, "learning_rate": 9.737134115111699e-05, "loss": 0.0177, "step": 14850 }, { "grad_norm": 0.17379656434059143, "learning_rate": 9.736604791190323e-05, "loss": 0.0197, "step": 14860 }, { "grad_norm": 0.23028334975242615, "learning_rate": 9.73607494928097e-05, "loss": 0.0195, "step": 14870 }, { "grad_norm": 0.2561046779155731, "learning_rate": 9.735544589441581e-05, "loss": 0.0215, "step": 14880 }, { "grad_norm": 0.2457779198884964, "learning_rate": 9.735013711730154e-05, "loss": 0.0212, "step": 14890 }, { "grad_norm": 0.24157774448394775, "learning_rate": 9.734482316204747e-05, "loss": 0.021, "step": 14900 }, { "grad_norm": 0.20148655772209167, "learning_rate": 9.733950402923473e-05, "loss": 0.0207, "step": 14910 }, { "grad_norm": 0.3076985776424408, "learning_rate": 9.7334179719445e-05, "loss": 0.0226, "step": 14920 }, { "grad_norm": 0.26085513830184937, "learning_rate": 9.732885023326053e-05, "loss": 0.0231, "step": 14930 }, { "grad_norm": 0.24818652868270874, "learning_rate": 9.732351557126418e-05, "loss": 0.0221, "step": 14940 }, { "grad_norm": 0.18103238940238953, "learning_rate": 9.731817573403929e-05, "loss": 0.0208, "step": 14950 }, { "grad_norm": 0.2112608104944229, "learning_rate": 9.731283072216985e-05, "loss": 0.0175, "step": 14960 }, { "grad_norm": 0.21441933512687683, "learning_rate": 9.730748053624039e-05, "loss": 0.0211, "step": 14970 }, { "grad_norm": 0.1691196858882904, "learning_rate": 9.730212517683598e-05, "loss": 0.0209, "step": 14980 }, { "grad_norm": 0.20634274184703827, "learning_rate": 9.729676464454228e-05, "loss": 0.0202, "step": 14990 }, { "grad_norm": 0.2685258090496063, "learning_rate": 9.72913989399455e-05, "loss": 0.0206, "step": 15000 }, { "grad_norm": 0.28702011704444885, "learning_rate": 9.728602806363242e-05, "loss": 0.0224, "step": 15010 }, { "grad_norm": 0.24726387858390808, "learning_rate": 9.728065201619043e-05, "loss": 0.0225, "step": 15020 }, { "grad_norm": 0.19746696949005127, "learning_rate": 9.727527079820742e-05, "loss": 0.0208, "step": 15030 }, { "grad_norm": 0.24102535843849182, "learning_rate": 9.726988441027186e-05, "loss": 0.0201, "step": 15040 }, { "grad_norm": 0.190145343542099, "learning_rate": 9.726449285297281e-05, "loss": 0.0215, "step": 15050 }, { "grad_norm": 0.21395111083984375, "learning_rate": 9.72590961268999e-05, "loss": 0.0208, "step": 15060 }, { "grad_norm": 0.27928635478019714, "learning_rate": 9.725369423264328e-05, "loss": 0.0228, "step": 15070 }, { "grad_norm": 0.2113877236843109, "learning_rate": 9.72482871707937e-05, "loss": 0.024, "step": 15080 }, { "grad_norm": 0.25476717948913574, "learning_rate": 9.724287494194247e-05, "loss": 0.0227, "step": 15090 }, { "grad_norm": 0.2455214262008667, "learning_rate": 9.723745754668147e-05, "loss": 0.0206, "step": 15100 }, { "grad_norm": 0.2963719964027405, "learning_rate": 9.723203498560313e-05, "loss": 0.0221, "step": 15110 }, { "grad_norm": 0.24788092076778412, "learning_rate": 9.722660725930046e-05, "loss": 0.0194, "step": 15120 }, { "grad_norm": 0.19946926832199097, "learning_rate": 9.722117436836702e-05, "loss": 0.0187, "step": 15130 }, { "grad_norm": 0.20256087183952332, "learning_rate": 9.721573631339696e-05, "loss": 0.0204, "step": 15140 }, { "grad_norm": 0.18806760013103485, "learning_rate": 9.721029309498494e-05, "loss": 0.0181, "step": 15150 }, { "grad_norm": 0.1851898580789566, "learning_rate": 9.720484471372627e-05, "loss": 0.0194, "step": 15160 }, { "grad_norm": 0.2366141825914383, "learning_rate": 9.719939117021673e-05, "loss": 0.0204, "step": 15170 }, { "grad_norm": 0.21909207105636597, "learning_rate": 9.719393246505275e-05, "loss": 0.0183, "step": 15180 }, { "grad_norm": 0.3690224289894104, "learning_rate": 9.718846859883128e-05, "loss": 0.0204, "step": 15190 }, { "grad_norm": 0.21037723124027252, "learning_rate": 9.718299957214982e-05, "loss": 0.0192, "step": 15200 }, { "grad_norm": 0.2995568513870239, "learning_rate": 9.717752538560646e-05, "loss": 0.023, "step": 15210 }, { "grad_norm": 0.2019372433423996, "learning_rate": 9.717204603979986e-05, "loss": 0.021, "step": 15220 }, { "grad_norm": 0.330737441778183, "learning_rate": 9.716656153532922e-05, "loss": 0.0211, "step": 15230 }, { "grad_norm": 0.2277233600616455, "learning_rate": 9.716107187279434e-05, "loss": 0.0217, "step": 15240 }, { "grad_norm": 0.2525290846824646, "learning_rate": 9.715557705279555e-05, "loss": 0.0198, "step": 15250 }, { "grad_norm": 0.2022363692522049, "learning_rate": 9.715007707593372e-05, "loss": 0.0215, "step": 15260 }, { "grad_norm": 0.25901323556900024, "learning_rate": 9.714457194281036e-05, "loss": 0.0232, "step": 15270 }, { "grad_norm": 0.2568235695362091, "learning_rate": 9.713906165402751e-05, "loss": 0.0215, "step": 15280 }, { "grad_norm": 0.21490631997585297, "learning_rate": 9.713354621018774e-05, "loss": 0.0215, "step": 15290 }, { "grad_norm": 0.19783972203731537, "learning_rate": 9.712802561189422e-05, "loss": 0.0224, "step": 15300 }, { "grad_norm": 0.23055291175842285, "learning_rate": 9.712249985975069e-05, "loss": 0.0201, "step": 15310 }, { "grad_norm": 0.20718377828598022, "learning_rate": 9.71169689543614e-05, "loss": 0.0208, "step": 15320 }, { "grad_norm": 0.259600430727005, "learning_rate": 9.711143289633123e-05, "loss": 0.0207, "step": 15330 }, { "grad_norm": 0.2503274083137512, "learning_rate": 9.710589168626561e-05, "loss": 0.02, "step": 15340 }, { "grad_norm": 0.2164829969406128, "learning_rate": 9.710034532477048e-05, "loss": 0.0198, "step": 15350 }, { "grad_norm": 0.23071232438087463, "learning_rate": 9.709479381245239e-05, "loss": 0.022, "step": 15360 }, { "grad_norm": 0.20073202252388, "learning_rate": 9.708923714991847e-05, "loss": 0.0207, "step": 15370 }, { "grad_norm": 0.29198384284973145, "learning_rate": 9.708367533777638e-05, "loss": 0.0205, "step": 15380 }, { "grad_norm": 0.18905988335609436, "learning_rate": 9.707810837663431e-05, "loss": 0.0222, "step": 15390 }, { "grad_norm": 0.16432884335517883, "learning_rate": 9.707253626710113e-05, "loss": 0.0219, "step": 15400 }, { "grad_norm": 0.23542965948581696, "learning_rate": 9.706695900978613e-05, "loss": 0.0184, "step": 15410 }, { "grad_norm": 0.2499498873949051, "learning_rate": 9.706137660529926e-05, "loss": 0.023, "step": 15420 }, { "grad_norm": 0.233810156583786, "learning_rate": 9.705578905425101e-05, "loss": 0.0208, "step": 15430 }, { "grad_norm": 0.25611117482185364, "learning_rate": 9.705019635725241e-05, "loss": 0.0249, "step": 15440 }, { "grad_norm": 0.2784821689128876, "learning_rate": 9.704459851491508e-05, "loss": 0.0206, "step": 15450 }, { "grad_norm": 0.2678648829460144, "learning_rate": 9.703899552785118e-05, "loss": 0.0209, "step": 15460 }, { "grad_norm": 0.1942865550518036, "learning_rate": 9.703338739667346e-05, "loss": 0.0188, "step": 15470 }, { "grad_norm": 0.19328482449054718, "learning_rate": 9.70277741219952e-05, "loss": 0.0205, "step": 15480 }, { "grad_norm": 0.22080621123313904, "learning_rate": 9.702215570443027e-05, "loss": 0.022, "step": 15490 }, { "grad_norm": 0.2417253702878952, "learning_rate": 9.701653214459309e-05, "loss": 0.0258, "step": 15500 }, { "grad_norm": 0.28885984420776367, "learning_rate": 9.701090344309865e-05, "loss": 0.0229, "step": 15510 }, { "grad_norm": 0.20248720049858093, "learning_rate": 9.700526960056247e-05, "loss": 0.0234, "step": 15520 }, { "grad_norm": 0.22687923908233643, "learning_rate": 9.699963061760068e-05, "loss": 0.0228, "step": 15530 }, { "grad_norm": 0.2283405214548111, "learning_rate": 9.699398649482997e-05, "loss": 0.0216, "step": 15540 }, { "grad_norm": 0.24686920642852783, "learning_rate": 9.698833723286753e-05, "loss": 0.0219, "step": 15550 }, { "grad_norm": 0.23625218868255615, "learning_rate": 9.698268283233118e-05, "loss": 0.0229, "step": 15560 }, { "grad_norm": 0.4861643612384796, "learning_rate": 9.697702329383929e-05, "loss": 0.0206, "step": 15570 }, { "grad_norm": 0.19332854449748993, "learning_rate": 9.697135861801074e-05, "loss": 0.0232, "step": 15580 }, { "grad_norm": 0.22179831564426422, "learning_rate": 9.696568880546505e-05, "loss": 0.0206, "step": 15590 }, { "grad_norm": 0.28695496916770935, "learning_rate": 9.696001385682223e-05, "loss": 0.0235, "step": 15600 }, { "grad_norm": 0.30458158254623413, "learning_rate": 9.695433377270291e-05, "loss": 0.0233, "step": 15610 }, { "grad_norm": 0.2214365303516388, "learning_rate": 9.694864855372824e-05, "loss": 0.024, "step": 15620 }, { "grad_norm": 0.5860043168067932, "learning_rate": 9.694295820051995e-05, "loss": 0.0201, "step": 15630 }, { "grad_norm": 0.18780511617660522, "learning_rate": 9.693726271370032e-05, "loss": 0.0229, "step": 15640 }, { "grad_norm": 0.23657380044460297, "learning_rate": 9.693156209389221e-05, "loss": 0.0215, "step": 15650 }, { "grad_norm": 0.273931086063385, "learning_rate": 9.692585634171905e-05, "loss": 0.0226, "step": 15660 }, { "grad_norm": 0.2521994113922119, "learning_rate": 9.692014545780476e-05, "loss": 0.0213, "step": 15670 }, { "grad_norm": 0.18997734785079956, "learning_rate": 9.691442944277393e-05, "loss": 0.0213, "step": 15680 }, { "grad_norm": 0.19197490811347961, "learning_rate": 9.690870829725162e-05, "loss": 0.0203, "step": 15690 }, { "grad_norm": 0.20479582250118256, "learning_rate": 9.69029820218635e-05, "loss": 0.021, "step": 15700 }, { "grad_norm": 0.19800084829330444, "learning_rate": 9.689725061723579e-05, "loss": 0.0216, "step": 15710 }, { "grad_norm": 0.2537260353565216, "learning_rate": 9.689151408399527e-05, "loss": 0.02, "step": 15720 }, { "grad_norm": 0.23731909692287445, "learning_rate": 9.688577242276924e-05, "loss": 0.0217, "step": 15730 }, { "grad_norm": 0.1985141932964325, "learning_rate": 9.688002563418566e-05, "loss": 0.0205, "step": 15740 }, { "grad_norm": 0.22227393090724945, "learning_rate": 9.687427371887293e-05, "loss": 0.0241, "step": 15750 }, { "grad_norm": 0.23430216312408447, "learning_rate": 9.686851667746012e-05, "loss": 0.0206, "step": 15760 }, { "grad_norm": 0.2192678600549698, "learning_rate": 9.686275451057677e-05, "loss": 0.0231, "step": 15770 }, { "grad_norm": 0.2831194996833801, "learning_rate": 9.685698721885308e-05, "loss": 0.0223, "step": 15780 }, { "grad_norm": 0.27679821848869324, "learning_rate": 9.68512148029197e-05, "loss": 0.0246, "step": 15790 }, { "grad_norm": 0.22922366857528687, "learning_rate": 9.684543726340791e-05, "loss": 0.0249, "step": 15800 }, { "grad_norm": 0.197278693318367, "learning_rate": 9.683965460094952e-05, "loss": 0.0207, "step": 15810 }, { "grad_norm": 0.1868634968996048, "learning_rate": 9.683386681617694e-05, "loss": 0.0207, "step": 15820 }, { "grad_norm": 0.21638499200344086, "learning_rate": 9.68280739097231e-05, "loss": 0.0221, "step": 15830 }, { "grad_norm": 0.2432202249765396, "learning_rate": 9.682227588222148e-05, "loss": 0.0203, "step": 15840 }, { "grad_norm": 0.24395494163036346, "learning_rate": 9.681647273430618e-05, "loss": 0.0221, "step": 15850 }, { "grad_norm": 0.2379877120256424, "learning_rate": 9.681066446661182e-05, "loss": 0.0224, "step": 15860 }, { "grad_norm": 0.27810344099998474, "learning_rate": 9.680485107977357e-05, "loss": 0.0215, "step": 15870 }, { "grad_norm": 0.25802183151245117, "learning_rate": 9.679903257442716e-05, "loss": 0.0204, "step": 15880 }, { "grad_norm": 0.17276686429977417, "learning_rate": 9.679320895120891e-05, "loss": 0.0209, "step": 15890 }, { "grad_norm": 0.19780103862285614, "learning_rate": 9.67873802107557e-05, "loss": 0.0206, "step": 15900 }, { "grad_norm": 0.23354214429855347, "learning_rate": 9.67815463537049e-05, "loss": 0.0183, "step": 15910 }, { "grad_norm": 0.255508691072464, "learning_rate": 9.677570738069457e-05, "loss": 0.0187, "step": 15920 }, { "grad_norm": 0.21826903522014618, "learning_rate": 9.676986329236318e-05, "loss": 0.0202, "step": 15930 }, { "grad_norm": 0.2075270563364029, "learning_rate": 9.676401408934987e-05, "loss": 0.0219, "step": 15940 }, { "grad_norm": 0.2291722297668457, "learning_rate": 9.675815977229428e-05, "loss": 0.0211, "step": 15950 }, { "grad_norm": 0.24321936070919037, "learning_rate": 9.675230034183664e-05, "loss": 0.018, "step": 15960 }, { "grad_norm": 0.22295410931110382, "learning_rate": 9.674643579861773e-05, "loss": 0.0206, "step": 15970 }, { "grad_norm": 0.19068121910095215, "learning_rate": 9.674056614327886e-05, "loss": 0.0204, "step": 15980 }, { "grad_norm": 0.21717481315135956, "learning_rate": 9.673469137646198e-05, "loss": 0.0202, "step": 15990 }, { "grad_norm": 0.17456015944480896, "learning_rate": 9.67288114988095e-05, "loss": 0.019, "step": 16000 }, { "grad_norm": 0.2642836570739746, "learning_rate": 9.672292651096447e-05, "loss": 0.0182, "step": 16010 }, { "grad_norm": 0.1977739781141281, "learning_rate": 9.671703641357042e-05, "loss": 0.0209, "step": 16020 }, { "grad_norm": 0.22819265723228455, "learning_rate": 9.67111412072715e-05, "loss": 0.0189, "step": 16030 }, { "grad_norm": 0.2790107727050781, "learning_rate": 9.670524089271242e-05, "loss": 0.0197, "step": 16040 }, { "grad_norm": 0.23032109439373016, "learning_rate": 9.669933547053842e-05, "loss": 0.0202, "step": 16050 }, { "grad_norm": 0.23996731638908386, "learning_rate": 9.669342494139531e-05, "loss": 0.021, "step": 16060 }, { "grad_norm": 0.22187693417072296, "learning_rate": 9.668750930592943e-05, "loss": 0.02, "step": 16070 }, { "grad_norm": 0.21456411480903625, "learning_rate": 9.668158856478775e-05, "loss": 0.0217, "step": 16080 }, { "grad_norm": 0.23308028280735016, "learning_rate": 9.66756627186177e-05, "loss": 0.0237, "step": 16090 }, { "grad_norm": 0.24341559410095215, "learning_rate": 9.666973176806737e-05, "loss": 0.0174, "step": 16100 }, { "grad_norm": 0.24625064432621002, "learning_rate": 9.666379571378534e-05, "loss": 0.0203, "step": 16110 }, { "grad_norm": 0.2307446300983429, "learning_rate": 9.665785455642076e-05, "loss": 0.021, "step": 16120 }, { "grad_norm": 0.17022331058979034, "learning_rate": 9.665190829662337e-05, "loss": 0.0201, "step": 16130 }, { "grad_norm": 0.25259023904800415, "learning_rate": 9.664595693504342e-05, "loss": 0.0198, "step": 16140 }, { "grad_norm": 0.19787131249904633, "learning_rate": 9.664000047233175e-05, "loss": 0.0186, "step": 16150 }, { "grad_norm": 0.2226700633764267, "learning_rate": 9.663403890913976e-05, "loss": 0.0187, "step": 16160 }, { "grad_norm": 0.24233348667621613, "learning_rate": 9.662807224611938e-05, "loss": 0.0203, "step": 16170 }, { "grad_norm": 0.20684215426445007, "learning_rate": 9.662210048392311e-05, "loss": 0.0203, "step": 16180 }, { "grad_norm": 0.22534430027008057, "learning_rate": 9.661612362320405e-05, "loss": 0.0195, "step": 16190 }, { "grad_norm": 0.2394012212753296, "learning_rate": 9.661014166461579e-05, "loss": 0.0227, "step": 16200 }, { "grad_norm": 0.3827306032180786, "learning_rate": 9.66041546088125e-05, "loss": 0.0209, "step": 16210 }, { "grad_norm": 0.27693161368370056, "learning_rate": 9.659816245644895e-05, "loss": 0.0224, "step": 16220 }, { "grad_norm": 0.23280221223831177, "learning_rate": 9.65921652081804e-05, "loss": 0.0199, "step": 16230 }, { "grad_norm": 0.2300579845905304, "learning_rate": 9.658616286466271e-05, "loss": 0.0183, "step": 16240 }, { "grad_norm": 0.2778612971305847, "learning_rate": 9.65801554265523e-05, "loss": 0.0202, "step": 16250 }, { "grad_norm": 0.1812124252319336, "learning_rate": 9.657414289450612e-05, "loss": 0.0211, "step": 16260 }, { "grad_norm": 0.17249128222465515, "learning_rate": 9.656812526918171e-05, "loss": 0.0221, "step": 16270 }, { "grad_norm": 0.21615447103977203, "learning_rate": 9.656210255123712e-05, "loss": 0.0212, "step": 16280 }, { "grad_norm": 0.19934727251529694, "learning_rate": 9.6556074741331e-05, "loss": 0.0248, "step": 16290 }, { "grad_norm": 0.2590319812297821, "learning_rate": 9.655004184012256e-05, "loss": 0.0233, "step": 16300 }, { "grad_norm": 0.2347499132156372, "learning_rate": 9.654400384827152e-05, "loss": 0.0189, "step": 16310 }, { "grad_norm": 0.20875591039657593, "learning_rate": 9.653796076643818e-05, "loss": 0.0186, "step": 16320 }, { "grad_norm": 0.17591890692710876, "learning_rate": 9.653191259528344e-05, "loss": 0.0199, "step": 16330 }, { "grad_norm": 0.15465781092643738, "learning_rate": 9.65258593354687e-05, "loss": 0.0199, "step": 16340 }, { "grad_norm": 0.18758659064769745, "learning_rate": 9.651980098765591e-05, "loss": 0.0176, "step": 16350 }, { "grad_norm": 0.2556658685207367, "learning_rate": 9.651373755250765e-05, "loss": 0.0213, "step": 16360 }, { "grad_norm": 0.19329044222831726, "learning_rate": 9.650766903068697e-05, "loss": 0.0184, "step": 16370 }, { "grad_norm": 0.25477707386016846, "learning_rate": 9.650159542285753e-05, "loss": 0.0187, "step": 16380 }, { "grad_norm": 0.22067856788635254, "learning_rate": 9.649551672968353e-05, "loss": 0.0211, "step": 16390 }, { "grad_norm": 0.23753546178340912, "learning_rate": 9.648943295182973e-05, "loss": 0.0198, "step": 16400 }, { "grad_norm": 0.24131402373313904, "learning_rate": 9.648334408996144e-05, "loss": 0.019, "step": 16410 }, { "grad_norm": 0.25543394684791565, "learning_rate": 9.647725014474452e-05, "loss": 0.0191, "step": 16420 }, { "grad_norm": 0.24886362254619598, "learning_rate": 9.64711511168454e-05, "loss": 0.0202, "step": 16430 }, { "grad_norm": 0.18455959856510162, "learning_rate": 9.646504700693108e-05, "loss": 0.0197, "step": 16440 }, { "grad_norm": 0.1789534091949463, "learning_rate": 9.645893781566907e-05, "loss": 0.021, "step": 16450 }, { "grad_norm": 0.30399906635284424, "learning_rate": 9.645282354372744e-05, "loss": 0.0238, "step": 16460 }, { "grad_norm": 0.2536199390888214, "learning_rate": 9.644670419177491e-05, "loss": 0.0233, "step": 16470 }, { "grad_norm": 0.2524515688419342, "learning_rate": 9.644057976048062e-05, "loss": 0.0259, "step": 16480 }, { "grad_norm": 0.270050972700119, "learning_rate": 9.643445025051435e-05, "loss": 0.0277, "step": 16490 }, { "grad_norm": 0.23729544878005981, "learning_rate": 9.642831566254641e-05, "loss": 0.0217, "step": 16500 }, { "grad_norm": 0.2064899504184723, "learning_rate": 9.642217599724769e-05, "loss": 0.0202, "step": 16510 }, { "grad_norm": 0.19685189425945282, "learning_rate": 9.64160312552896e-05, "loss": 0.0191, "step": 16520 }, { "grad_norm": 0.2717030346393585, "learning_rate": 9.64098814373441e-05, "loss": 0.021, "step": 16530 }, { "grad_norm": 0.162358820438385, "learning_rate": 9.640372654408374e-05, "loss": 0.018, "step": 16540 }, { "grad_norm": 0.21381990611553192, "learning_rate": 9.639756657618162e-05, "loss": 0.0212, "step": 16550 }, { "grad_norm": 0.19541823863983154, "learning_rate": 9.639140153431138e-05, "loss": 0.0184, "step": 16560 }, { "grad_norm": 0.2172422707080841, "learning_rate": 9.638523141914721e-05, "loss": 0.0183, "step": 16570 }, { "grad_norm": 0.2136610746383667, "learning_rate": 9.637905623136388e-05, "loss": 0.0211, "step": 16580 }, { "grad_norm": 0.24696744978427887, "learning_rate": 9.637287597163669e-05, "loss": 0.0228, "step": 16590 }, { "grad_norm": 0.2766513526439667, "learning_rate": 9.63666906406415e-05, "loss": 0.0222, "step": 16600 }, { "grad_norm": 0.3238585889339447, "learning_rate": 9.636050023905473e-05, "loss": 0.0195, "step": 16610 }, { "grad_norm": 0.1986362189054489, "learning_rate": 9.635430476755336e-05, "loss": 0.0196, "step": 16620 }, { "grad_norm": 0.17094337940216064, "learning_rate": 9.63481042268149e-05, "loss": 0.0191, "step": 16630 }, { "grad_norm": 0.26589131355285645, "learning_rate": 9.634189861751745e-05, "loss": 0.0238, "step": 16640 }, { "grad_norm": 0.22141382098197937, "learning_rate": 9.633568794033967e-05, "loss": 0.0211, "step": 16650 }, { "grad_norm": 0.251193106174469, "learning_rate": 9.63294721959607e-05, "loss": 0.021, "step": 16660 }, { "grad_norm": 0.2496751844882965, "learning_rate": 9.63232513850603e-05, "loss": 0.0211, "step": 16670 }, { "grad_norm": 0.2242109328508377, "learning_rate": 9.631702550831878e-05, "loss": 0.0206, "step": 16680 }, { "grad_norm": 0.26564934849739075, "learning_rate": 9.631079456641698e-05, "loss": 0.0257, "step": 16690 }, { "grad_norm": 0.184590682387352, "learning_rate": 9.630455856003632e-05, "loss": 0.021, "step": 16700 }, { "grad_norm": 0.19748735427856445, "learning_rate": 9.629831748985876e-05, "loss": 0.0196, "step": 16710 }, { "grad_norm": 0.20738063752651215, "learning_rate": 9.629207135656679e-05, "loss": 0.0182, "step": 16720 }, { "grad_norm": 0.18742023408412933, "learning_rate": 9.628582016084353e-05, "loss": 0.0197, "step": 16730 }, { "grad_norm": 0.17942194640636444, "learning_rate": 9.627956390337254e-05, "loss": 0.0173, "step": 16740 }, { "grad_norm": 0.23494401574134827, "learning_rate": 9.627330258483802e-05, "loss": 0.0226, "step": 16750 }, { "grad_norm": 0.22576503455638885, "learning_rate": 9.62670362059247e-05, "loss": 0.0221, "step": 16760 }, { "grad_norm": 0.2523803114891052, "learning_rate": 9.626076476731786e-05, "loss": 0.021, "step": 16770 }, { "grad_norm": 0.22027188539505005, "learning_rate": 9.625448826970336e-05, "loss": 0.0204, "step": 16780 }, { "grad_norm": 0.19644170999526978, "learning_rate": 9.624820671376755e-05, "loss": 0.021, "step": 16790 }, { "grad_norm": 0.2429218739271164, "learning_rate": 9.62419201001974e-05, "loss": 0.0238, "step": 16800 }, { "grad_norm": 0.32872098684310913, "learning_rate": 9.623562842968037e-05, "loss": 0.022, "step": 16810 }, { "grad_norm": 0.21454785764217377, "learning_rate": 9.622933170290454e-05, "loss": 0.0224, "step": 16820 }, { "grad_norm": 0.20090270042419434, "learning_rate": 9.622302992055849e-05, "loss": 0.0234, "step": 16830 }, { "grad_norm": 0.18320009112358093, "learning_rate": 9.62167230833314e-05, "loss": 0.02, "step": 16840 }, { "grad_norm": 0.2524435222148895, "learning_rate": 9.621041119191295e-05, "loss": 0.02, "step": 16850 }, { "grad_norm": 0.22792547941207886, "learning_rate": 9.620409424699342e-05, "loss": 0.0182, "step": 16860 }, { "grad_norm": 0.219133660197258, "learning_rate": 9.619777224926359e-05, "loss": 0.0199, "step": 16870 }, { "grad_norm": 0.1713394671678543, "learning_rate": 9.619144519941485e-05, "loss": 0.0193, "step": 16880 }, { "grad_norm": 0.19640329480171204, "learning_rate": 9.618511309813912e-05, "loss": 0.0183, "step": 16890 }, { "grad_norm": 0.2756732106208801, "learning_rate": 9.617877594612886e-05, "loss": 0.0217, "step": 16900 }, { "grad_norm": 0.214949369430542, "learning_rate": 9.617243374407707e-05, "loss": 0.018, "step": 16910 }, { "grad_norm": 0.2705856263637543, "learning_rate": 9.616608649267736e-05, "loss": 0.0193, "step": 16920 }, { "grad_norm": 0.22003982961177826, "learning_rate": 9.615973419262385e-05, "loss": 0.0166, "step": 16930 }, { "grad_norm": 0.22411014139652252, "learning_rate": 9.615337684461119e-05, "loss": 0.0175, "step": 16940 }, { "grad_norm": 0.2718609869480133, "learning_rate": 9.614701444933465e-05, "loss": 0.0192, "step": 16950 }, { "grad_norm": 0.21142642199993134, "learning_rate": 9.614064700748997e-05, "loss": 0.02, "step": 16960 }, { "grad_norm": 0.26185211539268494, "learning_rate": 9.613427451977352e-05, "loss": 0.0198, "step": 16970 }, { "grad_norm": 0.16603659093379974, "learning_rate": 9.612789698688216e-05, "loss": 0.0199, "step": 16980 }, { "grad_norm": 0.20625369250774384, "learning_rate": 9.612151440951334e-05, "loss": 0.0203, "step": 16990 }, { "grad_norm": 0.2505057752132416, "learning_rate": 9.611512678836506e-05, "loss": 0.0202, "step": 17000 }, { "grad_norm": 0.21913902461528778, "learning_rate": 9.610873412413584e-05, "loss": 0.0204, "step": 17010 }, { "grad_norm": 0.24435806274414062, "learning_rate": 9.610233641752476e-05, "loss": 0.0207, "step": 17020 }, { "grad_norm": 0.2118959277868271, "learning_rate": 9.609593366923151e-05, "loss": 0.0202, "step": 17030 }, { "grad_norm": 0.21876712143421173, "learning_rate": 9.608952587995625e-05, "loss": 0.0207, "step": 17040 }, { "grad_norm": 0.24819591641426086, "learning_rate": 9.608311305039972e-05, "loss": 0.0179, "step": 17050 }, { "grad_norm": 0.29566314816474915, "learning_rate": 9.607669518126326e-05, "loss": 0.0235, "step": 17060 }, { "grad_norm": 0.23932141065597534, "learning_rate": 9.607027227324866e-05, "loss": 0.0183, "step": 17070 }, { "grad_norm": 0.22111773490905762, "learning_rate": 9.606384432705837e-05, "loss": 0.0197, "step": 17080 }, { "grad_norm": 0.29612618684768677, "learning_rate": 9.60574113433953e-05, "loss": 0.0209, "step": 17090 }, { "grad_norm": 0.35042887926101685, "learning_rate": 9.6050973322963e-05, "loss": 0.0259, "step": 17100 }, { "grad_norm": 0.22942496836185455, "learning_rate": 9.604453026646547e-05, "loss": 0.0253, "step": 17110 }, { "grad_norm": 0.2932758331298828, "learning_rate": 9.603808217460735e-05, "loss": 0.0203, "step": 17120 }, { "grad_norm": 0.19292916357517242, "learning_rate": 9.603162904809377e-05, "loss": 0.0188, "step": 17130 }, { "grad_norm": 0.1962907612323761, "learning_rate": 9.602517088763045e-05, "loss": 0.02, "step": 17140 }, { "grad_norm": 0.21362580358982086, "learning_rate": 9.601870769392365e-05, "loss": 0.0195, "step": 17150 }, { "grad_norm": 0.19593797624111176, "learning_rate": 9.601223946768017e-05, "loss": 0.0187, "step": 17160 }, { "grad_norm": 0.20802552998065948, "learning_rate": 9.600576620960734e-05, "loss": 0.0183, "step": 17170 }, { "grad_norm": 0.2240247279405594, "learning_rate": 9.599928792041308e-05, "loss": 0.0196, "step": 17180 }, { "grad_norm": 0.14989353716373444, "learning_rate": 9.599280460080587e-05, "loss": 0.0178, "step": 17190 }, { "grad_norm": 0.17994150519371033, "learning_rate": 9.59863162514947e-05, "loss": 0.0186, "step": 17200 }, { "grad_norm": 0.2183752804994583, "learning_rate": 9.597982287318911e-05, "loss": 0.0179, "step": 17210 }, { "grad_norm": 0.27728089690208435, "learning_rate": 9.597332446659923e-05, "loss": 0.0212, "step": 17220 }, { "grad_norm": 0.24632467329502106, "learning_rate": 9.59668210324357e-05, "loss": 0.0194, "step": 17230 }, { "grad_norm": 0.22094063460826874, "learning_rate": 9.596031257140974e-05, "loss": 0.018, "step": 17240 }, { "grad_norm": 0.20690461993217468, "learning_rate": 9.59537990842331e-05, "loss": 0.0205, "step": 17250 }, { "grad_norm": 0.23724275827407837, "learning_rate": 9.594728057161806e-05, "loss": 0.0223, "step": 17260 }, { "grad_norm": 0.1511196345090866, "learning_rate": 9.594075703427752e-05, "loss": 0.0181, "step": 17270 }, { "grad_norm": 0.21676859259605408, "learning_rate": 9.593422847292486e-05, "loss": 0.0189, "step": 17280 }, { "grad_norm": 0.27231207489967346, "learning_rate": 9.592769488827402e-05, "loss": 0.0195, "step": 17290 }, { "grad_norm": 0.21202771365642548, "learning_rate": 9.592115628103952e-05, "loss": 0.0227, "step": 17300 }, { "grad_norm": 0.20033718645572662, "learning_rate": 9.591461265193643e-05, "loss": 0.0197, "step": 17310 }, { "grad_norm": 0.24514083564281464, "learning_rate": 9.590806400168032e-05, "loss": 0.02, "step": 17320 }, { "grad_norm": 0.2261505275964737, "learning_rate": 9.590151033098735e-05, "loss": 0.0175, "step": 17330 }, { "grad_norm": 0.23347195982933044, "learning_rate": 9.589495164057423e-05, "loss": 0.0155, "step": 17340 }, { "grad_norm": 0.19223015010356903, "learning_rate": 9.58883879311582e-05, "loss": 0.0165, "step": 17350 }, { "grad_norm": 0.25579485297203064, "learning_rate": 9.588181920345705e-05, "loss": 0.0163, "step": 17360 }, { "grad_norm": 0.22411948442459106, "learning_rate": 9.587524545818913e-05, "loss": 0.0186, "step": 17370 }, { "grad_norm": 0.30249306559562683, "learning_rate": 9.586866669607335e-05, "loss": 0.0191, "step": 17380 }, { "grad_norm": 0.239027202129364, "learning_rate": 9.586208291782915e-05, "loss": 0.0186, "step": 17390 }, { "grad_norm": 0.33048921823501587, "learning_rate": 9.58554941241765e-05, "loss": 0.0232, "step": 17400 }, { "grad_norm": 0.18572783470153809, "learning_rate": 9.584890031583596e-05, "loss": 0.0195, "step": 17410 }, { "grad_norm": 0.21883681416511536, "learning_rate": 9.584230149352861e-05, "loss": 0.0185, "step": 17420 }, { "grad_norm": 0.2121720165014267, "learning_rate": 9.58356976579761e-05, "loss": 0.0217, "step": 17430 }, { "grad_norm": 0.22139465808868408, "learning_rate": 9.58290888099006e-05, "loss": 0.0186, "step": 17440 }, { "grad_norm": 0.18778233230113983, "learning_rate": 9.582247495002486e-05, "loss": 0.0165, "step": 17450 }, { "grad_norm": 0.23170381784439087, "learning_rate": 9.581585607907214e-05, "loss": 0.0205, "step": 17460 }, { "grad_norm": 0.20222096145153046, "learning_rate": 9.580923219776628e-05, "loss": 0.0178, "step": 17470 }, { "grad_norm": 0.20426012575626373, "learning_rate": 9.580260330683167e-05, "loss": 0.0182, "step": 17480 }, { "grad_norm": 0.15214449167251587, "learning_rate": 9.579596940699322e-05, "loss": 0.0209, "step": 17490 }, { "grad_norm": 0.3681672215461731, "learning_rate": 9.578933049897643e-05, "loss": 0.0235, "step": 17500 }, { "grad_norm": 0.19352683424949646, "learning_rate": 9.578268658350728e-05, "loss": 0.0205, "step": 17510 }, { "grad_norm": 0.2560049295425415, "learning_rate": 9.577603766131235e-05, "loss": 0.0223, "step": 17520 }, { "grad_norm": 0.18930035829544067, "learning_rate": 9.576938373311878e-05, "loss": 0.0211, "step": 17530 }, { "grad_norm": 0.17716087400913239, "learning_rate": 9.576272479965421e-05, "loss": 0.02, "step": 17540 }, { "grad_norm": 0.16424833238124847, "learning_rate": 9.575606086164687e-05, "loss": 0.0172, "step": 17550 }, { "grad_norm": 0.22420711815357208, "learning_rate": 9.57493919198255e-05, "loss": 0.0184, "step": 17560 }, { "grad_norm": 0.21898788213729858, "learning_rate": 9.57427179749194e-05, "loss": 0.0201, "step": 17570 }, { "grad_norm": 0.22025446593761444, "learning_rate": 9.573603902765846e-05, "loss": 0.0187, "step": 17580 }, { "grad_norm": 0.19988276064395905, "learning_rate": 9.572935507877304e-05, "loss": 0.0192, "step": 17590 }, { "grad_norm": 0.16144073009490967, "learning_rate": 9.57226661289941e-05, "loss": 0.0182, "step": 17600 }, { "grad_norm": 0.2662699520587921, "learning_rate": 9.571597217905315e-05, "loss": 0.0165, "step": 17610 }, { "grad_norm": 0.2549988031387329, "learning_rate": 9.57092732296822e-05, "loss": 0.016, "step": 17620 }, { "grad_norm": 0.18918026983737946, "learning_rate": 9.570256928161385e-05, "loss": 0.0209, "step": 17630 }, { "grad_norm": 0.1893446147441864, "learning_rate": 9.569586033558126e-05, "loss": 0.0196, "step": 17640 }, { "grad_norm": 0.2391069084405899, "learning_rate": 9.568914639231807e-05, "loss": 0.0194, "step": 17650 }, { "grad_norm": 0.22150491178035736, "learning_rate": 9.568242745255852e-05, "loss": 0.0194, "step": 17660 }, { "grad_norm": 0.16077068448066711, "learning_rate": 9.567570351703739e-05, "loss": 0.0169, "step": 17670 }, { "grad_norm": 0.23557651042938232, "learning_rate": 9.566897458649001e-05, "loss": 0.0178, "step": 17680 }, { "grad_norm": 0.14722125232219696, "learning_rate": 9.566224066165221e-05, "loss": 0.0219, "step": 17690 }, { "grad_norm": 0.23239056766033173, "learning_rate": 9.565550174326043e-05, "loss": 0.016, "step": 17700 }, { "grad_norm": 0.21070794761180878, "learning_rate": 9.564875783205162e-05, "loss": 0.0166, "step": 17710 }, { "grad_norm": 0.14785446226596832, "learning_rate": 9.564200892876328e-05, "loss": 0.0154, "step": 17720 }, { "grad_norm": 0.18368950486183167, "learning_rate": 9.563525503413348e-05, "loss": 0.0158, "step": 17730 }, { "grad_norm": 0.21332208812236786, "learning_rate": 9.562849614890079e-05, "loss": 0.0153, "step": 17740 }, { "grad_norm": 0.18945570290088654, "learning_rate": 9.562173227380436e-05, "loss": 0.0162, "step": 17750 }, { "grad_norm": 0.18379883468151093, "learning_rate": 9.561496340958389e-05, "loss": 0.0151, "step": 17760 }, { "grad_norm": 0.2383102923631668, "learning_rate": 9.560818955697959e-05, "loss": 0.0158, "step": 17770 }, { "grad_norm": 0.20293830335140228, "learning_rate": 9.560141071673228e-05, "loss": 0.0167, "step": 17780 }, { "grad_norm": 0.23164251446723938, "learning_rate": 9.559462688958323e-05, "loss": 0.017, "step": 17790 }, { "grad_norm": 0.19271011650562286, "learning_rate": 9.558783807627434e-05, "loss": 0.0159, "step": 17800 }, { "grad_norm": 0.18442228436470032, "learning_rate": 9.558104427754801e-05, "loss": 0.0183, "step": 17810 }, { "grad_norm": 0.2395344078540802, "learning_rate": 9.557424549414722e-05, "loss": 0.0184, "step": 17820 }, { "grad_norm": 0.13902141153812408, "learning_rate": 9.556744172681546e-05, "loss": 0.0153, "step": 17830 }, { "grad_norm": 0.30367618799209595, "learning_rate": 9.556063297629677e-05, "loss": 0.0199, "step": 17840 }, { "grad_norm": 0.30957499146461487, "learning_rate": 9.555381924333578e-05, "loss": 0.018, "step": 17850 }, { "grad_norm": 0.17580969631671906, "learning_rate": 9.554700052867758e-05, "loss": 0.0176, "step": 17860 }, { "grad_norm": 0.30397462844848633, "learning_rate": 9.554017683306789e-05, "loss": 0.0193, "step": 17870 }, { "grad_norm": 0.17678430676460266, "learning_rate": 9.553334815725294e-05, "loss": 0.0175, "step": 17880 }, { "grad_norm": 0.1827615648508072, "learning_rate": 9.552651450197949e-05, "loss": 0.0174, "step": 17890 }, { "grad_norm": 0.1931280940771103, "learning_rate": 9.551967586799486e-05, "loss": 0.0175, "step": 17900 }, { "grad_norm": 0.15250609815120697, "learning_rate": 9.551283225604692e-05, "loss": 0.0164, "step": 17910 }, { "grad_norm": 0.19314433634281158, "learning_rate": 9.550598366688406e-05, "loss": 0.0181, "step": 17920 }, { "grad_norm": 0.1315947324037552, "learning_rate": 9.549913010125526e-05, "loss": 0.0166, "step": 17930 }, { "grad_norm": 0.19046826660633087, "learning_rate": 9.549227155990999e-05, "loss": 0.0169, "step": 17940 }, { "grad_norm": 0.2377397119998932, "learning_rate": 9.548540804359828e-05, "loss": 0.0163, "step": 17950 }, { "grad_norm": 0.1729600727558136, "learning_rate": 9.547853955307077e-05, "loss": 0.0191, "step": 17960 }, { "grad_norm": 0.18769097328186035, "learning_rate": 9.547166608907853e-05, "loss": 0.0196, "step": 17970 }, { "grad_norm": 0.20377226173877716, "learning_rate": 9.546478765237326e-05, "loss": 0.0162, "step": 17980 }, { "grad_norm": 0.23802639544010162, "learning_rate": 9.545790424370715e-05, "loss": 0.0174, "step": 17990 }, { "grad_norm": 0.22151705622673035, "learning_rate": 9.5451015863833e-05, "loss": 0.0184, "step": 18000 }, { "grad_norm": 0.19994814693927765, "learning_rate": 9.544412251350408e-05, "loss": 0.0197, "step": 18010 }, { "grad_norm": 0.27754905819892883, "learning_rate": 9.543722419347422e-05, "loss": 0.0184, "step": 18020 }, { "grad_norm": 0.23979242146015167, "learning_rate": 9.543032090449788e-05, "loss": 0.0186, "step": 18030 }, { "grad_norm": 0.20184850692749023, "learning_rate": 9.542341264732992e-05, "loss": 0.0186, "step": 18040 }, { "grad_norm": 0.19104304909706116, "learning_rate": 9.541649942272585e-05, "loss": 0.0192, "step": 18050 }, { "grad_norm": 0.2888917624950409, "learning_rate": 9.54095812314417e-05, "loss": 0.0176, "step": 18060 }, { "grad_norm": 0.28622496128082275, "learning_rate": 9.540265807423401e-05, "loss": 0.0178, "step": 18070 }, { "grad_norm": 0.2746996581554413, "learning_rate": 9.53957299518599e-05, "loss": 0.0196, "step": 18080 }, { "grad_norm": 0.20733922719955444, "learning_rate": 9.5388796865077e-05, "loss": 0.0217, "step": 18090 }, { "grad_norm": 0.15576250851154327, "learning_rate": 9.538185881464353e-05, "loss": 0.0164, "step": 18100 }, { "grad_norm": 0.16229552030563354, "learning_rate": 9.537491580131821e-05, "loss": 0.0175, "step": 18110 }, { "grad_norm": 0.18463191390037537, "learning_rate": 9.53679678258603e-05, "loss": 0.0161, "step": 18120 }, { "grad_norm": 0.22428540885448456, "learning_rate": 9.536101488902966e-05, "loss": 0.017, "step": 18130 }, { "grad_norm": 0.2331705242395401, "learning_rate": 9.535405699158663e-05, "loss": 0.022, "step": 18140 }, { "grad_norm": 0.17407146096229553, "learning_rate": 9.53470941342921e-05, "loss": 0.0172, "step": 18150 }, { "grad_norm": 0.17729178071022034, "learning_rate": 9.534012631790756e-05, "loss": 0.0172, "step": 18160 }, { "grad_norm": 0.16768188774585724, "learning_rate": 9.533315354319494e-05, "loss": 0.0209, "step": 18170 }, { "grad_norm": 0.1726006120443344, "learning_rate": 9.532617581091682e-05, "loss": 0.018, "step": 18180 }, { "grad_norm": 0.20292599499225616, "learning_rate": 9.531919312183629e-05, "loss": 0.0182, "step": 18190 }, { "grad_norm": 0.2891679108142853, "learning_rate": 9.531220547671688e-05, "loss": 0.0161, "step": 18200 }, { "grad_norm": 0.2275143563747406, "learning_rate": 9.530521287632285e-05, "loss": 0.0183, "step": 18210 }, { "grad_norm": 0.1836114078760147, "learning_rate": 9.529821532141884e-05, "loss": 0.0199, "step": 18220 }, { "grad_norm": 0.18553590774536133, "learning_rate": 9.52912128127701e-05, "loss": 0.0178, "step": 18230 }, { "grad_norm": 0.24814005196094513, "learning_rate": 9.528420535114244e-05, "loss": 0.0187, "step": 18240 }, { "grad_norm": 0.2305622100830078, "learning_rate": 9.527719293730215e-05, "loss": 0.0199, "step": 18250 }, { "grad_norm": 0.19716285169124603, "learning_rate": 9.527017557201611e-05, "loss": 0.0174, "step": 18260 }, { "grad_norm": 0.19420768320560455, "learning_rate": 9.526315325605176e-05, "loss": 0.0201, "step": 18270 }, { "grad_norm": 0.21262617409229279, "learning_rate": 9.525612599017699e-05, "loss": 0.0196, "step": 18280 }, { "grad_norm": 0.1635209172964096, "learning_rate": 9.524909377516033e-05, "loss": 0.0197, "step": 18290 }, { "grad_norm": 0.2814546525478363, "learning_rate": 9.524205661177081e-05, "loss": 0.0224, "step": 18300 }, { "grad_norm": 0.22925111651420593, "learning_rate": 9.523501450077801e-05, "loss": 0.0215, "step": 18310 }, { "grad_norm": 0.17809557914733887, "learning_rate": 9.522796744295202e-05, "loss": 0.0206, "step": 18320 }, { "grad_norm": 0.25463974475860596, "learning_rate": 9.522091543906352e-05, "loss": 0.0216, "step": 18330 }, { "grad_norm": 0.2306589037179947, "learning_rate": 9.521385848988369e-05, "loss": 0.0197, "step": 18340 }, { "grad_norm": 0.1671897917985916, "learning_rate": 9.520679659618428e-05, "loss": 0.0171, "step": 18350 }, { "grad_norm": 0.22007551789283752, "learning_rate": 9.519972975873754e-05, "loss": 0.0187, "step": 18360 }, { "grad_norm": 0.2906646728515625, "learning_rate": 9.519265797831633e-05, "loss": 0.0178, "step": 18370 }, { "grad_norm": 0.193692147731781, "learning_rate": 9.518558125569399e-05, "loss": 0.0223, "step": 18380 }, { "grad_norm": 0.22475601732730865, "learning_rate": 9.517849959164442e-05, "loss": 0.0183, "step": 18390 }, { "grad_norm": 0.19028013944625854, "learning_rate": 9.517141298694205e-05, "loss": 0.0167, "step": 18400 }, { "grad_norm": 0.23451295495033264, "learning_rate": 9.516432144236188e-05, "loss": 0.0206, "step": 18410 }, { "grad_norm": 0.1847449243068695, "learning_rate": 9.515722495867941e-05, "loss": 0.018, "step": 18420 }, { "grad_norm": 0.24425049126148224, "learning_rate": 9.515012353667072e-05, "loss": 0.0188, "step": 18430 }, { "grad_norm": 0.14163991808891296, "learning_rate": 9.51430171771124e-05, "loss": 0.0164, "step": 18440 }, { "grad_norm": 0.18584196269512177, "learning_rate": 9.513590588078159e-05, "loss": 0.0207, "step": 18450 }, { "grad_norm": 0.20658715069293976, "learning_rate": 9.512878964845597e-05, "loss": 0.0189, "step": 18460 }, { "grad_norm": 0.217320054769516, "learning_rate": 9.512166848091377e-05, "loss": 0.018, "step": 18470 }, { "grad_norm": 0.2847713530063629, "learning_rate": 9.511454237893376e-05, "loss": 0.0178, "step": 18480 }, { "grad_norm": 0.19421686232089996, "learning_rate": 9.51074113432952e-05, "loss": 0.0175, "step": 18490 }, { "grad_norm": 0.22236742079257965, "learning_rate": 9.510027537477797e-05, "loss": 0.0209, "step": 18500 }, { "grad_norm": 0.19965216517448425, "learning_rate": 9.509313447416242e-05, "loss": 0.0208, "step": 18510 }, { "grad_norm": 0.2712026536464691, "learning_rate": 9.508598864222949e-05, "loss": 0.0207, "step": 18520 }, { "grad_norm": 0.21910063922405243, "learning_rate": 9.507883787976062e-05, "loss": 0.0167, "step": 18530 }, { "grad_norm": 0.20230552554130554, "learning_rate": 9.507168218753781e-05, "loss": 0.0214, "step": 18540 }, { "grad_norm": 0.19903598725795746, "learning_rate": 9.506452156634362e-05, "loss": 0.0191, "step": 18550 }, { "grad_norm": 0.14654000103473663, "learning_rate": 9.505735601696109e-05, "loss": 0.0176, "step": 18560 }, { "grad_norm": 0.24047666788101196, "learning_rate": 9.505018554017385e-05, "loss": 0.0192, "step": 18570 }, { "grad_norm": 0.1679244041442871, "learning_rate": 9.504301013676604e-05, "loss": 0.0158, "step": 18580 }, { "grad_norm": 0.2029184103012085, "learning_rate": 9.503582980752238e-05, "loss": 0.0178, "step": 18590 }, { "grad_norm": 0.2345554679632187, "learning_rate": 9.502864455322809e-05, "loss": 0.0192, "step": 18600 }, { "grad_norm": 0.20634648203849792, "learning_rate": 9.502145437466891e-05, "loss": 0.0207, "step": 18610 }, { "grad_norm": 0.2119119018316269, "learning_rate": 9.501425927263116e-05, "loss": 0.0194, "step": 18620 }, { "grad_norm": 0.2116771936416626, "learning_rate": 9.500705924790172e-05, "loss": 0.0209, "step": 18630 }, { "grad_norm": 0.24501314759254456, "learning_rate": 9.499985430126794e-05, "loss": 0.018, "step": 18640 }, { "grad_norm": 0.19139371812343597, "learning_rate": 9.499264443351775e-05, "loss": 0.0171, "step": 18650 }, { "grad_norm": 0.21547216176986694, "learning_rate": 9.498542964543961e-05, "loss": 0.0191, "step": 18660 }, { "grad_norm": 0.19875456392765045, "learning_rate": 9.497820993782252e-05, "loss": 0.0177, "step": 18670 }, { "grad_norm": 0.234300434589386, "learning_rate": 9.497098531145601e-05, "loss": 0.0201, "step": 18680 }, { "grad_norm": 0.29516884684562683, "learning_rate": 9.496375576713017e-05, "loss": 0.0215, "step": 18690 }, { "grad_norm": 0.18457069993019104, "learning_rate": 9.49565213056356e-05, "loss": 0.0184, "step": 18700 }, { "grad_norm": 0.1447221040725708, "learning_rate": 9.494928192776342e-05, "loss": 0.0173, "step": 18710 }, { "grad_norm": 0.3421860933303833, "learning_rate": 9.494203763430538e-05, "loss": 0.018, "step": 18720 }, { "grad_norm": 0.22532832622528076, "learning_rate": 9.493478842605366e-05, "loss": 0.0171, "step": 18730 }, { "grad_norm": 0.20125004649162292, "learning_rate": 9.492753430380105e-05, "loss": 0.0197, "step": 18740 }, { "grad_norm": 0.21774035692214966, "learning_rate": 9.492027526834083e-05, "loss": 0.0185, "step": 18750 }, { "grad_norm": 0.2025020867586136, "learning_rate": 9.491301132046684e-05, "loss": 0.0176, "step": 18760 }, { "grad_norm": 0.21443401277065277, "learning_rate": 9.490574246097345e-05, "loss": 0.0176, "step": 18770 }, { "grad_norm": 0.264108806848526, "learning_rate": 9.48984686906556e-05, "loss": 0.018, "step": 18780 }, { "grad_norm": 0.15878139436244965, "learning_rate": 9.489119001030871e-05, "loss": 0.0186, "step": 18790 }, { "grad_norm": 0.14747664332389832, "learning_rate": 9.488390642072878e-05, "loss": 0.0178, "step": 18800 }, { "grad_norm": 0.1397467404603958, "learning_rate": 9.48766179227123e-05, "loss": 0.0185, "step": 18810 }, { "grad_norm": 0.21566489338874817, "learning_rate": 9.486932451705636e-05, "loss": 0.0174, "step": 18820 }, { "grad_norm": 0.2121308147907257, "learning_rate": 9.486202620455857e-05, "loss": 0.0181, "step": 18830 }, { "grad_norm": 0.23990708589553833, "learning_rate": 9.485472298601704e-05, "loss": 0.0187, "step": 18840 }, { "grad_norm": 0.20522013306617737, "learning_rate": 9.484741486223043e-05, "loss": 0.0182, "step": 18850 }, { "grad_norm": 0.18128934502601624, "learning_rate": 9.484010183399797e-05, "loss": 0.0183, "step": 18860 }, { "grad_norm": 0.18604658544063568, "learning_rate": 9.483278390211938e-05, "loss": 0.0161, "step": 18870 }, { "grad_norm": 0.18814049661159515, "learning_rate": 9.482546106739496e-05, "loss": 0.0172, "step": 18880 }, { "grad_norm": 0.1923237293958664, "learning_rate": 9.48181333306255e-05, "loss": 0.016, "step": 18890 }, { "grad_norm": 0.18210166692733765, "learning_rate": 9.481080069261237e-05, "loss": 0.0166, "step": 18900 }, { "grad_norm": 0.29728782176971436, "learning_rate": 9.480346315415745e-05, "loss": 0.0166, "step": 18910 }, { "grad_norm": 0.1859709471464157, "learning_rate": 9.479612071606314e-05, "loss": 0.0189, "step": 18920 }, { "grad_norm": 0.15074323117733002, "learning_rate": 9.478877337913244e-05, "loss": 0.0158, "step": 18930 }, { "grad_norm": 0.24126726388931274, "learning_rate": 9.478142114416881e-05, "loss": 0.0169, "step": 18940 }, { "grad_norm": 0.1970411241054535, "learning_rate": 9.47740640119763e-05, "loss": 0.0184, "step": 18950 }, { "grad_norm": 0.24666300415992737, "learning_rate": 9.476670198335947e-05, "loss": 0.015, "step": 18960 }, { "grad_norm": 0.18963056802749634, "learning_rate": 9.47593350591234e-05, "loss": 0.0179, "step": 18970 }, { "grad_norm": 0.17223267257213593, "learning_rate": 9.475196324007376e-05, "loss": 0.0164, "step": 18980 }, { "grad_norm": 0.1854415237903595, "learning_rate": 9.474458652701669e-05, "loss": 0.0175, "step": 18990 }, { "grad_norm": 0.20342692732810974, "learning_rate": 9.473720492075892e-05, "loss": 0.0158, "step": 19000 }, { "grad_norm": 0.18491366505622864, "learning_rate": 9.472981842210768e-05, "loss": 0.017, "step": 19010 }, { "grad_norm": 0.21477125585079193, "learning_rate": 9.472242703187074e-05, "loss": 0.0199, "step": 19020 }, { "grad_norm": 0.24174882471561432, "learning_rate": 9.471503075085643e-05, "loss": 0.0202, "step": 19030 }, { "grad_norm": 0.1663944274187088, "learning_rate": 9.470762957987359e-05, "loss": 0.0219, "step": 19040 }, { "grad_norm": 0.2383021116256714, "learning_rate": 9.470022351973158e-05, "loss": 0.0196, "step": 19050 }, { "grad_norm": 0.21939106285572052, "learning_rate": 9.469281257124034e-05, "loss": 0.0223, "step": 19060 }, { "grad_norm": 0.2509194016456604, "learning_rate": 9.46853967352103e-05, "loss": 0.0248, "step": 19070 }, { "grad_norm": 0.2217380404472351, "learning_rate": 9.467797601245246e-05, "loss": 0.0233, "step": 19080 }, { "grad_norm": 0.24457988142967224, "learning_rate": 9.467055040377834e-05, "loss": 0.0213, "step": 19090 }, { "grad_norm": 0.20973800122737885, "learning_rate": 9.466311990999999e-05, "loss": 0.0228, "step": 19100 }, { "grad_norm": 0.36577174067497253, "learning_rate": 9.465568453193e-05, "loss": 0.0236, "step": 19110 }, { "grad_norm": 0.2044205516576767, "learning_rate": 9.464824427038148e-05, "loss": 0.0204, "step": 19120 }, { "grad_norm": 0.20836369693279266, "learning_rate": 9.46407991261681e-05, "loss": 0.0189, "step": 19130 }, { "grad_norm": 0.2556374669075012, "learning_rate": 9.463334910010404e-05, "loss": 0.0185, "step": 19140 }, { "grad_norm": 0.2191242277622223, "learning_rate": 9.462589419300403e-05, "loss": 0.0199, "step": 19150 }, { "grad_norm": 0.1697823405265808, "learning_rate": 9.461843440568333e-05, "loss": 0.0174, "step": 19160 }, { "grad_norm": 0.17735818028450012, "learning_rate": 9.461096973895773e-05, "loss": 0.016, "step": 19170 }, { "grad_norm": 0.17599955201148987, "learning_rate": 9.460350019364355e-05, "loss": 0.0162, "step": 19180 }, { "grad_norm": 0.23274081945419312, "learning_rate": 9.459602577055764e-05, "loss": 0.018, "step": 19190 }, { "grad_norm": 0.20841842889785767, "learning_rate": 9.45885464705174e-05, "loss": 0.0187, "step": 19200 }, { "grad_norm": 0.18776901066303253, "learning_rate": 9.458106229434076e-05, "loss": 0.0181, "step": 19210 }, { "grad_norm": 0.1591925472021103, "learning_rate": 9.457357324284617e-05, "loss": 0.0169, "step": 19220 }, { "grad_norm": 0.25582557916641235, "learning_rate": 9.456607931685262e-05, "loss": 0.017, "step": 19230 }, { "grad_norm": 0.21360886096954346, "learning_rate": 9.455858051717965e-05, "loss": 0.0163, "step": 19240 }, { "grad_norm": 0.16530388593673706, "learning_rate": 9.45510768446473e-05, "loss": 0.0179, "step": 19250 }, { "grad_norm": 0.22231389582157135, "learning_rate": 9.454356830007618e-05, "loss": 0.0187, "step": 19260 }, { "grad_norm": 0.21394005417823792, "learning_rate": 9.45360548842874e-05, "loss": 0.0144, "step": 19270 }, { "grad_norm": 0.2957766056060791, "learning_rate": 9.452853659810261e-05, "loss": 0.0194, "step": 19280 }, { "grad_norm": 0.2191704362630844, "learning_rate": 9.452101344234401e-05, "loss": 0.0164, "step": 19290 }, { "grad_norm": 0.18462800979614258, "learning_rate": 9.451348541783431e-05, "loss": 0.0179, "step": 19300 }, { "grad_norm": 0.2569139301776886, "learning_rate": 9.450595252539678e-05, "loss": 0.017, "step": 19310 }, { "grad_norm": 0.26206162571907043, "learning_rate": 9.449841476585518e-05, "loss": 0.0185, "step": 19320 }, { "grad_norm": 0.25322502851486206, "learning_rate": 9.449087214003384e-05, "loss": 0.0169, "step": 19330 }, { "grad_norm": 0.212459534406662, "learning_rate": 9.448332464875765e-05, "loss": 0.0181, "step": 19340 }, { "grad_norm": 0.23163074254989624, "learning_rate": 9.447577229285192e-05, "loss": 0.0161, "step": 19350 }, { "grad_norm": 0.21578505635261536, "learning_rate": 9.446821507314261e-05, "loss": 0.0156, "step": 19360 }, { "grad_norm": 0.14656709134578705, "learning_rate": 9.446065299045617e-05, "loss": 0.0188, "step": 19370 }, { "grad_norm": 0.3014022409915924, "learning_rate": 9.445308604561955e-05, "loss": 0.0179, "step": 19380 }, { "grad_norm": 0.22198118269443512, "learning_rate": 9.444551423946028e-05, "loss": 0.02, "step": 19390 }, { "grad_norm": 0.16640998423099518, "learning_rate": 9.443793757280638e-05, "loss": 0.0169, "step": 19400 }, { "grad_norm": 0.21639180183410645, "learning_rate": 9.443035604648646e-05, "loss": 0.0186, "step": 19410 }, { "grad_norm": 0.17756833136081696, "learning_rate": 9.44227696613296e-05, "loss": 0.0166, "step": 19420 }, { "grad_norm": 0.18924081325531006, "learning_rate": 9.441517841816542e-05, "loss": 0.0148, "step": 19430 }, { "grad_norm": 0.1694682389497757, "learning_rate": 9.440758231782413e-05, "loss": 0.018, "step": 19440 }, { "grad_norm": 0.1937854290008545, "learning_rate": 9.439998136113639e-05, "loss": 0.0143, "step": 19450 }, { "grad_norm": 0.20806381106376648, "learning_rate": 9.439237554893344e-05, "loss": 0.0158, "step": 19460 }, { "grad_norm": 0.1647949516773224, "learning_rate": 9.438476488204705e-05, "loss": 0.0176, "step": 19470 }, { "grad_norm": 0.23938772082328796, "learning_rate": 9.43771493613095e-05, "loss": 0.0176, "step": 19480 }, { "grad_norm": 0.20173242688179016, "learning_rate": 9.436952898755362e-05, "loss": 0.0177, "step": 19490 }, { "grad_norm": 0.1616060733795166, "learning_rate": 9.436190376161276e-05, "loss": 0.0152, "step": 19500 }, { "grad_norm": 0.19843897223472595, "learning_rate": 9.43542736843208e-05, "loss": 0.0192, "step": 19510 }, { "grad_norm": 0.2200758010149002, "learning_rate": 9.434663875651216e-05, "loss": 0.0159, "step": 19520 }, { "grad_norm": 0.15101434290409088, "learning_rate": 9.433899897902177e-05, "loss": 0.0172, "step": 19530 }, { "grad_norm": 0.21893122792243958, "learning_rate": 9.433135435268511e-05, "loss": 0.015, "step": 19540 }, { "grad_norm": 0.289360910654068, "learning_rate": 9.432370487833819e-05, "loss": 0.0176, "step": 19550 }, { "grad_norm": 0.25483831763267517, "learning_rate": 9.431605055681756e-05, "loss": 0.0184, "step": 19560 }, { "grad_norm": 0.2395578920841217, "learning_rate": 9.430839138896026e-05, "loss": 0.0196, "step": 19570 }, { "grad_norm": 0.23401691019535065, "learning_rate": 9.43007273756039e-05, "loss": 0.018, "step": 19580 }, { "grad_norm": 0.2109822779893875, "learning_rate": 9.429305851758658e-05, "loss": 0.0162, "step": 19590 }, { "grad_norm": 0.18612521886825562, "learning_rate": 9.428538481574699e-05, "loss": 0.0176, "step": 19600 }, { "grad_norm": 0.18654687702655792, "learning_rate": 9.42777062709243e-05, "loss": 0.0176, "step": 19610 }, { "grad_norm": 0.21204081177711487, "learning_rate": 9.427002288395821e-05, "loss": 0.0169, "step": 19620 }, { "grad_norm": 0.18513274192810059, "learning_rate": 9.426233465568898e-05, "loss": 0.0163, "step": 19630 }, { "grad_norm": 0.1911458522081375, "learning_rate": 9.42546415869574e-05, "loss": 0.0154, "step": 19640 }, { "grad_norm": 0.17223003506660461, "learning_rate": 9.424694367860473e-05, "loss": 0.018, "step": 19650 }, { "grad_norm": 0.1988908052444458, "learning_rate": 9.423924093147284e-05, "loss": 0.0203, "step": 19660 }, { "grad_norm": 0.21098169684410095, "learning_rate": 9.423153334640407e-05, "loss": 0.0208, "step": 19670 }, { "grad_norm": 0.19095109403133392, "learning_rate": 9.42238209242413e-05, "loss": 0.0157, "step": 19680 }, { "grad_norm": 0.15375812351703644, "learning_rate": 9.421610366582798e-05, "loss": 0.0164, "step": 19690 }, { "grad_norm": 0.18067975342273712, "learning_rate": 9.420838157200803e-05, "loss": 0.0163, "step": 19700 }, { "grad_norm": 0.13773497939109802, "learning_rate": 9.420065464362594e-05, "loss": 0.0152, "step": 19710 }, { "grad_norm": 0.21106606721878052, "learning_rate": 9.419292288152673e-05, "loss": 0.0163, "step": 19720 }, { "grad_norm": 0.1613251268863678, "learning_rate": 9.418518628655588e-05, "loss": 0.0155, "step": 19730 }, { "grad_norm": 0.1691722869873047, "learning_rate": 9.417744485955951e-05, "loss": 0.0174, "step": 19740 }, { "grad_norm": 0.19017422199249268, "learning_rate": 9.41696986013842e-05, "loss": 0.0164, "step": 19750 }, { "grad_norm": 0.1983838975429535, "learning_rate": 9.416194751287705e-05, "loss": 0.0184, "step": 19760 }, { "grad_norm": 0.22746169567108154, "learning_rate": 9.415419159488572e-05, "loss": 0.0179, "step": 19770 }, { "grad_norm": 0.2173035889863968, "learning_rate": 9.414643084825837e-05, "loss": 0.0149, "step": 19780 }, { "grad_norm": 0.28867706656455994, "learning_rate": 9.413866527384372e-05, "loss": 0.0177, "step": 19790 }, { "grad_norm": 0.19239959120750427, "learning_rate": 9.4130894872491e-05, "loss": 0.0175, "step": 19800 }, { "grad_norm": 0.2677527666091919, "learning_rate": 9.412311964504998e-05, "loss": 0.0198, "step": 19810 }, { "grad_norm": 0.19167451560497284, "learning_rate": 9.411533959237091e-05, "loss": 0.0174, "step": 19820 }, { "grad_norm": 0.16400809586048126, "learning_rate": 9.410755471530464e-05, "loss": 0.0184, "step": 19830 }, { "grad_norm": 0.1723768711090088, "learning_rate": 9.40997650147025e-05, "loss": 0.0159, "step": 19840 }, { "grad_norm": 0.18770064413547516, "learning_rate": 9.409197049141637e-05, "loss": 0.0158, "step": 19850 }, { "grad_norm": 0.7500283122062683, "learning_rate": 9.408417114629863e-05, "loss": 0.0177, "step": 19860 }, { "grad_norm": 0.16307781636714935, "learning_rate": 9.40763669802022e-05, "loss": 0.0179, "step": 19870 }, { "grad_norm": 0.17782434821128845, "learning_rate": 9.406855799398056e-05, "loss": 0.0174, "step": 19880 }, { "grad_norm": 0.2787471115589142, "learning_rate": 9.406074418848767e-05, "loss": 0.0178, "step": 19890 }, { "grad_norm": 0.245181605219841, "learning_rate": 9.405292556457805e-05, "loss": 0.017, "step": 19900 }, { "grad_norm": 0.32919061183929443, "learning_rate": 9.404510212310671e-05, "loss": 0.0211, "step": 19910 }, { "grad_norm": 0.14686764776706696, "learning_rate": 9.403727386492924e-05, "loss": 0.0169, "step": 19920 }, { "grad_norm": 0.2256539762020111, "learning_rate": 9.40294407909017e-05, "loss": 0.0226, "step": 19930 }, { "grad_norm": 0.19732140004634857, "learning_rate": 9.40216029018807e-05, "loss": 0.018, "step": 19940 }, { "grad_norm": 0.18337401747703552, "learning_rate": 9.401376019872338e-05, "loss": 0.0172, "step": 19950 }, { "grad_norm": 0.18004414439201355, "learning_rate": 9.400591268228746e-05, "loss": 0.02, "step": 19960 }, { "grad_norm": 0.19885176420211792, "learning_rate": 9.399806035343106e-05, "loss": 0.015, "step": 19970 }, { "grad_norm": 0.146882101893425, "learning_rate": 9.399020321301294e-05, "loss": 0.0177, "step": 19980 }, { "grad_norm": 0.24955132603645325, "learning_rate": 9.398234126189234e-05, "loss": 0.017, "step": 19990 }, { "grad_norm": 0.15773814916610718, "learning_rate": 9.397447450092902e-05, "loss": 0.0181, "step": 20000 }, { "grad_norm": 0.20523935556411743, "learning_rate": 9.39666029309833e-05, "loss": 0.0175, "step": 20010 }, { "grad_norm": 0.22999702394008636, "learning_rate": 9.395872655291596e-05, "loss": 0.0178, "step": 20020 }, { "grad_norm": 0.2737760543823242, "learning_rate": 9.395084536758838e-05, "loss": 0.0226, "step": 20030 }, { "grad_norm": 0.33325493335723877, "learning_rate": 9.394295937586243e-05, "loss": 0.0212, "step": 20040 }, { "grad_norm": 0.2475341260433197, "learning_rate": 9.393506857860052e-05, "loss": 0.0212, "step": 20050 }, { "grad_norm": 0.1668471395969391, "learning_rate": 9.392717297666555e-05, "loss": 0.0201, "step": 20060 }, { "grad_norm": 0.2058703899383545, "learning_rate": 9.391927257092101e-05, "loss": 0.0207, "step": 20070 }, { "grad_norm": 0.22171859443187714, "learning_rate": 9.391136736223085e-05, "loss": 0.0191, "step": 20080 }, { "grad_norm": 0.17664875090122223, "learning_rate": 9.390345735145956e-05, "loss": 0.0192, "step": 20090 }, { "grad_norm": 0.270763635635376, "learning_rate": 9.389554253947219e-05, "loss": 0.0188, "step": 20100 }, { "grad_norm": 0.23265239596366882, "learning_rate": 9.388762292713428e-05, "loss": 0.0187, "step": 20110 }, { "grad_norm": 0.19637912511825562, "learning_rate": 9.38796985153119e-05, "loss": 0.0184, "step": 20120 }, { "grad_norm": 0.25750836730003357, "learning_rate": 9.387176930487169e-05, "loss": 0.0179, "step": 20130 }, { "grad_norm": 0.16442224383354187, "learning_rate": 9.386383529668072e-05, "loss": 0.0172, "step": 20140 }, { "grad_norm": 0.22046823799610138, "learning_rate": 9.385589649160669e-05, "loss": 0.0167, "step": 20150 }, { "grad_norm": 0.21673794090747833, "learning_rate": 9.384795289051775e-05, "loss": 0.0213, "step": 20160 }, { "grad_norm": 0.1934007704257965, "learning_rate": 9.384000449428261e-05, "loss": 0.0183, "step": 20170 }, { "grad_norm": 0.1889793574810028, "learning_rate": 9.383205130377048e-05, "loss": 0.0184, "step": 20180 }, { "grad_norm": 0.21070097386837006, "learning_rate": 9.382409331985114e-05, "loss": 0.0168, "step": 20190 }, { "grad_norm": 0.1881352663040161, "learning_rate": 9.381613054339482e-05, "loss": 0.016, "step": 20200 }, { "grad_norm": 0.22295355796813965, "learning_rate": 9.380816297527235e-05, "loss": 0.019, "step": 20210 }, { "grad_norm": 0.21643240749835968, "learning_rate": 9.380019061635506e-05, "loss": 0.0195, "step": 20220 }, { "grad_norm": 0.22671078145503998, "learning_rate": 9.379221346751474e-05, "loss": 0.0177, "step": 20230 }, { "grad_norm": 0.2145756632089615, "learning_rate": 9.378423152962382e-05, "loss": 0.0197, "step": 20240 }, { "grad_norm": 0.1862071454524994, "learning_rate": 9.377624480355517e-05, "loss": 0.0176, "step": 20250 }, { "grad_norm": 0.18893954157829285, "learning_rate": 9.376825329018219e-05, "loss": 0.019, "step": 20260 }, { "grad_norm": 0.22427870333194733, "learning_rate": 9.376025699037884e-05, "loss": 0.0169, "step": 20270 }, { "grad_norm": 0.20984524488449097, "learning_rate": 9.37522559050196e-05, "loss": 0.0169, "step": 20280 }, { "grad_norm": 0.27249962091445923, "learning_rate": 9.37442500349794e-05, "loss": 0.019, "step": 20290 }, { "grad_norm": 0.20999252796173096, "learning_rate": 9.373623938113381e-05, "loss": 0.0214, "step": 20300 }, { "grad_norm": 0.18340817093849182, "learning_rate": 9.372822394435883e-05, "loss": 0.0189, "step": 20310 }, { "grad_norm": 0.24918103218078613, "learning_rate": 9.372020372553102e-05, "loss": 0.016, "step": 20320 }, { "grad_norm": 0.16648739576339722, "learning_rate": 9.371217872552746e-05, "loss": 0.0181, "step": 20330 }, { "grad_norm": 0.18130090832710266, "learning_rate": 9.370414894522576e-05, "loss": 0.0169, "step": 20340 }, { "grad_norm": 0.23294763267040253, "learning_rate": 9.369611438550406e-05, "loss": 0.0151, "step": 20350 }, { "grad_norm": 0.21213160455226898, "learning_rate": 9.368807504724095e-05, "loss": 0.0188, "step": 20360 }, { "grad_norm": 0.1519702672958374, "learning_rate": 9.368003093131565e-05, "loss": 0.0177, "step": 20370 }, { "grad_norm": 0.224260613322258, "learning_rate": 9.367198203860785e-05, "loss": 0.0193, "step": 20380 }, { "grad_norm": 0.18850955367088318, "learning_rate": 9.366392836999774e-05, "loss": 0.0164, "step": 20390 }, { "grad_norm": 0.195068821310997, "learning_rate": 9.365586992636607e-05, "loss": 0.0203, "step": 20400 }, { "grad_norm": 0.2162700593471527, "learning_rate": 9.364780670859412e-05, "loss": 0.0183, "step": 20410 }, { "grad_norm": 0.1789066344499588, "learning_rate": 9.363973871756364e-05, "loss": 0.0177, "step": 20420 }, { "grad_norm": 0.21950770914554596, "learning_rate": 9.363166595415696e-05, "loss": 0.016, "step": 20430 }, { "grad_norm": 0.17951136827468872, "learning_rate": 9.362358841925686e-05, "loss": 0.0187, "step": 20440 }, { "grad_norm": 0.1811503767967224, "learning_rate": 9.361550611374674e-05, "loss": 0.0163, "step": 20450 }, { "grad_norm": 0.16936619579792023, "learning_rate": 9.360741903851043e-05, "loss": 0.0162, "step": 20460 }, { "grad_norm": 0.2238111048936844, "learning_rate": 9.359932719443236e-05, "loss": 0.0168, "step": 20470 }, { "grad_norm": 0.17534194886684418, "learning_rate": 9.35912305823974e-05, "loss": 0.0175, "step": 20480 }, { "grad_norm": 0.2648470997810364, "learning_rate": 9.358312920329101e-05, "loss": 0.0161, "step": 20490 }, { "grad_norm": 0.19706007838249207, "learning_rate": 9.357502305799914e-05, "loss": 0.0148, "step": 20500 }, { "grad_norm": 0.23800833523273468, "learning_rate": 9.356691214740824e-05, "loss": 0.0157, "step": 20510 }, { "grad_norm": 0.16240371763706207, "learning_rate": 9.355879647240535e-05, "loss": 0.0146, "step": 20520 }, { "grad_norm": 0.44909241795539856, "learning_rate": 9.355067603387798e-05, "loss": 0.0191, "step": 20530 }, { "grad_norm": 0.18572920560836792, "learning_rate": 9.354255083271412e-05, "loss": 0.0173, "step": 20540 }, { "grad_norm": 0.21569214761257172, "learning_rate": 9.353442086980239e-05, "loss": 0.0153, "step": 20550 }, { "grad_norm": 0.20006994903087616, "learning_rate": 9.352628614603185e-05, "loss": 0.0172, "step": 20560 }, { "grad_norm": 0.21641242504119873, "learning_rate": 9.351814666229209e-05, "loss": 0.017, "step": 20570 }, { "grad_norm": 0.191386878490448, "learning_rate": 9.351000241947324e-05, "loss": 0.0196, "step": 20580 }, { "grad_norm": 0.3473213315010071, "learning_rate": 9.350185341846594e-05, "loss": 0.0197, "step": 20590 }, { "grad_norm": 0.2059423178434372, "learning_rate": 9.349369966016134e-05, "loss": 0.0175, "step": 20600 }, { "grad_norm": 0.14682500064373016, "learning_rate": 9.348554114545117e-05, "loss": 0.0168, "step": 20610 }, { "grad_norm": 0.20478908717632294, "learning_rate": 9.347737787522758e-05, "loss": 0.0156, "step": 20620 }, { "grad_norm": 0.25563064217567444, "learning_rate": 9.346920985038332e-05, "loss": 0.0185, "step": 20630 }, { "grad_norm": 0.17245140671730042, "learning_rate": 9.346103707181162e-05, "loss": 0.0152, "step": 20640 }, { "grad_norm": 0.15131260454654694, "learning_rate": 9.345285954040626e-05, "loss": 0.0158, "step": 20650 }, { "grad_norm": 0.20326662063598633, "learning_rate": 9.34446772570615e-05, "loss": 0.0164, "step": 20660 }, { "grad_norm": 0.1266654133796692, "learning_rate": 9.343649022267214e-05, "loss": 0.0146, "step": 20670 }, { "grad_norm": 0.2042366862297058, "learning_rate": 9.342829843813353e-05, "loss": 0.0174, "step": 20680 }, { "grad_norm": 0.16359922289848328, "learning_rate": 9.342010190434149e-05, "loss": 0.0183, "step": 20690 }, { "grad_norm": 0.19196796417236328, "learning_rate": 9.34119006221924e-05, "loss": 0.0165, "step": 20700 }, { "grad_norm": 0.21871502697467804, "learning_rate": 9.340369459258313e-05, "loss": 0.0189, "step": 20710 }, { "grad_norm": 0.2596503496170044, "learning_rate": 9.339548381641106e-05, "loss": 0.0198, "step": 20720 }, { "grad_norm": 0.2188064604997635, "learning_rate": 9.338726829457413e-05, "loss": 0.0185, "step": 20730 }, { "grad_norm": 0.16356979310512543, "learning_rate": 9.337904802797078e-05, "loss": 0.0179, "step": 20740 }, { "grad_norm": 0.12877771258354187, "learning_rate": 9.337082301749993e-05, "loss": 0.0162, "step": 20750 }, { "grad_norm": 0.21969963610172272, "learning_rate": 9.336259326406109e-05, "loss": 0.0142, "step": 20760 }, { "grad_norm": 0.18539060652256012, "learning_rate": 9.335435876855427e-05, "loss": 0.0192, "step": 20770 }, { "grad_norm": 0.1933782547712326, "learning_rate": 9.334611953187994e-05, "loss": 0.0181, "step": 20780 }, { "grad_norm": 0.2091647982597351, "learning_rate": 9.333787555493914e-05, "loss": 0.0157, "step": 20790 }, { "grad_norm": 0.22404396533966064, "learning_rate": 9.332962683863345e-05, "loss": 0.0158, "step": 20800 }, { "grad_norm": 0.14390213787555695, "learning_rate": 9.332137338386489e-05, "loss": 0.0153, "step": 20810 }, { "grad_norm": 0.18621183931827545, "learning_rate": 9.33131151915361e-05, "loss": 0.0152, "step": 20820 }, { "grad_norm": 0.20362485945224762, "learning_rate": 9.330485226255012e-05, "loss": 0.017, "step": 20830 }, { "grad_norm": 0.16140982508659363, "learning_rate": 9.329658459781061e-05, "loss": 0.0148, "step": 20840 }, { "grad_norm": 0.20499077439308167, "learning_rate": 9.328831219822172e-05, "loss": 0.0179, "step": 20850 }, { "grad_norm": 0.1845170557498932, "learning_rate": 9.328003506468808e-05, "loss": 0.017, "step": 20860 }, { "grad_norm": 0.15030737221240997, "learning_rate": 9.327175319811488e-05, "loss": 0.0174, "step": 20870 }, { "grad_norm": 0.2233992964029312, "learning_rate": 9.326346659940781e-05, "loss": 0.0148, "step": 20880 }, { "grad_norm": 0.20339904725551605, "learning_rate": 9.325517526947308e-05, "loss": 0.017, "step": 20890 }, { "grad_norm": 0.2749623656272888, "learning_rate": 9.32468792092174e-05, "loss": 0.0198, "step": 20900 }, { "grad_norm": 0.20270535349845886, "learning_rate": 9.323857841954803e-05, "loss": 0.0213, "step": 20910 }, { "grad_norm": 0.19639144837856293, "learning_rate": 9.323027290137276e-05, "loss": 0.0207, "step": 20920 }, { "grad_norm": 0.24020186066627502, "learning_rate": 9.322196265559981e-05, "loss": 0.018, "step": 20930 }, { "grad_norm": 0.18105457723140717, "learning_rate": 9.321364768313803e-05, "loss": 0.0227, "step": 20940 }, { "grad_norm": 0.24439291656017303, "learning_rate": 9.32053279848967e-05, "loss": 0.0194, "step": 20950 }, { "grad_norm": 0.17443613708019257, "learning_rate": 9.319700356178567e-05, "loss": 0.0192, "step": 20960 }, { "grad_norm": 0.16424064338207245, "learning_rate": 9.318867441471527e-05, "loss": 0.0197, "step": 20970 }, { "grad_norm": 0.18633507192134857, "learning_rate": 9.318034054459637e-05, "loss": 0.0179, "step": 20980 }, { "grad_norm": 0.21805410087108612, "learning_rate": 9.317200195234034e-05, "loss": 0.0154, "step": 20990 }, { "grad_norm": 0.18224549293518066, "learning_rate": 9.316365863885909e-05, "loss": 0.0153, "step": 21000 }, { "grad_norm": 0.1802872121334076, "learning_rate": 9.315531060506502e-05, "loss": 0.0174, "step": 21010 }, { "grad_norm": 0.19327887892723083, "learning_rate": 9.314695785187108e-05, "loss": 0.0163, "step": 21020 }, { "grad_norm": 0.16977150738239288, "learning_rate": 9.313860038019069e-05, "loss": 0.0156, "step": 21030 }, { "grad_norm": 0.1369103640317917, "learning_rate": 9.313023819093782e-05, "loss": 0.0171, "step": 21040 }, { "grad_norm": 0.14754873514175415, "learning_rate": 9.312187128502695e-05, "loss": 0.0158, "step": 21050 }, { "grad_norm": 0.17694292962551117, "learning_rate": 9.311349966337307e-05, "loss": 0.0172, "step": 21060 }, { "grad_norm": 0.1926528811454773, "learning_rate": 9.310512332689169e-05, "loss": 0.0168, "step": 21070 }, { "grad_norm": 0.22471636533737183, "learning_rate": 9.309674227649883e-05, "loss": 0.0192, "step": 21080 }, { "grad_norm": 0.1885461062192917, "learning_rate": 9.308835651311103e-05, "loss": 0.0173, "step": 21090 }, { "grad_norm": 0.2065918743610382, "learning_rate": 9.307996603764533e-05, "loss": 0.0178, "step": 21100 }, { "grad_norm": 0.17188817262649536, "learning_rate": 9.307157085101932e-05, "loss": 0.0169, "step": 21110 }, { "grad_norm": 0.19136586785316467, "learning_rate": 9.306317095415109e-05, "loss": 0.0177, "step": 21120 }, { "grad_norm": 0.21570049226284027, "learning_rate": 9.305476634795922e-05, "loss": 0.0184, "step": 21130 }, { "grad_norm": 0.18403147161006927, "learning_rate": 9.304635703336284e-05, "loss": 0.0169, "step": 21140 }, { "grad_norm": 0.1520187109708786, "learning_rate": 9.303794301128157e-05, "loss": 0.0161, "step": 21150 }, { "grad_norm": 0.2485494613647461, "learning_rate": 9.302952428263555e-05, "loss": 0.0163, "step": 21160 }, { "grad_norm": 0.1980699598789215, "learning_rate": 9.302110084834545e-05, "loss": 0.0166, "step": 21170 }, { "grad_norm": 0.19088751077651978, "learning_rate": 9.301267270933245e-05, "loss": 0.0148, "step": 21180 }, { "grad_norm": 0.16523098945617676, "learning_rate": 9.300423986651823e-05, "loss": 0.0171, "step": 21190 }, { "grad_norm": 0.14677193760871887, "learning_rate": 9.299580232082501e-05, "loss": 0.016, "step": 21200 }, { "grad_norm": 0.1532076895236969, "learning_rate": 9.298736007317547e-05, "loss": 0.0145, "step": 21210 }, { "grad_norm": 0.12220583111047745, "learning_rate": 9.297891312449288e-05, "loss": 0.0143, "step": 21220 }, { "grad_norm": 0.1925041526556015, "learning_rate": 9.297046147570094e-05, "loss": 0.0165, "step": 21230 }, { "grad_norm": 0.2026623785495758, "learning_rate": 9.296200512772396e-05, "loss": 0.0169, "step": 21240 }, { "grad_norm": 0.17538127303123474, "learning_rate": 9.295354408148668e-05, "loss": 0.0161, "step": 21250 }, { "grad_norm": 0.1778459995985031, "learning_rate": 9.294507833791441e-05, "loss": 0.0144, "step": 21260 }, { "grad_norm": 0.17772826552391052, "learning_rate": 9.293660789793295e-05, "loss": 0.0165, "step": 21270 }, { "grad_norm": 0.365648478269577, "learning_rate": 9.292813276246858e-05, "loss": 0.0182, "step": 21280 }, { "grad_norm": 0.20045168697834015, "learning_rate": 9.291965293244816e-05, "loss": 0.0186, "step": 21290 }, { "grad_norm": 0.2113478183746338, "learning_rate": 9.291116840879904e-05, "loss": 0.0184, "step": 21300 }, { "grad_norm": 0.18939140439033508, "learning_rate": 9.290267919244904e-05, "loss": 0.0197, "step": 21310 }, { "grad_norm": 0.19493167102336884, "learning_rate": 9.289418528432655e-05, "loss": 0.0161, "step": 21320 }, { "grad_norm": 0.18284724652767181, "learning_rate": 9.288568668536045e-05, "loss": 0.0171, "step": 21330 }, { "grad_norm": 0.15547366440296173, "learning_rate": 9.287718339648013e-05, "loss": 0.0154, "step": 21340 }, { "grad_norm": 0.1905839443206787, "learning_rate": 9.28686754186155e-05, "loss": 0.0166, "step": 21350 }, { "grad_norm": 0.1827927827835083, "learning_rate": 9.286016275269698e-05, "loss": 0.0149, "step": 21360 }, { "grad_norm": 0.17418009042739868, "learning_rate": 9.285164539965551e-05, "loss": 0.0139, "step": 21370 }, { "grad_norm": 0.19021457433700562, "learning_rate": 9.284312336042251e-05, "loss": 0.0153, "step": 21380 }, { "grad_norm": 0.19211091101169586, "learning_rate": 9.283459663592996e-05, "loss": 0.0169, "step": 21390 }, { "grad_norm": 0.1749766767024994, "learning_rate": 9.282606522711033e-05, "loss": 0.02, "step": 21400 }, { "grad_norm": 0.17450399696826935, "learning_rate": 9.281752913489657e-05, "loss": 0.0173, "step": 21410 }, { "grad_norm": 0.19544605910778046, "learning_rate": 9.280898836022222e-05, "loss": 0.0178, "step": 21420 }, { "grad_norm": 0.1866329461336136, "learning_rate": 9.280044290402126e-05, "loss": 0.0175, "step": 21430 }, { "grad_norm": 0.16898752748966217, "learning_rate": 9.279189276722821e-05, "loss": 0.0158, "step": 21440 }, { "grad_norm": 0.2360808253288269, "learning_rate": 9.278333795077812e-05, "loss": 0.0154, "step": 21450 }, { "grad_norm": 0.2076064497232437, "learning_rate": 9.27747784556065e-05, "loss": 0.0151, "step": 21460 }, { "grad_norm": 0.137076735496521, "learning_rate": 9.276621428264942e-05, "loss": 0.0164, "step": 21470 }, { "grad_norm": 0.1528165638446808, "learning_rate": 9.275764543284345e-05, "loss": 0.0152, "step": 21480 }, { "grad_norm": 0.17475567758083344, "learning_rate": 9.274907190712566e-05, "loss": 0.0166, "step": 21490 }, { "grad_norm": 0.16111524403095245, "learning_rate": 9.274049370643363e-05, "loss": 0.0145, "step": 21500 }, { "grad_norm": 0.23917727172374725, "learning_rate": 9.273191083170547e-05, "loss": 0.0166, "step": 21510 }, { "grad_norm": 0.20000480115413666, "learning_rate": 9.27233232838798e-05, "loss": 0.0188, "step": 21520 }, { "grad_norm": 0.20364688336849213, "learning_rate": 9.27147310638957e-05, "loss": 0.0168, "step": 21530 }, { "grad_norm": 0.21096844971179962, "learning_rate": 9.270613417269286e-05, "loss": 0.0146, "step": 21540 }, { "grad_norm": 0.17063955962657928, "learning_rate": 9.269753261121138e-05, "loss": 0.0156, "step": 21550 }, { "grad_norm": 0.15303799510002136, "learning_rate": 9.268892638039194e-05, "loss": 0.0152, "step": 21560 }, { "grad_norm": 0.1709815412759781, "learning_rate": 9.268031548117569e-05, "loss": 0.0147, "step": 21570 }, { "grad_norm": 0.1557454615831375, "learning_rate": 9.26716999145043e-05, "loss": 0.0159, "step": 21580 }, { "grad_norm": 0.176423117518425, "learning_rate": 9.266307968131998e-05, "loss": 0.0151, "step": 21590 }, { "grad_norm": 0.2647111117839813, "learning_rate": 9.26544547825654e-05, "loss": 0.0143, "step": 21600 }, { "grad_norm": 0.21842451393604279, "learning_rate": 9.264582521918376e-05, "loss": 0.018, "step": 21610 }, { "grad_norm": 0.2094969004392624, "learning_rate": 9.263719099211881e-05, "loss": 0.0166, "step": 21620 }, { "grad_norm": 0.18374429643154144, "learning_rate": 9.262855210231476e-05, "loss": 0.0164, "step": 21630 }, { "grad_norm": 0.18718861043453217, "learning_rate": 9.261990855071633e-05, "loss": 0.0156, "step": 21640 }, { "grad_norm": 0.14975139498710632, "learning_rate": 9.261126033826878e-05, "loss": 0.0167, "step": 21650 }, { "grad_norm": 0.2210858315229416, "learning_rate": 9.260260746591786e-05, "loss": 0.0169, "step": 21660 }, { "grad_norm": 0.2253565937280655, "learning_rate": 9.259394993460985e-05, "loss": 0.0147, "step": 21670 }, { "grad_norm": 0.2268485575914383, "learning_rate": 9.258528774529151e-05, "loss": 0.017, "step": 21680 }, { "grad_norm": 0.19089844822883606, "learning_rate": 9.257662089891013e-05, "loss": 0.015, "step": 21690 }, { "grad_norm": 0.13438554108142853, "learning_rate": 9.25679493964135e-05, "loss": 0.0161, "step": 21700 }, { "grad_norm": 0.1558178812265396, "learning_rate": 9.255927323874994e-05, "loss": 0.0147, "step": 21710 }, { "grad_norm": 0.16576948761940002, "learning_rate": 9.255059242686822e-05, "loss": 0.0175, "step": 21720 }, { "grad_norm": 0.16248708963394165, "learning_rate": 9.254190696171769e-05, "loss": 0.0146, "step": 21730 }, { "grad_norm": 0.1504753679037094, "learning_rate": 9.25332168442482e-05, "loss": 0.0143, "step": 21740 }, { "grad_norm": 0.22642652690410614, "learning_rate": 9.252452207541004e-05, "loss": 0.0177, "step": 21750 }, { "grad_norm": 0.20941376686096191, "learning_rate": 9.251582265615409e-05, "loss": 0.0167, "step": 21760 }, { "grad_norm": 0.2205752283334732, "learning_rate": 9.250711858743169e-05, "loss": 0.0161, "step": 21770 }, { "grad_norm": 0.19886057078838348, "learning_rate": 9.24984098701947e-05, "loss": 0.018, "step": 21780 }, { "grad_norm": 0.200644388794899, "learning_rate": 9.248969650539552e-05, "loss": 0.0135, "step": 21790 }, { "grad_norm": 0.24673204123973846, "learning_rate": 9.2480978493987e-05, "loss": 0.0151, "step": 21800 }, { "grad_norm": 0.2382466197013855, "learning_rate": 9.247225583692256e-05, "loss": 0.0156, "step": 21810 }, { "grad_norm": 0.2182195633649826, "learning_rate": 9.246352853515607e-05, "loss": 0.0137, "step": 21820 }, { "grad_norm": 0.20392145216464996, "learning_rate": 9.245479658964194e-05, "loss": 0.0156, "step": 21830 }, { "grad_norm": 0.19300751388072968, "learning_rate": 9.244606000133507e-05, "loss": 0.0149, "step": 21840 }, { "grad_norm": 0.2046452760696411, "learning_rate": 9.24373187711909e-05, "loss": 0.0159, "step": 21850 }, { "grad_norm": 0.1508258432149887, "learning_rate": 9.242857290016537e-05, "loss": 0.0155, "step": 21860 }, { "grad_norm": 0.20380032062530518, "learning_rate": 9.241982238921488e-05, "loss": 0.0169, "step": 21870 }, { "grad_norm": 0.2268860936164856, "learning_rate": 9.24110672392964e-05, "loss": 0.0164, "step": 21880 }, { "grad_norm": 0.18432095646858215, "learning_rate": 9.240230745136737e-05, "loss": 0.0183, "step": 21890 }, { "grad_norm": 0.12721949815750122, "learning_rate": 9.239354302638575e-05, "loss": 0.0179, "step": 21900 }, { "grad_norm": 0.1872881054878235, "learning_rate": 9.238477396531e-05, "loss": 0.0163, "step": 21910 }, { "grad_norm": 0.18315410614013672, "learning_rate": 9.23760002690991e-05, "loss": 0.0163, "step": 21920 }, { "grad_norm": 0.22347137331962585, "learning_rate": 9.236722193871252e-05, "loss": 0.0178, "step": 21930 }, { "grad_norm": 0.13837184011936188, "learning_rate": 9.235843897511023e-05, "loss": 0.0159, "step": 21940 }, { "grad_norm": 0.14471738040447235, "learning_rate": 9.234965137925276e-05, "loss": 0.0141, "step": 21950 }, { "grad_norm": 0.20115181803703308, "learning_rate": 9.234085915210108e-05, "loss": 0.0139, "step": 21960 }, { "grad_norm": 0.17565861344337463, "learning_rate": 9.23320622946167e-05, "loss": 0.0153, "step": 21970 }, { "grad_norm": 0.20885133743286133, "learning_rate": 9.232326080776163e-05, "loss": 0.0149, "step": 21980 }, { "grad_norm": 0.18313805758953094, "learning_rate": 9.23144546924984e-05, "loss": 0.0148, "step": 21990 }, { "grad_norm": 0.16435211896896362, "learning_rate": 9.230564394979e-05, "loss": 0.0153, "step": 22000 }, { "grad_norm": 0.2330593317747116, "learning_rate": 9.22968285806e-05, "loss": 0.0146, "step": 22010 }, { "grad_norm": 0.17668496072292328, "learning_rate": 9.228800858589242e-05, "loss": 0.0155, "step": 22020 }, { "grad_norm": 0.1501563936471939, "learning_rate": 9.227918396663179e-05, "loss": 0.0139, "step": 22030 }, { "grad_norm": 0.23215588927268982, "learning_rate": 9.227035472378319e-05, "loss": 0.015, "step": 22040 }, { "grad_norm": 0.13978973031044006, "learning_rate": 9.226152085831213e-05, "loss": 0.0144, "step": 22050 }, { "grad_norm": 0.16563600301742554, "learning_rate": 9.22526823711847e-05, "loss": 0.0146, "step": 22060 }, { "grad_norm": 0.1067100465297699, "learning_rate": 9.224383926336745e-05, "loss": 0.0136, "step": 22070 }, { "grad_norm": 0.1285308301448822, "learning_rate": 9.223499153582744e-05, "loss": 0.0147, "step": 22080 }, { "grad_norm": 0.23047077655792236, "learning_rate": 9.222613918953226e-05, "loss": 0.0149, "step": 22090 }, { "grad_norm": 0.15540723502635956, "learning_rate": 9.221728222544999e-05, "loss": 0.0136, "step": 22100 }, { "grad_norm": 0.2229870706796646, "learning_rate": 9.22084206445492e-05, "loss": 0.0155, "step": 22110 }, { "grad_norm": 0.19358167052268982, "learning_rate": 9.2199554447799e-05, "loss": 0.0133, "step": 22120 }, { "grad_norm": 0.19038501381874084, "learning_rate": 9.219068363616897e-05, "loss": 0.0169, "step": 22130 }, { "grad_norm": 0.21706002950668335, "learning_rate": 9.218180821062919e-05, "loss": 0.016, "step": 22140 }, { "grad_norm": 0.22126100957393646, "learning_rate": 9.21729281721503e-05, "loss": 0.0182, "step": 22150 }, { "grad_norm": 0.19387121498584747, "learning_rate": 9.216404352170339e-05, "loss": 0.0178, "step": 22160 }, { "grad_norm": 0.20257128775119781, "learning_rate": 9.215515426026007e-05, "loss": 0.018, "step": 22170 }, { "grad_norm": 0.18570032715797424, "learning_rate": 9.214626038879246e-05, "loss": 0.0162, "step": 22180 }, { "grad_norm": 0.2181958109140396, "learning_rate": 9.21373619082732e-05, "loss": 0.0149, "step": 22190 }, { "grad_norm": 0.19365671277046204, "learning_rate": 9.212845881967535e-05, "loss": 0.0168, "step": 22200 }, { "grad_norm": 0.18445220589637756, "learning_rate": 9.211955112397262e-05, "loss": 0.0192, "step": 22210 }, { "grad_norm": 0.16986000537872314, "learning_rate": 9.211063882213909e-05, "loss": 0.0164, "step": 22220 }, { "grad_norm": 0.15770810842514038, "learning_rate": 9.210172191514942e-05, "loss": 0.0145, "step": 22230 }, { "grad_norm": 0.1383800357580185, "learning_rate": 9.209280040397874e-05, "loss": 0.0144, "step": 22240 }, { "grad_norm": 0.24897368252277374, "learning_rate": 9.208387428960268e-05, "loss": 0.0152, "step": 22250 }, { "grad_norm": 0.1715366393327713, "learning_rate": 9.20749435729974e-05, "loss": 0.0138, "step": 22260 }, { "grad_norm": 0.15715783834457397, "learning_rate": 9.206600825513957e-05, "loss": 0.0148, "step": 22270 }, { "grad_norm": 0.19736510515213013, "learning_rate": 9.20570683370063e-05, "loss": 0.0161, "step": 22280 }, { "grad_norm": 0.18268917500972748, "learning_rate": 9.204812381957528e-05, "loss": 0.0152, "step": 22290 }, { "grad_norm": 0.22585833072662354, "learning_rate": 9.203917470382465e-05, "loss": 0.0197, "step": 22300 }, { "grad_norm": 0.16597266495227814, "learning_rate": 9.203022099073309e-05, "loss": 0.0163, "step": 22310 }, { "grad_norm": 0.17204786837100983, "learning_rate": 9.202126268127976e-05, "loss": 0.0168, "step": 22320 }, { "grad_norm": 0.24093642830848694, "learning_rate": 9.20122997764443e-05, "loss": 0.0191, "step": 22330 }, { "grad_norm": 0.17719171941280365, "learning_rate": 9.200333227720692e-05, "loss": 0.0172, "step": 22340 }, { "grad_norm": 0.2091207206249237, "learning_rate": 9.199436018454826e-05, "loss": 0.0194, "step": 22350 }, { "grad_norm": 0.1939542591571808, "learning_rate": 9.198538349944952e-05, "loss": 0.0144, "step": 22360 }, { "grad_norm": 0.14851056039333344, "learning_rate": 9.197640222289234e-05, "loss": 0.0165, "step": 22370 }, { "grad_norm": 0.17741532623767853, "learning_rate": 9.196741635585895e-05, "loss": 0.0159, "step": 22380 }, { "grad_norm": 0.21890798211097717, "learning_rate": 9.195842589933199e-05, "loss": 0.0166, "step": 22390 }, { "grad_norm": 0.170749232172966, "learning_rate": 9.194943085429466e-05, "loss": 0.015, "step": 22400 }, { "grad_norm": 0.16809874773025513, "learning_rate": 9.194043122173065e-05, "loss": 0.0173, "step": 22410 }, { "grad_norm": 0.18786345422267914, "learning_rate": 9.193142700262413e-05, "loss": 0.0184, "step": 22420 }, { "grad_norm": 0.23779520392417908, "learning_rate": 9.192241819795979e-05, "loss": 0.0159, "step": 22430 }, { "grad_norm": 0.20212770998477936, "learning_rate": 9.191340480872284e-05, "loss": 0.017, "step": 22440 }, { "grad_norm": 0.1527472734451294, "learning_rate": 9.190438683589895e-05, "loss": 0.0175, "step": 22450 }, { "grad_norm": 0.17220237851142883, "learning_rate": 9.189536428047432e-05, "loss": 0.0168, "step": 22460 }, { "grad_norm": 0.21036145091056824, "learning_rate": 9.188633714343564e-05, "loss": 0.0139, "step": 22470 }, { "grad_norm": 0.1504630297422409, "learning_rate": 9.18773054257701e-05, "loss": 0.0148, "step": 22480 }, { "grad_norm": 0.21282751858234406, "learning_rate": 9.18682691284654e-05, "loss": 0.0149, "step": 22490 }, { "grad_norm": 0.19657540321350098, "learning_rate": 9.185922825250974e-05, "loss": 0.0166, "step": 22500 }, { "grad_norm": 0.20305252075195312, "learning_rate": 9.185018279889181e-05, "loss": 0.017, "step": 22510 }, { "grad_norm": 0.17074206471443176, "learning_rate": 9.184113276860082e-05, "loss": 0.0193, "step": 22520 }, { "grad_norm": 0.21488119661808014, "learning_rate": 9.183207816262645e-05, "loss": 0.0148, "step": 22530 }, { "grad_norm": 0.2505001425743103, "learning_rate": 9.182301898195891e-05, "loss": 0.018, "step": 22540 }, { "grad_norm": 0.1708076149225235, "learning_rate": 9.181395522758889e-05, "loss": 0.0173, "step": 22550 }, { "grad_norm": 0.18471737205982208, "learning_rate": 9.180488690050759e-05, "loss": 0.0138, "step": 22560 }, { "grad_norm": 0.12627013027668, "learning_rate": 9.179581400170671e-05, "loss": 0.0156, "step": 22570 }, { "grad_norm": 0.2614692151546478, "learning_rate": 9.178673653217845e-05, "loss": 0.0172, "step": 22580 }, { "grad_norm": 0.13330011069774628, "learning_rate": 9.177765449291551e-05, "loss": 0.0152, "step": 22590 }, { "grad_norm": 0.15846970677375793, "learning_rate": 9.176856788491109e-05, "loss": 0.0155, "step": 22600 }, { "grad_norm": 0.20028509199619293, "learning_rate": 9.175947670915887e-05, "loss": 0.0183, "step": 22610 }, { "grad_norm": 0.17175766825675964, "learning_rate": 9.175038096665309e-05, "loss": 0.0187, "step": 22620 }, { "grad_norm": 0.19228902459144592, "learning_rate": 9.17412806583884e-05, "loss": 0.017, "step": 22630 }, { "grad_norm": 0.1892799437046051, "learning_rate": 9.173217578536002e-05, "loss": 0.0172, "step": 22640 }, { "grad_norm": 0.18027061223983765, "learning_rate": 9.172306634856362e-05, "loss": 0.0194, "step": 22650 }, { "grad_norm": 0.14997483789920807, "learning_rate": 9.171395234899545e-05, "loss": 0.0179, "step": 22660 }, { "grad_norm": 0.16433629393577576, "learning_rate": 9.170483378765214e-05, "loss": 0.0178, "step": 22670 }, { "grad_norm": 0.22601105272769928, "learning_rate": 9.169571066553091e-05, "loss": 0.0162, "step": 22680 }, { "grad_norm": 0.257088303565979, "learning_rate": 9.168658298362946e-05, "loss": 0.0173, "step": 22690 }, { "grad_norm": 0.19082345068454742, "learning_rate": 9.167745074294598e-05, "loss": 0.0162, "step": 22700 }, { "grad_norm": 0.2272103726863861, "learning_rate": 9.166831394447913e-05, "loss": 0.0188, "step": 22710 }, { "grad_norm": 0.16010260581970215, "learning_rate": 9.165917258922812e-05, "loss": 0.0171, "step": 22720 }, { "grad_norm": 0.17270204424858093, "learning_rate": 9.165002667819262e-05, "loss": 0.0137, "step": 22730 }, { "grad_norm": 0.18302127718925476, "learning_rate": 9.164087621237282e-05, "loss": 0.0163, "step": 22740 }, { "grad_norm": 0.18445751070976257, "learning_rate": 9.163172119276942e-05, "loss": 0.0164, "step": 22750 }, { "grad_norm": 0.2583986520767212, "learning_rate": 9.162256162038358e-05, "loss": 0.0166, "step": 22760 }, { "grad_norm": 0.16836053133010864, "learning_rate": 9.161339749621698e-05, "loss": 0.0169, "step": 22770 }, { "grad_norm": 0.14385516941547394, "learning_rate": 9.160422882127177e-05, "loss": 0.0169, "step": 22780 }, { "grad_norm": 0.18710961937904358, "learning_rate": 9.159505559655069e-05, "loss": 0.0153, "step": 22790 }, { "grad_norm": 0.28898826241493225, "learning_rate": 9.158587782305684e-05, "loss": 0.0157, "step": 22800 }, { "grad_norm": 0.20413227379322052, "learning_rate": 9.157669550179391e-05, "loss": 0.0169, "step": 22810 }, { "grad_norm": 0.15906953811645508, "learning_rate": 9.156750863376609e-05, "loss": 0.0157, "step": 22820 }, { "grad_norm": 0.14292632043361664, "learning_rate": 9.155831721997801e-05, "loss": 0.0139, "step": 22830 }, { "grad_norm": 0.21716029942035675, "learning_rate": 9.154912126143484e-05, "loss": 0.0148, "step": 22840 }, { "grad_norm": 0.18626058101654053, "learning_rate": 9.153992075914224e-05, "loss": 0.0149, "step": 22850 }, { "grad_norm": 0.15120646357536316, "learning_rate": 9.153071571410635e-05, "loss": 0.0154, "step": 22860 }, { "grad_norm": 0.2093162089586258, "learning_rate": 9.152150612733384e-05, "loss": 0.0166, "step": 22870 }, { "grad_norm": 0.15708614885807037, "learning_rate": 9.151229199983184e-05, "loss": 0.0156, "step": 22880 }, { "grad_norm": 0.16625061631202698, "learning_rate": 9.150307333260802e-05, "loss": 0.0143, "step": 22890 }, { "grad_norm": 0.2393394410610199, "learning_rate": 9.149385012667048e-05, "loss": 0.0147, "step": 22900 }, { "grad_norm": 0.17317414283752441, "learning_rate": 9.148462238302788e-05, "loss": 0.0142, "step": 22910 }, { "grad_norm": 0.17278718948364258, "learning_rate": 9.147539010268936e-05, "loss": 0.0161, "step": 22920 }, { "grad_norm": 0.22838297486305237, "learning_rate": 9.14661532866645e-05, "loss": 0.0158, "step": 22930 }, { "grad_norm": 0.14751465618610382, "learning_rate": 9.145691193596348e-05, "loss": 0.0144, "step": 22940 }, { "grad_norm": 0.1769014298915863, "learning_rate": 9.144766605159691e-05, "loss": 0.0143, "step": 22950 }, { "grad_norm": 0.14361608028411865, "learning_rate": 9.14384156345759e-05, "loss": 0.0141, "step": 22960 }, { "grad_norm": 0.1348542720079422, "learning_rate": 9.142916068591204e-05, "loss": 0.014, "step": 22970 }, { "grad_norm": 0.14084450900554657, "learning_rate": 9.141990120661746e-05, "loss": 0.0153, "step": 22980 }, { "grad_norm": 0.16753756999969482, "learning_rate": 9.141063719770475e-05, "loss": 0.0169, "step": 22990 }, { "grad_norm": 0.18714888393878937, "learning_rate": 9.140136866018704e-05, "loss": 0.0139, "step": 23000 }, { "grad_norm": 0.15131986141204834, "learning_rate": 9.139209559507788e-05, "loss": 0.0139, "step": 23010 }, { "grad_norm": 0.14875860512256622, "learning_rate": 9.13828180033914e-05, "loss": 0.0144, "step": 23020 }, { "grad_norm": 0.18291009962558746, "learning_rate": 9.137353588614212e-05, "loss": 0.0167, "step": 23030 }, { "grad_norm": 0.14293095469474792, "learning_rate": 9.136424924434519e-05, "loss": 0.0155, "step": 23040 }, { "grad_norm": 0.1600731611251831, "learning_rate": 9.135495807901615e-05, "loss": 0.0178, "step": 23050 }, { "grad_norm": 0.17433688044548035, "learning_rate": 9.134566239117108e-05, "loss": 0.0148, "step": 23060 }, { "grad_norm": 0.16691544651985168, "learning_rate": 9.13363621818265e-05, "loss": 0.0154, "step": 23070 }, { "grad_norm": 0.15688230097293854, "learning_rate": 9.132705745199953e-05, "loss": 0.0146, "step": 23080 }, { "grad_norm": 0.19648142158985138, "learning_rate": 9.131774820270768e-05, "loss": 0.0153, "step": 23090 }, { "grad_norm": 0.1817525029182434, "learning_rate": 9.130843443496901e-05, "loss": 0.0179, "step": 23100 }, { "grad_norm": 0.14841023087501526, "learning_rate": 9.129911614980206e-05, "loss": 0.0153, "step": 23110 }, { "grad_norm": 0.27746880054473877, "learning_rate": 9.128979334822584e-05, "loss": 0.015, "step": 23120 }, { "grad_norm": 0.14160728454589844, "learning_rate": 9.128046603125992e-05, "loss": 0.0155, "step": 23130 }, { "grad_norm": 0.16435587406158447, "learning_rate": 9.12711341999243e-05, "loss": 0.014, "step": 23140 }, { "grad_norm": 0.1588134765625, "learning_rate": 9.12617978552395e-05, "loss": 0.02, "step": 23150 }, { "grad_norm": 0.1910264790058136, "learning_rate": 9.12524569982265e-05, "loss": 0.0169, "step": 23160 }, { "grad_norm": 0.18155641853809357, "learning_rate": 9.124311162990684e-05, "loss": 0.0147, "step": 23170 }, { "grad_norm": 0.1852271556854248, "learning_rate": 9.12337617513025e-05, "loss": 0.0162, "step": 23180 }, { "grad_norm": 0.18310458958148956, "learning_rate": 9.122440736343596e-05, "loss": 0.0201, "step": 23190 }, { "grad_norm": 0.27525949478149414, "learning_rate": 9.12150484673302e-05, "loss": 0.0184, "step": 23200 }, { "grad_norm": 0.23320946097373962, "learning_rate": 9.120568506400873e-05, "loss": 0.0156, "step": 23210 }, { "grad_norm": 0.22528652846813202, "learning_rate": 9.119631715449548e-05, "loss": 0.0163, "step": 23220 }, { "grad_norm": 0.16024360060691833, "learning_rate": 9.118694473981493e-05, "loss": 0.0144, "step": 23230 }, { "grad_norm": 0.17791740596294403, "learning_rate": 9.117756782099203e-05, "loss": 0.0146, "step": 23240 }, { "grad_norm": 0.14803177118301392, "learning_rate": 9.11681863990522e-05, "loss": 0.0141, "step": 23250 }, { "grad_norm": 0.22511185705661774, "learning_rate": 9.115880047502142e-05, "loss": 0.0163, "step": 23260 }, { "grad_norm": 0.16294094920158386, "learning_rate": 9.114941004992609e-05, "loss": 0.0158, "step": 23270 }, { "grad_norm": 0.24831002950668335, "learning_rate": 9.114001512479317e-05, "loss": 0.0176, "step": 23280 }, { "grad_norm": 0.15668918192386627, "learning_rate": 9.113061570065003e-05, "loss": 0.0144, "step": 23290 }, { "grad_norm": 0.12280838191509247, "learning_rate": 9.112121177852459e-05, "loss": 0.0149, "step": 23300 }, { "grad_norm": 0.16938936710357666, "learning_rate": 9.111180335944527e-05, "loss": 0.0149, "step": 23310 }, { "grad_norm": 0.1740434765815735, "learning_rate": 9.110239044444093e-05, "loss": 0.0145, "step": 23320 }, { "grad_norm": 0.15807616710662842, "learning_rate": 9.109297303454099e-05, "loss": 0.0138, "step": 23330 }, { "grad_norm": 0.19772066175937653, "learning_rate": 9.108355113077526e-05, "loss": 0.0177, "step": 23340 }, { "grad_norm": 0.16666391491889954, "learning_rate": 9.107412473417419e-05, "loss": 0.0154, "step": 23350 }, { "grad_norm": 0.1513725221157074, "learning_rate": 9.106469384576858e-05, "loss": 0.0151, "step": 23360 }, { "grad_norm": 0.18390443921089172, "learning_rate": 9.105525846658978e-05, "loss": 0.0165, "step": 23370 }, { "grad_norm": 0.13937805593013763, "learning_rate": 9.104581859766965e-05, "loss": 0.0152, "step": 23380 }, { "grad_norm": 0.1530487984418869, "learning_rate": 9.10363742400405e-05, "loss": 0.0145, "step": 23390 }, { "grad_norm": 0.17928412556648254, "learning_rate": 9.102692539473518e-05, "loss": 0.0137, "step": 23400 }, { "grad_norm": 0.26974421739578247, "learning_rate": 9.101747206278697e-05, "loss": 0.0148, "step": 23410 }, { "grad_norm": 0.13277652859687805, "learning_rate": 9.100801424522968e-05, "loss": 0.0173, "step": 23420 }, { "grad_norm": 0.21112312376499176, "learning_rate": 9.099855194309762e-05, "loss": 0.0161, "step": 23430 }, { "grad_norm": 0.17406871914863586, "learning_rate": 9.098908515742554e-05, "loss": 0.0164, "step": 23440 }, { "grad_norm": 0.187938392162323, "learning_rate": 9.097961388924873e-05, "loss": 0.015, "step": 23450 }, { "grad_norm": 0.20118971168994904, "learning_rate": 9.097013813960298e-05, "loss": 0.0161, "step": 23460 }, { "grad_norm": 0.1665053367614746, "learning_rate": 9.09606579095245e-05, "loss": 0.0194, "step": 23470 }, { "grad_norm": 0.19544152915477753, "learning_rate": 9.095117320005008e-05, "loss": 0.0162, "step": 23480 }, { "grad_norm": 0.4770030975341797, "learning_rate": 9.094168401221691e-05, "loss": 0.0138, "step": 23490 }, { "grad_norm": 0.1747511625289917, "learning_rate": 9.093219034706273e-05, "loss": 0.0199, "step": 23500 }, { "grad_norm": 0.16516132652759552, "learning_rate": 9.092269220562577e-05, "loss": 0.0204, "step": 23510 }, { "grad_norm": 0.15803222358226776, "learning_rate": 9.09131895889447e-05, "loss": 0.0184, "step": 23520 }, { "grad_norm": 0.29419583082199097, "learning_rate": 9.090368249805873e-05, "loss": 0.0175, "step": 23530 }, { "grad_norm": 0.1687716841697693, "learning_rate": 9.089417093400754e-05, "loss": 0.0192, "step": 23540 }, { "grad_norm": 0.142417311668396, "learning_rate": 9.088465489783131e-05, "loss": 0.0164, "step": 23550 }, { "grad_norm": 0.1564468890428543, "learning_rate": 9.087513439057068e-05, "loss": 0.0165, "step": 23560 }, { "grad_norm": 0.18856556713581085, "learning_rate": 9.08656094132668e-05, "loss": 0.0159, "step": 23570 }, { "grad_norm": 0.23392660915851593, "learning_rate": 9.085607996696134e-05, "loss": 0.0158, "step": 23580 }, { "grad_norm": 0.27784523367881775, "learning_rate": 9.084654605269639e-05, "loss": 0.016, "step": 23590 }, { "grad_norm": 0.30660930275917053, "learning_rate": 9.083700767151457e-05, "loss": 0.0171, "step": 23600 }, { "grad_norm": 0.18891815841197968, "learning_rate": 9.082746482445898e-05, "loss": 0.0189, "step": 23610 }, { "grad_norm": 0.17812001705169678, "learning_rate": 9.081791751257325e-05, "loss": 0.0165, "step": 23620 }, { "grad_norm": 0.19721940159797668, "learning_rate": 9.080836573690142e-05, "loss": 0.0178, "step": 23630 }, { "grad_norm": 0.13736510276794434, "learning_rate": 9.079880949848805e-05, "loss": 0.015, "step": 23640 }, { "grad_norm": 0.23869450390338898, "learning_rate": 9.078924879837822e-05, "loss": 0.0165, "step": 23650 }, { "grad_norm": 0.1698514223098755, "learning_rate": 9.077968363761747e-05, "loss": 0.017, "step": 23660 }, { "grad_norm": 0.21746650338172913, "learning_rate": 9.077011401725182e-05, "loss": 0.0165, "step": 23670 }, { "grad_norm": 0.15803660452365875, "learning_rate": 9.07605399383278e-05, "loss": 0.0136, "step": 23680 }, { "grad_norm": 0.16691593825817108, "learning_rate": 9.075096140189243e-05, "loss": 0.0155, "step": 23690 }, { "grad_norm": 0.19942960143089294, "learning_rate": 9.074137840899318e-05, "loss": 0.0158, "step": 23700 }, { "grad_norm": 0.18284036219120026, "learning_rate": 9.073179096067804e-05, "loss": 0.0145, "step": 23710 }, { "grad_norm": 0.23580434918403625, "learning_rate": 9.072219905799549e-05, "loss": 0.0172, "step": 23720 }, { "grad_norm": 0.16619688272476196, "learning_rate": 9.071260270199447e-05, "loss": 0.0159, "step": 23730 }, { "grad_norm": 0.11420824378728867, "learning_rate": 9.070300189372441e-05, "loss": 0.015, "step": 23740 }, { "grad_norm": 0.15955479443073273, "learning_rate": 9.069339663423528e-05, "loss": 0.0153, "step": 23750 }, { "grad_norm": 0.1596159040927887, "learning_rate": 9.068378692457747e-05, "loss": 0.017, "step": 23760 }, { "grad_norm": 0.15047100186347961, "learning_rate": 9.067417276580189e-05, "loss": 0.015, "step": 23770 }, { "grad_norm": 0.2278054803609848, "learning_rate": 9.066455415895993e-05, "loss": 0.0146, "step": 23780 }, { "grad_norm": 0.1570626050233841, "learning_rate": 9.065493110510346e-05, "loss": 0.0137, "step": 23790 }, { "grad_norm": 0.18404139578342438, "learning_rate": 9.064530360528484e-05, "loss": 0.0148, "step": 23800 }, { "grad_norm": 0.2013808786869049, "learning_rate": 9.063567166055695e-05, "loss": 0.0168, "step": 23810 }, { "grad_norm": 0.15588954091072083, "learning_rate": 9.062603527197308e-05, "loss": 0.0144, "step": 23820 }, { "grad_norm": 0.19740836322307587, "learning_rate": 9.06163944405871e-05, "loss": 0.0157, "step": 23830 }, { "grad_norm": 0.235897496342659, "learning_rate": 9.060674916745327e-05, "loss": 0.0147, "step": 23840 }, { "grad_norm": 0.16710400581359863, "learning_rate": 9.05970994536264e-05, "loss": 0.016, "step": 23850 }, { "grad_norm": 0.21956558525562286, "learning_rate": 9.05874453001618e-05, "loss": 0.0146, "step": 23860 }, { "grad_norm": 0.1282535195350647, "learning_rate": 9.057778670811517e-05, "loss": 0.014, "step": 23870 }, { "grad_norm": 0.20983844995498657, "learning_rate": 9.056812367854281e-05, "loss": 0.0139, "step": 23880 }, { "grad_norm": 0.21455012261867523, "learning_rate": 9.055845621250143e-05, "loss": 0.0144, "step": 23890 }, { "grad_norm": 0.19147683680057526, "learning_rate": 9.054878431104825e-05, "loss": 0.0165, "step": 23900 }, { "grad_norm": 0.19117975234985352, "learning_rate": 9.0539107975241e-05, "loss": 0.015, "step": 23910 }, { "grad_norm": 0.25795432925224304, "learning_rate": 9.052942720613784e-05, "loss": 0.0148, "step": 23920 }, { "grad_norm": 0.1680642068386078, "learning_rate": 9.051974200479745e-05, "loss": 0.0148, "step": 23930 }, { "grad_norm": 0.11823713034391403, "learning_rate": 9.051005237227901e-05, "loss": 0.0136, "step": 23940 }, { "grad_norm": 0.11665131151676178, "learning_rate": 9.050035830964215e-05, "loss": 0.0144, "step": 23950 }, { "grad_norm": 0.28598830103874207, "learning_rate": 9.049065981794698e-05, "loss": 0.0171, "step": 23960 }, { "grad_norm": 0.1680479645729065, "learning_rate": 9.048095689825414e-05, "loss": 0.0176, "step": 23970 }, { "grad_norm": 0.1662273406982422, "learning_rate": 9.047124955162472e-05, "loss": 0.0155, "step": 23980 }, { "grad_norm": 0.141438826918602, "learning_rate": 9.046153777912028e-05, "loss": 0.0136, "step": 23990 }, { "grad_norm": 0.20999601483345032, "learning_rate": 9.045182158180292e-05, "loss": 0.014, "step": 24000 }, { "grad_norm": 0.19691471755504608, "learning_rate": 9.044210096073516e-05, "loss": 0.0183, "step": 24010 }, { "grad_norm": 0.14377352595329285, "learning_rate": 9.043237591698004e-05, "loss": 0.0155, "step": 24020 }, { "grad_norm": 0.2030940055847168, "learning_rate": 9.04226464516011e-05, "loss": 0.0147, "step": 24030 }, { "grad_norm": 0.3570360243320465, "learning_rate": 9.041291256566229e-05, "loss": 0.0136, "step": 24040 }, { "grad_norm": 0.2040369063615799, "learning_rate": 9.040317426022814e-05, "loss": 0.0155, "step": 24050 }, { "grad_norm": 0.18643756210803986, "learning_rate": 9.03934315363636e-05, "loss": 0.0165, "step": 24060 }, { "grad_norm": 0.19406500458717346, "learning_rate": 9.038368439513409e-05, "loss": 0.0154, "step": 24070 }, { "grad_norm": 0.13540434837341309, "learning_rate": 9.03739328376056e-05, "loss": 0.0151, "step": 24080 }, { "grad_norm": 0.22744829952716827, "learning_rate": 9.036417686484451e-05, "loss": 0.015, "step": 24090 }, { "grad_norm": 0.15306907892227173, "learning_rate": 9.035441647791773e-05, "loss": 0.0166, "step": 24100 }, { "grad_norm": 0.19598853588104248, "learning_rate": 9.034465167789263e-05, "loss": 0.0168, "step": 24110 }, { "grad_norm": 0.15921373665332794, "learning_rate": 9.033488246583706e-05, "loss": 0.0137, "step": 24120 }, { "grad_norm": 0.18612277507781982, "learning_rate": 9.032510884281941e-05, "loss": 0.0158, "step": 24130 }, { "grad_norm": 0.1904963105916977, "learning_rate": 9.031533080990848e-05, "loss": 0.016, "step": 24140 }, { "grad_norm": 0.14110973477363586, "learning_rate": 9.030554836817358e-05, "loss": 0.0151, "step": 24150 }, { "grad_norm": 0.1778392344713211, "learning_rate": 9.029576151868451e-05, "loss": 0.0184, "step": 24160 }, { "grad_norm": 0.16313841938972473, "learning_rate": 9.028597026251155e-05, "loss": 0.0159, "step": 24170 }, { "grad_norm": 0.16602948307991028, "learning_rate": 9.027617460072547e-05, "loss": 0.0152, "step": 24180 }, { "grad_norm": 0.2339215874671936, "learning_rate": 9.026637453439745e-05, "loss": 0.0137, "step": 24190 }, { "grad_norm": 0.15197812020778656, "learning_rate": 9.025657006459927e-05, "loss": 0.0165, "step": 24200 }, { "grad_norm": 0.16929180920124054, "learning_rate": 9.024676119240311e-05, "loss": 0.0157, "step": 24210 }, { "grad_norm": 0.146244078874588, "learning_rate": 9.023694791888166e-05, "loss": 0.0169, "step": 24220 }, { "grad_norm": 0.16114163398742676, "learning_rate": 9.022713024510808e-05, "loss": 0.0137, "step": 24230 }, { "grad_norm": 0.18619300425052643, "learning_rate": 9.021730817215601e-05, "loss": 0.0126, "step": 24240 }, { "grad_norm": 0.1897222101688385, "learning_rate": 9.02074817010996e-05, "loss": 0.0133, "step": 24250 }, { "grad_norm": 0.1662486493587494, "learning_rate": 9.019765083301342e-05, "loss": 0.0133, "step": 24260 }, { "grad_norm": 0.11123771965503693, "learning_rate": 9.01878155689726e-05, "loss": 0.014, "step": 24270 }, { "grad_norm": 0.15936465561389923, "learning_rate": 9.017797591005268e-05, "loss": 0.0151, "step": 24280 }, { "grad_norm": 0.15545552968978882, "learning_rate": 9.016813185732972e-05, "loss": 0.0155, "step": 24290 }, { "grad_norm": 0.20297545194625854, "learning_rate": 9.015828341188027e-05, "loss": 0.0159, "step": 24300 }, { "grad_norm": 0.21410584449768066, "learning_rate": 9.01484305747813e-05, "loss": 0.018, "step": 24310 }, { "grad_norm": 0.3123021721839905, "learning_rate": 9.013857334711033e-05, "loss": 0.0158, "step": 24320 }, { "grad_norm": 0.1581624150276184, "learning_rate": 9.012871172994534e-05, "loss": 0.0151, "step": 24330 }, { "grad_norm": 0.18008553981781006, "learning_rate": 9.011884572436476e-05, "loss": 0.0159, "step": 24340 }, { "grad_norm": 0.17992453277111053, "learning_rate": 9.010897533144754e-05, "loss": 0.0157, "step": 24350 }, { "grad_norm": 0.19231243431568146, "learning_rate": 9.009910055227306e-05, "loss": 0.015, "step": 24360 }, { "grad_norm": 0.22180762887001038, "learning_rate": 9.008922138792124e-05, "loss": 0.0149, "step": 24370 }, { "grad_norm": 0.13532654941082, "learning_rate": 9.007933783947244e-05, "loss": 0.0133, "step": 24380 }, { "grad_norm": 0.18333721160888672, "learning_rate": 9.006944990800752e-05, "loss": 0.0157, "step": 24390 }, { "grad_norm": 0.197284996509552, "learning_rate": 9.005955759460779e-05, "loss": 0.0159, "step": 24400 }, { "grad_norm": 0.16886131465435028, "learning_rate": 9.004966090035508e-05, "loss": 0.017, "step": 24410 }, { "grad_norm": 0.1867254227399826, "learning_rate": 9.003975982633166e-05, "loss": 0.0136, "step": 24420 }, { "grad_norm": 0.16434744000434875, "learning_rate": 9.00298543736203e-05, "loss": 0.0149, "step": 24430 }, { "grad_norm": 0.19547270238399506, "learning_rate": 9.001994454330427e-05, "loss": 0.0172, "step": 24440 }, { "grad_norm": 0.16151757538318634, "learning_rate": 9.001003033646727e-05, "loss": 0.0187, "step": 24450 }, { "grad_norm": 0.13536648452281952, "learning_rate": 9.00001117541935e-05, "loss": 0.0125, "step": 24460 }, { "grad_norm": 0.13120931386947632, "learning_rate": 8.999018879756764e-05, "loss": 0.0128, "step": 24470 }, { "grad_norm": 0.16685034334659576, "learning_rate": 8.998026146767487e-05, "loss": 0.0137, "step": 24480 }, { "grad_norm": 0.1654520332813263, "learning_rate": 8.99703297656008e-05, "loss": 0.0122, "step": 24490 }, { "grad_norm": 0.15154746174812317, "learning_rate": 8.996039369243156e-05, "loss": 0.0147, "step": 24500 }, { "grad_norm": 0.16329853236675262, "learning_rate": 8.995045324925378e-05, "loss": 0.0143, "step": 24510 }, { "grad_norm": 0.20847934484481812, "learning_rate": 8.994050843715448e-05, "loss": 0.0137, "step": 24520 }, { "grad_norm": 0.15884122252464294, "learning_rate": 8.993055925722121e-05, "loss": 0.0138, "step": 24530 }, { "grad_norm": 0.18771220743656158, "learning_rate": 8.992060571054202e-05, "loss": 0.0131, "step": 24540 }, { "grad_norm": 0.16656629741191864, "learning_rate": 8.991064779820542e-05, "loss": 0.0149, "step": 24550 }, { "grad_norm": 0.1505620777606964, "learning_rate": 8.990068552130036e-05, "loss": 0.0145, "step": 24560 }, { "grad_norm": 0.22291453182697296, "learning_rate": 8.989071888091634e-05, "loss": 0.0142, "step": 24570 }, { "grad_norm": 0.1658901423215866, "learning_rate": 8.988074787814329e-05, "loss": 0.0146, "step": 24580 }, { "grad_norm": 0.1350901573896408, "learning_rate": 8.987077251407158e-05, "loss": 0.0125, "step": 24590 }, { "grad_norm": 0.14145642518997192, "learning_rate": 8.986079278979216e-05, "loss": 0.0127, "step": 24600 }, { "grad_norm": 0.144391268491745, "learning_rate": 8.985080870639635e-05, "loss": 0.0129, "step": 24610 }, { "grad_norm": 0.18388204276561737, "learning_rate": 8.984082026497603e-05, "loss": 0.0133, "step": 24620 }, { "grad_norm": 0.17427577078342438, "learning_rate": 8.98308274666235e-05, "loss": 0.0133, "step": 24630 }, { "grad_norm": 0.16116395592689514, "learning_rate": 8.982083031243155e-05, "loss": 0.0141, "step": 24640 }, { "grad_norm": 0.2086905837059021, "learning_rate": 8.98108288034935e-05, "loss": 0.0127, "step": 24650 }, { "grad_norm": 0.1924665868282318, "learning_rate": 8.980082294090305e-05, "loss": 0.0145, "step": 24660 }, { "grad_norm": 0.19992028176784515, "learning_rate": 8.979081272575443e-05, "loss": 0.0156, "step": 24670 }, { "grad_norm": 0.23915007710456848, "learning_rate": 8.978079815914236e-05, "loss": 0.0146, "step": 24680 }, { "grad_norm": 0.22803056240081787, "learning_rate": 8.977077924216202e-05, "loss": 0.016, "step": 24690 }, { "grad_norm": 0.2094060480594635, "learning_rate": 8.976075597590905e-05, "loss": 0.0154, "step": 24700 }, { "grad_norm": 0.19170644879341125, "learning_rate": 8.975072836147958e-05, "loss": 0.0174, "step": 24710 }, { "grad_norm": 0.11964245140552521, "learning_rate": 8.974069639997025e-05, "loss": 0.0131, "step": 24720 }, { "grad_norm": 0.16825418174266815, "learning_rate": 8.973066009247808e-05, "loss": 0.0127, "step": 24730 }, { "grad_norm": 0.1341848522424698, "learning_rate": 8.972061944010066e-05, "loss": 0.014, "step": 24740 }, { "grad_norm": 0.1148313358426094, "learning_rate": 8.971057444393603e-05, "loss": 0.0134, "step": 24750 }, { "grad_norm": 0.16556906700134277, "learning_rate": 8.970052510508268e-05, "loss": 0.0135, "step": 24760 }, { "grad_norm": 0.16219742596149445, "learning_rate": 8.969047142463959e-05, "loss": 0.0175, "step": 24770 }, { "grad_norm": 0.13056118786334991, "learning_rate": 8.968041340370621e-05, "loss": 0.0148, "step": 24780 }, { "grad_norm": 0.13738936185836792, "learning_rate": 8.96703510433825e-05, "loss": 0.0152, "step": 24790 }, { "grad_norm": 0.16609907150268555, "learning_rate": 8.966028434476883e-05, "loss": 0.0142, "step": 24800 }, { "grad_norm": 0.19253994524478912, "learning_rate": 8.96502133089661e-05, "loss": 0.013, "step": 24810 }, { "grad_norm": 0.17113007605075836, "learning_rate": 8.964013793707564e-05, "loss": 0.0152, "step": 24820 }, { "grad_norm": 0.18656478822231293, "learning_rate": 8.963005823019932e-05, "loss": 0.0152, "step": 24830 }, { "grad_norm": 0.1753595471382141, "learning_rate": 8.961997418943939e-05, "loss": 0.0129, "step": 24840 }, { "grad_norm": 0.17350715398788452, "learning_rate": 8.960988581589865e-05, "loss": 0.0144, "step": 24850 }, { "grad_norm": 0.12778112292289734, "learning_rate": 8.959979311068037e-05, "loss": 0.018, "step": 24860 }, { "grad_norm": 0.33147069811820984, "learning_rate": 8.958969607488823e-05, "loss": 0.0134, "step": 24870 }, { "grad_norm": 0.17481157183647156, "learning_rate": 8.957959470962647e-05, "loss": 0.0151, "step": 24880 }, { "grad_norm": 0.2483697086572647, "learning_rate": 8.956948901599971e-05, "loss": 0.015, "step": 24890 }, { "grad_norm": 0.1625925600528717, "learning_rate": 8.955937899511315e-05, "loss": 0.0137, "step": 24900 }, { "grad_norm": 0.2087913602590561, "learning_rate": 8.954926464807238e-05, "loss": 0.0161, "step": 24910 }, { "grad_norm": 0.17154495418071747, "learning_rate": 8.953914597598347e-05, "loss": 0.0159, "step": 24920 }, { "grad_norm": 0.22425541281700134, "learning_rate": 8.952902297995303e-05, "loss": 0.0156, "step": 24930 }, { "grad_norm": 0.22471408545970917, "learning_rate": 8.951889566108804e-05, "loss": 0.0175, "step": 24940 }, { "grad_norm": 0.1953902691602707, "learning_rate": 8.950876402049606e-05, "loss": 0.0127, "step": 24950 }, { "grad_norm": 0.1454121172428131, "learning_rate": 8.949862805928504e-05, "loss": 0.0169, "step": 24960 }, { "grad_norm": 0.19048653542995453, "learning_rate": 8.948848777856343e-05, "loss": 0.0163, "step": 24970 }, { "grad_norm": 0.17456325888633728, "learning_rate": 8.947834317944017e-05, "loss": 0.0139, "step": 24980 }, { "grad_norm": 0.13818149268627167, "learning_rate": 8.946819426302466e-05, "loss": 0.013, "step": 24990 }, { "grad_norm": 0.15762744843959808, "learning_rate": 8.945804103042676e-05, "loss": 0.0172, "step": 25000 }, { "grad_norm": 0.15810526907444, "learning_rate": 8.944788348275681e-05, "loss": 0.0137, "step": 25010 }, { "grad_norm": 0.17377498745918274, "learning_rate": 8.943772162112565e-05, "loss": 0.0142, "step": 25020 }, { "grad_norm": 0.15704262256622314, "learning_rate": 8.942755544664454e-05, "loss": 0.015, "step": 25030 }, { "grad_norm": 0.13766498863697052, "learning_rate": 8.941738496042525e-05, "loss": 0.0148, "step": 25040 }, { "grad_norm": 0.15120889246463776, "learning_rate": 8.940721016357999e-05, "loss": 0.0133, "step": 25050 }, { "grad_norm": 0.13207770884037018, "learning_rate": 8.939703105722148e-05, "loss": 0.0149, "step": 25060 }, { "grad_norm": 0.15262392163276672, "learning_rate": 8.93868476424629e-05, "loss": 0.0129, "step": 25070 }, { "grad_norm": 0.20610322058200836, "learning_rate": 8.937665992041786e-05, "loss": 0.0178, "step": 25080 }, { "grad_norm": 0.17035339772701263, "learning_rate": 8.93664678922005e-05, "loss": 0.0151, "step": 25090 }, { "grad_norm": 0.21174900233745575, "learning_rate": 8.93562715589254e-05, "loss": 0.0149, "step": 25100 }, { "grad_norm": 0.23231254518032074, "learning_rate": 8.934607092170762e-05, "loss": 0.0167, "step": 25110 }, { "grad_norm": 0.14680244028568268, "learning_rate": 8.933586598166266e-05, "loss": 0.0144, "step": 25120 }, { "grad_norm": 0.1779378056526184, "learning_rate": 8.932565673990655e-05, "loss": 0.0152, "step": 25130 }, { "grad_norm": 0.1593616008758545, "learning_rate": 8.931544319755574e-05, "loss": 0.0161, "step": 25140 }, { "grad_norm": 0.19209204614162445, "learning_rate": 8.930522535572718e-05, "loss": 0.0142, "step": 25150 }, { "grad_norm": 0.17815297842025757, "learning_rate": 8.929500321553826e-05, "loss": 0.0142, "step": 25160 }, { "grad_norm": 0.1654997617006302, "learning_rate": 8.928477677810686e-05, "loss": 0.0137, "step": 25170 }, { "grad_norm": 0.2081628292798996, "learning_rate": 8.927454604455137e-05, "loss": 0.0135, "step": 25180 }, { "grad_norm": 0.19332073628902435, "learning_rate": 8.926431101599053e-05, "loss": 0.0148, "step": 25190 }, { "grad_norm": 0.17210239171981812, "learning_rate": 8.925407169354369e-05, "loss": 0.0162, "step": 25200 }, { "grad_norm": 0.15434573590755463, "learning_rate": 8.92438280783306e-05, "loss": 0.0154, "step": 25210 }, { "grad_norm": 0.1942235678434372, "learning_rate": 8.923358017147146e-05, "loss": 0.0144, "step": 25220 }, { "grad_norm": 0.13992135226726532, "learning_rate": 8.922332797408697e-05, "loss": 0.0145, "step": 25230 }, { "grad_norm": 0.152179554104805, "learning_rate": 8.921307148729831e-05, "loss": 0.0138, "step": 25240 }, { "grad_norm": 0.18476350605487823, "learning_rate": 8.920281071222712e-05, "loss": 0.0145, "step": 25250 }, { "grad_norm": 0.19394861161708832, "learning_rate": 8.919254564999548e-05, "loss": 0.013, "step": 25260 }, { "grad_norm": 0.16560395061969757, "learning_rate": 8.918227630172598e-05, "loss": 0.0144, "step": 25270 }, { "grad_norm": 0.18055380880832672, "learning_rate": 8.917200266854165e-05, "loss": 0.0133, "step": 25280 }, { "grad_norm": 0.1347871720790863, "learning_rate": 8.9161724751566e-05, "loss": 0.0145, "step": 25290 }, { "grad_norm": 0.26533934473991394, "learning_rate": 8.915144255192302e-05, "loss": 0.0129, "step": 25300 }, { "grad_norm": 0.1771220713853836, "learning_rate": 8.914115607073714e-05, "loss": 0.0146, "step": 25310 }, { "grad_norm": 0.1383904665708542, "learning_rate": 8.913086530913327e-05, "loss": 0.0143, "step": 25320 }, { "grad_norm": 0.21341799199581146, "learning_rate": 8.912057026823681e-05, "loss": 0.0148, "step": 25330 }, { "grad_norm": 0.19655311107635498, "learning_rate": 8.91102709491736e-05, "loss": 0.014, "step": 25340 }, { "grad_norm": 0.1518326997756958, "learning_rate": 8.909996735306996e-05, "loss": 0.0157, "step": 25350 }, { "grad_norm": 0.16192488372325897, "learning_rate": 8.908965948105268e-05, "loss": 0.0137, "step": 25360 }, { "grad_norm": 0.2131769061088562, "learning_rate": 8.907934733424901e-05, "loss": 0.0143, "step": 25370 }, { "grad_norm": 0.13152842223644257, "learning_rate": 8.906903091378666e-05, "loss": 0.0135, "step": 25380 }, { "grad_norm": 0.13549137115478516, "learning_rate": 8.905871022079384e-05, "loss": 0.0133, "step": 25390 }, { "grad_norm": 0.13296404480934143, "learning_rate": 8.90483852563992e-05, "loss": 0.0121, "step": 25400 }, { "grad_norm": 0.1629398763179779, "learning_rate": 8.903805602173185e-05, "loss": 0.0132, "step": 25410 }, { "grad_norm": 0.21930858492851257, "learning_rate": 8.902772251792137e-05, "loss": 0.0167, "step": 25420 }, { "grad_norm": 0.14190679788589478, "learning_rate": 8.901738474609786e-05, "loss": 0.0162, "step": 25430 }, { "grad_norm": 0.15768441557884216, "learning_rate": 8.900704270739179e-05, "loss": 0.0148, "step": 25440 }, { "grad_norm": 0.18714043498039246, "learning_rate": 8.89966964029342e-05, "loss": 0.017, "step": 25450 }, { "grad_norm": 0.22762639820575714, "learning_rate": 8.898634583385652e-05, "loss": 0.0135, "step": 25460 }, { "grad_norm": 0.3789268732070923, "learning_rate": 8.897599100129065e-05, "loss": 0.0153, "step": 25470 }, { "grad_norm": 0.1948314905166626, "learning_rate": 8.896563190636903e-05, "loss": 0.0172, "step": 25480 }, { "grad_norm": 0.17147351801395416, "learning_rate": 8.895526855022448e-05, "loss": 0.0143, "step": 25490 }, { "grad_norm": 0.1640327125787735, "learning_rate": 8.894490093399033e-05, "loss": 0.0167, "step": 25500 }, { "grad_norm": 0.16514074802398682, "learning_rate": 8.893452905880035e-05, "loss": 0.0135, "step": 25510 }, { "grad_norm": 0.18633274734020233, "learning_rate": 8.892415292578883e-05, "loss": 0.0137, "step": 25520 }, { "grad_norm": 0.18782217800617218, "learning_rate": 8.891377253609046e-05, "loss": 0.0158, "step": 25530 }, { "grad_norm": 0.14007389545440674, "learning_rate": 8.890338789084043e-05, "loss": 0.0133, "step": 25540 }, { "grad_norm": 0.21584759652614594, "learning_rate": 8.88929989911744e-05, "loss": 0.0154, "step": 25550 }, { "grad_norm": 0.1351122260093689, "learning_rate": 8.888260583822847e-05, "loss": 0.0153, "step": 25560 }, { "grad_norm": 0.1671961545944214, "learning_rate": 8.887220843313921e-05, "loss": 0.0159, "step": 25570 }, { "grad_norm": 0.203411266207695, "learning_rate": 8.88618067770437e-05, "loss": 0.013, "step": 25580 }, { "grad_norm": 0.14970968663692474, "learning_rate": 8.885140087107942e-05, "loss": 0.0166, "step": 25590 }, { "grad_norm": 0.22833922505378723, "learning_rate": 8.884099071638436e-05, "loss": 0.0142, "step": 25600 }, { "grad_norm": 0.15484176576137543, "learning_rate": 8.883057631409695e-05, "loss": 0.0142, "step": 25610 }, { "grad_norm": 0.20981831848621368, "learning_rate": 8.882015766535608e-05, "loss": 0.0143, "step": 25620 }, { "grad_norm": 0.1901450753211975, "learning_rate": 8.880973477130115e-05, "loss": 0.0161, "step": 25630 }, { "grad_norm": 0.16175951063632965, "learning_rate": 8.879930763307197e-05, "loss": 0.013, "step": 25640 }, { "grad_norm": 0.21175344288349152, "learning_rate": 8.878887625180884e-05, "loss": 0.0146, "step": 25650 }, { "grad_norm": 0.28172484040260315, "learning_rate": 8.877844062865253e-05, "loss": 0.0145, "step": 25660 }, { "grad_norm": 0.14544568955898285, "learning_rate": 8.876800076474424e-05, "loss": 0.0184, "step": 25670 }, { "grad_norm": 0.1552160382270813, "learning_rate": 8.875755666122568e-05, "loss": 0.018, "step": 25680 }, { "grad_norm": 0.208025261759758, "learning_rate": 8.8747108319239e-05, "loss": 0.0144, "step": 25690 }, { "grad_norm": 0.19828978180885315, "learning_rate": 8.87366557399268e-05, "loss": 0.0141, "step": 25700 }, { "grad_norm": 0.17111636698246002, "learning_rate": 8.872619892443217e-05, "loss": 0.0147, "step": 25710 }, { "grad_norm": 0.1698504537343979, "learning_rate": 8.871573787389865e-05, "loss": 0.0133, "step": 25720 }, { "grad_norm": 0.16457729041576385, "learning_rate": 8.870527258947024e-05, "loss": 0.0117, "step": 25730 }, { "grad_norm": 0.16671092808246613, "learning_rate": 8.869480307229143e-05, "loss": 0.0136, "step": 25740 }, { "grad_norm": 0.19058218598365784, "learning_rate": 8.868432932350712e-05, "loss": 0.0139, "step": 25750 }, { "grad_norm": 0.14281991124153137, "learning_rate": 8.867385134426272e-05, "loss": 0.0143, "step": 25760 }, { "grad_norm": 0.14961548149585724, "learning_rate": 8.866336913570407e-05, "loss": 0.015, "step": 25770 }, { "grad_norm": 0.16481536626815796, "learning_rate": 8.865288269897751e-05, "loss": 0.0124, "step": 25780 }, { "grad_norm": 0.1408732831478119, "learning_rate": 8.864239203522981e-05, "loss": 0.0161, "step": 25790 }, { "grad_norm": 0.1592196226119995, "learning_rate": 8.863189714560822e-05, "loss": 0.0134, "step": 25800 }, { "grad_norm": 0.16659727692604065, "learning_rate": 8.862139803126043e-05, "loss": 0.0117, "step": 25810 }, { "grad_norm": 0.2179621309041977, "learning_rate": 8.861089469333463e-05, "loss": 0.0139, "step": 25820 }, { "grad_norm": 0.12630246579647064, "learning_rate": 8.860038713297944e-05, "loss": 0.0114, "step": 25830 }, { "grad_norm": 0.15875530242919922, "learning_rate": 8.858987535134394e-05, "loss": 0.0137, "step": 25840 }, { "grad_norm": 0.13083048164844513, "learning_rate": 8.857935934957769e-05, "loss": 0.0149, "step": 25850 }, { "grad_norm": 0.16407807171344757, "learning_rate": 8.856883912883071e-05, "loss": 0.0138, "step": 25860 }, { "grad_norm": 0.1890053004026413, "learning_rate": 8.855831469025346e-05, "loss": 0.0131, "step": 25870 }, { "grad_norm": 0.1890653371810913, "learning_rate": 8.854778603499689e-05, "loss": 0.0134, "step": 25880 }, { "grad_norm": 0.16729597747325897, "learning_rate": 8.85372531642124e-05, "loss": 0.0154, "step": 25890 }, { "grad_norm": 0.1394842267036438, "learning_rate": 8.852671607905185e-05, "loss": 0.0154, "step": 25900 }, { "grad_norm": 0.1463070660829544, "learning_rate": 8.851617478066754e-05, "loss": 0.0141, "step": 25910 }, { "grad_norm": 0.1700734943151474, "learning_rate": 8.850562927021227e-05, "loss": 0.0135, "step": 25920 }, { "grad_norm": 0.2367638796567917, "learning_rate": 8.849507954883928e-05, "loss": 0.0153, "step": 25930 }, { "grad_norm": 0.17399314045906067, "learning_rate": 8.848452561770226e-05, "loss": 0.0144, "step": 25940 }, { "grad_norm": 0.12258937954902649, "learning_rate": 8.847396747795538e-05, "loss": 0.0135, "step": 25950 }, { "grad_norm": 0.19011500477790833, "learning_rate": 8.846340513075327e-05, "loss": 0.0135, "step": 25960 }, { "grad_norm": 0.3145267963409424, "learning_rate": 8.845283857725099e-05, "loss": 0.0124, "step": 25970 }, { "grad_norm": 0.16089296340942383, "learning_rate": 8.844226781860409e-05, "loss": 0.0162, "step": 25980 }, { "grad_norm": 0.15661783516407013, "learning_rate": 8.84316928559686e-05, "loss": 0.0167, "step": 25990 }, { "grad_norm": 0.17289067804813385, "learning_rate": 8.842111369050094e-05, "loss": 0.0131, "step": 26000 }, { "grad_norm": 0.1683846414089203, "learning_rate": 8.841053032335808e-05, "loss": 0.0152, "step": 26010 }, { "grad_norm": 0.16175717115402222, "learning_rate": 8.839994275569735e-05, "loss": 0.0131, "step": 26020 }, { "grad_norm": 0.12406405061483383, "learning_rate": 8.838935098867662e-05, "loss": 0.014, "step": 26030 }, { "grad_norm": 0.43878912925720215, "learning_rate": 8.837875502345418e-05, "loss": 0.0159, "step": 26040 }, { "grad_norm": 0.17318928241729736, "learning_rate": 8.83681548611888e-05, "loss": 0.0134, "step": 26050 }, { "grad_norm": 0.15945562720298767, "learning_rate": 8.835755050303969e-05, "loss": 0.0173, "step": 26060 }, { "grad_norm": 0.1378997564315796, "learning_rate": 8.834694195016653e-05, "loss": 0.0147, "step": 26070 }, { "grad_norm": 0.19684959948062897, "learning_rate": 8.833632920372942e-05, "loss": 0.015, "step": 26080 }, { "grad_norm": 0.1836560219526291, "learning_rate": 8.832571226488903e-05, "loss": 0.0125, "step": 26090 }, { "grad_norm": 0.1514016091823578, "learning_rate": 8.831509113480634e-05, "loss": 0.0124, "step": 26100 }, { "grad_norm": 0.16397987306118011, "learning_rate": 8.83044658146429e-05, "loss": 0.0149, "step": 26110 }, { "grad_norm": 0.19256886839866638, "learning_rate": 8.829383630556067e-05, "loss": 0.0158, "step": 26120 }, { "grad_norm": 0.20490841567516327, "learning_rate": 8.828320260872207e-05, "loss": 0.0146, "step": 26130 }, { "grad_norm": 0.20316798985004425, "learning_rate": 8.827256472529e-05, "loss": 0.0148, "step": 26140 }, { "grad_norm": 0.19748811423778534, "learning_rate": 8.826192265642778e-05, "loss": 0.0149, "step": 26150 }, { "grad_norm": 0.16976407170295715, "learning_rate": 8.825127640329923e-05, "loss": 0.0151, "step": 26160 }, { "grad_norm": 0.16722974181175232, "learning_rate": 8.824062596706861e-05, "loss": 0.0151, "step": 26170 }, { "grad_norm": 0.14228473603725433, "learning_rate": 8.822997134890062e-05, "loss": 0.0129, "step": 26180 }, { "grad_norm": 0.15005740523338318, "learning_rate": 8.821931254996044e-05, "loss": 0.0127, "step": 26190 }, { "grad_norm": 0.1341763585805893, "learning_rate": 8.82086495714137e-05, "loss": 0.0139, "step": 26200 }, { "grad_norm": 0.22302795946598053, "learning_rate": 8.81979824144265e-05, "loss": 0.0123, "step": 26210 }, { "grad_norm": 0.1833508461713791, "learning_rate": 8.818731108016536e-05, "loss": 0.0136, "step": 26220 }, { "grad_norm": 0.24204157292842865, "learning_rate": 8.81766355697973e-05, "loss": 0.0152, "step": 26230 }, { "grad_norm": 0.17180116474628448, "learning_rate": 8.816595588448977e-05, "loss": 0.0137, "step": 26240 }, { "grad_norm": 0.14754708111286163, "learning_rate": 8.81552720254107e-05, "loss": 0.0119, "step": 26250 }, { "grad_norm": 0.1728232204914093, "learning_rate": 8.814458399372842e-05, "loss": 0.0167, "step": 26260 }, { "grad_norm": 0.12322212010622025, "learning_rate": 8.813389179061181e-05, "loss": 0.0129, "step": 26270 }, { "grad_norm": 0.1868961751461029, "learning_rate": 8.812319541723012e-05, "loss": 0.014, "step": 26280 }, { "grad_norm": 0.21988466382026672, "learning_rate": 8.811249487475309e-05, "loss": 0.0129, "step": 26290 }, { "grad_norm": 0.11636942625045776, "learning_rate": 8.810179016435092e-05, "loss": 0.0119, "step": 26300 }, { "grad_norm": 0.17232176661491394, "learning_rate": 8.809108128719428e-05, "loss": 0.0177, "step": 26310 }, { "grad_norm": 0.18752874433994293, "learning_rate": 8.808036824445424e-05, "loss": 0.0142, "step": 26320 }, { "grad_norm": 0.12122940272092819, "learning_rate": 8.806965103730238e-05, "loss": 0.0132, "step": 26330 }, { "grad_norm": 0.14179709553718567, "learning_rate": 8.805892966691074e-05, "loss": 0.0157, "step": 26340 }, { "grad_norm": 0.12303551286458969, "learning_rate": 8.804820413445175e-05, "loss": 0.0128, "step": 26350 }, { "grad_norm": 0.14592838287353516, "learning_rate": 8.803747444109837e-05, "loss": 0.0116, "step": 26360 }, { "grad_norm": 0.16238832473754883, "learning_rate": 8.802674058802399e-05, "loss": 0.0124, "step": 26370 }, { "grad_norm": 0.2000068575143814, "learning_rate": 8.801600257640241e-05, "loss": 0.0132, "step": 26380 }, { "grad_norm": 0.1805112510919571, "learning_rate": 8.800526040740795e-05, "loss": 0.0158, "step": 26390 }, { "grad_norm": 0.16385161876678467, "learning_rate": 8.799451408221535e-05, "loss": 0.0157, "step": 26400 }, { "grad_norm": 0.13099201023578644, "learning_rate": 8.798376360199982e-05, "loss": 0.012, "step": 26410 }, { "grad_norm": 0.15675801038742065, "learning_rate": 8.797300896793701e-05, "loss": 0.0139, "step": 26420 }, { "grad_norm": 0.1781521588563919, "learning_rate": 8.796225018120302e-05, "loss": 0.0134, "step": 26430 }, { "grad_norm": 0.3231700658798218, "learning_rate": 8.795148724297444e-05, "loss": 0.0156, "step": 26440 }, { "grad_norm": 0.11732740700244904, "learning_rate": 8.794072015442825e-05, "loss": 0.0133, "step": 26450 }, { "grad_norm": 0.16854816675186157, "learning_rate": 8.792994891674198e-05, "loss": 0.0129, "step": 26460 }, { "grad_norm": 0.20890137553215027, "learning_rate": 8.79191735310935e-05, "loss": 0.0152, "step": 26470 }, { "grad_norm": 0.1859498769044876, "learning_rate": 8.790839399866122e-05, "loss": 0.0146, "step": 26480 }, { "grad_norm": 0.1624213308095932, "learning_rate": 8.789761032062397e-05, "loss": 0.0143, "step": 26490 }, { "grad_norm": 0.19413645565509796, "learning_rate": 8.788682249816103e-05, "loss": 0.016, "step": 26500 }, { "grad_norm": 0.18683944642543793, "learning_rate": 8.787603053245215e-05, "loss": 0.0179, "step": 26510 }, { "grad_norm": 0.18133610486984253, "learning_rate": 8.78652344246775e-05, "loss": 0.0183, "step": 26520 }, { "grad_norm": 0.23820045590400696, "learning_rate": 8.785443417601776e-05, "loss": 0.0238, "step": 26530 }, { "grad_norm": 0.2157655656337738, "learning_rate": 8.784362978765401e-05, "loss": 0.0213, "step": 26540 }, { "grad_norm": 0.2470443844795227, "learning_rate": 8.783282126076779e-05, "loss": 0.0184, "step": 26550 }, { "grad_norm": 0.19009967148303986, "learning_rate": 8.782200859654112e-05, "loss": 0.0159, "step": 26560 }, { "grad_norm": 0.1266106516122818, "learning_rate": 8.781119179615646e-05, "loss": 0.0183, "step": 26570 }, { "grad_norm": 0.1742323935031891, "learning_rate": 8.780037086079674e-05, "loss": 0.0195, "step": 26580 }, { "grad_norm": 0.14387035369873047, "learning_rate": 8.778954579164527e-05, "loss": 0.0154, "step": 26590 }, { "grad_norm": 0.18551096320152283, "learning_rate": 8.777871658988588e-05, "loss": 0.0192, "step": 26600 }, { "grad_norm": 0.14861774444580078, "learning_rate": 8.776788325670285e-05, "loss": 0.0152, "step": 26610 }, { "grad_norm": 0.5492379665374756, "learning_rate": 8.775704579328089e-05, "loss": 0.0163, "step": 26620 }, { "grad_norm": 0.17954064905643463, "learning_rate": 8.774620420080517e-05, "loss": 0.0251, "step": 26630 }, { "grad_norm": 0.2396083027124405, "learning_rate": 8.773535848046131e-05, "loss": 0.0201, "step": 26640 }, { "grad_norm": 0.18189400434494019, "learning_rate": 8.772450863343538e-05, "loss": 0.0186, "step": 26650 }, { "grad_norm": 0.2620042562484741, "learning_rate": 8.77136546609139e-05, "loss": 0.0165, "step": 26660 }, { "grad_norm": 0.1429552584886551, "learning_rate": 8.770279656408385e-05, "loss": 0.0154, "step": 26670 }, { "grad_norm": 0.16702714562416077, "learning_rate": 8.769193434413265e-05, "loss": 0.0143, "step": 26680 }, { "grad_norm": 0.1755668818950653, "learning_rate": 8.76810680022482e-05, "loss": 0.0161, "step": 26690 }, { "grad_norm": 0.16496317088603973, "learning_rate": 8.767019753961878e-05, "loss": 0.0142, "step": 26700 }, { "grad_norm": 0.13783229887485504, "learning_rate": 8.765932295743321e-05, "loss": 0.0141, "step": 26710 }, { "grad_norm": 0.19714590907096863, "learning_rate": 8.764844425688068e-05, "loss": 0.0136, "step": 26720 }, { "grad_norm": 0.22539617121219635, "learning_rate": 8.763756143915092e-05, "loss": 0.0172, "step": 26730 }, { "grad_norm": 0.15034396946430206, "learning_rate": 8.7626674505434e-05, "loss": 0.0138, "step": 26740 }, { "grad_norm": 0.1387532353401184, "learning_rate": 8.761578345692053e-05, "loss": 0.0135, "step": 26750 }, { "grad_norm": 0.14573301374912262, "learning_rate": 8.760488829480156e-05, "loss": 0.0151, "step": 26760 }, { "grad_norm": 0.10778256505727768, "learning_rate": 8.759398902026854e-05, "loss": 0.0155, "step": 26770 }, { "grad_norm": 0.12577269971370697, "learning_rate": 8.758308563451339e-05, "loss": 0.0126, "step": 26780 }, { "grad_norm": 0.15108303725719452, "learning_rate": 8.75721781387285e-05, "loss": 0.0176, "step": 26790 }, { "grad_norm": 0.1301661878824234, "learning_rate": 8.75612665341067e-05, "loss": 0.0152, "step": 26800 }, { "grad_norm": 0.20260779559612274, "learning_rate": 8.755035082184126e-05, "loss": 0.0128, "step": 26810 }, { "grad_norm": 0.1430332362651825, "learning_rate": 8.753943100312592e-05, "loss": 0.0117, "step": 26820 }, { "grad_norm": 0.1735258400440216, "learning_rate": 8.752850707915484e-05, "loss": 0.0124, "step": 26830 }, { "grad_norm": 0.15361101925373077, "learning_rate": 8.751757905112264e-05, "loss": 0.0114, "step": 26840 }, { "grad_norm": 0.2120923101902008, "learning_rate": 8.75066469202244e-05, "loss": 0.0134, "step": 26850 }, { "grad_norm": 0.1524815708398819, "learning_rate": 8.749571068765567e-05, "loss": 0.0141, "step": 26860 }, { "grad_norm": 0.18183289468288422, "learning_rate": 8.748477035461238e-05, "loss": 0.0144, "step": 26870 }, { "grad_norm": 0.1598903089761734, "learning_rate": 8.747382592229095e-05, "loss": 0.0146, "step": 26880 }, { "grad_norm": 0.1865309178829193, "learning_rate": 8.746287739188828e-05, "loss": 0.013, "step": 26890 }, { "grad_norm": 0.14521116018295288, "learning_rate": 8.745192476460165e-05, "loss": 0.0137, "step": 26900 }, { "grad_norm": 0.1254732757806778, "learning_rate": 8.744096804162882e-05, "loss": 0.015, "step": 26910 }, { "grad_norm": 0.18387557566165924, "learning_rate": 8.743000722416804e-05, "loss": 0.0146, "step": 26920 }, { "grad_norm": 0.16627806425094604, "learning_rate": 8.741904231341793e-05, "loss": 0.0155, "step": 26930 }, { "grad_norm": 0.1699886918067932, "learning_rate": 8.740807331057762e-05, "loss": 0.0142, "step": 26940 }, { "grad_norm": 0.3145321309566498, "learning_rate": 8.739710021684667e-05, "loss": 0.0144, "step": 26950 }, { "grad_norm": 0.18291877210140228, "learning_rate": 8.738612303342503e-05, "loss": 0.0135, "step": 26960 }, { "grad_norm": 0.15391644835472107, "learning_rate": 8.73751417615132e-05, "loss": 0.0132, "step": 26970 }, { "grad_norm": 0.2415708750486374, "learning_rate": 8.736415640231208e-05, "loss": 0.0164, "step": 26980 }, { "grad_norm": 0.18148095905780792, "learning_rate": 8.735316695702297e-05, "loss": 0.0158, "step": 26990 }, { "grad_norm": 0.17670036852359772, "learning_rate": 8.734217342684769e-05, "loss": 0.0156, "step": 27000 }, { "grad_norm": 0.15604887902736664, "learning_rate": 8.733117581298847e-05, "loss": 0.0136, "step": 27010 }, { "grad_norm": 0.1444244235754013, "learning_rate": 8.732017411664796e-05, "loss": 0.015, "step": 27020 }, { "grad_norm": 0.19069288671016693, "learning_rate": 8.730916833902936e-05, "loss": 0.0146, "step": 27030 }, { "grad_norm": 0.17733822762966156, "learning_rate": 8.729815848133618e-05, "loss": 0.0134, "step": 27040 }, { "grad_norm": 0.15842729806900024, "learning_rate": 8.728714454477247e-05, "loss": 0.0136, "step": 27050 }, { "grad_norm": 0.12495780736207962, "learning_rate": 8.727612653054269e-05, "loss": 0.0155, "step": 27060 }, { "grad_norm": 0.15520064532756805, "learning_rate": 8.726510443985176e-05, "loss": 0.0167, "step": 27070 }, { "grad_norm": 0.16746138036251068, "learning_rate": 8.725407827390503e-05, "loss": 0.0154, "step": 27080 }, { "grad_norm": 0.15860357880592346, "learning_rate": 8.724304803390833e-05, "loss": 0.0148, "step": 27090 }, { "grad_norm": 0.28173205256462097, "learning_rate": 8.723201372106788e-05, "loss": 0.0144, "step": 27100 }, { "grad_norm": 0.18888592720031738, "learning_rate": 8.722097533659038e-05, "loss": 0.0144, "step": 27110 }, { "grad_norm": 0.16786423325538635, "learning_rate": 8.720993288168299e-05, "loss": 0.0157, "step": 27120 }, { "grad_norm": 0.14802730083465576, "learning_rate": 8.719888635755327e-05, "loss": 0.0155, "step": 27130 }, { "grad_norm": 0.1707679033279419, "learning_rate": 8.718783576540928e-05, "loss": 0.0131, "step": 27140 }, { "grad_norm": 0.15003399550914764, "learning_rate": 8.717678110645948e-05, "loss": 0.0128, "step": 27150 }, { "grad_norm": 0.151815727353096, "learning_rate": 8.716572238191279e-05, "loss": 0.013, "step": 27160 }, { "grad_norm": 0.16563552618026733, "learning_rate": 8.715465959297857e-05, "loss": 0.0137, "step": 27170 }, { "grad_norm": 0.17668750882148743, "learning_rate": 8.714359274086665e-05, "loss": 0.015, "step": 27180 }, { "grad_norm": 0.1696666181087494, "learning_rate": 8.713252182678726e-05, "loss": 0.0139, "step": 27190 }, { "grad_norm": 0.15330009162425995, "learning_rate": 8.712144685195112e-05, "loss": 0.0141, "step": 27200 }, { "grad_norm": 0.16689835488796234, "learning_rate": 8.711036781756936e-05, "loss": 0.0143, "step": 27210 }, { "grad_norm": 0.13306541740894318, "learning_rate": 8.709928472485357e-05, "loss": 0.0137, "step": 27220 }, { "grad_norm": 0.17493611574172974, "learning_rate": 8.708819757501579e-05, "loss": 0.0133, "step": 27230 }, { "grad_norm": 0.1566699594259262, "learning_rate": 8.707710636926846e-05, "loss": 0.0131, "step": 27240 }, { "grad_norm": 0.14880502223968506, "learning_rate": 8.706601110882455e-05, "loss": 0.0157, "step": 27250 }, { "grad_norm": 0.15814851224422455, "learning_rate": 8.705491179489738e-05, "loss": 0.0134, "step": 27260 }, { "grad_norm": 0.1350773125886917, "learning_rate": 8.704380842870077e-05, "loss": 0.0122, "step": 27270 }, { "grad_norm": 0.2841539680957794, "learning_rate": 8.703270101144895e-05, "loss": 0.0132, "step": 27280 }, { "grad_norm": 0.17575357854366302, "learning_rate": 8.702158954435664e-05, "loss": 0.0127, "step": 27290 }, { "grad_norm": 0.1441464126110077, "learning_rate": 8.701047402863896e-05, "loss": 0.0147, "step": 27300 }, { "grad_norm": 0.22432826459407806, "learning_rate": 8.699935446551148e-05, "loss": 0.0125, "step": 27310 }, { "grad_norm": 0.16315817832946777, "learning_rate": 8.698823085619022e-05, "loss": 0.0131, "step": 27320 }, { "grad_norm": 0.1319894790649414, "learning_rate": 8.697710320189166e-05, "loss": 0.0135, "step": 27330 }, { "grad_norm": 0.1186215803027153, "learning_rate": 8.696597150383268e-05, "loss": 0.0127, "step": 27340 }, { "grad_norm": 0.13460366427898407, "learning_rate": 8.695483576323063e-05, "loss": 0.012, "step": 27350 }, { "grad_norm": 0.14772248268127441, "learning_rate": 8.69436959813033e-05, "loss": 0.0125, "step": 27360 }, { "grad_norm": 0.14976291358470917, "learning_rate": 8.693255215926892e-05, "loss": 0.0119, "step": 27370 }, { "grad_norm": 0.1893281489610672, "learning_rate": 8.692140429834617e-05, "loss": 0.0131, "step": 27380 }, { "grad_norm": 0.14257991313934326, "learning_rate": 8.691025239975415e-05, "loss": 0.0138, "step": 27390 }, { "grad_norm": 0.1676955372095108, "learning_rate": 8.689909646471243e-05, "loss": 0.0124, "step": 27400 }, { "grad_norm": 0.1475636065006256, "learning_rate": 8.688793649444099e-05, "loss": 0.0125, "step": 27410 }, { "grad_norm": 0.17478252947330475, "learning_rate": 8.687677249016029e-05, "loss": 0.012, "step": 27420 }, { "grad_norm": 0.1433403342962265, "learning_rate": 8.686560445309118e-05, "loss": 0.0142, "step": 27430 }, { "grad_norm": 0.1925298422574997, "learning_rate": 8.685443238445499e-05, "loss": 0.0147, "step": 27440 }, { "grad_norm": 0.15507732331752777, "learning_rate": 8.68432562854735e-05, "loss": 0.0162, "step": 27450 }, { "grad_norm": 0.19090650975704193, "learning_rate": 8.683207615736887e-05, "loss": 0.0157, "step": 27460 }, { "grad_norm": 0.14898347854614258, "learning_rate": 8.682089200136379e-05, "loss": 0.0137, "step": 27470 }, { "grad_norm": 0.1808159500360489, "learning_rate": 8.680970381868132e-05, "loss": 0.0138, "step": 27480 }, { "grad_norm": 0.15645021200180054, "learning_rate": 8.679851161054498e-05, "loss": 0.0149, "step": 27490 }, { "grad_norm": 0.21642586588859558, "learning_rate": 8.678731537817873e-05, "loss": 0.0129, "step": 27500 }, { "grad_norm": 0.14294524490833282, "learning_rate": 8.677611512280697e-05, "loss": 0.0136, "step": 27510 }, { "grad_norm": 0.19001899659633636, "learning_rate": 8.676491084565457e-05, "loss": 0.0168, "step": 27520 }, { "grad_norm": 0.14975331723690033, "learning_rate": 8.675370254794678e-05, "loss": 0.0136, "step": 27530 }, { "grad_norm": 0.17463050782680511, "learning_rate": 8.674249023090935e-05, "loss": 0.0159, "step": 27540 }, { "grad_norm": 0.19681817293167114, "learning_rate": 8.673127389576843e-05, "loss": 0.0143, "step": 27550 }, { "grad_norm": 0.18209584057331085, "learning_rate": 8.67200535437506e-05, "loss": 0.0178, "step": 27560 }, { "grad_norm": 0.14973926544189453, "learning_rate": 8.670882917608296e-05, "loss": 0.0143, "step": 27570 }, { "grad_norm": 0.12403535842895508, "learning_rate": 8.669760079399292e-05, "loss": 0.0117, "step": 27580 }, { "grad_norm": 0.16048577427864075, "learning_rate": 8.668636839870845e-05, "loss": 0.0123, "step": 27590 }, { "grad_norm": 0.1687929332256317, "learning_rate": 8.667513199145789e-05, "loss": 0.0135, "step": 27600 }, { "grad_norm": 0.1809857040643692, "learning_rate": 8.666389157347002e-05, "loss": 0.0125, "step": 27610 }, { "grad_norm": 0.23706144094467163, "learning_rate": 8.66526471459741e-05, "loss": 0.0145, "step": 27620 }, { "grad_norm": 0.2250324934720993, "learning_rate": 8.66413987101998e-05, "loss": 0.0138, "step": 27630 }, { "grad_norm": 0.13431277871131897, "learning_rate": 8.663014626737723e-05, "loss": 0.014, "step": 27640 }, { "grad_norm": 0.14226512610912323, "learning_rate": 8.661888981873691e-05, "loss": 0.0129, "step": 27650 }, { "grad_norm": 0.150750070810318, "learning_rate": 8.660762936550988e-05, "loss": 0.0125, "step": 27660 }, { "grad_norm": 0.14320579171180725, "learning_rate": 8.659636490892753e-05, "loss": 0.0116, "step": 27670 }, { "grad_norm": 0.1672540307044983, "learning_rate": 8.658509645022174e-05, "loss": 0.0123, "step": 27680 }, { "grad_norm": 0.20210884511470795, "learning_rate": 8.657382399062481e-05, "loss": 0.0152, "step": 27690 }, { "grad_norm": 0.16516059637069702, "learning_rate": 8.656254753136946e-05, "loss": 0.0132, "step": 27700 }, { "grad_norm": 0.16253624856472015, "learning_rate": 8.655126707368891e-05, "loss": 0.0156, "step": 27710 }, { "grad_norm": 0.16278108954429626, "learning_rate": 8.653998261881672e-05, "loss": 0.0155, "step": 27720 }, { "grad_norm": 0.14735659956932068, "learning_rate": 8.652869416798699e-05, "loss": 0.0132, "step": 27730 }, { "grad_norm": 0.18020853400230408, "learning_rate": 8.651740172243417e-05, "loss": 0.012, "step": 27740 }, { "grad_norm": 0.14891494810581207, "learning_rate": 8.65061052833932e-05, "loss": 0.014, "step": 27750 }, { "grad_norm": 0.18085671961307526, "learning_rate": 8.649480485209945e-05, "loss": 0.0131, "step": 27760 }, { "grad_norm": 0.15392711758613586, "learning_rate": 8.64835004297887e-05, "loss": 0.0126, "step": 27770 }, { "grad_norm": 0.15925581753253937, "learning_rate": 8.64721920176972e-05, "loss": 0.0146, "step": 27780 }, { "grad_norm": 0.17254731059074402, "learning_rate": 8.646087961706164e-05, "loss": 0.0133, "step": 27790 }, { "grad_norm": 0.18677696585655212, "learning_rate": 8.644956322911908e-05, "loss": 0.0141, "step": 27800 }, { "grad_norm": 0.16475193202495575, "learning_rate": 8.643824285510709e-05, "loss": 0.0115, "step": 27810 }, { "grad_norm": 0.1696990430355072, "learning_rate": 8.642691849626364e-05, "loss": 0.0133, "step": 27820 }, { "grad_norm": 0.15827561914920807, "learning_rate": 8.641559015382717e-05, "loss": 0.0125, "step": 27830 }, { "grad_norm": 0.14624549448490143, "learning_rate": 8.640425782903649e-05, "loss": 0.0109, "step": 27840 }, { "grad_norm": 0.134428471326828, "learning_rate": 8.639292152313091e-05, "loss": 0.0116, "step": 27850 }, { "grad_norm": 0.20224210619926453, "learning_rate": 8.638158123735015e-05, "loss": 0.0118, "step": 27860 }, { "grad_norm": 0.12731145322322845, "learning_rate": 8.637023697293436e-05, "loss": 0.0126, "step": 27870 }, { "grad_norm": 0.14388194680213928, "learning_rate": 8.635888873112414e-05, "loss": 0.0142, "step": 27880 }, { "grad_norm": 0.14709746837615967, "learning_rate": 8.634753651316052e-05, "loss": 0.0127, "step": 27890 }, { "grad_norm": 0.17321869730949402, "learning_rate": 8.633618032028496e-05, "loss": 0.014, "step": 27900 }, { "grad_norm": 0.17485997080802917, "learning_rate": 8.632482015373934e-05, "loss": 0.0156, "step": 27910 }, { "grad_norm": 0.18500344455242157, "learning_rate": 8.6313456014766e-05, "loss": 0.0154, "step": 27920 }, { "grad_norm": 0.13401436805725098, "learning_rate": 8.630208790460771e-05, "loss": 0.0115, "step": 27930 }, { "grad_norm": 0.23553155362606049, "learning_rate": 8.629071582450768e-05, "loss": 0.0144, "step": 27940 }, { "grad_norm": 0.11105620115995407, "learning_rate": 8.62793397757095e-05, "loss": 0.0113, "step": 27950 }, { "grad_norm": 0.18637864291667938, "learning_rate": 8.626795975945729e-05, "loss": 0.0133, "step": 27960 }, { "grad_norm": 0.11735573410987854, "learning_rate": 8.625657577699551e-05, "loss": 0.0119, "step": 27970 }, { "grad_norm": 0.12070386856794357, "learning_rate": 8.624518782956914e-05, "loss": 0.0115, "step": 27980 }, { "grad_norm": 0.1922685205936432, "learning_rate": 8.62337959184235e-05, "loss": 0.0148, "step": 27990 }, { "grad_norm": 0.13726183772087097, "learning_rate": 8.622240004480441e-05, "loss": 0.0139, "step": 28000 }, { "grad_norm": 0.17814424633979797, "learning_rate": 8.621100020995814e-05, "loss": 0.0141, "step": 28010 }, { "grad_norm": 0.1318991780281067, "learning_rate": 8.619959641513132e-05, "loss": 0.0139, "step": 28020 }, { "grad_norm": 0.15275602042675018, "learning_rate": 8.618818866157105e-05, "loss": 0.0116, "step": 28030 }, { "grad_norm": 0.1673671305179596, "learning_rate": 8.617677695052487e-05, "loss": 0.0139, "step": 28040 }, { "grad_norm": 0.15358924865722656, "learning_rate": 8.616536128324078e-05, "loss": 0.0126, "step": 28050 }, { "grad_norm": 0.13301682472229004, "learning_rate": 8.615394166096712e-05, "loss": 0.0157, "step": 28060 }, { "grad_norm": 0.18360206484794617, "learning_rate": 8.614251808495279e-05, "loss": 0.015, "step": 28070 }, { "grad_norm": 0.18988189101219177, "learning_rate": 8.6131090556447e-05, "loss": 0.0165, "step": 28080 }, { "grad_norm": 0.19685247540473938, "learning_rate": 8.611965907669947e-05, "loss": 0.015, "step": 28090 }, { "grad_norm": 0.12149042636156082, "learning_rate": 8.610822364696034e-05, "loss": 0.0132, "step": 28100 }, { "grad_norm": 0.13201682269573212, "learning_rate": 8.609678426848015e-05, "loss": 0.0151, "step": 28110 }, { "grad_norm": 0.19781428575515747, "learning_rate": 8.60853409425099e-05, "loss": 0.0133, "step": 28120 }, { "grad_norm": 0.12043266743421555, "learning_rate": 8.607389367030104e-05, "loss": 0.0142, "step": 28130 }, { "grad_norm": 0.2262231558561325, "learning_rate": 8.606244245310538e-05, "loss": 0.0147, "step": 28140 }, { "grad_norm": 0.18359491229057312, "learning_rate": 8.605098729217525e-05, "loss": 0.0131, "step": 28150 }, { "grad_norm": 0.12566286325454712, "learning_rate": 8.603952818876335e-05, "loss": 0.0121, "step": 28160 }, { "grad_norm": 0.1924346387386322, "learning_rate": 8.602806514412281e-05, "loss": 0.013, "step": 28170 }, { "grad_norm": 0.1424052119255066, "learning_rate": 8.601659815950726e-05, "loss": 0.0141, "step": 28180 }, { "grad_norm": 0.1579381376504898, "learning_rate": 8.600512723617067e-05, "loss": 0.0132, "step": 28190 }, { "grad_norm": 0.18124617636203766, "learning_rate": 8.59936523753675e-05, "loss": 0.011, "step": 28200 }, { "grad_norm": 0.12558405101299286, "learning_rate": 8.598217357835264e-05, "loss": 0.0123, "step": 28210 }, { "grad_norm": 0.11609473079442978, "learning_rate": 8.597069084638135e-05, "loss": 0.0112, "step": 28220 }, { "grad_norm": 0.13079029321670532, "learning_rate": 8.595920418070939e-05, "loss": 0.0108, "step": 28230 }, { "grad_norm": 0.17218275368213654, "learning_rate": 8.594771358259295e-05, "loss": 0.0129, "step": 28240 }, { "grad_norm": 0.18806016445159912, "learning_rate": 8.593621905328858e-05, "loss": 0.0153, "step": 28250 }, { "grad_norm": 0.18667477369308472, "learning_rate": 8.592472059405333e-05, "loss": 0.0154, "step": 28260 }, { "grad_norm": 0.15124547481536865, "learning_rate": 8.591321820614464e-05, "loss": 0.0135, "step": 28270 }, { "grad_norm": 0.1267842948436737, "learning_rate": 8.590171189082041e-05, "loss": 0.0106, "step": 28280 }, { "grad_norm": 0.13195455074310303, "learning_rate": 8.589020164933894e-05, "loss": 0.0126, "step": 28290 }, { "grad_norm": 0.1324000060558319, "learning_rate": 8.587868748295898e-05, "loss": 0.0127, "step": 28300 }, { "grad_norm": 0.14167585968971252, "learning_rate": 8.586716939293971e-05, "loss": 0.0135, "step": 28310 }, { "grad_norm": 0.1495344042778015, "learning_rate": 8.58556473805407e-05, "loss": 0.0127, "step": 28320 }, { "grad_norm": 0.17245355248451233, "learning_rate": 8.584412144702202e-05, "loss": 0.0114, "step": 28330 }, { "grad_norm": 0.22281454503536224, "learning_rate": 8.58325915936441e-05, "loss": 0.0131, "step": 28340 }, { "grad_norm": 0.12137514352798462, "learning_rate": 8.582105782166783e-05, "loss": 0.0116, "step": 28350 }, { "grad_norm": 0.14655359089374542, "learning_rate": 8.580952013235455e-05, "loss": 0.014, "step": 28360 }, { "grad_norm": 0.1385953724384308, "learning_rate": 8.579797852696596e-05, "loss": 0.0122, "step": 28370 }, { "grad_norm": 0.17409314215183258, "learning_rate": 8.578643300676428e-05, "loss": 0.0127, "step": 28380 }, { "grad_norm": 0.23926077783107758, "learning_rate": 8.577488357301209e-05, "loss": 0.0149, "step": 28390 }, { "grad_norm": 0.1637771874666214, "learning_rate": 8.576333022697242e-05, "loss": 0.012, "step": 28400 }, { "grad_norm": 0.15380516648292542, "learning_rate": 8.575177296990873e-05, "loss": 0.0123, "step": 28410 }, { "grad_norm": 0.16937613487243652, "learning_rate": 8.574021180308489e-05, "loss": 0.0112, "step": 28420 }, { "grad_norm": 0.12101522088050842, "learning_rate": 8.572864672776523e-05, "loss": 0.0123, "step": 28430 }, { "grad_norm": 0.15246492624282837, "learning_rate": 8.571707774521447e-05, "loss": 0.0123, "step": 28440 }, { "grad_norm": 0.17406323552131653, "learning_rate": 8.57055048566978e-05, "loss": 0.0123, "step": 28450 }, { "grad_norm": 0.15927386283874512, "learning_rate": 8.569392806348078e-05, "loss": 0.0118, "step": 28460 }, { "grad_norm": 0.14251770079135895, "learning_rate": 8.568234736682947e-05, "loss": 0.0128, "step": 28470 }, { "grad_norm": 0.17030173540115356, "learning_rate": 8.567076276801029e-05, "loss": 0.0127, "step": 28480 }, { "grad_norm": 0.13707029819488525, "learning_rate": 8.565917426829013e-05, "loss": 0.0123, "step": 28490 }, { "grad_norm": 0.1840977966785431, "learning_rate": 8.564758186893628e-05, "loss": 0.0143, "step": 28500 }, { "grad_norm": 0.11694247275590897, "learning_rate": 8.563598557121649e-05, "loss": 0.0118, "step": 28510 }, { "grad_norm": 0.15134891867637634, "learning_rate": 8.562438537639888e-05, "loss": 0.0141, "step": 28520 }, { "grad_norm": 0.2030319720506668, "learning_rate": 8.561278128575206e-05, "loss": 0.0122, "step": 28530 }, { "grad_norm": 0.20798452198505402, "learning_rate": 8.5601173300545e-05, "loss": 0.0136, "step": 28540 }, { "grad_norm": 0.19842205941677094, "learning_rate": 8.558956142204717e-05, "loss": 0.0121, "step": 28550 }, { "grad_norm": 0.17803144454956055, "learning_rate": 8.55779456515284e-05, "loss": 0.0151, "step": 28560 }, { "grad_norm": 0.22453637421131134, "learning_rate": 8.556632599025898e-05, "loss": 0.0157, "step": 28570 }, { "grad_norm": 0.22419705986976624, "learning_rate": 8.555470243950964e-05, "loss": 0.0168, "step": 28580 }, { "grad_norm": 0.1640215367078781, "learning_rate": 8.554307500055148e-05, "loss": 0.0134, "step": 28590 }, { "grad_norm": 0.13371628522872925, "learning_rate": 8.553144367465609e-05, "loss": 0.0141, "step": 28600 }, { "grad_norm": 0.12464933842420578, "learning_rate": 8.551980846309544e-05, "loss": 0.0106, "step": 28610 }, { "grad_norm": 0.1392587274312973, "learning_rate": 8.550816936714193e-05, "loss": 0.0113, "step": 28620 }, { "grad_norm": 0.16825175285339355, "learning_rate": 8.549652638806841e-05, "loss": 0.0139, "step": 28630 }, { "grad_norm": 0.22818684577941895, "learning_rate": 8.548487952714812e-05, "loss": 0.0133, "step": 28640 }, { "grad_norm": 0.132620707154274, "learning_rate": 8.547322878565478e-05, "loss": 0.013, "step": 28650 }, { "grad_norm": 0.1671474426984787, "learning_rate": 8.546157416486245e-05, "loss": 0.0122, "step": 28660 }, { "grad_norm": 0.1503165364265442, "learning_rate": 8.54499156660457e-05, "loss": 0.0157, "step": 28670 }, { "grad_norm": 0.13155457377433777, "learning_rate": 8.543825329047947e-05, "loss": 0.0114, "step": 28680 }, { "grad_norm": 0.14215175807476044, "learning_rate": 8.542658703943913e-05, "loss": 0.0128, "step": 28690 }, { "grad_norm": 0.13939189910888672, "learning_rate": 8.541491691420051e-05, "loss": 0.0117, "step": 28700 }, { "grad_norm": 0.16983924806118011, "learning_rate": 8.54032429160398e-05, "loss": 0.0106, "step": 28710 }, { "grad_norm": 0.17057552933692932, "learning_rate": 8.539156504623369e-05, "loss": 0.0127, "step": 28720 }, { "grad_norm": 0.19225935637950897, "learning_rate": 8.537988330605923e-05, "loss": 0.0116, "step": 28730 }, { "grad_norm": 0.16340236365795135, "learning_rate": 8.536819769679393e-05, "loss": 0.0126, "step": 28740 }, { "grad_norm": 0.15799210965633392, "learning_rate": 8.53565082197157e-05, "loss": 0.0142, "step": 28750 }, { "grad_norm": 0.16338905692100525, "learning_rate": 8.534481487610289e-05, "loss": 0.0153, "step": 28760 }, { "grad_norm": 0.177665114402771, "learning_rate": 8.533311766723428e-05, "loss": 0.0115, "step": 28770 }, { "grad_norm": 0.19193854928016663, "learning_rate": 8.532141659438901e-05, "loss": 0.0132, "step": 28780 }, { "grad_norm": 0.14941850304603577, "learning_rate": 8.530971165884675e-05, "loss": 0.0127, "step": 28790 }, { "grad_norm": 0.1661720722913742, "learning_rate": 8.529800286188752e-05, "loss": 0.0129, "step": 28800 }, { "grad_norm": 0.13504604995250702, "learning_rate": 8.528629020479175e-05, "loss": 0.0155, "step": 28810 }, { "grad_norm": 0.15011493861675262, "learning_rate": 8.527457368884033e-05, "loss": 0.0164, "step": 28820 }, { "grad_norm": 0.20429985225200653, "learning_rate": 8.526285331531458e-05, "loss": 0.0144, "step": 28830 }, { "grad_norm": 0.18378804624080658, "learning_rate": 8.525112908549621e-05, "loss": 0.0144, "step": 28840 }, { "grad_norm": 0.1732630729675293, "learning_rate": 8.523940100066735e-05, "loss": 0.015, "step": 28850 }, { "grad_norm": 0.18302330374717712, "learning_rate": 8.52276690621106e-05, "loss": 0.0143, "step": 28860 }, { "grad_norm": 0.16589133441448212, "learning_rate": 8.521593327110889e-05, "loss": 0.0155, "step": 28870 }, { "grad_norm": 0.16692399978637695, "learning_rate": 8.520419362894569e-05, "loss": 0.0157, "step": 28880 }, { "grad_norm": 0.32635900378227234, "learning_rate": 8.51924501369048e-05, "loss": 0.0128, "step": 28890 }, { "grad_norm": 0.20900222659111023, "learning_rate": 8.518070279627047e-05, "loss": 0.0153, "step": 28900 }, { "grad_norm": 0.15020520985126495, "learning_rate": 8.516895160832737e-05, "loss": 0.017, "step": 28910 }, { "grad_norm": 0.1831512749195099, "learning_rate": 8.515719657436061e-05, "loss": 0.0154, "step": 28920 }, { "grad_norm": 0.16978897154331207, "learning_rate": 8.514543769565568e-05, "loss": 0.0153, "step": 28930 }, { "grad_norm": 0.20891831815242767, "learning_rate": 8.513367497349853e-05, "loss": 0.0142, "step": 28940 }, { "grad_norm": 0.14379824697971344, "learning_rate": 8.51219084091755e-05, "loss": 0.016, "step": 28950 }, { "grad_norm": 0.15163525938987732, "learning_rate": 8.511013800397338e-05, "loss": 0.0157, "step": 28960 }, { "grad_norm": 0.1816369593143463, "learning_rate": 8.509836375917937e-05, "loss": 0.0125, "step": 28970 }, { "grad_norm": 0.16180403530597687, "learning_rate": 8.508658567608104e-05, "loss": 0.0123, "step": 28980 }, { "grad_norm": 0.1524973213672638, "learning_rate": 8.507480375596647e-05, "loss": 0.0123, "step": 28990 }, { "grad_norm": 0.13457314670085907, "learning_rate": 8.506301800012408e-05, "loss": 0.013, "step": 29000 }, { "grad_norm": 0.1369864046573639, "learning_rate": 8.505122840984278e-05, "loss": 0.0129, "step": 29010 }, { "grad_norm": 0.17067955434322357, "learning_rate": 8.503943498641182e-05, "loss": 0.0128, "step": 29020 }, { "grad_norm": 0.16256661713123322, "learning_rate": 8.502763773112095e-05, "loss": 0.0123, "step": 29030 }, { "grad_norm": 0.14999167621135712, "learning_rate": 8.501583664526026e-05, "loss": 0.0129, "step": 29040 }, { "grad_norm": 0.14318066835403442, "learning_rate": 8.500403173012032e-05, "loss": 0.014, "step": 29050 }, { "grad_norm": 0.181613489985466, "learning_rate": 8.499222298699211e-05, "loss": 0.0133, "step": 29060 }, { "grad_norm": 0.16127416491508484, "learning_rate": 8.498041041716701e-05, "loss": 0.0124, "step": 29070 }, { "grad_norm": 0.13144956529140472, "learning_rate": 8.496859402193681e-05, "loss": 0.0131, "step": 29080 }, { "grad_norm": 0.09718598425388336, "learning_rate": 8.495677380259374e-05, "loss": 0.0109, "step": 29090 }, { "grad_norm": 0.13915804028511047, "learning_rate": 8.494494976043045e-05, "loss": 0.011, "step": 29100 }, { "grad_norm": 0.1638541966676712, "learning_rate": 8.493312189673998e-05, "loss": 0.0111, "step": 29110 }, { "grad_norm": 0.2243192493915558, "learning_rate": 8.492129021281584e-05, "loss": 0.0125, "step": 29120 }, { "grad_norm": 0.12760591506958008, "learning_rate": 8.490945470995188e-05, "loss": 0.0115, "step": 29130 }, { "grad_norm": 0.14743392169475555, "learning_rate": 8.489761538944247e-05, "loss": 0.0135, "step": 29140 }, { "grad_norm": 0.16830772161483765, "learning_rate": 8.48857722525823e-05, "loss": 0.0133, "step": 29150 }, { "grad_norm": 0.19362030923366547, "learning_rate": 8.487392530066652e-05, "loss": 0.0128, "step": 29160 }, { "grad_norm": 0.16843605041503906, "learning_rate": 8.486207453499069e-05, "loss": 0.0138, "step": 29170 }, { "grad_norm": 0.1435796320438385, "learning_rate": 8.485021995685082e-05, "loss": 0.0153, "step": 29180 }, { "grad_norm": 0.34304797649383545, "learning_rate": 8.483836156754328e-05, "loss": 0.0129, "step": 29190 }, { "grad_norm": 0.2069987952709198, "learning_rate": 8.482649936836491e-05, "loss": 0.0135, "step": 29200 }, { "grad_norm": 0.1354082077741623, "learning_rate": 8.481463336061293e-05, "loss": 0.0155, "step": 29210 }, { "grad_norm": 0.15525321662425995, "learning_rate": 8.480276354558496e-05, "loss": 0.0138, "step": 29220 }, { "grad_norm": 0.14375914633274078, "learning_rate": 8.479088992457913e-05, "loss": 0.013, "step": 29230 }, { "grad_norm": 0.14629273116588593, "learning_rate": 8.477901249889387e-05, "loss": 0.0131, "step": 29240 }, { "grad_norm": 0.10920692980289459, "learning_rate": 8.47671312698281e-05, "loss": 0.0107, "step": 29250 }, { "grad_norm": 0.12635871767997742, "learning_rate": 8.475524623868112e-05, "loss": 0.0128, "step": 29260 }, { "grad_norm": 0.11560235917568207, "learning_rate": 8.474335740675266e-05, "loss": 0.0127, "step": 29270 }, { "grad_norm": 0.16395650804042816, "learning_rate": 8.473146477534289e-05, "loss": 0.0131, "step": 29280 }, { "grad_norm": 0.1411684900522232, "learning_rate": 8.471956834575232e-05, "loss": 0.013, "step": 29290 }, { "grad_norm": 0.12489911168813705, "learning_rate": 8.470766811928197e-05, "loss": 0.0138, "step": 29300 }, { "grad_norm": 0.15850749611854553, "learning_rate": 8.469576409723323e-05, "loss": 0.0119, "step": 29310 }, { "grad_norm": 0.18267732858657837, "learning_rate": 8.468385628090788e-05, "loss": 0.015, "step": 29320 }, { "grad_norm": 0.15435177087783813, "learning_rate": 8.467194467160815e-05, "loss": 0.0128, "step": 29330 }, { "grad_norm": 0.15674437582492828, "learning_rate": 8.466002927063667e-05, "loss": 0.0146, "step": 29340 }, { "grad_norm": 0.12738460302352905, "learning_rate": 8.464811007929651e-05, "loss": 0.0153, "step": 29350 }, { "grad_norm": 0.14487527310848236, "learning_rate": 8.463618709889114e-05, "loss": 0.0124, "step": 29360 }, { "grad_norm": 0.21825405955314636, "learning_rate": 8.462426033072442e-05, "loss": 0.0134, "step": 29370 }, { "grad_norm": 0.16846932470798492, "learning_rate": 8.461232977610061e-05, "loss": 0.0117, "step": 29380 }, { "grad_norm": 0.15969346463680267, "learning_rate": 8.46003954363245e-05, "loss": 0.0131, "step": 29390 }, { "grad_norm": 0.13675865530967712, "learning_rate": 8.458845731270115e-05, "loss": 0.0112, "step": 29400 }, { "grad_norm": 0.15531137585639954, "learning_rate": 8.45765154065361e-05, "loss": 0.0122, "step": 29410 }, { "grad_norm": 0.2238418012857437, "learning_rate": 8.456456971913532e-05, "loss": 0.012, "step": 29420 }, { "grad_norm": 0.21651491522789001, "learning_rate": 8.455262025180517e-05, "loss": 0.0152, "step": 29430 }, { "grad_norm": 0.1414167284965515, "learning_rate": 8.454066700585242e-05, "loss": 0.0112, "step": 29440 }, { "grad_norm": 0.1759072095155716, "learning_rate": 8.452870998258423e-05, "loss": 0.0118, "step": 29450 }, { "grad_norm": 0.16935263574123383, "learning_rate": 8.451674918330825e-05, "loss": 0.0145, "step": 29460 }, { "grad_norm": 0.1353074312210083, "learning_rate": 8.450478460933246e-05, "loss": 0.0133, "step": 29470 }, { "grad_norm": 0.14936739206314087, "learning_rate": 8.449281626196532e-05, "loss": 0.0151, "step": 29480 }, { "grad_norm": 0.21359184384346008, "learning_rate": 8.448084414251564e-05, "loss": 0.0138, "step": 29490 }, { "grad_norm": 0.16342231631278992, "learning_rate": 8.446886825229271e-05, "loss": 0.0132, "step": 29500 }, { "grad_norm": 0.14000219106674194, "learning_rate": 8.445688859260615e-05, "loss": 0.0134, "step": 29510 }, { "grad_norm": 0.1724701076745987, "learning_rate": 8.444490516476606e-05, "loss": 0.0137, "step": 29520 }, { "grad_norm": 0.13305898010730743, "learning_rate": 8.443291797008293e-05, "loss": 0.0136, "step": 29530 }, { "grad_norm": 0.15158256888389587, "learning_rate": 8.442092700986765e-05, "loss": 0.0132, "step": 29540 }, { "grad_norm": 0.14151377975940704, "learning_rate": 8.440893228543156e-05, "loss": 0.0117, "step": 29550 }, { "grad_norm": 0.16831330955028534, "learning_rate": 8.439693379808638e-05, "loss": 0.015, "step": 29560 }, { "grad_norm": 0.17867861688137054, "learning_rate": 8.43849315491442e-05, "loss": 0.0136, "step": 29570 }, { "grad_norm": 0.16651926934719086, "learning_rate": 8.437292553991763e-05, "loss": 0.0159, "step": 29580 }, { "grad_norm": 0.14821653068065643, "learning_rate": 8.436091577171959e-05, "loss": 0.0138, "step": 29590 }, { "grad_norm": 0.16842854022979736, "learning_rate": 8.434890224586347e-05, "loss": 0.0124, "step": 29600 }, { "grad_norm": 0.16567377746105194, "learning_rate": 8.433688496366303e-05, "loss": 0.0133, "step": 29610 }, { "grad_norm": 0.1890573352575302, "learning_rate": 8.432486392643248e-05, "loss": 0.0142, "step": 29620 }, { "grad_norm": 0.13654296100139618, "learning_rate": 8.431283913548643e-05, "loss": 0.0152, "step": 29630 }, { "grad_norm": 0.16339534521102905, "learning_rate": 8.430081059213985e-05, "loss": 0.0147, "step": 29640 }, { "grad_norm": 0.16617585718631744, "learning_rate": 8.428877829770823e-05, "loss": 0.013, "step": 29650 }, { "grad_norm": 0.16260094940662384, "learning_rate": 8.427674225350735e-05, "loss": 0.0124, "step": 29660 }, { "grad_norm": 0.14441093802452087, "learning_rate": 8.426470246085347e-05, "loss": 0.0129, "step": 29670 }, { "grad_norm": 0.18765726685523987, "learning_rate": 8.425265892106324e-05, "loss": 0.0129, "step": 29680 }, { "grad_norm": 0.18774385750293732, "learning_rate": 8.424061163545374e-05, "loss": 0.0126, "step": 29690 }, { "grad_norm": 0.17323805391788483, "learning_rate": 8.422856060534243e-05, "loss": 0.014, "step": 29700 }, { "grad_norm": 0.3182329833507538, "learning_rate": 8.421650583204718e-05, "loss": 0.0137, "step": 29710 }, { "grad_norm": 0.17902757227420807, "learning_rate": 8.420444731688633e-05, "loss": 0.0123, "step": 29720 }, { "grad_norm": 0.17406302690505981, "learning_rate": 8.419238506117852e-05, "loss": 0.0126, "step": 29730 }, { "grad_norm": 0.1615445762872696, "learning_rate": 8.418031906624289e-05, "loss": 0.0125, "step": 29740 }, { "grad_norm": 0.17484687268733978, "learning_rate": 8.416824933339898e-05, "loss": 0.0128, "step": 29750 }, { "grad_norm": 0.1747833788394928, "learning_rate": 8.415617586396667e-05, "loss": 0.0138, "step": 29760 }, { "grad_norm": 0.2585110664367676, "learning_rate": 8.414409865926632e-05, "loss": 0.0151, "step": 29770 }, { "grad_norm": 0.15886589884757996, "learning_rate": 8.413201772061867e-05, "loss": 0.0166, "step": 29780 }, { "grad_norm": 0.17591941356658936, "learning_rate": 8.411993304934488e-05, "loss": 0.0131, "step": 29790 }, { "grad_norm": 0.15572869777679443, "learning_rate": 8.410784464676654e-05, "loss": 0.0151, "step": 29800 }, { "grad_norm": 0.20559628307819366, "learning_rate": 8.409575251420556e-05, "loss": 0.0154, "step": 29810 }, { "grad_norm": 0.1735704094171524, "learning_rate": 8.408365665298435e-05, "loss": 0.0129, "step": 29820 }, { "grad_norm": 0.14700978994369507, "learning_rate": 8.40715570644257e-05, "loss": 0.0133, "step": 29830 }, { "grad_norm": 0.13212448358535767, "learning_rate": 8.40594537498528e-05, "loss": 0.0132, "step": 29840 }, { "grad_norm": 0.17581957578659058, "learning_rate": 8.404734671058924e-05, "loss": 0.0147, "step": 29850 }, { "grad_norm": 0.1630561202764511, "learning_rate": 8.403523594795902e-05, "loss": 0.0125, "step": 29860 }, { "grad_norm": 0.1045127883553505, "learning_rate": 8.402312146328659e-05, "loss": 0.012, "step": 29870 }, { "grad_norm": 0.14453844726085663, "learning_rate": 8.401100325789675e-05, "loss": 0.0124, "step": 29880 }, { "grad_norm": 0.13114358484745026, "learning_rate": 8.399888133311472e-05, "loss": 0.0133, "step": 29890 }, { "grad_norm": 0.14429640769958496, "learning_rate": 8.398675569026613e-05, "loss": 0.0112, "step": 29900 }, { "grad_norm": 0.1527191549539566, "learning_rate": 8.397462633067705e-05, "loss": 0.0132, "step": 29910 }, { "grad_norm": 0.1488574743270874, "learning_rate": 8.396249325567392e-05, "loss": 0.013, "step": 29920 }, { "grad_norm": 0.16466951370239258, "learning_rate": 8.395035646658357e-05, "loss": 0.0139, "step": 29930 }, { "grad_norm": 0.2306169718503952, "learning_rate": 8.39382159647333e-05, "loss": 0.0146, "step": 29940 }, { "grad_norm": 0.15245473384857178, "learning_rate": 8.392607175145075e-05, "loss": 0.014, "step": 29950 }, { "grad_norm": 0.22287827730178833, "learning_rate": 8.3913923828064e-05, "loss": 0.0137, "step": 29960 }, { "grad_norm": 0.15969301760196686, "learning_rate": 8.390177219590152e-05, "loss": 0.0145, "step": 29970 }, { "grad_norm": 0.1612761914730072, "learning_rate": 8.388961685629222e-05, "loss": 0.0146, "step": 29980 }, { "grad_norm": 0.15750856697559357, "learning_rate": 8.387745781056536e-05, "loss": 0.0135, "step": 29990 }, { "grad_norm": 0.1343231499195099, "learning_rate": 8.386529506005065e-05, "loss": 0.0131, "step": 30000 }, { "grad_norm": 0.16116441786289215, "learning_rate": 8.38531286060782e-05, "loss": 0.0127, "step": 30010 }, { "grad_norm": 0.13903965055942535, "learning_rate": 8.384095844997849e-05, "loss": 0.0133, "step": 30020 }, { "grad_norm": 0.13499976694583893, "learning_rate": 8.382878459308245e-05, "loss": 0.0148, "step": 30030 }, { "grad_norm": 0.13645632565021515, "learning_rate": 8.381660703672138e-05, "loss": 0.0131, "step": 30040 }, { "grad_norm": 0.11443363130092621, "learning_rate": 8.380442578222702e-05, "loss": 0.0118, "step": 30050 }, { "grad_norm": 0.10619191825389862, "learning_rate": 8.379224083093146e-05, "loss": 0.0117, "step": 30060 }, { "grad_norm": 0.15934884548187256, "learning_rate": 8.378005218416727e-05, "loss": 0.0142, "step": 30070 }, { "grad_norm": 0.11457697302103043, "learning_rate": 8.376785984326735e-05, "loss": 0.0118, "step": 30080 }, { "grad_norm": 0.15261933207511902, "learning_rate": 8.375566380956506e-05, "loss": 0.0132, "step": 30090 }, { "grad_norm": 0.13176953792572021, "learning_rate": 8.374346408439411e-05, "loss": 0.0165, "step": 30100 }, { "grad_norm": 0.16677045822143555, "learning_rate": 8.373126066908868e-05, "loss": 0.0146, "step": 30110 }, { "grad_norm": 0.13838325440883636, "learning_rate": 8.371905356498326e-05, "loss": 0.0119, "step": 30120 }, { "grad_norm": 0.15027286112308502, "learning_rate": 8.370684277341288e-05, "loss": 0.014, "step": 30130 }, { "grad_norm": 0.16468487679958344, "learning_rate": 8.369462829571282e-05, "loss": 0.0124, "step": 30140 }, { "grad_norm": 0.20245929062366486, "learning_rate": 8.36824101332189e-05, "loss": 0.0135, "step": 30150 }, { "grad_norm": 0.1557636708021164, "learning_rate": 8.367018828726721e-05, "loss": 0.0116, "step": 30160 }, { "grad_norm": 0.12082719802856445, "learning_rate": 8.365796275919438e-05, "loss": 0.0126, "step": 30170 }, { "grad_norm": 0.14367012679576874, "learning_rate": 8.364573355033734e-05, "loss": 0.0134, "step": 30180 }, { "grad_norm": 0.1005294919013977, "learning_rate": 8.363350066203346e-05, "loss": 0.0095, "step": 30190 }, { "grad_norm": 0.18258535861968994, "learning_rate": 8.362126409562053e-05, "loss": 0.0125, "step": 30200 }, { "grad_norm": 0.1709388643503189, "learning_rate": 8.360902385243667e-05, "loss": 0.0125, "step": 30210 }, { "grad_norm": 0.22610993683338165, "learning_rate": 8.359677993382052e-05, "loss": 0.0108, "step": 30220 }, { "grad_norm": 0.190086230635643, "learning_rate": 8.358453234111103e-05, "loss": 0.0124, "step": 30230 }, { "grad_norm": 0.20007847249507904, "learning_rate": 8.357228107564756e-05, "loss": 0.015, "step": 30240 }, { "grad_norm": 0.15028293430805206, "learning_rate": 8.356002613876993e-05, "loss": 0.0128, "step": 30250 }, { "grad_norm": 0.13054905831813812, "learning_rate": 8.35477675318183e-05, "loss": 0.011, "step": 30260 }, { "grad_norm": 0.13114430010318756, "learning_rate": 8.353550525613323e-05, "loss": 0.0137, "step": 30270 }, { "grad_norm": 0.11678507924079895, "learning_rate": 8.352323931305572e-05, "loss": 0.0108, "step": 30280 }, { "grad_norm": 0.15887485444545746, "learning_rate": 8.351096970392717e-05, "loss": 0.0099, "step": 30290 }, { "grad_norm": 0.1263047158718109, "learning_rate": 8.349869643008937e-05, "loss": 0.0127, "step": 30300 }, { "grad_norm": 0.1873418390750885, "learning_rate": 8.348641949288449e-05, "loss": 0.0117, "step": 30310 }, { "grad_norm": 0.14841510355472565, "learning_rate": 8.34741388936551e-05, "loss": 0.0129, "step": 30320 }, { "grad_norm": 0.13611403107643127, "learning_rate": 8.346185463374423e-05, "loss": 0.0155, "step": 30330 }, { "grad_norm": 0.13641661405563354, "learning_rate": 8.344956671449524e-05, "loss": 0.0124, "step": 30340 }, { "grad_norm": 0.1240399181842804, "learning_rate": 8.343727513725192e-05, "loss": 0.0131, "step": 30350 }, { "grad_norm": 0.1758313924074173, "learning_rate": 8.342497990335847e-05, "loss": 0.0117, "step": 30360 }, { "grad_norm": 0.14645129442214966, "learning_rate": 8.341268101415946e-05, "loss": 0.0122, "step": 30370 }, { "grad_norm": 0.14995801448822021, "learning_rate": 8.34003784709999e-05, "loss": 0.0119, "step": 30380 }, { "grad_norm": 0.14118479192256927, "learning_rate": 8.338807227522516e-05, "loss": 0.0126, "step": 30390 }, { "grad_norm": 0.16911768913269043, "learning_rate": 8.337576242818103e-05, "loss": 0.0168, "step": 30400 }, { "grad_norm": 0.17874568700790405, "learning_rate": 8.336344893121372e-05, "loss": 0.0152, "step": 30410 }, { "grad_norm": 0.16723622381687164, "learning_rate": 8.335113178566977e-05, "loss": 0.0143, "step": 30420 }, { "grad_norm": 0.16883496940135956, "learning_rate": 8.33388109928962e-05, "loss": 0.0122, "step": 30430 }, { "grad_norm": 0.24471724033355713, "learning_rate": 8.33264865542404e-05, "loss": 0.0115, "step": 30440 }, { "grad_norm": 0.17434433102607727, "learning_rate": 8.331415847105013e-05, "loss": 0.0126, "step": 30450 }, { "grad_norm": 0.16028167307376862, "learning_rate": 8.330182674467357e-05, "loss": 0.0126, "step": 30460 }, { "grad_norm": 0.1077122837305069, "learning_rate": 8.32894913764593e-05, "loss": 0.0119, "step": 30470 }, { "grad_norm": 0.11480924487113953, "learning_rate": 8.327715236775633e-05, "loss": 0.0132, "step": 30480 }, { "grad_norm": 0.14405156672000885, "learning_rate": 8.326480971991398e-05, "loss": 0.0103, "step": 30490 }, { "grad_norm": 0.1019057035446167, "learning_rate": 8.325246343428206e-05, "loss": 0.0109, "step": 30500 }, { "grad_norm": 0.15320910513401031, "learning_rate": 8.324011351221072e-05, "loss": 0.0106, "step": 30510 }, { "grad_norm": 0.10240097343921661, "learning_rate": 8.322775995505057e-05, "loss": 0.0128, "step": 30520 }, { "grad_norm": 0.1821458786725998, "learning_rate": 8.321540276415254e-05, "loss": 0.0106, "step": 30530 }, { "grad_norm": 0.17870697379112244, "learning_rate": 8.320304194086798e-05, "loss": 0.0119, "step": 30540 }, { "grad_norm": 0.13354991376399994, "learning_rate": 8.31906774865487e-05, "loss": 0.0107, "step": 30550 }, { "grad_norm": 0.1350393146276474, "learning_rate": 8.317830940254682e-05, "loss": 0.013, "step": 30560 }, { "grad_norm": 0.18086539208889008, "learning_rate": 8.316593769021491e-05, "loss": 0.0125, "step": 30570 }, { "grad_norm": 0.1397487372159958, "learning_rate": 8.315356235090592e-05, "loss": 0.0103, "step": 30580 }, { "grad_norm": 0.20958739519119263, "learning_rate": 8.314118338597319e-05, "loss": 0.0124, "step": 30590 }, { "grad_norm": 0.19749200344085693, "learning_rate": 8.312880079677048e-05, "loss": 0.0123, "step": 30600 }, { "grad_norm": 0.19527150690555573, "learning_rate": 8.311641458465191e-05, "loss": 0.0124, "step": 30610 }, { "grad_norm": 0.23504509031772614, "learning_rate": 8.310402475097205e-05, "loss": 0.0126, "step": 30620 }, { "grad_norm": 0.26333048939704895, "learning_rate": 8.309163129708581e-05, "loss": 0.0131, "step": 30630 }, { "grad_norm": 0.2586197257041931, "learning_rate": 8.307923422434852e-05, "loss": 0.0137, "step": 30640 }, { "grad_norm": 0.1770649403333664, "learning_rate": 8.30668335341159e-05, "loss": 0.0114, "step": 30650 }, { "grad_norm": 0.14682114124298096, "learning_rate": 8.305442922774408e-05, "loss": 0.0115, "step": 30660 }, { "grad_norm": 0.1327262669801712, "learning_rate": 8.304202130658959e-05, "loss": 0.013, "step": 30670 }, { "grad_norm": 0.15007884800434113, "learning_rate": 8.302960977200931e-05, "loss": 0.014, "step": 30680 }, { "grad_norm": 0.17204205691814423, "learning_rate": 8.301719462536058e-05, "loss": 0.0113, "step": 30690 }, { "grad_norm": 0.17201076447963715, "learning_rate": 8.300477586800108e-05, "loss": 0.0117, "step": 30700 }, { "grad_norm": 0.13055875897407532, "learning_rate": 8.299235350128892e-05, "loss": 0.0137, "step": 30710 }, { "grad_norm": 0.2544955313205719, "learning_rate": 8.297992752658258e-05, "loss": 0.0128, "step": 30720 }, { "grad_norm": 0.16507311165332794, "learning_rate": 8.296749794524094e-05, "loss": 0.0148, "step": 30730 }, { "grad_norm": 0.1482509970664978, "learning_rate": 8.295506475862332e-05, "loss": 0.0115, "step": 30740 }, { "grad_norm": 0.13723617792129517, "learning_rate": 8.294262796808933e-05, "loss": 0.0134, "step": 30750 }, { "grad_norm": 0.13551650941371918, "learning_rate": 8.293018757499909e-05, "loss": 0.0129, "step": 30760 }, { "grad_norm": 0.19772593677043915, "learning_rate": 8.291774358071305e-05, "loss": 0.0121, "step": 30770 }, { "grad_norm": 0.17313708364963531, "learning_rate": 8.290529598659205e-05, "loss": 0.0114, "step": 30780 }, { "grad_norm": 0.15032817423343658, "learning_rate": 8.289284479399738e-05, "loss": 0.0107, "step": 30790 }, { "grad_norm": 0.18015192449092865, "learning_rate": 8.288039000429064e-05, "loss": 0.0113, "step": 30800 }, { "grad_norm": 0.19276273250579834, "learning_rate": 8.286793161883388e-05, "loss": 0.011, "step": 30810 }, { "grad_norm": 0.1924446076154709, "learning_rate": 8.285546963898954e-05, "loss": 0.0119, "step": 30820 }, { "grad_norm": 0.20252668857574463, "learning_rate": 8.284300406612044e-05, "loss": 0.012, "step": 30830 }, { "grad_norm": 0.17336001992225647, "learning_rate": 8.283053490158978e-05, "loss": 0.0105, "step": 30840 }, { "grad_norm": 0.18034757673740387, "learning_rate": 8.28180621467612e-05, "loss": 0.0109, "step": 30850 }, { "grad_norm": 0.13270996510982513, "learning_rate": 8.280558580299867e-05, "loss": 0.0119, "step": 30860 }, { "grad_norm": 0.1365586817264557, "learning_rate": 8.279310587166661e-05, "loss": 0.0098, "step": 30870 }, { "grad_norm": 0.24617333710193634, "learning_rate": 8.278062235412978e-05, "loss": 0.012, "step": 30880 }, { "grad_norm": 0.11856210976839066, "learning_rate": 8.276813525175339e-05, "loss": 0.013, "step": 30890 }, { "grad_norm": 0.17105337977409363, "learning_rate": 8.2755644565903e-05, "loss": 0.0167, "step": 30900 }, { "grad_norm": 0.1392563432455063, "learning_rate": 8.274315029794454e-05, "loss": 0.011, "step": 30910 }, { "grad_norm": 0.11883650720119476, "learning_rate": 8.273065244924443e-05, "loss": 0.0112, "step": 30920 }, { "grad_norm": 0.1786852329969406, "learning_rate": 8.271815102116936e-05, "loss": 0.0118, "step": 30930 }, { "grad_norm": 0.20018649101257324, "learning_rate": 8.270564601508648e-05, "loss": 0.0134, "step": 30940 }, { "grad_norm": 0.17795008420944214, "learning_rate": 8.269313743236333e-05, "loss": 0.0127, "step": 30950 }, { "grad_norm": 0.1506868153810501, "learning_rate": 8.268062527436783e-05, "loss": 0.0122, "step": 30960 }, { "grad_norm": 0.13646835088729858, "learning_rate": 8.266810954246828e-05, "loss": 0.01, "step": 30970 }, { "grad_norm": 0.2264244705438614, "learning_rate": 8.265559023803338e-05, "loss": 0.0116, "step": 30980 }, { "grad_norm": 0.19713248312473297, "learning_rate": 8.264306736243223e-05, "loss": 0.0127, "step": 30990 }, { "grad_norm": 0.18150660395622253, "learning_rate": 8.263054091703432e-05, "loss": 0.0117, "step": 31000 }, { "grad_norm": 0.18005535006523132, "learning_rate": 8.26180109032095e-05, "loss": 0.0129, "step": 31010 }, { "grad_norm": 0.14760135114192963, "learning_rate": 8.260547732232807e-05, "loss": 0.0123, "step": 31020 }, { "grad_norm": 0.20682889223098755, "learning_rate": 8.259294017576063e-05, "loss": 0.0136, "step": 31030 }, { "grad_norm": 0.17720820009708405, "learning_rate": 8.258039946487828e-05, "loss": 0.0112, "step": 31040 }, { "grad_norm": 0.10973090678453445, "learning_rate": 8.256785519105241e-05, "loss": 0.0125, "step": 31050 }, { "grad_norm": 0.15468870103359222, "learning_rate": 8.255530735565488e-05, "loss": 0.0143, "step": 31060 }, { "grad_norm": 0.11586618423461914, "learning_rate": 8.254275596005787e-05, "loss": 0.0105, "step": 31070 }, { "grad_norm": 0.15021835267543793, "learning_rate": 8.2530201005634e-05, "loss": 0.0142, "step": 31080 }, { "grad_norm": 0.1708829402923584, "learning_rate": 8.251764249375626e-05, "loss": 0.0127, "step": 31090 }, { "grad_norm": 0.19696266949176788, "learning_rate": 8.250508042579803e-05, "loss": 0.0138, "step": 31100 }, { "grad_norm": 0.16547061502933502, "learning_rate": 8.249251480313307e-05, "loss": 0.0123, "step": 31110 }, { "grad_norm": 0.15460914373397827, "learning_rate": 8.247994562713555e-05, "loss": 0.0144, "step": 31120 }, { "grad_norm": 0.16799496114253998, "learning_rate": 8.246737289917998e-05, "loss": 0.0119, "step": 31130 }, { "grad_norm": 0.16284628212451935, "learning_rate": 8.245479662064135e-05, "loss": 0.0121, "step": 31140 }, { "grad_norm": 0.172336146235466, "learning_rate": 8.244221679289496e-05, "loss": 0.0134, "step": 31150 }, { "grad_norm": 0.13170871138572693, "learning_rate": 8.242963341731652e-05, "loss": 0.0114, "step": 31160 }, { "grad_norm": 0.13633814454078674, "learning_rate": 8.24170464952821e-05, "loss": 0.0118, "step": 31170 }, { "grad_norm": 0.16768988966941833, "learning_rate": 8.240445602816824e-05, "loss": 0.0135, "step": 31180 }, { "grad_norm": 0.16099078953266144, "learning_rate": 8.239186201735179e-05, "loss": 0.0145, "step": 31190 }, { "grad_norm": 0.18312810361385345, "learning_rate": 8.237926446420998e-05, "loss": 0.0121, "step": 31200 }, { "grad_norm": 0.14757925271987915, "learning_rate": 8.236666337012052e-05, "loss": 0.0113, "step": 31210 }, { "grad_norm": 0.16477538645267487, "learning_rate": 8.23540587364614e-05, "loss": 0.0134, "step": 31220 }, { "grad_norm": 0.15533848106861115, "learning_rate": 8.234145056461107e-05, "loss": 0.0101, "step": 31230 }, { "grad_norm": 0.15258532762527466, "learning_rate": 8.232883885594832e-05, "loss": 0.0102, "step": 31240 }, { "grad_norm": 0.17324113845825195, "learning_rate": 8.231622361185236e-05, "loss": 0.0113, "step": 31250 }, { "grad_norm": 0.19270800054073334, "learning_rate": 8.230360483370278e-05, "loss": 0.0114, "step": 31260 }, { "grad_norm": 0.18391332030296326, "learning_rate": 8.229098252287953e-05, "loss": 0.0136, "step": 31270 }, { "grad_norm": 0.22331249713897705, "learning_rate": 8.2278356680763e-05, "loss": 0.0117, "step": 31280 }, { "grad_norm": 0.20335382223129272, "learning_rate": 8.22657273087339e-05, "loss": 0.014, "step": 31290 }, { "grad_norm": 0.15605036914348602, "learning_rate": 8.225309440817336e-05, "loss": 0.0125, "step": 31300 }, { "grad_norm": 0.17187654972076416, "learning_rate": 8.224045798046293e-05, "loss": 0.0144, "step": 31310 }, { "grad_norm": 0.14592556655406952, "learning_rate": 8.22278180269845e-05, "loss": 0.0119, "step": 31320 }, { "grad_norm": 0.13807083666324615, "learning_rate": 8.221517454912031e-05, "loss": 0.0147, "step": 31330 }, { "grad_norm": 0.1444624811410904, "learning_rate": 8.220252754825308e-05, "loss": 0.0139, "step": 31340 }, { "grad_norm": 0.13463321328163147, "learning_rate": 8.218987702576586e-05, "loss": 0.0116, "step": 31350 }, { "grad_norm": 0.1415698379278183, "learning_rate": 8.217722298304207e-05, "loss": 0.0107, "step": 31360 }, { "grad_norm": 0.2151644080877304, "learning_rate": 8.216456542146557e-05, "loss": 0.0142, "step": 31370 }, { "grad_norm": 0.13111701607704163, "learning_rate": 8.215190434242055e-05, "loss": 0.0108, "step": 31380 }, { "grad_norm": 0.16711246967315674, "learning_rate": 8.213923974729161e-05, "loss": 0.0135, "step": 31390 }, { "grad_norm": 0.1580238938331604, "learning_rate": 8.212657163746373e-05, "loss": 0.0125, "step": 31400 }, { "grad_norm": 0.15582096576690674, "learning_rate": 8.211390001432227e-05, "loss": 0.0138, "step": 31410 }, { "grad_norm": 0.1306331902742386, "learning_rate": 8.210122487925297e-05, "loss": 0.0107, "step": 31420 }, { "grad_norm": 0.12264642864465714, "learning_rate": 8.208854623364202e-05, "loss": 0.0121, "step": 31430 }, { "grad_norm": 0.1705573946237564, "learning_rate": 8.207586407887589e-05, "loss": 0.0131, "step": 31440 }, { "grad_norm": 0.14720042049884796, "learning_rate": 8.206317841634148e-05, "loss": 0.0126, "step": 31450 }, { "grad_norm": 0.16630849242210388, "learning_rate": 8.205048924742609e-05, "loss": 0.0106, "step": 31460 }, { "grad_norm": 0.1299455165863037, "learning_rate": 8.203779657351738e-05, "loss": 0.0129, "step": 31470 }, { "grad_norm": 0.1812528520822525, "learning_rate": 8.20251003960034e-05, "loss": 0.0138, "step": 31480 }, { "grad_norm": 0.13948987424373627, "learning_rate": 8.201240071627258e-05, "loss": 0.0124, "step": 31490 }, { "grad_norm": 0.1378105878829956, "learning_rate": 8.199969753571377e-05, "loss": 0.0132, "step": 31500 }, { "grad_norm": 0.16211166977882385, "learning_rate": 8.198699085571615e-05, "loss": 0.0131, "step": 31510 }, { "grad_norm": 0.20617851614952087, "learning_rate": 8.197428067766928e-05, "loss": 0.0128, "step": 31520 }, { "grad_norm": 0.1681348830461502, "learning_rate": 8.196156700296316e-05, "loss": 0.0133, "step": 31530 }, { "grad_norm": 0.16421940922737122, "learning_rate": 8.194884983298814e-05, "loss": 0.0109, "step": 31540 }, { "grad_norm": 0.11378016322851181, "learning_rate": 8.193612916913491e-05, "loss": 0.0127, "step": 31550 }, { "grad_norm": 0.13172899186611176, "learning_rate": 8.192340501279463e-05, "loss": 0.0103, "step": 31560 }, { "grad_norm": 0.1965823918581009, "learning_rate": 8.191067736535876e-05, "loss": 0.0125, "step": 31570 }, { "grad_norm": 0.1458885371685028, "learning_rate": 8.18979462282192e-05, "loss": 0.0116, "step": 31580 }, { "grad_norm": 0.11348222941160202, "learning_rate": 8.188521160276819e-05, "loss": 0.0119, "step": 31590 }, { "grad_norm": 0.1598658412694931, "learning_rate": 8.187247349039837e-05, "loss": 0.0118, "step": 31600 }, { "grad_norm": 0.12127669155597687, "learning_rate": 8.185973189250278e-05, "loss": 0.012, "step": 31610 }, { "grad_norm": 0.1731811761856079, "learning_rate": 8.184698681047482e-05, "loss": 0.0118, "step": 31620 }, { "grad_norm": 0.15669308602809906, "learning_rate": 8.183423824570827e-05, "loss": 0.013, "step": 31630 }, { "grad_norm": 0.14725248515605927, "learning_rate": 8.182148619959725e-05, "loss": 0.0101, "step": 31640 }, { "grad_norm": 0.1625305861234665, "learning_rate": 8.180873067353636e-05, "loss": 0.0122, "step": 31650 }, { "grad_norm": 0.16451556980609894, "learning_rate": 8.179597166892052e-05, "loss": 0.0101, "step": 31660 }, { "grad_norm": 0.22128501534461975, "learning_rate": 8.178320918714501e-05, "loss": 0.0123, "step": 31670 }, { "grad_norm": 0.25974804162979126, "learning_rate": 8.177044322960554e-05, "loss": 0.013, "step": 31680 }, { "grad_norm": 0.17788580060005188, "learning_rate": 8.175767379769816e-05, "loss": 0.0124, "step": 31690 }, { "grad_norm": 0.15582889318466187, "learning_rate": 8.174490089281932e-05, "loss": 0.0117, "step": 31700 }, { "grad_norm": 0.11924584954977036, "learning_rate": 8.173212451636584e-05, "loss": 0.0105, "step": 31710 }, { "grad_norm": 0.15269650518894196, "learning_rate": 8.171934466973493e-05, "loss": 0.0104, "step": 31720 }, { "grad_norm": 0.18507234752178192, "learning_rate": 8.170656135432418e-05, "loss": 0.0112, "step": 31730 }, { "grad_norm": 0.13725723326206207, "learning_rate": 8.169377457153155e-05, "loss": 0.0135, "step": 31740 }, { "grad_norm": 0.14586111903190613, "learning_rate": 8.168098432275539e-05, "loss": 0.0128, "step": 31750 }, { "grad_norm": 0.15460680425167084, "learning_rate": 8.166819060939442e-05, "loss": 0.0116, "step": 31760 }, { "grad_norm": 0.12913639843463898, "learning_rate": 8.165539343284772e-05, "loss": 0.0129, "step": 31770 }, { "grad_norm": 0.1372508555650711, "learning_rate": 8.16425927945148e-05, "loss": 0.0136, "step": 31780 }, { "grad_norm": 0.19442778825759888, "learning_rate": 8.162978869579551e-05, "loss": 0.0158, "step": 31790 }, { "grad_norm": 0.11799158900976181, "learning_rate": 8.161698113809007e-05, "loss": 0.0118, "step": 31800 }, { "grad_norm": 0.14173771440982819, "learning_rate": 8.160417012279911e-05, "loss": 0.0121, "step": 31810 }, { "grad_norm": 0.18235865235328674, "learning_rate": 8.159135565132363e-05, "loss": 0.0141, "step": 31820 }, { "grad_norm": 0.15990804135799408, "learning_rate": 8.157853772506498e-05, "loss": 0.0138, "step": 31830 }, { "grad_norm": 0.12907475233078003, "learning_rate": 8.156571634542494e-05, "loss": 0.0132, "step": 31840 }, { "grad_norm": 0.18845267593860626, "learning_rate": 8.15528915138056e-05, "loss": 0.0131, "step": 31850 }, { "grad_norm": 0.15708455443382263, "learning_rate": 8.154006323160949e-05, "loss": 0.0139, "step": 31860 }, { "grad_norm": 0.11737435311079025, "learning_rate": 8.152723150023949e-05, "loss": 0.0112, "step": 31870 }, { "grad_norm": 0.12453577667474747, "learning_rate": 8.151439632109886e-05, "loss": 0.0118, "step": 31880 }, { "grad_norm": 0.15369151532649994, "learning_rate": 8.150155769559122e-05, "loss": 0.0136, "step": 31890 }, { "grad_norm": 0.11813392490148544, "learning_rate": 8.148871562512058e-05, "loss": 0.0122, "step": 31900 }, { "grad_norm": 0.136738583445549, "learning_rate": 8.147587011109136e-05, "loss": 0.0114, "step": 31910 }, { "grad_norm": 0.1452249437570572, "learning_rate": 8.14630211549083e-05, "loss": 0.0105, "step": 31920 }, { "grad_norm": 0.10223682224750519, "learning_rate": 8.145016875797655e-05, "loss": 0.0091, "step": 31930 }, { "grad_norm": 0.1660185307264328, "learning_rate": 8.143731292170164e-05, "loss": 0.0123, "step": 31940 }, { "grad_norm": 0.10231483727693558, "learning_rate": 8.142445364748944e-05, "loss": 0.0145, "step": 31950 }, { "grad_norm": 0.15114715695381165, "learning_rate": 8.141159093674624e-05, "loss": 0.0109, "step": 31960 }, { "grad_norm": 0.15019017457962036, "learning_rate": 8.139872479087869e-05, "loss": 0.0136, "step": 31970 }, { "grad_norm": 0.12995104491710663, "learning_rate": 8.13858552112938e-05, "loss": 0.0119, "step": 31980 }, { "grad_norm": 0.1634363830089569, "learning_rate": 8.137298219939895e-05, "loss": 0.0111, "step": 31990 }, { "grad_norm": 0.11568544059991837, "learning_rate": 8.136010575660196e-05, "loss": 0.0108, "step": 32000 }, { "grad_norm": 0.15176308155059814, "learning_rate": 8.134722588431095e-05, "loss": 0.0129, "step": 32010 }, { "grad_norm": 0.12589915096759796, "learning_rate": 8.133434258393444e-05, "loss": 0.0124, "step": 32020 }, { "grad_norm": 0.17221879959106445, "learning_rate": 8.132145585688134e-05, "loss": 0.0112, "step": 32030 }, { "grad_norm": 0.21360404789447784, "learning_rate": 8.130856570456093e-05, "loss": 0.0139, "step": 32040 }, { "grad_norm": 0.17703422904014587, "learning_rate": 8.129567212838283e-05, "loss": 0.0137, "step": 32050 }, { "grad_norm": 0.11048224568367004, "learning_rate": 8.128277512975708e-05, "loss": 0.0103, "step": 32060 }, { "grad_norm": 0.13961587846279144, "learning_rate": 8.126987471009408e-05, "loss": 0.0108, "step": 32070 }, { "grad_norm": 0.12089591473340988, "learning_rate": 8.125697087080459e-05, "loss": 0.0106, "step": 32080 }, { "grad_norm": 0.23816350102424622, "learning_rate": 8.124406361329976e-05, "loss": 0.0118, "step": 32090 }, { "grad_norm": 0.19809743762016296, "learning_rate": 8.123115293899111e-05, "loss": 0.0124, "step": 32100 }, { "grad_norm": 0.1614002287387848, "learning_rate": 8.121823884929055e-05, "loss": 0.012, "step": 32110 }, { "grad_norm": 0.12727531790733337, "learning_rate": 8.12053213456103e-05, "loss": 0.012, "step": 32120 }, { "grad_norm": 0.1691824495792389, "learning_rate": 8.119240042936303e-05, "loss": 0.013, "step": 32130 }, { "grad_norm": 0.1654146909713745, "learning_rate": 8.117947610196175e-05, "loss": 0.0124, "step": 32140 }, { "grad_norm": 0.13209140300750732, "learning_rate": 8.116654836481982e-05, "loss": 0.0108, "step": 32150 }, { "grad_norm": 0.15877868235111237, "learning_rate": 8.115361721935106e-05, "loss": 0.0118, "step": 32160 }, { "grad_norm": 0.1459459662437439, "learning_rate": 8.114068266696953e-05, "loss": 0.0114, "step": 32170 }, { "grad_norm": 0.09950617700815201, "learning_rate": 8.112774470908978e-05, "loss": 0.0115, "step": 32180 }, { "grad_norm": 0.16845135390758514, "learning_rate": 8.111480334712665e-05, "loss": 0.0123, "step": 32190 }, { "grad_norm": 0.1636127084493637, "learning_rate": 8.110185858249542e-05, "loss": 0.0121, "step": 32200 }, { "grad_norm": 0.18417073786258698, "learning_rate": 8.108891041661168e-05, "loss": 0.012, "step": 32210 }, { "grad_norm": 0.16435818374156952, "learning_rate": 8.107595885089146e-05, "loss": 0.0149, "step": 32220 }, { "grad_norm": 0.1621507704257965, "learning_rate": 8.106300388675108e-05, "loss": 0.0116, "step": 32230 }, { "grad_norm": 0.18036288022994995, "learning_rate": 8.105004552560731e-05, "loss": 0.0133, "step": 32240 }, { "grad_norm": 0.11283852905035019, "learning_rate": 8.103708376887724e-05, "loss": 0.0112, "step": 32250 }, { "grad_norm": 0.18173019587993622, "learning_rate": 8.102411861797836e-05, "loss": 0.0115, "step": 32260 }, { "grad_norm": 0.13390988111495972, "learning_rate": 8.101115007432851e-05, "loss": 0.0109, "step": 32270 }, { "grad_norm": 0.11378847807645798, "learning_rate": 8.09981781393459e-05, "loss": 0.0113, "step": 32280 }, { "grad_norm": 0.15389464795589447, "learning_rate": 8.098520281444915e-05, "loss": 0.0124, "step": 32290 }, { "grad_norm": 0.14991767704486847, "learning_rate": 8.09722241010572e-05, "loss": 0.0121, "step": 32300 }, { "grad_norm": 0.15920165181159973, "learning_rate": 8.095924200058939e-05, "loss": 0.0118, "step": 32310 }, { "grad_norm": 0.11653949320316315, "learning_rate": 8.094625651446541e-05, "loss": 0.0113, "step": 32320 }, { "grad_norm": 0.11657053977251053, "learning_rate": 8.093326764410536e-05, "loss": 0.0104, "step": 32330 }, { "grad_norm": 0.14376699924468994, "learning_rate": 8.092027539092966e-05, "loss": 0.0125, "step": 32340 }, { "grad_norm": 0.13569097220897675, "learning_rate": 8.090727975635913e-05, "loss": 0.0121, "step": 32350 }, { "grad_norm": 0.14162878692150116, "learning_rate": 8.089428074181497e-05, "loss": 0.0107, "step": 32360 }, { "grad_norm": 0.11215102672576904, "learning_rate": 8.088127834871871e-05, "loss": 0.0105, "step": 32370 }, { "grad_norm": 0.1647324115037918, "learning_rate": 8.086827257849226e-05, "loss": 0.0115, "step": 32380 }, { "grad_norm": 0.15918117761611938, "learning_rate": 8.085526343255795e-05, "loss": 0.0105, "step": 32390 }, { "grad_norm": 0.15283234417438507, "learning_rate": 8.084225091233842e-05, "loss": 0.0111, "step": 32400 }, { "grad_norm": 0.0944812148809433, "learning_rate": 8.082923501925668e-05, "loss": 0.0104, "step": 32410 }, { "grad_norm": 0.19965504109859467, "learning_rate": 8.081621575473617e-05, "loss": 0.0159, "step": 32420 }, { "grad_norm": 0.20103245973587036, "learning_rate": 8.080319312020064e-05, "loss": 0.0128, "step": 32430 }, { "grad_norm": 0.1895793378353119, "learning_rate": 8.079016711707421e-05, "loss": 0.012, "step": 32440 }, { "grad_norm": 0.1342589259147644, "learning_rate": 8.077713774678139e-05, "loss": 0.0109, "step": 32450 }, { "grad_norm": 0.14996030926704407, "learning_rate": 8.076410501074707e-05, "loss": 0.0109, "step": 32460 }, { "grad_norm": 0.20099280774593353, "learning_rate": 8.075106891039647e-05, "loss": 0.0125, "step": 32470 }, { "grad_norm": 0.1824033409357071, "learning_rate": 8.073802944715523e-05, "loss": 0.0117, "step": 32480 }, { "grad_norm": 0.13544602692127228, "learning_rate": 8.072498662244929e-05, "loss": 0.0114, "step": 32490 }, { "grad_norm": 0.19684891402721405, "learning_rate": 8.0711940437705e-05, "loss": 0.0139, "step": 32500 }, { "grad_norm": 0.26100513339042664, "learning_rate": 8.06988908943491e-05, "loss": 0.012, "step": 32510 }, { "grad_norm": 0.24949978291988373, "learning_rate": 8.068583799380863e-05, "loss": 0.0123, "step": 32520 }, { "grad_norm": 0.13359379768371582, "learning_rate": 8.067278173751104e-05, "loss": 0.0125, "step": 32530 }, { "grad_norm": 0.16508108377456665, "learning_rate": 8.065972212688417e-05, "loss": 0.0125, "step": 32540 }, { "grad_norm": 0.1461336463689804, "learning_rate": 8.064665916335618e-05, "loss": 0.0119, "step": 32550 }, { "grad_norm": 0.1554577648639679, "learning_rate": 8.063359284835564e-05, "loss": 0.0128, "step": 32560 }, { "grad_norm": 0.12712819874286652, "learning_rate": 8.062052318331142e-05, "loss": 0.014, "step": 32570 }, { "grad_norm": 0.18360695242881775, "learning_rate": 8.060745016965283e-05, "loss": 0.0128, "step": 32580 }, { "grad_norm": 0.14996649324893951, "learning_rate": 8.059437380880952e-05, "loss": 0.0108, "step": 32590 }, { "grad_norm": 0.14650768041610718, "learning_rate": 8.058129410221146e-05, "loss": 0.0117, "step": 32600 }, { "grad_norm": 0.181073859333992, "learning_rate": 8.056821105128908e-05, "loss": 0.0132, "step": 32610 }, { "grad_norm": 0.12544268369674683, "learning_rate": 8.05551246574731e-05, "loss": 0.0117, "step": 32620 }, { "grad_norm": 0.16393500566482544, "learning_rate": 8.05420349221946e-05, "loss": 0.0126, "step": 32630 }, { "grad_norm": 0.1604364961385727, "learning_rate": 8.05289418468851e-05, "loss": 0.0122, "step": 32640 }, { "grad_norm": 0.12747803330421448, "learning_rate": 8.051584543297642e-05, "loss": 0.0122, "step": 32650 }, { "grad_norm": 0.13679896295070648, "learning_rate": 8.050274568190074e-05, "loss": 0.0112, "step": 32660 }, { "grad_norm": 0.11909572035074234, "learning_rate": 8.048964259509067e-05, "loss": 0.0122, "step": 32670 }, { "grad_norm": 0.20617741346359253, "learning_rate": 8.047653617397914e-05, "loss": 0.0132, "step": 32680 }, { "grad_norm": 0.11630840599536896, "learning_rate": 8.046342641999941e-05, "loss": 0.012, "step": 32690 }, { "grad_norm": 0.1351146101951599, "learning_rate": 8.045031333458517e-05, "loss": 0.0124, "step": 32700 }, { "grad_norm": 0.1434357762336731, "learning_rate": 8.043719691917047e-05, "loss": 0.0126, "step": 32710 }, { "grad_norm": 0.1491200476884842, "learning_rate": 8.042407717518966e-05, "loss": 0.0117, "step": 32720 }, { "grad_norm": 0.12742772698402405, "learning_rate": 8.041095410407751e-05, "loss": 0.0131, "step": 32730 }, { "grad_norm": 0.16751566529273987, "learning_rate": 8.039782770726913e-05, "loss": 0.0133, "step": 32740 }, { "grad_norm": 0.16251599788665771, "learning_rate": 8.038469798620004e-05, "loss": 0.0129, "step": 32750 }, { "grad_norm": 0.23833516240119934, "learning_rate": 8.037156494230604e-05, "loss": 0.0155, "step": 32760 }, { "grad_norm": 0.16285452246665955, "learning_rate": 8.035842857702338e-05, "loss": 0.0147, "step": 32770 }, { "grad_norm": 0.12706656754016876, "learning_rate": 8.03452888917886e-05, "loss": 0.0151, "step": 32780 }, { "grad_norm": 0.17193813621997833, "learning_rate": 8.033214588803866e-05, "loss": 0.0148, "step": 32790 }, { "grad_norm": 0.14152395725250244, "learning_rate": 8.031899956721083e-05, "loss": 0.0111, "step": 32800 }, { "grad_norm": 0.1752479076385498, "learning_rate": 8.030584993074282e-05, "loss": 0.0132, "step": 32810 }, { "grad_norm": 0.2240334153175354, "learning_rate": 8.02926969800726e-05, "loss": 0.0126, "step": 32820 }, { "grad_norm": 0.14086821675300598, "learning_rate": 8.027954071663859e-05, "loss": 0.0103, "step": 32830 }, { "grad_norm": 0.15673667192459106, "learning_rate": 8.026638114187954e-05, "loss": 0.0131, "step": 32840 }, { "grad_norm": 0.11697594821453094, "learning_rate": 8.025321825723456e-05, "loss": 0.0148, "step": 32850 }, { "grad_norm": 0.14613932371139526, "learning_rate": 8.02400520641431e-05, "loss": 0.0114, "step": 32860 }, { "grad_norm": 0.13846710324287415, "learning_rate": 8.022688256404501e-05, "loss": 0.0118, "step": 32870 }, { "grad_norm": 0.18473058938980103, "learning_rate": 8.02137097583805e-05, "loss": 0.0109, "step": 32880 }, { "grad_norm": 0.14406107366085052, "learning_rate": 8.02005336485901e-05, "loss": 0.0127, "step": 32890 }, { "grad_norm": 0.15404511988162994, "learning_rate": 8.018735423611476e-05, "loss": 0.0131, "step": 32900 }, { "grad_norm": 0.14052507281303406, "learning_rate": 8.017417152239574e-05, "loss": 0.0126, "step": 32910 }, { "grad_norm": 0.1214413121342659, "learning_rate": 8.01609855088747e-05, "loss": 0.0094, "step": 32920 }, { "grad_norm": 0.138331338763237, "learning_rate": 8.014779619699362e-05, "loss": 0.0125, "step": 32930 }, { "grad_norm": 0.13416031002998352, "learning_rate": 8.013460358819489e-05, "loss": 0.0108, "step": 32940 }, { "grad_norm": 0.16954199969768524, "learning_rate": 8.01214076839212e-05, "loss": 0.0119, "step": 32950 }, { "grad_norm": 0.37385568022727966, "learning_rate": 8.010820848561565e-05, "loss": 0.0132, "step": 32960 }, { "grad_norm": 0.14850148558616638, "learning_rate": 8.009500599472171e-05, "loss": 0.0145, "step": 32970 }, { "grad_norm": 0.1671857386827469, "learning_rate": 8.008180021268314e-05, "loss": 0.0118, "step": 32980 }, { "grad_norm": 0.16060112416744232, "learning_rate": 8.006859114094414e-05, "loss": 0.0133, "step": 32990 }, { "grad_norm": 0.17625658214092255, "learning_rate": 8.005537878094921e-05, "loss": 0.0132, "step": 33000 }, { "grad_norm": 0.10258032381534576, "learning_rate": 8.004216313414323e-05, "loss": 0.0117, "step": 33010 }, { "grad_norm": 0.13309742510318756, "learning_rate": 8.002894420197149e-05, "loss": 0.0111, "step": 33020 }, { "grad_norm": 0.11693089455366135, "learning_rate": 8.001572198587954e-05, "loss": 0.011, "step": 33030 }, { "grad_norm": 0.1663568764925003, "learning_rate": 8.000249648731338e-05, "loss": 0.0105, "step": 33040 }, { "grad_norm": 0.16879455745220184, "learning_rate": 7.998926770771928e-05, "loss": 0.0121, "step": 33050 }, { "grad_norm": 0.1647137850522995, "learning_rate": 7.997603564854397e-05, "loss": 0.0111, "step": 33060 }, { "grad_norm": 0.1264939159154892, "learning_rate": 7.996280031123448e-05, "loss": 0.0106, "step": 33070 }, { "grad_norm": 0.12107323110103607, "learning_rate": 7.994956169723818e-05, "loss": 0.0111, "step": 33080 }, { "grad_norm": 0.20435599982738495, "learning_rate": 7.993631980800285e-05, "loss": 0.0114, "step": 33090 }, { "grad_norm": 0.12630248069763184, "learning_rate": 7.992307464497659e-05, "loss": 0.0114, "step": 33100 }, { "grad_norm": 0.229514941573143, "learning_rate": 7.990982620960787e-05, "loss": 0.0114, "step": 33110 }, { "grad_norm": 0.13770535588264465, "learning_rate": 7.989657450334554e-05, "loss": 0.0093, "step": 33120 }, { "grad_norm": 0.15539708733558655, "learning_rate": 7.988331952763877e-05, "loss": 0.0125, "step": 33130 }, { "grad_norm": 0.15475381910800934, "learning_rate": 7.987006128393709e-05, "loss": 0.0103, "step": 33140 }, { "grad_norm": 0.14972229301929474, "learning_rate": 7.985679977369043e-05, "loss": 0.0123, "step": 33150 }, { "grad_norm": 0.155063658952713, "learning_rate": 7.984353499834902e-05, "loss": 0.0111, "step": 33160 }, { "grad_norm": 0.1625460535287857, "learning_rate": 7.983026695936351e-05, "loss": 0.0117, "step": 33170 }, { "grad_norm": 0.16782861948013306, "learning_rate": 7.981699565818486e-05, "loss": 0.0113, "step": 33180 }, { "grad_norm": 0.13377909362316132, "learning_rate": 7.980372109626437e-05, "loss": 0.0109, "step": 33190 }, { "grad_norm": 0.12164154648780823, "learning_rate": 7.979044327505375e-05, "loss": 0.0111, "step": 33200 }, { "grad_norm": 0.19074060022830963, "learning_rate": 7.977716219600506e-05, "loss": 0.0142, "step": 33210 }, { "grad_norm": 0.14532074332237244, "learning_rate": 7.97638778605707e-05, "loss": 0.0142, "step": 33220 }, { "grad_norm": 0.138208270072937, "learning_rate": 7.975059027020338e-05, "loss": 0.0138, "step": 33230 }, { "grad_norm": 0.13341470062732697, "learning_rate": 7.973729942635623e-05, "loss": 0.0139, "step": 33240 }, { "grad_norm": 0.16009393334388733, "learning_rate": 7.972400533048273e-05, "loss": 0.0134, "step": 33250 }, { "grad_norm": 0.12102293968200684, "learning_rate": 7.97107079840367e-05, "loss": 0.0134, "step": 33260 }, { "grad_norm": 0.12445271760225296, "learning_rate": 7.969740738847231e-05, "loss": 0.0126, "step": 33270 }, { "grad_norm": 0.15712764859199524, "learning_rate": 7.968410354524411e-05, "loss": 0.012, "step": 33280 }, { "grad_norm": 0.15415765345096588, "learning_rate": 7.967079645580697e-05, "loss": 0.0128, "step": 33290 }, { "grad_norm": 0.1374630331993103, "learning_rate": 7.965748612161612e-05, "loss": 0.0109, "step": 33300 }, { "grad_norm": 0.14468267560005188, "learning_rate": 7.96441725441272e-05, "loss": 0.0116, "step": 33310 }, { "grad_norm": 0.10124620050191879, "learning_rate": 7.963085572479614e-05, "loss": 0.0114, "step": 33320 }, { "grad_norm": 0.12093482911586761, "learning_rate": 7.961753566507924e-05, "loss": 0.0113, "step": 33330 }, { "grad_norm": 0.15175452828407288, "learning_rate": 7.960421236643316e-05, "loss": 0.0113, "step": 33340 }, { "grad_norm": 0.14209888875484467, "learning_rate": 7.959088583031496e-05, "loss": 0.0133, "step": 33350 }, { "grad_norm": 0.1728595495223999, "learning_rate": 7.957755605818194e-05, "loss": 0.0116, "step": 33360 }, { "grad_norm": 0.2002336084842682, "learning_rate": 7.956422305149185e-05, "loss": 0.0105, "step": 33370 }, { "grad_norm": 0.12097574025392532, "learning_rate": 7.95508868117028e-05, "loss": 0.011, "step": 33380 }, { "grad_norm": 0.17079001665115356, "learning_rate": 7.953754734027318e-05, "loss": 0.0106, "step": 33390 }, { "grad_norm": 0.17094619572162628, "learning_rate": 7.952420463866182e-05, "loss": 0.0111, "step": 33400 }, { "grad_norm": 0.17487095296382904, "learning_rate": 7.951085870832782e-05, "loss": 0.0123, "step": 33410 }, { "grad_norm": 0.18722258508205414, "learning_rate": 7.949750955073067e-05, "loss": 0.0125, "step": 33420 }, { "grad_norm": 0.20953713357448578, "learning_rate": 7.948415716733022e-05, "loss": 0.0121, "step": 33430 }, { "grad_norm": 0.15018917620182037, "learning_rate": 7.947080155958669e-05, "loss": 0.0136, "step": 33440 }, { "grad_norm": 0.1607559472322464, "learning_rate": 7.94574427289606e-05, "loss": 0.0126, "step": 33450 }, { "grad_norm": 0.1679074913263321, "learning_rate": 7.944408067691284e-05, "loss": 0.0138, "step": 33460 }, { "grad_norm": 0.1504315733909607, "learning_rate": 7.943071540490473e-05, "loss": 0.0131, "step": 33470 }, { "grad_norm": 0.15416133403778076, "learning_rate": 7.94173469143978e-05, "loss": 0.0132, "step": 33480 }, { "grad_norm": 0.12950602173805237, "learning_rate": 7.940397520685406e-05, "loss": 0.0126, "step": 33490 }, { "grad_norm": 0.13744843006134033, "learning_rate": 7.939060028373577e-05, "loss": 0.0114, "step": 33500 }, { "grad_norm": 0.22454798221588135, "learning_rate": 7.937722214650565e-05, "loss": 0.0122, "step": 33510 }, { "grad_norm": 0.1526069939136505, "learning_rate": 7.936384079662666e-05, "loss": 0.0116, "step": 33520 }, { "grad_norm": 0.13697843253612518, "learning_rate": 7.93504562355622e-05, "loss": 0.0126, "step": 33530 }, { "grad_norm": 0.12948451936244965, "learning_rate": 7.933706846477599e-05, "loss": 0.0121, "step": 33540 }, { "grad_norm": 0.11916784197092056, "learning_rate": 7.932367748573206e-05, "loss": 0.0123, "step": 33550 }, { "grad_norm": 0.16176027059555054, "learning_rate": 7.931028329989485e-05, "loss": 0.0114, "step": 33560 }, { "grad_norm": 0.14933082461357117, "learning_rate": 7.929688590872913e-05, "loss": 0.0104, "step": 33570 }, { "grad_norm": 0.13228215277194977, "learning_rate": 7.928348531370003e-05, "loss": 0.0119, "step": 33580 }, { "grad_norm": 0.16053159534931183, "learning_rate": 7.927008151627297e-05, "loss": 0.0107, "step": 33590 }, { "grad_norm": 0.11093468964099884, "learning_rate": 7.925667451791383e-05, "loss": 0.0105, "step": 33600 }, { "grad_norm": 0.13051241636276245, "learning_rate": 7.924326432008874e-05, "loss": 0.0106, "step": 33610 }, { "grad_norm": 0.1346251368522644, "learning_rate": 7.922985092426422e-05, "loss": 0.0124, "step": 33620 }, { "grad_norm": 0.11517778784036636, "learning_rate": 7.921643433190717e-05, "loss": 0.0136, "step": 33630 }, { "grad_norm": 0.1325438916683197, "learning_rate": 7.920301454448478e-05, "loss": 0.0112, "step": 33640 }, { "grad_norm": 0.13802850246429443, "learning_rate": 7.918959156346461e-05, "loss": 0.0113, "step": 33650 }, { "grad_norm": 0.17025494575500488, "learning_rate": 7.91761653903146e-05, "loss": 0.0103, "step": 33660 }, { "grad_norm": 0.18346615135669708, "learning_rate": 7.916273602650302e-05, "loss": 0.0105, "step": 33670 }, { "grad_norm": 0.1439111828804016, "learning_rate": 7.914930347349847e-05, "loss": 0.0113, "step": 33680 }, { "grad_norm": 0.13152466714382172, "learning_rate": 7.913586773276992e-05, "loss": 0.0107, "step": 33690 }, { "grad_norm": 0.1562637835741043, "learning_rate": 7.912242880578667e-05, "loss": 0.0161, "step": 33700 }, { "grad_norm": 0.1527283489704132, "learning_rate": 7.910898669401839e-05, "loss": 0.012, "step": 33710 }, { "grad_norm": 0.18028807640075684, "learning_rate": 7.909554139893511e-05, "loss": 0.0119, "step": 33720 }, { "grad_norm": 0.15019260346889496, "learning_rate": 7.908209292200715e-05, "loss": 0.009, "step": 33730 }, { "grad_norm": 0.12769098579883575, "learning_rate": 7.906864126470523e-05, "loss": 0.0121, "step": 33740 }, { "grad_norm": 0.13923513889312744, "learning_rate": 7.905518642850041e-05, "loss": 0.0108, "step": 33750 }, { "grad_norm": 0.15654374659061432, "learning_rate": 7.904172841486409e-05, "loss": 0.011, "step": 33760 }, { "grad_norm": 0.12612292170524597, "learning_rate": 7.902826722526801e-05, "loss": 0.0098, "step": 33770 }, { "grad_norm": 0.1643865555524826, "learning_rate": 7.901480286118427e-05, "loss": 0.0128, "step": 33780 }, { "grad_norm": 0.16023068130016327, "learning_rate": 7.900133532408531e-05, "loss": 0.012, "step": 33790 }, { "grad_norm": 0.14435240626335144, "learning_rate": 7.898786461544395e-05, "loss": 0.0113, "step": 33800 }, { "grad_norm": 0.11031337827444077, "learning_rate": 7.897439073673325e-05, "loss": 0.0113, "step": 33810 }, { "grad_norm": 0.11919629573822021, "learning_rate": 7.896091368942677e-05, "loss": 0.0095, "step": 33820 }, { "grad_norm": 0.22036591172218323, "learning_rate": 7.894743347499832e-05, "loss": 0.0109, "step": 33830 }, { "grad_norm": 0.14461186528205872, "learning_rate": 7.893395009492203e-05, "loss": 0.0111, "step": 33840 }, { "grad_norm": 0.1359805017709732, "learning_rate": 7.892046355067248e-05, "loss": 0.011, "step": 33850 }, { "grad_norm": 0.1421215981245041, "learning_rate": 7.890697384372451e-05, "loss": 0.0125, "step": 33860 }, { "grad_norm": 0.1175529807806015, "learning_rate": 7.889348097555336e-05, "loss": 0.0131, "step": 33870 }, { "grad_norm": 0.1477239578962326, "learning_rate": 7.887998494763455e-05, "loss": 0.0108, "step": 33880 }, { "grad_norm": 0.15932422876358032, "learning_rate": 7.886648576144404e-05, "loss": 0.0138, "step": 33890 }, { "grad_norm": 0.16494467854499817, "learning_rate": 7.885298341845802e-05, "loss": 0.0128, "step": 33900 }, { "grad_norm": 0.15969157218933105, "learning_rate": 7.883947792015311e-05, "loss": 0.0116, "step": 33910 }, { "grad_norm": 0.15213799476623535, "learning_rate": 7.882596926800628e-05, "loss": 0.0118, "step": 33920 }, { "grad_norm": 0.13587099313735962, "learning_rate": 7.881245746349477e-05, "loss": 0.0116, "step": 33930 }, { "grad_norm": 0.17782653868198395, "learning_rate": 7.879894250809623e-05, "loss": 0.0138, "step": 33940 }, { "grad_norm": 0.17340359091758728, "learning_rate": 7.878542440328865e-05, "loss": 0.0131, "step": 33950 }, { "grad_norm": 0.12769386172294617, "learning_rate": 7.877190315055031e-05, "loss": 0.0122, "step": 33960 }, { "grad_norm": 0.13291224837303162, "learning_rate": 7.875837875135991e-05, "loss": 0.0123, "step": 33970 }, { "grad_norm": 0.1660536676645279, "learning_rate": 7.874485120719646e-05, "loss": 0.0124, "step": 33980 }, { "grad_norm": 0.198098823428154, "learning_rate": 7.873132051953928e-05, "loss": 0.0124, "step": 33990 }, { "grad_norm": 0.1287168562412262, "learning_rate": 7.87177866898681e-05, "loss": 0.0115, "step": 34000 }, { "grad_norm": 0.15384098887443542, "learning_rate": 7.870424971966294e-05, "loss": 0.0122, "step": 34010 }, { "grad_norm": 0.10043301433324814, "learning_rate": 7.869070961040419e-05, "loss": 0.0103, "step": 34020 }, { "grad_norm": 0.15991076827049255, "learning_rate": 7.867716636357257e-05, "loss": 0.0127, "step": 34030 }, { "grad_norm": 0.1604485809803009, "learning_rate": 7.866361998064915e-05, "loss": 0.0112, "step": 34040 }, { "grad_norm": 0.19076235592365265, "learning_rate": 7.865007046311534e-05, "loss": 0.0134, "step": 34050 }, { "grad_norm": 0.2297777533531189, "learning_rate": 7.86365178124529e-05, "loss": 0.012, "step": 34060 }, { "grad_norm": 0.2252601981163025, "learning_rate": 7.862296203014394e-05, "loss": 0.0099, "step": 34070 }, { "grad_norm": 0.10669116675853729, "learning_rate": 7.860940311767088e-05, "loss": 0.0102, "step": 34080 }, { "grad_norm": 0.14347150921821594, "learning_rate": 7.85958410765165e-05, "loss": 0.0125, "step": 34090 }, { "grad_norm": 0.12829945981502533, "learning_rate": 7.858227590816394e-05, "loss": 0.0099, "step": 34100 }, { "grad_norm": 0.1254660189151764, "learning_rate": 7.856870761409664e-05, "loss": 0.0121, "step": 34110 }, { "grad_norm": 0.15435108542442322, "learning_rate": 7.855513619579846e-05, "loss": 0.0115, "step": 34120 }, { "grad_norm": 0.21605230867862701, "learning_rate": 7.85415616547535e-05, "loss": 0.0126, "step": 34130 }, { "grad_norm": 0.15806248784065247, "learning_rate": 7.852798399244627e-05, "loss": 0.014, "step": 34140 }, { "grad_norm": 0.13898658752441406, "learning_rate": 7.851440321036161e-05, "loss": 0.0107, "step": 34150 }, { "grad_norm": 0.10457503795623779, "learning_rate": 7.850081930998468e-05, "loss": 0.011, "step": 34160 }, { "grad_norm": 0.16282548010349274, "learning_rate": 7.8487232292801e-05, "loss": 0.0107, "step": 34170 }, { "grad_norm": 0.1561887264251709, "learning_rate": 7.847364216029642e-05, "loss": 0.012, "step": 34180 }, { "grad_norm": 0.16667014360427856, "learning_rate": 7.846004891395716e-05, "loss": 0.0107, "step": 34190 }, { "grad_norm": 0.13883937895298004, "learning_rate": 7.844645255526972e-05, "loss": 0.0135, "step": 34200 }, { "grad_norm": 0.14150437712669373, "learning_rate": 7.843285308572101e-05, "loss": 0.011, "step": 34210 }, { "grad_norm": 0.1488899439573288, "learning_rate": 7.841925050679823e-05, "loss": 0.01, "step": 34220 }, { "grad_norm": 0.13702212274074554, "learning_rate": 7.840564481998895e-05, "loss": 0.0132, "step": 34230 }, { "grad_norm": 0.1504431664943695, "learning_rate": 7.839203602678105e-05, "loss": 0.0103, "step": 34240 }, { "grad_norm": 0.20448416471481323, "learning_rate": 7.837842412866279e-05, "loss": 0.0106, "step": 34250 }, { "grad_norm": 0.09248075634241104, "learning_rate": 7.836480912712272e-05, "loss": 0.0111, "step": 34260 }, { "grad_norm": 0.11210644990205765, "learning_rate": 7.835119102364976e-05, "loss": 0.0108, "step": 34270 }, { "grad_norm": 0.16422156989574432, "learning_rate": 7.83375698197332e-05, "loss": 0.0119, "step": 34280 }, { "grad_norm": 0.12679395079612732, "learning_rate": 7.83239455168626e-05, "loss": 0.0109, "step": 34290 }, { "grad_norm": 0.13431809842586517, "learning_rate": 7.83103181165279e-05, "loss": 0.0125, "step": 34300 }, { "grad_norm": 0.21357543766498566, "learning_rate": 7.829668762021937e-05, "loss": 0.0133, "step": 34310 }, { "grad_norm": 0.18224744498729706, "learning_rate": 7.828305402942764e-05, "loss": 0.0124, "step": 34320 }, { "grad_norm": 0.21659961342811584, "learning_rate": 7.826941734564363e-05, "loss": 0.0138, "step": 34330 }, { "grad_norm": 0.23615416884422302, "learning_rate": 7.825577757035865e-05, "loss": 0.0149, "step": 34340 }, { "grad_norm": 0.1796749085187912, "learning_rate": 7.824213470506431e-05, "loss": 0.0135, "step": 34350 }, { "grad_norm": 0.1228233277797699, "learning_rate": 7.822848875125257e-05, "loss": 0.0118, "step": 34360 }, { "grad_norm": 0.14872558414936066, "learning_rate": 7.821483971041576e-05, "loss": 0.0107, "step": 34370 }, { "grad_norm": 0.13254648447036743, "learning_rate": 7.820118758404649e-05, "loss": 0.0108, "step": 34380 }, { "grad_norm": 0.12420923262834549, "learning_rate": 7.818753237363776e-05, "loss": 0.0117, "step": 34390 }, { "grad_norm": 0.12184958159923553, "learning_rate": 7.817387408068286e-05, "loss": 0.0117, "step": 34400 }, { "grad_norm": 0.13922566175460815, "learning_rate": 7.816021270667544e-05, "loss": 0.0133, "step": 34410 }, { "grad_norm": 0.1373566836118698, "learning_rate": 7.81465482531095e-05, "loss": 0.0103, "step": 34420 }, { "grad_norm": 0.14207789301872253, "learning_rate": 7.813288072147938e-05, "loss": 0.0098, "step": 34430 }, { "grad_norm": 0.12786561250686646, "learning_rate": 7.811921011327972e-05, "loss": 0.011, "step": 34440 }, { "grad_norm": 0.11982887983322144, "learning_rate": 7.810553643000549e-05, "loss": 0.0098, "step": 34450 }, { "grad_norm": 0.1313762366771698, "learning_rate": 7.809185967315206e-05, "loss": 0.0097, "step": 34460 }, { "grad_norm": 0.1718865931034088, "learning_rate": 7.80781798442151e-05, "loss": 0.0106, "step": 34470 }, { "grad_norm": 0.20974299311637878, "learning_rate": 7.806449694469061e-05, "loss": 0.011, "step": 34480 }, { "grad_norm": 0.1428612470626831, "learning_rate": 7.805081097607492e-05, "loss": 0.0104, "step": 34490 }, { "grad_norm": 0.1609840989112854, "learning_rate": 7.803712193986474e-05, "loss": 0.0137, "step": 34500 }, { "grad_norm": 0.16343477368354797, "learning_rate": 7.802342983755702e-05, "loss": 0.0122, "step": 34510 }, { "grad_norm": 0.12158852815628052, "learning_rate": 7.800973467064918e-05, "loss": 0.0102, "step": 34520 }, { "grad_norm": 0.2957444190979004, "learning_rate": 7.799603644063884e-05, "loss": 0.0126, "step": 34530 }, { "grad_norm": 0.23376312851905823, "learning_rate": 7.798233514902405e-05, "loss": 0.0168, "step": 34540 }, { "grad_norm": 0.13508401811122894, "learning_rate": 7.796863079730318e-05, "loss": 0.0143, "step": 34550 }, { "grad_norm": 0.1772470474243164, "learning_rate": 7.795492338697488e-05, "loss": 0.0128, "step": 34560 }, { "grad_norm": 0.1450004279613495, "learning_rate": 7.794121291953819e-05, "loss": 0.0112, "step": 34570 }, { "grad_norm": 0.10506004840135574, "learning_rate": 7.792749939649246e-05, "loss": 0.0107, "step": 34580 }, { "grad_norm": 0.17977941036224365, "learning_rate": 7.79137828193374e-05, "loss": 0.0134, "step": 34590 }, { "grad_norm": 0.1330857276916504, "learning_rate": 7.790006318957301e-05, "loss": 0.0136, "step": 34600 }, { "grad_norm": 0.13932077586650848, "learning_rate": 7.788634050869965e-05, "loss": 0.011, "step": 34610 }, { "grad_norm": 0.13654087483882904, "learning_rate": 7.787261477821803e-05, "loss": 0.0113, "step": 34620 }, { "grad_norm": 0.133998841047287, "learning_rate": 7.785888599962916e-05, "loss": 0.0104, "step": 34630 }, { "grad_norm": 0.16328461468219757, "learning_rate": 7.784515417443439e-05, "loss": 0.0103, "step": 34640 }, { "grad_norm": 0.1344054490327835, "learning_rate": 7.783141930413545e-05, "loss": 0.0113, "step": 34650 }, { "grad_norm": 0.15222002565860748, "learning_rate": 7.78176813902343e-05, "loss": 0.0114, "step": 34660 }, { "grad_norm": 0.13218648731708527, "learning_rate": 7.780394043423336e-05, "loss": 0.0101, "step": 34670 }, { "grad_norm": 0.1539812833070755, "learning_rate": 7.77901964376353e-05, "loss": 0.0112, "step": 34680 }, { "grad_norm": 0.1521044373512268, "learning_rate": 7.777644940194316e-05, "loss": 0.0102, "step": 34690 }, { "grad_norm": 0.1299661546945572, "learning_rate": 7.776269932866023e-05, "loss": 0.0112, "step": 34700 }, { "grad_norm": 0.17472447454929352, "learning_rate": 7.774894621929026e-05, "loss": 0.0101, "step": 34710 }, { "grad_norm": 0.10899759829044342, "learning_rate": 7.773519007533725e-05, "loss": 0.0112, "step": 34720 }, { "grad_norm": 0.12867052853107452, "learning_rate": 7.772143089830556e-05, "loss": 0.0107, "step": 34730 }, { "grad_norm": 0.1619603931903839, "learning_rate": 7.770766868969985e-05, "loss": 0.0129, "step": 34740 }, { "grad_norm": 0.17893965542316437, "learning_rate": 7.769390345102518e-05, "loss": 0.0125, "step": 34750 }, { "grad_norm": 0.21391339600086212, "learning_rate": 7.768013518378683e-05, "loss": 0.0122, "step": 34760 }, { "grad_norm": 0.16474007070064545, "learning_rate": 7.766636388949053e-05, "loss": 0.0113, "step": 34770 }, { "grad_norm": 0.13967272639274597, "learning_rate": 7.765258956964229e-05, "loss": 0.0111, "step": 34780 }, { "grad_norm": 0.1182127520442009, "learning_rate": 7.76388122257484e-05, "loss": 0.0124, "step": 34790 }, { "grad_norm": 0.16504347324371338, "learning_rate": 7.762503185931558e-05, "loss": 0.0105, "step": 34800 }, { "grad_norm": 0.16046015918254852, "learning_rate": 7.76112484718508e-05, "loss": 0.0126, "step": 34810 }, { "grad_norm": 0.16330182552337646, "learning_rate": 7.75974620648614e-05, "loss": 0.0107, "step": 34820 }, { "grad_norm": 0.17728254199028015, "learning_rate": 7.758367263985503e-05, "loss": 0.011, "step": 34830 }, { "grad_norm": 0.1696014255285263, "learning_rate": 7.75698801983397e-05, "loss": 0.0098, "step": 34840 }, { "grad_norm": 0.09349232912063599, "learning_rate": 7.755608474182372e-05, "loss": 0.0117, "step": 34850 }, { "grad_norm": 0.16289322078227997, "learning_rate": 7.754228627181574e-05, "loss": 0.012, "step": 34860 }, { "grad_norm": 0.17023411393165588, "learning_rate": 7.752848478982476e-05, "loss": 0.0109, "step": 34870 }, { "grad_norm": 0.11833900213241577, "learning_rate": 7.751468029736006e-05, "loss": 0.0119, "step": 34880 }, { "grad_norm": 0.13342244923114777, "learning_rate": 7.750087279593129e-05, "loss": 0.0115, "step": 34890 }, { "grad_norm": 0.13755573332309723, "learning_rate": 7.748706228704843e-05, "loss": 0.01, "step": 34900 }, { "grad_norm": 0.1375216543674469, "learning_rate": 7.747324877222176e-05, "loss": 0.011, "step": 34910 }, { "grad_norm": 0.17984838783740997, "learning_rate": 7.745943225296188e-05, "loss": 0.0109, "step": 34920 }, { "grad_norm": 0.12811215221881866, "learning_rate": 7.744561273077981e-05, "loss": 0.0094, "step": 34930 }, { "grad_norm": 0.1923747956752777, "learning_rate": 7.743179020718678e-05, "loss": 0.0114, "step": 34940 }, { "grad_norm": 0.14157232642173767, "learning_rate": 7.741796468369443e-05, "loss": 0.0105, "step": 34950 }, { "grad_norm": 0.13804744184017181, "learning_rate": 7.740413616181466e-05, "loss": 0.0095, "step": 34960 }, { "grad_norm": 0.10709305852651596, "learning_rate": 7.739030464305978e-05, "loss": 0.0123, "step": 34970 }, { "grad_norm": 0.12273090332746506, "learning_rate": 7.737647012894235e-05, "loss": 0.0095, "step": 34980 }, { "grad_norm": 0.14762462675571442, "learning_rate": 7.736263262097532e-05, "loss": 0.0109, "step": 34990 }, { "grad_norm": 0.2231183797121048, "learning_rate": 7.734879212067192e-05, "loss": 0.0113, "step": 35000 }, { "grad_norm": 0.17017905414104462, "learning_rate": 7.733494862954573e-05, "loss": 0.0104, "step": 35010 }, { "grad_norm": 0.18139766156673431, "learning_rate": 7.732110214911066e-05, "loss": 0.011, "step": 35020 }, { "grad_norm": 0.15973511338233948, "learning_rate": 7.730725268088094e-05, "loss": 0.0116, "step": 35030 }, { "grad_norm": 0.1795911341905594, "learning_rate": 7.729340022637112e-05, "loss": 0.0119, "step": 35040 }, { "grad_norm": 0.1281694620847702, "learning_rate": 7.727954478709607e-05, "loss": 0.0138, "step": 35050 }, { "grad_norm": 0.12219791859388351, "learning_rate": 7.726568636457103e-05, "loss": 0.0114, "step": 35060 }, { "grad_norm": 0.15979886054992676, "learning_rate": 7.725182496031153e-05, "loss": 0.0126, "step": 35070 }, { "grad_norm": 0.17359767854213715, "learning_rate": 7.723796057583341e-05, "loss": 0.0126, "step": 35080 }, { "grad_norm": 0.1337163746356964, "learning_rate": 7.722409321265287e-05, "loss": 0.0126, "step": 35090 }, { "grad_norm": 0.13413600623607635, "learning_rate": 7.721022287228645e-05, "loss": 0.0117, "step": 35100 }, { "grad_norm": 0.1560927778482437, "learning_rate": 7.719634955625094e-05, "loss": 0.0106, "step": 35110 }, { "grad_norm": 0.1479455679655075, "learning_rate": 7.718247326606355e-05, "loss": 0.0111, "step": 35120 }, { "grad_norm": 0.15838703513145447, "learning_rate": 7.716859400324175e-05, "loss": 0.009, "step": 35130 }, { "grad_norm": 0.16036692261695862, "learning_rate": 7.715471176930335e-05, "loss": 0.0105, "step": 35140 }, { "grad_norm": 0.13775116205215454, "learning_rate": 7.714082656576651e-05, "loss": 0.0102, "step": 35150 }, { "grad_norm": 0.13460373878479004, "learning_rate": 7.712693839414968e-05, "loss": 0.0127, "step": 35160 }, { "grad_norm": 0.13452446460723877, "learning_rate": 7.711304725597164e-05, "loss": 0.0089, "step": 35170 }, { "grad_norm": 0.15906961262226105, "learning_rate": 7.709915315275151e-05, "loss": 0.0118, "step": 35180 }, { "grad_norm": 0.14757758378982544, "learning_rate": 7.708525608600876e-05, "loss": 0.0101, "step": 35190 }, { "grad_norm": 0.10706810653209686, "learning_rate": 7.707135605726311e-05, "loss": 0.0123, "step": 35200 }, { "grad_norm": 0.18983244895935059, "learning_rate": 7.705745306803466e-05, "loss": 0.0119, "step": 35210 }, { "grad_norm": 0.15149502456188202, "learning_rate": 7.704354711984383e-05, "loss": 0.0121, "step": 35220 }, { "grad_norm": 0.12077837437391281, "learning_rate": 7.702963821421133e-05, "loss": 0.0117, "step": 35230 }, { "grad_norm": 0.13382947444915771, "learning_rate": 7.701572635265826e-05, "loss": 0.0132, "step": 35240 }, { "grad_norm": 0.15849682688713074, "learning_rate": 7.700181153670596e-05, "loss": 0.0092, "step": 35250 }, { "grad_norm": 0.13148730993270874, "learning_rate": 7.698789376787616e-05, "loss": 0.0107, "step": 35260 }, { "grad_norm": 0.12273906916379929, "learning_rate": 7.697397304769085e-05, "loss": 0.0094, "step": 35270 }, { "grad_norm": 0.11442413926124573, "learning_rate": 7.696004937767241e-05, "loss": 0.0138, "step": 35280 }, { "grad_norm": 0.12113160640001297, "learning_rate": 7.694612275934352e-05, "loss": 0.011, "step": 35290 }, { "grad_norm": 0.17893429100513458, "learning_rate": 7.693219319422714e-05, "loss": 0.0124, "step": 35300 }, { "grad_norm": 0.13876789808273315, "learning_rate": 7.69182606838466e-05, "loss": 0.0109, "step": 35310 }, { "grad_norm": 0.16825594007968903, "learning_rate": 7.690432522972558e-05, "loss": 0.0112, "step": 35320 }, { "grad_norm": 0.14696142077445984, "learning_rate": 7.689038683338796e-05, "loss": 0.0118, "step": 35330 }, { "grad_norm": 0.14389288425445557, "learning_rate": 7.687644549635808e-05, "loss": 0.0106, "step": 35340 }, { "grad_norm": 0.11662405729293823, "learning_rate": 7.686250122016053e-05, "loss": 0.0111, "step": 35350 }, { "grad_norm": 0.16543936729431152, "learning_rate": 7.684855400632022e-05, "loss": 0.0115, "step": 35360 }, { "grad_norm": 0.13331367075443268, "learning_rate": 7.683460385636243e-05, "loss": 0.0124, "step": 35370 }, { "grad_norm": 0.22421403229236603, "learning_rate": 7.68206507718127e-05, "loss": 0.0114, "step": 35380 }, { "grad_norm": 0.1700383424758911, "learning_rate": 7.680669475419693e-05, "loss": 0.0098, "step": 35390 }, { "grad_norm": 0.14008015394210815, "learning_rate": 7.679273580504132e-05, "loss": 0.0119, "step": 35400 }, { "grad_norm": 0.18222753703594208, "learning_rate": 7.67787739258724e-05, "loss": 0.0128, "step": 35410 }, { "grad_norm": 0.17727544903755188, "learning_rate": 7.676480911821705e-05, "loss": 0.0116, "step": 35420 }, { "grad_norm": 0.16433009505271912, "learning_rate": 7.675084138360238e-05, "loss": 0.0107, "step": 35430 }, { "grad_norm": 0.13170213997364044, "learning_rate": 7.673687072355592e-05, "loss": 0.0108, "step": 35440 }, { "grad_norm": 0.16612687706947327, "learning_rate": 7.67228971396055e-05, "loss": 0.0108, "step": 35450 }, { "grad_norm": 0.14997176826000214, "learning_rate": 7.670892063327922e-05, "loss": 0.0102, "step": 35460 }, { "grad_norm": 0.1091354712843895, "learning_rate": 7.669494120610552e-05, "loss": 0.0111, "step": 35470 }, { "grad_norm": 0.13665954768657684, "learning_rate": 7.66809588596132e-05, "loss": 0.0091, "step": 35480 }, { "grad_norm": 0.146434023976326, "learning_rate": 7.666697359533132e-05, "loss": 0.0101, "step": 35490 }, { "grad_norm": 0.12278398126363754, "learning_rate": 7.665298541478932e-05, "loss": 0.0098, "step": 35500 }, { "grad_norm": 0.13557763397693634, "learning_rate": 7.66389943195169e-05, "loss": 0.0101, "step": 35510 }, { "grad_norm": 0.15046048164367676, "learning_rate": 7.66250003110441e-05, "loss": 0.0124, "step": 35520 }, { "grad_norm": 0.1451561599969864, "learning_rate": 7.66110033909013e-05, "loss": 0.01, "step": 35530 }, { "grad_norm": 0.16074945032596588, "learning_rate": 7.659700356061918e-05, "loss": 0.0114, "step": 35540 }, { "grad_norm": 0.12644363939762115, "learning_rate": 7.658300082172875e-05, "loss": 0.01, "step": 35550 }, { "grad_norm": 0.14639979600906372, "learning_rate": 7.65689951757613e-05, "loss": 0.0114, "step": 35560 }, { "grad_norm": 0.12367402017116547, "learning_rate": 7.655498662424849e-05, "loss": 0.0091, "step": 35570 }, { "grad_norm": 0.22325140237808228, "learning_rate": 7.654097516872227e-05, "loss": 0.0111, "step": 35580 }, { "grad_norm": 0.15956485271453857, "learning_rate": 7.65269608107149e-05, "loss": 0.0106, "step": 35590 }, { "grad_norm": 0.18038439750671387, "learning_rate": 7.6512943551759e-05, "loss": 0.0121, "step": 35600 }, { "grad_norm": 0.13512718677520752, "learning_rate": 7.649892339338743e-05, "loss": 0.0111, "step": 35610 }, { "grad_norm": 0.12366130948066711, "learning_rate": 7.648490033713344e-05, "loss": 0.0097, "step": 35620 }, { "grad_norm": 0.12125259637832642, "learning_rate": 7.647087438453058e-05, "loss": 0.0094, "step": 35630 }, { "grad_norm": 0.1175665631890297, "learning_rate": 7.64568455371127e-05, "loss": 0.0107, "step": 35640 }, { "grad_norm": 0.15542946755886078, "learning_rate": 7.644281379641396e-05, "loss": 0.0106, "step": 35650 }, { "grad_norm": 0.16877663135528564, "learning_rate": 7.642877916396887e-05, "loss": 0.0106, "step": 35660 }, { "grad_norm": 0.1419563591480255, "learning_rate": 7.641474164131221e-05, "loss": 0.0113, "step": 35670 }, { "grad_norm": 0.14016549289226532, "learning_rate": 7.640070122997913e-05, "loss": 0.0103, "step": 35680 }, { "grad_norm": 0.1301950067281723, "learning_rate": 7.638665793150505e-05, "loss": 0.0106, "step": 35690 }, { "grad_norm": 0.14422248303890228, "learning_rate": 7.637261174742574e-05, "loss": 0.0105, "step": 35700 }, { "grad_norm": 0.13672122359275818, "learning_rate": 7.635856267927725e-05, "loss": 0.0112, "step": 35710 }, { "grad_norm": 0.11223320662975311, "learning_rate": 7.634451072859597e-05, "loss": 0.0117, "step": 35720 }, { "grad_norm": 0.11870058625936508, "learning_rate": 7.633045589691863e-05, "loss": 0.0092, "step": 35730 }, { "grad_norm": 0.11994944512844086, "learning_rate": 7.63163981857822e-05, "loss": 0.0104, "step": 35740 }, { "grad_norm": 0.16595995426177979, "learning_rate": 7.630233759672403e-05, "loss": 0.0106, "step": 35750 }, { "grad_norm": 0.1523841768503189, "learning_rate": 7.628827413128177e-05, "loss": 0.0096, "step": 35760 }, { "grad_norm": 0.11077187955379486, "learning_rate": 7.627420779099338e-05, "loss": 0.0099, "step": 35770 }, { "grad_norm": 0.1723436862230301, "learning_rate": 7.626013857739711e-05, "loss": 0.0099, "step": 35780 }, { "grad_norm": 0.11113743484020233, "learning_rate": 7.624606649203158e-05, "loss": 0.0091, "step": 35790 }, { "grad_norm": 0.15706384181976318, "learning_rate": 7.623199153643569e-05, "loss": 0.0101, "step": 35800 }, { "grad_norm": 0.15566758811473846, "learning_rate": 7.621791371214863e-05, "loss": 0.0105, "step": 35810 }, { "grad_norm": 0.13932378590106964, "learning_rate": 7.620383302070995e-05, "loss": 0.0107, "step": 35820 }, { "grad_norm": 0.19229398667812347, "learning_rate": 7.61897494636595e-05, "loss": 0.0092, "step": 35830 }, { "grad_norm": 0.17592935264110565, "learning_rate": 7.617566304253739e-05, "loss": 0.0095, "step": 35840 }, { "grad_norm": 0.20315086841583252, "learning_rate": 7.616157375888416e-05, "loss": 0.0105, "step": 35850 }, { "grad_norm": 0.15984544157981873, "learning_rate": 7.614748161424053e-05, "loss": 0.0101, "step": 35860 }, { "grad_norm": 0.20877669751644135, "learning_rate": 7.613338661014763e-05, "loss": 0.0122, "step": 35870 }, { "grad_norm": 0.13708248734474182, "learning_rate": 7.611928874814686e-05, "loss": 0.0099, "step": 35880 }, { "grad_norm": 0.13732680678367615, "learning_rate": 7.610518802977996e-05, "loss": 0.0097, "step": 35890 }, { "grad_norm": 0.11649176478385925, "learning_rate": 7.609108445658893e-05, "loss": 0.0107, "step": 35900 }, { "grad_norm": 0.18081092834472656, "learning_rate": 7.607697803011612e-05, "loss": 0.0101, "step": 35910 }, { "grad_norm": 0.10043992102146149, "learning_rate": 7.606286875190421e-05, "loss": 0.0121, "step": 35920 }, { "grad_norm": 0.1376436948776245, "learning_rate": 7.604875662349617e-05, "loss": 0.0109, "step": 35930 }, { "grad_norm": 0.13415907323360443, "learning_rate": 7.603464164643525e-05, "loss": 0.0107, "step": 35940 }, { "grad_norm": 0.13561126589775085, "learning_rate": 7.602052382226507e-05, "loss": 0.0108, "step": 35950 }, { "grad_norm": 0.15267148613929749, "learning_rate": 7.600640315252954e-05, "loss": 0.0094, "step": 35960 }, { "grad_norm": 0.12957975268363953, "learning_rate": 7.599227963877284e-05, "loss": 0.0127, "step": 35970 }, { "grad_norm": 0.18379555642604828, "learning_rate": 7.597815328253954e-05, "loss": 0.0125, "step": 35980 }, { "grad_norm": 0.14158572256565094, "learning_rate": 7.596402408537443e-05, "loss": 0.0102, "step": 35990 }, { "grad_norm": 0.18544574081897736, "learning_rate": 7.59498920488227e-05, "loss": 0.0129, "step": 36000 }, { "grad_norm": 0.16106607019901276, "learning_rate": 7.593575717442979e-05, "loss": 0.0091, "step": 36010 }, { "grad_norm": 0.17095458507537842, "learning_rate": 7.592161946374147e-05, "loss": 0.0097, "step": 36020 }, { "grad_norm": 0.10579903423786163, "learning_rate": 7.590747891830381e-05, "loss": 0.0098, "step": 36030 }, { "grad_norm": 0.13750064373016357, "learning_rate": 7.58933355396632e-05, "loss": 0.0097, "step": 36040 }, { "grad_norm": 0.13704639673233032, "learning_rate": 7.587918932936636e-05, "loss": 0.0102, "step": 36050 }, { "grad_norm": 0.14654622972011566, "learning_rate": 7.586504028896028e-05, "loss": 0.0106, "step": 36060 }, { "grad_norm": 0.13749848306179047, "learning_rate": 7.585088841999228e-05, "loss": 0.0107, "step": 36070 }, { "grad_norm": 0.12182675302028656, "learning_rate": 7.583673372400999e-05, "loss": 0.0116, "step": 36080 }, { "grad_norm": 0.14949962496757507, "learning_rate": 7.582257620256134e-05, "loss": 0.0117, "step": 36090 }, { "grad_norm": 0.15495535731315613, "learning_rate": 7.580841585719458e-05, "loss": 0.0114, "step": 36100 }, { "grad_norm": 0.20467418432235718, "learning_rate": 7.579425268945825e-05, "loss": 0.011, "step": 36110 }, { "grad_norm": 0.13495561480522156, "learning_rate": 7.578008670090127e-05, "loss": 0.009, "step": 36120 }, { "grad_norm": 0.1404642015695572, "learning_rate": 7.576591789307272e-05, "loss": 0.0092, "step": 36130 }, { "grad_norm": 0.12789955735206604, "learning_rate": 7.575174626752216e-05, "loss": 0.0097, "step": 36140 }, { "grad_norm": 0.14540764689445496, "learning_rate": 7.573757182579934e-05, "loss": 0.0105, "step": 36150 }, { "grad_norm": 0.11107618361711502, "learning_rate": 7.572339456945435e-05, "loss": 0.0116, "step": 36160 }, { "grad_norm": 0.1946449875831604, "learning_rate": 7.570921450003762e-05, "loss": 0.012, "step": 36170 }, { "grad_norm": 0.1559486836194992, "learning_rate": 7.569503161909983e-05, "loss": 0.013, "step": 36180 }, { "grad_norm": 0.21455422043800354, "learning_rate": 7.568084592819202e-05, "loss": 0.0117, "step": 36190 }, { "grad_norm": 0.1428179293870926, "learning_rate": 7.566665742886551e-05, "loss": 0.0113, "step": 36200 }, { "grad_norm": 0.11145174503326416, "learning_rate": 7.565246612267194e-05, "loss": 0.011, "step": 36210 }, { "grad_norm": 0.13562309741973877, "learning_rate": 7.563827201116325e-05, "loss": 0.0113, "step": 36220 }, { "grad_norm": 0.16962866485118866, "learning_rate": 7.562407509589164e-05, "loss": 0.011, "step": 36230 }, { "grad_norm": 0.16603614389896393, "learning_rate": 7.560987537840973e-05, "loss": 0.0129, "step": 36240 }, { "grad_norm": 0.20123232901096344, "learning_rate": 7.559567286027036e-05, "loss": 0.0116, "step": 36250 }, { "grad_norm": 0.16488224267959595, "learning_rate": 7.558146754302668e-05, "loss": 0.0099, "step": 36260 }, { "grad_norm": 0.15909087657928467, "learning_rate": 7.556725942823217e-05, "loss": 0.0118, "step": 36270 }, { "grad_norm": 0.13224583864212036, "learning_rate": 7.555304851744061e-05, "loss": 0.0125, "step": 36280 }, { "grad_norm": 0.1265774965286255, "learning_rate": 7.553883481220608e-05, "loss": 0.0145, "step": 36290 }, { "grad_norm": 0.11812838166952133, "learning_rate": 7.552461831408298e-05, "loss": 0.0111, "step": 36300 }, { "grad_norm": 0.18223567306995392, "learning_rate": 7.551039902462599e-05, "loss": 0.0103, "step": 36310 }, { "grad_norm": 0.18947860598564148, "learning_rate": 7.549617694539011e-05, "loss": 0.0088, "step": 36320 }, { "grad_norm": 0.18640057742595673, "learning_rate": 7.548195207793066e-05, "loss": 0.0118, "step": 36330 }, { "grad_norm": 0.14255528151988983, "learning_rate": 7.546772442380323e-05, "loss": 0.0105, "step": 36340 }, { "grad_norm": 0.12747013568878174, "learning_rate": 7.545349398456376e-05, "loss": 0.0114, "step": 36350 }, { "grad_norm": 0.10754729062318802, "learning_rate": 7.543926076176845e-05, "loss": 0.0101, "step": 36360 }, { "grad_norm": 0.14265291392803192, "learning_rate": 7.542502475697385e-05, "loss": 0.0122, "step": 36370 }, { "grad_norm": 0.13887472450733185, "learning_rate": 7.541078597173675e-05, "loss": 0.0121, "step": 36380 }, { "grad_norm": 0.1502595990896225, "learning_rate": 7.539654440761431e-05, "loss": 0.0144, "step": 36390 }, { "grad_norm": 0.16021116077899933, "learning_rate": 7.538230006616395e-05, "loss": 0.0132, "step": 36400 }, { "grad_norm": 0.11503848433494568, "learning_rate": 7.536805294894342e-05, "loss": 0.0102, "step": 36410 }, { "grad_norm": 0.1554562896490097, "learning_rate": 7.535380305751076e-05, "loss": 0.0108, "step": 36420 }, { "grad_norm": 0.10708872973918915, "learning_rate": 7.533955039342431e-05, "loss": 0.0094, "step": 36430 }, { "grad_norm": 0.1476583033800125, "learning_rate": 7.532529495824274e-05, "loss": 0.0129, "step": 36440 }, { "grad_norm": 0.132149338722229, "learning_rate": 7.531103675352497e-05, "loss": 0.0107, "step": 36450 }, { "grad_norm": 0.16411609947681427, "learning_rate": 7.529677578083029e-05, "loss": 0.0118, "step": 36460 }, { "grad_norm": 0.13389113545417786, "learning_rate": 7.528251204171823e-05, "loss": 0.0107, "step": 36470 }, { "grad_norm": 0.11684640496969223, "learning_rate": 7.526824553774867e-05, "loss": 0.0117, "step": 36480 }, { "grad_norm": 0.1797723025083542, "learning_rate": 7.525397627048178e-05, "loss": 0.013, "step": 36490 }, { "grad_norm": 0.15680678188800812, "learning_rate": 7.523970424147802e-05, "loss": 0.012, "step": 36500 }, { "grad_norm": 0.12826062738895416, "learning_rate": 7.522542945229813e-05, "loss": 0.0105, "step": 36510 }, { "grad_norm": 0.1411985605955124, "learning_rate": 7.521115190450321e-05, "loss": 0.0106, "step": 36520 }, { "grad_norm": 0.1256747841835022, "learning_rate": 7.519687159965462e-05, "loss": 0.0126, "step": 36530 }, { "grad_norm": 0.15403956174850464, "learning_rate": 7.518258853931403e-05, "loss": 0.01, "step": 36540 }, { "grad_norm": 0.14263184368610382, "learning_rate": 7.516830272504342e-05, "loss": 0.013, "step": 36550 }, { "grad_norm": 0.17609566450119019, "learning_rate": 7.515401415840509e-05, "loss": 0.0106, "step": 36560 }, { "grad_norm": 0.16122747957706451, "learning_rate": 7.513972284096155e-05, "loss": 0.0113, "step": 36570 }, { "grad_norm": 0.16397352516651154, "learning_rate": 7.512542877427574e-05, "loss": 0.013, "step": 36580 }, { "grad_norm": 0.1565074324607849, "learning_rate": 7.51111319599108e-05, "loss": 0.0117, "step": 36590 }, { "grad_norm": 0.12678273022174835, "learning_rate": 7.509683239943024e-05, "loss": 0.0109, "step": 36600 }, { "grad_norm": 0.1311219483613968, "learning_rate": 7.50825300943978e-05, "loss": 0.0103, "step": 36610 }, { "grad_norm": 0.1200203225016594, "learning_rate": 7.50682250463776e-05, "loss": 0.0105, "step": 36620 }, { "grad_norm": 0.13956785202026367, "learning_rate": 7.5053917256934e-05, "loss": 0.0121, "step": 36630 }, { "grad_norm": 0.15416884422302246, "learning_rate": 7.503960672763166e-05, "loss": 0.0122, "step": 36640 }, { "grad_norm": 0.14583218097686768, "learning_rate": 7.502529346003559e-05, "loss": 0.0107, "step": 36650 }, { "grad_norm": 0.1346564143896103, "learning_rate": 7.501097745571106e-05, "loss": 0.0109, "step": 36660 }, { "grad_norm": 0.13046866655349731, "learning_rate": 7.499665871622361e-05, "loss": 0.0098, "step": 36670 }, { "grad_norm": 0.19265645742416382, "learning_rate": 7.498233724313917e-05, "loss": 0.011, "step": 36680 }, { "grad_norm": 0.16276028752326965, "learning_rate": 7.496801303802387e-05, "loss": 0.0099, "step": 36690 }, { "grad_norm": 0.1404310017824173, "learning_rate": 7.495368610244423e-05, "loss": 0.0121, "step": 36700 }, { "grad_norm": 0.12081442028284073, "learning_rate": 7.493935643796697e-05, "loss": 0.0132, "step": 36710 }, { "grad_norm": 0.16339337825775146, "learning_rate": 7.49250240461592e-05, "loss": 0.0091, "step": 36720 }, { "grad_norm": 0.16148538887500763, "learning_rate": 7.491068892858826e-05, "loss": 0.0114, "step": 36730 }, { "grad_norm": 0.14188668131828308, "learning_rate": 7.489635108682184e-05, "loss": 0.0106, "step": 36740 }, { "grad_norm": 0.14290989935398102, "learning_rate": 7.48820105224279e-05, "loss": 0.011, "step": 36750 }, { "grad_norm": 0.14192089438438416, "learning_rate": 7.48676672369747e-05, "loss": 0.0094, "step": 36760 }, { "grad_norm": 0.19001302123069763, "learning_rate": 7.485332123203079e-05, "loss": 0.014, "step": 36770 }, { "grad_norm": 0.1659563034772873, "learning_rate": 7.483897250916504e-05, "loss": 0.0097, "step": 36780 }, { "grad_norm": 0.2107146829366684, "learning_rate": 7.48246210699466e-05, "loss": 0.0128, "step": 36790 }, { "grad_norm": 0.15963679552078247, "learning_rate": 7.481026691594492e-05, "loss": 0.0094, "step": 36800 }, { "grad_norm": 0.14899307489395142, "learning_rate": 7.479591004872975e-05, "loss": 0.0135, "step": 36810 }, { "grad_norm": 0.19212669134140015, "learning_rate": 7.478155046987114e-05, "loss": 0.0097, "step": 36820 }, { "grad_norm": 0.13953708112239838, "learning_rate": 7.476718818093944e-05, "loss": 0.0104, "step": 36830 }, { "grad_norm": 0.1542671024799347, "learning_rate": 7.475282318350527e-05, "loss": 0.0101, "step": 36840 }, { "grad_norm": 0.1713830977678299, "learning_rate": 7.473845547913957e-05, "loss": 0.0097, "step": 36850 }, { "grad_norm": 0.15175840258598328, "learning_rate": 7.472408506941357e-05, "loss": 0.0113, "step": 36860 }, { "grad_norm": 0.18223193287849426, "learning_rate": 7.470971195589881e-05, "loss": 0.0127, "step": 36870 }, { "grad_norm": 0.13081799447536469, "learning_rate": 7.46953361401671e-05, "loss": 0.0106, "step": 36880 }, { "grad_norm": 0.11052925139665604, "learning_rate": 7.468095762379055e-05, "loss": 0.0114, "step": 36890 }, { "grad_norm": 0.14817363023757935, "learning_rate": 7.466657640834158e-05, "loss": 0.0114, "step": 36900 }, { "grad_norm": 0.2188487946987152, "learning_rate": 7.46521924953929e-05, "loss": 0.0097, "step": 36910 }, { "grad_norm": 0.15184703469276428, "learning_rate": 7.463780588651752e-05, "loss": 0.0111, "step": 36920 }, { "grad_norm": 0.1457892656326294, "learning_rate": 7.462341658328873e-05, "loss": 0.0119, "step": 36930 }, { "grad_norm": 0.16729341447353363, "learning_rate": 7.460902458728012e-05, "loss": 0.012, "step": 36940 }, { "grad_norm": 0.12706144154071808, "learning_rate": 7.459462990006558e-05, "loss": 0.0096, "step": 36950 }, { "grad_norm": 0.13792794942855835, "learning_rate": 7.458023252321929e-05, "loss": 0.0098, "step": 36960 }, { "grad_norm": 0.10977089405059814, "learning_rate": 7.456583245831573e-05, "loss": 0.0099, "step": 36970 }, { "grad_norm": 0.16889767348766327, "learning_rate": 7.455142970692968e-05, "loss": 0.01, "step": 36980 }, { "grad_norm": 0.13307319581508636, "learning_rate": 7.453702427063615e-05, "loss": 0.0107, "step": 36990 }, { "grad_norm": 0.13470634818077087, "learning_rate": 7.452261615101057e-05, "loss": 0.0115, "step": 37000 }, { "grad_norm": 0.13023094832897186, "learning_rate": 7.450820534962856e-05, "loss": 0.0114, "step": 37010 }, { "grad_norm": 0.11042971909046173, "learning_rate": 7.449379186806604e-05, "loss": 0.0101, "step": 37020 }, { "grad_norm": 0.1456446796655655, "learning_rate": 7.44793757078993e-05, "loss": 0.0107, "step": 37030 }, { "grad_norm": 0.20125630497932434, "learning_rate": 7.44649568707048e-05, "loss": 0.0131, "step": 37040 }, { "grad_norm": 0.1698841005563736, "learning_rate": 7.445053535805942e-05, "loss": 0.0119, "step": 37050 }, { "grad_norm": 0.18834659457206726, "learning_rate": 7.443611117154024e-05, "loss": 0.0134, "step": 37060 }, { "grad_norm": 0.10577626526355743, "learning_rate": 7.442168431272469e-05, "loss": 0.0118, "step": 37070 }, { "grad_norm": 0.1596524715423584, "learning_rate": 7.440725478319044e-05, "loss": 0.0112, "step": 37080 }, { "grad_norm": 0.14189572632312775, "learning_rate": 7.439282258451553e-05, "loss": 0.0103, "step": 37090 }, { "grad_norm": 0.1325758695602417, "learning_rate": 7.43783877182782e-05, "loss": 0.0108, "step": 37100 }, { "grad_norm": 0.13457711040973663, "learning_rate": 7.436395018605707e-05, "loss": 0.0095, "step": 37110 }, { "grad_norm": 0.14208854734897614, "learning_rate": 7.434950998943094e-05, "loss": 0.0109, "step": 37120 }, { "grad_norm": 0.18461859226226807, "learning_rate": 7.433506712997902e-05, "loss": 0.0121, "step": 37130 }, { "grad_norm": 0.12043607980012894, "learning_rate": 7.432062160928077e-05, "loss": 0.0096, "step": 37140 }, { "grad_norm": 0.1264055073261261, "learning_rate": 7.430617342891588e-05, "loss": 0.01, "step": 37150 }, { "grad_norm": 0.15270020067691803, "learning_rate": 7.429172259046442e-05, "loss": 0.0106, "step": 37160 }, { "grad_norm": 0.11317281424999237, "learning_rate": 7.42772690955067e-05, "loss": 0.009, "step": 37170 }, { "grad_norm": 0.17756184935569763, "learning_rate": 7.426281294562332e-05, "loss": 0.0107, "step": 37180 }, { "grad_norm": 0.12659107148647308, "learning_rate": 7.424835414239519e-05, "loss": 0.013, "step": 37190 }, { "grad_norm": 0.14535324275493622, "learning_rate": 7.423389268740352e-05, "loss": 0.0107, "step": 37200 }, { "grad_norm": 0.10559704899787903, "learning_rate": 7.42194285822298e-05, "loss": 0.0098, "step": 37210 }, { "grad_norm": 0.17022307217121124, "learning_rate": 7.420496182845576e-05, "loss": 0.0113, "step": 37220 }, { "grad_norm": 0.12284902483224869, "learning_rate": 7.419049242766348e-05, "loss": 0.0104, "step": 37230 }, { "grad_norm": 0.11555720865726471, "learning_rate": 7.417602038143533e-05, "loss": 0.01, "step": 37240 }, { "grad_norm": 0.1314229816198349, "learning_rate": 7.416154569135393e-05, "loss": 0.0107, "step": 37250 }, { "grad_norm": 0.1487732082605362, "learning_rate": 7.414706835900222e-05, "loss": 0.0101, "step": 37260 }, { "grad_norm": 0.11168593168258667, "learning_rate": 7.413258838596345e-05, "loss": 0.01, "step": 37270 }, { "grad_norm": 0.1333465874195099, "learning_rate": 7.411810577382106e-05, "loss": 0.0103, "step": 37280 }, { "grad_norm": 0.12300271540880203, "learning_rate": 7.410362052415889e-05, "loss": 0.011, "step": 37290 }, { "grad_norm": 0.16724279522895813, "learning_rate": 7.408913263856102e-05, "loss": 0.0115, "step": 37300 }, { "grad_norm": 0.1419370472431183, "learning_rate": 7.407464211861183e-05, "loss": 0.0118, "step": 37310 }, { "grad_norm": 0.1528281271457672, "learning_rate": 7.406014896589597e-05, "loss": 0.0121, "step": 37320 }, { "grad_norm": 0.1756996214389801, "learning_rate": 7.404565318199838e-05, "loss": 0.0121, "step": 37330 }, { "grad_norm": 0.14163772761821747, "learning_rate": 7.403115476850432e-05, "loss": 0.0126, "step": 37340 }, { "grad_norm": 0.15029984712600708, "learning_rate": 7.401665372699932e-05, "loss": 0.0111, "step": 37350 }, { "grad_norm": 0.13967743515968323, "learning_rate": 7.400215005906916e-05, "loss": 0.0109, "step": 37360 }, { "grad_norm": 0.1459863930940628, "learning_rate": 7.398764376629996e-05, "loss": 0.0101, "step": 37370 }, { "grad_norm": 0.11653845012187958, "learning_rate": 7.39731348502781e-05, "loss": 0.0099, "step": 37380 }, { "grad_norm": 0.1480705887079239, "learning_rate": 7.395862331259028e-05, "loss": 0.011, "step": 37390 }, { "grad_norm": 0.11730021238327026, "learning_rate": 7.394410915482341e-05, "loss": 0.0098, "step": 37400 }, { "grad_norm": 0.16157856583595276, "learning_rate": 7.392959237856476e-05, "loss": 0.0102, "step": 37410 }, { "grad_norm": 0.1158551573753357, "learning_rate": 7.391507298540187e-05, "loss": 0.011, "step": 37420 }, { "grad_norm": 0.11181676387786865, "learning_rate": 7.390055097692257e-05, "loss": 0.0085, "step": 37430 }, { "grad_norm": 0.14327101409435272, "learning_rate": 7.388602635471494e-05, "loss": 0.0094, "step": 37440 }, { "grad_norm": 0.12493198364973068, "learning_rate": 7.387149912036738e-05, "loss": 0.012, "step": 37450 }, { "grad_norm": 0.12968780100345612, "learning_rate": 7.385696927546854e-05, "loss": 0.0108, "step": 37460 }, { "grad_norm": 0.13755778968334198, "learning_rate": 7.384243682160743e-05, "loss": 0.0092, "step": 37470 }, { "grad_norm": 0.11785422265529633, "learning_rate": 7.382790176037329e-05, "loss": 0.0094, "step": 37480 }, { "grad_norm": 0.11702658981084824, "learning_rate": 7.381336409335562e-05, "loss": 0.0097, "step": 37490 }, { "grad_norm": 0.12472257763147354, "learning_rate": 7.379882382214426e-05, "loss": 0.0112, "step": 37500 }, { "grad_norm": 0.14525681734085083, "learning_rate": 7.378428094832931e-05, "loss": 0.0142, "step": 37510 }, { "grad_norm": 0.16371119022369385, "learning_rate": 7.376973547350114e-05, "loss": 0.0121, "step": 37520 }, { "grad_norm": 0.13768422603607178, "learning_rate": 7.375518739925046e-05, "loss": 0.0106, "step": 37530 }, { "grad_norm": 0.2303709089756012, "learning_rate": 7.374063672716818e-05, "loss": 0.0116, "step": 37540 }, { "grad_norm": 0.12230005115270615, "learning_rate": 7.372608345884558e-05, "loss": 0.0097, "step": 37550 }, { "grad_norm": 0.12580399215221405, "learning_rate": 7.371152759587415e-05, "loss": 0.0106, "step": 37560 }, { "grad_norm": 0.26378098130226135, "learning_rate": 7.369696913984572e-05, "loss": 0.0135, "step": 37570 }, { "grad_norm": 0.14797116816043854, "learning_rate": 7.368240809235237e-05, "loss": 0.0103, "step": 37580 }, { "grad_norm": 0.14286890625953674, "learning_rate": 7.366784445498649e-05, "loss": 0.015, "step": 37590 }, { "grad_norm": 0.14388731122016907, "learning_rate": 7.36532782293407e-05, "loss": 0.0135, "step": 37600 }, { "grad_norm": 0.20088954269886017, "learning_rate": 7.363870941700797e-05, "loss": 0.0115, "step": 37610 }, { "grad_norm": 0.1460791379213333, "learning_rate": 7.362413801958152e-05, "loss": 0.0089, "step": 37620 }, { "grad_norm": 0.1503024399280548, "learning_rate": 7.360956403865486e-05, "loss": 0.0112, "step": 37630 }, { "grad_norm": 0.13613581657409668, "learning_rate": 7.359498747582177e-05, "loss": 0.0098, "step": 37640 }, { "grad_norm": 0.14373767375946045, "learning_rate": 7.358040833267634e-05, "loss": 0.0102, "step": 37650 }, { "grad_norm": 0.10927711427211761, "learning_rate": 7.356582661081288e-05, "loss": 0.0089, "step": 37660 }, { "grad_norm": 0.11211799830198288, "learning_rate": 7.355124231182607e-05, "loss": 0.0091, "step": 37670 }, { "grad_norm": 0.12435052543878555, "learning_rate": 7.353665543731083e-05, "loss": 0.0098, "step": 37680 }, { "grad_norm": 0.12881505489349365, "learning_rate": 7.352206598886229e-05, "loss": 0.0109, "step": 37690 }, { "grad_norm": 0.10713236778974533, "learning_rate": 7.350747396807601e-05, "loss": 0.0098, "step": 37700 }, { "grad_norm": 0.13755732774734497, "learning_rate": 7.349287937654772e-05, "loss": 0.0115, "step": 37710 }, { "grad_norm": 0.1243123859167099, "learning_rate": 7.347828221587345e-05, "loss": 0.0095, "step": 37720 }, { "grad_norm": 0.10877574980258942, "learning_rate": 7.346368248764955e-05, "loss": 0.0096, "step": 37730 }, { "grad_norm": 0.1646163910627365, "learning_rate": 7.34490801934726e-05, "loss": 0.0135, "step": 37740 }, { "grad_norm": 0.11910659819841385, "learning_rate": 7.343447533493947e-05, "loss": 0.0104, "step": 37750 }, { "grad_norm": 0.14546334743499756, "learning_rate": 7.341986791364739e-05, "loss": 0.0102, "step": 37760 }, { "grad_norm": 0.15096065402030945, "learning_rate": 7.340525793119374e-05, "loss": 0.0107, "step": 37770 }, { "grad_norm": 0.13174466788768768, "learning_rate": 7.33906453891763e-05, "loss": 0.0109, "step": 37780 }, { "grad_norm": 0.1098717525601387, "learning_rate": 7.337603028919303e-05, "loss": 0.0112, "step": 37790 }, { "grad_norm": 0.13157112896442413, "learning_rate": 7.336141263284225e-05, "loss": 0.0102, "step": 37800 }, { "grad_norm": 0.20019567012786865, "learning_rate": 7.334679242172249e-05, "loss": 0.0107, "step": 37810 }, { "grad_norm": 0.1633611023426056, "learning_rate": 7.33321696574326e-05, "loss": 0.0117, "step": 37820 }, { "grad_norm": 0.1145094633102417, "learning_rate": 7.331754434157174e-05, "loss": 0.0121, "step": 37830 }, { "grad_norm": 0.1367945671081543, "learning_rate": 7.330291647573928e-05, "loss": 0.0105, "step": 37840 }, { "grad_norm": 0.13869412243366241, "learning_rate": 7.32882860615349e-05, "loss": 0.0111, "step": 37850 }, { "grad_norm": 0.12208381295204163, "learning_rate": 7.327365310055858e-05, "loss": 0.0096, "step": 37860 }, { "grad_norm": 0.12298337370157242, "learning_rate": 7.325901759441055e-05, "loss": 0.0105, "step": 37870 }, { "grad_norm": 0.13236045837402344, "learning_rate": 7.324437954469132e-05, "loss": 0.0109, "step": 37880 }, { "grad_norm": 0.1053890809416771, "learning_rate": 7.32297389530017e-05, "loss": 0.0078, "step": 37890 }, { "grad_norm": 0.12697763741016388, "learning_rate": 7.321509582094274e-05, "loss": 0.0095, "step": 37900 }, { "grad_norm": 0.18713173270225525, "learning_rate": 7.320045015011582e-05, "loss": 0.0088, "step": 37910 }, { "grad_norm": 0.11987674236297607, "learning_rate": 7.318580194212253e-05, "loss": 0.0116, "step": 37920 }, { "grad_norm": 0.14224949479103088, "learning_rate": 7.317115119856483e-05, "loss": 0.0111, "step": 37930 }, { "grad_norm": 0.13477014005184174, "learning_rate": 7.315649792104485e-05, "loss": 0.0099, "step": 37940 }, { "grad_norm": 0.11388635635375977, "learning_rate": 7.314184211116507e-05, "loss": 0.0102, "step": 37950 }, { "grad_norm": 0.13408714532852173, "learning_rate": 7.312718377052823e-05, "loss": 0.0107, "step": 37960 }, { "grad_norm": 0.12793824076652527, "learning_rate": 7.311252290073737e-05, "loss": 0.0113, "step": 37970 }, { "grad_norm": 0.12109455466270447, "learning_rate": 7.309785950339572e-05, "loss": 0.0097, "step": 37980 }, { "grad_norm": 0.14761123061180115, "learning_rate": 7.308319358010692e-05, "loss": 0.0098, "step": 37990 }, { "grad_norm": 0.20394489169120789, "learning_rate": 7.306852513247474e-05, "loss": 0.0093, "step": 38000 }, { "grad_norm": 0.16344839334487915, "learning_rate": 7.305385416210336e-05, "loss": 0.0091, "step": 38010 }, { "grad_norm": 0.14088553190231323, "learning_rate": 7.303918067059711e-05, "loss": 0.0096, "step": 38020 }, { "grad_norm": 0.13236427307128906, "learning_rate": 7.302450465956075e-05, "loss": 0.0091, "step": 38030 }, { "grad_norm": 0.12932781875133514, "learning_rate": 7.300982613059914e-05, "loss": 0.0096, "step": 38040 }, { "grad_norm": 0.14014293253421783, "learning_rate": 7.299514508531757e-05, "loss": 0.0101, "step": 38050 }, { "grad_norm": 0.15923157334327698, "learning_rate": 7.298046152532148e-05, "loss": 0.0101, "step": 38060 }, { "grad_norm": 0.12510305643081665, "learning_rate": 7.296577545221668e-05, "loss": 0.0102, "step": 38070 }, { "grad_norm": 0.1445050686597824, "learning_rate": 7.29510868676092e-05, "loss": 0.0117, "step": 38080 }, { "grad_norm": 0.115688256919384, "learning_rate": 7.293639577310538e-05, "loss": 0.0096, "step": 38090 }, { "grad_norm": 0.12976358830928802, "learning_rate": 7.292170217031179e-05, "loss": 0.009, "step": 38100 }, { "grad_norm": 0.1533052921295166, "learning_rate": 7.290700606083532e-05, "loss": 0.0086, "step": 38110 }, { "grad_norm": 0.09912477433681488, "learning_rate": 7.28923074462831e-05, "loss": 0.0094, "step": 38120 }, { "grad_norm": 0.11743106693029404, "learning_rate": 7.287760632826255e-05, "loss": 0.0107, "step": 38130 }, { "grad_norm": 0.143304705619812, "learning_rate": 7.286290270838139e-05, "loss": 0.0084, "step": 38140 }, { "grad_norm": 0.1218482255935669, "learning_rate": 7.284819658824756e-05, "loss": 0.0105, "step": 38150 }, { "grad_norm": 0.13819967210292816, "learning_rate": 7.283348796946929e-05, "loss": 0.0101, "step": 38160 }, { "grad_norm": 0.13875247538089752, "learning_rate": 7.28187768536551e-05, "loss": 0.0101, "step": 38170 }, { "grad_norm": 0.14028960466384888, "learning_rate": 7.280406324241379e-05, "loss": 0.0113, "step": 38180 }, { "grad_norm": 0.15393000841140747, "learning_rate": 7.278934713735441e-05, "loss": 0.011, "step": 38190 }, { "grad_norm": 0.13016943633556366, "learning_rate": 7.277462854008629e-05, "loss": 0.0089, "step": 38200 }, { "grad_norm": 0.1826547235250473, "learning_rate": 7.275990745221904e-05, "loss": 0.0111, "step": 38210 }, { "grad_norm": 0.1699093133211136, "learning_rate": 7.274518387536252e-05, "loss": 0.0099, "step": 38220 }, { "grad_norm": 0.11782753467559814, "learning_rate": 7.273045781112689e-05, "loss": 0.0093, "step": 38230 }, { "grad_norm": 0.14564476907253265, "learning_rate": 7.271572926112257e-05, "loss": 0.0085, "step": 38240 }, { "grad_norm": 0.14656378328800201, "learning_rate": 7.270099822696024e-05, "loss": 0.0086, "step": 38250 }, { "grad_norm": 0.13060392439365387, "learning_rate": 7.26862647102509e-05, "loss": 0.0084, "step": 38260 }, { "grad_norm": 0.16133370995521545, "learning_rate": 7.267152871260573e-05, "loss": 0.0101, "step": 38270 }, { "grad_norm": 0.11835252493619919, "learning_rate": 7.265679023563629e-05, "loss": 0.0098, "step": 38280 }, { "grad_norm": 0.14637306332588196, "learning_rate": 7.264204928095433e-05, "loss": 0.0097, "step": 38290 }, { "grad_norm": 0.1349828541278839, "learning_rate": 7.262730585017188e-05, "loss": 0.0113, "step": 38300 }, { "grad_norm": 0.22094914317131042, "learning_rate": 7.261255994490131e-05, "loss": 0.0107, "step": 38310 }, { "grad_norm": 0.12940840423107147, "learning_rate": 7.259781156675517e-05, "loss": 0.0104, "step": 38320 }, { "grad_norm": 0.09485846012830734, "learning_rate": 7.258306071734633e-05, "loss": 0.0094, "step": 38330 }, { "grad_norm": 0.13263022899627686, "learning_rate": 7.256830739828792e-05, "loss": 0.0109, "step": 38340 }, { "grad_norm": 0.1215464398264885, "learning_rate": 7.255355161119336e-05, "loss": 0.0089, "step": 38350 }, { "grad_norm": 0.1268603801727295, "learning_rate": 7.253879335767628e-05, "loss": 0.0082, "step": 38360 }, { "grad_norm": 0.09489360451698303, "learning_rate": 7.252403263935067e-05, "loss": 0.0084, "step": 38370 }, { "grad_norm": 0.09729727357625961, "learning_rate": 7.25092694578307e-05, "loss": 0.0072, "step": 38380 }, { "grad_norm": 0.14992406964302063, "learning_rate": 7.249450381473087e-05, "loss": 0.0101, "step": 38390 }, { "grad_norm": 0.17152200639247894, "learning_rate": 7.247973571166593e-05, "loss": 0.0093, "step": 38400 }, { "grad_norm": 0.1253615766763687, "learning_rate": 7.246496515025089e-05, "loss": 0.0089, "step": 38410 }, { "grad_norm": 0.15439772605895996, "learning_rate": 7.245019213210101e-05, "loss": 0.0095, "step": 38420 }, { "grad_norm": 0.13488458096981049, "learning_rate": 7.24354166588319e-05, "loss": 0.0107, "step": 38430 }, { "grad_norm": 0.14441129565238953, "learning_rate": 7.242063873205935e-05, "loss": 0.0088, "step": 38440 }, { "grad_norm": 0.17189304530620575, "learning_rate": 7.240585835339946e-05, "loss": 0.0117, "step": 38450 }, { "grad_norm": 0.16226299107074738, "learning_rate": 7.239107552446857e-05, "loss": 0.0095, "step": 38460 }, { "grad_norm": 0.12986645102500916, "learning_rate": 7.237629024688334e-05, "loss": 0.0095, "step": 38470 }, { "grad_norm": 0.12094607949256897, "learning_rate": 7.236150252226064e-05, "loss": 0.0101, "step": 38480 }, { "grad_norm": 0.17346929013729095, "learning_rate": 7.234671235221765e-05, "loss": 0.01, "step": 38490 }, { "grad_norm": 0.11348312348127365, "learning_rate": 7.233191973837179e-05, "loss": 0.0088, "step": 38500 }, { "grad_norm": 0.1373233050107956, "learning_rate": 7.231712468234078e-05, "loss": 0.0097, "step": 38510 }, { "grad_norm": 0.1258324235677719, "learning_rate": 7.230232718574254e-05, "loss": 0.0087, "step": 38520 }, { "grad_norm": 0.12163852155208588, "learning_rate": 7.228752725019535e-05, "loss": 0.0096, "step": 38530 }, { "grad_norm": 0.14187105000019073, "learning_rate": 7.227272487731769e-05, "loss": 0.0092, "step": 38540 }, { "grad_norm": 0.1347804218530655, "learning_rate": 7.225792006872831e-05, "loss": 0.0095, "step": 38550 }, { "grad_norm": 0.16450197994709015, "learning_rate": 7.224311282604628e-05, "loss": 0.011, "step": 38560 }, { "grad_norm": 0.11120938509702682, "learning_rate": 7.222830315089085e-05, "loss": 0.0113, "step": 38570 }, { "grad_norm": 0.14461404085159302, "learning_rate": 7.22134910448816e-05, "loss": 0.0086, "step": 38580 }, { "grad_norm": 0.14831486344337463, "learning_rate": 7.219867650963839e-05, "loss": 0.0117, "step": 38590 }, { "grad_norm": 0.12678062915802002, "learning_rate": 7.21838595467813e-05, "loss": 0.0107, "step": 38600 }, { "grad_norm": 0.11983811110258102, "learning_rate": 7.216904015793067e-05, "loss": 0.0104, "step": 38610 }, { "grad_norm": 0.1138136237859726, "learning_rate": 7.215421834470713e-05, "loss": 0.0085, "step": 38620 }, { "grad_norm": 0.13661040365695953, "learning_rate": 7.213939410873158e-05, "loss": 0.0082, "step": 38630 }, { "grad_norm": 0.19724708795547485, "learning_rate": 7.212456745162519e-05, "loss": 0.0097, "step": 38640 }, { "grad_norm": 0.15158748626708984, "learning_rate": 7.210973837500937e-05, "loss": 0.0104, "step": 38650 }, { "grad_norm": 0.11422715336084366, "learning_rate": 7.209490688050578e-05, "loss": 0.0089, "step": 38660 }, { "grad_norm": 0.1329239010810852, "learning_rate": 7.208007296973641e-05, "loss": 0.0085, "step": 38670 }, { "grad_norm": 0.11401889473199844, "learning_rate": 7.206523664432345e-05, "loss": 0.0115, "step": 38680 }, { "grad_norm": 0.1330534815788269, "learning_rate": 7.205039790588939e-05, "loss": 0.0081, "step": 38690 }, { "grad_norm": 0.1360759735107422, "learning_rate": 7.203555675605697e-05, "loss": 0.0104, "step": 38700 }, { "grad_norm": 0.10742847621440887, "learning_rate": 7.202071319644917e-05, "loss": 0.0082, "step": 38710 }, { "grad_norm": 0.12036053836345673, "learning_rate": 7.200586722868932e-05, "loss": 0.0085, "step": 38720 }, { "grad_norm": 0.1598995178937912, "learning_rate": 7.199101885440088e-05, "loss": 0.0106, "step": 38730 }, { "grad_norm": 0.11962694674730301, "learning_rate": 7.197616807520771e-05, "loss": 0.0102, "step": 38740 }, { "grad_norm": 0.1556045114994049, "learning_rate": 7.196131489273381e-05, "loss": 0.0098, "step": 38750 }, { "grad_norm": 0.13294032216072083, "learning_rate": 7.194645930860355e-05, "loss": 0.0102, "step": 38760 }, { "grad_norm": 0.10999123007059097, "learning_rate": 7.19316013244415e-05, "loss": 0.0094, "step": 38770 }, { "grad_norm": 0.11863435059785843, "learning_rate": 7.191674094187248e-05, "loss": 0.0092, "step": 38780 }, { "grad_norm": 0.10928715765476227, "learning_rate": 7.190187816252165e-05, "loss": 0.0089, "step": 38790 }, { "grad_norm": 0.11019770056009293, "learning_rate": 7.188701298801435e-05, "loss": 0.0088, "step": 38800 }, { "grad_norm": 0.13417692482471466, "learning_rate": 7.18721454199762e-05, "loss": 0.0095, "step": 38810 }, { "grad_norm": 0.12378418445587158, "learning_rate": 7.185727546003312e-05, "loss": 0.0124, "step": 38820 }, { "grad_norm": 0.1074063628911972, "learning_rate": 7.184240310981126e-05, "loss": 0.0122, "step": 38830 }, { "grad_norm": 0.13049401342868805, "learning_rate": 7.182752837093704e-05, "loss": 0.0093, "step": 38840 }, { "grad_norm": 0.09763085842132568, "learning_rate": 7.181265124503711e-05, "loss": 0.0108, "step": 38850 }, { "grad_norm": 0.10592558234930038, "learning_rate": 7.179777173373847e-05, "loss": 0.0076, "step": 38860 }, { "grad_norm": 0.14368267357349396, "learning_rate": 7.178288983866826e-05, "loss": 0.0117, "step": 38870 }, { "grad_norm": 0.1337391883134842, "learning_rate": 7.176800556145397e-05, "loss": 0.0104, "step": 38880 }, { "grad_norm": 0.12866462767124176, "learning_rate": 7.175311890372334e-05, "loss": 0.0098, "step": 38890 }, { "grad_norm": 0.14606256783008575, "learning_rate": 7.17382298671043e-05, "loss": 0.0105, "step": 38900 }, { "grad_norm": 0.13129745423793793, "learning_rate": 7.172333845322515e-05, "loss": 0.0097, "step": 38910 }, { "grad_norm": 0.131935253739357, "learning_rate": 7.170844466371436e-05, "loss": 0.0095, "step": 38920 }, { "grad_norm": 0.13399279117584229, "learning_rate": 7.16935485002007e-05, "loss": 0.0113, "step": 38930 }, { "grad_norm": 0.13360731303691864, "learning_rate": 7.167864996431319e-05, "loss": 0.0108, "step": 38940 }, { "grad_norm": 0.12041705846786499, "learning_rate": 7.166374905768111e-05, "loss": 0.0102, "step": 38950 }, { "grad_norm": 0.15098974108695984, "learning_rate": 7.164884578193401e-05, "loss": 0.0125, "step": 38960 }, { "grad_norm": 0.1496061384677887, "learning_rate": 7.163394013870168e-05, "loss": 0.0112, "step": 38970 }, { "grad_norm": 0.1475055068731308, "learning_rate": 7.161903212961419e-05, "loss": 0.0104, "step": 38980 }, { "grad_norm": 0.13127604126930237, "learning_rate": 7.160412175630183e-05, "loss": 0.0138, "step": 38990 }, { "grad_norm": 0.12329775094985962, "learning_rate": 7.158920902039521e-05, "loss": 0.0114, "step": 39000 }, { "grad_norm": 0.1805853396654129, "learning_rate": 7.157429392352514e-05, "loss": 0.0098, "step": 39010 }, { "grad_norm": 0.12549430131912231, "learning_rate": 7.155937646732275e-05, "loss": 0.0114, "step": 39020 }, { "grad_norm": 0.14716628193855286, "learning_rate": 7.154445665341933e-05, "loss": 0.0117, "step": 39030 }, { "grad_norm": 0.18871568143367767, "learning_rate": 7.152953448344654e-05, "loss": 0.0121, "step": 39040 }, { "grad_norm": 0.17644819617271423, "learning_rate": 7.151460995903624e-05, "loss": 0.0103, "step": 39050 }, { "grad_norm": 0.11153683811426163, "learning_rate": 7.149968308182052e-05, "loss": 0.0093, "step": 39060 }, { "grad_norm": 0.20965780317783356, "learning_rate": 7.14847538534318e-05, "loss": 0.0107, "step": 39070 }, { "grad_norm": 0.10983291268348694, "learning_rate": 7.14698222755027e-05, "loss": 0.0093, "step": 39080 }, { "grad_norm": 0.11320537328720093, "learning_rate": 7.14548883496661e-05, "loss": 0.0101, "step": 39090 }, { "grad_norm": 0.11177431792020798, "learning_rate": 7.143995207755517e-05, "loss": 0.0096, "step": 39100 }, { "grad_norm": 0.14490750432014465, "learning_rate": 7.142501346080333e-05, "loss": 0.0101, "step": 39110 }, { "grad_norm": 0.15384350717067719, "learning_rate": 7.141007250104421e-05, "loss": 0.0103, "step": 39120 }, { "grad_norm": 0.13358190655708313, "learning_rate": 7.139512919991176e-05, "loss": 0.0111, "step": 39130 }, { "grad_norm": 0.1536152958869934, "learning_rate": 7.138018355904014e-05, "loss": 0.0112, "step": 39140 }, { "grad_norm": 0.21668493747711182, "learning_rate": 7.13652355800638e-05, "loss": 0.0123, "step": 39150 }, { "grad_norm": 0.1530970335006714, "learning_rate": 7.135028526461743e-05, "loss": 0.01, "step": 39160 }, { "grad_norm": 0.15176817774772644, "learning_rate": 7.133533261433594e-05, "loss": 0.0104, "step": 39170 }, { "grad_norm": 0.3045442998409271, "learning_rate": 7.132037763085457e-05, "loss": 0.0126, "step": 39180 }, { "grad_norm": 0.11527860909700394, "learning_rate": 7.130542031580875e-05, "loss": 0.0101, "step": 39190 }, { "grad_norm": 0.11472759395837784, "learning_rate": 7.12904606708342e-05, "loss": 0.0095, "step": 39200 }, { "grad_norm": 0.14225398004055023, "learning_rate": 7.127549869756687e-05, "loss": 0.0116, "step": 39210 }, { "grad_norm": 0.11979532986879349, "learning_rate": 7.126053439764299e-05, "loss": 0.0108, "step": 39220 }, { "grad_norm": 0.12192260473966599, "learning_rate": 7.124556777269904e-05, "loss": 0.0111, "step": 39230 }, { "grad_norm": 0.11492267996072769, "learning_rate": 7.123059882437174e-05, "loss": 0.0096, "step": 39240 }, { "grad_norm": 0.18655802309513092, "learning_rate": 7.121562755429807e-05, "loss": 0.0103, "step": 39250 }, { "grad_norm": 0.12545378506183624, "learning_rate": 7.120065396411528e-05, "loss": 0.0111, "step": 39260 }, { "grad_norm": 0.13157624006271362, "learning_rate": 7.118567805546084e-05, "loss": 0.0099, "step": 39270 }, { "grad_norm": 0.15272170305252075, "learning_rate": 7.117069982997248e-05, "loss": 0.0094, "step": 39280 }, { "grad_norm": 0.1215849295258522, "learning_rate": 7.115571928928825e-05, "loss": 0.0095, "step": 39290 }, { "grad_norm": 0.120669886469841, "learning_rate": 7.114073643504635e-05, "loss": 0.0086, "step": 39300 }, { "grad_norm": 0.1584613174200058, "learning_rate": 7.11257512688853e-05, "loss": 0.0141, "step": 39310 }, { "grad_norm": 0.12052890658378601, "learning_rate": 7.111076379244384e-05, "loss": 0.0105, "step": 39320 }, { "grad_norm": 0.14568080008029938, "learning_rate": 7.109577400736101e-05, "loss": 0.0099, "step": 39330 }, { "grad_norm": 0.11339178681373596, "learning_rate": 7.108078191527605e-05, "loss": 0.0112, "step": 39340 }, { "grad_norm": 0.11172246187925339, "learning_rate": 7.106578751782847e-05, "loss": 0.0101, "step": 39350 }, { "grad_norm": 0.16746492683887482, "learning_rate": 7.105079081665803e-05, "loss": 0.0105, "step": 39360 }, { "grad_norm": 0.12485601007938385, "learning_rate": 7.103579181340476e-05, "loss": 0.0113, "step": 39370 }, { "grad_norm": 0.1493850201368332, "learning_rate": 7.102079050970893e-05, "loss": 0.0097, "step": 39380 }, { "grad_norm": 0.14962950348854065, "learning_rate": 7.100578690721105e-05, "loss": 0.01, "step": 39390 }, { "grad_norm": 0.14666610956192017, "learning_rate": 7.09907810075519e-05, "loss": 0.0086, "step": 39400 }, { "grad_norm": 0.16214682161808014, "learning_rate": 7.097577281237249e-05, "loss": 0.01, "step": 39410 }, { "grad_norm": 0.20277626812458038, "learning_rate": 7.09607623233141e-05, "loss": 0.0107, "step": 39420 }, { "grad_norm": 0.09513887017965317, "learning_rate": 7.094574954201827e-05, "loss": 0.0095, "step": 39430 }, { "grad_norm": 0.14943400025367737, "learning_rate": 7.093073447012675e-05, "loss": 0.0098, "step": 39440 }, { "grad_norm": 0.15382255613803864, "learning_rate": 7.09157171092816e-05, "loss": 0.0112, "step": 39450 }, { "grad_norm": 0.1216951236128807, "learning_rate": 7.090069746112504e-05, "loss": 0.0116, "step": 39460 }, { "grad_norm": 0.1581277698278427, "learning_rate": 7.088567552729965e-05, "loss": 0.0103, "step": 39470 }, { "grad_norm": 0.14053533971309662, "learning_rate": 7.087065130944818e-05, "loss": 0.0092, "step": 39480 }, { "grad_norm": 0.12304804474115372, "learning_rate": 7.085562480921366e-05, "loss": 0.01, "step": 39490 }, { "grad_norm": 0.11482831090688705, "learning_rate": 7.084059602823937e-05, "loss": 0.0089, "step": 39500 }, { "grad_norm": 0.14632682502269745, "learning_rate": 7.082556496816882e-05, "loss": 0.0096, "step": 39510 }, { "grad_norm": 0.116937056183815, "learning_rate": 7.081053163064582e-05, "loss": 0.0079, "step": 39520 }, { "grad_norm": 0.13513793051242828, "learning_rate": 7.079549601731434e-05, "loss": 0.0095, "step": 39530 }, { "grad_norm": 0.13977216184139252, "learning_rate": 7.07804581298187e-05, "loss": 0.0107, "step": 39540 }, { "grad_norm": 0.14279748499393463, "learning_rate": 7.07654179698034e-05, "loss": 0.0097, "step": 39550 }, { "grad_norm": 0.12917137145996094, "learning_rate": 7.075037553891321e-05, "loss": 0.01, "step": 39560 }, { "grad_norm": 0.14080508053302765, "learning_rate": 7.073533083879315e-05, "loss": 0.0118, "step": 39570 }, { "grad_norm": 0.15693534910678864, "learning_rate": 7.072028387108849e-05, "loss": 0.0099, "step": 39580 }, { "grad_norm": 0.12860311567783356, "learning_rate": 7.070523463744472e-05, "loss": 0.0086, "step": 39590 }, { "grad_norm": 0.15200625360012054, "learning_rate": 7.069018313950763e-05, "loss": 0.0106, "step": 39600 }, { "grad_norm": 0.11253681778907776, "learning_rate": 7.067512937892322e-05, "loss": 0.0105, "step": 39610 }, { "grad_norm": 0.1314319521188736, "learning_rate": 7.066007335733775e-05, "loss": 0.0115, "step": 39620 }, { "grad_norm": 0.10697077214717865, "learning_rate": 7.064501507639772e-05, "loss": 0.0086, "step": 39630 }, { "grad_norm": 0.1486346274614334, "learning_rate": 7.062995453774987e-05, "loss": 0.0089, "step": 39640 }, { "grad_norm": 0.10225558280944824, "learning_rate": 7.061489174304121e-05, "loss": 0.0104, "step": 39650 }, { "grad_norm": 0.11129708588123322, "learning_rate": 7.0599826693919e-05, "loss": 0.01, "step": 39660 }, { "grad_norm": 0.08640456199645996, "learning_rate": 7.05847593920307e-05, "loss": 0.0092, "step": 39670 }, { "grad_norm": 0.12397174537181854, "learning_rate": 7.056968983902406e-05, "loss": 0.0097, "step": 39680 }, { "grad_norm": 0.10513842105865479, "learning_rate": 7.055461803654706e-05, "loss": 0.0125, "step": 39690 }, { "grad_norm": 0.1177745908498764, "learning_rate": 7.053954398624794e-05, "loss": 0.0105, "step": 39700 }, { "grad_norm": 0.14815083146095276, "learning_rate": 7.052446768977518e-05, "loss": 0.0111, "step": 39710 }, { "grad_norm": 0.11124234646558762, "learning_rate": 7.050938914877748e-05, "loss": 0.0101, "step": 39720 }, { "grad_norm": 0.13896970450878143, "learning_rate": 7.049430836490382e-05, "loss": 0.0101, "step": 39730 }, { "grad_norm": 0.11658047884702682, "learning_rate": 7.047922533980342e-05, "loss": 0.01, "step": 39740 }, { "grad_norm": 0.1708943247795105, "learning_rate": 7.046414007512571e-05, "loss": 0.0088, "step": 39750 }, { "grad_norm": 0.17432977259159088, "learning_rate": 7.044905257252042e-05, "loss": 0.0099, "step": 39760 }, { "grad_norm": 0.10220562666654587, "learning_rate": 7.04339628336375e-05, "loss": 0.011, "step": 39770 }, { "grad_norm": 0.10850781947374344, "learning_rate": 7.041887086012711e-05, "loss": 0.0097, "step": 39780 }, { "grad_norm": 0.12659364938735962, "learning_rate": 7.040377665363969e-05, "loss": 0.0125, "step": 39790 }, { "grad_norm": 0.18710817396640778, "learning_rate": 7.038868021582594e-05, "loss": 0.0133, "step": 39800 }, { "grad_norm": 0.16080397367477417, "learning_rate": 7.037358154833679e-05, "loss": 0.0096, "step": 39810 }, { "grad_norm": 0.1440207064151764, "learning_rate": 7.035848065282339e-05, "loss": 0.0142, "step": 39820 }, { "grad_norm": 0.14945609867572784, "learning_rate": 7.034337753093714e-05, "loss": 0.0101, "step": 39830 }, { "grad_norm": 0.16594265401363373, "learning_rate": 7.032827218432972e-05, "loss": 0.01, "step": 39840 }, { "grad_norm": 0.10445286333560944, "learning_rate": 7.031316461465302e-05, "loss": 0.0104, "step": 39850 }, { "grad_norm": 0.12529881298542023, "learning_rate": 7.029805482355915e-05, "loss": 0.0098, "step": 39860 }, { "grad_norm": 0.10381494462490082, "learning_rate": 7.028294281270055e-05, "loss": 0.0096, "step": 39870 }, { "grad_norm": 0.16153578460216522, "learning_rate": 7.026782858372982e-05, "loss": 0.0102, "step": 39880 }, { "grad_norm": 0.11661561578512192, "learning_rate": 7.025271213829982e-05, "loss": 0.0093, "step": 39890 }, { "grad_norm": 0.19374234974384308, "learning_rate": 7.023759347806366e-05, "loss": 0.0099, "step": 39900 }, { "grad_norm": 0.16844139993190765, "learning_rate": 7.022247260467472e-05, "loss": 0.011, "step": 39910 }, { "grad_norm": 0.2068820744752884, "learning_rate": 7.020734951978658e-05, "loss": 0.0087, "step": 39920 }, { "grad_norm": 0.16961811482906342, "learning_rate": 7.019222422505307e-05, "loss": 0.0094, "step": 39930 }, { "grad_norm": 0.1158720925450325, "learning_rate": 7.017709672212827e-05, "loss": 0.008, "step": 39940 }, { "grad_norm": 0.110651396214962, "learning_rate": 7.016196701266652e-05, "loss": 0.0086, "step": 39950 }, { "grad_norm": 0.14522449672222137, "learning_rate": 7.014683509832238e-05, "loss": 0.0084, "step": 39960 }, { "grad_norm": 0.11826502531766891, "learning_rate": 7.013170098075063e-05, "loss": 0.0089, "step": 39970 }, { "grad_norm": 0.18113753199577332, "learning_rate": 7.011656466160632e-05, "loss": 0.011, "step": 39980 }, { "grad_norm": 0.1195218563079834, "learning_rate": 7.010142614254475e-05, "loss": 0.0096, "step": 39990 }, { "grad_norm": 0.16102835536003113, "learning_rate": 7.008628542522147e-05, "loss": 0.0115, "step": 40000 }, { "grad_norm": 0.15799543261528015, "learning_rate": 7.007114251129218e-05, "loss": 0.0112, "step": 40010 }, { "grad_norm": 0.12198711931705475, "learning_rate": 7.005599740241293e-05, "loss": 0.0096, "step": 40020 }, { "grad_norm": 0.12917131185531616, "learning_rate": 7.004085010023996e-05, "loss": 0.0113, "step": 40030 }, { "grad_norm": 0.14670690894126892, "learning_rate": 7.002570060642976e-05, "loss": 0.0097, "step": 40040 }, { "grad_norm": 0.11504791676998138, "learning_rate": 7.001054892263903e-05, "loss": 0.0101, "step": 40050 }, { "grad_norm": 0.16912667453289032, "learning_rate": 6.999539505052477e-05, "loss": 0.01, "step": 40060 }, { "grad_norm": 0.10589632391929626, "learning_rate": 6.998023899174418e-05, "loss": 0.0098, "step": 40070 }, { "grad_norm": 0.16209714114665985, "learning_rate": 6.996508074795467e-05, "loss": 0.0138, "step": 40080 }, { "grad_norm": 0.15230631828308105, "learning_rate": 6.994992032081396e-05, "loss": 0.0096, "step": 40090 }, { "grad_norm": 0.12454051524400711, "learning_rate": 6.993475771197995e-05, "loss": 0.0109, "step": 40100 }, { "grad_norm": 0.12607070803642273, "learning_rate": 6.991959292311082e-05, "loss": 0.0102, "step": 40110 }, { "grad_norm": 0.12894229590892792, "learning_rate": 6.990442595586495e-05, "loss": 0.0091, "step": 40120 }, { "grad_norm": 0.15016092360019684, "learning_rate": 6.988925681190098e-05, "loss": 0.0098, "step": 40130 }, { "grad_norm": 0.1251852959394455, "learning_rate": 6.987408549287778e-05, "loss": 0.0095, "step": 40140 }, { "grad_norm": 0.26848655939102173, "learning_rate": 6.985891200045449e-05, "loss": 0.0115, "step": 40150 }, { "grad_norm": 0.12539048492908478, "learning_rate": 6.984373633629045e-05, "loss": 0.0113, "step": 40160 }, { "grad_norm": 0.15762107074260712, "learning_rate": 6.98285585020452e-05, "loss": 0.0112, "step": 40170 }, { "grad_norm": 0.09438429772853851, "learning_rate": 6.981337849937864e-05, "loss": 0.008, "step": 40180 }, { "grad_norm": 0.19624587893486023, "learning_rate": 6.979819632995078e-05, "loss": 0.0113, "step": 40190 }, { "grad_norm": 0.12214267253875732, "learning_rate": 6.978301199542193e-05, "loss": 0.0081, "step": 40200 }, { "grad_norm": 0.11746125668287277, "learning_rate": 6.976782549745263e-05, "loss": 0.0094, "step": 40210 }, { "grad_norm": 0.14566053450107574, "learning_rate": 6.975263683770367e-05, "loss": 0.0095, "step": 40220 }, { "grad_norm": 0.11329670995473862, "learning_rate": 6.973744601783604e-05, "loss": 0.0091, "step": 40230 }, { "grad_norm": 0.13334466516971588, "learning_rate": 6.972225303951097e-05, "loss": 0.0131, "step": 40240 }, { "grad_norm": 0.15011358261108398, "learning_rate": 6.970705790438998e-05, "loss": 0.0129, "step": 40250 }, { "grad_norm": 0.1636441946029663, "learning_rate": 6.969186061413477e-05, "loss": 0.0101, "step": 40260 }, { "grad_norm": 0.14093555510044098, "learning_rate": 6.967666117040727e-05, "loss": 0.009, "step": 40270 }, { "grad_norm": 0.17840932309627533, "learning_rate": 6.966145957486972e-05, "loss": 0.0098, "step": 40280 }, { "grad_norm": 0.15017662942409515, "learning_rate": 6.964625582918449e-05, "loss": 0.0114, "step": 40290 }, { "grad_norm": 0.11810848116874695, "learning_rate": 6.963104993501425e-05, "loss": 0.0126, "step": 40300 }, { "grad_norm": 0.13506504893302917, "learning_rate": 6.961584189402192e-05, "loss": 0.0121, "step": 40310 }, { "grad_norm": 0.13275142014026642, "learning_rate": 6.96006317078706e-05, "loss": 0.0088, "step": 40320 }, { "grad_norm": 0.12384451925754547, "learning_rate": 6.958541937822367e-05, "loss": 0.0106, "step": 40330 }, { "grad_norm": 0.12306469678878784, "learning_rate": 6.957020490674473e-05, "loss": 0.0094, "step": 40340 }, { "grad_norm": 0.13083341717720032, "learning_rate": 6.95549882950976e-05, "loss": 0.0098, "step": 40350 }, { "grad_norm": 0.14690549671649933, "learning_rate": 6.953976954494635e-05, "loss": 0.0109, "step": 40360 }, { "grad_norm": 0.1539240926504135, "learning_rate": 6.952454865795528e-05, "loss": 0.0109, "step": 40370 }, { "grad_norm": 0.1386067122220993, "learning_rate": 6.950932563578892e-05, "loss": 0.0107, "step": 40380 }, { "grad_norm": 0.152873694896698, "learning_rate": 6.949410048011206e-05, "loss": 0.01, "step": 40390 }, { "grad_norm": 0.12137063592672348, "learning_rate": 6.947887319258966e-05, "loss": 0.009, "step": 40400 }, { "grad_norm": 0.12783190608024597, "learning_rate": 6.946364377488696e-05, "loss": 0.0099, "step": 40410 }, { "grad_norm": 0.14877048134803772, "learning_rate": 6.944841222866947e-05, "loss": 0.0114, "step": 40420 }, { "grad_norm": 0.18495935201644897, "learning_rate": 6.943317855560284e-05, "loss": 0.0104, "step": 40430 }, { "grad_norm": 0.14622749388217926, "learning_rate": 6.941794275735302e-05, "loss": 0.0104, "step": 40440 }, { "grad_norm": 0.17579875886440277, "learning_rate": 6.94027048355862e-05, "loss": 0.0084, "step": 40450 }, { "grad_norm": 0.15628668665885925, "learning_rate": 6.938746479196871e-05, "loss": 0.0088, "step": 40460 }, { "grad_norm": 0.17809364199638367, "learning_rate": 6.937222262816724e-05, "loss": 0.0096, "step": 40470 }, { "grad_norm": 0.12516066431999207, "learning_rate": 6.935697834584865e-05, "loss": 0.0089, "step": 40480 }, { "grad_norm": 0.11602812260389328, "learning_rate": 6.934173194667998e-05, "loss": 0.0104, "step": 40490 }, { "grad_norm": 0.11162146925926208, "learning_rate": 6.93264834323286e-05, "loss": 0.0084, "step": 40500 }, { "grad_norm": 0.133936807513237, "learning_rate": 6.931123280446206e-05, "loss": 0.0085, "step": 40510 }, { "grad_norm": 0.11858701705932617, "learning_rate": 6.929598006474811e-05, "loss": 0.008, "step": 40520 }, { "grad_norm": 0.16370686888694763, "learning_rate": 6.92807252148548e-05, "loss": 0.0083, "step": 40530 }, { "grad_norm": 0.14914710819721222, "learning_rate": 6.926546825645037e-05, "loss": 0.0117, "step": 40540 }, { "grad_norm": 0.10431324690580368, "learning_rate": 6.92502091912033e-05, "loss": 0.0095, "step": 40550 }, { "grad_norm": 0.1379423588514328, "learning_rate": 6.923494802078229e-05, "loss": 0.0097, "step": 40560 }, { "grad_norm": 0.1590953767299652, "learning_rate": 6.921968474685628e-05, "loss": 0.0096, "step": 40570 }, { "grad_norm": 0.10451463609933853, "learning_rate": 6.920441937109445e-05, "loss": 0.009, "step": 40580 }, { "grad_norm": 0.13225442171096802, "learning_rate": 6.91891518951662e-05, "loss": 0.0072, "step": 40590 }, { "grad_norm": 0.10203120112419128, "learning_rate": 6.917388232074114e-05, "loss": 0.0087, "step": 40600 }, { "grad_norm": 0.12551720440387726, "learning_rate": 6.915861064948914e-05, "loss": 0.0099, "step": 40610 }, { "grad_norm": 0.1785145401954651, "learning_rate": 6.914333688308025e-05, "loss": 0.0101, "step": 40620 }, { "grad_norm": 0.1285540908575058, "learning_rate": 6.912806102318487e-05, "loss": 0.0082, "step": 40630 }, { "grad_norm": 0.15342940390110016, "learning_rate": 6.911278307147347e-05, "loss": 0.0097, "step": 40640 }, { "grad_norm": 0.10102871805429459, "learning_rate": 6.909750302961684e-05, "loss": 0.0083, "step": 40650 }, { "grad_norm": 0.28779006004333496, "learning_rate": 6.9082220899286e-05, "loss": 0.013, "step": 40660 }, { "grad_norm": 0.18111300468444824, "learning_rate": 6.906693668215219e-05, "loss": 0.0089, "step": 40670 }, { "grad_norm": 0.13282273709774017, "learning_rate": 6.905165037988683e-05, "loss": 0.0112, "step": 40680 }, { "grad_norm": 0.10275973379611969, "learning_rate": 6.903636199416164e-05, "loss": 0.0097, "step": 40690 }, { "grad_norm": 0.13977791368961334, "learning_rate": 6.902107152664851e-05, "loss": 0.0094, "step": 40700 }, { "grad_norm": 0.10865958780050278, "learning_rate": 6.90057789790196e-05, "loss": 0.0105, "step": 40710 }, { "grad_norm": 0.12557446956634521, "learning_rate": 6.899048435294728e-05, "loss": 0.0104, "step": 40720 }, { "grad_norm": 0.12920477986335754, "learning_rate": 6.897518765010415e-05, "loss": 0.0115, "step": 40730 }, { "grad_norm": 0.17169933021068573, "learning_rate": 6.895988887216302e-05, "loss": 0.0105, "step": 40740 }, { "grad_norm": 0.12857742607593536, "learning_rate": 6.894458802079694e-05, "loss": 0.0081, "step": 40750 }, { "grad_norm": 0.1337050199508667, "learning_rate": 6.892928509767922e-05, "loss": 0.0111, "step": 40760 }, { "grad_norm": 0.1654847413301468, "learning_rate": 6.891398010448333e-05, "loss": 0.0105, "step": 40770 }, { "grad_norm": 0.1344292312860489, "learning_rate": 6.8898673042883e-05, "loss": 0.0095, "step": 40780 }, { "grad_norm": 0.1466086059808731, "learning_rate": 6.888336391455222e-05, "loss": 0.0081, "step": 40790 }, { "grad_norm": 0.14607839286327362, "learning_rate": 6.886805272116513e-05, "loss": 0.0096, "step": 40800 }, { "grad_norm": 0.13457851111888885, "learning_rate": 6.885273946439617e-05, "loss": 0.0085, "step": 40810 }, { "grad_norm": 0.1518298089504242, "learning_rate": 6.883742414591998e-05, "loss": 0.0107, "step": 40820 }, { "grad_norm": 0.16749873757362366, "learning_rate": 6.88221067674114e-05, "loss": 0.0101, "step": 40830 }, { "grad_norm": 0.16068531572818756, "learning_rate": 6.88067873305455e-05, "loss": 0.0099, "step": 40840 }, { "grad_norm": 0.09126166999340057, "learning_rate": 6.879146583699765e-05, "loss": 0.0089, "step": 40850 }, { "grad_norm": 0.11251199245452881, "learning_rate": 6.877614228844334e-05, "loss": 0.009, "step": 40860 }, { "grad_norm": 0.11737541854381561, "learning_rate": 6.876081668655832e-05, "loss": 0.0105, "step": 40870 }, { "grad_norm": 0.12146054953336716, "learning_rate": 6.874548903301861e-05, "loss": 0.0116, "step": 40880 }, { "grad_norm": 0.14514471590518951, "learning_rate": 6.873015932950039e-05, "loss": 0.0105, "step": 40890 }, { "grad_norm": 0.10953646153211594, "learning_rate": 6.871482757768012e-05, "loss": 0.0099, "step": 40900 }, { "grad_norm": 0.12441994994878769, "learning_rate": 6.869949377923442e-05, "loss": 0.01, "step": 40910 }, { "grad_norm": 0.16945499181747437, "learning_rate": 6.868415793584022e-05, "loss": 0.0097, "step": 40920 }, { "grad_norm": 0.15317723155021667, "learning_rate": 6.86688200491746e-05, "loss": 0.0104, "step": 40930 }, { "grad_norm": 0.13122175633907318, "learning_rate": 6.865348012091488e-05, "loss": 0.0127, "step": 40940 }, { "grad_norm": 0.17727938294410706, "learning_rate": 6.86381381527386e-05, "loss": 0.0126, "step": 40950 }, { "grad_norm": 0.16562342643737793, "learning_rate": 6.862279414632359e-05, "loss": 0.0101, "step": 40960 }, { "grad_norm": 0.16281606256961823, "learning_rate": 6.860744810334778e-05, "loss": 0.011, "step": 40970 }, { "grad_norm": 0.16400299966335297, "learning_rate": 6.859210002548943e-05, "loss": 0.0114, "step": 40980 }, { "grad_norm": 0.1515265554189682, "learning_rate": 6.8576749914427e-05, "loss": 0.0092, "step": 40990 }, { "grad_norm": 0.12368754297494888, "learning_rate": 6.85613977718391e-05, "loss": 0.0101, "step": 41000 }, { "grad_norm": 0.11923303455114365, "learning_rate": 6.854604359940464e-05, "loss": 0.0105, "step": 41010 }, { "grad_norm": 0.13364934921264648, "learning_rate": 6.853068739880276e-05, "loss": 0.0088, "step": 41020 }, { "grad_norm": 0.11573543399572372, "learning_rate": 6.851532917171276e-05, "loss": 0.0093, "step": 41030 }, { "grad_norm": 0.13734504580497742, "learning_rate": 6.849996891981418e-05, "loss": 0.0115, "step": 41040 }, { "grad_norm": 0.11700484156608582, "learning_rate": 6.848460664478684e-05, "loss": 0.0107, "step": 41050 }, { "grad_norm": 0.11958379298448563, "learning_rate": 6.846924234831069e-05, "loss": 0.0091, "step": 41060 }, { "grad_norm": 0.11651346832513809, "learning_rate": 6.845387603206597e-05, "loss": 0.0088, "step": 41070 }, { "grad_norm": 0.12594622373580933, "learning_rate": 6.843850769773312e-05, "loss": 0.0085, "step": 41080 }, { "grad_norm": 0.13140179216861725, "learning_rate": 6.842313734699278e-05, "loss": 0.0097, "step": 41090 }, { "grad_norm": 0.13464853167533875, "learning_rate": 6.840776498152584e-05, "loss": 0.009, "step": 41100 }, { "grad_norm": 0.15388478338718414, "learning_rate": 6.83923906030134e-05, "loss": 0.0125, "step": 41110 }, { "grad_norm": 0.1526181548833847, "learning_rate": 6.837701421313677e-05, "loss": 0.0116, "step": 41120 }, { "grad_norm": 0.1162279024720192, "learning_rate": 6.83616358135775e-05, "loss": 0.0092, "step": 41130 }, { "grad_norm": 0.1163998693227768, "learning_rate": 6.834625540601734e-05, "loss": 0.0095, "step": 41140 }, { "grad_norm": 0.134515181183815, "learning_rate": 6.833087299213829e-05, "loss": 0.0086, "step": 41150 }, { "grad_norm": 0.1254095584154129, "learning_rate": 6.83154885736225e-05, "loss": 0.008, "step": 41160 }, { "grad_norm": 0.14031721651554108, "learning_rate": 6.830010215215244e-05, "loss": 0.0081, "step": 41170 }, { "grad_norm": 0.1444164663553238, "learning_rate": 6.828471372941072e-05, "loss": 0.0114, "step": 41180 }, { "grad_norm": 0.13363544642925262, "learning_rate": 6.826932330708018e-05, "loss": 0.0095, "step": 41190 }, { "grad_norm": 0.1734083890914917, "learning_rate": 6.825393088684393e-05, "loss": 0.0089, "step": 41200 }, { "grad_norm": 0.11671289801597595, "learning_rate": 6.823853647038524e-05, "loss": 0.0083, "step": 41210 }, { "grad_norm": 0.10070858150720596, "learning_rate": 6.822314005938762e-05, "loss": 0.0088, "step": 41220 }, { "grad_norm": 0.14407607913017273, "learning_rate": 6.820774165553479e-05, "loss": 0.0082, "step": 41230 }, { "grad_norm": 0.14623261988162994, "learning_rate": 6.819234126051072e-05, "loss": 0.009, "step": 41240 }, { "grad_norm": 0.09739971905946732, "learning_rate": 6.817693887599956e-05, "loss": 0.0072, "step": 41250 }, { "grad_norm": 0.09896792471408844, "learning_rate": 6.816153450368568e-05, "loss": 0.0098, "step": 41260 }, { "grad_norm": 0.11694931983947754, "learning_rate": 6.814612814525372e-05, "loss": 0.0094, "step": 41270 }, { "grad_norm": 0.1448153555393219, "learning_rate": 6.813071980238846e-05, "loss": 0.0089, "step": 41280 }, { "grad_norm": 0.13190552592277527, "learning_rate": 6.811530947677492e-05, "loss": 0.0094, "step": 41290 }, { "grad_norm": 0.2059209942817688, "learning_rate": 6.809989717009839e-05, "loss": 0.0111, "step": 41300 }, { "grad_norm": 0.15208059549331665, "learning_rate": 6.808448288404431e-05, "loss": 0.01, "step": 41310 }, { "grad_norm": 0.1661173552274704, "learning_rate": 6.806906662029838e-05, "loss": 0.0107, "step": 41320 }, { "grad_norm": 0.15112465620040894, "learning_rate": 6.805364838054647e-05, "loss": 0.011, "step": 41330 }, { "grad_norm": 0.15834006667137146, "learning_rate": 6.803822816647474e-05, "loss": 0.0106, "step": 41340 }, { "grad_norm": 0.16973280906677246, "learning_rate": 6.802280597976949e-05, "loss": 0.0105, "step": 41350 }, { "grad_norm": 0.1478353589773178, "learning_rate": 6.800738182211727e-05, "loss": 0.0098, "step": 41360 }, { "grad_norm": 0.1234484538435936, "learning_rate": 6.799195569520485e-05, "loss": 0.01, "step": 41370 }, { "grad_norm": 0.11488749831914902, "learning_rate": 6.79765276007192e-05, "loss": 0.0098, "step": 41380 }, { "grad_norm": 0.16424715518951416, "learning_rate": 6.796109754034753e-05, "loss": 0.0099, "step": 41390 }, { "grad_norm": 0.11559536308050156, "learning_rate": 6.794566551577724e-05, "loss": 0.0102, "step": 41400 }, { "grad_norm": 0.15860459208488464, "learning_rate": 6.793023152869594e-05, "loss": 0.0094, "step": 41410 }, { "grad_norm": 0.09884310513734818, "learning_rate": 6.791479558079147e-05, "loss": 0.0077, "step": 41420 }, { "grad_norm": 0.14017631113529205, "learning_rate": 6.789935767375191e-05, "loss": 0.0098, "step": 41430 }, { "grad_norm": 0.13639384508132935, "learning_rate": 6.78839178092655e-05, "loss": 0.0094, "step": 41440 }, { "grad_norm": 0.10109999775886536, "learning_rate": 6.786847598902072e-05, "loss": 0.0081, "step": 41450 }, { "grad_norm": 0.1022055521607399, "learning_rate": 6.785303221470627e-05, "loss": 0.0075, "step": 41460 }, { "grad_norm": 0.11865945905447006, "learning_rate": 6.783758648801108e-05, "loss": 0.0107, "step": 41470 }, { "grad_norm": 0.11779831349849701, "learning_rate": 6.782213881062422e-05, "loss": 0.0093, "step": 41480 }, { "grad_norm": 0.20831629633903503, "learning_rate": 6.780668918423508e-05, "loss": 0.0097, "step": 41490 }, { "grad_norm": 0.14396917819976807, "learning_rate": 6.779123761053317e-05, "loss": 0.0087, "step": 41500 }, { "grad_norm": 0.12804917991161346, "learning_rate": 6.777578409120826e-05, "loss": 0.0089, "step": 41510 }, { "grad_norm": 0.20178452134132385, "learning_rate": 6.776032862795033e-05, "loss": 0.0097, "step": 41520 }, { "grad_norm": 0.11603474617004395, "learning_rate": 6.774487122244956e-05, "loss": 0.0099, "step": 41530 }, { "grad_norm": 0.15190263092517853, "learning_rate": 6.772941187639637e-05, "loss": 0.0088, "step": 41540 }, { "grad_norm": 0.12042257934808731, "learning_rate": 6.771395059148134e-05, "loss": 0.0107, "step": 41550 }, { "grad_norm": 0.19421154260635376, "learning_rate": 6.76984873693953e-05, "loss": 0.0093, "step": 41560 }, { "grad_norm": 0.1349613070487976, "learning_rate": 6.768302221182931e-05, "loss": 0.0107, "step": 41570 }, { "grad_norm": 0.1279705911874771, "learning_rate": 6.766755512047457e-05, "loss": 0.01, "step": 41580 }, { "grad_norm": 0.16793076694011688, "learning_rate": 6.765208609702259e-05, "loss": 0.0118, "step": 41590 }, { "grad_norm": 0.1831832379102707, "learning_rate": 6.763661514316499e-05, "loss": 0.0098, "step": 41600 }, { "grad_norm": 0.14683346450328827, "learning_rate": 6.76211422605937e-05, "loss": 0.0095, "step": 41610 }, { "grad_norm": 0.10406367480754852, "learning_rate": 6.760566745100076e-05, "loss": 0.0103, "step": 41620 }, { "grad_norm": 0.13165397942066193, "learning_rate": 6.75901907160785e-05, "loss": 0.0099, "step": 41630 }, { "grad_norm": 0.17401407659053802, "learning_rate": 6.757471205751943e-05, "loss": 0.0094, "step": 41640 }, { "grad_norm": 0.14627785980701447, "learning_rate": 6.75592314770163e-05, "loss": 0.0076, "step": 41650 }, { "grad_norm": 0.1647762656211853, "learning_rate": 6.7543748976262e-05, "loss": 0.0099, "step": 41660 }, { "grad_norm": 0.11554840952157974, "learning_rate": 6.752826455694968e-05, "loss": 0.0096, "step": 41670 }, { "grad_norm": 0.12815512716770172, "learning_rate": 6.751277822077271e-05, "loss": 0.0079, "step": 41680 }, { "grad_norm": 0.11281295865774155, "learning_rate": 6.749728996942465e-05, "loss": 0.01, "step": 41690 }, { "grad_norm": 0.16630129516124725, "learning_rate": 6.748179980459924e-05, "loss": 0.0095, "step": 41700 }, { "grad_norm": 0.11748658120632172, "learning_rate": 6.746630772799052e-05, "loss": 0.0091, "step": 41710 }, { "grad_norm": 0.12878300249576569, "learning_rate": 6.745081374129262e-05, "loss": 0.0101, "step": 41720 }, { "grad_norm": 0.1424444168806076, "learning_rate": 6.743531784619998e-05, "loss": 0.0095, "step": 41730 }, { "grad_norm": 0.10853073000907898, "learning_rate": 6.741982004440719e-05, "loss": 0.0092, "step": 41740 }, { "grad_norm": 0.11282794177532196, "learning_rate": 6.740432033760907e-05, "loss": 0.008, "step": 41750 }, { "grad_norm": 0.10613839328289032, "learning_rate": 6.738881872750066e-05, "loss": 0.0097, "step": 41760 }, { "grad_norm": 0.16082151234149933, "learning_rate": 6.737331521577715e-05, "loss": 0.0102, "step": 41770 }, { "grad_norm": 0.16773289442062378, "learning_rate": 6.735780980413403e-05, "loss": 0.0119, "step": 41780 }, { "grad_norm": 0.10503093153238297, "learning_rate": 6.734230249426692e-05, "loss": 0.0102, "step": 41790 }, { "grad_norm": 0.14817707240581512, "learning_rate": 6.732679328787168e-05, "loss": 0.0085, "step": 41800 }, { "grad_norm": 0.14405235648155212, "learning_rate": 6.731128218664438e-05, "loss": 0.0078, "step": 41810 }, { "grad_norm": 0.14906777441501617, "learning_rate": 6.729576919228129e-05, "loss": 0.0088, "step": 41820 }, { "grad_norm": 0.13520784676074982, "learning_rate": 6.728025430647888e-05, "loss": 0.01, "step": 41830 }, { "grad_norm": 0.15000049769878387, "learning_rate": 6.726473753093383e-05, "loss": 0.0095, "step": 41840 }, { "grad_norm": 0.10638737678527832, "learning_rate": 6.724921886734305e-05, "loss": 0.0101, "step": 41850 }, { "grad_norm": 0.13258931040763855, "learning_rate": 6.723369831740363e-05, "loss": 0.0092, "step": 41860 }, { "grad_norm": 0.11256223917007446, "learning_rate": 6.721817588281288e-05, "loss": 0.008, "step": 41870 }, { "grad_norm": 0.14002500474452972, "learning_rate": 6.720265156526828e-05, "loss": 0.0099, "step": 41880 }, { "grad_norm": 0.1980847865343094, "learning_rate": 6.718712536646758e-05, "loss": 0.0082, "step": 41890 }, { "grad_norm": 0.14956849813461304, "learning_rate": 6.71715972881087e-05, "loss": 0.0106, "step": 41900 }, { "grad_norm": 0.13820819556713104, "learning_rate": 6.715606733188974e-05, "loss": 0.0092, "step": 41910 }, { "grad_norm": 0.12600503861904144, "learning_rate": 6.714053549950909e-05, "loss": 0.0076, "step": 41920 }, { "grad_norm": 0.12165175378322601, "learning_rate": 6.71250017926652e-05, "loss": 0.0091, "step": 41930 }, { "grad_norm": 0.1362849473953247, "learning_rate": 6.71094662130569e-05, "loss": 0.0089, "step": 41940 }, { "grad_norm": 0.14686621725559235, "learning_rate": 6.709392876238307e-05, "loss": 0.0104, "step": 41950 }, { "grad_norm": 0.10674037039279938, "learning_rate": 6.707838944234289e-05, "loss": 0.0098, "step": 41960 }, { "grad_norm": 0.14315295219421387, "learning_rate": 6.706284825463574e-05, "loss": 0.0111, "step": 41970 }, { "grad_norm": 0.11289066076278687, "learning_rate": 6.704730520096114e-05, "loss": 0.0094, "step": 41980 }, { "grad_norm": 0.11673729866743088, "learning_rate": 6.703176028301888e-05, "loss": 0.0078, "step": 41990 }, { "grad_norm": 0.09116089344024658, "learning_rate": 6.701621350250892e-05, "loss": 0.0077, "step": 42000 }, { "grad_norm": 0.12021001428365707, "learning_rate": 6.700066486113144e-05, "loss": 0.0095, "step": 42010 }, { "grad_norm": 0.14358514547348022, "learning_rate": 6.69851143605868e-05, "loss": 0.0093, "step": 42020 }, { "grad_norm": 0.1522197127342224, "learning_rate": 6.696956200257561e-05, "loss": 0.0086, "step": 42030 }, { "grad_norm": 0.10782351344823837, "learning_rate": 6.695400778879863e-05, "loss": 0.0082, "step": 42040 }, { "grad_norm": 0.11555954068899155, "learning_rate": 6.693845172095683e-05, "loss": 0.0085, "step": 42050 }, { "grad_norm": 0.12744192779064178, "learning_rate": 6.692289380075142e-05, "loss": 0.0109, "step": 42060 }, { "grad_norm": 0.13212579488754272, "learning_rate": 6.690733402988379e-05, "loss": 0.013, "step": 42070 }, { "grad_norm": 0.13182687759399414, "learning_rate": 6.689177241005553e-05, "loss": 0.0103, "step": 42080 }, { "grad_norm": 0.10961862653493881, "learning_rate": 6.687620894296844e-05, "loss": 0.0079, "step": 42090 }, { "grad_norm": 0.15292328596115112, "learning_rate": 6.686064363032451e-05, "loss": 0.0087, "step": 42100 }, { "grad_norm": 0.12195099890232086, "learning_rate": 6.684507647382596e-05, "loss": 0.0105, "step": 42110 }, { "grad_norm": 0.11141618341207504, "learning_rate": 6.682950747517513e-05, "loss": 0.0079, "step": 42120 }, { "grad_norm": 0.11622217297554016, "learning_rate": 6.68139366360747e-05, "loss": 0.0083, "step": 42130 }, { "grad_norm": 0.1419483870267868, "learning_rate": 6.679836395822744e-05, "loss": 0.0091, "step": 42140 }, { "grad_norm": 0.17394565045833588, "learning_rate": 6.678278944333633e-05, "loss": 0.01, "step": 42150 }, { "grad_norm": 0.11621260643005371, "learning_rate": 6.676721309310462e-05, "loss": 0.0097, "step": 42160 }, { "grad_norm": 0.1481286734342575, "learning_rate": 6.67516349092357e-05, "loss": 0.0106, "step": 42170 }, { "grad_norm": 0.151369109749794, "learning_rate": 6.673605489343315e-05, "loss": 0.0103, "step": 42180 }, { "grad_norm": 0.14645911753177643, "learning_rate": 6.672047304740081e-05, "loss": 0.0103, "step": 42190 }, { "grad_norm": 0.10993500798940659, "learning_rate": 6.670488937284268e-05, "loss": 0.01, "step": 42200 }, { "grad_norm": 0.09861057996749878, "learning_rate": 6.668930387146294e-05, "loss": 0.0094, "step": 42210 }, { "grad_norm": 0.11963243037462234, "learning_rate": 6.667371654496605e-05, "loss": 0.0096, "step": 42220 }, { "grad_norm": 0.11783411353826523, "learning_rate": 6.665812739505659e-05, "loss": 0.0082, "step": 42230 }, { "grad_norm": 0.1460486799478531, "learning_rate": 6.664253642343935e-05, "loss": 0.0084, "step": 42240 }, { "grad_norm": 0.1144520714879036, "learning_rate": 6.662694363181935e-05, "loss": 0.0082, "step": 42250 }, { "grad_norm": 0.12866201996803284, "learning_rate": 6.66113490219018e-05, "loss": 0.0107, "step": 42260 }, { "grad_norm": 0.1370762139558792, "learning_rate": 6.659575259539207e-05, "loss": 0.009, "step": 42270 }, { "grad_norm": 0.12253224849700928, "learning_rate": 6.65801543539958e-05, "loss": 0.0088, "step": 42280 }, { "grad_norm": 0.13917654752731323, "learning_rate": 6.65645542994188e-05, "loss": 0.0094, "step": 42290 }, { "grad_norm": 0.14651641249656677, "learning_rate": 6.654895243336702e-05, "loss": 0.0085, "step": 42300 }, { "grad_norm": 0.16497090458869934, "learning_rate": 6.65333487575467e-05, "loss": 0.0112, "step": 42310 }, { "grad_norm": 0.11036188155412674, "learning_rate": 6.651774327366422e-05, "loss": 0.0112, "step": 42320 }, { "grad_norm": 0.16559049487113953, "learning_rate": 6.650213598342613e-05, "loss": 0.0115, "step": 42330 }, { "grad_norm": 0.15614348649978638, "learning_rate": 6.648652688853931e-05, "loss": 0.0095, "step": 42340 }, { "grad_norm": 0.10486141592264175, "learning_rate": 6.647091599071066e-05, "loss": 0.0092, "step": 42350 }, { "grad_norm": 0.13050051033496857, "learning_rate": 6.645530329164743e-05, "loss": 0.0089, "step": 42360 }, { "grad_norm": 0.13935919106006622, "learning_rate": 6.643968879305694e-05, "loss": 0.0109, "step": 42370 }, { "grad_norm": 0.1432427018880844, "learning_rate": 6.642407249664683e-05, "loss": 0.0101, "step": 42380 }, { "grad_norm": 0.1379433423280716, "learning_rate": 6.640845440412483e-05, "loss": 0.009, "step": 42390 }, { "grad_norm": 0.10986980050802231, "learning_rate": 6.639283451719893e-05, "loss": 0.009, "step": 42400 }, { "grad_norm": 0.11804080009460449, "learning_rate": 6.637721283757729e-05, "loss": 0.0094, "step": 42410 }, { "grad_norm": 0.16039511561393738, "learning_rate": 6.636158936696828e-05, "loss": 0.0106, "step": 42420 }, { "grad_norm": 0.1302924007177353, "learning_rate": 6.634596410708047e-05, "loss": 0.0117, "step": 42430 }, { "grad_norm": 0.15018241107463837, "learning_rate": 6.633033705962257e-05, "loss": 0.0109, "step": 42440 }, { "grad_norm": 0.17914637923240662, "learning_rate": 6.631470822630359e-05, "loss": 0.0111, "step": 42450 }, { "grad_norm": 0.12773412466049194, "learning_rate": 6.629907760883263e-05, "loss": 0.0113, "step": 42460 }, { "grad_norm": 0.11301779001951218, "learning_rate": 6.628344520891905e-05, "loss": 0.0108, "step": 42470 }, { "grad_norm": 0.12389296293258667, "learning_rate": 6.62678110282724e-05, "loss": 0.0093, "step": 42480 }, { "grad_norm": 0.190986767411232, "learning_rate": 6.625217506860237e-05, "loss": 0.0112, "step": 42490 }, { "grad_norm": 0.16125260293483734, "learning_rate": 6.623653733161892e-05, "loss": 0.0105, "step": 42500 }, { "grad_norm": 0.10772192478179932, "learning_rate": 6.622089781903217e-05, "loss": 0.0087, "step": 42510 }, { "grad_norm": 0.14447453618049622, "learning_rate": 6.620525653255241e-05, "loss": 0.0109, "step": 42520 }, { "grad_norm": 0.13638867437839508, "learning_rate": 6.618961347389015e-05, "loss": 0.0108, "step": 42530 }, { "grad_norm": 0.13878415524959564, "learning_rate": 6.617396864475613e-05, "loss": 0.01, "step": 42540 }, { "grad_norm": 0.11926313489675522, "learning_rate": 6.61583220468612e-05, "loss": 0.0108, "step": 42550 }, { "grad_norm": 0.16355443000793457, "learning_rate": 6.614267368191645e-05, "loss": 0.0102, "step": 42560 }, { "grad_norm": 0.14913854002952576, "learning_rate": 6.61270235516332e-05, "loss": 0.0091, "step": 42570 }, { "grad_norm": 0.10545974969863892, "learning_rate": 6.611137165772287e-05, "loss": 0.0096, "step": 42580 }, { "grad_norm": 0.1351899802684784, "learning_rate": 6.60957180018972e-05, "loss": 0.0095, "step": 42590 }, { "grad_norm": 0.14528702199459076, "learning_rate": 6.608006258586797e-05, "loss": 0.0082, "step": 42600 }, { "grad_norm": 0.19139450788497925, "learning_rate": 6.606440541134731e-05, "loss": 0.01, "step": 42610 }, { "grad_norm": 0.11675820499658585, "learning_rate": 6.60487464800474e-05, "loss": 0.0078, "step": 42620 }, { "grad_norm": 0.13771386444568634, "learning_rate": 6.603308579368071e-05, "loss": 0.0095, "step": 42630 }, { "grad_norm": 0.15265627205371857, "learning_rate": 6.601742335395987e-05, "loss": 0.0101, "step": 42640 }, { "grad_norm": 0.1012314185500145, "learning_rate": 6.600175916259769e-05, "loss": 0.0078, "step": 42650 }, { "grad_norm": 0.16655468940734863, "learning_rate": 6.598609322130718e-05, "loss": 0.0099, "step": 42660 }, { "grad_norm": 0.12415699660778046, "learning_rate": 6.597042553180155e-05, "loss": 0.009, "step": 42670 }, { "grad_norm": 0.11017493158578873, "learning_rate": 6.59547560957942e-05, "loss": 0.0094, "step": 42680 }, { "grad_norm": 0.11261948198080063, "learning_rate": 6.59390849149987e-05, "loss": 0.0103, "step": 42690 }, { "grad_norm": 0.1510520875453949, "learning_rate": 6.592341199112886e-05, "loss": 0.0114, "step": 42700 }, { "grad_norm": 0.11460937559604645, "learning_rate": 6.590773732589861e-05, "loss": 0.0089, "step": 42710 }, { "grad_norm": 0.13019606471061707, "learning_rate": 6.589206092102213e-05, "loss": 0.0099, "step": 42720 }, { "grad_norm": 0.11815770715475082, "learning_rate": 6.587638277821375e-05, "loss": 0.0092, "step": 42730 }, { "grad_norm": 0.11982028931379318, "learning_rate": 6.586070289918803e-05, "loss": 0.0082, "step": 42740 }, { "grad_norm": 0.1517113894224167, "learning_rate": 6.584502128565968e-05, "loss": 0.01, "step": 42750 }, { "grad_norm": 0.1415819674730301, "learning_rate": 6.582933793934363e-05, "loss": 0.0078, "step": 42760 }, { "grad_norm": 0.12380024790763855, "learning_rate": 6.581365286195498e-05, "loss": 0.0085, "step": 42770 }, { "grad_norm": 0.14630115032196045, "learning_rate": 6.579796605520903e-05, "loss": 0.0113, "step": 42780 }, { "grad_norm": 0.15587159991264343, "learning_rate": 6.578227752082127e-05, "loss": 0.01, "step": 42790 }, { "grad_norm": 0.14192728698253632, "learning_rate": 6.576658726050735e-05, "loss": 0.0082, "step": 42800 }, { "grad_norm": 0.13600578904151917, "learning_rate": 6.575089527598316e-05, "loss": 0.0097, "step": 42810 }, { "grad_norm": 0.13410457968711853, "learning_rate": 6.573520156896476e-05, "loss": 0.0104, "step": 42820 }, { "grad_norm": 0.15178070962429047, "learning_rate": 6.571950614116835e-05, "loss": 0.0097, "step": 42830 }, { "grad_norm": 0.1910736858844757, "learning_rate": 6.570380899431039e-05, "loss": 0.0109, "step": 42840 }, { "grad_norm": 0.1292385458946228, "learning_rate": 6.568811013010749e-05, "loss": 0.0117, "step": 42850 }, { "grad_norm": 0.12635958194732666, "learning_rate": 6.567240955027645e-05, "loss": 0.0082, "step": 42860 }, { "grad_norm": 0.21810804307460785, "learning_rate": 6.565670725653427e-05, "loss": 0.0082, "step": 42870 }, { "grad_norm": 0.16796168684959412, "learning_rate": 6.564100325059812e-05, "loss": 0.0085, "step": 42880 }, { "grad_norm": 0.09749284386634827, "learning_rate": 6.562529753418538e-05, "loss": 0.0087, "step": 42890 }, { "grad_norm": 0.14011994004249573, "learning_rate": 6.56095901090136e-05, "loss": 0.0087, "step": 42900 }, { "grad_norm": 0.16207991540431976, "learning_rate": 6.55938809768005e-05, "loss": 0.0085, "step": 42910 }, { "grad_norm": 0.11686359345912933, "learning_rate": 6.557817013926402e-05, "loss": 0.0108, "step": 42920 }, { "grad_norm": 0.09591418504714966, "learning_rate": 6.556245759812227e-05, "loss": 0.0082, "step": 42930 }, { "grad_norm": 0.12515181303024292, "learning_rate": 6.554674335509357e-05, "loss": 0.0079, "step": 42940 }, { "grad_norm": 0.12878544628620148, "learning_rate": 6.553102741189638e-05, "loss": 0.0089, "step": 42950 }, { "grad_norm": 0.13251662254333496, "learning_rate": 6.55153097702494e-05, "loss": 0.0091, "step": 42960 }, { "grad_norm": 0.13871948421001434, "learning_rate": 6.549959043187146e-05, "loss": 0.0126, "step": 42970 }, { "grad_norm": 0.12313912063837051, "learning_rate": 6.54838693984816e-05, "loss": 0.0098, "step": 42980 }, { "grad_norm": 0.10033681243658066, "learning_rate": 6.546814667179909e-05, "loss": 0.0101, "step": 42990 }, { "grad_norm": 0.11822658777236938, "learning_rate": 6.545242225354328e-05, "loss": 0.0108, "step": 43000 }, { "grad_norm": 0.1283096969127655, "learning_rate": 6.543669614543383e-05, "loss": 0.0104, "step": 43010 }, { "grad_norm": 0.16403470933437347, "learning_rate": 6.542096834919049e-05, "loss": 0.0108, "step": 43020 }, { "grad_norm": 0.13657037913799286, "learning_rate": 6.540523886653324e-05, "loss": 0.0102, "step": 43030 }, { "grad_norm": 0.1255374550819397, "learning_rate": 6.538950769918222e-05, "loss": 0.0107, "step": 43040 }, { "grad_norm": 0.1326923817396164, "learning_rate": 6.537377484885779e-05, "loss": 0.01, "step": 43050 }, { "grad_norm": 0.1494813859462738, "learning_rate": 6.535804031728045e-05, "loss": 0.0091, "step": 43060 }, { "grad_norm": 0.1931881159543991, "learning_rate": 6.534230410617092e-05, "loss": 0.0112, "step": 43070 }, { "grad_norm": 0.15942563116550446, "learning_rate": 6.532656621725007e-05, "loss": 0.0094, "step": 43080 }, { "grad_norm": 0.13609446585178375, "learning_rate": 6.531082665223898e-05, "loss": 0.0107, "step": 43090 }, { "grad_norm": 0.11931334435939789, "learning_rate": 6.529508541285889e-05, "loss": 0.009, "step": 43100 }, { "grad_norm": 0.1173621341586113, "learning_rate": 6.527934250083128e-05, "loss": 0.0077, "step": 43110 }, { "grad_norm": 0.11961666494607925, "learning_rate": 6.526359791787775e-05, "loss": 0.0092, "step": 43120 }, { "grad_norm": 0.1350565105676651, "learning_rate": 6.524785166572009e-05, "loss": 0.009, "step": 43130 }, { "grad_norm": 0.13578559458255768, "learning_rate": 6.523210374608029e-05, "loss": 0.0096, "step": 43140 }, { "grad_norm": 0.11543253809213638, "learning_rate": 6.521635416068054e-05, "loss": 0.0102, "step": 43150 }, { "grad_norm": 0.139207124710083, "learning_rate": 6.520060291124317e-05, "loss": 0.0108, "step": 43160 }, { "grad_norm": 0.2594671845436096, "learning_rate": 6.518484999949072e-05, "loss": 0.0118, "step": 43170 }, { "grad_norm": 0.13598863780498505, "learning_rate": 6.516909542714591e-05, "loss": 0.0093, "step": 43180 }, { "grad_norm": 0.1265650987625122, "learning_rate": 6.515333919593161e-05, "loss": 0.0087, "step": 43190 }, { "grad_norm": 0.15671491622924805, "learning_rate": 6.513758130757094e-05, "loss": 0.0105, "step": 43200 }, { "grad_norm": 0.15209674835205078, "learning_rate": 6.512182176378713e-05, "loss": 0.0096, "step": 43210 }, { "grad_norm": 0.10997172445058823, "learning_rate": 6.510606056630362e-05, "loss": 0.0109, "step": 43220 }, { "grad_norm": 0.08889241516590118, "learning_rate": 6.509029771684405e-05, "loss": 0.0103, "step": 43230 }, { "grad_norm": 0.11025245487689972, "learning_rate": 6.50745332171322e-05, "loss": 0.0096, "step": 43240 }, { "grad_norm": 0.19050423800945282, "learning_rate": 6.505876706889207e-05, "loss": 0.0085, "step": 43250 }, { "grad_norm": 0.1254732459783554, "learning_rate": 6.504299927384781e-05, "loss": 0.009, "step": 43260 }, { "grad_norm": 0.12297430634498596, "learning_rate": 6.502722983372378e-05, "loss": 0.0102, "step": 43270 }, { "grad_norm": 0.14020170271396637, "learning_rate": 6.501145875024446e-05, "loss": 0.0105, "step": 43280 }, { "grad_norm": 0.12477561086416245, "learning_rate": 6.499568602513459e-05, "loss": 0.0084, "step": 43290 }, { "grad_norm": 0.1323445588350296, "learning_rate": 6.497991166011903e-05, "loss": 0.0086, "step": 43300 }, { "grad_norm": 0.14025913178920746, "learning_rate": 6.49641356569229e-05, "loss": 0.0079, "step": 43310 }, { "grad_norm": 0.1389862596988678, "learning_rate": 6.494835801727135e-05, "loss": 0.0102, "step": 43320 }, { "grad_norm": 0.15170271694660187, "learning_rate": 6.493257874288986e-05, "loss": 0.0103, "step": 43330 }, { "grad_norm": 0.1527126580476761, "learning_rate": 6.491679783550399e-05, "loss": 0.0086, "step": 43340 }, { "grad_norm": 0.1125045046210289, "learning_rate": 6.490101529683957e-05, "loss": 0.009, "step": 43350 }, { "grad_norm": 0.13279038667678833, "learning_rate": 6.488523112862248e-05, "loss": 0.0083, "step": 43360 }, { "grad_norm": 0.10353706032037735, "learning_rate": 6.486944533257891e-05, "loss": 0.0096, "step": 43370 }, { "grad_norm": 0.15680591762065887, "learning_rate": 6.485365791043517e-05, "loss": 0.0101, "step": 43380 }, { "grad_norm": 0.09820320457220078, "learning_rate": 6.483786886391772e-05, "loss": 0.0086, "step": 43390 }, { "grad_norm": 0.11309076845645905, "learning_rate": 6.482207819475323e-05, "loss": 0.0083, "step": 43400 }, { "grad_norm": 0.12172795087099075, "learning_rate": 6.480628590466857e-05, "loss": 0.0091, "step": 43410 }, { "grad_norm": 0.4593230187892914, "learning_rate": 6.479049199539074e-05, "loss": 0.0084, "step": 43420 }, { "grad_norm": 0.13455000519752502, "learning_rate": 6.477469646864693e-05, "loss": 0.0075, "step": 43430 }, { "grad_norm": 0.09679724276065826, "learning_rate": 6.475889932616454e-05, "loss": 0.0082, "step": 43440 }, { "grad_norm": 0.1407289057970047, "learning_rate": 6.474310056967111e-05, "loss": 0.0082, "step": 43450 }, { "grad_norm": 0.1143527626991272, "learning_rate": 6.472730020089437e-05, "loss": 0.0081, "step": 43460 }, { "grad_norm": 0.13527539372444153, "learning_rate": 6.471149822156222e-05, "loss": 0.0078, "step": 43470 }, { "grad_norm": 0.0921812430024147, "learning_rate": 6.469569463340275e-05, "loss": 0.009, "step": 43480 }, { "grad_norm": 0.1286616027355194, "learning_rate": 6.467988943814418e-05, "loss": 0.0086, "step": 43490 }, { "grad_norm": 0.17935244739055634, "learning_rate": 6.4664082637515e-05, "loss": 0.0085, "step": 43500 }, { "grad_norm": 0.18075579404830933, "learning_rate": 6.464827423324377e-05, "loss": 0.0092, "step": 43510 }, { "grad_norm": 0.10215975344181061, "learning_rate": 6.463246422705931e-05, "loss": 0.0083, "step": 43520 }, { "grad_norm": 0.14709056913852692, "learning_rate": 6.461665262069055e-05, "loss": 0.0092, "step": 43530 }, { "grad_norm": 0.11439597606658936, "learning_rate": 6.460083941586665e-05, "loss": 0.0085, "step": 43540 }, { "grad_norm": 0.13454288244247437, "learning_rate": 6.45850246143169e-05, "loss": 0.0098, "step": 43550 }, { "grad_norm": 0.17783717811107635, "learning_rate": 6.456920821777077e-05, "loss": 0.0107, "step": 43560 }, { "grad_norm": 0.12967626750469208, "learning_rate": 6.455339022795796e-05, "loss": 0.0092, "step": 43570 }, { "grad_norm": 0.09902764856815338, "learning_rate": 6.453757064660825e-05, "loss": 0.0083, "step": 43580 }, { "grad_norm": 0.15856638550758362, "learning_rate": 6.452174947545169e-05, "loss": 0.0097, "step": 43590 }, { "grad_norm": 0.12587550282478333, "learning_rate": 6.450592671621842e-05, "loss": 0.0096, "step": 43600 }, { "grad_norm": 0.11344518512487411, "learning_rate": 6.449010237063882e-05, "loss": 0.0085, "step": 43610 }, { "grad_norm": 0.12241800874471664, "learning_rate": 6.447427644044343e-05, "loss": 0.0094, "step": 43620 }, { "grad_norm": 0.13825467228889465, "learning_rate": 6.44584489273629e-05, "loss": 0.0091, "step": 43630 }, { "grad_norm": 0.131740540266037, "learning_rate": 6.444261983312817e-05, "loss": 0.0086, "step": 43640 }, { "grad_norm": 0.12904445827007294, "learning_rate": 6.442678915947023e-05, "loss": 0.0093, "step": 43650 }, { "grad_norm": 0.11655455082654953, "learning_rate": 6.441095690812032e-05, "loss": 0.0105, "step": 43660 }, { "grad_norm": 0.1341511756181717, "learning_rate": 6.439512308080982e-05, "loss": 0.0088, "step": 43670 }, { "grad_norm": 0.12453373521566391, "learning_rate": 6.43792876792703e-05, "loss": 0.0083, "step": 43680 }, { "grad_norm": 0.13608671724796295, "learning_rate": 6.43634507052335e-05, "loss": 0.0107, "step": 43690 }, { "grad_norm": 0.12309335917234421, "learning_rate": 6.434761216043135e-05, "loss": 0.0099, "step": 43700 }, { "grad_norm": 0.1794234812259674, "learning_rate": 6.433177204659589e-05, "loss": 0.0097, "step": 43710 }, { "grad_norm": 0.11929603666067123, "learning_rate": 6.431593036545938e-05, "loss": 0.0099, "step": 43720 }, { "grad_norm": 0.1362350434064865, "learning_rate": 6.430008711875426e-05, "loss": 0.0092, "step": 43730 }, { "grad_norm": 0.17378774285316467, "learning_rate": 6.42842423082131e-05, "loss": 0.0086, "step": 43740 }, { "grad_norm": 0.13058747351169586, "learning_rate": 6.426839593556868e-05, "loss": 0.01, "step": 43750 }, { "grad_norm": 0.1634233593940735, "learning_rate": 6.425254800255392e-05, "loss": 0.011, "step": 43760 }, { "grad_norm": 0.14272448420524597, "learning_rate": 6.423669851090196e-05, "loss": 0.0092, "step": 43770 }, { "grad_norm": 0.12721750140190125, "learning_rate": 6.422084746234604e-05, "loss": 0.0102, "step": 43780 }, { "grad_norm": 0.12886539101600647, "learning_rate": 6.420499485861965e-05, "loss": 0.0089, "step": 43790 }, { "grad_norm": 0.10174175351858139, "learning_rate": 6.418914070145636e-05, "loss": 0.0118, "step": 43800 }, { "grad_norm": 0.11529391258955002, "learning_rate": 6.417328499258996e-05, "loss": 0.0101, "step": 43810 }, { "grad_norm": 0.1682642251253128, "learning_rate": 6.415742773375446e-05, "loss": 0.0082, "step": 43820 }, { "grad_norm": 0.13248033821582794, "learning_rate": 6.414156892668393e-05, "loss": 0.0088, "step": 43830 }, { "grad_norm": 0.10360054671764374, "learning_rate": 6.412570857311267e-05, "loss": 0.0105, "step": 43840 }, { "grad_norm": 0.14063489437103271, "learning_rate": 6.410984667477518e-05, "loss": 0.0105, "step": 43850 }, { "grad_norm": 0.1260899156332016, "learning_rate": 6.409398323340607e-05, "loss": 0.0091, "step": 43860 }, { "grad_norm": 0.15215545892715454, "learning_rate": 6.407811825074012e-05, "loss": 0.0087, "step": 43870 }, { "grad_norm": 0.10509326308965683, "learning_rate": 6.406225172851234e-05, "loss": 0.0091, "step": 43880 }, { "grad_norm": 0.12519733607769012, "learning_rate": 6.404638366845786e-05, "loss": 0.0097, "step": 43890 }, { "grad_norm": 0.11049649864435196, "learning_rate": 6.403051407231196e-05, "loss": 0.0097, "step": 43900 }, { "grad_norm": 0.14329761266708374, "learning_rate": 6.401464294181016e-05, "loss": 0.0109, "step": 43910 }, { "grad_norm": 0.130090594291687, "learning_rate": 6.399877027868808e-05, "loss": 0.0098, "step": 43920 }, { "grad_norm": 0.12205823510885239, "learning_rate": 6.39828960846815e-05, "loss": 0.0087, "step": 43930 }, { "grad_norm": 0.13228945434093475, "learning_rate": 6.396702036152644e-05, "loss": 0.0091, "step": 43940 }, { "grad_norm": 0.12083250284194946, "learning_rate": 6.395114311095905e-05, "loss": 0.0091, "step": 43950 }, { "grad_norm": 0.09560081362724304, "learning_rate": 6.393526433471559e-05, "loss": 0.008, "step": 43960 }, { "grad_norm": 0.11678052693605423, "learning_rate": 6.391938403453262e-05, "loss": 0.0082, "step": 43970 }, { "grad_norm": 0.12931261956691742, "learning_rate": 6.390350221214671e-05, "loss": 0.0091, "step": 43980 }, { "grad_norm": 0.16706381738185883, "learning_rate": 6.38876188692947e-05, "loss": 0.0096, "step": 43990 }, { "grad_norm": 0.11385173350572586, "learning_rate": 6.387173400771358e-05, "loss": 0.008, "step": 44000 }, { "grad_norm": 0.11057750135660172, "learning_rate": 6.385584762914047e-05, "loss": 0.0073, "step": 44010 }, { "grad_norm": 0.0962032675743103, "learning_rate": 6.38399597353127e-05, "loss": 0.0072, "step": 44020 }, { "grad_norm": 0.11170385032892227, "learning_rate": 6.382407032796775e-05, "loss": 0.0094, "step": 44030 }, { "grad_norm": 0.09056857228279114, "learning_rate": 6.380817940884325e-05, "loss": 0.008, "step": 44040 }, { "grad_norm": 0.1324126422405243, "learning_rate": 6.379228697967702e-05, "loss": 0.0089, "step": 44050 }, { "grad_norm": 0.14941275119781494, "learning_rate": 6.3776393042207e-05, "loss": 0.0083, "step": 44060 }, { "grad_norm": 0.10833548754453659, "learning_rate": 6.376049759817137e-05, "loss": 0.0082, "step": 44070 }, { "grad_norm": 0.09763576835393906, "learning_rate": 6.37446006493084e-05, "loss": 0.0087, "step": 44080 }, { "grad_norm": 0.13520127534866333, "learning_rate": 6.372870219735656e-05, "loss": 0.0087, "step": 44090 }, { "grad_norm": 0.1286962628364563, "learning_rate": 6.371280224405452e-05, "loss": 0.0087, "step": 44100 }, { "grad_norm": 0.10016630589962006, "learning_rate": 6.369690079114101e-05, "loss": 0.01, "step": 44110 }, { "grad_norm": 0.09802474081516266, "learning_rate": 6.368099784035504e-05, "loss": 0.0085, "step": 44120 }, { "grad_norm": 0.13511115312576294, "learning_rate": 6.366509339343572e-05, "loss": 0.0083, "step": 44130 }, { "grad_norm": 0.15260308980941772, "learning_rate": 6.364918745212232e-05, "loss": 0.0079, "step": 44140 }, { "grad_norm": 0.13737444579601288, "learning_rate": 6.363328001815431e-05, "loss": 0.0097, "step": 44150 }, { "grad_norm": 0.17781402170658112, "learning_rate": 6.361737109327128e-05, "loss": 0.0108, "step": 44160 }, { "grad_norm": 0.12373196333646774, "learning_rate": 6.360146067921301e-05, "loss": 0.0105, "step": 44170 }, { "grad_norm": 0.12301576882600784, "learning_rate": 6.35855487777195e-05, "loss": 0.0079, "step": 44180 }, { "grad_norm": 0.12281149625778198, "learning_rate": 6.356963539053076e-05, "loss": 0.0068, "step": 44190 }, { "grad_norm": 0.10989195853471756, "learning_rate": 6.355372051938711e-05, "loss": 0.0077, "step": 44200 }, { "grad_norm": 0.11636555194854736, "learning_rate": 6.353780416602894e-05, "loss": 0.01, "step": 44210 }, { "grad_norm": 0.11716995388269424, "learning_rate": 6.352188633219689e-05, "loss": 0.0091, "step": 44220 }, { "grad_norm": 0.16048145294189453, "learning_rate": 6.350596701963166e-05, "loss": 0.0095, "step": 44230 }, { "grad_norm": 0.15837779641151428, "learning_rate": 6.349004623007419e-05, "loss": 0.0113, "step": 44240 }, { "grad_norm": 0.15082035958766937, "learning_rate": 6.347412396526555e-05, "loss": 0.0095, "step": 44250 }, { "grad_norm": 0.16540366411209106, "learning_rate": 6.345820022694696e-05, "loss": 0.0101, "step": 44260 }, { "grad_norm": 0.12118475139141083, "learning_rate": 6.344227501685984e-05, "loss": 0.0105, "step": 44270 }, { "grad_norm": 0.13853423297405243, "learning_rate": 6.342634833674572e-05, "loss": 0.0111, "step": 44280 }, { "grad_norm": 0.20339789986610413, "learning_rate": 6.341042018834635e-05, "loss": 0.0098, "step": 44290 }, { "grad_norm": 0.1462336778640747, "learning_rate": 6.339449057340359e-05, "loss": 0.0086, "step": 44300 }, { "grad_norm": 0.12818998098373413, "learning_rate": 6.337855949365945e-05, "loss": 0.0085, "step": 44310 }, { "grad_norm": 0.12330733239650726, "learning_rate": 6.336262695085619e-05, "loss": 0.0106, "step": 44320 }, { "grad_norm": 0.10955051332712173, "learning_rate": 6.334669294673612e-05, "loss": 0.0089, "step": 44330 }, { "grad_norm": 0.106901153922081, "learning_rate": 6.333075748304179e-05, "loss": 0.0086, "step": 44340 }, { "grad_norm": 0.13672317564487457, "learning_rate": 6.331482056151585e-05, "loss": 0.0076, "step": 44350 }, { "grad_norm": 0.12499171495437622, "learning_rate": 6.329888218390117e-05, "loss": 0.0081, "step": 44360 }, { "grad_norm": 0.11652278155088425, "learning_rate": 6.328294235194072e-05, "loss": 0.0076, "step": 44370 }, { "grad_norm": 0.1285354197025299, "learning_rate": 6.326700106737765e-05, "loss": 0.0087, "step": 44380 }, { "grad_norm": 0.11558280885219574, "learning_rate": 6.325105833195531e-05, "loss": 0.0084, "step": 44390 }, { "grad_norm": 0.11560893058776855, "learning_rate": 6.323511414741715e-05, "loss": 0.0077, "step": 44400 }, { "grad_norm": 0.15229371190071106, "learning_rate": 6.321916851550678e-05, "loss": 0.008, "step": 44410 }, { "grad_norm": 0.13944435119628906, "learning_rate": 6.320322143796806e-05, "loss": 0.0093, "step": 44420 }, { "grad_norm": 0.14105670154094696, "learning_rate": 6.318727291654488e-05, "loss": 0.0077, "step": 44430 }, { "grad_norm": 0.11121176928281784, "learning_rate": 6.317132295298134e-05, "loss": 0.0093, "step": 44440 }, { "grad_norm": 0.12213848531246185, "learning_rate": 6.315537154902173e-05, "loss": 0.0094, "step": 44450 }, { "grad_norm": 0.09287367761135101, "learning_rate": 6.313941870641048e-05, "loss": 0.0076, "step": 44460 }, { "grad_norm": 0.14526966214179993, "learning_rate": 6.312346442689214e-05, "loss": 0.0086, "step": 44470 }, { "grad_norm": 0.1267419308423996, "learning_rate": 6.310750871221147e-05, "loss": 0.0083, "step": 44480 }, { "grad_norm": 0.099644735455513, "learning_rate": 6.309155156411335e-05, "loss": 0.0067, "step": 44490 }, { "grad_norm": 0.11071322858333588, "learning_rate": 6.307559298434284e-05, "loss": 0.0095, "step": 44500 }, { "grad_norm": 0.1283435970544815, "learning_rate": 6.305963297464512e-05, "loss": 0.0074, "step": 44510 }, { "grad_norm": 0.08965405076742172, "learning_rate": 6.304367153676561e-05, "loss": 0.0072, "step": 44520 }, { "grad_norm": 0.17037856578826904, "learning_rate": 6.302770867244978e-05, "loss": 0.0103, "step": 44530 }, { "grad_norm": 0.14335854351520538, "learning_rate": 6.301174438344328e-05, "loss": 0.0091, "step": 44540 }, { "grad_norm": 0.1345544308423996, "learning_rate": 6.299577867149202e-05, "loss": 0.0095, "step": 44550 }, { "grad_norm": 0.14531093835830688, "learning_rate": 6.297981153834192e-05, "loss": 0.0085, "step": 44560 }, { "grad_norm": 0.10801568627357483, "learning_rate": 6.296384298573916e-05, "loss": 0.0116, "step": 44570 }, { "grad_norm": 0.13795505464076996, "learning_rate": 6.294787301543001e-05, "loss": 0.0099, "step": 44580 }, { "grad_norm": 0.15126265585422516, "learning_rate": 6.293190162916095e-05, "loss": 0.011, "step": 44590 }, { "grad_norm": 0.11032062768936157, "learning_rate": 6.291592882867855e-05, "loss": 0.008, "step": 44600 }, { "grad_norm": 0.13788369297981262, "learning_rate": 6.28999546157296e-05, "loss": 0.0089, "step": 44610 }, { "grad_norm": 0.10699989646673203, "learning_rate": 6.288397899206102e-05, "loss": 0.0079, "step": 44620 }, { "grad_norm": 0.12580841779708862, "learning_rate": 6.286800195941984e-05, "loss": 0.009, "step": 44630 }, { "grad_norm": 0.09577760845422745, "learning_rate": 6.285202351955334e-05, "loss": 0.0081, "step": 44640 }, { "grad_norm": 0.09499794989824295, "learning_rate": 6.283604367420887e-05, "loss": 0.0091, "step": 44650 }, { "grad_norm": 0.09129393845796585, "learning_rate": 6.282006242513394e-05, "loss": 0.0086, "step": 44660 }, { "grad_norm": 0.14082437753677368, "learning_rate": 6.280407977407628e-05, "loss": 0.0087, "step": 44670 }, { "grad_norm": 0.1329466849565506, "learning_rate": 6.27880957227837e-05, "loss": 0.0109, "step": 44680 }, { "grad_norm": 0.13599301874637604, "learning_rate": 6.27721102730042e-05, "loss": 0.0098, "step": 44690 }, { "grad_norm": 0.1276695430278778, "learning_rate": 6.275612342648592e-05, "loss": 0.0094, "step": 44700 }, { "grad_norm": 0.1744931936264038, "learning_rate": 6.274013518497716e-05, "loss": 0.0078, "step": 44710 }, { "grad_norm": 0.11479440331459045, "learning_rate": 6.272414555022636e-05, "loss": 0.0096, "step": 44720 }, { "grad_norm": 0.16043506562709808, "learning_rate": 6.270815452398215e-05, "loss": 0.0096, "step": 44730 }, { "grad_norm": 0.16717906296253204, "learning_rate": 6.269216210799326e-05, "loss": 0.0082, "step": 44740 }, { "grad_norm": 0.12877298891544342, "learning_rate": 6.26761683040086e-05, "loss": 0.0071, "step": 44750 }, { "grad_norm": 0.14093102514743805, "learning_rate": 6.266017311377723e-05, "loss": 0.0075, "step": 44760 }, { "grad_norm": 0.12279339879751205, "learning_rate": 6.264417653904839e-05, "loss": 0.0075, "step": 44770 }, { "grad_norm": 0.12535664439201355, "learning_rate": 6.262817858157139e-05, "loss": 0.0085, "step": 44780 }, { "grad_norm": 0.12608979642391205, "learning_rate": 6.261217924309576e-05, "loss": 0.0088, "step": 44790 }, { "grad_norm": 0.12339222431182861, "learning_rate": 6.259617852537118e-05, "loss": 0.0076, "step": 44800 }, { "grad_norm": 0.14016763865947723, "learning_rate": 6.258017643014747e-05, "loss": 0.0101, "step": 44810 }, { "grad_norm": 0.11443784087896347, "learning_rate": 6.256417295917456e-05, "loss": 0.0095, "step": 44820 }, { "grad_norm": 0.1484367847442627, "learning_rate": 6.254816811420258e-05, "loss": 0.0095, "step": 44830 }, { "grad_norm": 0.16709816455841064, "learning_rate": 6.253216189698183e-05, "loss": 0.0094, "step": 44840 }, { "grad_norm": 0.13510790467262268, "learning_rate": 6.251615430926267e-05, "loss": 0.0091, "step": 44850 }, { "grad_norm": 0.10556292533874512, "learning_rate": 6.25001453527957e-05, "loss": 0.0077, "step": 44860 }, { "grad_norm": 0.1253904551267624, "learning_rate": 6.248413502933164e-05, "loss": 0.0078, "step": 44870 }, { "grad_norm": 0.12601400911808014, "learning_rate": 6.246812334062133e-05, "loss": 0.0088, "step": 44880 }, { "grad_norm": 0.13366277515888214, "learning_rate": 6.245211028841579e-05, "loss": 0.0086, "step": 44890 }, { "grad_norm": 0.12233315408229828, "learning_rate": 6.24360958744662e-05, "loss": 0.0106, "step": 44900 }, { "grad_norm": 0.09890329837799072, "learning_rate": 6.242008010052387e-05, "loss": 0.0089, "step": 44910 }, { "grad_norm": 0.08720498532056808, "learning_rate": 6.240406296834024e-05, "loss": 0.01, "step": 44920 }, { "grad_norm": 0.13776350021362305, "learning_rate": 6.238804447966694e-05, "loss": 0.0093, "step": 44930 }, { "grad_norm": 0.15092068910598755, "learning_rate": 6.237202463625573e-05, "loss": 0.0094, "step": 44940 }, { "grad_norm": 0.16053512692451477, "learning_rate": 6.235600343985848e-05, "loss": 0.009, "step": 44950 }, { "grad_norm": 0.11308183521032333, "learning_rate": 6.233998089222729e-05, "loss": 0.008, "step": 44960 }, { "grad_norm": 0.11455865204334259, "learning_rate": 6.232395699511433e-05, "loss": 0.0074, "step": 44970 }, { "grad_norm": 0.0851215124130249, "learning_rate": 6.230793175027199e-05, "loss": 0.0072, "step": 44980 }, { "grad_norm": 0.13108578324317932, "learning_rate": 6.22919051594527e-05, "loss": 0.0094, "step": 44990 }, { "grad_norm": 0.14002008736133575, "learning_rate": 6.227587722440917e-05, "loss": 0.0092, "step": 45000 }, { "grad_norm": 0.13148006796836853, "learning_rate": 6.225984794689414e-05, "loss": 0.0102, "step": 45010 }, { "grad_norm": 0.11260627210140228, "learning_rate": 6.22438173286606e-05, "loss": 0.0104, "step": 45020 }, { "grad_norm": 0.09901494532823563, "learning_rate": 6.222778537146159e-05, "loss": 0.0087, "step": 45030 }, { "grad_norm": 0.11413498222827911, "learning_rate": 6.221175207705037e-05, "loss": 0.0115, "step": 45040 }, { "grad_norm": 0.12777076661586761, "learning_rate": 6.219571744718027e-05, "loss": 0.0087, "step": 45050 }, { "grad_norm": 0.08981657028198242, "learning_rate": 6.217968148360487e-05, "loss": 0.0078, "step": 45060 }, { "grad_norm": 0.13041207194328308, "learning_rate": 6.216364418807782e-05, "loss": 0.0091, "step": 45070 }, { "grad_norm": 0.08542083203792572, "learning_rate": 6.214760556235292e-05, "loss": 0.0075, "step": 45080 }, { "grad_norm": 0.13764925301074982, "learning_rate": 6.213156560818414e-05, "loss": 0.0092, "step": 45090 }, { "grad_norm": 0.14575545489788055, "learning_rate": 6.211552432732559e-05, "loss": 0.0088, "step": 45100 }, { "grad_norm": 0.12199574708938599, "learning_rate": 6.20994817215315e-05, "loss": 0.0087, "step": 45110 }, { "grad_norm": 0.12559950351715088, "learning_rate": 6.208343779255628e-05, "loss": 0.01, "step": 45120 }, { "grad_norm": 0.10713984817266464, "learning_rate": 6.206739254215449e-05, "loss": 0.0089, "step": 45130 }, { "grad_norm": 0.13470050692558289, "learning_rate": 6.205134597208077e-05, "loss": 0.0094, "step": 45140 }, { "grad_norm": 0.1067788228392601, "learning_rate": 6.203529808409e-05, "loss": 0.0081, "step": 45150 }, { "grad_norm": 0.10010344535112381, "learning_rate": 6.201924887993712e-05, "loss": 0.0078, "step": 45160 }, { "grad_norm": 0.11748166382312775, "learning_rate": 6.200319836137725e-05, "loss": 0.0075, "step": 45170 }, { "grad_norm": 0.08742474019527435, "learning_rate": 6.198714653016565e-05, "loss": 0.0078, "step": 45180 }, { "grad_norm": 0.09744371473789215, "learning_rate": 6.197109338805774e-05, "loss": 0.0077, "step": 45190 }, { "grad_norm": 0.14650902152061462, "learning_rate": 6.195503893680903e-05, "loss": 0.0089, "step": 45200 }, { "grad_norm": 0.11765716969966888, "learning_rate": 6.193898317817524e-05, "loss": 0.0089, "step": 45210 }, { "grad_norm": 0.13581104576587677, "learning_rate": 6.192292611391221e-05, "loss": 0.008, "step": 45220 }, { "grad_norm": 0.16713786125183105, "learning_rate": 6.190686774577591e-05, "loss": 0.0074, "step": 45230 }, { "grad_norm": 0.09814456105232239, "learning_rate": 6.189080807552245e-05, "loss": 0.008, "step": 45240 }, { "grad_norm": 0.09880416095256805, "learning_rate": 6.187474710490809e-05, "loss": 0.0091, "step": 45250 }, { "grad_norm": 0.11076050996780396, "learning_rate": 6.185868483568926e-05, "loss": 0.0092, "step": 45260 }, { "grad_norm": 0.14614374935626984, "learning_rate": 6.184262126962245e-05, "loss": 0.0098, "step": 45270 }, { "grad_norm": 0.10915612429380417, "learning_rate": 6.182655640846442e-05, "loss": 0.0083, "step": 45280 }, { "grad_norm": 0.13278698921203613, "learning_rate": 6.181049025397196e-05, "loss": 0.009, "step": 45290 }, { "grad_norm": 0.15251553058624268, "learning_rate": 6.179442280790202e-05, "loss": 0.0094, "step": 45300 }, { "grad_norm": 0.1389922946691513, "learning_rate": 6.177835407201174e-05, "loss": 0.0115, "step": 45310 }, { "grad_norm": 0.11823350191116333, "learning_rate": 6.176228404805839e-05, "loss": 0.0085, "step": 45320 }, { "grad_norm": 0.14622902870178223, "learning_rate": 6.174621273779932e-05, "loss": 0.0078, "step": 45330 }, { "grad_norm": 0.17178963124752045, "learning_rate": 6.17301401429921e-05, "loss": 0.0086, "step": 45340 }, { "grad_norm": 0.11282476782798767, "learning_rate": 6.171406626539436e-05, "loss": 0.0097, "step": 45350 }, { "grad_norm": 0.1400856375694275, "learning_rate": 6.169799110676398e-05, "loss": 0.0073, "step": 45360 }, { "grad_norm": 0.11701991409063339, "learning_rate": 6.168191466885885e-05, "loss": 0.0091, "step": 45370 }, { "grad_norm": 0.10705975443124771, "learning_rate": 6.166583695343711e-05, "loss": 0.0073, "step": 45380 }, { "grad_norm": 0.10819289833307266, "learning_rate": 6.164975796225698e-05, "loss": 0.0078, "step": 45390 }, { "grad_norm": 0.13893060386180878, "learning_rate": 6.163367769707683e-05, "loss": 0.0079, "step": 45400 }, { "grad_norm": 0.12047088891267776, "learning_rate": 6.161759615965519e-05, "loss": 0.0083, "step": 45410 }, { "grad_norm": 0.09701039642095566, "learning_rate": 6.16015133517507e-05, "loss": 0.0081, "step": 45420 }, { "grad_norm": 0.11958597600460052, "learning_rate": 6.158542927512214e-05, "loss": 0.0084, "step": 45430 }, { "grad_norm": 0.09586010128259659, "learning_rate": 6.156934393152846e-05, "loss": 0.0081, "step": 45440 }, { "grad_norm": 0.16031301021575928, "learning_rate": 6.15532573227287e-05, "loss": 0.0094, "step": 45450 }, { "grad_norm": 0.11077429354190826, "learning_rate": 6.153716945048212e-05, "loss": 0.013, "step": 45460 }, { "grad_norm": 0.13623389601707458, "learning_rate": 6.152108031654802e-05, "loss": 0.0098, "step": 45470 }, { "grad_norm": 0.11187588423490524, "learning_rate": 6.15049899226859e-05, "loss": 0.0101, "step": 45480 }, { "grad_norm": 0.08961106091737747, "learning_rate": 6.148889827065537e-05, "loss": 0.0093, "step": 45490 }, { "grad_norm": 0.14782926440238953, "learning_rate": 6.147280536221622e-05, "loss": 0.009, "step": 45500 }, { "grad_norm": 0.11522569507360458, "learning_rate": 6.145671119912832e-05, "loss": 0.0093, "step": 45510 }, { "grad_norm": 0.13496176898479462, "learning_rate": 6.144061578315169e-05, "loss": 0.0116, "step": 45520 }, { "grad_norm": 0.15451820194721222, "learning_rate": 6.142451911604654e-05, "loss": 0.0089, "step": 45530 }, { "grad_norm": 0.1189359799027443, "learning_rate": 6.140842119957315e-05, "loss": 0.009, "step": 45540 }, { "grad_norm": 0.1381499320268631, "learning_rate": 6.139232203549197e-05, "loss": 0.0109, "step": 45550 }, { "grad_norm": 0.13785439729690552, "learning_rate": 6.137622162556357e-05, "loss": 0.0083, "step": 45560 }, { "grad_norm": 0.09443533420562744, "learning_rate": 6.13601199715487e-05, "loss": 0.0082, "step": 45570 }, { "grad_norm": 0.08966489136219025, "learning_rate": 6.134401707520816e-05, "loss": 0.007, "step": 45580 }, { "grad_norm": 0.12197697162628174, "learning_rate": 6.132791293830298e-05, "loss": 0.0082, "step": 45590 }, { "grad_norm": 0.17245621979236603, "learning_rate": 6.131180756259428e-05, "loss": 0.0077, "step": 45600 }, { "grad_norm": 0.1285085529088974, "learning_rate": 6.129570094984331e-05, "loss": 0.0081, "step": 45610 }, { "grad_norm": 0.14452339708805084, "learning_rate": 6.127959310181145e-05, "loss": 0.0068, "step": 45620 }, { "grad_norm": 0.15087354183197021, "learning_rate": 6.126348402026026e-05, "loss": 0.0086, "step": 45630 }, { "grad_norm": 0.11578407883644104, "learning_rate": 6.12473737069514e-05, "loss": 0.0109, "step": 45640 }, { "grad_norm": 0.12465910613536835, "learning_rate": 6.123126216364665e-05, "loss": 0.0071, "step": 45650 }, { "grad_norm": 0.12496256083250046, "learning_rate": 6.121514939210797e-05, "loss": 0.0084, "step": 45660 }, { "grad_norm": 0.09888666868209839, "learning_rate": 6.119903539409741e-05, "loss": 0.0073, "step": 45670 }, { "grad_norm": 0.13877862691879272, "learning_rate": 6.118292017137716e-05, "loss": 0.0082, "step": 45680 }, { "grad_norm": 0.09043490141630173, "learning_rate": 6.116680372570959e-05, "loss": 0.0073, "step": 45690 }, { "grad_norm": 0.14398375153541565, "learning_rate": 6.115068605885713e-05, "loss": 0.008, "step": 45700 }, { "grad_norm": 0.20775413513183594, "learning_rate": 6.113456717258243e-05, "loss": 0.0109, "step": 45710 }, { "grad_norm": 0.10546897351741791, "learning_rate": 6.11184470686482e-05, "loss": 0.0075, "step": 45720 }, { "grad_norm": 0.1680457442998886, "learning_rate": 6.110232574881731e-05, "loss": 0.0073, "step": 45730 }, { "grad_norm": 0.11049354821443558, "learning_rate": 6.108620321485277e-05, "loss": 0.0079, "step": 45740 }, { "grad_norm": 0.15349464118480682, "learning_rate": 6.107007946851773e-05, "loss": 0.0078, "step": 45750 }, { "grad_norm": 0.13541342318058014, "learning_rate": 6.105395451157542e-05, "loss": 0.0115, "step": 45760 }, { "grad_norm": 0.10838962346315384, "learning_rate": 6.103782834578928e-05, "loss": 0.0075, "step": 45770 }, { "grad_norm": 0.10248343646526337, "learning_rate": 6.102170097292281e-05, "loss": 0.0088, "step": 45780 }, { "grad_norm": 0.1335373818874359, "learning_rate": 6.1005572394739716e-05, "loss": 0.008, "step": 45790 }, { "grad_norm": 0.11422990262508392, "learning_rate": 6.0989442613003765e-05, "loss": 0.0084, "step": 45800 }, { "grad_norm": 0.10285012423992157, "learning_rate": 6.097331162947888e-05, "loss": 0.0096, "step": 45810 }, { "grad_norm": 0.12256220728158951, "learning_rate": 6.095717944592914e-05, "loss": 0.009, "step": 45820 }, { "grad_norm": 0.10431566089391708, "learning_rate": 6.094104606411873e-05, "loss": 0.0081, "step": 45830 }, { "grad_norm": 0.13476738333702087, "learning_rate": 6.0924911485811966e-05, "loss": 0.0093, "step": 45840 }, { "grad_norm": 0.15271487832069397, "learning_rate": 6.090877571277331e-05, "loss": 0.0093, "step": 45850 }, { "grad_norm": 0.12324309349060059, "learning_rate": 6.089263874676736e-05, "loss": 0.0075, "step": 45860 }, { "grad_norm": 0.13048051297664642, "learning_rate": 6.0876500589558796e-05, "loss": 0.0087, "step": 45870 }, { "grad_norm": 0.11936928331851959, "learning_rate": 6.086036124291248e-05, "loss": 0.009, "step": 45880 }, { "grad_norm": 0.11128173023462296, "learning_rate": 6.084422070859339e-05, "loss": 0.0079, "step": 45890 }, { "grad_norm": 0.16105128824710846, "learning_rate": 6.082807898836663e-05, "loss": 0.0102, "step": 45900 }, { "grad_norm": 0.14621156454086304, "learning_rate": 6.081193608399742e-05, "loss": 0.0089, "step": 45910 }, { "grad_norm": 0.10724072903394699, "learning_rate": 6.0795791997251164e-05, "loss": 0.0084, "step": 45920 }, { "grad_norm": 0.13607896864414215, "learning_rate": 6.0779646729893294e-05, "loss": 0.008, "step": 45930 }, { "grad_norm": 0.11050952970981598, "learning_rate": 6.0763500283689476e-05, "loss": 0.0094, "step": 45940 }, { "grad_norm": 0.09290427714586258, "learning_rate": 6.0747352660405455e-05, "loss": 0.0078, "step": 45950 }, { "grad_norm": 0.10210949182510376, "learning_rate": 6.073120386180709e-05, "loss": 0.0082, "step": 45960 }, { "grad_norm": 0.15061351656913757, "learning_rate": 6.0715053889660425e-05, "loss": 0.0081, "step": 45970 }, { "grad_norm": 0.13863517343997955, "learning_rate": 6.069890274573157e-05, "loss": 0.0108, "step": 45980 }, { "grad_norm": 0.12028411030769348, "learning_rate": 6.068275043178679e-05, "loss": 0.0079, "step": 45990 }, { "grad_norm": 0.10820169746875763, "learning_rate": 6.066659694959248e-05, "loss": 0.0094, "step": 46000 }, { "grad_norm": 0.1269049048423767, "learning_rate": 6.0650442300915176e-05, "loss": 0.0082, "step": 46010 }, { "grad_norm": 0.14853963255882263, "learning_rate": 6.063428648752152e-05, "loss": 0.009, "step": 46020 }, { "grad_norm": 0.11377892643213272, "learning_rate": 6.0618129511178266e-05, "loss": 0.0077, "step": 46030 }, { "grad_norm": 0.2044520080089569, "learning_rate": 6.060197137365234e-05, "loss": 0.01, "step": 46040 }, { "grad_norm": 0.145135298371315, "learning_rate": 6.058581207671077e-05, "loss": 0.0097, "step": 46050 }, { "grad_norm": 0.11787262558937073, "learning_rate": 6.056965162212072e-05, "loss": 0.0077, "step": 46060 }, { "grad_norm": 0.12575022876262665, "learning_rate": 6.055349001164943e-05, "loss": 0.0074, "step": 46070 }, { "grad_norm": 0.13711273670196533, "learning_rate": 6.0537327247064347e-05, "loss": 0.0077, "step": 46080 }, { "grad_norm": 0.12322324514389038, "learning_rate": 6.0521163330133014e-05, "loss": 0.0073, "step": 46090 }, { "grad_norm": 0.1329810470342636, "learning_rate": 6.050499826262306e-05, "loss": 0.007, "step": 46100 }, { "grad_norm": 0.13620413839817047, "learning_rate": 6.0488832046302294e-05, "loss": 0.0096, "step": 46110 }, { "grad_norm": 0.12335189431905746, "learning_rate": 6.0472664682938626e-05, "loss": 0.0081, "step": 46120 }, { "grad_norm": 0.12183085083961487, "learning_rate": 6.045649617430009e-05, "loss": 0.0071, "step": 46130 }, { "grad_norm": 0.1472957283258438, "learning_rate": 6.0440326522154866e-05, "loss": 0.0076, "step": 46140 }, { "grad_norm": 0.09927580505609512, "learning_rate": 6.0424155728271224e-05, "loss": 0.0086, "step": 46150 }, { "grad_norm": 0.1285392791032791, "learning_rate": 6.040798379441758e-05, "loss": 0.0093, "step": 46160 }, { "grad_norm": 0.14245586097240448, "learning_rate": 6.0391810722362485e-05, "loss": 0.01, "step": 46170 }, { "grad_norm": 0.15386179089546204, "learning_rate": 6.037563651387458e-05, "loss": 0.0084, "step": 46180 }, { "grad_norm": 0.08299406617879868, "learning_rate": 6.0359461170722666e-05, "loss": 0.0071, "step": 46190 }, { "grad_norm": 0.13130472600460052, "learning_rate": 6.034328469467566e-05, "loss": 0.0077, "step": 46200 }, { "grad_norm": 0.13369712233543396, "learning_rate": 6.0327107087502596e-05, "loss": 0.0074, "step": 46210 }, { "grad_norm": 0.18558214604854584, "learning_rate": 6.031092835097262e-05, "loss": 0.0081, "step": 46220 }, { "grad_norm": 0.11689114570617676, "learning_rate": 6.0294748486855024e-05, "loss": 0.0088, "step": 46230 }, { "grad_norm": 0.157941997051239, "learning_rate": 6.0278567496919216e-05, "loss": 0.009, "step": 46240 }, { "grad_norm": 0.13927961885929108, "learning_rate": 6.026238538293472e-05, "loss": 0.0066, "step": 46250 }, { "grad_norm": 0.12983077764511108, "learning_rate": 6.024620214667118e-05, "loss": 0.0074, "step": 46260 }, { "grad_norm": 0.11298079043626785, "learning_rate": 6.0230017789898384e-05, "loss": 0.0067, "step": 46270 }, { "grad_norm": 0.1330292969942093, "learning_rate": 6.021383231438622e-05, "loss": 0.0067, "step": 46280 }, { "grad_norm": 0.09170514345169067, "learning_rate": 6.0197645721904704e-05, "loss": 0.007, "step": 46290 }, { "grad_norm": 0.11861136555671692, "learning_rate": 6.0181458014224e-05, "loss": 0.0116, "step": 46300 }, { "grad_norm": 0.12237086892127991, "learning_rate": 6.016526919311435e-05, "loss": 0.0068, "step": 46310 }, { "grad_norm": 0.10376919060945511, "learning_rate": 6.014907926034613e-05, "loss": 0.0082, "step": 46320 }, { "grad_norm": 0.09472054243087769, "learning_rate": 6.0132888217689875e-05, "loss": 0.0077, "step": 46330 }, { "grad_norm": 0.15828967094421387, "learning_rate": 6.01166960669162e-05, "loss": 0.0086, "step": 46340 }, { "grad_norm": 0.11710699647665024, "learning_rate": 6.0100502809795845e-05, "loss": 0.0065, "step": 46350 }, { "grad_norm": 0.1438378393650055, "learning_rate": 6.00843084480997e-05, "loss": 0.0085, "step": 46360 }, { "grad_norm": 0.14868266880512238, "learning_rate": 6.0068112983598736e-05, "loss": 0.008, "step": 46370 }, { "grad_norm": 0.11606743186712265, "learning_rate": 6.005191641806407e-05, "loss": 0.0076, "step": 46380 }, { "grad_norm": 0.10644466429948807, "learning_rate": 6.003571875326694e-05, "loss": 0.008, "step": 46390 }, { "grad_norm": 0.1263505220413208, "learning_rate": 6.00195199909787e-05, "loss": 0.0082, "step": 46400 }, { "grad_norm": 0.14682820439338684, "learning_rate": 6.000332013297082e-05, "loss": 0.0078, "step": 46410 }, { "grad_norm": 0.11364655196666718, "learning_rate": 5.998711918101487e-05, "loss": 0.0095, "step": 46420 }, { "grad_norm": 0.14251765608787537, "learning_rate": 5.997091713688261e-05, "loss": 0.0094, "step": 46430 }, { "grad_norm": 0.13694168627262115, "learning_rate": 5.9954714002345836e-05, "loss": 0.008, "step": 46440 }, { "grad_norm": 0.13084624707698822, "learning_rate": 5.993850977917649e-05, "loss": 0.0079, "step": 46450 }, { "grad_norm": 0.12731221318244934, "learning_rate": 5.992230446914667e-05, "loss": 0.0076, "step": 46460 }, { "grad_norm": 0.1241479441523552, "learning_rate": 5.990609807402855e-05, "loss": 0.0117, "step": 46470 }, { "grad_norm": 0.16349342465400696, "learning_rate": 5.988989059559443e-05, "loss": 0.0113, "step": 46480 }, { "grad_norm": 0.1318705677986145, "learning_rate": 5.987368203561675e-05, "loss": 0.0075, "step": 46490 }, { "grad_norm": 0.13976499438285828, "learning_rate": 5.9857472395868055e-05, "loss": 0.0083, "step": 46500 }, { "grad_norm": 0.227463498711586, "learning_rate": 5.9841261678120983e-05, "loss": 0.0074, "step": 46510 }, { "grad_norm": 0.14108900725841522, "learning_rate": 5.982504988414834e-05, "loss": 0.0084, "step": 46520 }, { "grad_norm": 0.12515589594841003, "learning_rate": 5.9808837015723015e-05, "loss": 0.0085, "step": 46530 }, { "grad_norm": 0.10862135887145996, "learning_rate": 5.9792623074618016e-05, "loss": 0.0108, "step": 46540 }, { "grad_norm": 0.14223167300224304, "learning_rate": 5.977640806260648e-05, "loss": 0.0091, "step": 46550 }, { "grad_norm": 0.15082304179668427, "learning_rate": 5.976019198146166e-05, "loss": 0.0091, "step": 46560 }, { "grad_norm": 0.16274186968803406, "learning_rate": 5.974397483295692e-05, "loss": 0.0078, "step": 46570 }, { "grad_norm": 0.10864441096782684, "learning_rate": 5.972775661886575e-05, "loss": 0.0083, "step": 46580 }, { "grad_norm": 0.13657885789871216, "learning_rate": 5.971153734096173e-05, "loss": 0.0097, "step": 46590 }, { "grad_norm": 0.11490820348262787, "learning_rate": 5.969531700101859e-05, "loss": 0.0113, "step": 46600 }, { "grad_norm": 0.18682490289211273, "learning_rate": 5.9679095600810155e-05, "loss": 0.0082, "step": 46610 }, { "grad_norm": 0.16993360221385956, "learning_rate": 5.9662873142110384e-05, "loss": 0.0098, "step": 46620 }, { "grad_norm": 0.18293198943138123, "learning_rate": 5.964664962669333e-05, "loss": 0.0084, "step": 46630 }, { "grad_norm": 0.12231598049402237, "learning_rate": 5.9630425056333186e-05, "loss": 0.0091, "step": 46640 }, { "grad_norm": 0.13689924776554108, "learning_rate": 5.961419943280422e-05, "loss": 0.0089, "step": 46650 }, { "grad_norm": 0.09372460842132568, "learning_rate": 5.959797275788087e-05, "loss": 0.0078, "step": 46660 }, { "grad_norm": 0.11764165759086609, "learning_rate": 5.958174503333765e-05, "loss": 0.0074, "step": 46670 }, { "grad_norm": 0.09382785856723785, "learning_rate": 5.9565516260949195e-05, "loss": 0.0092, "step": 46680 }, { "grad_norm": 0.1261090338230133, "learning_rate": 5.9549286442490273e-05, "loss": 0.0073, "step": 46690 }, { "grad_norm": 0.09567022323608398, "learning_rate": 5.953305557973572e-05, "loss": 0.0073, "step": 46700 }, { "grad_norm": 0.10476703196763992, "learning_rate": 5.9516823674460564e-05, "loss": 0.0068, "step": 46710 }, { "grad_norm": 0.14723971486091614, "learning_rate": 5.9500590728439875e-05, "loss": 0.0088, "step": 46720 }, { "grad_norm": 0.12937873601913452, "learning_rate": 5.948435674344886e-05, "loss": 0.0096, "step": 46730 }, { "grad_norm": 0.32091525197029114, "learning_rate": 5.946812172126285e-05, "loss": 0.0069, "step": 46740 }, { "grad_norm": 0.25321394205093384, "learning_rate": 5.945188566365729e-05, "loss": 0.0099, "step": 46750 }, { "grad_norm": 0.10012616217136383, "learning_rate": 5.943564857240773e-05, "loss": 0.0082, "step": 46760 }, { "grad_norm": 0.14960990846157074, "learning_rate": 5.941941044928981e-05, "loss": 0.0071, "step": 46770 }, { "grad_norm": 0.1301879584789276, "learning_rate": 5.940317129607935e-05, "loss": 0.011, "step": 46780 }, { "grad_norm": 0.08247994631528854, "learning_rate": 5.9386931114552204e-05, "loss": 0.0072, "step": 46790 }, { "grad_norm": 0.1745523363351822, "learning_rate": 5.937068990648438e-05, "loss": 0.009, "step": 46800 }, { "grad_norm": 0.10098381340503693, "learning_rate": 5.935444767365199e-05, "loss": 0.0099, "step": 46810 }, { "grad_norm": 0.12101773917675018, "learning_rate": 5.933820441783129e-05, "loss": 0.0091, "step": 46820 }, { "grad_norm": 0.11769971251487732, "learning_rate": 5.932196014079857e-05, "loss": 0.0069, "step": 46830 }, { "grad_norm": 0.12164237350225449, "learning_rate": 5.930571484433032e-05, "loss": 0.0068, "step": 46840 }, { "grad_norm": 0.13351117074489594, "learning_rate": 5.928946853020309e-05, "loss": 0.0079, "step": 46850 }, { "grad_norm": 0.08248280733823776, "learning_rate": 5.927322120019352e-05, "loss": 0.0079, "step": 46860 }, { "grad_norm": 0.12256035953760147, "learning_rate": 5.925697285607844e-05, "loss": 0.0073, "step": 46870 }, { "grad_norm": 0.13431870937347412, "learning_rate": 5.924072349963472e-05, "loss": 0.0087, "step": 46880 }, { "grad_norm": 0.11086253076791763, "learning_rate": 5.922447313263938e-05, "loss": 0.0073, "step": 46890 }, { "grad_norm": 0.11245767772197723, "learning_rate": 5.920822175686952e-05, "loss": 0.0104, "step": 46900 }, { "grad_norm": 0.12692010402679443, "learning_rate": 5.919196937410239e-05, "loss": 0.0087, "step": 46910 }, { "grad_norm": 0.11798861622810364, "learning_rate": 5.917571598611529e-05, "loss": 0.0094, "step": 46920 }, { "grad_norm": 0.13382479548454285, "learning_rate": 5.9159461594685686e-05, "loss": 0.009, "step": 46930 }, { "grad_norm": 0.12914063036441803, "learning_rate": 5.914320620159114e-05, "loss": 0.0089, "step": 46940 }, { "grad_norm": 0.1383446902036667, "learning_rate": 5.912694980860931e-05, "loss": 0.0093, "step": 46950 }, { "grad_norm": 0.09743987023830414, "learning_rate": 5.9110692417517964e-05, "loss": 0.0065, "step": 46960 }, { "grad_norm": 0.11350000649690628, "learning_rate": 5.9094434030095e-05, "loss": 0.0083, "step": 46970 }, { "grad_norm": 0.1060764417052269, "learning_rate": 5.907817464811841e-05, "loss": 0.0074, "step": 46980 }, { "grad_norm": 0.19596917927265167, "learning_rate": 5.906191427336627e-05, "loss": 0.0094, "step": 46990 }, { "grad_norm": 0.14060081541538239, "learning_rate": 5.9045652907616835e-05, "loss": 0.0084, "step": 47000 }, { "grad_norm": 0.14323817193508148, "learning_rate": 5.902939055264838e-05, "loss": 0.0082, "step": 47010 }, { "grad_norm": 0.17025625705718994, "learning_rate": 5.901312721023935e-05, "loss": 0.0086, "step": 47020 }, { "grad_norm": 0.13150982558727264, "learning_rate": 5.8996862882168294e-05, "loss": 0.0072, "step": 47030 }, { "grad_norm": 0.11626773327589035, "learning_rate": 5.8980597570213826e-05, "loss": 0.0075, "step": 47040 }, { "grad_norm": 0.11707132309675217, "learning_rate": 5.896433127615471e-05, "loss": 0.0072, "step": 47050 }, { "grad_norm": 0.11654925346374512, "learning_rate": 5.894806400176981e-05, "loss": 0.008, "step": 47060 }, { "grad_norm": 0.13566099107265472, "learning_rate": 5.893179574883808e-05, "loss": 0.0076, "step": 47070 }, { "grad_norm": 0.11215544492006302, "learning_rate": 5.8915526519138585e-05, "loss": 0.0078, "step": 47080 }, { "grad_norm": 0.12853766977787018, "learning_rate": 5.889925631445053e-05, "loss": 0.0075, "step": 47090 }, { "grad_norm": 0.11960022151470184, "learning_rate": 5.888298513655318e-05, "loss": 0.0071, "step": 47100 }, { "grad_norm": 0.1092238500714302, "learning_rate": 5.886671298722592e-05, "loss": 0.007, "step": 47110 }, { "grad_norm": 0.10305158793926239, "learning_rate": 5.885043986824828e-05, "loss": 0.0083, "step": 47120 }, { "grad_norm": 0.13844694197177887, "learning_rate": 5.883416578139982e-05, "loss": 0.008, "step": 47130 }, { "grad_norm": 0.12071342021226883, "learning_rate": 5.881789072846029e-05, "loss": 0.0091, "step": 47140 }, { "grad_norm": 0.1717335730791092, "learning_rate": 5.8801614711209474e-05, "loss": 0.01, "step": 47150 }, { "grad_norm": 0.0921245738863945, "learning_rate": 5.878533773142732e-05, "loss": 0.0075, "step": 47160 }, { "grad_norm": 0.09052607417106628, "learning_rate": 5.8769059790893834e-05, "loss": 0.0089, "step": 47170 }, { "grad_norm": 0.12564927339553833, "learning_rate": 5.8752780891389145e-05, "loss": 0.0099, "step": 47180 }, { "grad_norm": 0.1283184140920639, "learning_rate": 5.873650103469351e-05, "loss": 0.0078, "step": 47190 }, { "grad_norm": 0.12188898026943207, "learning_rate": 5.872022022258725e-05, "loss": 0.0082, "step": 47200 }, { "grad_norm": 0.126999631524086, "learning_rate": 5.8703938456850814e-05, "loss": 0.0084, "step": 47210 }, { "grad_norm": 0.12089104950428009, "learning_rate": 5.868765573926477e-05, "loss": 0.0073, "step": 47220 }, { "grad_norm": 0.17928028106689453, "learning_rate": 5.867137207160974e-05, "loss": 0.0081, "step": 47230 }, { "grad_norm": 0.14506419003009796, "learning_rate": 5.86550874556665e-05, "loss": 0.0077, "step": 47240 }, { "grad_norm": 0.1495184749364853, "learning_rate": 5.863880189321592e-05, "loss": 0.0087, "step": 47250 }, { "grad_norm": 0.11078839749097824, "learning_rate": 5.862251538603894e-05, "loss": 0.0083, "step": 47260 }, { "grad_norm": 0.17518483102321625, "learning_rate": 5.8606227935916656e-05, "loss": 0.01, "step": 47270 }, { "grad_norm": 0.14238899946212769, "learning_rate": 5.858993954463021e-05, "loss": 0.0075, "step": 47280 }, { "grad_norm": 0.11607160419225693, "learning_rate": 5.8573650213960896e-05, "loss": 0.0088, "step": 47290 }, { "grad_norm": 0.1130557432770729, "learning_rate": 5.8557359945690084e-05, "loss": 0.0075, "step": 47300 }, { "grad_norm": 0.10943333059549332, "learning_rate": 5.8541068741599246e-05, "loss": 0.0059, "step": 47310 }, { "grad_norm": 0.14684607088565826, "learning_rate": 5.8524776603469985e-05, "loss": 0.0078, "step": 47320 }, { "grad_norm": 0.1911349892616272, "learning_rate": 5.850848353308397e-05, "loss": 0.0083, "step": 47330 }, { "grad_norm": 0.09352846443653107, "learning_rate": 5.849218953222297e-05, "loss": 0.0082, "step": 47340 }, { "grad_norm": 0.12445368617773056, "learning_rate": 5.847589460266891e-05, "loss": 0.0077, "step": 47350 }, { "grad_norm": 0.09433910250663757, "learning_rate": 5.845959874620375e-05, "loss": 0.0077, "step": 47360 }, { "grad_norm": 0.13488440215587616, "learning_rate": 5.84433019646096e-05, "loss": 0.0085, "step": 47370 }, { "grad_norm": 0.12687541544437408, "learning_rate": 5.842700425966863e-05, "loss": 0.0087, "step": 47380 }, { "grad_norm": 0.16302061080932617, "learning_rate": 5.841070563316315e-05, "loss": 0.0093, "step": 47390 }, { "grad_norm": 0.12482970207929611, "learning_rate": 5.839440608687554e-05, "loss": 0.0086, "step": 47400 }, { "grad_norm": 0.15212690830230713, "learning_rate": 5.837810562258831e-05, "loss": 0.0081, "step": 47410 }, { "grad_norm": 0.10957866907119751, "learning_rate": 5.836180424208405e-05, "loss": 0.0072, "step": 47420 }, { "grad_norm": 0.10416482388973236, "learning_rate": 5.834550194714543e-05, "loss": 0.0096, "step": 47430 }, { "grad_norm": 0.10911073535680771, "learning_rate": 5.832919873955528e-05, "loss": 0.0098, "step": 47440 }, { "grad_norm": 0.11834420263767242, "learning_rate": 5.831289462109648e-05, "loss": 0.0076, "step": 47450 }, { "grad_norm": 0.1707545965909958, "learning_rate": 5.829658959355202e-05, "loss": 0.0078, "step": 47460 }, { "grad_norm": 0.1291811764240265, "learning_rate": 5.8280283658704995e-05, "loss": 0.0092, "step": 47470 }, { "grad_norm": 0.13651707768440247, "learning_rate": 5.826397681833859e-05, "loss": 0.0078, "step": 47480 }, { "grad_norm": 0.16080299019813538, "learning_rate": 5.824766907423612e-05, "loss": 0.0095, "step": 47490 }, { "grad_norm": 0.1641121357679367, "learning_rate": 5.8231360428180945e-05, "loss": 0.0075, "step": 47500 }, { "grad_norm": 0.13641594350337982, "learning_rate": 5.821505088195658e-05, "loss": 0.0097, "step": 47510 }, { "grad_norm": 0.1395675390958786, "learning_rate": 5.819874043734661e-05, "loss": 0.0086, "step": 47520 }, { "grad_norm": 0.1530526578426361, "learning_rate": 5.8182429096134695e-05, "loss": 0.0086, "step": 47530 }, { "grad_norm": 0.12969398498535156, "learning_rate": 5.816611686010465e-05, "loss": 0.0082, "step": 47540 }, { "grad_norm": 0.14393433928489685, "learning_rate": 5.814980373104033e-05, "loss": 0.0101, "step": 47550 }, { "grad_norm": 0.12521523237228394, "learning_rate": 5.813348971072572e-05, "loss": 0.0078, "step": 47560 }, { "grad_norm": 0.13059617578983307, "learning_rate": 5.811717480094492e-05, "loss": 0.0074, "step": 47570 }, { "grad_norm": 0.1400575339794159, "learning_rate": 5.810085900348209e-05, "loss": 0.0104, "step": 47580 }, { "grad_norm": 0.13080690801143646, "learning_rate": 5.8084542320121483e-05, "loss": 0.0087, "step": 47590 }, { "grad_norm": 0.12268677353858948, "learning_rate": 5.8068224752647497e-05, "loss": 0.0083, "step": 47600 }, { "grad_norm": 0.12951825559139252, "learning_rate": 5.805190630284457e-05, "loss": 0.0076, "step": 47610 }, { "grad_norm": 0.10567948967218399, "learning_rate": 5.803558697249729e-05, "loss": 0.0096, "step": 47620 }, { "grad_norm": 0.16430731117725372, "learning_rate": 5.8019266763390276e-05, "loss": 0.0099, "step": 47630 }, { "grad_norm": 0.17519839107990265, "learning_rate": 5.800294567730833e-05, "loss": 0.008, "step": 47640 }, { "grad_norm": 0.17556102573871613, "learning_rate": 5.798662371603628e-05, "loss": 0.0073, "step": 47650 }, { "grad_norm": 0.17139780521392822, "learning_rate": 5.797030088135904e-05, "loss": 0.0078, "step": 47660 }, { "grad_norm": 0.13571257889270782, "learning_rate": 5.7953977175061705e-05, "loss": 0.0082, "step": 47670 }, { "grad_norm": 0.11155650019645691, "learning_rate": 5.793765259892938e-05, "loss": 0.0077, "step": 47680 }, { "grad_norm": 0.24144338071346283, "learning_rate": 5.792132715474729e-05, "loss": 0.0075, "step": 47690 }, { "grad_norm": 0.13312353193759918, "learning_rate": 5.790500084430078e-05, "loss": 0.0072, "step": 47700 }, { "grad_norm": 0.08880084753036499, "learning_rate": 5.7888673669375274e-05, "loss": 0.0071, "step": 47710 }, { "grad_norm": 0.10432054102420807, "learning_rate": 5.787234563175625e-05, "loss": 0.0078, "step": 47720 }, { "grad_norm": 0.11302538961172104, "learning_rate": 5.7856016733229355e-05, "loss": 0.008, "step": 47730 }, { "grad_norm": 0.10833285748958588, "learning_rate": 5.7839686975580297e-05, "loss": 0.0082, "step": 47740 }, { "grad_norm": 0.11758387833833694, "learning_rate": 5.782335636059484e-05, "loss": 0.0073, "step": 47750 }, { "grad_norm": 0.10208304226398468, "learning_rate": 5.780702489005889e-05, "loss": 0.0072, "step": 47760 }, { "grad_norm": 0.12914590537548065, "learning_rate": 5.779069256575845e-05, "loss": 0.009, "step": 47770 }, { "grad_norm": 0.11540062725543976, "learning_rate": 5.7774359389479574e-05, "loss": 0.0083, "step": 47780 }, { "grad_norm": 0.11426030099391937, "learning_rate": 5.775802536300845e-05, "loss": 0.0083, "step": 47790 }, { "grad_norm": 0.17236724495887756, "learning_rate": 5.774169048813134e-05, "loss": 0.0077, "step": 47800 }, { "grad_norm": 0.10873443633317947, "learning_rate": 5.77253547666346e-05, "loss": 0.0079, "step": 47810 }, { "grad_norm": 0.1305207461118698, "learning_rate": 5.770901820030465e-05, "loss": 0.0084, "step": 47820 }, { "grad_norm": 0.1176699623465538, "learning_rate": 5.769268079092809e-05, "loss": 0.0077, "step": 47830 }, { "grad_norm": 0.1050996407866478, "learning_rate": 5.767634254029151e-05, "loss": 0.0084, "step": 47840 }, { "grad_norm": 0.08786119520664215, "learning_rate": 5.7660003450181655e-05, "loss": 0.0066, "step": 47850 }, { "grad_norm": 0.12121069431304932, "learning_rate": 5.764366352238534e-05, "loss": 0.0084, "step": 47860 }, { "grad_norm": 0.10942602157592773, "learning_rate": 5.7627322758689474e-05, "loss": 0.0084, "step": 47870 }, { "grad_norm": 0.12096872925758362, "learning_rate": 5.761098116088105e-05, "loss": 0.007, "step": 47880 }, { "grad_norm": 0.16852082312107086, "learning_rate": 5.759463873074717e-05, "loss": 0.009, "step": 47890 }, { "grad_norm": 0.11052926629781723, "learning_rate": 5.757829547007504e-05, "loss": 0.0098, "step": 47900 }, { "grad_norm": 0.16046380996704102, "learning_rate": 5.756195138065189e-05, "loss": 0.0087, "step": 47910 }, { "grad_norm": 0.17832641303539276, "learning_rate": 5.754560646426511e-05, "loss": 0.0089, "step": 47920 }, { "grad_norm": 0.11910083889961243, "learning_rate": 5.752926072270216e-05, "loss": 0.0082, "step": 47930 }, { "grad_norm": 0.11938546597957611, "learning_rate": 5.7512914157750563e-05, "loss": 0.0099, "step": 47940 }, { "grad_norm": 0.11094260960817337, "learning_rate": 5.749656677119798e-05, "loss": 0.0099, "step": 47950 }, { "grad_norm": 0.13617685437202454, "learning_rate": 5.7480218564832125e-05, "loss": 0.008, "step": 47960 }, { "grad_norm": 0.16361649334430695, "learning_rate": 5.746386954044082e-05, "loss": 0.0099, "step": 47970 }, { "grad_norm": 0.11777731776237488, "learning_rate": 5.744751969981195e-05, "loss": 0.0076, "step": 47980 }, { "grad_norm": 0.11340173333883286, "learning_rate": 5.7431169044733526e-05, "loss": 0.0076, "step": 47990 }, { "grad_norm": 0.15709079802036285, "learning_rate": 5.741481757699364e-05, "loss": 0.0089, "step": 48000 }, { "grad_norm": 0.10937074571847916, "learning_rate": 5.7398465298380434e-05, "loss": 0.0077, "step": 48010 }, { "grad_norm": 0.12106423079967499, "learning_rate": 5.7382112210682193e-05, "loss": 0.0092, "step": 48020 }, { "grad_norm": 0.24641618132591248, "learning_rate": 5.7365758315687266e-05, "loss": 0.0071, "step": 48030 }, { "grad_norm": 0.15955936908721924, "learning_rate": 5.734940361518407e-05, "loss": 0.0093, "step": 48040 }, { "grad_norm": 0.11478165537118912, "learning_rate": 5.733304811096116e-05, "loss": 0.0062, "step": 48050 }, { "grad_norm": 0.14484958350658417, "learning_rate": 5.731669180480713e-05, "loss": 0.0101, "step": 48060 }, { "grad_norm": 0.10300091654062271, "learning_rate": 5.730033469851067e-05, "loss": 0.0079, "step": 48070 }, { "grad_norm": 0.09870713204145432, "learning_rate": 5.7283976793860607e-05, "loss": 0.0087, "step": 48080 }, { "grad_norm": 0.11033929884433746, "learning_rate": 5.726761809264577e-05, "loss": 0.0085, "step": 48090 }, { "grad_norm": 0.11181586235761642, "learning_rate": 5.7251258596655155e-05, "loss": 0.0089, "step": 48100 }, { "grad_norm": 0.12524999678134918, "learning_rate": 5.72348983076778e-05, "loss": 0.0102, "step": 48110 }, { "grad_norm": 0.14387156069278717, "learning_rate": 5.7218537227502854e-05, "loss": 0.0098, "step": 48120 }, { "grad_norm": 0.17268067598342896, "learning_rate": 5.720217535791951e-05, "loss": 0.0072, "step": 48130 }, { "grad_norm": 0.17767633497714996, "learning_rate": 5.718581270071711e-05, "loss": 0.0102, "step": 48140 }, { "grad_norm": 0.12940339744091034, "learning_rate": 5.716944925768505e-05, "loss": 0.0098, "step": 48150 }, { "grad_norm": 0.08657202124595642, "learning_rate": 5.7153085030612786e-05, "loss": 0.0074, "step": 48160 }, { "grad_norm": 0.0859854444861412, "learning_rate": 5.713672002128989e-05, "loss": 0.0081, "step": 48170 }, { "grad_norm": 0.11686799675226212, "learning_rate": 5.712035423150602e-05, "loss": 0.0098, "step": 48180 }, { "grad_norm": 0.10869535803794861, "learning_rate": 5.710398766305094e-05, "loss": 0.0084, "step": 48190 }, { "grad_norm": 0.163962259888649, "learning_rate": 5.708762031771442e-05, "loss": 0.0065, "step": 48200 }, { "grad_norm": 0.12245748937129974, "learning_rate": 5.707125219728642e-05, "loss": 0.0092, "step": 48210 }, { "grad_norm": 0.12413489073514938, "learning_rate": 5.7054883303556905e-05, "loss": 0.008, "step": 48220 }, { "grad_norm": 0.1124139055609703, "learning_rate": 5.703851363831595e-05, "loss": 0.0085, "step": 48230 }, { "grad_norm": 0.12135805189609528, "learning_rate": 5.702214320335374e-05, "loss": 0.0081, "step": 48240 }, { "grad_norm": 0.15013788640499115, "learning_rate": 5.7005772000460514e-05, "loss": 0.008, "step": 48250 }, { "grad_norm": 0.1363166868686676, "learning_rate": 5.698940003142656e-05, "loss": 0.0084, "step": 48260 }, { "grad_norm": 0.11650702357292175, "learning_rate": 5.697302729804236e-05, "loss": 0.0088, "step": 48270 }, { "grad_norm": 0.10564930737018585, "learning_rate": 5.695665380209837e-05, "loss": 0.0072, "step": 48280 }, { "grad_norm": 0.10843051970005035, "learning_rate": 5.6940279545385165e-05, "loss": 0.01, "step": 48290 }, { "grad_norm": 0.09002815186977386, "learning_rate": 5.692390452969344e-05, "loss": 0.0073, "step": 48300 }, { "grad_norm": 0.1269981414079666, "learning_rate": 5.690752875681392e-05, "loss": 0.0079, "step": 48310 }, { "grad_norm": 0.12900133430957794, "learning_rate": 5.6891152228537435e-05, "loss": 0.0082, "step": 48320 }, { "grad_norm": 0.1164681687951088, "learning_rate": 5.687477494665492e-05, "loss": 0.0068, "step": 48330 }, { "grad_norm": 0.07722514122724533, "learning_rate": 5.685839691295733e-05, "loss": 0.0063, "step": 48340 }, { "grad_norm": 0.13309326767921448, "learning_rate": 5.6842018129235786e-05, "loss": 0.0069, "step": 48350 }, { "grad_norm": 0.13345901668071747, "learning_rate": 5.6825638597281404e-05, "loss": 0.0069, "step": 48360 }, { "grad_norm": 0.09911473840475082, "learning_rate": 5.680925831888546e-05, "loss": 0.008, "step": 48370 }, { "grad_norm": 0.119156613945961, "learning_rate": 5.6792877295839274e-05, "loss": 0.0072, "step": 48380 }, { "grad_norm": 0.14118537306785583, "learning_rate": 5.6776495529934224e-05, "loss": 0.0066, "step": 48390 }, { "grad_norm": 0.10278524458408356, "learning_rate": 5.6760113022961824e-05, "loss": 0.0071, "step": 48400 }, { "grad_norm": 0.1503407508134842, "learning_rate": 5.6743729776713617e-05, "loss": 0.0078, "step": 48410 }, { "grad_norm": 0.10129792988300323, "learning_rate": 5.672734579298126e-05, "loss": 0.007, "step": 48420 }, { "grad_norm": 0.1337733417749405, "learning_rate": 5.671096107355649e-05, "loss": 0.0076, "step": 48430 }, { "grad_norm": 0.12847165763378143, "learning_rate": 5.66945756202311e-05, "loss": 0.0079, "step": 48440 }, { "grad_norm": 0.14455513656139374, "learning_rate": 5.667818943479699e-05, "loss": 0.0088, "step": 48450 }, { "grad_norm": 0.14632147550582886, "learning_rate": 5.666180251904612e-05, "loss": 0.0062, "step": 48460 }, { "grad_norm": 0.12102360278367996, "learning_rate": 5.6645414874770555e-05, "loss": 0.0076, "step": 48470 }, { "grad_norm": 0.10775408893823624, "learning_rate": 5.66290265037624e-05, "loss": 0.0082, "step": 48480 }, { "grad_norm": 0.0898808091878891, "learning_rate": 5.661263740781386e-05, "loss": 0.0075, "step": 48490 }, { "grad_norm": 0.1148604080080986, "learning_rate": 5.6596247588717254e-05, "loss": 0.0087, "step": 48500 }, { "grad_norm": 0.1362563967704773, "learning_rate": 5.6579857048264926e-05, "loss": 0.0081, "step": 48510 }, { "grad_norm": 0.10675854980945587, "learning_rate": 5.6563465788249314e-05, "loss": 0.0072, "step": 48520 }, { "grad_norm": 0.09077155590057373, "learning_rate": 5.6547073810462956e-05, "loss": 0.0075, "step": 48530 }, { "grad_norm": 0.08170604705810547, "learning_rate": 5.653068111669846e-05, "loss": 0.0065, "step": 48540 }, { "grad_norm": 0.14370794594287872, "learning_rate": 5.651428770874848e-05, "loss": 0.0094, "step": 48550 }, { "grad_norm": 0.09480728209018707, "learning_rate": 5.64978935884058e-05, "loss": 0.0103, "step": 48560 }, { "grad_norm": 0.15089000761508942, "learning_rate": 5.6481498757463244e-05, "loss": 0.009, "step": 48570 }, { "grad_norm": 0.12338230013847351, "learning_rate": 5.646510321771373e-05, "loss": 0.0079, "step": 48580 }, { "grad_norm": 0.12670232355594635, "learning_rate": 5.644870697095024e-05, "loss": 0.0073, "step": 48590 }, { "grad_norm": 0.12560541927814484, "learning_rate": 5.643231001896586e-05, "loss": 0.0105, "step": 48600 }, { "grad_norm": 0.15912479162216187, "learning_rate": 5.6415912363553726e-05, "loss": 0.0083, "step": 48610 }, { "grad_norm": 0.1429179608821869, "learning_rate": 5.639951400650706e-05, "loss": 0.0078, "step": 48620 }, { "grad_norm": 0.15294282138347626, "learning_rate": 5.6383114949619165e-05, "loss": 0.0083, "step": 48630 }, { "grad_norm": 0.12087556719779968, "learning_rate": 5.636671519468342e-05, "loss": 0.0085, "step": 48640 }, { "grad_norm": 0.12404084950685501, "learning_rate": 5.635031474349327e-05, "loss": 0.01, "step": 48650 }, { "grad_norm": 0.09119891375303268, "learning_rate": 5.6333913597842246e-05, "loss": 0.0073, "step": 48660 }, { "grad_norm": 0.09129021316766739, "learning_rate": 5.6317511759523955e-05, "loss": 0.0067, "step": 48670 }, { "grad_norm": 0.09320438653230667, "learning_rate": 5.630110923033207e-05, "loss": 0.0067, "step": 48680 }, { "grad_norm": 0.1264292597770691, "learning_rate": 5.628470601206036e-05, "loss": 0.0074, "step": 48690 }, { "grad_norm": 0.09141099452972412, "learning_rate": 5.626830210650266e-05, "loss": 0.0083, "step": 48700 }, { "grad_norm": 0.11399321258068085, "learning_rate": 5.625189751545285e-05, "loss": 0.0072, "step": 48710 }, { "grad_norm": 0.11119510233402252, "learning_rate": 5.6235492240704936e-05, "loss": 0.0072, "step": 48720 }, { "grad_norm": 0.11012963205575943, "learning_rate": 5.621908628405296e-05, "loss": 0.0083, "step": 48730 }, { "grad_norm": 0.10664508491754532, "learning_rate": 5.620267964729106e-05, "loss": 0.0072, "step": 48740 }, { "grad_norm": 0.13803540170192719, "learning_rate": 5.618627233221344e-05, "loss": 0.0086, "step": 48750 }, { "grad_norm": 0.12187236547470093, "learning_rate": 5.616986434061438e-05, "loss": 0.0089, "step": 48760 }, { "grad_norm": 0.12868206202983856, "learning_rate": 5.615345567428822e-05, "loss": 0.0084, "step": 48770 }, { "grad_norm": 0.11364526301622391, "learning_rate": 5.613704633502941e-05, "loss": 0.0094, "step": 48780 }, { "grad_norm": 0.14713642001152039, "learning_rate": 5.612063632463245e-05, "loss": 0.0095, "step": 48790 }, { "grad_norm": 0.15311171114444733, "learning_rate": 5.610422564489188e-05, "loss": 0.0096, "step": 48800 }, { "grad_norm": 0.09973517805337906, "learning_rate": 5.608781429760239e-05, "loss": 0.0097, "step": 48810 }, { "grad_norm": 0.0991683304309845, "learning_rate": 5.607140228455866e-05, "loss": 0.0079, "step": 48820 }, { "grad_norm": 0.14886415004730225, "learning_rate": 5.605498960755553e-05, "loss": 0.0084, "step": 48830 }, { "grad_norm": 0.13686342537403107, "learning_rate": 5.603857626838782e-05, "loss": 0.0083, "step": 48840 }, { "grad_norm": 0.12803828716278076, "learning_rate": 5.60221622688505e-05, "loss": 0.0083, "step": 48850 }, { "grad_norm": 0.11861039698123932, "learning_rate": 5.6005747610738565e-05, "loss": 0.008, "step": 48860 }, { "grad_norm": 0.0915553867816925, "learning_rate": 5.59893322958471e-05, "loss": 0.006, "step": 48870 }, { "grad_norm": 0.1356211006641388, "learning_rate": 5.5972916325971256e-05, "loss": 0.0086, "step": 48880 }, { "grad_norm": 0.09840536117553711, "learning_rate": 5.595649970290628e-05, "loss": 0.0066, "step": 48890 }, { "grad_norm": 0.15195494890213013, "learning_rate": 5.5940082428447426e-05, "loss": 0.0079, "step": 48900 }, { "grad_norm": 0.1339908242225647, "learning_rate": 5.592366450439012e-05, "loss": 0.0064, "step": 48910 }, { "grad_norm": 0.09569555521011353, "learning_rate": 5.590724593252975e-05, "loss": 0.007, "step": 48920 }, { "grad_norm": 0.10590275377035141, "learning_rate": 5.589082671466184e-05, "loss": 0.0084, "step": 48930 }, { "grad_norm": 0.09708893299102783, "learning_rate": 5.587440685258199e-05, "loss": 0.0067, "step": 48940 }, { "grad_norm": 0.1184043437242508, "learning_rate": 5.585798634808583e-05, "loss": 0.0069, "step": 48950 }, { "grad_norm": 0.10826774686574936, "learning_rate": 5.584156520296909e-05, "loss": 0.0092, "step": 48960 }, { "grad_norm": 0.10714881867170334, "learning_rate": 5.582514341902757e-05, "loss": 0.0076, "step": 48970 }, { "grad_norm": 0.11400695890188217, "learning_rate": 5.580872099805713e-05, "loss": 0.009, "step": 48980 }, { "grad_norm": 0.14405634999275208, "learning_rate": 5.5792297941853674e-05, "loss": 0.0083, "step": 48990 }, { "grad_norm": 0.12499181181192398, "learning_rate": 5.5775874252213247e-05, "loss": 0.0067, "step": 49000 }, { "grad_norm": 0.10167109966278076, "learning_rate": 5.575944993093189e-05, "loss": 0.0086, "step": 49010 }, { "grad_norm": 0.11456967145204544, "learning_rate": 5.574302497980574e-05, "loss": 0.0118, "step": 49020 }, { "grad_norm": 0.17067888379096985, "learning_rate": 5.5726599400631005e-05, "loss": 0.0095, "step": 49030 }, { "grad_norm": 0.13336338102817535, "learning_rate": 5.571017319520399e-05, "loss": 0.0077, "step": 49040 }, { "grad_norm": 0.13306009769439697, "learning_rate": 5.5693746365320984e-05, "loss": 0.0102, "step": 49050 }, { "grad_norm": 0.14418716728687286, "learning_rate": 5.567731891277848e-05, "loss": 0.0116, "step": 49060 }, { "grad_norm": 0.1536301076412201, "learning_rate": 5.566089083937288e-05, "loss": 0.0103, "step": 49070 }, { "grad_norm": 0.11730632185935974, "learning_rate": 5.564446214690079e-05, "loss": 0.0091, "step": 49080 }, { "grad_norm": 0.09590926766395569, "learning_rate": 5.56280328371588e-05, "loss": 0.0078, "step": 49090 }, { "grad_norm": 0.1279791295528412, "learning_rate": 5.5611602911943596e-05, "loss": 0.0089, "step": 49100 }, { "grad_norm": 0.14919592440128326, "learning_rate": 5.559517237305194e-05, "loss": 0.0086, "step": 49110 }, { "grad_norm": 0.10418150573968887, "learning_rate": 5.557874122228064e-05, "loss": 0.0073, "step": 49120 }, { "grad_norm": 0.12159912288188934, "learning_rate": 5.55623094614266e-05, "loss": 0.0073, "step": 49130 }, { "grad_norm": 0.15408699214458466, "learning_rate": 5.5545877092286743e-05, "loss": 0.0068, "step": 49140 }, { "grad_norm": 0.12102033197879791, "learning_rate": 5.552944411665811e-05, "loss": 0.008, "step": 49150 }, { "grad_norm": 0.1368131935596466, "learning_rate": 5.5513010536337795e-05, "loss": 0.008, "step": 49160 }, { "grad_norm": 0.13384795188903809, "learning_rate": 5.5496576353122934e-05, "loss": 0.0087, "step": 49170 }, { "grad_norm": 0.16086642444133759, "learning_rate": 5.548014156881074e-05, "loss": 0.0078, "step": 49180 }, { "grad_norm": 0.1160585880279541, "learning_rate": 5.5463706185198494e-05, "loss": 0.0077, "step": 49190 }, { "grad_norm": 0.13566669821739197, "learning_rate": 5.544727020408358e-05, "loss": 0.0093, "step": 49200 }, { "grad_norm": 0.17170804738998413, "learning_rate": 5.5430833627263366e-05, "loss": 0.0096, "step": 49210 }, { "grad_norm": 0.10139906406402588, "learning_rate": 5.5414396456535354e-05, "loss": 0.0064, "step": 49220 }, { "grad_norm": 0.12546499073505402, "learning_rate": 5.539795869369709e-05, "loss": 0.0082, "step": 49230 }, { "grad_norm": 0.12277894467115402, "learning_rate": 5.538152034054618e-05, "loss": 0.0077, "step": 49240 }, { "grad_norm": 0.10608778893947601, "learning_rate": 5.536508139888028e-05, "loss": 0.007, "step": 49250 }, { "grad_norm": 0.07904994487762451, "learning_rate": 5.534864187049716e-05, "loss": 0.0073, "step": 49260 }, { "grad_norm": 0.14007234573364258, "learning_rate": 5.5332201757194615e-05, "loss": 0.01, "step": 49270 }, { "grad_norm": 0.10086888074874878, "learning_rate": 5.531576106077049e-05, "loss": 0.0077, "step": 49280 }, { "grad_norm": 0.11483555287122726, "learning_rate": 5.529931978302272e-05, "loss": 0.0095, "step": 49290 }, { "grad_norm": 0.15164752304553986, "learning_rate": 5.528287792574932e-05, "loss": 0.0094, "step": 49300 }, { "grad_norm": 0.15530359745025635, "learning_rate": 5.5266435490748324e-05, "loss": 0.0061, "step": 49310 }, { "grad_norm": 0.11662673950195312, "learning_rate": 5.524999247981787e-05, "loss": 0.0072, "step": 49320 }, { "grad_norm": 0.10972960293292999, "learning_rate": 5.523354889475613e-05, "loss": 0.0082, "step": 49330 }, { "grad_norm": 0.11394335329532623, "learning_rate": 5.521710473736134e-05, "loss": 0.0077, "step": 49340 }, { "grad_norm": 0.10676778107881546, "learning_rate": 5.5200660009431826e-05, "loss": 0.0068, "step": 49350 }, { "grad_norm": 0.10106196999549866, "learning_rate": 5.518421471276596e-05, "loss": 0.0081, "step": 49360 }, { "grad_norm": 0.09984240680932999, "learning_rate": 5.5167768849162174e-05, "loss": 0.0072, "step": 49370 }, { "grad_norm": 0.10560698062181473, "learning_rate": 5.515132242041893e-05, "loss": 0.0072, "step": 49380 }, { "grad_norm": 0.11182389408349991, "learning_rate": 5.513487542833483e-05, "loss": 0.0069, "step": 49390 }, { "grad_norm": 0.10493390262126923, "learning_rate": 5.5118427874708466e-05, "loss": 0.0063, "step": 49400 }, { "grad_norm": 0.1391921043395996, "learning_rate": 5.510197976133853e-05, "loss": 0.0086, "step": 49410 }, { "grad_norm": 0.11023185402154922, "learning_rate": 5.508553109002376e-05, "loss": 0.0066, "step": 49420 }, { "grad_norm": 0.12570008635520935, "learning_rate": 5.5069081862562957e-05, "loss": 0.0076, "step": 49430 }, { "grad_norm": 0.08499845117330551, "learning_rate": 5.5052632080754965e-05, "loss": 0.0075, "step": 49440 }, { "grad_norm": 0.10498549044132233, "learning_rate": 5.503618174639874e-05, "loss": 0.007, "step": 49450 }, { "grad_norm": 0.10718947649002075, "learning_rate": 5.501973086129325e-05, "loss": 0.0065, "step": 49460 }, { "grad_norm": 0.11503767967224121, "learning_rate": 5.500327942723753e-05, "loss": 0.0072, "step": 49470 }, { "grad_norm": 0.11931972950696945, "learning_rate": 5.498682744603071e-05, "loss": 0.0074, "step": 49480 }, { "grad_norm": 0.1149197369813919, "learning_rate": 5.497037491947194e-05, "loss": 0.0085, "step": 49490 }, { "grad_norm": 0.07871997356414795, "learning_rate": 5.4953921849360424e-05, "loss": 0.0065, "step": 49500 }, { "grad_norm": 0.09809578955173492, "learning_rate": 5.493746823749547e-05, "loss": 0.0061, "step": 49510 }, { "grad_norm": 0.13380353152751923, "learning_rate": 5.49210140856764e-05, "loss": 0.0072, "step": 49520 }, { "grad_norm": 0.12898124754428864, "learning_rate": 5.4904559395702647e-05, "loss": 0.0071, "step": 49530 }, { "grad_norm": 0.10034490376710892, "learning_rate": 5.488810416937364e-05, "loss": 0.0075, "step": 49540 }, { "grad_norm": 0.12386369705200195, "learning_rate": 5.4871648408488926e-05, "loss": 0.0084, "step": 49550 }, { "grad_norm": 0.19832555949687958, "learning_rate": 5.485519211484807e-05, "loss": 0.0079, "step": 49560 }, { "grad_norm": 0.1448063850402832, "learning_rate": 5.4838735290250676e-05, "loss": 0.0076, "step": 49570 }, { "grad_norm": 0.11774647980928421, "learning_rate": 5.482227793649649e-05, "loss": 0.0088, "step": 49580 }, { "grad_norm": 0.12724970281124115, "learning_rate": 5.480582005538524e-05, "loss": 0.0089, "step": 49590 }, { "grad_norm": 0.11640866845846176, "learning_rate": 5.478936164871671e-05, "loss": 0.0073, "step": 49600 }, { "grad_norm": 0.12501853704452515, "learning_rate": 5.477290271829081e-05, "loss": 0.0074, "step": 49610 }, { "grad_norm": 0.1101895421743393, "learning_rate": 5.475644326590744e-05, "loss": 0.0072, "step": 49620 }, { "grad_norm": 0.11681435257196426, "learning_rate": 5.473998329336658e-05, "loss": 0.008, "step": 49630 }, { "grad_norm": 0.08980178087949753, "learning_rate": 5.4723522802468286e-05, "loss": 0.0103, "step": 49640 }, { "grad_norm": 0.10893316566944122, "learning_rate": 5.470706179501264e-05, "loss": 0.0078, "step": 49650 }, { "grad_norm": 0.11148808151483536, "learning_rate": 5.4690600272799776e-05, "loss": 0.0077, "step": 49660 }, { "grad_norm": 0.13288681209087372, "learning_rate": 5.467413823762993e-05, "loss": 0.0071, "step": 49670 }, { "grad_norm": 0.12644045054912567, "learning_rate": 5.465767569130335e-05, "loss": 0.0069, "step": 49680 }, { "grad_norm": 0.11375681310892105, "learning_rate": 5.464121263562036e-05, "loss": 0.0069, "step": 49690 }, { "grad_norm": 0.1390065848827362, "learning_rate": 5.4624749072381343e-05, "loss": 0.0079, "step": 49700 }, { "grad_norm": 0.11371461302042007, "learning_rate": 5.460828500338672e-05, "loss": 0.0064, "step": 49710 }, { "grad_norm": 0.13284237682819366, "learning_rate": 5.459182043043698e-05, "loss": 0.0079, "step": 49720 }, { "grad_norm": 0.15084634721279144, "learning_rate": 5.457535535533265e-05, "loss": 0.008, "step": 49730 }, { "grad_norm": 0.10327810794115067, "learning_rate": 5.4558889779874334e-05, "loss": 0.0055, "step": 49740 }, { "grad_norm": 0.07924386858940125, "learning_rate": 5.454242370586269e-05, "loss": 0.0082, "step": 49750 }, { "grad_norm": 0.11624611914157867, "learning_rate": 5.452595713509843e-05, "loss": 0.0072, "step": 49760 }, { "grad_norm": 0.09631162136793137, "learning_rate": 5.4509490069382274e-05, "loss": 0.0071, "step": 49770 }, { "grad_norm": 0.11582771688699722, "learning_rate": 5.449302251051509e-05, "loss": 0.0088, "step": 49780 }, { "grad_norm": 0.12504427134990692, "learning_rate": 5.44765544602977e-05, "loss": 0.0081, "step": 49790 }, { "grad_norm": 0.12355445325374603, "learning_rate": 5.446008592053107e-05, "loss": 0.0082, "step": 49800 }, { "grad_norm": 0.11303366720676422, "learning_rate": 5.4443616893016135e-05, "loss": 0.0081, "step": 49810 }, { "grad_norm": 0.10228270292282104, "learning_rate": 5.4427147379553934e-05, "loss": 0.01, "step": 49820 }, { "grad_norm": 0.0984843447804451, "learning_rate": 5.441067738194556e-05, "loss": 0.0095, "step": 49830 }, { "grad_norm": 0.09836787730455399, "learning_rate": 5.439420690199214e-05, "loss": 0.0069, "step": 49840 }, { "grad_norm": 0.18476708233356476, "learning_rate": 5.437773594149484e-05, "loss": 0.0083, "step": 49850 }, { "grad_norm": 0.09439171105623245, "learning_rate": 5.436126450225495e-05, "loss": 0.0095, "step": 49860 }, { "grad_norm": 0.07528919726610184, "learning_rate": 5.434479258607371e-05, "loss": 0.0074, "step": 49870 }, { "grad_norm": 0.13574284315109253, "learning_rate": 5.43283201947525e-05, "loss": 0.0073, "step": 49880 }, { "grad_norm": 0.11229805648326874, "learning_rate": 5.431184733009268e-05, "loss": 0.0079, "step": 49890 }, { "grad_norm": 0.13793551921844482, "learning_rate": 5.4295373993895736e-05, "loss": 0.009, "step": 49900 }, { "grad_norm": 0.10786230117082596, "learning_rate": 5.4278900187963157e-05, "loss": 0.0071, "step": 49910 }, { "grad_norm": 0.11431339383125305, "learning_rate": 5.426242591409646e-05, "loss": 0.009, "step": 49920 }, { "grad_norm": 0.1384664624929428, "learning_rate": 5.42459511740973e-05, "loss": 0.0087, "step": 49930 }, { "grad_norm": 0.24672669172286987, "learning_rate": 5.422947596976729e-05, "loss": 0.0104, "step": 49940 }, { "grad_norm": 0.0851411446928978, "learning_rate": 5.4213000302908134e-05, "loss": 0.0085, "step": 49950 }, { "grad_norm": 0.1344834715127945, "learning_rate": 5.419652417532162e-05, "loss": 0.0073, "step": 49960 }, { "grad_norm": 0.0986708402633667, "learning_rate": 5.4180047588809534e-05, "loss": 0.0073, "step": 49970 }, { "grad_norm": 0.12980592250823975, "learning_rate": 5.4163570545173704e-05, "loss": 0.0078, "step": 49980 }, { "grad_norm": 0.09912262111902237, "learning_rate": 5.414709304621608e-05, "loss": 0.0076, "step": 49990 }, { "grad_norm": 0.11323543637990952, "learning_rate": 5.413061509373858e-05, "loss": 0.0073, "step": 50000 }, { "grad_norm": 0.10197105258703232, "learning_rate": 5.411413668954324e-05, "loss": 0.0078, "step": 50010 }, { "grad_norm": 0.12447931617498398, "learning_rate": 5.4097657835432083e-05, "loss": 0.0082, "step": 50020 }, { "grad_norm": 0.09024901688098907, "learning_rate": 5.408117853320723e-05, "loss": 0.0061, "step": 50030 }, { "grad_norm": 0.13368751108646393, "learning_rate": 5.406469878467084e-05, "loss": 0.0088, "step": 50040 }, { "grad_norm": 0.12775294482707977, "learning_rate": 5.404821859162509e-05, "loss": 0.0079, "step": 50050 }, { "grad_norm": 0.16053181886672974, "learning_rate": 5.403173795587225e-05, "loss": 0.0089, "step": 50060 }, { "grad_norm": 0.13509531319141388, "learning_rate": 5.4015256879214606e-05, "loss": 0.0106, "step": 50070 }, { "grad_norm": 0.10540369153022766, "learning_rate": 5.39987753634545e-05, "loss": 0.0088, "step": 50080 }, { "grad_norm": 0.10227992385625839, "learning_rate": 5.3982293410394335e-05, "loss": 0.0068, "step": 50090 }, { "grad_norm": 0.13216997683048248, "learning_rate": 5.396581102183655e-05, "loss": 0.0076, "step": 50100 }, { "grad_norm": 0.10522474348545074, "learning_rate": 5.3949328199583634e-05, "loss": 0.008, "step": 50110 }, { "grad_norm": 0.13778546452522278, "learning_rate": 5.393284494543813e-05, "loss": 0.0085, "step": 50120 }, { "grad_norm": 0.14291337132453918, "learning_rate": 5.391636126120262e-05, "loss": 0.009, "step": 50130 }, { "grad_norm": 0.14628663659095764, "learning_rate": 5.389987714867971e-05, "loss": 0.0096, "step": 50140 }, { "grad_norm": 0.12206447869539261, "learning_rate": 5.3883392609672114e-05, "loss": 0.0091, "step": 50150 }, { "grad_norm": 0.10824067145586014, "learning_rate": 5.3866907645982545e-05, "loss": 0.0071, "step": 50160 }, { "grad_norm": 0.09216737002134323, "learning_rate": 5.385042225941375e-05, "loss": 0.0091, "step": 50170 }, { "grad_norm": 0.14004041254520416, "learning_rate": 5.3833936451768585e-05, "loss": 0.0085, "step": 50180 }, { "grad_norm": 0.09411123394966125, "learning_rate": 5.3817450224849884e-05, "loss": 0.009, "step": 50190 }, { "grad_norm": 0.10891849547624588, "learning_rate": 5.380096358046056e-05, "loss": 0.0095, "step": 50200 }, { "grad_norm": 0.11818989366292953, "learning_rate": 5.378447652040359e-05, "loss": 0.0068, "step": 50210 }, { "grad_norm": 0.11016495525836945, "learning_rate": 5.376798904648195e-05, "loss": 0.0077, "step": 50220 }, { "grad_norm": 0.12545368075370789, "learning_rate": 5.375150116049869e-05, "loss": 0.008, "step": 50230 }, { "grad_norm": 0.11391474306583405, "learning_rate": 5.373501286425691e-05, "loss": 0.0064, "step": 50240 }, { "grad_norm": 0.09399053454399109, "learning_rate": 5.3718524159559726e-05, "loss": 0.0056, "step": 50250 }, { "grad_norm": 0.1270333081483841, "learning_rate": 5.370203504821034e-05, "loss": 0.0072, "step": 50260 }, { "grad_norm": 0.1509568691253662, "learning_rate": 5.368554553201196e-05, "loss": 0.0091, "step": 50270 }, { "grad_norm": 0.13258443772792816, "learning_rate": 5.3669055612767874e-05, "loss": 0.008, "step": 50280 }, { "grad_norm": 0.10823394358158112, "learning_rate": 5.3652565292281376e-05, "loss": 0.0072, "step": 50290 }, { "grad_norm": 0.1237456426024437, "learning_rate": 5.3636074572355824e-05, "loss": 0.0072, "step": 50300 }, { "grad_norm": 0.11638401448726654, "learning_rate": 5.361958345479463e-05, "loss": 0.0064, "step": 50310 }, { "grad_norm": 0.12386294454336166, "learning_rate": 5.360309194140123e-05, "loss": 0.0085, "step": 50320 }, { "grad_norm": 0.10697276145219803, "learning_rate": 5.3586600033979106e-05, "loss": 0.0067, "step": 50330 }, { "grad_norm": 0.10902923345565796, "learning_rate": 5.357010773433181e-05, "loss": 0.0076, "step": 50340 }, { "grad_norm": 0.10591767728328705, "learning_rate": 5.3553615044262895e-05, "loss": 0.0065, "step": 50350 }, { "grad_norm": 0.116392120718956, "learning_rate": 5.3537121965575964e-05, "loss": 0.0063, "step": 50360 }, { "grad_norm": 0.14380685985088348, "learning_rate": 5.352062850007471e-05, "loss": 0.0085, "step": 50370 }, { "grad_norm": 0.1494550108909607, "learning_rate": 5.350413464956282e-05, "loss": 0.007, "step": 50380 }, { "grad_norm": 0.1233399286866188, "learning_rate": 5.348764041584403e-05, "loss": 0.0073, "step": 50390 }, { "grad_norm": 0.13739190995693207, "learning_rate": 5.347114580072212e-05, "loss": 0.0072, "step": 50400 }, { "grad_norm": 0.11893120408058167, "learning_rate": 5.345465080600093e-05, "loss": 0.0083, "step": 50410 }, { "grad_norm": 0.1763337403535843, "learning_rate": 5.3438155433484326e-05, "loss": 0.0078, "step": 50420 }, { "grad_norm": 0.13362959027290344, "learning_rate": 5.3421659684976197e-05, "loss": 0.0073, "step": 50430 }, { "grad_norm": 0.08828156441450119, "learning_rate": 5.340516356228052e-05, "loss": 0.0065, "step": 50440 }, { "grad_norm": 0.12269148975610733, "learning_rate": 5.338866706720128e-05, "loss": 0.0083, "step": 50450 }, { "grad_norm": 0.12975814938545227, "learning_rate": 5.337217020154249e-05, "loss": 0.009, "step": 50460 }, { "grad_norm": 0.12157153338193893, "learning_rate": 5.335567296710825e-05, "loss": 0.0108, "step": 50470 }, { "grad_norm": 0.09501282125711441, "learning_rate": 5.333917536570265e-05, "loss": 0.0083, "step": 50480 }, { "grad_norm": 0.14768700301647186, "learning_rate": 5.332267739912986e-05, "loss": 0.0073, "step": 50490 }, { "grad_norm": 0.10247260332107544, "learning_rate": 5.330617906919405e-05, "loss": 0.0066, "step": 50500 }, { "grad_norm": 0.12501023709774017, "learning_rate": 5.328968037769949e-05, "loss": 0.0083, "step": 50510 }, { "grad_norm": 0.11878281086683273, "learning_rate": 5.327318132645042e-05, "loss": 0.007, "step": 50520 }, { "grad_norm": 0.13634929060935974, "learning_rate": 5.325668191725116e-05, "loss": 0.0066, "step": 50530 }, { "grad_norm": 0.12956200540065765, "learning_rate": 5.324018215190606e-05, "loss": 0.0073, "step": 50540 }, { "grad_norm": 0.09281910955905914, "learning_rate": 5.3223682032219515e-05, "loss": 0.0063, "step": 50550 }, { "grad_norm": 0.12302006036043167, "learning_rate": 5.320718155999595e-05, "loss": 0.0074, "step": 50560 }, { "grad_norm": 0.13284669816493988, "learning_rate": 5.3190680737039835e-05, "loss": 0.008, "step": 50570 }, { "grad_norm": 0.12234926223754883, "learning_rate": 5.317417956515567e-05, "loss": 0.0063, "step": 50580 }, { "grad_norm": 0.11297574639320374, "learning_rate": 5.315767804614798e-05, "loss": 0.0074, "step": 50590 }, { "grad_norm": 0.11877802014350891, "learning_rate": 5.3141176181821395e-05, "loss": 0.0074, "step": 50600 }, { "grad_norm": 0.10327239334583282, "learning_rate": 5.31246739739805e-05, "loss": 0.0091, "step": 50610 }, { "grad_norm": 0.11630271375179291, "learning_rate": 5.310817142442994e-05, "loss": 0.0076, "step": 50620 }, { "grad_norm": 0.09944736957550049, "learning_rate": 5.309166853497445e-05, "loss": 0.0069, "step": 50630 }, { "grad_norm": 0.08324733376502991, "learning_rate": 5.307516530741873e-05, "loss": 0.0079, "step": 50640 }, { "grad_norm": 0.12145577371120453, "learning_rate": 5.305866174356754e-05, "loss": 0.0089, "step": 50650 }, { "grad_norm": 0.1235630065202713, "learning_rate": 5.304215784522571e-05, "loss": 0.0056, "step": 50660 }, { "grad_norm": 0.10293228924274445, "learning_rate": 5.302565361419808e-05, "loss": 0.0069, "step": 50670 }, { "grad_norm": 0.11646381765604019, "learning_rate": 5.3009149052289507e-05, "loss": 0.007, "step": 50680 }, { "grad_norm": 0.1279471218585968, "learning_rate": 5.299264416130493e-05, "loss": 0.0062, "step": 50690 }, { "grad_norm": 0.10489293932914734, "learning_rate": 5.297613894304928e-05, "loss": 0.0076, "step": 50700 }, { "grad_norm": 0.10943470150232315, "learning_rate": 5.2959633399327534e-05, "loss": 0.0071, "step": 50710 }, { "grad_norm": 0.11744818836450577, "learning_rate": 5.294312753194476e-05, "loss": 0.0094, "step": 50720 }, { "grad_norm": 0.12632669508457184, "learning_rate": 5.292662134270596e-05, "loss": 0.0076, "step": 50730 }, { "grad_norm": 0.12011095136404037, "learning_rate": 5.291011483341626e-05, "loss": 0.0064, "step": 50740 }, { "grad_norm": 0.10209272056818008, "learning_rate": 5.2893608005880767e-05, "loss": 0.0076, "step": 50750 }, { "grad_norm": 0.1162542924284935, "learning_rate": 5.287710086190467e-05, "loss": 0.0073, "step": 50760 }, { "grad_norm": 0.11127622425556183, "learning_rate": 5.286059340329314e-05, "loss": 0.0081, "step": 50770 }, { "grad_norm": 0.14879606664180756, "learning_rate": 5.284408563185141e-05, "loss": 0.0093, "step": 50780 }, { "grad_norm": 0.16978897154331207, "learning_rate": 5.282757754938476e-05, "loss": 0.0086, "step": 50790 }, { "grad_norm": 0.12024236470460892, "learning_rate": 5.281106915769849e-05, "loss": 0.0091, "step": 50800 }, { "grad_norm": 0.11224010586738586, "learning_rate": 5.2794560458597897e-05, "loss": 0.0059, "step": 50810 }, { "grad_norm": 0.10703389346599579, "learning_rate": 5.27780514538884e-05, "loss": 0.0098, "step": 50820 }, { "grad_norm": 0.09390374273061752, "learning_rate": 5.2761542145375365e-05, "loss": 0.0074, "step": 50830 }, { "grad_norm": 0.10188932716846466, "learning_rate": 5.274503253486421e-05, "loss": 0.0068, "step": 50840 }, { "grad_norm": 0.10089115053415298, "learning_rate": 5.272852262416046e-05, "loss": 0.0062, "step": 50850 }, { "grad_norm": 0.11556477099657059, "learning_rate": 5.2712012415069555e-05, "loss": 0.0089, "step": 50860 }, { "grad_norm": 0.09779496490955353, "learning_rate": 5.269550190939705e-05, "loss": 0.0072, "step": 50870 }, { "grad_norm": 0.1219988614320755, "learning_rate": 5.267899110894852e-05, "loss": 0.0076, "step": 50880 }, { "grad_norm": 0.12894068658351898, "learning_rate": 5.266248001552955e-05, "loss": 0.0074, "step": 50890 }, { "grad_norm": 0.15899167954921722, "learning_rate": 5.264596863094575e-05, "loss": 0.0073, "step": 50900 }, { "grad_norm": 0.0752955749630928, "learning_rate": 5.262945695700282e-05, "loss": 0.0074, "step": 50910 }, { "grad_norm": 0.09523538500070572, "learning_rate": 5.261294499550643e-05, "loss": 0.0082, "step": 50920 }, { "grad_norm": 0.11954891681671143, "learning_rate": 5.25964327482623e-05, "loss": 0.0094, "step": 50930 }, { "grad_norm": 0.11360318958759308, "learning_rate": 5.257992021707617e-05, "loss": 0.0069, "step": 50940 }, { "grad_norm": 0.12479811906814575, "learning_rate": 5.256340740375387e-05, "loss": 0.0082, "step": 50950 }, { "grad_norm": 0.13516080379486084, "learning_rate": 5.254689431010117e-05, "loss": 0.0078, "step": 50960 }, { "grad_norm": 0.12615756690502167, "learning_rate": 5.253038093792395e-05, "loss": 0.0065, "step": 50970 }, { "grad_norm": 0.10402563959360123, "learning_rate": 5.251386728902806e-05, "loss": 0.0089, "step": 50980 }, { "grad_norm": 0.14497220516204834, "learning_rate": 5.2497353365219446e-05, "loss": 0.0076, "step": 50990 }, { "grad_norm": 0.1061200425028801, "learning_rate": 5.2480839168304e-05, "loss": 0.0074, "step": 51000 }, { "grad_norm": 0.10565248876810074, "learning_rate": 5.246432470008773e-05, "loss": 0.0067, "step": 51010 }, { "grad_norm": 0.09415991604328156, "learning_rate": 5.2447809962376616e-05, "loss": 0.0065, "step": 51020 }, { "grad_norm": 0.14689849317073822, "learning_rate": 5.243129495697666e-05, "loss": 0.0078, "step": 51030 }, { "grad_norm": 0.10472677648067474, "learning_rate": 5.241477968569397e-05, "loss": 0.0073, "step": 51040 }, { "grad_norm": 0.11897967010736465, "learning_rate": 5.23982641503346e-05, "loss": 0.0072, "step": 51050 }, { "grad_norm": 0.12270940095186234, "learning_rate": 5.238174835270465e-05, "loss": 0.008, "step": 51060 }, { "grad_norm": 0.15388749539852142, "learning_rate": 5.23652322946103e-05, "loss": 0.0086, "step": 51070 }, { "grad_norm": 0.09850005805492401, "learning_rate": 5.234871597785771e-05, "loss": 0.0078, "step": 51080 }, { "grad_norm": 0.114398293197155, "learning_rate": 5.233219940425306e-05, "loss": 0.0075, "step": 51090 }, { "grad_norm": 0.21838240325450897, "learning_rate": 5.2315682575602576e-05, "loss": 0.0088, "step": 51100 }, { "grad_norm": 0.15210293233394623, "learning_rate": 5.229916549371254e-05, "loss": 0.0092, "step": 51110 }, { "grad_norm": 0.1417931169271469, "learning_rate": 5.228264816038924e-05, "loss": 0.0073, "step": 51120 }, { "grad_norm": 0.11199631541967392, "learning_rate": 5.2266130577438935e-05, "loss": 0.0078, "step": 51130 }, { "grad_norm": 0.0930638313293457, "learning_rate": 5.224961274666802e-05, "loss": 0.0066, "step": 51140 }, { "grad_norm": 0.09415660798549652, "learning_rate": 5.223309466988283e-05, "loss": 0.0072, "step": 51150 }, { "grad_norm": 0.12282320111989975, "learning_rate": 5.221657634888976e-05, "loss": 0.0078, "step": 51160 }, { "grad_norm": 0.1348143219947815, "learning_rate": 5.2200057785495235e-05, "loss": 0.0093, "step": 51170 }, { "grad_norm": 0.16877859830856323, "learning_rate": 5.2183538981505706e-05, "loss": 0.0082, "step": 51180 }, { "grad_norm": 0.14137619733810425, "learning_rate": 5.216701993872762e-05, "loss": 0.0075, "step": 51190 }, { "grad_norm": 0.11979001015424728, "learning_rate": 5.215050065896752e-05, "loss": 0.0071, "step": 51200 }, { "grad_norm": 0.1457090973854065, "learning_rate": 5.2133981144031875e-05, "loss": 0.0069, "step": 51210 }, { "grad_norm": 0.1234879121184349, "learning_rate": 5.211746139572727e-05, "loss": 0.008, "step": 51220 }, { "grad_norm": 0.1285405308008194, "learning_rate": 5.210094141586026e-05, "loss": 0.0078, "step": 51230 }, { "grad_norm": 0.11687616258859634, "learning_rate": 5.2084421206237476e-05, "loss": 0.0075, "step": 51240 }, { "grad_norm": 0.13387563824653625, "learning_rate": 5.20679007686655e-05, "loss": 0.0082, "step": 51250 }, { "grad_norm": 0.22956760227680206, "learning_rate": 5.205138010495102e-05, "loss": 0.0087, "step": 51260 }, { "grad_norm": 0.1636827439069748, "learning_rate": 5.20348592169007e-05, "loss": 0.0087, "step": 51270 }, { "grad_norm": 0.11392360925674438, "learning_rate": 5.201833810632123e-05, "loss": 0.0086, "step": 51280 }, { "grad_norm": 0.13185438513755798, "learning_rate": 5.200181677501933e-05, "loss": 0.0073, "step": 51290 }, { "grad_norm": 0.09126608818769455, "learning_rate": 5.1985295224801777e-05, "loss": 0.0088, "step": 51300 }, { "grad_norm": 0.11725602298974991, "learning_rate": 5.196877345747531e-05, "loss": 0.0075, "step": 51310 }, { "grad_norm": 0.09102214127779007, "learning_rate": 5.195225147484674e-05, "loss": 0.0072, "step": 51320 }, { "grad_norm": 0.37099722027778625, "learning_rate": 5.193572927872288e-05, "loss": 0.0071, "step": 51330 }, { "grad_norm": 0.11040888726711273, "learning_rate": 5.191920687091059e-05, "loss": 0.0079, "step": 51340 }, { "grad_norm": 0.10897205024957657, "learning_rate": 5.1902684253216695e-05, "loss": 0.0077, "step": 51350 }, { "grad_norm": 0.11997353285551071, "learning_rate": 5.1886161427448134e-05, "loss": 0.0075, "step": 51360 }, { "grad_norm": 0.12480968236923218, "learning_rate": 5.186963839541179e-05, "loss": 0.0083, "step": 51370 }, { "grad_norm": 0.14976684749126434, "learning_rate": 5.185311515891459e-05, "loss": 0.0104, "step": 51380 }, { "grad_norm": 0.1932750940322876, "learning_rate": 5.1836591719763516e-05, "loss": 0.0078, "step": 51390 }, { "grad_norm": 0.10499188303947449, "learning_rate": 5.182006807976554e-05, "loss": 0.0079, "step": 51400 }, { "grad_norm": 0.12035854160785675, "learning_rate": 5.180354424072763e-05, "loss": 0.0071, "step": 51410 }, { "grad_norm": 0.11804483085870743, "learning_rate": 5.178702020445686e-05, "loss": 0.0077, "step": 51420 }, { "grad_norm": 0.09803120791912079, "learning_rate": 5.177049597276024e-05, "loss": 0.0079, "step": 51430 }, { "grad_norm": 0.20123162865638733, "learning_rate": 5.175397154744482e-05, "loss": 0.0077, "step": 51440 }, { "grad_norm": 0.15547233819961548, "learning_rate": 5.1737446930317736e-05, "loss": 0.0082, "step": 51450 }, { "grad_norm": 0.11203790456056595, "learning_rate": 5.172092212318604e-05, "loss": 0.008, "step": 51460 }, { "grad_norm": 0.10982844978570938, "learning_rate": 5.170439712785692e-05, "loss": 0.0078, "step": 51470 }, { "grad_norm": 0.11439083516597748, "learning_rate": 5.1687871946137467e-05, "loss": 0.0081, "step": 51480 }, { "grad_norm": 0.10995069891214371, "learning_rate": 5.16713465798349e-05, "loss": 0.0082, "step": 51490 }, { "grad_norm": 0.0945575013756752, "learning_rate": 5.165482103075637e-05, "loss": 0.007, "step": 51500 }, { "grad_norm": 0.1157839223742485, "learning_rate": 5.163829530070909e-05, "loss": 0.0066, "step": 51510 }, { "grad_norm": 0.09521379321813583, "learning_rate": 5.162176939150032e-05, "loss": 0.0069, "step": 51520 }, { "grad_norm": 0.12099765986204147, "learning_rate": 5.160524330493728e-05, "loss": 0.0063, "step": 51530 }, { "grad_norm": 0.091921865940094, "learning_rate": 5.158871704282724e-05, "loss": 0.0104, "step": 51540 }, { "grad_norm": 0.10630559176206589, "learning_rate": 5.15721906069775e-05, "loss": 0.0069, "step": 51550 }, { "grad_norm": 0.1221473440527916, "learning_rate": 5.155566399919536e-05, "loss": 0.0069, "step": 51560 }, { "grad_norm": 0.11338113993406296, "learning_rate": 5.153913722128813e-05, "loss": 0.007, "step": 51570 }, { "grad_norm": 0.12075787037611008, "learning_rate": 5.1522610275063176e-05, "loss": 0.0069, "step": 51580 }, { "grad_norm": 0.09338491410017014, "learning_rate": 5.150608316232786e-05, "loss": 0.007, "step": 51590 }, { "grad_norm": 0.09470497071743011, "learning_rate": 5.148955588488955e-05, "loss": 0.0072, "step": 51600 }, { "grad_norm": 0.15804915130138397, "learning_rate": 5.147302844455564e-05, "loss": 0.0098, "step": 51610 }, { "grad_norm": 0.16869382560253143, "learning_rate": 5.145650084313356e-05, "loss": 0.0085, "step": 51620 }, { "grad_norm": 0.10940574109554291, "learning_rate": 5.1439973082430734e-05, "loss": 0.0101, "step": 51630 }, { "grad_norm": 0.1543593406677246, "learning_rate": 5.142344516425461e-05, "loss": 0.0091, "step": 51640 }, { "grad_norm": 0.11941030621528625, "learning_rate": 5.1406917090412673e-05, "loss": 0.0089, "step": 51650 }, { "grad_norm": 0.12603560090065002, "learning_rate": 5.1390388862712404e-05, "loss": 0.0078, "step": 51660 }, { "grad_norm": 0.11978060752153397, "learning_rate": 5.137386048296129e-05, "loss": 0.009, "step": 51670 }, { "grad_norm": 0.16985154151916504, "learning_rate": 5.135733195296686e-05, "loss": 0.0075, "step": 51680 }, { "grad_norm": 0.12258169054985046, "learning_rate": 5.134080327453664e-05, "loss": 0.0073, "step": 51690 }, { "grad_norm": 0.17476427555084229, "learning_rate": 5.1324274449478204e-05, "loss": 0.0069, "step": 51700 }, { "grad_norm": 0.11227120459079742, "learning_rate": 5.130774547959909e-05, "loss": 0.0068, "step": 51710 }, { "grad_norm": 0.11604069173336029, "learning_rate": 5.129121636670692e-05, "loss": 0.0087, "step": 51720 }, { "grad_norm": 0.10399151593446732, "learning_rate": 5.1274687112609246e-05, "loss": 0.0076, "step": 51730 }, { "grad_norm": 0.0973660796880722, "learning_rate": 5.125815771911372e-05, "loss": 0.0072, "step": 51740 }, { "grad_norm": 0.0799189954996109, "learning_rate": 5.1241628188027956e-05, "loss": 0.0071, "step": 51750 }, { "grad_norm": 0.09778250753879547, "learning_rate": 5.1225098521159585e-05, "loss": 0.0059, "step": 51760 }, { "grad_norm": 0.06890701502561569, "learning_rate": 5.12085687203163e-05, "loss": 0.007, "step": 51770 }, { "grad_norm": 0.10330969095230103, "learning_rate": 5.119203878730575e-05, "loss": 0.0057, "step": 51780 }, { "grad_norm": 0.10861076414585114, "learning_rate": 5.117550872393564e-05, "loss": 0.0077, "step": 51790 }, { "grad_norm": 0.08859899640083313, "learning_rate": 5.1158978532013645e-05, "loss": 0.0054, "step": 51800 }, { "grad_norm": 0.11545345187187195, "learning_rate": 5.114244821334751e-05, "loss": 0.0063, "step": 51810 }, { "grad_norm": 0.1389436274766922, "learning_rate": 5.112591776974497e-05, "loss": 0.0063, "step": 51820 }, { "grad_norm": 0.1318819671869278, "learning_rate": 5.110938720301374e-05, "loss": 0.0067, "step": 51830 }, { "grad_norm": 0.13695287704467773, "learning_rate": 5.1092856514961595e-05, "loss": 0.0069, "step": 51840 }, { "grad_norm": 0.14111725986003876, "learning_rate": 5.107632570739631e-05, "loss": 0.0076, "step": 51850 }, { "grad_norm": 0.11050890386104584, "learning_rate": 5.1059794782125645e-05, "loss": 0.0074, "step": 51860 }, { "grad_norm": 0.12948457896709442, "learning_rate": 5.104326374095744e-05, "loss": 0.0091, "step": 51870 }, { "grad_norm": 0.13417790830135345, "learning_rate": 5.102673258569947e-05, "loss": 0.0069, "step": 51880 }, { "grad_norm": 0.08984477818012238, "learning_rate": 5.101020131815956e-05, "loss": 0.0062, "step": 51890 }, { "grad_norm": 0.12276940047740936, "learning_rate": 5.099366994014556e-05, "loss": 0.0065, "step": 51900 }, { "grad_norm": 0.12216682732105255, "learning_rate": 5.097713845346532e-05, "loss": 0.0065, "step": 51910 }, { "grad_norm": 0.11343276500701904, "learning_rate": 5.096060685992668e-05, "loss": 0.0063, "step": 51920 }, { "grad_norm": 0.08726805448532104, "learning_rate": 5.0944075161337514e-05, "loss": 0.0078, "step": 51930 }, { "grad_norm": 0.1307234764099121, "learning_rate": 5.09275433595057e-05, "loss": 0.0079, "step": 51940 }, { "grad_norm": 0.1482861191034317, "learning_rate": 5.0911011456239157e-05, "loss": 0.0062, "step": 51950 }, { "grad_norm": 0.0944008156657219, "learning_rate": 5.089447945334574e-05, "loss": 0.0084, "step": 51960 }, { "grad_norm": 0.14643219113349915, "learning_rate": 5.087794735263343e-05, "loss": 0.0062, "step": 51970 }, { "grad_norm": 0.15736916661262512, "learning_rate": 5.086141515591011e-05, "loss": 0.0085, "step": 51980 }, { "grad_norm": 0.13040560483932495, "learning_rate": 5.084488286498371e-05, "loss": 0.006, "step": 51990 }, { "grad_norm": 0.1922534555196762, "learning_rate": 5.08283504816622e-05, "loss": 0.0089, "step": 52000 }, { "grad_norm": 0.11159082502126694, "learning_rate": 5.081181800775353e-05, "loss": 0.0066, "step": 52010 }, { "grad_norm": 0.14010168612003326, "learning_rate": 5.079528544506565e-05, "loss": 0.006, "step": 52020 }, { "grad_norm": 0.11759559065103531, "learning_rate": 5.077875279540657e-05, "loss": 0.007, "step": 52030 }, { "grad_norm": 0.11398293823003769, "learning_rate": 5.076222006058424e-05, "loss": 0.0086, "step": 52040 }, { "grad_norm": 0.10994013398885727, "learning_rate": 5.074568724240667e-05, "loss": 0.0058, "step": 52050 }, { "grad_norm": 0.09920304268598557, "learning_rate": 5.072915434268187e-05, "loss": 0.0064, "step": 52060 }, { "grad_norm": 0.10778865218162537, "learning_rate": 5.071262136321785e-05, "loss": 0.0061, "step": 52070 }, { "grad_norm": 0.11429785192012787, "learning_rate": 5.069608830582262e-05, "loss": 0.0052, "step": 52080 }, { "grad_norm": 0.09465669095516205, "learning_rate": 5.067955517230423e-05, "loss": 0.008, "step": 52090 }, { "grad_norm": 0.11730226129293442, "learning_rate": 5.06630219644707e-05, "loss": 0.0063, "step": 52100 }, { "grad_norm": 0.12193343043327332, "learning_rate": 5.064648868413008e-05, "loss": 0.0065, "step": 52110 }, { "grad_norm": 0.09554997831583023, "learning_rate": 5.062995533309044e-05, "loss": 0.0073, "step": 52120 }, { "grad_norm": 0.15233108401298523, "learning_rate": 5.061342191315983e-05, "loss": 0.0078, "step": 52130 }, { "grad_norm": 0.0749034509062767, "learning_rate": 5.0596888426146325e-05, "loss": 0.0072, "step": 52140 }, { "grad_norm": 0.19272708892822266, "learning_rate": 5.058035487385798e-05, "loss": 0.0069, "step": 52150 }, { "grad_norm": 0.0895005464553833, "learning_rate": 5.0563821258102915e-05, "loss": 0.0067, "step": 52160 }, { "grad_norm": 0.11209453642368317, "learning_rate": 5.05472875806892e-05, "loss": 0.0067, "step": 52170 }, { "grad_norm": 0.11203105747699738, "learning_rate": 5.053075384342494e-05, "loss": 0.0066, "step": 52180 }, { "grad_norm": 0.122292160987854, "learning_rate": 5.0514220048118234e-05, "loss": 0.006, "step": 52190 }, { "grad_norm": 0.08336066454648972, "learning_rate": 5.049768619657721e-05, "loss": 0.0079, "step": 52200 }, { "grad_norm": 0.11180561780929565, "learning_rate": 5.048115229060996e-05, "loss": 0.0071, "step": 52210 }, { "grad_norm": 0.11848894506692886, "learning_rate": 5.046461833202462e-05, "loss": 0.0074, "step": 52220 }, { "grad_norm": 0.09497872740030289, "learning_rate": 5.044808432262932e-05, "loss": 0.0081, "step": 52230 }, { "grad_norm": 0.10889280587434769, "learning_rate": 5.043155026423219e-05, "loss": 0.0064, "step": 52240 }, { "grad_norm": 0.1172417551279068, "learning_rate": 5.041501615864137e-05, "loss": 0.0063, "step": 52250 }, { "grad_norm": 0.1563367247581482, "learning_rate": 5.039848200766502e-05, "loss": 0.0073, "step": 52260 }, { "grad_norm": 0.13128718733787537, "learning_rate": 5.038194781311125e-05, "loss": 0.0067, "step": 52270 }, { "grad_norm": 0.10903235524892807, "learning_rate": 5.0365413576788264e-05, "loss": 0.0062, "step": 52280 }, { "grad_norm": 0.08749507367610931, "learning_rate": 5.0348879300504194e-05, "loss": 0.0073, "step": 52290 }, { "grad_norm": 0.10938242077827454, "learning_rate": 5.033234498606721e-05, "loss": 0.0078, "step": 52300 }, { "grad_norm": 0.0939393937587738, "learning_rate": 5.0315810635285476e-05, "loss": 0.006, "step": 52310 }, { "grad_norm": 0.09510872513055801, "learning_rate": 5.0299276249967164e-05, "loss": 0.0074, "step": 52320 }, { "grad_norm": 0.14273355901241302, "learning_rate": 5.0282741831920454e-05, "loss": 0.0078, "step": 52330 }, { "grad_norm": 0.08332997560501099, "learning_rate": 5.026620738295351e-05, "loss": 0.0067, "step": 52340 }, { "grad_norm": 0.13117820024490356, "learning_rate": 5.024967290487455e-05, "loss": 0.0072, "step": 52350 }, { "grad_norm": 0.10564810782670975, "learning_rate": 5.023313839949172e-05, "loss": 0.0075, "step": 52360 }, { "grad_norm": 0.12033029645681381, "learning_rate": 5.0216603868613224e-05, "loss": 0.008, "step": 52370 }, { "grad_norm": 0.138861283659935, "learning_rate": 5.020006931404726e-05, "loss": 0.0074, "step": 52380 }, { "grad_norm": 0.12398722767829895, "learning_rate": 5.0183534737602014e-05, "loss": 0.0079, "step": 52390 }, { "grad_norm": 0.12980332970619202, "learning_rate": 5.0167000141085686e-05, "loss": 0.0092, "step": 52400 }, { "grad_norm": 0.16220824420452118, "learning_rate": 5.015046552630647e-05, "loss": 0.0084, "step": 52410 }, { "grad_norm": 0.14559896290302277, "learning_rate": 5.013393089507258e-05, "loss": 0.0097, "step": 52420 }, { "grad_norm": 0.12550269067287445, "learning_rate": 5.011739624919219e-05, "loss": 0.0084, "step": 52430 }, { "grad_norm": 0.11921011656522751, "learning_rate": 5.0100861590473545e-05, "loss": 0.0076, "step": 52440 }, { "grad_norm": 0.121326744556427, "learning_rate": 5.008432692072482e-05, "loss": 0.0083, "step": 52450 }, { "grad_norm": 0.13873432576656342, "learning_rate": 5.006779224175423e-05, "loss": 0.0067, "step": 52460 }, { "grad_norm": 0.11032495647668839, "learning_rate": 5.005125755536998e-05, "loss": 0.0067, "step": 52470 }, { "grad_norm": 0.11739496886730194, "learning_rate": 5.003472286338029e-05, "loss": 0.006, "step": 52480 }, { "grad_norm": 0.1671290397644043, "learning_rate": 5.001818816759335e-05, "loss": 0.0079, "step": 52490 }, { "grad_norm": 0.13592487573623657, "learning_rate": 5.000165346981738e-05, "loss": 0.0075, "step": 52500 }, { "grad_norm": 0.09880144149065018, "learning_rate": 4.9985118771860595e-05, "loss": 0.0058, "step": 52510 }, { "grad_norm": 0.13338376581668854, "learning_rate": 4.996858407553119e-05, "loss": 0.0075, "step": 52520 }, { "grad_norm": 0.1268010437488556, "learning_rate": 4.99520493826374e-05, "loss": 0.0073, "step": 52530 }, { "grad_norm": 0.12451041489839554, "learning_rate": 4.9935514694987384e-05, "loss": 0.0084, "step": 52540 }, { "grad_norm": 0.12394366413354874, "learning_rate": 4.991898001438941e-05, "loss": 0.006, "step": 52550 }, { "grad_norm": 0.10370052605867386, "learning_rate": 4.990244534265166e-05, "loss": 0.0075, "step": 52560 }, { "grad_norm": 0.13994285464286804, "learning_rate": 4.988591068158234e-05, "loss": 0.0088, "step": 52570 }, { "grad_norm": 0.11846677213907242, "learning_rate": 4.986937603298965e-05, "loss": 0.0071, "step": 52580 }, { "grad_norm": 0.12037656456232071, "learning_rate": 4.985284139868181e-05, "loss": 0.0078, "step": 52590 }, { "grad_norm": 0.12196843326091766, "learning_rate": 4.9836306780467e-05, "loss": 0.0076, "step": 52600 }, { "grad_norm": 0.12719756364822388, "learning_rate": 4.981977218015343e-05, "loss": 0.0061, "step": 52610 }, { "grad_norm": 0.1610134392976761, "learning_rate": 4.980323759954932e-05, "loss": 0.0078, "step": 52620 }, { "grad_norm": 0.12119810283184052, "learning_rate": 4.978670304046284e-05, "loss": 0.0067, "step": 52630 }, { "grad_norm": 0.13552412390708923, "learning_rate": 4.9770168504702195e-05, "loss": 0.0081, "step": 52640 }, { "grad_norm": 0.1655338704586029, "learning_rate": 4.975363399407559e-05, "loss": 0.0063, "step": 52650 }, { "grad_norm": 0.14462381601333618, "learning_rate": 4.9737099510391194e-05, "loss": 0.0078, "step": 52660 }, { "grad_norm": 0.12045623362064362, "learning_rate": 4.972056505545719e-05, "loss": 0.008, "step": 52670 }, { "grad_norm": 0.10679804533720016, "learning_rate": 4.9704030631081775e-05, "loss": 0.0097, "step": 52680 }, { "grad_norm": 0.09828725457191467, "learning_rate": 4.968749623907313e-05, "loss": 0.0073, "step": 52690 }, { "grad_norm": 0.10835091024637222, "learning_rate": 4.967096188123943e-05, "loss": 0.0074, "step": 52700 }, { "grad_norm": 0.12603944540023804, "learning_rate": 4.965442755938884e-05, "loss": 0.0071, "step": 52710 }, { "grad_norm": 0.08552689850330353, "learning_rate": 4.963789327532954e-05, "loss": 0.0075, "step": 52720 }, { "grad_norm": 0.10486353933811188, "learning_rate": 4.962135903086968e-05, "loss": 0.0069, "step": 52730 }, { "grad_norm": 0.08751782774925232, "learning_rate": 4.960482482781741e-05, "loss": 0.0062, "step": 52740 }, { "grad_norm": 0.08666560798883438, "learning_rate": 4.958829066798092e-05, "loss": 0.0073, "step": 52750 }, { "grad_norm": 0.11460672318935394, "learning_rate": 4.957175655316835e-05, "loss": 0.0071, "step": 52760 }, { "grad_norm": 0.10892410576343536, "learning_rate": 4.9555222485187834e-05, "loss": 0.0063, "step": 52770 }, { "grad_norm": 0.14243806898593903, "learning_rate": 4.953868846584752e-05, "loss": 0.0069, "step": 52780 }, { "grad_norm": 0.09388865530490875, "learning_rate": 4.9522154496955544e-05, "loss": 0.0066, "step": 52790 }, { "grad_norm": 0.10119765251874924, "learning_rate": 4.950562058032001e-05, "loss": 0.0074, "step": 52800 }, { "grad_norm": 0.08509047329425812, "learning_rate": 4.948908671774909e-05, "loss": 0.0064, "step": 52810 }, { "grad_norm": 0.1274937391281128, "learning_rate": 4.947255291105087e-05, "loss": 0.0083, "step": 52820 }, { "grad_norm": 0.1729574054479599, "learning_rate": 4.945601916203348e-05, "loss": 0.007, "step": 52830 }, { "grad_norm": 0.15558607876300812, "learning_rate": 4.9439485472505016e-05, "loss": 0.0071, "step": 52840 }, { "grad_norm": 0.12485872209072113, "learning_rate": 4.942295184427356e-05, "loss": 0.0066, "step": 52850 }, { "grad_norm": 0.12880048155784607, "learning_rate": 4.9406418279147234e-05, "loss": 0.006, "step": 52860 }, { "grad_norm": 0.10871722549200058, "learning_rate": 4.9389884778934116e-05, "loss": 0.0067, "step": 52870 }, { "grad_norm": 0.11912588775157928, "learning_rate": 4.937335134544228e-05, "loss": 0.0067, "step": 52880 }, { "grad_norm": 0.1247008815407753, "learning_rate": 4.935681798047979e-05, "loss": 0.0063, "step": 52890 }, { "grad_norm": 0.12165455520153046, "learning_rate": 4.934028468585473e-05, "loss": 0.0087, "step": 52900 }, { "grad_norm": 0.1467680186033249, "learning_rate": 4.932375146337514e-05, "loss": 0.0072, "step": 52910 }, { "grad_norm": 0.12965308129787445, "learning_rate": 4.930721831484908e-05, "loss": 0.0059, "step": 52920 }, { "grad_norm": 0.11501111090183258, "learning_rate": 4.929068524208456e-05, "loss": 0.0063, "step": 52930 }, { "grad_norm": 0.12610657513141632, "learning_rate": 4.927415224688965e-05, "loss": 0.0089, "step": 52940 }, { "grad_norm": 0.10731358826160431, "learning_rate": 4.9257619331072375e-05, "loss": 0.0069, "step": 52950 }, { "grad_norm": 0.12301993370056152, "learning_rate": 4.924108649644072e-05, "loss": 0.0068, "step": 52960 }, { "grad_norm": 0.1490866243839264, "learning_rate": 4.922455374480271e-05, "loss": 0.0071, "step": 52970 }, { "grad_norm": 0.12783557176589966, "learning_rate": 4.920802107796634e-05, "loss": 0.0063, "step": 52980 }, { "grad_norm": 0.10616815090179443, "learning_rate": 4.9191488497739564e-05, "loss": 0.0068, "step": 52990 }, { "grad_norm": 0.11872578412294388, "learning_rate": 4.917495600593042e-05, "loss": 0.0063, "step": 53000 }, { "grad_norm": 0.10869342088699341, "learning_rate": 4.915842360434684e-05, "loss": 0.0065, "step": 53010 }, { "grad_norm": 0.11160720139741898, "learning_rate": 4.914189129479679e-05, "loss": 0.0065, "step": 53020 }, { "grad_norm": 0.11141690611839294, "learning_rate": 4.912535907908822e-05, "loss": 0.0071, "step": 53030 }, { "grad_norm": 0.10948404669761658, "learning_rate": 4.910882695902907e-05, "loss": 0.0099, "step": 53040 }, { "grad_norm": 0.13637834787368774, "learning_rate": 4.9092294936427234e-05, "loss": 0.0075, "step": 53050 }, { "grad_norm": 0.142482727766037, "learning_rate": 4.9075763013090685e-05, "loss": 0.0076, "step": 53060 }, { "grad_norm": 0.10551179945468903, "learning_rate": 4.9059231190827305e-05, "loss": 0.0068, "step": 53070 }, { "grad_norm": 0.1329972892999649, "learning_rate": 4.904269947144499e-05, "loss": 0.008, "step": 53080 }, { "grad_norm": 0.13407038152217865, "learning_rate": 4.902616785675161e-05, "loss": 0.0059, "step": 53090 }, { "grad_norm": 0.10628270357847214, "learning_rate": 4.900963634855505e-05, "loss": 0.0061, "step": 53100 }, { "grad_norm": 0.09800257533788681, "learning_rate": 4.8993104948663173e-05, "loss": 0.0059, "step": 53110 }, { "grad_norm": 0.09112841635942459, "learning_rate": 4.8976573658883823e-05, "loss": 0.0088, "step": 53120 }, { "grad_norm": 0.1174067035317421, "learning_rate": 4.896004248102484e-05, "loss": 0.0072, "step": 53130 }, { "grad_norm": 0.09313046187162399, "learning_rate": 4.8943511416894045e-05, "loss": 0.0064, "step": 53140 }, { "grad_norm": 0.09695074707269669, "learning_rate": 4.892698046829926e-05, "loss": 0.0048, "step": 53150 }, { "grad_norm": 0.10024099797010422, "learning_rate": 4.8910449637048275e-05, "loss": 0.0053, "step": 53160 }, { "grad_norm": 0.09704606980085373, "learning_rate": 4.8893918924948884e-05, "loss": 0.0055, "step": 53170 }, { "grad_norm": 0.10783886164426804, "learning_rate": 4.8877388333808835e-05, "loss": 0.0076, "step": 53180 }, { "grad_norm": 0.11636677384376526, "learning_rate": 4.8860857865435925e-05, "loss": 0.0062, "step": 53190 }, { "grad_norm": 0.12337029725313187, "learning_rate": 4.884432752163789e-05, "loss": 0.0078, "step": 53200 }, { "grad_norm": 0.13670480251312256, "learning_rate": 4.8827797304222466e-05, "loss": 0.0068, "step": 53210 }, { "grad_norm": 0.11913454532623291, "learning_rate": 4.8811267214997366e-05, "loss": 0.0065, "step": 53220 }, { "grad_norm": 0.11421020328998566, "learning_rate": 4.879473725577029e-05, "loss": 0.0067, "step": 53230 }, { "grad_norm": 0.10812385380268097, "learning_rate": 4.877820742834891e-05, "loss": 0.0078, "step": 53240 }, { "grad_norm": 0.1289556473493576, "learning_rate": 4.8761677734540956e-05, "loss": 0.0069, "step": 53250 }, { "grad_norm": 0.1368246078491211, "learning_rate": 4.874514817615407e-05, "loss": 0.0067, "step": 53260 }, { "grad_norm": 0.17054328322410583, "learning_rate": 4.872861875499588e-05, "loss": 0.0085, "step": 53270 }, { "grad_norm": 0.12213419377803802, "learning_rate": 4.871208947287404e-05, "loss": 0.0058, "step": 53280 }, { "grad_norm": 0.1295585334300995, "learning_rate": 4.869556033159615e-05, "loss": 0.0072, "step": 53290 }, { "grad_norm": 0.17005224525928497, "learning_rate": 4.8679031332969835e-05, "loss": 0.0066, "step": 53300 }, { "grad_norm": 0.12959155440330505, "learning_rate": 4.8662502478802635e-05, "loss": 0.0069, "step": 53310 }, { "grad_norm": 0.11220602691173553, "learning_rate": 4.8645973770902176e-05, "loss": 0.0106, "step": 53320 }, { "grad_norm": 0.10517818480730057, "learning_rate": 4.862944521107599e-05, "loss": 0.0067, "step": 53330 }, { "grad_norm": 0.10040471702814102, "learning_rate": 4.861291680113162e-05, "loss": 0.0066, "step": 53340 }, { "grad_norm": 0.14003823697566986, "learning_rate": 4.8596388542876563e-05, "loss": 0.0071, "step": 53350 }, { "grad_norm": 0.14688637852668762, "learning_rate": 4.857986043811836e-05, "loss": 0.0086, "step": 53360 }, { "grad_norm": 0.10607142001390457, "learning_rate": 4.856333248866448e-05, "loss": 0.0068, "step": 53370 }, { "grad_norm": 0.1107422262430191, "learning_rate": 4.85468046963224e-05, "loss": 0.0056, "step": 53380 }, { "grad_norm": 0.1330953687429428, "learning_rate": 4.853027706289957e-05, "loss": 0.0066, "step": 53390 }, { "grad_norm": 0.11899375170469284, "learning_rate": 4.851374959020344e-05, "loss": 0.0065, "step": 53400 }, { "grad_norm": 0.10299453139305115, "learning_rate": 4.849722228004141e-05, "loss": 0.006, "step": 53410 }, { "grad_norm": 0.11782743781805038, "learning_rate": 4.848069513422091e-05, "loss": 0.007, "step": 53420 }, { "grad_norm": 0.10637117922306061, "learning_rate": 4.8464168154549294e-05, "loss": 0.0061, "step": 53430 }, { "grad_norm": 0.13352756202220917, "learning_rate": 4.844764134283392e-05, "loss": 0.0064, "step": 53440 }, { "grad_norm": 0.1533103883266449, "learning_rate": 4.843111470088218e-05, "loss": 0.0066, "step": 53450 }, { "grad_norm": 0.1137852892279625, "learning_rate": 4.841458823050137e-05, "loss": 0.0057, "step": 53460 }, { "grad_norm": 0.11258727312088013, "learning_rate": 4.8398061933498816e-05, "loss": 0.0079, "step": 53470 }, { "grad_norm": 0.11270162463188171, "learning_rate": 4.83815358116818e-05, "loss": 0.0064, "step": 53480 }, { "grad_norm": 0.10664818435907364, "learning_rate": 4.8365009866857584e-05, "loss": 0.006, "step": 53490 }, { "grad_norm": 0.10935094207525253, "learning_rate": 4.834848410083342e-05, "loss": 0.0073, "step": 53500 }, { "grad_norm": 0.10581173002719879, "learning_rate": 4.833195851541657e-05, "loss": 0.0069, "step": 53510 }, { "grad_norm": 0.15596918761730194, "learning_rate": 4.831543311241423e-05, "loss": 0.0083, "step": 53520 }, { "grad_norm": 0.08930517733097076, "learning_rate": 4.829890789363359e-05, "loss": 0.006, "step": 53530 }, { "grad_norm": 0.12569786608219147, "learning_rate": 4.828238286088182e-05, "loss": 0.0067, "step": 53540 }, { "grad_norm": 0.10805946588516235, "learning_rate": 4.826585801596609e-05, "loss": 0.0068, "step": 53550 }, { "grad_norm": 0.12961769104003906, "learning_rate": 4.8249333360693486e-05, "loss": 0.0073, "step": 53560 }, { "grad_norm": 0.100100576877594, "learning_rate": 4.8232808896871185e-05, "loss": 0.0068, "step": 53570 }, { "grad_norm": 0.11059994995594025, "learning_rate": 4.821628462630624e-05, "loss": 0.0065, "step": 53580 }, { "grad_norm": 0.07419499009847641, "learning_rate": 4.8199760550805735e-05, "loss": 0.0063, "step": 53590 }, { "grad_norm": 0.10547150671482086, "learning_rate": 4.818323667217669e-05, "loss": 0.0061, "step": 53600 }, { "grad_norm": 0.12237140536308289, "learning_rate": 4.8166712992226165e-05, "loss": 0.0075, "step": 53610 }, { "grad_norm": 0.10133945941925049, "learning_rate": 4.8150189512761154e-05, "loss": 0.0067, "step": 53620 }, { "grad_norm": 0.12291896343231201, "learning_rate": 4.813366623558862e-05, "loss": 0.0069, "step": 53630 }, { "grad_norm": 0.11136201024055481, "learning_rate": 4.811714316251554e-05, "loss": 0.0062, "step": 53640 }, { "grad_norm": 0.08564203232526779, "learning_rate": 4.810062029534886e-05, "loss": 0.006, "step": 53650 }, { "grad_norm": 0.09893742948770523, "learning_rate": 4.8084097635895505e-05, "loss": 0.0075, "step": 53660 }, { "grad_norm": 0.1965431421995163, "learning_rate": 4.8067575185962335e-05, "loss": 0.0064, "step": 53670 }, { "grad_norm": 0.09970777481794357, "learning_rate": 4.805105294735625e-05, "loss": 0.0084, "step": 53680 }, { "grad_norm": 0.09476940333843231, "learning_rate": 4.803453092188406e-05, "loss": 0.0068, "step": 53690 }, { "grad_norm": 0.09390272945165634, "learning_rate": 4.801800911135263e-05, "loss": 0.0065, "step": 53700 }, { "grad_norm": 0.11562617868185043, "learning_rate": 4.8001487517568744e-05, "loss": 0.0072, "step": 53710 }, { "grad_norm": 0.16557660698890686, "learning_rate": 4.798496614233918e-05, "loss": 0.008, "step": 53720 }, { "grad_norm": 0.13018539547920227, "learning_rate": 4.796844498747069e-05, "loss": 0.008, "step": 53730 }, { "grad_norm": 0.11879299581050873, "learning_rate": 4.795192405477e-05, "loss": 0.0072, "step": 53740 }, { "grad_norm": 0.08004464954137802, "learning_rate": 4.79354033460438e-05, "loss": 0.0066, "step": 53750 }, { "grad_norm": 0.1619029939174652, "learning_rate": 4.79188828630988e-05, "loss": 0.0078, "step": 53760 }, { "grad_norm": 0.11981730163097382, "learning_rate": 4.790236260774165e-05, "loss": 0.0081, "step": 53770 }, { "grad_norm": 0.11902585625648499, "learning_rate": 4.788584258177896e-05, "loss": 0.0064, "step": 53780 }, { "grad_norm": 0.1020059883594513, "learning_rate": 4.7869322787017356e-05, "loss": 0.0073, "step": 53790 }, { "grad_norm": 0.10283257067203522, "learning_rate": 4.78528032252634e-05, "loss": 0.0067, "step": 53800 }, { "grad_norm": 0.15456919372081757, "learning_rate": 4.783628389832366e-05, "loss": 0.0074, "step": 53810 }, { "grad_norm": 0.13239671289920807, "learning_rate": 4.7819764808004644e-05, "loss": 0.0071, "step": 53820 }, { "grad_norm": 0.15426908433437347, "learning_rate": 4.780324595611289e-05, "loss": 0.0077, "step": 53830 }, { "grad_norm": 0.1068490669131279, "learning_rate": 4.7786727344454854e-05, "loss": 0.0066, "step": 53840 }, { "grad_norm": 0.10064338892698288, "learning_rate": 4.777020897483697e-05, "loss": 0.0077, "step": 53850 }, { "grad_norm": 0.1131686344742775, "learning_rate": 4.775369084906569e-05, "loss": 0.0064, "step": 53860 }, { "grad_norm": 0.11480990052223206, "learning_rate": 4.773717296894742e-05, "loss": 0.0083, "step": 53870 }, { "grad_norm": 0.16431207954883575, "learning_rate": 4.772065533628848e-05, "loss": 0.0068, "step": 53880 }, { "grad_norm": 0.13935740292072296, "learning_rate": 4.7704137952895265e-05, "loss": 0.0076, "step": 53890 }, { "grad_norm": 0.1347801685333252, "learning_rate": 4.768762082057407e-05, "loss": 0.0075, "step": 53900 }, { "grad_norm": 0.10817495733499527, "learning_rate": 4.7671103941131186e-05, "loss": 0.0075, "step": 53910 }, { "grad_norm": 0.12181304395198822, "learning_rate": 4.765458731637287e-05, "loss": 0.0082, "step": 53920 }, { "grad_norm": 0.14113859832286835, "learning_rate": 4.763807094810537e-05, "loss": 0.0073, "step": 53930 }, { "grad_norm": 0.1083519235253334, "learning_rate": 4.762155483813487e-05, "loss": 0.0075, "step": 53940 }, { "grad_norm": 0.1468496024608612, "learning_rate": 4.760503898826754e-05, "loss": 0.0068, "step": 53950 }, { "grad_norm": 0.10419291257858276, "learning_rate": 4.758852340030957e-05, "loss": 0.0068, "step": 53960 }, { "grad_norm": 0.16679631173610687, "learning_rate": 4.757200807606705e-05, "loss": 0.0093, "step": 53970 }, { "grad_norm": 0.0968371257185936, "learning_rate": 4.755549301734607e-05, "loss": 0.0076, "step": 53980 }, { "grad_norm": 0.12192840129137039, "learning_rate": 4.753897822595271e-05, "loss": 0.0067, "step": 53990 }, { "grad_norm": 0.09896842390298843, "learning_rate": 4.7522463703692994e-05, "loss": 0.0072, "step": 54000 }, { "grad_norm": 0.14316323399543762, "learning_rate": 4.75059494523729e-05, "loss": 0.0077, "step": 54010 }, { "grad_norm": 0.09702915698289871, "learning_rate": 4.748943547379843e-05, "loss": 0.0081, "step": 54020 }, { "grad_norm": 0.13531526923179626, "learning_rate": 4.7472921769775535e-05, "loss": 0.0092, "step": 54030 }, { "grad_norm": 0.15632028877735138, "learning_rate": 4.745640834211011e-05, "loss": 0.0076, "step": 54040 }, { "grad_norm": 0.09241635352373123, "learning_rate": 4.7439895192608054e-05, "loss": 0.0064, "step": 54050 }, { "grad_norm": 0.10629532486200333, "learning_rate": 4.7423382323075186e-05, "loss": 0.0074, "step": 54060 }, { "grad_norm": 0.0890287384390831, "learning_rate": 4.740686973531737e-05, "loss": 0.0071, "step": 54070 }, { "grad_norm": 0.1406719833612442, "learning_rate": 4.7390357431140375e-05, "loss": 0.0069, "step": 54080 }, { "grad_norm": 0.08650407940149307, "learning_rate": 4.737384541234998e-05, "loss": 0.0081, "step": 54090 }, { "grad_norm": 0.11718206852674484, "learning_rate": 4.735733368075189e-05, "loss": 0.0067, "step": 54100 }, { "grad_norm": 0.12065975368022919, "learning_rate": 4.734082223815182e-05, "loss": 0.0069, "step": 54110 }, { "grad_norm": 0.11799885332584381, "learning_rate": 4.7324311086355436e-05, "loss": 0.0071, "step": 54120 }, { "grad_norm": 0.09515857696533203, "learning_rate": 4.730780022716836e-05, "loss": 0.0072, "step": 54130 }, { "grad_norm": 0.08469264954328537, "learning_rate": 4.729128966239619e-05, "loss": 0.0064, "step": 54140 }, { "grad_norm": 0.09930417686700821, "learning_rate": 4.7274779393844536e-05, "loss": 0.0067, "step": 54150 }, { "grad_norm": 0.12665031850337982, "learning_rate": 4.725826942331891e-05, "loss": 0.0059, "step": 54160 }, { "grad_norm": 0.12579137086868286, "learning_rate": 4.724175975262481e-05, "loss": 0.0064, "step": 54170 }, { "grad_norm": 0.12215418368577957, "learning_rate": 4.722525038356772e-05, "loss": 0.0058, "step": 54180 }, { "grad_norm": 0.09953506290912628, "learning_rate": 4.7208741317953075e-05, "loss": 0.0053, "step": 54190 }, { "grad_norm": 0.12361528724431992, "learning_rate": 4.7192232557586264e-05, "loss": 0.0073, "step": 54200 }, { "grad_norm": 0.10514666140079498, "learning_rate": 4.717572410427269e-05, "loss": 0.0053, "step": 54210 }, { "grad_norm": 0.0748090147972107, "learning_rate": 4.71592159598177e-05, "loss": 0.0066, "step": 54220 }, { "grad_norm": 0.1352488249540329, "learning_rate": 4.714270812602657e-05, "loss": 0.0058, "step": 54230 }, { "grad_norm": 0.13462130725383759, "learning_rate": 4.712620060470458e-05, "loss": 0.0072, "step": 54240 }, { "grad_norm": 0.10260847210884094, "learning_rate": 4.7109693397656986e-05, "loss": 0.0078, "step": 54250 }, { "grad_norm": 0.11430685967206955, "learning_rate": 4.709318650668894e-05, "loss": 0.0067, "step": 54260 }, { "grad_norm": 0.10899457335472107, "learning_rate": 4.7076679933605676e-05, "loss": 0.0063, "step": 54270 }, { "grad_norm": 0.1226300299167633, "learning_rate": 4.70601736802123e-05, "loss": 0.0068, "step": 54280 }, { "grad_norm": 0.15451939404010773, "learning_rate": 4.704366774831391e-05, "loss": 0.0076, "step": 54290 }, { "grad_norm": 0.1398211419582367, "learning_rate": 4.702716213971557e-05, "loss": 0.0076, "step": 54300 }, { "grad_norm": 0.10245462507009506, "learning_rate": 4.70106568562223e-05, "loss": 0.0054, "step": 54310 }, { "grad_norm": 0.10156078636646271, "learning_rate": 4.6994151899639105e-05, "loss": 0.0075, "step": 54320 }, { "grad_norm": 0.11903741210699081, "learning_rate": 4.697764727177093e-05, "loss": 0.007, "step": 54330 }, { "grad_norm": 0.12242664396762848, "learning_rate": 4.696114297442272e-05, "loss": 0.0077, "step": 54340 }, { "grad_norm": 0.12165627628564835, "learning_rate": 4.694463900939933e-05, "loss": 0.0081, "step": 54350 }, { "grad_norm": 0.11163878440856934, "learning_rate": 4.692813537850564e-05, "loss": 0.0066, "step": 54360 }, { "grad_norm": 0.08696279674768448, "learning_rate": 4.6911632083546445e-05, "loss": 0.0068, "step": 54370 }, { "grad_norm": 0.10187064856290817, "learning_rate": 4.6895129126326525e-05, "loss": 0.0056, "step": 54380 }, { "grad_norm": 0.11072559654712677, "learning_rate": 4.687862650865059e-05, "loss": 0.0064, "step": 54390 }, { "grad_norm": 0.1169719398021698, "learning_rate": 4.686212423232339e-05, "loss": 0.007, "step": 54400 }, { "grad_norm": 0.11286597698926926, "learning_rate": 4.6845622299149565e-05, "loss": 0.0057, "step": 54410 }, { "grad_norm": 0.13909317553043365, "learning_rate": 4.682912071093374e-05, "loss": 0.0076, "step": 54420 }, { "grad_norm": 0.12217023223638535, "learning_rate": 4.681261946948052e-05, "loss": 0.0064, "step": 54430 }, { "grad_norm": 0.1397188901901245, "learning_rate": 4.679611857659443e-05, "loss": 0.0067, "step": 54440 }, { "grad_norm": 0.12149283289909363, "learning_rate": 4.6779618034079975e-05, "loss": 0.0073, "step": 54450 }, { "grad_norm": 0.16876095533370972, "learning_rate": 4.6763117843741675e-05, "loss": 0.0076, "step": 54460 }, { "grad_norm": 0.15527410805225372, "learning_rate": 4.674661800738394e-05, "loss": 0.0087, "step": 54470 }, { "grad_norm": 0.08889617770910263, "learning_rate": 4.673011852681117e-05, "loss": 0.0079, "step": 54480 }, { "grad_norm": 0.15325917303562164, "learning_rate": 4.671361940382772e-05, "loss": 0.0078, "step": 54490 }, { "grad_norm": 0.13798534870147705, "learning_rate": 4.669712064023791e-05, "loss": 0.0071, "step": 54500 }, { "grad_norm": 0.11924093216657639, "learning_rate": 4.668062223784602e-05, "loss": 0.0083, "step": 54510 }, { "grad_norm": 0.08550718426704407, "learning_rate": 4.6664124198456274e-05, "loss": 0.0067, "step": 54520 }, { "grad_norm": 0.08064242452383041, "learning_rate": 4.6647626523872905e-05, "loss": 0.0062, "step": 54530 }, { "grad_norm": 0.1068728044629097, "learning_rate": 4.6631129215900066e-05, "loss": 0.0064, "step": 54540 }, { "grad_norm": 0.1323966532945633, "learning_rate": 4.661463227634187e-05, "loss": 0.0067, "step": 54550 }, { "grad_norm": 0.11800637096166611, "learning_rate": 4.659813570700238e-05, "loss": 0.0072, "step": 54560 }, { "grad_norm": 0.1880008578300476, "learning_rate": 4.658163950968568e-05, "loss": 0.008, "step": 54570 }, { "grad_norm": 0.12501221895217896, "learning_rate": 4.6565143686195726e-05, "loss": 0.0064, "step": 54580 }, { "grad_norm": 0.09739259630441666, "learning_rate": 4.654864823833649e-05, "loss": 0.0058, "step": 54590 }, { "grad_norm": 0.11593911051750183, "learning_rate": 4.65321531679119e-05, "loss": 0.0072, "step": 54600 }, { "grad_norm": 0.09991506487131119, "learning_rate": 4.6515658476725834e-05, "loss": 0.0065, "step": 54610 }, { "grad_norm": 0.08772820234298706, "learning_rate": 4.649916416658212e-05, "loss": 0.0083, "step": 54620 }, { "grad_norm": 0.10244404524564743, "learning_rate": 4.648267023928455e-05, "loss": 0.0059, "step": 54630 }, { "grad_norm": 0.12817472219467163, "learning_rate": 4.646617669663687e-05, "loss": 0.0076, "step": 54640 }, { "grad_norm": 0.14704585075378418, "learning_rate": 4.644968354044277e-05, "loss": 0.0071, "step": 54650 }, { "grad_norm": 0.1316150724887848, "learning_rate": 4.643319077250596e-05, "loss": 0.0061, "step": 54660 }, { "grad_norm": 0.1298944354057312, "learning_rate": 4.6416698394630055e-05, "loss": 0.0067, "step": 54670 }, { "grad_norm": 0.1114848405122757, "learning_rate": 4.640020640861862e-05, "loss": 0.0076, "step": 54680 }, { "grad_norm": 0.1409515142440796, "learning_rate": 4.6383714816275195e-05, "loss": 0.0057, "step": 54690 }, { "grad_norm": 0.13041578233242035, "learning_rate": 4.636722361940329e-05, "loss": 0.0065, "step": 54700 }, { "grad_norm": 0.1170765608549118, "learning_rate": 4.635073281980632e-05, "loss": 0.0072, "step": 54710 }, { "grad_norm": 0.11178386211395264, "learning_rate": 4.633424241928775e-05, "loss": 0.0054, "step": 54720 }, { "grad_norm": 0.11073552817106247, "learning_rate": 4.6317752419650914e-05, "loss": 0.0085, "step": 54730 }, { "grad_norm": 0.10663538426160812, "learning_rate": 4.630126282269914e-05, "loss": 0.0057, "step": 54740 }, { "grad_norm": 0.1258409470319748, "learning_rate": 4.628477363023569e-05, "loss": 0.0088, "step": 54750 }, { "grad_norm": 0.11163073778152466, "learning_rate": 4.626828484406383e-05, "loss": 0.0069, "step": 54760 }, { "grad_norm": 0.11263561248779297, "learning_rate": 4.625179646598669e-05, "loss": 0.0073, "step": 54770 }, { "grad_norm": 0.14635765552520752, "learning_rate": 4.6235308497807487e-05, "loss": 0.0073, "step": 54780 }, { "grad_norm": 0.1629319190979004, "learning_rate": 4.6218820941329276e-05, "loss": 0.0061, "step": 54790 }, { "grad_norm": 0.08096757531166077, "learning_rate": 4.620233379835513e-05, "loss": 0.0061, "step": 54800 }, { "grad_norm": 0.09497646242380142, "learning_rate": 4.618584707068804e-05, "loss": 0.0069, "step": 54810 }, { "grad_norm": 0.1419491171836853, "learning_rate": 4.616936076013099e-05, "loss": 0.0077, "step": 54820 }, { "grad_norm": 0.11493430286645889, "learning_rate": 4.615287486848688e-05, "loss": 0.007, "step": 54830 }, { "grad_norm": 0.12311230599880219, "learning_rate": 4.6136389397558584e-05, "loss": 0.008, "step": 54840 }, { "grad_norm": 0.0997779592871666, "learning_rate": 4.6119904349148935e-05, "loss": 0.0067, "step": 54850 }, { "grad_norm": 0.12646611034870148, "learning_rate": 4.6103419725060726e-05, "loss": 0.0058, "step": 54860 }, { "grad_norm": 0.16815225780010223, "learning_rate": 4.608693552709667e-05, "loss": 0.0067, "step": 54870 }, { "grad_norm": 0.11414208263158798, "learning_rate": 4.607045175705947e-05, "loss": 0.0077, "step": 54880 }, { "grad_norm": 0.1269253045320511, "learning_rate": 4.605396841675176e-05, "loss": 0.0087, "step": 54890 }, { "grad_norm": 0.11684801429510117, "learning_rate": 4.6037485507976094e-05, "loss": 0.008, "step": 54900 }, { "grad_norm": 0.0882311686873436, "learning_rate": 4.602100303253509e-05, "loss": 0.0065, "step": 54910 }, { "grad_norm": 0.10679232329130173, "learning_rate": 4.600452099223122e-05, "loss": 0.008, "step": 54920 }, { "grad_norm": 0.10157350450754166, "learning_rate": 4.598803938886691e-05, "loss": 0.0063, "step": 54930 }, { "grad_norm": 0.10212410241365433, "learning_rate": 4.597155822424459e-05, "loss": 0.0058, "step": 54940 }, { "grad_norm": 0.1271563321352005, "learning_rate": 4.59550775001666e-05, "loss": 0.0091, "step": 54950 }, { "grad_norm": 0.13534961640834808, "learning_rate": 4.5938597218435234e-05, "loss": 0.0066, "step": 54960 }, { "grad_norm": 0.10985907167196274, "learning_rate": 4.592211738085279e-05, "loss": 0.0059, "step": 54970 }, { "grad_norm": 0.08678975701332092, "learning_rate": 4.590563798922146e-05, "loss": 0.0062, "step": 54980 }, { "grad_norm": 0.0735415518283844, "learning_rate": 4.5889159045343404e-05, "loss": 0.0068, "step": 54990 }, { "grad_norm": 0.12076045572757721, "learning_rate": 4.5872680551020734e-05, "loss": 0.0071, "step": 55000 }, { "grad_norm": 0.09909190982580185, "learning_rate": 4.585620250805552e-05, "loss": 0.0062, "step": 55010 }, { "grad_norm": 0.10208762437105179, "learning_rate": 4.583972491824975e-05, "loss": 0.008, "step": 55020 }, { "grad_norm": 0.1135522723197937, "learning_rate": 4.582324778340541e-05, "loss": 0.0073, "step": 55030 }, { "grad_norm": 0.14113089442253113, "learning_rate": 4.5806771105324424e-05, "loss": 0.008, "step": 55040 }, { "grad_norm": 0.11995197087526321, "learning_rate": 4.579029488580865e-05, "loss": 0.0063, "step": 55050 }, { "grad_norm": 0.0868738666176796, "learning_rate": 4.5773819126659876e-05, "loss": 0.006, "step": 55060 }, { "grad_norm": 0.1331024020910263, "learning_rate": 4.575734382967991e-05, "loss": 0.0063, "step": 55070 }, { "grad_norm": 0.07806427031755447, "learning_rate": 4.574086899667044e-05, "loss": 0.0072, "step": 55080 }, { "grad_norm": 0.1037714034318924, "learning_rate": 4.5724394629433106e-05, "loss": 0.0075, "step": 55090 }, { "grad_norm": 0.09084364026784897, "learning_rate": 4.570792072976958e-05, "loss": 0.0059, "step": 55100 }, { "grad_norm": 0.1371610462665558, "learning_rate": 4.569144729948138e-05, "loss": 0.0062, "step": 55110 }, { "grad_norm": 0.11816564202308655, "learning_rate": 4.5674974340370033e-05, "loss": 0.0083, "step": 55120 }, { "grad_norm": 0.09001524746417999, "learning_rate": 4.565850185423699e-05, "loss": 0.0061, "step": 55130 }, { "grad_norm": 0.10970395058393478, "learning_rate": 4.564202984288365e-05, "loss": 0.0068, "step": 55140 }, { "grad_norm": 0.12041973322629929, "learning_rate": 4.5625558308111356e-05, "loss": 0.0083, "step": 55150 }, { "grad_norm": 0.11823595315217972, "learning_rate": 4.560908725172145e-05, "loss": 0.006, "step": 55160 }, { "grad_norm": 0.11636026948690414, "learning_rate": 4.559261667551516e-05, "loss": 0.0073, "step": 55170 }, { "grad_norm": 0.09964968264102936, "learning_rate": 4.5576146581293685e-05, "loss": 0.0071, "step": 55180 }, { "grad_norm": 0.12584809958934784, "learning_rate": 4.5559676970858164e-05, "loss": 0.0086, "step": 55190 }, { "grad_norm": 0.15389196574687958, "learning_rate": 4.5543207846009705e-05, "loss": 0.0077, "step": 55200 }, { "grad_norm": 0.10310234129428864, "learning_rate": 4.552673920854933e-05, "loss": 0.0055, "step": 55210 }, { "grad_norm": 0.1217404305934906, "learning_rate": 4.5510271060278006e-05, "loss": 0.007, "step": 55220 }, { "grad_norm": 0.09334185719490051, "learning_rate": 4.5493803402996704e-05, "loss": 0.0068, "step": 55230 }, { "grad_norm": 0.09250258654356003, "learning_rate": 4.5477336238506295e-05, "loss": 0.0055, "step": 55240 }, { "grad_norm": 0.12272397428750992, "learning_rate": 4.546086956860759e-05, "loss": 0.0071, "step": 55250 }, { "grad_norm": 0.11932175606489182, "learning_rate": 4.5444403395101374e-05, "loss": 0.0069, "step": 55260 }, { "grad_norm": 0.10720393806695938, "learning_rate": 4.542793771978834e-05, "loss": 0.0057, "step": 55270 }, { "grad_norm": 0.11125679314136505, "learning_rate": 4.541147254446916e-05, "loss": 0.0082, "step": 55280 }, { "grad_norm": 0.138787642121315, "learning_rate": 4.539500787094445e-05, "loss": 0.0069, "step": 55290 }, { "grad_norm": 0.11802167445421219, "learning_rate": 4.537854370101475e-05, "loss": 0.0076, "step": 55300 }, { "grad_norm": 0.1543211191892624, "learning_rate": 4.5362080036480566e-05, "loss": 0.0082, "step": 55310 }, { "grad_norm": 0.11658447235822678, "learning_rate": 4.534561687914235e-05, "loss": 0.0063, "step": 55320 }, { "grad_norm": 0.1045459732413292, "learning_rate": 4.5329154230800466e-05, "loss": 0.0072, "step": 55330 }, { "grad_norm": 0.08828118443489075, "learning_rate": 4.531269209325525e-05, "loss": 0.0058, "step": 55340 }, { "grad_norm": 0.0941946730017662, "learning_rate": 4.529623046830696e-05, "loss": 0.0073, "step": 55350 }, { "grad_norm": 0.11982325464487076, "learning_rate": 4.5279769357755866e-05, "loss": 0.0061, "step": 55360 }, { "grad_norm": 0.13693466782569885, "learning_rate": 4.5263308763402084e-05, "loss": 0.007, "step": 55370 }, { "grad_norm": 0.07674757391214371, "learning_rate": 4.524684868704574e-05, "loss": 0.006, "step": 55380 }, { "grad_norm": 0.08577580004930496, "learning_rate": 4.523038913048688e-05, "loss": 0.0075, "step": 55390 }, { "grad_norm": 0.13844379782676697, "learning_rate": 4.521393009552549e-05, "loss": 0.0074, "step": 55400 }, { "grad_norm": 0.1102699562907219, "learning_rate": 4.5197471583961484e-05, "loss": 0.0078, "step": 55410 }, { "grad_norm": 0.14137014746665955, "learning_rate": 4.51810135975948e-05, "loss": 0.0065, "step": 55420 }, { "grad_norm": 0.10131020843982697, "learning_rate": 4.516455613822521e-05, "loss": 0.0059, "step": 55430 }, { "grad_norm": 0.07249718904495239, "learning_rate": 4.514809920765249e-05, "loss": 0.0052, "step": 55440 }, { "grad_norm": 0.12331831455230713, "learning_rate": 4.5131642807676356e-05, "loss": 0.006, "step": 55450 }, { "grad_norm": 0.10360369831323624, "learning_rate": 4.511518694009644e-05, "loss": 0.0058, "step": 55460 }, { "grad_norm": 0.11940261721611023, "learning_rate": 4.509873160671231e-05, "loss": 0.0069, "step": 55470 }, { "grad_norm": 0.08965547382831573, "learning_rate": 4.508227680932356e-05, "loss": 0.0054, "step": 55480 }, { "grad_norm": 0.0854005366563797, "learning_rate": 4.5065822549729605e-05, "loss": 0.007, "step": 55490 }, { "grad_norm": 0.10793864727020264, "learning_rate": 4.504936882972989e-05, "loss": 0.0082, "step": 55500 }, { "grad_norm": 0.11095478385686874, "learning_rate": 4.503291565112374e-05, "loss": 0.0061, "step": 55510 }, { "grad_norm": 0.1090691015124321, "learning_rate": 4.501646301571048e-05, "loss": 0.0066, "step": 55520 }, { "grad_norm": 0.12079361826181412, "learning_rate": 4.500001092528933e-05, "loss": 0.0057, "step": 55530 }, { "grad_norm": 0.14177972078323364, "learning_rate": 4.4983559381659455e-05, "loss": 0.0061, "step": 55540 }, { "grad_norm": 0.11131097376346588, "learning_rate": 4.496710838661999e-05, "loss": 0.0064, "step": 55550 }, { "grad_norm": 0.12280488759279251, "learning_rate": 4.495065794196999e-05, "loss": 0.0059, "step": 55560 }, { "grad_norm": 0.13112899661064148, "learning_rate": 4.4934208049508455e-05, "loss": 0.0083, "step": 55570 }, { "grad_norm": 0.11285070329904556, "learning_rate": 4.4917758711034303e-05, "loss": 0.0069, "step": 55580 }, { "grad_norm": 0.09529121220111847, "learning_rate": 4.490130992834642e-05, "loss": 0.0074, "step": 55590 }, { "grad_norm": 0.18692737817764282, "learning_rate": 4.488486170324359e-05, "loss": 0.0073, "step": 55600 }, { "grad_norm": 0.10650507360696793, "learning_rate": 4.486841403752463e-05, "loss": 0.0084, "step": 55610 }, { "grad_norm": 0.1405324935913086, "learning_rate": 4.4851966932988174e-05, "loss": 0.0076, "step": 55620 }, { "grad_norm": 0.10742930322885513, "learning_rate": 4.4835520391432894e-05, "loss": 0.0074, "step": 55630 }, { "grad_norm": 0.13724209368228912, "learning_rate": 4.481907441465733e-05, "loss": 0.0066, "step": 55640 }, { "grad_norm": 0.1324542611837387, "learning_rate": 4.480262900446e-05, "loss": 0.0067, "step": 55650 }, { "grad_norm": 0.09658858180046082, "learning_rate": 4.478618416263933e-05, "loss": 0.006, "step": 55660 }, { "grad_norm": 0.09337836503982544, "learning_rate": 4.476973989099373e-05, "loss": 0.0083, "step": 55670 }, { "grad_norm": 0.1256222426891327, "learning_rate": 4.475329619132153e-05, "loss": 0.0071, "step": 55680 }, { "grad_norm": 0.09859729558229446, "learning_rate": 4.4736853065420956e-05, "loss": 0.0064, "step": 55690 }, { "grad_norm": 0.10932765156030655, "learning_rate": 4.472041051509023e-05, "loss": 0.0072, "step": 55700 }, { "grad_norm": 0.1108761578798294, "learning_rate": 4.470396854212746e-05, "loss": 0.0066, "step": 55710 }, { "grad_norm": 0.1278061717748642, "learning_rate": 4.468752714833072e-05, "loss": 0.0055, "step": 55720 }, { "grad_norm": 0.09174669533967972, "learning_rate": 4.467108633549803e-05, "loss": 0.0083, "step": 55730 }, { "grad_norm": 0.1415708363056183, "learning_rate": 4.4654646105427335e-05, "loss": 0.0074, "step": 55740 }, { "grad_norm": 0.1193116158246994, "learning_rate": 4.463820645991651e-05, "loss": 0.0066, "step": 55750 }, { "grad_norm": 0.11700417846441269, "learning_rate": 4.462176740076334e-05, "loss": 0.0075, "step": 55760 }, { "grad_norm": 0.21230600774288177, "learning_rate": 4.4605328929765624e-05, "loss": 0.0066, "step": 55770 }, { "grad_norm": 0.12813255190849304, "learning_rate": 4.458889104872102e-05, "loss": 0.0066, "step": 55780 }, { "grad_norm": 0.1173926442861557, "learning_rate": 4.457245375942715e-05, "loss": 0.0069, "step": 55790 }, { "grad_norm": 0.09517871588468552, "learning_rate": 4.455601706368157e-05, "loss": 0.0067, "step": 55800 }, { "grad_norm": 0.11226703971624374, "learning_rate": 4.453958096328179e-05, "loss": 0.0054, "step": 55810 }, { "grad_norm": 0.130946084856987, "learning_rate": 4.452314546002523e-05, "loss": 0.007, "step": 55820 }, { "grad_norm": 0.1222633346915245, "learning_rate": 4.4506710555709246e-05, "loss": 0.0077, "step": 55830 }, { "grad_norm": 0.10857568681240082, "learning_rate": 4.449027625213114e-05, "loss": 0.0065, "step": 55840 }, { "grad_norm": 0.11199848353862762, "learning_rate": 4.4473842551088135e-05, "loss": 0.0062, "step": 55850 }, { "grad_norm": 0.13203030824661255, "learning_rate": 4.445740945437738e-05, "loss": 0.0077, "step": 55860 }, { "grad_norm": 0.12213315814733505, "learning_rate": 4.444097696379602e-05, "loss": 0.0064, "step": 55870 }, { "grad_norm": 0.10997752845287323, "learning_rate": 4.442454508114106e-05, "loss": 0.0065, "step": 55880 }, { "grad_norm": 0.10916455835103989, "learning_rate": 4.4408113808209465e-05, "loss": 0.0065, "step": 55890 }, { "grad_norm": 0.12097803503274918, "learning_rate": 4.439168314679813e-05, "loss": 0.0061, "step": 55900 }, { "grad_norm": 0.11025518923997879, "learning_rate": 4.43752530987039e-05, "loss": 0.0069, "step": 55910 }, { "grad_norm": 0.12428605556488037, "learning_rate": 4.4358823665723515e-05, "loss": 0.007, "step": 55920 }, { "grad_norm": 0.14625929296016693, "learning_rate": 4.434239484965371e-05, "loss": 0.0063, "step": 55930 }, { "grad_norm": 0.11967112869024277, "learning_rate": 4.4325966652291103e-05, "loss": 0.0062, "step": 55940 }, { "grad_norm": 0.1256553679704666, "learning_rate": 4.430953907543225e-05, "loss": 0.007, "step": 55950 }, { "grad_norm": 0.1044684424996376, "learning_rate": 4.4293112120873645e-05, "loss": 0.0079, "step": 55960 }, { "grad_norm": 0.09606550633907318, "learning_rate": 4.427668579041172e-05, "loss": 0.0054, "step": 55970 }, { "grad_norm": 0.10578922927379608, "learning_rate": 4.4260260085842816e-05, "loss": 0.0073, "step": 55980 }, { "grad_norm": 0.13791203498840332, "learning_rate": 4.4243835008963256e-05, "loss": 0.0073, "step": 55990 }, { "grad_norm": 0.11400335282087326, "learning_rate": 4.422741056156925e-05, "loss": 0.0071, "step": 56000 }, { "grad_norm": 0.11496065557003021, "learning_rate": 4.421098674545693e-05, "loss": 0.0072, "step": 56010 }, { "grad_norm": 0.11831231415271759, "learning_rate": 4.419456356242241e-05, "loss": 0.0076, "step": 56020 }, { "grad_norm": 0.1175859123468399, "learning_rate": 4.4178141014261685e-05, "loss": 0.0073, "step": 56030 }, { "grad_norm": 0.1153288260102272, "learning_rate": 4.416171910277071e-05, "loss": 0.009, "step": 56040 }, { "grad_norm": 0.1144522875547409, "learning_rate": 4.414529782974533e-05, "loss": 0.0078, "step": 56050 }, { "grad_norm": 0.11052558571100235, "learning_rate": 4.41288771969814e-05, "loss": 0.0071, "step": 56060 }, { "grad_norm": 0.1019623801112175, "learning_rate": 4.411245720627464e-05, "loss": 0.0085, "step": 56070 }, { "grad_norm": 0.1036878228187561, "learning_rate": 4.40960378594207e-05, "loss": 0.007, "step": 56080 }, { "grad_norm": 0.11975505948066711, "learning_rate": 4.407961915821519e-05, "loss": 0.0093, "step": 56090 }, { "grad_norm": 0.11620526760816574, "learning_rate": 4.406320110445362e-05, "loss": 0.0071, "step": 56100 }, { "grad_norm": 0.14629165828227997, "learning_rate": 4.404678369993144e-05, "loss": 0.0065, "step": 56110 }, { "grad_norm": 0.11118653416633606, "learning_rate": 4.403036694644406e-05, "loss": 0.0065, "step": 56120 }, { "grad_norm": 0.130654975771904, "learning_rate": 4.4013950845786764e-05, "loss": 0.0084, "step": 56130 }, { "grad_norm": 0.09382301568984985, "learning_rate": 4.399753539975482e-05, "loss": 0.0086, "step": 56140 }, { "grad_norm": 0.14205987751483917, "learning_rate": 4.398112061014337e-05, "loss": 0.0078, "step": 56150 }, { "grad_norm": 0.11215009540319443, "learning_rate": 4.396470647874753e-05, "loss": 0.006, "step": 56160 }, { "grad_norm": 0.12294259667396545, "learning_rate": 4.394829300736229e-05, "loss": 0.0063, "step": 56170 }, { "grad_norm": 0.13001219928264618, "learning_rate": 4.393188019778265e-05, "loss": 0.0079, "step": 56180 }, { "grad_norm": 0.10798157006502151, "learning_rate": 4.391546805180347e-05, "loss": 0.006, "step": 56190 }, { "grad_norm": 0.142380490899086, "learning_rate": 4.389905657121955e-05, "loss": 0.0065, "step": 56200 }, { "grad_norm": 0.1125456914305687, "learning_rate": 4.3882645757825635e-05, "loss": 0.0072, "step": 56210 }, { "grad_norm": 0.13664577901363373, "learning_rate": 4.386623561341637e-05, "loss": 0.0084, "step": 56220 }, { "grad_norm": 0.10352405905723572, "learning_rate": 4.384982613978637e-05, "loss": 0.0064, "step": 56230 }, { "grad_norm": 0.12165210396051407, "learning_rate": 4.383341733873012e-05, "loss": 0.0056, "step": 56240 }, { "grad_norm": 0.10225563496351242, "learning_rate": 4.381700921204209e-05, "loss": 0.0059, "step": 56250 }, { "grad_norm": 0.10513672232627869, "learning_rate": 4.380060176151663e-05, "loss": 0.0077, "step": 56260 }, { "grad_norm": 0.11262010037899017, "learning_rate": 4.378419498894805e-05, "loss": 0.0077, "step": 56270 }, { "grad_norm": 0.1114250048995018, "learning_rate": 4.376778889613056e-05, "loss": 0.0057, "step": 56280 }, { "grad_norm": 0.10148158669471741, "learning_rate": 4.375138348485831e-05, "loss": 0.0067, "step": 56290 }, { "grad_norm": 0.09162231534719467, "learning_rate": 4.3734978756925346e-05, "loss": 0.008, "step": 56300 }, { "grad_norm": 0.13576707243919373, "learning_rate": 4.3718574714125706e-05, "loss": 0.0092, "step": 56310 }, { "grad_norm": 0.10417993366718292, "learning_rate": 4.370217135825329e-05, "loss": 0.0083, "step": 56320 }, { "grad_norm": 0.10655663162469864, "learning_rate": 4.368576869110194e-05, "loss": 0.0059, "step": 56330 }, { "grad_norm": 0.09075412154197693, "learning_rate": 4.366936671446544e-05, "loss": 0.0058, "step": 56340 }, { "grad_norm": 0.11769462376832962, "learning_rate": 4.3652965430137474e-05, "loss": 0.0065, "step": 56350 }, { "grad_norm": 0.15403403341770172, "learning_rate": 4.3636564839911646e-05, "loss": 0.007, "step": 56360 }, { "grad_norm": 0.08336088806390762, "learning_rate": 4.3620164945581545e-05, "loss": 0.0068, "step": 56370 }, { "grad_norm": 0.09953853487968445, "learning_rate": 4.360376574894061e-05, "loss": 0.0059, "step": 56380 }, { "grad_norm": 0.12012682855129242, "learning_rate": 4.358736725178224e-05, "loss": 0.0064, "step": 56390 }, { "grad_norm": 0.11791420727968216, "learning_rate": 4.357096945589974e-05, "loss": 0.0077, "step": 56400 }, { "grad_norm": 0.1237574890255928, "learning_rate": 4.355457236308636e-05, "loss": 0.0075, "step": 56410 }, { "grad_norm": 0.10629960894584656, "learning_rate": 4.353817597513526e-05, "loss": 0.0078, "step": 56420 }, { "grad_norm": 0.13434439897537231, "learning_rate": 4.352178029383948e-05, "loss": 0.0076, "step": 56430 }, { "grad_norm": 0.10224287956953049, "learning_rate": 4.3505385320992105e-05, "loss": 0.0081, "step": 56440 }, { "grad_norm": 0.09810726344585419, "learning_rate": 4.348899105838602e-05, "loss": 0.0064, "step": 56450 }, { "grad_norm": 0.11641434580087662, "learning_rate": 4.3472597507814087e-05, "loss": 0.0068, "step": 56460 }, { "grad_norm": 0.08047681301832199, "learning_rate": 4.3456204671069066e-05, "loss": 0.0063, "step": 56470 }, { "grad_norm": 0.10596291720867157, "learning_rate": 4.343981254994367e-05, "loss": 0.0053, "step": 56480 }, { "grad_norm": 0.13658404350280762, "learning_rate": 4.34234211462305e-05, "loss": 0.0061, "step": 56490 }, { "grad_norm": 0.11040370166301727, "learning_rate": 4.340703046172213e-05, "loss": 0.0061, "step": 56500 }, { "grad_norm": 0.1097048744559288, "learning_rate": 4.339064049821097e-05, "loss": 0.0057, "step": 56510 }, { "grad_norm": 0.12530751526355743, "learning_rate": 4.3374251257489446e-05, "loss": 0.0075, "step": 56520 }, { "grad_norm": 0.10982491075992584, "learning_rate": 4.335786274134984e-05, "loss": 0.0066, "step": 56530 }, { "grad_norm": 0.09726940095424652, "learning_rate": 4.334147495158438e-05, "loss": 0.0078, "step": 56540 }, { "grad_norm": 0.11469588428735733, "learning_rate": 4.33250878899852e-05, "loss": 0.0062, "step": 56550 }, { "grad_norm": 0.10125516355037689, "learning_rate": 4.330870155834437e-05, "loss": 0.0054, "step": 56560 }, { "grad_norm": 0.13119512796401978, "learning_rate": 4.3292315958453884e-05, "loss": 0.0068, "step": 56570 }, { "grad_norm": 0.12412481755018234, "learning_rate": 4.3275931092105646e-05, "loss": 0.0057, "step": 56580 }, { "grad_norm": 0.10027851164340973, "learning_rate": 4.325954696109147e-05, "loss": 0.0071, "step": 56590 }, { "grad_norm": 0.11122427880764008, "learning_rate": 4.32431635672031e-05, "loss": 0.0058, "step": 56600 }, { "grad_norm": 0.11162281036376953, "learning_rate": 4.3226780912232214e-05, "loss": 0.0062, "step": 56610 }, { "grad_norm": 0.13130316138267517, "learning_rate": 4.321039899797036e-05, "loss": 0.0075, "step": 56620 }, { "grad_norm": 0.11080505698919296, "learning_rate": 4.319401782620908e-05, "loss": 0.0068, "step": 56630 }, { "grad_norm": 0.08298809826374054, "learning_rate": 4.317763739873978e-05, "loss": 0.0062, "step": 56640 }, { "grad_norm": 0.09110712260007858, "learning_rate": 4.31612577173538e-05, "loss": 0.0071, "step": 56650 }, { "grad_norm": 0.16520732641220093, "learning_rate": 4.314487878384239e-05, "loss": 0.0073, "step": 56660 }, { "grad_norm": 0.15460270643234253, "learning_rate": 4.3128500599996736e-05, "loss": 0.0075, "step": 56670 }, { "grad_norm": 0.0871281772851944, "learning_rate": 4.31121231676079e-05, "loss": 0.0063, "step": 56680 }, { "grad_norm": 0.0938567966222763, "learning_rate": 4.309574648846694e-05, "loss": 0.005, "step": 56690 }, { "grad_norm": 0.10845202952623367, "learning_rate": 4.3079370564364755e-05, "loss": 0.0056, "step": 56700 }, { "grad_norm": 0.11697247624397278, "learning_rate": 4.306299539709221e-05, "loss": 0.0065, "step": 56710 }, { "grad_norm": 0.10711189359426498, "learning_rate": 4.3046620988440046e-05, "loss": 0.0064, "step": 56720 }, { "grad_norm": 0.09357878565788269, "learning_rate": 4.303024734019897e-05, "loss": 0.0063, "step": 56730 }, { "grad_norm": 0.13217441737651825, "learning_rate": 4.3013874454159565e-05, "loss": 0.0064, "step": 56740 }, { "grad_norm": 0.10358576476573944, "learning_rate": 4.299750233211233e-05, "loss": 0.006, "step": 56750 }, { "grad_norm": 0.12171294540166855, "learning_rate": 4.2981130975847715e-05, "loss": 0.0068, "step": 56760 }, { "grad_norm": 0.10463415831327438, "learning_rate": 4.2964760387156076e-05, "loss": 0.0063, "step": 56770 }, { "grad_norm": 0.08436111360788345, "learning_rate": 4.294839056782767e-05, "loss": 0.0053, "step": 56780 }, { "grad_norm": 0.09605041891336441, "learning_rate": 4.293202151965266e-05, "loss": 0.0068, "step": 56790 }, { "grad_norm": 0.08254580199718475, "learning_rate": 4.291565324442115e-05, "loss": 0.0061, "step": 56800 }, { "grad_norm": 0.1260201632976532, "learning_rate": 4.2899285743923135e-05, "loss": 0.006, "step": 56810 }, { "grad_norm": 0.11705236881971359, "learning_rate": 4.288291901994857e-05, "loss": 0.0068, "step": 56820 }, { "grad_norm": 0.10694178193807602, "learning_rate": 4.286655307428729e-05, "loss": 0.0059, "step": 56830 }, { "grad_norm": 0.10731728374958038, "learning_rate": 4.285018790872903e-05, "loss": 0.007, "step": 56840 }, { "grad_norm": 0.11901775747537613, "learning_rate": 4.283382352506347e-05, "loss": 0.0067, "step": 56850 }, { "grad_norm": 0.13496500253677368, "learning_rate": 4.2817459925080205e-05, "loss": 0.0057, "step": 56860 }, { "grad_norm": 0.09604617953300476, "learning_rate": 4.28010971105687e-05, "loss": 0.0067, "step": 56870 }, { "grad_norm": 0.11840730160474777, "learning_rate": 4.278473508331841e-05, "loss": 0.0076, "step": 56880 }, { "grad_norm": 0.0984293594956398, "learning_rate": 4.276837384511864e-05, "loss": 0.007, "step": 56890 }, { "grad_norm": 0.11740535497665405, "learning_rate": 4.275201339775864e-05, "loss": 0.006, "step": 56900 }, { "grad_norm": 0.1576748490333557, "learning_rate": 4.273565374302756e-05, "loss": 0.0061, "step": 56910 }, { "grad_norm": 0.09750498831272125, "learning_rate": 4.2719294882714454e-05, "loss": 0.0058, "step": 56920 }, { "grad_norm": 0.12573809921741486, "learning_rate": 4.270293681860831e-05, "loss": 0.0062, "step": 56930 }, { "grad_norm": 0.10627594590187073, "learning_rate": 4.268657955249802e-05, "loss": 0.0068, "step": 56940 }, { "grad_norm": 0.12986795604228973, "learning_rate": 4.267022308617241e-05, "loss": 0.0073, "step": 56950 }, { "grad_norm": 0.08334340900182724, "learning_rate": 4.265386742142018e-05, "loss": 0.0061, "step": 56960 }, { "grad_norm": 0.10783644765615463, "learning_rate": 4.263751256002996e-05, "loss": 0.0067, "step": 56970 }, { "grad_norm": 0.14031866192817688, "learning_rate": 4.26211585037903e-05, "loss": 0.0074, "step": 56980 }, { "grad_norm": 0.06617404520511627, "learning_rate": 4.2604805254489656e-05, "loss": 0.0067, "step": 56990 }, { "grad_norm": 0.11763538420200348, "learning_rate": 4.258845281391638e-05, "loss": 0.0073, "step": 57000 }, { "grad_norm": 0.1132822334766388, "learning_rate": 4.257210118385876e-05, "loss": 0.0059, "step": 57010 }, { "grad_norm": 0.09979046136140823, "learning_rate": 4.2555750366104994e-05, "loss": 0.0067, "step": 57020 }, { "grad_norm": 0.09016510099172592, "learning_rate": 4.2539400362443184e-05, "loss": 0.0058, "step": 57030 }, { "grad_norm": 0.106588214635849, "learning_rate": 4.252305117466133e-05, "loss": 0.0083, "step": 57040 }, { "grad_norm": 0.1000777930021286, "learning_rate": 4.2506702804547356e-05, "loss": 0.0056, "step": 57050 }, { "grad_norm": 0.08935286849737167, "learning_rate": 4.249035525388907e-05, "loss": 0.0058, "step": 57060 }, { "grad_norm": 0.08126911520957947, "learning_rate": 4.247400852447427e-05, "loss": 0.0073, "step": 57070 }, { "grad_norm": 0.10496309399604797, "learning_rate": 4.245766261809059e-05, "loss": 0.0065, "step": 57080 }, { "grad_norm": 0.15749481320381165, "learning_rate": 4.244131753652558e-05, "loss": 0.0074, "step": 57090 }, { "grad_norm": 0.12972986698150635, "learning_rate": 4.242497328156672e-05, "loss": 0.0072, "step": 57100 }, { "grad_norm": 0.1345498114824295, "learning_rate": 4.240862985500139e-05, "loss": 0.0077, "step": 57110 }, { "grad_norm": 0.10654494166374207, "learning_rate": 4.2392287258616894e-05, "loss": 0.007, "step": 57120 }, { "grad_norm": 0.1104971319437027, "learning_rate": 4.2375945494200396e-05, "loss": 0.0071, "step": 57130 }, { "grad_norm": 0.12319052219390869, "learning_rate": 4.2359604563539054e-05, "loss": 0.0059, "step": 57140 }, { "grad_norm": 0.15037016570568085, "learning_rate": 4.2343264468419885e-05, "loss": 0.0068, "step": 57150 }, { "grad_norm": 0.14100396633148193, "learning_rate": 4.232692521062979e-05, "loss": 0.0081, "step": 57160 }, { "grad_norm": 0.1507474035024643, "learning_rate": 4.231058679195562e-05, "loss": 0.0071, "step": 57170 }, { "grad_norm": 0.1326330155134201, "learning_rate": 4.229424921418411e-05, "loss": 0.0063, "step": 57180 }, { "grad_norm": 0.13428372144699097, "learning_rate": 4.2277912479101916e-05, "loss": 0.0058, "step": 57190 }, { "grad_norm": 0.1613362431526184, "learning_rate": 4.226157658849562e-05, "loss": 0.0074, "step": 57200 }, { "grad_norm": 0.1127590537071228, "learning_rate": 4.2245241544151664e-05, "loss": 0.0066, "step": 57210 }, { "grad_norm": 0.11835408955812454, "learning_rate": 4.2228907347856425e-05, "loss": 0.006, "step": 57220 }, { "grad_norm": 0.12031228840351105, "learning_rate": 4.2212574001396214e-05, "loss": 0.0094, "step": 57230 }, { "grad_norm": 0.08524278551340103, "learning_rate": 4.21962415065572e-05, "loss": 0.0058, "step": 57240 }, { "grad_norm": 0.07925297319889069, "learning_rate": 4.217990986512548e-05, "loss": 0.0068, "step": 57250 }, { "grad_norm": 0.09733537584543228, "learning_rate": 4.216357907888704e-05, "loss": 0.007, "step": 57260 }, { "grad_norm": 0.09186708182096481, "learning_rate": 4.2147249149627824e-05, "loss": 0.0059, "step": 57270 }, { "grad_norm": 0.1598810851573944, "learning_rate": 4.213092007913364e-05, "loss": 0.0077, "step": 57280 }, { "grad_norm": 0.08337001502513885, "learning_rate": 4.2114591869190205e-05, "loss": 0.0057, "step": 57290 }, { "grad_norm": 0.10291951149702072, "learning_rate": 4.209826452158315e-05, "loss": 0.0059, "step": 57300 }, { "grad_norm": 0.11396501958370209, "learning_rate": 4.2081938038098e-05, "loss": 0.0064, "step": 57310 }, { "grad_norm": 0.11187098920345306, "learning_rate": 4.2065612420520185e-05, "loss": 0.0065, "step": 57320 }, { "grad_norm": 0.10494215786457062, "learning_rate": 4.204928767063509e-05, "loss": 0.006, "step": 57330 }, { "grad_norm": 0.11226435005664825, "learning_rate": 4.2032963790227944e-05, "loss": 0.0068, "step": 57340 }, { "grad_norm": 0.11948125064373016, "learning_rate": 4.2016640781083887e-05, "loss": 0.0065, "step": 57350 }, { "grad_norm": 0.17222094535827637, "learning_rate": 4.2000318644988004e-05, "loss": 0.0075, "step": 57360 }, { "grad_norm": 0.13175521790981293, "learning_rate": 4.198399738372524e-05, "loss": 0.0082, "step": 57370 }, { "grad_norm": 0.1381402462720871, "learning_rate": 4.196767699908045e-05, "loss": 0.0063, "step": 57380 }, { "grad_norm": 0.12255235016345978, "learning_rate": 4.195135749283845e-05, "loss": 0.0067, "step": 57390 }, { "grad_norm": 0.11785043776035309, "learning_rate": 4.1935038866783894e-05, "loss": 0.0087, "step": 57400 }, { "grad_norm": 0.12472867220640182, "learning_rate": 4.1918721122701355e-05, "loss": 0.0073, "step": 57410 }, { "grad_norm": 0.13036298751831055, "learning_rate": 4.190240426237533e-05, "loss": 0.0068, "step": 57420 }, { "grad_norm": 0.12452436983585358, "learning_rate": 4.188608828759019e-05, "loss": 0.0065, "step": 57430 }, { "grad_norm": 0.12999315559864044, "learning_rate": 4.1869773200130246e-05, "loss": 0.0062, "step": 57440 }, { "grad_norm": 0.11581087857484818, "learning_rate": 4.185345900177967e-05, "loss": 0.0067, "step": 57450 }, { "grad_norm": 0.11492796987295151, "learning_rate": 4.183714569432258e-05, "loss": 0.0057, "step": 57460 }, { "grad_norm": 0.08655540645122528, "learning_rate": 4.182083327954295e-05, "loss": 0.0056, "step": 57470 }, { "grad_norm": 0.13048022985458374, "learning_rate": 4.1804521759224714e-05, "loss": 0.0066, "step": 57480 }, { "grad_norm": 0.09585462510585785, "learning_rate": 4.178821113515165e-05, "loss": 0.006, "step": 57490 }, { "grad_norm": 0.1281941682100296, "learning_rate": 4.1771901409107476e-05, "loss": 0.0055, "step": 57500 }, { "grad_norm": 0.16225090622901917, "learning_rate": 4.175559258287577e-05, "loss": 0.0067, "step": 57510 }, { "grad_norm": 0.12651513516902924, "learning_rate": 4.173928465824008e-05, "loss": 0.0058, "step": 57520 }, { "grad_norm": 0.1026979312300682, "learning_rate": 4.17229776369838e-05, "loss": 0.0055, "step": 57530 }, { "grad_norm": 0.09824515134096146, "learning_rate": 4.170667152089025e-05, "loss": 0.0063, "step": 57540 }, { "grad_norm": 0.13478223979473114, "learning_rate": 4.1690366311742626e-05, "loss": 0.0058, "step": 57550 }, { "grad_norm": 0.1088535413146019, "learning_rate": 4.1674062011324055e-05, "loss": 0.0074, "step": 57560 }, { "grad_norm": 0.10375197976827621, "learning_rate": 4.165775862141752e-05, "loss": 0.0055, "step": 57570 }, { "grad_norm": 0.1107097715139389, "learning_rate": 4.164145614380598e-05, "loss": 0.0049, "step": 57580 }, { "grad_norm": 0.09485648572444916, "learning_rate": 4.162515458027223e-05, "loss": 0.0069, "step": 57590 }, { "grad_norm": 0.08737044781446457, "learning_rate": 4.160885393259899e-05, "loss": 0.0053, "step": 57600 }, { "grad_norm": 0.13415232300758362, "learning_rate": 4.159255420256886e-05, "loss": 0.0066, "step": 57610 }, { "grad_norm": 0.13642114400863647, "learning_rate": 4.157625539196436e-05, "loss": 0.0063, "step": 57620 }, { "grad_norm": 0.12347769737243652, "learning_rate": 4.155995750256791e-05, "loss": 0.0055, "step": 57630 }, { "grad_norm": 0.11418895423412323, "learning_rate": 4.15436605361618e-05, "loss": 0.006, "step": 57640 }, { "grad_norm": 0.09281837940216064, "learning_rate": 4.152736449452827e-05, "loss": 0.0055, "step": 57650 }, { "grad_norm": 0.09471675753593445, "learning_rate": 4.151106937944943e-05, "loss": 0.0056, "step": 57660 }, { "grad_norm": 0.13835640251636505, "learning_rate": 4.149477519270727e-05, "loss": 0.0067, "step": 57670 }, { "grad_norm": 0.0969737321138382, "learning_rate": 4.147848193608371e-05, "loss": 0.0063, "step": 57680 }, { "grad_norm": 0.13188865780830383, "learning_rate": 4.146218961136056e-05, "loss": 0.0058, "step": 57690 }, { "grad_norm": 0.18105390667915344, "learning_rate": 4.14458982203195e-05, "loss": 0.0081, "step": 57700 }, { "grad_norm": 0.14978349208831787, "learning_rate": 4.1429607764742165e-05, "loss": 0.0062, "step": 57710 }, { "grad_norm": 0.10884452611207962, "learning_rate": 4.141331824641003e-05, "loss": 0.006, "step": 57720 }, { "grad_norm": 0.10390938073396683, "learning_rate": 4.139702966710452e-05, "loss": 0.0063, "step": 57730 }, { "grad_norm": 0.09459622204303741, "learning_rate": 4.13807420286069e-05, "loss": 0.0057, "step": 57740 }, { "grad_norm": 0.10314320772886276, "learning_rate": 4.136445533269837e-05, "loss": 0.0065, "step": 57750 }, { "grad_norm": 0.13103625178337097, "learning_rate": 4.134816958116003e-05, "loss": 0.0059, "step": 57760 }, { "grad_norm": 0.1039150133728981, "learning_rate": 4.133188477577283e-05, "loss": 0.0058, "step": 57770 }, { "grad_norm": 0.10690154135227203, "learning_rate": 4.131560091831771e-05, "loss": 0.0051, "step": 57780 }, { "grad_norm": 0.2138950377702713, "learning_rate": 4.1299318010575405e-05, "loss": 0.0069, "step": 57790 }, { "grad_norm": 0.11112163960933685, "learning_rate": 4.128303605432661e-05, "loss": 0.0061, "step": 57800 }, { "grad_norm": 0.1315818727016449, "learning_rate": 4.126675505135188e-05, "loss": 0.0049, "step": 57810 }, { "grad_norm": 0.08670145273208618, "learning_rate": 4.125047500343169e-05, "loss": 0.0047, "step": 57820 }, { "grad_norm": 0.12199147045612335, "learning_rate": 4.1234195912346376e-05, "loss": 0.0057, "step": 57830 }, { "grad_norm": 0.10647493600845337, "learning_rate": 4.1217917779876235e-05, "loss": 0.0056, "step": 57840 }, { "grad_norm": 0.13969632983207703, "learning_rate": 4.120164060780141e-05, "loss": 0.0071, "step": 57850 }, { "grad_norm": 0.1282356083393097, "learning_rate": 4.1185364397901935e-05, "loss": 0.006, "step": 57860 }, { "grad_norm": 0.1614932268857956, "learning_rate": 4.116908915195776e-05, "loss": 0.0059, "step": 57870 }, { "grad_norm": 0.12866373360157013, "learning_rate": 4.115281487174872e-05, "loss": 0.0053, "step": 57880 }, { "grad_norm": 0.15441899001598358, "learning_rate": 4.113654155905452e-05, "loss": 0.0068, "step": 57890 }, { "grad_norm": 0.15642079710960388, "learning_rate": 4.112026921565483e-05, "loss": 0.0064, "step": 57900 }, { "grad_norm": 0.09511853009462357, "learning_rate": 4.1103997843329147e-05, "loss": 0.0058, "step": 57910 }, { "grad_norm": 0.0832308977842331, "learning_rate": 4.108772744385689e-05, "loss": 0.0071, "step": 57920 }, { "grad_norm": 0.10156688094139099, "learning_rate": 4.1071458019017346e-05, "loss": 0.0053, "step": 57930 }, { "grad_norm": 0.11067791283130646, "learning_rate": 4.1055189570589734e-05, "loss": 0.0058, "step": 57940 }, { "grad_norm": 0.1300879567861557, "learning_rate": 4.103892210035315e-05, "loss": 0.0057, "step": 57950 }, { "grad_norm": 0.10769638419151306, "learning_rate": 4.102265561008656e-05, "loss": 0.0063, "step": 57960 }, { "grad_norm": 0.12320544570684433, "learning_rate": 4.100639010156885e-05, "loss": 0.0063, "step": 57970 }, { "grad_norm": 0.11709413677453995, "learning_rate": 4.099012557657882e-05, "loss": 0.0057, "step": 57980 }, { "grad_norm": 0.13432341814041138, "learning_rate": 4.09738620368951e-05, "loss": 0.0058, "step": 57990 }, { "grad_norm": 0.15200112760066986, "learning_rate": 4.095759948429626e-05, "loss": 0.0057, "step": 58000 }, { "grad_norm": 0.11217443645000458, "learning_rate": 4.094133792056075e-05, "loss": 0.0056, "step": 58010 }, { "grad_norm": 0.08401450514793396, "learning_rate": 4.092507734746688e-05, "loss": 0.0054, "step": 58020 }, { "grad_norm": 0.10088995844125748, "learning_rate": 4.090881776679293e-05, "loss": 0.0055, "step": 58030 }, { "grad_norm": 0.10567250847816467, "learning_rate": 4.0892559180316994e-05, "loss": 0.0065, "step": 58040 }, { "grad_norm": 0.11113817989826202, "learning_rate": 4.08763015898171e-05, "loss": 0.0074, "step": 58050 }, { "grad_norm": 0.10349566489458084, "learning_rate": 4.0860044997071145e-05, "loss": 0.0066, "step": 58060 }, { "grad_norm": 0.09639541804790497, "learning_rate": 4.084378940385693e-05, "loss": 0.0067, "step": 58070 }, { "grad_norm": 0.09004424512386322, "learning_rate": 4.082753481195211e-05, "loss": 0.0054, "step": 58080 }, { "grad_norm": 0.09312469512224197, "learning_rate": 4.0811281223134324e-05, "loss": 0.0054, "step": 58090 }, { "grad_norm": 0.12949110567569733, "learning_rate": 4.0795028639181e-05, "loss": 0.0093, "step": 58100 }, { "grad_norm": 0.1074109598994255, "learning_rate": 4.077877706186951e-05, "loss": 0.0059, "step": 58110 }, { "grad_norm": 0.10595256090164185, "learning_rate": 4.076252649297709e-05, "loss": 0.0059, "step": 58120 }, { "grad_norm": 0.09632682800292969, "learning_rate": 4.07462769342809e-05, "loss": 0.0067, "step": 58130 }, { "grad_norm": 0.12691956758499146, "learning_rate": 4.073002838755793e-05, "loss": 0.006, "step": 58140 }, { "grad_norm": 0.10059105604887009, "learning_rate": 4.0713780854585115e-05, "loss": 0.005, "step": 58150 }, { "grad_norm": 0.09045354276895523, "learning_rate": 4.069753433713927e-05, "loss": 0.0079, "step": 58160 }, { "grad_norm": 0.11825306713581085, "learning_rate": 4.068128883699709e-05, "loss": 0.0063, "step": 58170 }, { "grad_norm": 0.08860639482736588, "learning_rate": 4.0665044355935155e-05, "loss": 0.0062, "step": 58180 }, { "grad_norm": 0.1254737377166748, "learning_rate": 4.064880089572993e-05, "loss": 0.007, "step": 58190 }, { "grad_norm": 0.11704118549823761, "learning_rate": 4.063255845815779e-05, "loss": 0.0084, "step": 58200 }, { "grad_norm": 0.11571390926837921, "learning_rate": 4.0616317044994945e-05, "loss": 0.0078, "step": 58210 }, { "grad_norm": 0.10801474750041962, "learning_rate": 4.0600076658017585e-05, "loss": 0.0063, "step": 58220 }, { "grad_norm": 0.1260145753622055, "learning_rate": 4.0583837299001706e-05, "loss": 0.0067, "step": 58230 }, { "grad_norm": 0.10420211404561996, "learning_rate": 4.056759896972322e-05, "loss": 0.0057, "step": 58240 }, { "grad_norm": 0.07713592797517776, "learning_rate": 4.055136167195793e-05, "loss": 0.0073, "step": 58250 }, { "grad_norm": 0.0911090224981308, "learning_rate": 4.053512540748152e-05, "loss": 0.0056, "step": 58260 }, { "grad_norm": 0.09513361752033234, "learning_rate": 4.051889017806955e-05, "loss": 0.0049, "step": 58270 }, { "grad_norm": 0.08725281059741974, "learning_rate": 4.0502655985497516e-05, "loss": 0.0047, "step": 58280 }, { "grad_norm": 0.08889582753181458, "learning_rate": 4.048642283154074e-05, "loss": 0.0053, "step": 58290 }, { "grad_norm": 0.13969221711158752, "learning_rate": 4.0470190717974464e-05, "loss": 0.0079, "step": 58300 }, { "grad_norm": 0.15194864571094513, "learning_rate": 4.045395964657381e-05, "loss": 0.0074, "step": 58310 }, { "grad_norm": 0.12219763547182083, "learning_rate": 4.043772961911376e-05, "loss": 0.0066, "step": 58320 }, { "grad_norm": 0.1478470265865326, "learning_rate": 4.042150063736924e-05, "loss": 0.0068, "step": 58330 }, { "grad_norm": 0.12836545705795288, "learning_rate": 4.0405272703114976e-05, "loss": 0.007, "step": 58340 }, { "grad_norm": 0.10797368735074997, "learning_rate": 4.038904581812569e-05, "loss": 0.0066, "step": 58350 }, { "grad_norm": 0.09301430732011795, "learning_rate": 4.03728199841759e-05, "loss": 0.0059, "step": 58360 }, { "grad_norm": 0.0919533222913742, "learning_rate": 4.0356595203040044e-05, "loss": 0.0056, "step": 58370 }, { "grad_norm": 0.10407698899507523, "learning_rate": 4.0340371476492426e-05, "loss": 0.0057, "step": 58380 }, { "grad_norm": 0.10167603939771652, "learning_rate": 4.032414880630727e-05, "loss": 0.0055, "step": 58390 }, { "grad_norm": 0.13193581998348236, "learning_rate": 4.0307927194258646e-05, "loss": 0.0076, "step": 58400 }, { "grad_norm": 0.09311363846063614, "learning_rate": 4.029170664212054e-05, "loss": 0.0068, "step": 58410 }, { "grad_norm": 0.08765625953674316, "learning_rate": 4.0275487151666794e-05, "loss": 0.0052, "step": 58420 }, { "grad_norm": 0.11616057902574539, "learning_rate": 4.025926872467116e-05, "loss": 0.0063, "step": 58430 }, { "grad_norm": 0.1099938377737999, "learning_rate": 4.024305136290725e-05, "loss": 0.0066, "step": 58440 }, { "grad_norm": 0.11619488894939423, "learning_rate": 4.022683506814858e-05, "loss": 0.0051, "step": 58450 }, { "grad_norm": 0.11949317157268524, "learning_rate": 4.021061984216853e-05, "loss": 0.007, "step": 58460 }, { "grad_norm": 0.10854317992925644, "learning_rate": 4.019440568674035e-05, "loss": 0.0053, "step": 58470 }, { "grad_norm": 0.11169932037591934, "learning_rate": 4.017819260363724e-05, "loss": 0.0073, "step": 58480 }, { "grad_norm": 0.10538572818040848, "learning_rate": 4.016198059463223e-05, "loss": 0.005, "step": 58490 }, { "grad_norm": 0.1137530505657196, "learning_rate": 4.0145769661498233e-05, "loss": 0.0058, "step": 58500 }, { "grad_norm": 0.12907330691814423, "learning_rate": 4.012955980600804e-05, "loss": 0.006, "step": 58510 }, { "grad_norm": 0.11845795065164566, "learning_rate": 4.0113351029934355e-05, "loss": 0.0054, "step": 58520 }, { "grad_norm": 0.10726477950811386, "learning_rate": 4.0097143335049714e-05, "loss": 0.0065, "step": 58530 }, { "grad_norm": 0.098094642162323, "learning_rate": 4.0080936723126606e-05, "loss": 0.0084, "step": 58540 }, { "grad_norm": 0.10012064129114151, "learning_rate": 4.006473119593735e-05, "loss": 0.007, "step": 58550 }, { "grad_norm": 0.09336435794830322, "learning_rate": 4.004852675525415e-05, "loss": 0.0061, "step": 58560 }, { "grad_norm": 0.14394350349903107, "learning_rate": 4.00323234028491e-05, "loss": 0.0074, "step": 58570 }, { "grad_norm": 0.1123717725276947, "learning_rate": 4.0016121140494186e-05, "loss": 0.0062, "step": 58580 }, { "grad_norm": 0.12112783640623093, "learning_rate": 3.999991996996122e-05, "loss": 0.0061, "step": 58590 }, { "grad_norm": 0.1454196572303772, "learning_rate": 3.9983719893022e-05, "loss": 0.0059, "step": 58600 }, { "grad_norm": 0.11278092861175537, "learning_rate": 3.996752091144812e-05, "loss": 0.0065, "step": 58610 }, { "grad_norm": 0.10803455114364624, "learning_rate": 3.9951323027011055e-05, "loss": 0.0059, "step": 58620 }, { "grad_norm": 0.11135552823543549, "learning_rate": 3.993512624148219e-05, "loss": 0.0063, "step": 58630 }, { "grad_norm": 0.08837126940488815, "learning_rate": 3.99189305566328e-05, "loss": 0.005, "step": 58640 }, { "grad_norm": 0.09576472640037537, "learning_rate": 3.990273597423401e-05, "loss": 0.005, "step": 58650 }, { "grad_norm": 0.09212444722652435, "learning_rate": 3.988654249605681e-05, "loss": 0.0057, "step": 58660 }, { "grad_norm": 0.0998363122344017, "learning_rate": 3.987035012387213e-05, "loss": 0.0072, "step": 58670 }, { "grad_norm": 0.13263019919395447, "learning_rate": 3.9854158859450735e-05, "loss": 0.0074, "step": 58680 }, { "grad_norm": 0.13183537125587463, "learning_rate": 3.983796870456328e-05, "loss": 0.0076, "step": 58690 }, { "grad_norm": 0.10991086065769196, "learning_rate": 3.982177966098028e-05, "loss": 0.006, "step": 58700 }, { "grad_norm": 0.09604376554489136, "learning_rate": 3.980559173047216e-05, "loss": 0.0064, "step": 58710 }, { "grad_norm": 0.12739667296409607, "learning_rate": 3.9789404914809176e-05, "loss": 0.006, "step": 58720 }, { "grad_norm": 0.08626261353492737, "learning_rate": 3.9773219215761544e-05, "loss": 0.0054, "step": 58730 }, { "grad_norm": 0.10295593738555908, "learning_rate": 3.975703463509928e-05, "loss": 0.0068, "step": 58740 }, { "grad_norm": 0.12939168512821198, "learning_rate": 3.974085117459231e-05, "loss": 0.0059, "step": 58750 }, { "grad_norm": 0.16380758583545685, "learning_rate": 3.972466883601043e-05, "loss": 0.0062, "step": 58760 }, { "grad_norm": 0.22287969291210175, "learning_rate": 3.9708487621123316e-05, "loss": 0.0074, "step": 58770 }, { "grad_norm": 0.11357270926237106, "learning_rate": 3.969230753170051e-05, "loss": 0.0065, "step": 58780 }, { "grad_norm": 0.0981830507516861, "learning_rate": 3.967612856951146e-05, "loss": 0.0076, "step": 58790 }, { "grad_norm": 0.11503850668668747, "learning_rate": 3.9659950736325476e-05, "loss": 0.0079, "step": 58800 }, { "grad_norm": 0.14465266466140747, "learning_rate": 3.964377403391174e-05, "loss": 0.0079, "step": 58810 }, { "grad_norm": 0.17976322770118713, "learning_rate": 3.9627598464039303e-05, "loss": 0.0065, "step": 58820 }, { "grad_norm": 0.09280620515346527, "learning_rate": 3.9611424028477096e-05, "loss": 0.0061, "step": 58830 }, { "grad_norm": 0.10965973883867264, "learning_rate": 3.959525072899394e-05, "loss": 0.0063, "step": 58840 }, { "grad_norm": 0.10285502672195435, "learning_rate": 3.9579078567358506e-05, "loss": 0.0062, "step": 58850 }, { "grad_norm": 0.09145697951316833, "learning_rate": 3.956290754533939e-05, "loss": 0.0053, "step": 58860 }, { "grad_norm": 0.10273883491754532, "learning_rate": 3.954673766470501e-05, "loss": 0.0048, "step": 58870 }, { "grad_norm": 0.07352414727210999, "learning_rate": 3.9530568927223674e-05, "loss": 0.0068, "step": 58880 }, { "grad_norm": 0.10984569042921066, "learning_rate": 3.95144013346636e-05, "loss": 0.0069, "step": 58890 }, { "grad_norm": 0.1167561337351799, "learning_rate": 3.949823488879282e-05, "loss": 0.0055, "step": 58900 }, { "grad_norm": 0.12687599658966064, "learning_rate": 3.9482069591379276e-05, "loss": 0.0058, "step": 58910 }, { "grad_norm": 0.10878806561231613, "learning_rate": 3.9465905444190784e-05, "loss": 0.0061, "step": 58920 }, { "grad_norm": 0.10765349119901657, "learning_rate": 3.944974244899505e-05, "loss": 0.008, "step": 58930 }, { "grad_norm": 0.1974044293165207, "learning_rate": 3.9433580607559616e-05, "loss": 0.007, "step": 58940 }, { "grad_norm": 0.07601813971996307, "learning_rate": 3.941741992165193e-05, "loss": 0.0069, "step": 58950 }, { "grad_norm": 0.09295031428337097, "learning_rate": 3.940126039303928e-05, "loss": 0.0055, "step": 58960 }, { "grad_norm": 0.11829929053783417, "learning_rate": 3.938510202348886e-05, "loss": 0.0049, "step": 58970 }, { "grad_norm": 0.07665284723043442, "learning_rate": 3.93689448147677e-05, "loss": 0.0049, "step": 58980 }, { "grad_norm": 0.08492536097764969, "learning_rate": 3.935278876864278e-05, "loss": 0.0055, "step": 58990 }, { "grad_norm": 0.10459799319505692, "learning_rate": 3.933663388688087e-05, "loss": 0.0053, "step": 59000 }, { "grad_norm": 0.12163112312555313, "learning_rate": 3.9320480171248645e-05, "loss": 0.0063, "step": 59010 }, { "grad_norm": 0.10795152932405472, "learning_rate": 3.930432762351265e-05, "loss": 0.0059, "step": 59020 }, { "grad_norm": 0.11662419140338898, "learning_rate": 3.928817624543931e-05, "loss": 0.0059, "step": 59030 }, { "grad_norm": 0.0993797555565834, "learning_rate": 3.9272026038794885e-05, "loss": 0.0065, "step": 59040 }, { "grad_norm": 0.10778540372848511, "learning_rate": 3.9255877005345586e-05, "loss": 0.0065, "step": 59050 }, { "grad_norm": 0.11964922398328781, "learning_rate": 3.923972914685743e-05, "loss": 0.0067, "step": 59060 }, { "grad_norm": 0.1095697283744812, "learning_rate": 3.922358246509631e-05, "loss": 0.0056, "step": 59070 }, { "grad_norm": 0.10668525844812393, "learning_rate": 3.920743696182801e-05, "loss": 0.0073, "step": 59080 }, { "grad_norm": 0.07794781774282455, "learning_rate": 3.9191292638818164e-05, "loss": 0.0062, "step": 59090 }, { "grad_norm": 0.0958440750837326, "learning_rate": 3.9175149497832316e-05, "loss": 0.0047, "step": 59100 }, { "grad_norm": 0.10993307083845139, "learning_rate": 3.915900754063584e-05, "loss": 0.007, "step": 59110 }, { "grad_norm": 0.11867816746234894, "learning_rate": 3.9142866768994014e-05, "loss": 0.005, "step": 59120 }, { "grad_norm": 0.15022683143615723, "learning_rate": 3.912672718467193e-05, "loss": 0.0057, "step": 59130 }, { "grad_norm": 0.12751828134059906, "learning_rate": 3.911058878943463e-05, "loss": 0.0071, "step": 59140 }, { "grad_norm": 0.08241754025220871, "learning_rate": 3.9094451585046974e-05, "loss": 0.0052, "step": 59150 }, { "grad_norm": 0.07998914271593094, "learning_rate": 3.90783155732737e-05, "loss": 0.0061, "step": 59160 }, { "grad_norm": 0.12433065474033356, "learning_rate": 3.9062180755879373e-05, "loss": 0.0068, "step": 59170 }, { "grad_norm": 0.10183548182249069, "learning_rate": 3.904604713462855e-05, "loss": 0.0059, "step": 59180 }, { "grad_norm": 0.08107001334428787, "learning_rate": 3.902991471128554e-05, "loss": 0.0058, "step": 59190 }, { "grad_norm": 0.12025853246450424, "learning_rate": 3.901378348761456e-05, "loss": 0.0108, "step": 59200 }, { "grad_norm": 0.10913590341806412, "learning_rate": 3.899765346537971e-05, "loss": 0.0067, "step": 59210 }, { "grad_norm": 0.09399019926786423, "learning_rate": 3.898152464634493e-05, "loss": 0.0056, "step": 59220 }, { "grad_norm": 0.08017481118440628, "learning_rate": 3.896539703227402e-05, "loss": 0.005, "step": 59230 }, { "grad_norm": 0.14626292884349823, "learning_rate": 3.894927062493073e-05, "loss": 0.0049, "step": 59240 }, { "grad_norm": 0.10865037143230438, "learning_rate": 3.893314542607859e-05, "loss": 0.0058, "step": 59250 }, { "grad_norm": 0.123155876994133, "learning_rate": 3.891702143748103e-05, "loss": 0.0048, "step": 59260 }, { "grad_norm": 0.12456324696540833, "learning_rate": 3.890089866090133e-05, "loss": 0.0076, "step": 59270 }, { "grad_norm": 0.13344378769397736, "learning_rate": 3.888477709810267e-05, "loss": 0.0067, "step": 59280 }, { "grad_norm": 0.12100569903850555, "learning_rate": 3.886865675084806e-05, "loss": 0.0065, "step": 59290 }, { "grad_norm": 0.09918437153100967, "learning_rate": 3.885253762090043e-05, "loss": 0.0047, "step": 59300 }, { "grad_norm": 0.08803583681583405, "learning_rate": 3.8836419710022534e-05, "loss": 0.0059, "step": 59310 }, { "grad_norm": 0.1026521548628807, "learning_rate": 3.882030301997698e-05, "loss": 0.0056, "step": 59320 }, { "grad_norm": 0.12560531497001648, "learning_rate": 3.880418755252629e-05, "loss": 0.006, "step": 59330 }, { "grad_norm": 0.0930902361869812, "learning_rate": 3.87880733094328e-05, "loss": 0.0067, "step": 59340 }, { "grad_norm": 0.12333928048610687, "learning_rate": 3.877196029245877e-05, "loss": 0.0062, "step": 59350 }, { "grad_norm": 0.11066307872533798, "learning_rate": 3.875584850336627e-05, "loss": 0.0074, "step": 59360 }, { "grad_norm": 0.12756121158599854, "learning_rate": 3.873973794391728e-05, "loss": 0.0062, "step": 59370 }, { "grad_norm": 0.10896018892526627, "learning_rate": 3.872362861587361e-05, "loss": 0.0046, "step": 59380 }, { "grad_norm": 0.10178018361330032, "learning_rate": 3.870752052099698e-05, "loss": 0.0055, "step": 59390 }, { "grad_norm": 0.08724460005760193, "learning_rate": 3.8691413661048916e-05, "loss": 0.0064, "step": 59400 }, { "grad_norm": 0.10327573865652084, "learning_rate": 3.867530803779085e-05, "loss": 0.0078, "step": 59410 }, { "grad_norm": 0.10369937866926193, "learning_rate": 3.8659203652984054e-05, "loss": 0.0066, "step": 59420 }, { "grad_norm": 0.07743075489997864, "learning_rate": 3.864310050838972e-05, "loss": 0.0045, "step": 59430 }, { "grad_norm": 0.0877649337053299, "learning_rate": 3.8626998605768835e-05, "loss": 0.0065, "step": 59440 }, { "grad_norm": 0.12412947416305542, "learning_rate": 3.8610897946882294e-05, "loss": 0.0056, "step": 59450 }, { "grad_norm": 0.09889531135559082, "learning_rate": 3.859479853349082e-05, "loss": 0.0056, "step": 59460 }, { "grad_norm": 0.12652382254600525, "learning_rate": 3.857870036735503e-05, "loss": 0.0063, "step": 59470 }, { "grad_norm": 0.1058785617351532, "learning_rate": 3.856260345023538e-05, "loss": 0.0081, "step": 59480 }, { "grad_norm": 0.12121522426605225, "learning_rate": 3.854650778389224e-05, "loss": 0.0066, "step": 59490 }, { "grad_norm": 0.120024174451828, "learning_rate": 3.8530413370085786e-05, "loss": 0.0062, "step": 59500 }, { "grad_norm": 0.09616527706384659, "learning_rate": 3.851432021057608e-05, "loss": 0.0048, "step": 59510 }, { "grad_norm": 0.1002257689833641, "learning_rate": 3.849822830712306e-05, "loss": 0.0077, "step": 59520 }, { "grad_norm": 0.0955672413110733, "learning_rate": 3.848213766148649e-05, "loss": 0.007, "step": 59530 }, { "grad_norm": 0.1144445538520813, "learning_rate": 3.846604827542602e-05, "loss": 0.0062, "step": 59540 }, { "grad_norm": 0.10035938024520874, "learning_rate": 3.844996015070117e-05, "loss": 0.0068, "step": 59550 }, { "grad_norm": 0.0947442278265953, "learning_rate": 3.843387328907132e-05, "loss": 0.0063, "step": 59560 }, { "grad_norm": 0.1397147923707962, "learning_rate": 3.8417787692295696e-05, "loss": 0.007, "step": 59570 }, { "grad_norm": 0.11662054061889648, "learning_rate": 3.84017033621334e-05, "loss": 0.0067, "step": 59580 }, { "grad_norm": 0.15942084789276123, "learning_rate": 3.838562030034338e-05, "loss": 0.0064, "step": 59590 }, { "grad_norm": 0.10395941883325577, "learning_rate": 3.836953850868447e-05, "loss": 0.0067, "step": 59600 }, { "grad_norm": 0.10134189575910568, "learning_rate": 3.835345798891533e-05, "loss": 0.0064, "step": 59610 }, { "grad_norm": 0.08957819640636444, "learning_rate": 3.833737874279452e-05, "loss": 0.0062, "step": 59620 }, { "grad_norm": 0.1276673674583435, "learning_rate": 3.8321300772080433e-05, "loss": 0.0055, "step": 59630 }, { "grad_norm": 0.07570105046033859, "learning_rate": 3.8305224078531345e-05, "loss": 0.0052, "step": 59640 }, { "grad_norm": 0.08044780045747757, "learning_rate": 3.828914866390537e-05, "loss": 0.0057, "step": 59650 }, { "grad_norm": 0.09320417046546936, "learning_rate": 3.827307452996048e-05, "loss": 0.0056, "step": 59660 }, { "grad_norm": 0.0743551105260849, "learning_rate": 3.825700167845454e-05, "loss": 0.0057, "step": 59670 }, { "grad_norm": 0.13001909852027893, "learning_rate": 3.8240930111145214e-05, "loss": 0.0052, "step": 59680 }, { "grad_norm": 0.09777049720287323, "learning_rate": 3.8224859829790116e-05, "loss": 0.0055, "step": 59690 }, { "grad_norm": 0.09082768857479095, "learning_rate": 3.820879083614664e-05, "loss": 0.0057, "step": 59700 }, { "grad_norm": 0.11161638051271439, "learning_rate": 3.819272313197208e-05, "loss": 0.006, "step": 59710 }, { "grad_norm": 0.09318078309297562, "learning_rate": 3.817665671902355e-05, "loss": 0.0052, "step": 59720 }, { "grad_norm": 0.09332044422626495, "learning_rate": 3.816059159905808e-05, "loss": 0.0051, "step": 59730 }, { "grad_norm": 0.09936102479696274, "learning_rate": 3.8144527773832476e-05, "loss": 0.0056, "step": 59740 }, { "grad_norm": 0.09798534214496613, "learning_rate": 3.812846524510352e-05, "loss": 0.005, "step": 59750 }, { "grad_norm": 0.11436246335506439, "learning_rate": 3.811240401462775e-05, "loss": 0.006, "step": 59760 }, { "grad_norm": 0.12271882593631744, "learning_rate": 3.809634408416162e-05, "loss": 0.0069, "step": 59770 }, { "grad_norm": 0.12930120527744293, "learning_rate": 3.8080285455461394e-05, "loss": 0.0057, "step": 59780 }, { "grad_norm": 0.13966645300388336, "learning_rate": 3.806422813028323e-05, "loss": 0.0058, "step": 59790 }, { "grad_norm": 0.08560376614332199, "learning_rate": 3.8048172110383114e-05, "loss": 0.0049, "step": 59800 }, { "grad_norm": 0.11406752467155457, "learning_rate": 3.8032117397516944e-05, "loss": 0.0063, "step": 59810 }, { "grad_norm": 0.11859290301799774, "learning_rate": 3.801606399344042e-05, "loss": 0.0057, "step": 59820 }, { "grad_norm": 0.12445257604122162, "learning_rate": 3.800001189990913e-05, "loss": 0.0072, "step": 59830 }, { "grad_norm": 0.08912872523069382, "learning_rate": 3.798396111867847e-05, "loss": 0.0056, "step": 59840 }, { "grad_norm": 0.09197422862052917, "learning_rate": 3.796791165150378e-05, "loss": 0.0047, "step": 59850 }, { "grad_norm": 0.11976989358663559, "learning_rate": 3.795186350014018e-05, "loss": 0.0066, "step": 59860 }, { "grad_norm": 0.11186239123344421, "learning_rate": 3.793581666634266e-05, "loss": 0.0068, "step": 59870 }, { "grad_norm": 0.07729663699865341, "learning_rate": 3.791977115186609e-05, "loss": 0.0053, "step": 59880 }, { "grad_norm": 0.10846829414367676, "learning_rate": 3.79037269584652e-05, "loss": 0.0056, "step": 59890 }, { "grad_norm": 0.11628066748380661, "learning_rate": 3.7887684087894545e-05, "loss": 0.0061, "step": 59900 }, { "grad_norm": 0.10102793574333191, "learning_rate": 3.787164254190854e-05, "loss": 0.0061, "step": 59910 }, { "grad_norm": 0.10852844268083572, "learning_rate": 3.7855602322261473e-05, "loss": 0.0064, "step": 59920 }, { "grad_norm": 0.09689880907535553, "learning_rate": 3.783956343070746e-05, "loss": 0.005, "step": 59930 }, { "grad_norm": 0.11885309964418411, "learning_rate": 3.7823525869000526e-05, "loss": 0.0057, "step": 59940 }, { "grad_norm": 0.137803852558136, "learning_rate": 3.78074896388945e-05, "loss": 0.0064, "step": 59950 }, { "grad_norm": 0.08718176931142807, "learning_rate": 3.7791454742143075e-05, "loss": 0.0053, "step": 59960 }, { "grad_norm": 0.13815364241600037, "learning_rate": 3.7775421180499795e-05, "loss": 0.0072, "step": 59970 }, { "grad_norm": 0.09240107983350754, "learning_rate": 3.7759388955718086e-05, "loss": 0.0055, "step": 59980 }, { "grad_norm": 0.11859089136123657, "learning_rate": 3.774335806955117e-05, "loss": 0.0062, "step": 59990 }, { "grad_norm": 0.07446644455194473, "learning_rate": 3.7727328523752215e-05, "loss": 0.0059, "step": 60000 }, { "grad_norm": 0.10655452311038971, "learning_rate": 3.771130032007415e-05, "loss": 0.0072, "step": 60010 }, { "grad_norm": 0.09323269128799438, "learning_rate": 3.769527346026982e-05, "loss": 0.0055, "step": 60020 }, { "grad_norm": 0.10444744676351547, "learning_rate": 3.767924794609188e-05, "loss": 0.0065, "step": 60030 }, { "grad_norm": 0.1389145404100418, "learning_rate": 3.7663223779292854e-05, "loss": 0.0054, "step": 60040 }, { "grad_norm": 0.16270224750041962, "learning_rate": 3.764720096162512e-05, "loss": 0.0053, "step": 60050 }, { "grad_norm": 0.06812910735607147, "learning_rate": 3.763117949484092e-05, "loss": 0.0053, "step": 60060 }, { "grad_norm": 0.1246718242764473, "learning_rate": 3.7615159380692344e-05, "loss": 0.0065, "step": 60070 }, { "grad_norm": 0.08776536583900452, "learning_rate": 3.759914062093132e-05, "loss": 0.0049, "step": 60080 }, { "grad_norm": 0.11106762290000916, "learning_rate": 3.7583123217309615e-05, "loss": 0.0052, "step": 60090 }, { "grad_norm": 0.1164855808019638, "learning_rate": 3.7567107171578904e-05, "loss": 0.0066, "step": 60100 }, { "grad_norm": 0.121422678232193, "learning_rate": 3.755109248549066e-05, "loss": 0.0073, "step": 60110 }, { "grad_norm": 0.11699814349412918, "learning_rate": 3.75350791607962e-05, "loss": 0.0053, "step": 60120 }, { "grad_norm": 0.08918627351522446, "learning_rate": 3.751906719924676e-05, "loss": 0.0058, "step": 60130 }, { "grad_norm": 0.12670491635799408, "learning_rate": 3.750305660259337e-05, "loss": 0.0073, "step": 60140 }, { "grad_norm": 0.08766663819551468, "learning_rate": 3.748704737258693e-05, "loss": 0.0053, "step": 60150 }, { "grad_norm": 0.0956849530339241, "learning_rate": 3.747103951097816e-05, "loss": 0.0058, "step": 60160 }, { "grad_norm": 0.08820415288209915, "learning_rate": 3.745503301951768e-05, "loss": 0.005, "step": 60170 }, { "grad_norm": 0.0852598026394844, "learning_rate": 3.743902789995592e-05, "loss": 0.0057, "step": 60180 }, { "grad_norm": 0.08665917813777924, "learning_rate": 3.7423024154043183e-05, "loss": 0.0047, "step": 60190 }, { "grad_norm": 0.08320575207471848, "learning_rate": 3.740702178352963e-05, "loss": 0.0064, "step": 60200 }, { "grad_norm": 0.11256080120801926, "learning_rate": 3.739102079016523e-05, "loss": 0.0059, "step": 60210 }, { "grad_norm": 0.12806819379329681, "learning_rate": 3.7375021175699846e-05, "loss": 0.0067, "step": 60220 }, { "grad_norm": 0.08694878965616226, "learning_rate": 3.735902294188317e-05, "loss": 0.0051, "step": 60230 }, { "grad_norm": 0.10959047824144363, "learning_rate": 3.734302609046473e-05, "loss": 0.0068, "step": 60240 }, { "grad_norm": 0.10951108485460281, "learning_rate": 3.732703062319391e-05, "loss": 0.0057, "step": 60250 }, { "grad_norm": 0.06898360699415207, "learning_rate": 3.731103654181999e-05, "loss": 0.007, "step": 60260 }, { "grad_norm": 0.09675011038780212, "learning_rate": 3.729504384809203e-05, "loss": 0.0064, "step": 60270 }, { "grad_norm": 0.08703022450208664, "learning_rate": 3.727905254375898e-05, "loss": 0.0054, "step": 60280 }, { "grad_norm": 0.09308961033821106, "learning_rate": 3.7263062630569614e-05, "loss": 0.0055, "step": 60290 }, { "grad_norm": 0.09285279363393784, "learning_rate": 3.724707411027256e-05, "loss": 0.0053, "step": 60300 }, { "grad_norm": 0.10243218392133713, "learning_rate": 3.723108698461631e-05, "loss": 0.0052, "step": 60310 }, { "grad_norm": 0.13703571259975433, "learning_rate": 3.72151012553492e-05, "loss": 0.0052, "step": 60320 }, { "grad_norm": 0.09408984333276749, "learning_rate": 3.719911692421939e-05, "loss": 0.0055, "step": 60330 }, { "grad_norm": 0.10790714621543884, "learning_rate": 3.71831339929749e-05, "loss": 0.0057, "step": 60340 }, { "grad_norm": 0.09419838339090347, "learning_rate": 3.716715246336362e-05, "loss": 0.0062, "step": 60350 }, { "grad_norm": 0.13101890683174133, "learning_rate": 3.715117233713324e-05, "loss": 0.007, "step": 60360 }, { "grad_norm": 0.1119229793548584, "learning_rate": 3.713519361603135e-05, "loss": 0.0094, "step": 60370 }, { "grad_norm": 0.07704105228185654, "learning_rate": 3.711921630180532e-05, "loss": 0.0071, "step": 60380 }, { "grad_norm": 0.10984918475151062, "learning_rate": 3.710324039620245e-05, "loss": 0.0083, "step": 60390 }, { "grad_norm": 0.10326418280601501, "learning_rate": 3.708726590096982e-05, "loss": 0.0059, "step": 60400 }, { "grad_norm": 0.09732617437839508, "learning_rate": 3.707129281785437e-05, "loss": 0.0061, "step": 60410 }, { "grad_norm": 0.09905219823122025, "learning_rate": 3.705532114860291e-05, "loss": 0.0051, "step": 60420 }, { "grad_norm": 0.10013336688280106, "learning_rate": 3.703935089496207e-05, "loss": 0.0085, "step": 60430 }, { "grad_norm": 0.08357701450586319, "learning_rate": 3.702338205867829e-05, "loss": 0.0086, "step": 60440 }, { "grad_norm": 0.07999374717473984, "learning_rate": 3.7007414641497976e-05, "loss": 0.0061, "step": 60450 }, { "grad_norm": 0.10455209016799927, "learning_rate": 3.699144864516726e-05, "loss": 0.0059, "step": 60460 }, { "grad_norm": 0.0830470472574234, "learning_rate": 3.6975484071432153e-05, "loss": 0.0074, "step": 60470 }, { "grad_norm": 0.09043987840414047, "learning_rate": 3.695952092203854e-05, "loss": 0.0051, "step": 60480 }, { "grad_norm": 0.12024213373661041, "learning_rate": 3.69435591987321e-05, "loss": 0.0056, "step": 60490 }, { "grad_norm": 0.08564987778663635, "learning_rate": 3.6927598903258374e-05, "loss": 0.0047, "step": 60500 }, { "grad_norm": 0.1449645310640335, "learning_rate": 3.691164003736281e-05, "loss": 0.0067, "step": 60510 }, { "grad_norm": 0.13217893242835999, "learning_rate": 3.689568260279061e-05, "loss": 0.0057, "step": 60520 }, { "grad_norm": 0.1356850564479828, "learning_rate": 3.687972660128686e-05, "loss": 0.0062, "step": 60530 }, { "grad_norm": 0.10479087382555008, "learning_rate": 3.686377203459648e-05, "loss": 0.0064, "step": 60540 }, { "grad_norm": 0.08423815667629242, "learning_rate": 3.6847818904464226e-05, "loss": 0.0048, "step": 60550 }, { "grad_norm": 0.09129633754491806, "learning_rate": 3.683186721263474e-05, "loss": 0.0062, "step": 60560 }, { "grad_norm": 0.10648217052221298, "learning_rate": 3.6815916960852435e-05, "loss": 0.0072, "step": 60570 }, { "grad_norm": 0.13020111620426178, "learning_rate": 3.679996815086165e-05, "loss": 0.0057, "step": 60580 }, { "grad_norm": 0.0959603562951088, "learning_rate": 3.678402078440649e-05, "loss": 0.0056, "step": 60590 }, { "grad_norm": 0.12122093886137009, "learning_rate": 3.676807486323096e-05, "loss": 0.0049, "step": 60600 }, { "grad_norm": 0.07983621209859848, "learning_rate": 3.6752130389078864e-05, "loss": 0.0051, "step": 60610 }, { "grad_norm": 0.11104491353034973, "learning_rate": 3.673618736369388e-05, "loss": 0.0071, "step": 60620 }, { "grad_norm": 0.09617263078689575, "learning_rate": 3.6720245788819474e-05, "loss": 0.0069, "step": 60630 }, { "grad_norm": 0.13393914699554443, "learning_rate": 3.6704305666199044e-05, "loss": 0.0059, "step": 60640 }, { "grad_norm": 0.12225785106420517, "learning_rate": 3.668836699757576e-05, "loss": 0.0078, "step": 60650 }, { "grad_norm": 0.10963711887598038, "learning_rate": 3.667242978469265e-05, "loss": 0.0047, "step": 60660 }, { "grad_norm": 0.14397290349006653, "learning_rate": 3.6656494029292575e-05, "loss": 0.006, "step": 60670 }, { "grad_norm": 0.11237888038158417, "learning_rate": 3.664055973311825e-05, "loss": 0.0059, "step": 60680 }, { "grad_norm": 0.09187003970146179, "learning_rate": 3.662462689791221e-05, "loss": 0.007, "step": 60690 }, { "grad_norm": 0.09289684891700745, "learning_rate": 3.660869552541689e-05, "loss": 0.0052, "step": 60700 }, { "grad_norm": 0.15152175724506378, "learning_rate": 3.659276561737448e-05, "loss": 0.0062, "step": 60710 }, { "grad_norm": 0.08710607141256332, "learning_rate": 3.657683717552708e-05, "loss": 0.0058, "step": 60720 }, { "grad_norm": 0.10866423696279526, "learning_rate": 3.6560910201616574e-05, "loss": 0.0058, "step": 60730 }, { "grad_norm": 0.10873210430145264, "learning_rate": 3.654498469738472e-05, "loss": 0.0063, "step": 60740 }, { "grad_norm": 0.09474363178014755, "learning_rate": 3.6529060664573126e-05, "loss": 0.0073, "step": 60750 }, { "grad_norm": 0.10301627963781357, "learning_rate": 3.6513138104923176e-05, "loss": 0.0074, "step": 60760 }, { "grad_norm": 0.11111614108085632, "learning_rate": 3.6497217020176176e-05, "loss": 0.0049, "step": 60770 }, { "grad_norm": 0.10326512902975082, "learning_rate": 3.648129741207323e-05, "loss": 0.0064, "step": 60780 }, { "grad_norm": 0.15862521529197693, "learning_rate": 3.646537928235527e-05, "loss": 0.007, "step": 60790 }, { "grad_norm": 0.11922838538885117, "learning_rate": 3.6449462632763075e-05, "loss": 0.006, "step": 60800 }, { "grad_norm": 0.09564152359962463, "learning_rate": 3.643354746503729e-05, "loss": 0.0057, "step": 60810 }, { "grad_norm": 0.09808968007564545, "learning_rate": 3.6417633780918345e-05, "loss": 0.0062, "step": 60820 }, { "grad_norm": 0.10176081955432892, "learning_rate": 3.640172158214655e-05, "loss": 0.0062, "step": 60830 }, { "grad_norm": 0.08235935121774673, "learning_rate": 3.6385810870462035e-05, "loss": 0.0052, "step": 60840 }, { "grad_norm": 0.1510922908782959, "learning_rate": 3.636990164760479e-05, "loss": 0.0059, "step": 60850 }, { "grad_norm": 0.10306186228990555, "learning_rate": 3.6353993915314596e-05, "loss": 0.006, "step": 60860 }, { "grad_norm": 0.09088820964097977, "learning_rate": 3.6338087675331124e-05, "loss": 0.0048, "step": 60870 }, { "grad_norm": 0.08577850461006165, "learning_rate": 3.632218292939383e-05, "loss": 0.0061, "step": 60880 }, { "grad_norm": 0.1318112015724182, "learning_rate": 3.6306279679242033e-05, "loss": 0.0065, "step": 60890 }, { "grad_norm": 0.14387668669223785, "learning_rate": 3.629037792661492e-05, "loss": 0.0062, "step": 60900 }, { "grad_norm": 0.15897974371910095, "learning_rate": 3.627447767325147e-05, "loss": 0.0056, "step": 60910 }, { "grad_norm": 0.12239181250333786, "learning_rate": 3.62585789208905e-05, "loss": 0.005, "step": 60920 }, { "grad_norm": 0.15219280123710632, "learning_rate": 3.624268167127068e-05, "loss": 0.0053, "step": 60930 }, { "grad_norm": 0.11290594190359116, "learning_rate": 3.622678592613051e-05, "loss": 0.0058, "step": 60940 }, { "grad_norm": 0.11952400952577591, "learning_rate": 3.6210891687208304e-05, "loss": 0.007, "step": 60950 }, { "grad_norm": 0.0907423123717308, "learning_rate": 3.619499895624228e-05, "loss": 0.0057, "step": 60960 }, { "grad_norm": 0.12188044935464859, "learning_rate": 3.61791077349704e-05, "loss": 0.0066, "step": 60970 }, { "grad_norm": 0.11006130278110504, "learning_rate": 3.616321802513053e-05, "loss": 0.0043, "step": 60980 }, { "grad_norm": 0.10209494829177856, "learning_rate": 3.6147329828460325e-05, "loss": 0.0045, "step": 60990 }, { "grad_norm": 0.10878945142030716, "learning_rate": 3.6131443146697306e-05, "loss": 0.0064, "step": 61000 }, { "grad_norm": 0.09197011590003967, "learning_rate": 3.611555798157879e-05, "loss": 0.0065, "step": 61010 }, { "grad_norm": 0.09913847595453262, "learning_rate": 3.6099674334842e-05, "loss": 0.0072, "step": 61020 }, { "grad_norm": 0.14608731865882874, "learning_rate": 3.6083792208223934e-05, "loss": 0.0075, "step": 61030 }, { "grad_norm": 0.1418754607439041, "learning_rate": 3.606791160346142e-05, "loss": 0.0068, "step": 61040 }, { "grad_norm": 0.09568615257740021, "learning_rate": 3.6052032522291134e-05, "loss": 0.0068, "step": 61050 }, { "grad_norm": 0.1360521763563156, "learning_rate": 3.603615496644962e-05, "loss": 0.0058, "step": 61060 }, { "grad_norm": 0.10921019315719604, "learning_rate": 3.60202789376732e-05, "loss": 0.0058, "step": 61070 }, { "grad_norm": 0.16341255605220795, "learning_rate": 3.6004404437698036e-05, "loss": 0.0064, "step": 61080 }, { "grad_norm": 0.1601221263408661, "learning_rate": 3.5988531468260184e-05, "loss": 0.0057, "step": 61090 }, { "grad_norm": 0.10792120546102524, "learning_rate": 3.597266003109546e-05, "loss": 0.005, "step": 61100 }, { "grad_norm": 0.1410602480173111, "learning_rate": 3.5956790127939557e-05, "loss": 0.0057, "step": 61110 }, { "grad_norm": 0.07817249000072479, "learning_rate": 3.594092176052796e-05, "loss": 0.0053, "step": 61120 }, { "grad_norm": 0.11602982133626938, "learning_rate": 3.592505493059603e-05, "loss": 0.0052, "step": 61130 }, { "grad_norm": 0.10793928802013397, "learning_rate": 3.5909189639878905e-05, "loss": 0.0061, "step": 61140 }, { "grad_norm": 0.1208905428647995, "learning_rate": 3.5893325890111626e-05, "loss": 0.0066, "step": 61150 }, { "grad_norm": 0.12716200947761536, "learning_rate": 3.5877463683029034e-05, "loss": 0.0059, "step": 61160 }, { "grad_norm": 0.12092209607362747, "learning_rate": 3.586160302036578e-05, "loss": 0.0053, "step": 61170 }, { "grad_norm": 0.1371377855539322, "learning_rate": 3.584574390385636e-05, "loss": 0.0091, "step": 61180 }, { "grad_norm": 0.1034461259841919, "learning_rate": 3.582988633523511e-05, "loss": 0.005, "step": 61190 }, { "grad_norm": 0.08704489469528198, "learning_rate": 3.581403031623616e-05, "loss": 0.0062, "step": 61200 }, { "grad_norm": 0.12118525058031082, "learning_rate": 3.579817584859354e-05, "loss": 0.0077, "step": 61210 }, { "grad_norm": 0.1097278743982315, "learning_rate": 3.578232293404107e-05, "loss": 0.0058, "step": 61220 }, { "grad_norm": 0.12096497416496277, "learning_rate": 3.5766471574312385e-05, "loss": 0.0065, "step": 61230 }, { "grad_norm": 0.12292058020830154, "learning_rate": 3.5750621771140964e-05, "loss": 0.0049, "step": 61240 }, { "grad_norm": 0.13812826573848724, "learning_rate": 3.5734773526260115e-05, "loss": 0.0062, "step": 61250 }, { "grad_norm": 0.11392564326524734, "learning_rate": 3.5718926841402996e-05, "loss": 0.0074, "step": 61260 }, { "grad_norm": 0.11064722388982773, "learning_rate": 3.5703081718302555e-05, "loss": 0.0072, "step": 61270 }, { "grad_norm": 0.12178954482078552, "learning_rate": 3.568723815869162e-05, "loss": 0.0058, "step": 61280 }, { "grad_norm": 0.11875439435243607, "learning_rate": 3.5671396164302777e-05, "loss": 0.0056, "step": 61290 }, { "grad_norm": 0.09567902982234955, "learning_rate": 3.565555573686853e-05, "loss": 0.0072, "step": 61300 }, { "grad_norm": 0.07293132692575455, "learning_rate": 3.563971687812113e-05, "loss": 0.0046, "step": 61310 }, { "grad_norm": 0.10181144624948502, "learning_rate": 3.5623879589792706e-05, "loss": 0.0048, "step": 61320 }, { "grad_norm": 0.11834069341421127, "learning_rate": 3.5608043873615185e-05, "loss": 0.0051, "step": 61330 }, { "grad_norm": 0.09753533452749252, "learning_rate": 3.559220973132035e-05, "loss": 0.0061, "step": 61340 }, { "grad_norm": 0.10105733573436737, "learning_rate": 3.557637716463981e-05, "loss": 0.0064, "step": 61350 }, { "grad_norm": 0.09639612585306168, "learning_rate": 3.556054617530497e-05, "loss": 0.0063, "step": 61360 }, { "grad_norm": 0.07558410614728928, "learning_rate": 3.5544716765047106e-05, "loss": 0.0048, "step": 61370 }, { "grad_norm": 0.08827099949121475, "learning_rate": 3.5528888935597274e-05, "loss": 0.0062, "step": 61380 }, { "grad_norm": 0.09965735673904419, "learning_rate": 3.551306268868636e-05, "loss": 0.0051, "step": 61390 }, { "grad_norm": 0.07689246535301208, "learning_rate": 3.5497238026045166e-05, "loss": 0.0077, "step": 61400 }, { "grad_norm": 0.11576641350984573, "learning_rate": 3.548141494940421e-05, "loss": 0.0052, "step": 61410 }, { "grad_norm": 0.09902197122573853, "learning_rate": 3.546559346049388e-05, "loss": 0.0059, "step": 61420 }, { "grad_norm": 0.1297139674425125, "learning_rate": 3.544977356104441e-05, "loss": 0.0055, "step": 61430 }, { "grad_norm": 0.09273114055395126, "learning_rate": 3.543395525278582e-05, "loss": 0.0061, "step": 61440 }, { "grad_norm": 0.10696340352296829, "learning_rate": 3.5418138537447985e-05, "loss": 0.0063, "step": 61450 }, { "grad_norm": 0.14855380356311798, "learning_rate": 3.5402323416760574e-05, "loss": 0.0067, "step": 61460 }, { "grad_norm": 0.16061244904994965, "learning_rate": 3.538650989245316e-05, "loss": 0.0065, "step": 61470 }, { "grad_norm": 0.10625765472650528, "learning_rate": 3.537069796625504e-05, "loss": 0.0062, "step": 61480 }, { "grad_norm": 0.09493063390254974, "learning_rate": 3.53548876398954e-05, "loss": 0.0063, "step": 61490 }, { "grad_norm": 0.08883731067180634, "learning_rate": 3.5339078915103214e-05, "loss": 0.0056, "step": 61500 }, { "grad_norm": 0.18457692861557007, "learning_rate": 3.532327179360733e-05, "loss": 0.0054, "step": 61510 }, { "grad_norm": 0.09497077018022537, "learning_rate": 3.530746627713636e-05, "loss": 0.0053, "step": 61520 }, { "grad_norm": 0.16097614169120789, "learning_rate": 3.5291662367418805e-05, "loss": 0.0066, "step": 61530 }, { "grad_norm": 0.1256483495235443, "learning_rate": 3.527586006618293e-05, "loss": 0.0069, "step": 61540 }, { "grad_norm": 0.10753141343593597, "learning_rate": 3.526005937515687e-05, "loss": 0.0057, "step": 61550 }, { "grad_norm": 0.08107404410839081, "learning_rate": 3.524426029606856e-05, "loss": 0.0046, "step": 61560 }, { "grad_norm": 0.10541848838329315, "learning_rate": 3.5228462830645746e-05, "loss": 0.0048, "step": 61570 }, { "grad_norm": 0.08646702021360397, "learning_rate": 3.5212666980616036e-05, "loss": 0.0066, "step": 61580 }, { "grad_norm": 0.14430874586105347, "learning_rate": 3.519687274770681e-05, "loss": 0.005, "step": 61590 }, { "grad_norm": 0.11265552043914795, "learning_rate": 3.518108013364535e-05, "loss": 0.0052, "step": 61600 }, { "grad_norm": 0.09418538957834244, "learning_rate": 3.516528914015868e-05, "loss": 0.0047, "step": 61610 }, { "grad_norm": 0.10827504843473434, "learning_rate": 3.5149499768973694e-05, "loss": 0.0062, "step": 61620 }, { "grad_norm": 0.09334459900856018, "learning_rate": 3.513371202181708e-05, "loss": 0.0047, "step": 61630 }, { "grad_norm": 0.08266305178403854, "learning_rate": 3.511792590041537e-05, "loss": 0.0047, "step": 61640 }, { "grad_norm": 0.10171569883823395, "learning_rate": 3.510214140649488e-05, "loss": 0.0049, "step": 61650 }, { "grad_norm": 0.2763062119483948, "learning_rate": 3.508635854178184e-05, "loss": 0.0052, "step": 61660 }, { "grad_norm": 0.10407113283872604, "learning_rate": 3.50705773080022e-05, "loss": 0.0057, "step": 61670 }, { "grad_norm": 0.08716286718845367, "learning_rate": 3.505479770688178e-05, "loss": 0.0087, "step": 61680 }, { "grad_norm": 0.1287417858839035, "learning_rate": 3.503901974014622e-05, "loss": 0.0048, "step": 61690 }, { "grad_norm": 0.0912509486079216, "learning_rate": 3.502324340952095e-05, "loss": 0.0054, "step": 61700 }, { "grad_norm": 0.07942947745323181, "learning_rate": 3.500746871673126e-05, "loss": 0.0057, "step": 61710 }, { "grad_norm": 0.0925874263048172, "learning_rate": 3.499169566350227e-05, "loss": 0.0058, "step": 61720 }, { "grad_norm": 0.11692814528942108, "learning_rate": 3.4975924251558876e-05, "loss": 0.0073, "step": 61730 }, { "grad_norm": 0.11249503493309021, "learning_rate": 3.496015448262582e-05, "loss": 0.0052, "step": 61740 }, { "grad_norm": 0.09729213267564774, "learning_rate": 3.494438635842764e-05, "loss": 0.0063, "step": 61750 }, { "grad_norm": 0.10964075475931168, "learning_rate": 3.492861988068874e-05, "loss": 0.0061, "step": 61760 }, { "grad_norm": 0.07334958016872406, "learning_rate": 3.491285505113332e-05, "loss": 0.0047, "step": 61770 }, { "grad_norm": 0.09292836487293243, "learning_rate": 3.4897091871485376e-05, "loss": 0.0047, "step": 61780 }, { "grad_norm": 0.1624806672334671, "learning_rate": 3.488133034346875e-05, "loss": 0.0057, "step": 61790 }, { "grad_norm": 0.10004787147045135, "learning_rate": 3.486557046880713e-05, "loss": 0.0058, "step": 61800 }, { "grad_norm": 0.1867230236530304, "learning_rate": 3.484981224922395e-05, "loss": 0.0061, "step": 61810 }, { "grad_norm": 0.1041266992688179, "learning_rate": 3.4834055686442544e-05, "loss": 0.0073, "step": 61820 }, { "grad_norm": 0.13361656665802002, "learning_rate": 3.4818300782186e-05, "loss": 0.0083, "step": 61830 }, { "grad_norm": 0.0878835991024971, "learning_rate": 3.4802547538177236e-05, "loss": 0.0057, "step": 61840 }, { "grad_norm": 0.09018256515264511, "learning_rate": 3.478679595613905e-05, "loss": 0.0058, "step": 61850 }, { "grad_norm": 0.09468291699886322, "learning_rate": 3.4771046037793975e-05, "loss": 0.0054, "step": 61860 }, { "grad_norm": 0.10388767719268799, "learning_rate": 3.4755297784864416e-05, "loss": 0.0063, "step": 61870 }, { "grad_norm": 0.09912009537220001, "learning_rate": 3.473955119907257e-05, "loss": 0.0056, "step": 61880 }, { "grad_norm": 0.05472316965460777, "learning_rate": 3.472380628214046e-05, "loss": 0.0047, "step": 61890 }, { "grad_norm": 0.10068140923976898, "learning_rate": 3.4708063035789906e-05, "loss": 0.0058, "step": 61900 }, { "grad_norm": 0.10170300304889679, "learning_rate": 3.4692321461742614e-05, "loss": 0.0058, "step": 61910 }, { "grad_norm": 0.12877576053142548, "learning_rate": 3.4676581561720023e-05, "loss": 0.0043, "step": 61920 }, { "grad_norm": 0.09133081883192062, "learning_rate": 3.466084333744344e-05, "loss": 0.0053, "step": 61930 }, { "grad_norm": 0.11236114054918289, "learning_rate": 3.4645106790633964e-05, "loss": 0.0063, "step": 61940 }, { "grad_norm": 0.11238575726747513, "learning_rate": 3.4629371923012526e-05, "loss": 0.0052, "step": 61950 }, { "grad_norm": 0.10428775101900101, "learning_rate": 3.4613638736299856e-05, "loss": 0.0061, "step": 61960 }, { "grad_norm": 0.10259920358657837, "learning_rate": 3.459790723221652e-05, "loss": 0.0058, "step": 61970 }, { "grad_norm": 0.13606396317481995, "learning_rate": 3.45821774124829e-05, "loss": 0.005, "step": 61980 }, { "grad_norm": 0.09786480665206909, "learning_rate": 3.456644927881918e-05, "loss": 0.0052, "step": 61990 }, { "grad_norm": 0.08704452961683273, "learning_rate": 3.4550722832945354e-05, "loss": 0.0049, "step": 62000 }, { "grad_norm": 0.10287962853908539, "learning_rate": 3.453499807658126e-05, "loss": 0.0051, "step": 62010 }, { "grad_norm": 0.0777563527226448, "learning_rate": 3.451927501144653e-05, "loss": 0.0053, "step": 62020 }, { "grad_norm": 0.08613824099302292, "learning_rate": 3.450355363926059e-05, "loss": 0.0056, "step": 62030 }, { "grad_norm": 0.09821005910634995, "learning_rate": 3.448783396174273e-05, "loss": 0.0049, "step": 62040 }, { "grad_norm": 0.1103195995092392, "learning_rate": 3.4472115980612034e-05, "loss": 0.0065, "step": 62050 }, { "grad_norm": 0.06076015904545784, "learning_rate": 3.4456399697587396e-05, "loss": 0.0045, "step": 62060 }, { "grad_norm": 0.16730272769927979, "learning_rate": 3.4440685114387506e-05, "loss": 0.0072, "step": 62070 }, { "grad_norm": 0.10441739112138748, "learning_rate": 3.442497223273091e-05, "loss": 0.0055, "step": 62080 }, { "grad_norm": 0.08347618579864502, "learning_rate": 3.4409261054335926e-05, "loss": 0.0056, "step": 62090 }, { "grad_norm": 0.12498229742050171, "learning_rate": 3.439355158092069e-05, "loss": 0.0051, "step": 62100 }, { "grad_norm": 0.09760264307260513, "learning_rate": 3.437784381420322e-05, "loss": 0.0057, "step": 62110 }, { "grad_norm": 0.16037246584892273, "learning_rate": 3.436213775590126e-05, "loss": 0.0051, "step": 62120 }, { "grad_norm": 0.13934087753295898, "learning_rate": 3.4346433407732405e-05, "loss": 0.0048, "step": 62130 }, { "grad_norm": 0.089459627866745, "learning_rate": 3.433073077141405e-05, "loss": 0.0058, "step": 62140 }, { "grad_norm": 0.11113452166318893, "learning_rate": 3.431502984866343e-05, "loss": 0.0055, "step": 62150 }, { "grad_norm": 0.10644355416297913, "learning_rate": 3.429933064119753e-05, "loss": 0.0053, "step": 62160 }, { "grad_norm": 0.09875094145536423, "learning_rate": 3.4283633150733255e-05, "loss": 0.0047, "step": 62170 }, { "grad_norm": 0.09441158920526505, "learning_rate": 3.4267937378987225e-05, "loss": 0.006, "step": 62180 }, { "grad_norm": 0.11947499960660934, "learning_rate": 3.425224332767592e-05, "loss": 0.0058, "step": 62190 }, { "grad_norm": 0.1209132969379425, "learning_rate": 3.4236550998515595e-05, "loss": 0.0061, "step": 62200 }, { "grad_norm": 0.07314227521419525, "learning_rate": 3.422086039322235e-05, "loss": 0.0051, "step": 62210 }, { "grad_norm": 0.10986104607582092, "learning_rate": 3.420517151351209e-05, "loss": 0.0058, "step": 62220 }, { "grad_norm": 0.12104925513267517, "learning_rate": 3.418948436110053e-05, "loss": 0.0068, "step": 62230 }, { "grad_norm": 0.08890191465616226, "learning_rate": 3.417379893770319e-05, "loss": 0.0051, "step": 62240 }, { "grad_norm": 0.11777760088443756, "learning_rate": 3.41581152450354e-05, "loss": 0.0058, "step": 62250 }, { "grad_norm": 0.09100904315710068, "learning_rate": 3.414243328481231e-05, "loss": 0.0046, "step": 62260 }, { "grad_norm": 0.11902303248643875, "learning_rate": 3.412675305874888e-05, "loss": 0.0059, "step": 62270 }, { "grad_norm": 0.10353318601846695, "learning_rate": 3.4111074568559876e-05, "loss": 0.0061, "step": 62280 }, { "grad_norm": 0.09768933802843094, "learning_rate": 3.4095397815959846e-05, "loss": 0.0067, "step": 62290 }, { "grad_norm": 0.10038170963525772, "learning_rate": 3.407972280266322e-05, "loss": 0.0054, "step": 62300 }, { "grad_norm": 0.10002917051315308, "learning_rate": 3.4064049530384166e-05, "loss": 0.0046, "step": 62310 }, { "grad_norm": 0.0858633890748024, "learning_rate": 3.40483780008367e-05, "loss": 0.0053, "step": 62320 }, { "grad_norm": 0.10297329723834991, "learning_rate": 3.4032708215734635e-05, "loss": 0.0067, "step": 62330 }, { "grad_norm": 0.12011414766311646, "learning_rate": 3.4017040176791584e-05, "loss": 0.0065, "step": 62340 }, { "grad_norm": 0.07067780196666718, "learning_rate": 3.4001373885720976e-05, "loss": 0.0055, "step": 62350 }, { "grad_norm": 0.11216568201780319, "learning_rate": 3.398570934423608e-05, "loss": 0.0053, "step": 62360 }, { "grad_norm": 0.11170762032270432, "learning_rate": 3.397004655404994e-05, "loss": 0.0055, "step": 62370 }, { "grad_norm": 0.0993647500872612, "learning_rate": 3.395438551687541e-05, "loss": 0.0059, "step": 62380 }, { "grad_norm": 0.10870376974344254, "learning_rate": 3.393872623442515e-05, "loss": 0.0069, "step": 62390 }, { "grad_norm": 0.07147697359323502, "learning_rate": 3.392306870841164e-05, "loss": 0.0053, "step": 62400 }, { "grad_norm": 0.0773923471570015, "learning_rate": 3.390741294054715e-05, "loss": 0.0042, "step": 62410 }, { "grad_norm": 0.0842074528336525, "learning_rate": 3.38917589325438e-05, "loss": 0.0053, "step": 62420 }, { "grad_norm": 0.10395722836256027, "learning_rate": 3.387610668611349e-05, "loss": 0.007, "step": 62430 }, { "grad_norm": 0.10837456583976746, "learning_rate": 3.38604562029679e-05, "loss": 0.0064, "step": 62440 }, { "grad_norm": 0.15859301388263702, "learning_rate": 3.384480748481855e-05, "loss": 0.0072, "step": 62450 }, { "grad_norm": 0.12502050399780273, "learning_rate": 3.382916053337676e-05, "loss": 0.0068, "step": 62460 }, { "grad_norm": 0.08552031964063644, "learning_rate": 3.381351535035366e-05, "loss": 0.0044, "step": 62470 }, { "grad_norm": 0.0807204395532608, "learning_rate": 3.379787193746019e-05, "loss": 0.0053, "step": 62480 }, { "grad_norm": 0.07780604809522629, "learning_rate": 3.378223029640709e-05, "loss": 0.0071, "step": 62490 }, { "grad_norm": 0.09410929679870605, "learning_rate": 3.376659042890489e-05, "loss": 0.0055, "step": 62500 }, { "grad_norm": 0.08467679470777512, "learning_rate": 3.375095233666397e-05, "loss": 0.006, "step": 62510 }, { "grad_norm": 0.05062735080718994, "learning_rate": 3.373531602139447e-05, "loss": 0.0048, "step": 62520 }, { "grad_norm": 0.09364766627550125, "learning_rate": 3.371968148480635e-05, "loss": 0.0049, "step": 62530 }, { "grad_norm": 0.13241344690322876, "learning_rate": 3.370404872860937e-05, "loss": 0.0056, "step": 62540 }, { "grad_norm": 0.10747424513101578, "learning_rate": 3.3688417754513145e-05, "loss": 0.0058, "step": 62550 }, { "grad_norm": 0.11189000308513641, "learning_rate": 3.367278856422703e-05, "loss": 0.0061, "step": 62560 }, { "grad_norm": 0.07629857212305069, "learning_rate": 3.365716115946021e-05, "loss": 0.0052, "step": 62570 }, { "grad_norm": 0.07994486391544342, "learning_rate": 3.3641535541921666e-05, "loss": 0.0052, "step": 62580 }, { "grad_norm": 0.11250322312116623, "learning_rate": 3.36259117133202e-05, "loss": 0.0061, "step": 62590 }, { "grad_norm": 0.09037989377975464, "learning_rate": 3.36102896753644e-05, "loss": 0.0051, "step": 62600 }, { "grad_norm": 0.13443408906459808, "learning_rate": 3.35946694297627e-05, "loss": 0.0062, "step": 62610 }, { "grad_norm": 0.11756874620914459, "learning_rate": 3.357905097822328e-05, "loss": 0.0057, "step": 62620 }, { "grad_norm": 0.11904001981019974, "learning_rate": 3.356343432245416e-05, "loss": 0.0063, "step": 62630 }, { "grad_norm": 0.097943976521492, "learning_rate": 3.354781946416315e-05, "loss": 0.0058, "step": 62640 }, { "grad_norm": 0.13131260871887207, "learning_rate": 3.3532206405057864e-05, "loss": 0.0051, "step": 62650 }, { "grad_norm": 0.10270214825868607, "learning_rate": 3.351659514684573e-05, "loss": 0.0045, "step": 62660 }, { "grad_norm": 0.10715050995349884, "learning_rate": 3.350098569123395e-05, "loss": 0.008, "step": 62670 }, { "grad_norm": 0.07880665361881256, "learning_rate": 3.348537803992958e-05, "loss": 0.0054, "step": 62680 }, { "grad_norm": 0.1452426314353943, "learning_rate": 3.346977219463945e-05, "loss": 0.0056, "step": 62690 }, { "grad_norm": 0.10012497752904892, "learning_rate": 3.345416815707017e-05, "loss": 0.0049, "step": 62700 }, { "grad_norm": 0.1158832237124443, "learning_rate": 3.3438565928928185e-05, "loss": 0.0096, "step": 62710 }, { "grad_norm": 0.44107648730278015, "learning_rate": 3.3422965511919736e-05, "loss": 0.0057, "step": 62720 }, { "grad_norm": 0.0908970832824707, "learning_rate": 3.3407366907750846e-05, "loss": 0.0063, "step": 62730 }, { "grad_norm": 0.1000109314918518, "learning_rate": 3.339177011812737e-05, "loss": 0.0059, "step": 62740 }, { "grad_norm": 0.10705020278692245, "learning_rate": 3.3376175144754935e-05, "loss": 0.009, "step": 62750 }, { "grad_norm": 0.09379396587610245, "learning_rate": 3.3360581989339004e-05, "loss": 0.0062, "step": 62760 }, { "grad_norm": 0.10797853022813797, "learning_rate": 3.3344990653584804e-05, "loss": 0.0066, "step": 62770 }, { "grad_norm": 0.07921292632818222, "learning_rate": 3.332940113919739e-05, "loss": 0.0045, "step": 62780 }, { "grad_norm": 0.13731862604618073, "learning_rate": 3.331381344788161e-05, "loss": 0.0062, "step": 62790 }, { "grad_norm": 0.1039913222193718, "learning_rate": 3.3298227581342076e-05, "loss": 0.0058, "step": 62800 }, { "grad_norm": 0.12375742197036743, "learning_rate": 3.328264354128329e-05, "loss": 0.0047, "step": 62810 }, { "grad_norm": 0.11380947381258011, "learning_rate": 3.3267061329409465e-05, "loss": 0.0057, "step": 62820 }, { "grad_norm": 0.14197903871536255, "learning_rate": 3.325148094742466e-05, "loss": 0.005, "step": 62830 }, { "grad_norm": 0.11421457678079605, "learning_rate": 3.323590239703271e-05, "loss": 0.0061, "step": 62840 }, { "grad_norm": 0.09896033257246017, "learning_rate": 3.322032567993728e-05, "loss": 0.0052, "step": 62850 }, { "grad_norm": 0.08462632447481155, "learning_rate": 3.3204750797841774e-05, "loss": 0.0045, "step": 62860 }, { "grad_norm": 0.12609311938285828, "learning_rate": 3.31891777524495e-05, "loss": 0.0071, "step": 62870 }, { "grad_norm": 0.10693735629320145, "learning_rate": 3.317360654546347e-05, "loss": 0.0059, "step": 62880 }, { "grad_norm": 0.13051852583885193, "learning_rate": 3.3158037178586534e-05, "loss": 0.0044, "step": 62890 }, { "grad_norm": 0.09902596473693848, "learning_rate": 3.3142469653521316e-05, "loss": 0.0051, "step": 62900 }, { "grad_norm": 0.06528617441654205, "learning_rate": 3.3126903971970284e-05, "loss": 0.0053, "step": 62910 }, { "grad_norm": 0.10662052035331726, "learning_rate": 3.311134013563564e-05, "loss": 0.0063, "step": 62920 }, { "grad_norm": 0.11549826711416245, "learning_rate": 3.309577814621947e-05, "loss": 0.005, "step": 62930 }, { "grad_norm": 0.08462731540203094, "learning_rate": 3.308021800542359e-05, "loss": 0.0051, "step": 62940 }, { "grad_norm": 0.09481456130743027, "learning_rate": 3.306465971494963e-05, "loss": 0.0055, "step": 62950 }, { "grad_norm": 0.1054568737745285, "learning_rate": 3.304910327649901e-05, "loss": 0.0056, "step": 62960 }, { "grad_norm": 0.10492590814828873, "learning_rate": 3.303354869177297e-05, "loss": 0.0049, "step": 62970 }, { "grad_norm": 0.11059510707855225, "learning_rate": 3.301799596247255e-05, "loss": 0.0064, "step": 62980 }, { "grad_norm": 0.14067962765693665, "learning_rate": 3.300244509029854e-05, "loss": 0.0056, "step": 62990 }, { "grad_norm": 0.12153972685337067, "learning_rate": 3.2986896076951585e-05, "loss": 0.0097, "step": 63000 }, { "grad_norm": 0.12764549255371094, "learning_rate": 3.297134892413211e-05, "loss": 0.0051, "step": 63010 }, { "grad_norm": 0.14496496319770813, "learning_rate": 3.295580363354031e-05, "loss": 0.0084, "step": 63020 }, { "grad_norm": 0.08697361499071121, "learning_rate": 3.2940260206876194e-05, "loss": 0.0064, "step": 63030 }, { "grad_norm": 0.11780789494514465, "learning_rate": 3.292471864583958e-05, "loss": 0.0068, "step": 63040 }, { "grad_norm": 0.09859151393175125, "learning_rate": 3.290917895213004e-05, "loss": 0.0056, "step": 63050 }, { "grad_norm": 0.066735178232193, "learning_rate": 3.289364112744701e-05, "loss": 0.0056, "step": 63060 }, { "grad_norm": 0.12042662501335144, "learning_rate": 3.287810517348967e-05, "loss": 0.0053, "step": 63070 }, { "grad_norm": 0.09563872963190079, "learning_rate": 3.2862571091957014e-05, "loss": 0.0053, "step": 63080 }, { "grad_norm": 0.12171870470046997, "learning_rate": 3.284703888454781e-05, "loss": 0.0057, "step": 63090 }, { "grad_norm": 0.1021769642829895, "learning_rate": 3.2831508552960646e-05, "loss": 0.0053, "step": 63100 }, { "grad_norm": 0.13304443657398224, "learning_rate": 3.281598009889387e-05, "loss": 0.0056, "step": 63110 }, { "grad_norm": 0.11599399149417877, "learning_rate": 3.28004535240457e-05, "loss": 0.0045, "step": 63120 }, { "grad_norm": 0.09906195104122162, "learning_rate": 3.278492883011407e-05, "loss": 0.0056, "step": 63130 }, { "grad_norm": 0.0997566506266594, "learning_rate": 3.276940601879675e-05, "loss": 0.0073, "step": 63140 }, { "grad_norm": 0.11473030596971512, "learning_rate": 3.275388509179127e-05, "loss": 0.0064, "step": 63150 }, { "grad_norm": 0.1160876676440239, "learning_rate": 3.2738366050794985e-05, "loss": 0.0052, "step": 63160 }, { "grad_norm": 0.09783050417900085, "learning_rate": 3.2722848897505034e-05, "loss": 0.0043, "step": 63170 }, { "grad_norm": 0.14807231724262238, "learning_rate": 3.2707333633618356e-05, "loss": 0.0068, "step": 63180 }, { "grad_norm": 0.09893446415662766, "learning_rate": 3.269182026083167e-05, "loss": 0.0056, "step": 63190 }, { "grad_norm": 0.09231063723564148, "learning_rate": 3.2676308780841516e-05, "loss": 0.005, "step": 63200 }, { "grad_norm": 0.14461490511894226, "learning_rate": 3.266079919534416e-05, "loss": 0.0054, "step": 63210 }, { "grad_norm": 0.09685889631509781, "learning_rate": 3.264529150603575e-05, "loss": 0.005, "step": 63220 }, { "grad_norm": 0.11475295573472977, "learning_rate": 3.262978571461217e-05, "loss": 0.0063, "step": 63230 }, { "grad_norm": 0.11911875009536743, "learning_rate": 3.26142818227691e-05, "loss": 0.0061, "step": 63240 }, { "grad_norm": 0.10456934571266174, "learning_rate": 3.259877983220203e-05, "loss": 0.0047, "step": 63250 }, { "grad_norm": 0.11059944331645966, "learning_rate": 3.2583279744606246e-05, "loss": 0.0066, "step": 63260 }, { "grad_norm": 0.10782498866319656, "learning_rate": 3.25677815616768e-05, "loss": 0.0051, "step": 63270 }, { "grad_norm": 0.08094654232263565, "learning_rate": 3.255228528510856e-05, "loss": 0.0056, "step": 63280 }, { "grad_norm": 0.1037868857383728, "learning_rate": 3.253679091659617e-05, "loss": 0.0058, "step": 63290 }, { "grad_norm": 0.15035785734653473, "learning_rate": 3.252129845783404e-05, "loss": 0.0052, "step": 63300 }, { "grad_norm": 0.14290499687194824, "learning_rate": 3.2505807910516456e-05, "loss": 0.0058, "step": 63310 }, { "grad_norm": 0.1142578125, "learning_rate": 3.249031927633742e-05, "loss": 0.005, "step": 63320 }, { "grad_norm": 0.08903232216835022, "learning_rate": 3.2474832556990737e-05, "loss": 0.005, "step": 63330 }, { "grad_norm": 0.09269145876169205, "learning_rate": 3.245934775417002e-05, "loss": 0.0054, "step": 63340 }, { "grad_norm": 0.0925459936261177, "learning_rate": 3.244386486956866e-05, "loss": 0.0068, "step": 63350 }, { "grad_norm": 0.10505666583776474, "learning_rate": 3.2428383904879846e-05, "loss": 0.0058, "step": 63360 }, { "grad_norm": 0.14866891503334045, "learning_rate": 3.241290486179652e-05, "loss": 0.0064, "step": 63370 }, { "grad_norm": 0.08150693029165268, "learning_rate": 3.239742774201151e-05, "loss": 0.0058, "step": 63380 }, { "grad_norm": 0.07868361473083496, "learning_rate": 3.238195254721733e-05, "loss": 0.0066, "step": 63390 }, { "grad_norm": 0.0872034952044487, "learning_rate": 3.236647927910633e-05, "loss": 0.0054, "step": 63400 }, { "grad_norm": 0.10419797152280807, "learning_rate": 3.2351007939370646e-05, "loss": 0.0054, "step": 63410 }, { "grad_norm": 0.12918969988822937, "learning_rate": 3.2335538529702184e-05, "loss": 0.0061, "step": 63420 }, { "grad_norm": 0.10587289929389954, "learning_rate": 3.232007105179268e-05, "loss": 0.0046, "step": 63430 }, { "grad_norm": 0.09635192900896072, "learning_rate": 3.230460550733363e-05, "loss": 0.0058, "step": 63440 }, { "grad_norm": 0.1133507788181305, "learning_rate": 3.2289141898016315e-05, "loss": 0.006, "step": 63450 }, { "grad_norm": 0.09720560163259506, "learning_rate": 3.227368022553181e-05, "loss": 0.0052, "step": 63460 }, { "grad_norm": 0.07131998986005783, "learning_rate": 3.225822049157099e-05, "loss": 0.0052, "step": 63470 }, { "grad_norm": 0.1213199570775032, "learning_rate": 3.2242762697824494e-05, "loss": 0.006, "step": 63480 }, { "grad_norm": 0.08885696530342102, "learning_rate": 3.222730684598279e-05, "loss": 0.005, "step": 63490 }, { "grad_norm": 0.10501116514205933, "learning_rate": 3.221185293773605e-05, "loss": 0.0057, "step": 63500 }, { "grad_norm": 0.08387468010187149, "learning_rate": 3.2196400974774354e-05, "loss": 0.0052, "step": 63510 }, { "grad_norm": 0.11918772757053375, "learning_rate": 3.2180950958787485e-05, "loss": 0.0057, "step": 63520 }, { "grad_norm": 0.11934752762317657, "learning_rate": 3.2165502891465024e-05, "loss": 0.0056, "step": 63530 }, { "grad_norm": 0.11554441601037979, "learning_rate": 3.215005677449636e-05, "loss": 0.0054, "step": 63540 }, { "grad_norm": 0.06993415951728821, "learning_rate": 3.2134612609570644e-05, "loss": 0.0054, "step": 63550 }, { "grad_norm": 0.11244675517082214, "learning_rate": 3.211917039837682e-05, "loss": 0.005, "step": 63560 }, { "grad_norm": 0.11390374600887299, "learning_rate": 3.210373014260365e-05, "loss": 0.005, "step": 63570 }, { "grad_norm": 0.08058850467205048, "learning_rate": 3.208829184393965e-05, "loss": 0.0057, "step": 63580 }, { "grad_norm": 0.12046165019273758, "learning_rate": 3.207285550407312e-05, "loss": 0.006, "step": 63590 }, { "grad_norm": 0.11044591665267944, "learning_rate": 3.205742112469217e-05, "loss": 0.0053, "step": 63600 }, { "grad_norm": 0.11976883560419083, "learning_rate": 3.204198870748465e-05, "loss": 0.0056, "step": 63610 }, { "grad_norm": 0.07274606078863144, "learning_rate": 3.202655825413824e-05, "loss": 0.0052, "step": 63620 }, { "grad_norm": 0.10348246991634369, "learning_rate": 3.2011129766340416e-05, "loss": 0.0056, "step": 63630 }, { "grad_norm": 0.16564659774303436, "learning_rate": 3.1995703245778384e-05, "loss": 0.0049, "step": 63640 }, { "grad_norm": 0.09996943920850754, "learning_rate": 3.1980278694139186e-05, "loss": 0.007, "step": 63650 }, { "grad_norm": 0.09845250099897385, "learning_rate": 3.196485611310962e-05, "loss": 0.0064, "step": 63660 }, { "grad_norm": 0.10238105803728104, "learning_rate": 3.1949435504376257e-05, "loss": 0.005, "step": 63670 }, { "grad_norm": 0.08097702264785767, "learning_rate": 3.1934016869625505e-05, "loss": 0.0058, "step": 63680 }, { "grad_norm": 0.08523419499397278, "learning_rate": 3.191860021054349e-05, "loss": 0.0068, "step": 63690 }, { "grad_norm": 0.10762660950422287, "learning_rate": 3.190318552881618e-05, "loss": 0.006, "step": 63700 }, { "grad_norm": 0.11661182343959808, "learning_rate": 3.1887772826129285e-05, "loss": 0.0063, "step": 63710 }, { "grad_norm": 0.11794517189264297, "learning_rate": 3.1872362104168335e-05, "loss": 0.0044, "step": 63720 }, { "grad_norm": 0.1186721995472908, "learning_rate": 3.185695336461861e-05, "loss": 0.006, "step": 63730 }, { "grad_norm": 0.1267823427915573, "learning_rate": 3.184154660916519e-05, "loss": 0.0055, "step": 63740 }, { "grad_norm": 0.1414884477853775, "learning_rate": 3.1826141839492906e-05, "loss": 0.007, "step": 63750 }, { "grad_norm": 0.12814532220363617, "learning_rate": 3.181073905728644e-05, "loss": 0.0055, "step": 63760 }, { "grad_norm": 0.1086755022406578, "learning_rate": 3.1795338264230214e-05, "loss": 0.0066, "step": 63770 }, { "grad_norm": 0.15874402225017548, "learning_rate": 3.1779939462008426e-05, "loss": 0.0048, "step": 63780 }, { "grad_norm": 0.14085066318511963, "learning_rate": 3.1764542652305066e-05, "loss": 0.0061, "step": 63790 }, { "grad_norm": 0.08715219050645828, "learning_rate": 3.17491478368039e-05, "loss": 0.0047, "step": 63800 }, { "grad_norm": 0.10975825786590576, "learning_rate": 3.173375501718847e-05, "loss": 0.0056, "step": 63810 }, { "grad_norm": 0.20078808069229126, "learning_rate": 3.171836419514215e-05, "loss": 0.0068, "step": 63820 }, { "grad_norm": 0.10468294471502304, "learning_rate": 3.170297537234803e-05, "loss": 0.0062, "step": 63830 }, { "grad_norm": 0.1253698617219925, "learning_rate": 3.168758855048901e-05, "loss": 0.0045, "step": 63840 }, { "grad_norm": 0.15145640075206757, "learning_rate": 3.1672203731247785e-05, "loss": 0.0069, "step": 63850 }, { "grad_norm": 0.09975692629814148, "learning_rate": 3.16568209163068e-05, "loss": 0.0048, "step": 63860 }, { "grad_norm": 0.10739956051111221, "learning_rate": 3.1641440107348295e-05, "loss": 0.0049, "step": 63870 }, { "grad_norm": 0.09580706804990768, "learning_rate": 3.162606130605429e-05, "loss": 0.0059, "step": 63880 }, { "grad_norm": 0.08942179381847382, "learning_rate": 3.16106845141066e-05, "loss": 0.005, "step": 63890 }, { "grad_norm": 0.09876454621553421, "learning_rate": 3.159530973318682e-05, "loss": 0.0056, "step": 63900 }, { "grad_norm": 0.10431499034166336, "learning_rate": 3.1579936964976286e-05, "loss": 0.0058, "step": 63910 }, { "grad_norm": 0.0993288978934288, "learning_rate": 3.156456621115615e-05, "loss": 0.0073, "step": 63920 }, { "grad_norm": 0.07942304760217667, "learning_rate": 3.154919747340734e-05, "loss": 0.0049, "step": 63930 }, { "grad_norm": 0.10164418816566467, "learning_rate": 3.153383075341054e-05, "loss": 0.0046, "step": 63940 }, { "grad_norm": 0.09248335659503937, "learning_rate": 3.1518466052846264e-05, "loss": 0.0043, "step": 63950 }, { "grad_norm": 0.10332813858985901, "learning_rate": 3.150310337339474e-05, "loss": 0.0046, "step": 63960 }, { "grad_norm": 0.10358206927776337, "learning_rate": 3.148774271673603e-05, "loss": 0.0059, "step": 63970 }, { "grad_norm": 0.08700186014175415, "learning_rate": 3.147238408454995e-05, "loss": 0.0056, "step": 63980 }, { "grad_norm": 0.09988123178482056, "learning_rate": 3.1457027478516086e-05, "loss": 0.0052, "step": 63990 }, { "grad_norm": 0.08650734275579453, "learning_rate": 3.144167290031382e-05, "loss": 0.0055, "step": 64000 }, { "grad_norm": 0.08463041484355927, "learning_rate": 3.142632035162227e-05, "loss": 0.0061, "step": 64010 }, { "grad_norm": 0.10465054959058762, "learning_rate": 3.141096983412044e-05, "loss": 0.0055, "step": 64020 }, { "grad_norm": 0.09460516273975372, "learning_rate": 3.139562134948699e-05, "loss": 0.0052, "step": 64030 }, { "grad_norm": 0.10151559114456177, "learning_rate": 3.1380274899400424e-05, "loss": 0.0058, "step": 64040 }, { "grad_norm": 0.14593209326267242, "learning_rate": 3.1364930485539e-05, "loss": 0.0067, "step": 64050 }, { "grad_norm": 0.10879883170127869, "learning_rate": 3.134958810958076e-05, "loss": 0.005, "step": 64060 }, { "grad_norm": 0.09913550317287445, "learning_rate": 3.13342477732035e-05, "loss": 0.0055, "step": 64070 }, { "grad_norm": 0.1116049662232399, "learning_rate": 3.131890947808487e-05, "loss": 0.0066, "step": 64080 }, { "grad_norm": 0.09000049531459808, "learning_rate": 3.13035732259022e-05, "loss": 0.006, "step": 64090 }, { "grad_norm": 0.09431730210781097, "learning_rate": 3.1288239018332656e-05, "loss": 0.0048, "step": 64100 }, { "grad_norm": 0.08382528275251389, "learning_rate": 3.1272906857053164e-05, "loss": 0.0065, "step": 64110 }, { "grad_norm": 0.1038660779595375, "learning_rate": 3.125757674374042e-05, "loss": 0.0052, "step": 64120 }, { "grad_norm": 0.0897035300731659, "learning_rate": 3.124224868007088e-05, "loss": 0.0049, "step": 64130 }, { "grad_norm": 0.08031240105628967, "learning_rate": 3.1226922667720854e-05, "loss": 0.0046, "step": 64140 }, { "grad_norm": 0.09532197564840317, "learning_rate": 3.121159870836633e-05, "loss": 0.0068, "step": 64150 }, { "grad_norm": 0.10542458295822144, "learning_rate": 3.119627680368312e-05, "loss": 0.0074, "step": 64160 }, { "grad_norm": 0.07355271279811859, "learning_rate": 3.1180956955346816e-05, "loss": 0.0054, "step": 64170 }, { "grad_norm": 0.10060237348079681, "learning_rate": 3.116563916503276e-05, "loss": 0.0064, "step": 64180 }, { "grad_norm": 0.1294902265071869, "learning_rate": 3.1150323434416095e-05, "loss": 0.0044, "step": 64190 }, { "grad_norm": 0.09226904064416885, "learning_rate": 3.1135009765171696e-05, "loss": 0.0051, "step": 64200 }, { "grad_norm": 0.0696597546339035, "learning_rate": 3.111969815897429e-05, "loss": 0.0046, "step": 64210 }, { "grad_norm": 0.1071970984339714, "learning_rate": 3.1104388617498295e-05, "loss": 0.0049, "step": 64220 }, { "grad_norm": 0.09039390832185745, "learning_rate": 3.1089081142417966e-05, "loss": 0.0051, "step": 64230 }, { "grad_norm": 0.13544665277004242, "learning_rate": 3.1073775735407285e-05, "loss": 0.006, "step": 64240 }, { "grad_norm": 0.09643164277076721, "learning_rate": 3.1058472398140035e-05, "loss": 0.0052, "step": 64250 }, { "grad_norm": 0.1483229696750641, "learning_rate": 3.1043171132289736e-05, "loss": 0.0049, "step": 64260 }, { "grad_norm": 0.09177297353744507, "learning_rate": 3.1027871939529774e-05, "loss": 0.005, "step": 64270 }, { "grad_norm": 0.09453818947076797, "learning_rate": 3.1012574821533214e-05, "loss": 0.0048, "step": 64280 }, { "grad_norm": 0.08876287192106247, "learning_rate": 3.099727977997291e-05, "loss": 0.005, "step": 64290 }, { "grad_norm": 0.10253390669822693, "learning_rate": 3.098198681652154e-05, "loss": 0.0044, "step": 64300 }, { "grad_norm": 0.10035434365272522, "learning_rate": 3.0966695932851475e-05, "loss": 0.0058, "step": 64310 }, { "grad_norm": 0.0767717957496643, "learning_rate": 3.0951407130634926e-05, "loss": 0.0051, "step": 64320 }, { "grad_norm": 0.1771135926246643, "learning_rate": 3.0936120411543864e-05, "loss": 0.0052, "step": 64330 }, { "grad_norm": 0.1298486590385437, "learning_rate": 3.0920835777250015e-05, "loss": 0.005, "step": 64340 }, { "grad_norm": 0.09384767711162567, "learning_rate": 3.090555322942488e-05, "loss": 0.0049, "step": 64350 }, { "grad_norm": 0.0913667306303978, "learning_rate": 3.089027276973973e-05, "loss": 0.0065, "step": 64360 }, { "grad_norm": 0.0969666987657547, "learning_rate": 3.0874994399865615e-05, "loss": 0.0076, "step": 64370 }, { "grad_norm": 0.11246152222156525, "learning_rate": 3.085971812147337e-05, "loss": 0.0057, "step": 64380 }, { "grad_norm": 0.12155112624168396, "learning_rate": 3.0844443936233557e-05, "loss": 0.0053, "step": 64390 }, { "grad_norm": 0.10215446352958679, "learning_rate": 3.082917184581657e-05, "loss": 0.006, "step": 64400 }, { "grad_norm": 0.1139005720615387, "learning_rate": 3.0813901851892513e-05, "loss": 0.0051, "step": 64410 }, { "grad_norm": 0.10888485610485077, "learning_rate": 3.0798633956131316e-05, "loss": 0.0054, "step": 64420 }, { "grad_norm": 0.13833054900169373, "learning_rate": 3.078336816020264e-05, "loss": 0.0058, "step": 64430 }, { "grad_norm": 0.07174107432365417, "learning_rate": 3.0768104465775935e-05, "loss": 0.0052, "step": 64440 }, { "grad_norm": 0.10731959342956543, "learning_rate": 3.0752842874520396e-05, "loss": 0.0053, "step": 64450 }, { "grad_norm": 0.09127786755561829, "learning_rate": 3.073758338810503e-05, "loss": 0.0046, "step": 64460 }, { "grad_norm": 0.07959745824337006, "learning_rate": 3.072232600819859e-05, "loss": 0.0061, "step": 64470 }, { "grad_norm": 0.10147430002689362, "learning_rate": 3.070707073646959e-05, "loss": 0.0044, "step": 64480 }, { "grad_norm": 0.07404763996601105, "learning_rate": 3.069181757458633e-05, "loss": 0.0053, "step": 64490 }, { "grad_norm": 0.13855785131454468, "learning_rate": 3.067656652421687e-05, "loss": 0.0048, "step": 64500 }, { "grad_norm": 0.0893900990486145, "learning_rate": 3.0661317587029025e-05, "loss": 0.0037, "step": 64510 }, { "grad_norm": 0.14768101274967194, "learning_rate": 3.064607076469043e-05, "loss": 0.0053, "step": 64520 }, { "grad_norm": 0.09144081175327301, "learning_rate": 3.063082605886845e-05, "loss": 0.0049, "step": 64530 }, { "grad_norm": 0.09537585824728012, "learning_rate": 3.061558347123019e-05, "loss": 0.0051, "step": 64540 }, { "grad_norm": 0.09451092779636383, "learning_rate": 3.060034300344259e-05, "loss": 0.0058, "step": 64550 }, { "grad_norm": 0.07291284203529358, "learning_rate": 3.058510465717231e-05, "loss": 0.0039, "step": 64560 }, { "grad_norm": 0.13877452909946442, "learning_rate": 3.0569868434085804e-05, "loss": 0.0044, "step": 64570 }, { "grad_norm": 0.09354803711175919, "learning_rate": 3.055463433584924e-05, "loss": 0.0044, "step": 64580 }, { "grad_norm": 0.11045698076486588, "learning_rate": 3.0539402364128656e-05, "loss": 0.0061, "step": 64590 }, { "grad_norm": 0.08369310200214386, "learning_rate": 3.052417252058977e-05, "loss": 0.0052, "step": 64600 }, { "grad_norm": 0.09436926990747452, "learning_rate": 3.0508944806898092e-05, "loss": 0.0046, "step": 64610 }, { "grad_norm": 0.06452476978302002, "learning_rate": 3.04937192247189e-05, "loss": 0.0052, "step": 64620 }, { "grad_norm": 0.11091132462024689, "learning_rate": 3.047849577571725e-05, "loss": 0.0041, "step": 64630 }, { "grad_norm": 0.0861063152551651, "learning_rate": 3.046327446155795e-05, "loss": 0.0058, "step": 64640 }, { "grad_norm": 0.0985976904630661, "learning_rate": 3.0448055283905587e-05, "loss": 0.0056, "step": 64650 }, { "grad_norm": 0.10718854516744614, "learning_rate": 3.0432838244424494e-05, "loss": 0.0049, "step": 64660 }, { "grad_norm": 0.10826730728149414, "learning_rate": 3.04176233447788e-05, "loss": 0.0068, "step": 64670 }, { "grad_norm": 0.1592046618461609, "learning_rate": 3.040241058663238e-05, "loss": 0.006, "step": 64680 }, { "grad_norm": 0.1052166223526001, "learning_rate": 3.038719997164887e-05, "loss": 0.0051, "step": 64690 }, { "grad_norm": 0.09234809875488281, "learning_rate": 3.0371991501491683e-05, "loss": 0.0055, "step": 64700 }, { "grad_norm": 0.0736313909292221, "learning_rate": 3.0356785177823975e-05, "loss": 0.0061, "step": 64710 }, { "grad_norm": 0.11632674932479858, "learning_rate": 3.0341581002308726e-05, "loss": 0.0044, "step": 64720 }, { "grad_norm": 0.1480734795331955, "learning_rate": 3.0326378976608627e-05, "loss": 0.0052, "step": 64730 }, { "grad_norm": 0.13833236694335938, "learning_rate": 3.0311179102386146e-05, "loss": 0.0044, "step": 64740 }, { "grad_norm": 0.16542017459869385, "learning_rate": 3.0295981381303513e-05, "loss": 0.0052, "step": 64750 }, { "grad_norm": 0.13881538808345795, "learning_rate": 3.0280785815022727e-05, "loss": 0.0044, "step": 64760 }, { "grad_norm": 0.1424434632062912, "learning_rate": 3.0265592405205544e-05, "loss": 0.0054, "step": 64770 }, { "grad_norm": 0.10564786195755005, "learning_rate": 3.0250401153513517e-05, "loss": 0.005, "step": 64780 }, { "grad_norm": 0.09213455766439438, "learning_rate": 3.0235212061607936e-05, "loss": 0.0049, "step": 64790 }, { "grad_norm": 0.13037928938865662, "learning_rate": 3.0220025131149842e-05, "loss": 0.0065, "step": 64800 }, { "grad_norm": 0.13164369761943817, "learning_rate": 3.020484036380007e-05, "loss": 0.0054, "step": 64810 }, { "grad_norm": 0.11664678156375885, "learning_rate": 3.0189657761219188e-05, "loss": 0.0058, "step": 64820 }, { "grad_norm": 0.09386168420314789, "learning_rate": 3.017447732506753e-05, "loss": 0.006, "step": 64830 }, { "grad_norm": 0.10716244578361511, "learning_rate": 3.015929905700524e-05, "loss": 0.0056, "step": 64840 }, { "grad_norm": 0.1264292299747467, "learning_rate": 3.014412295869218e-05, "loss": 0.0054, "step": 64850 }, { "grad_norm": 0.12332793325185776, "learning_rate": 3.0128949031787977e-05, "loss": 0.0055, "step": 64860 }, { "grad_norm": 0.10770146548748016, "learning_rate": 3.0113777277952022e-05, "loss": 0.0047, "step": 64870 }, { "grad_norm": 0.07365764677524567, "learning_rate": 3.0098607698843496e-05, "loss": 0.0064, "step": 64880 }, { "grad_norm": 0.06986749172210693, "learning_rate": 3.0083440296121312e-05, "loss": 0.0041, "step": 64890 }, { "grad_norm": 0.09401991963386536, "learning_rate": 3.0068275071444142e-05, "loss": 0.005, "step": 64900 }, { "grad_norm": 0.07227044552564621, "learning_rate": 3.0053112026470442e-05, "loss": 0.0038, "step": 64910 }, { "grad_norm": 0.13259519636631012, "learning_rate": 3.0037951162858434e-05, "loss": 0.0046, "step": 64920 }, { "grad_norm": 0.10287059843540192, "learning_rate": 3.002279248226607e-05, "loss": 0.0043, "step": 64930 }, { "grad_norm": 0.10762032121419907, "learning_rate": 3.000763598635109e-05, "loss": 0.0047, "step": 64940 }, { "grad_norm": 0.09382081031799316, "learning_rate": 2.999248167677098e-05, "loss": 0.0041, "step": 64950 }, { "grad_norm": 0.08227770775556564, "learning_rate": 2.9977329555182976e-05, "loss": 0.0046, "step": 64960 }, { "grad_norm": 0.09774808585643768, "learning_rate": 2.9962179623244124e-05, "loss": 0.005, "step": 64970 }, { "grad_norm": 0.13203728199005127, "learning_rate": 2.9947031882611175e-05, "loss": 0.0069, "step": 64980 }, { "grad_norm": 0.09476730227470398, "learning_rate": 2.9931886334940685e-05, "loss": 0.0068, "step": 64990 }, { "grad_norm": 0.10051452368497849, "learning_rate": 2.9916742981888923e-05, "loss": 0.0052, "step": 65000 }, { "grad_norm": 0.112612284719944, "learning_rate": 2.990160182511196e-05, "loss": 0.0052, "step": 65010 }, { "grad_norm": 0.12183788418769836, "learning_rate": 2.9886462866265574e-05, "loss": 0.0064, "step": 65020 }, { "grad_norm": 0.12639935314655304, "learning_rate": 2.9871326107005392e-05, "loss": 0.0061, "step": 65030 }, { "grad_norm": 0.07612375915050507, "learning_rate": 2.9856191548986735e-05, "loss": 0.0048, "step": 65040 }, { "grad_norm": 0.10580785572528839, "learning_rate": 2.9841059193864672e-05, "loss": 0.0043, "step": 65050 }, { "grad_norm": 0.08455266058444977, "learning_rate": 2.982592904329407e-05, "loss": 0.0046, "step": 65060 }, { "grad_norm": 0.09848279505968094, "learning_rate": 2.981080109892954e-05, "loss": 0.0048, "step": 65070 }, { "grad_norm": 0.10745671391487122, "learning_rate": 2.9795675362425434e-05, "loss": 0.0047, "step": 65080 }, { "grad_norm": 0.08140632510185242, "learning_rate": 2.97805518354359e-05, "loss": 0.0059, "step": 65090 }, { "grad_norm": 0.07966754585504532, "learning_rate": 2.9765430519614825e-05, "loss": 0.0047, "step": 65100 }, { "grad_norm": 0.07842402160167694, "learning_rate": 2.975031141661584e-05, "loss": 0.0043, "step": 65110 }, { "grad_norm": 0.08746632933616638, "learning_rate": 2.9735194528092347e-05, "loss": 0.0041, "step": 65120 }, { "grad_norm": 0.0987563282251358, "learning_rate": 2.9720079855697525e-05, "loss": 0.0049, "step": 65130 }, { "grad_norm": 0.13169392943382263, "learning_rate": 2.970496740108427e-05, "loss": 0.0055, "step": 65140 }, { "grad_norm": 0.1273338496685028, "learning_rate": 2.968985716590527e-05, "loss": 0.0045, "step": 65150 }, { "grad_norm": 0.09483788162469864, "learning_rate": 2.9674749151812953e-05, "loss": 0.0055, "step": 65160 }, { "grad_norm": 0.05829548090696335, "learning_rate": 2.965964336045951e-05, "loss": 0.0049, "step": 65170 }, { "grad_norm": 0.07624022662639618, "learning_rate": 2.9644539793496894e-05, "loss": 0.005, "step": 65180 }, { "grad_norm": 0.11587405204772949, "learning_rate": 2.9629438452576807e-05, "loss": 0.005, "step": 65190 }, { "grad_norm": 0.07127361744642258, "learning_rate": 2.9614339339350698e-05, "loss": 0.0055, "step": 65200 }, { "grad_norm": 0.09415899217128754, "learning_rate": 2.959924245546979e-05, "loss": 0.0055, "step": 65210 }, { "grad_norm": 0.09339301288127899, "learning_rate": 2.9584147802585038e-05, "loss": 0.0054, "step": 65220 }, { "grad_norm": 0.07945983856916428, "learning_rate": 2.956905538234721e-05, "loss": 0.006, "step": 65230 }, { "grad_norm": 0.09814894944429398, "learning_rate": 2.9553965196406764e-05, "loss": 0.0046, "step": 65240 }, { "grad_norm": 0.07641246914863586, "learning_rate": 2.9538877246413943e-05, "loss": 0.0051, "step": 65250 }, { "grad_norm": 0.07389496266841888, "learning_rate": 2.952379153401874e-05, "loss": 0.0063, "step": 65260 }, { "grad_norm": 0.06544334441423416, "learning_rate": 2.9508708060870914e-05, "loss": 0.0051, "step": 65270 }, { "grad_norm": 0.07515592873096466, "learning_rate": 2.9493626828619948e-05, "loss": 0.0044, "step": 65280 }, { "grad_norm": 0.0816500186920166, "learning_rate": 2.9478547838915137e-05, "loss": 0.0068, "step": 65290 }, { "grad_norm": 0.11430968344211578, "learning_rate": 2.946347109340548e-05, "loss": 0.0056, "step": 65300 }, { "grad_norm": 0.07504954934120178, "learning_rate": 2.944839659373976e-05, "loss": 0.0057, "step": 65310 }, { "grad_norm": 0.09127532690763474, "learning_rate": 2.943332434156647e-05, "loss": 0.0081, "step": 65320 }, { "grad_norm": 0.07890423387289047, "learning_rate": 2.9418254338533912e-05, "loss": 0.0046, "step": 65330 }, { "grad_norm": 0.08283782750368118, "learning_rate": 2.940318658629011e-05, "loss": 0.0054, "step": 65340 }, { "grad_norm": 0.0846962258219719, "learning_rate": 2.938812108648287e-05, "loss": 0.0063, "step": 65350 }, { "grad_norm": 0.10917764902114868, "learning_rate": 2.937305784075972e-05, "loss": 0.0053, "step": 65360 }, { "grad_norm": 0.1130499467253685, "learning_rate": 2.9357996850767937e-05, "loss": 0.0049, "step": 65370 }, { "grad_norm": 0.09799850732088089, "learning_rate": 2.9342938118154595e-05, "loss": 0.0047, "step": 65380 }, { "grad_norm": 0.09750165790319443, "learning_rate": 2.9327881644566485e-05, "loss": 0.0056, "step": 65390 }, { "grad_norm": 0.09083191305398941, "learning_rate": 2.9312827431650157e-05, "loss": 0.0048, "step": 65400 }, { "grad_norm": 0.09102604538202286, "learning_rate": 2.92977754810519e-05, "loss": 0.0062, "step": 65410 }, { "grad_norm": 0.0746687799692154, "learning_rate": 2.928272579441781e-05, "loss": 0.0053, "step": 65420 }, { "grad_norm": 0.12026578933000565, "learning_rate": 2.9267678373393676e-05, "loss": 0.0056, "step": 65430 }, { "grad_norm": 0.11989213526248932, "learning_rate": 2.925263321962507e-05, "loss": 0.0047, "step": 65440 }, { "grad_norm": 0.08642037212848663, "learning_rate": 2.9237590334757292e-05, "loss": 0.0063, "step": 65450 }, { "grad_norm": 0.1312083750963211, "learning_rate": 2.922254972043543e-05, "loss": 0.0056, "step": 65460 }, { "grad_norm": 0.07199497520923615, "learning_rate": 2.9207511378304265e-05, "loss": 0.0056, "step": 65470 }, { "grad_norm": 0.126570463180542, "learning_rate": 2.9192475310008428e-05, "loss": 0.0062, "step": 65480 }, { "grad_norm": 0.09711654484272003, "learning_rate": 2.917744151719219e-05, "loss": 0.0052, "step": 65490 }, { "grad_norm": 0.11014813929796219, "learning_rate": 2.916241000149964e-05, "loss": 0.0049, "step": 65500 }, { "grad_norm": 0.12266258895397186, "learning_rate": 2.9147380764574622e-05, "loss": 0.0048, "step": 65510 }, { "grad_norm": 0.10738515108823776, "learning_rate": 2.9132353808060674e-05, "loss": 0.0045, "step": 65520 }, { "grad_norm": 0.10648931562900543, "learning_rate": 2.9117329133601134e-05, "loss": 0.0046, "step": 65530 }, { "grad_norm": 0.1261463314294815, "learning_rate": 2.9102306742839093e-05, "loss": 0.0051, "step": 65540 }, { "grad_norm": 0.0966348648071289, "learning_rate": 2.9087286637417375e-05, "loss": 0.0039, "step": 65550 }, { "grad_norm": 0.09542479366064072, "learning_rate": 2.9072268818978537e-05, "loss": 0.0043, "step": 65560 }, { "grad_norm": 0.11167803406715393, "learning_rate": 2.905725328916493e-05, "loss": 0.0046, "step": 65570 }, { "grad_norm": 0.0885029062628746, "learning_rate": 2.9042240049618608e-05, "loss": 0.0049, "step": 65580 }, { "grad_norm": 0.09825702756643295, "learning_rate": 2.9027229101981413e-05, "loss": 0.0043, "step": 65590 }, { "grad_norm": 0.10044395923614502, "learning_rate": 2.901222044789488e-05, "loss": 0.0044, "step": 65600 }, { "grad_norm": 0.09230300784111023, "learning_rate": 2.89972140890004e-05, "loss": 0.005, "step": 65610 }, { "grad_norm": 0.09508354961872101, "learning_rate": 2.8982210026938988e-05, "loss": 0.0048, "step": 65620 }, { "grad_norm": 0.07813174277544022, "learning_rate": 2.896720826335151e-05, "loss": 0.0077, "step": 65630 }, { "grad_norm": 0.09058742225170135, "learning_rate": 2.8952208799878493e-05, "loss": 0.0055, "step": 65640 }, { "grad_norm": 0.13929735124111176, "learning_rate": 2.89372116381603e-05, "loss": 0.0046, "step": 65650 }, { "grad_norm": 0.10924115031957626, "learning_rate": 2.8922216779836932e-05, "loss": 0.0048, "step": 65660 }, { "grad_norm": 0.11482841521501541, "learning_rate": 2.8907224226548285e-05, "loss": 0.0051, "step": 65670 }, { "grad_norm": 0.11288274824619293, "learning_rate": 2.8892233979933868e-05, "loss": 0.0052, "step": 65680 }, { "grad_norm": 0.10632627457380295, "learning_rate": 2.8877246041633026e-05, "loss": 0.005, "step": 65690 }, { "grad_norm": 0.11439049988985062, "learning_rate": 2.886226041328478e-05, "loss": 0.007, "step": 65700 }, { "grad_norm": 0.1077895313501358, "learning_rate": 2.8847277096527957e-05, "loss": 0.0059, "step": 65710 }, { "grad_norm": 0.09189283847808838, "learning_rate": 2.8832296093001105e-05, "loss": 0.0066, "step": 65720 }, { "grad_norm": 0.09872551262378693, "learning_rate": 2.8817317404342547e-05, "loss": 0.0057, "step": 65730 }, { "grad_norm": 0.10894526541233063, "learning_rate": 2.880234103219029e-05, "loss": 0.0044, "step": 65740 }, { "grad_norm": 0.12735679745674133, "learning_rate": 2.878736697818215e-05, "loss": 0.0048, "step": 65750 }, { "grad_norm": 0.08372726291418076, "learning_rate": 2.877239524395568e-05, "loss": 0.0057, "step": 65760 }, { "grad_norm": 0.09646160900592804, "learning_rate": 2.8757425831148134e-05, "loss": 0.0059, "step": 65770 }, { "grad_norm": 0.10999931395053864, "learning_rate": 2.874245874139658e-05, "loss": 0.0062, "step": 65780 }, { "grad_norm": 0.12411486357450485, "learning_rate": 2.8727493976337737e-05, "loss": 0.0058, "step": 65790 }, { "grad_norm": 0.07938947528600693, "learning_rate": 2.8712531537608218e-05, "loss": 0.0047, "step": 65800 }, { "grad_norm": 0.10132192820310593, "learning_rate": 2.8697571426844217e-05, "loss": 0.0044, "step": 65810 }, { "grad_norm": 0.1064969003200531, "learning_rate": 2.86826136456818e-05, "loss": 0.0049, "step": 65820 }, { "grad_norm": 0.12343538552522659, "learning_rate": 2.8667658195756685e-05, "loss": 0.005, "step": 65830 }, { "grad_norm": 0.17324842512607574, "learning_rate": 2.8652705078704412e-05, "loss": 0.0057, "step": 65840 }, { "grad_norm": 0.08657460659742355, "learning_rate": 2.863775429616018e-05, "loss": 0.004, "step": 65850 }, { "grad_norm": 0.10391625016927719, "learning_rate": 2.862280584975906e-05, "loss": 0.0044, "step": 65860 }, { "grad_norm": 0.09560059756040573, "learning_rate": 2.8607859741135735e-05, "loss": 0.0048, "step": 65870 }, { "grad_norm": 0.08108483254909515, "learning_rate": 2.8592915971924723e-05, "loss": 0.0066, "step": 65880 }, { "grad_norm": 0.08405718207359314, "learning_rate": 2.8577974543760212e-05, "loss": 0.0057, "step": 65890 }, { "grad_norm": 0.09648674726486206, "learning_rate": 2.8563035458276222e-05, "loss": 0.0045, "step": 65900 }, { "grad_norm": 0.08834703266620636, "learning_rate": 2.8548098717106424e-05, "loss": 0.0049, "step": 65910 }, { "grad_norm": 0.09932474046945572, "learning_rate": 2.85331643218843e-05, "loss": 0.0053, "step": 65920 }, { "grad_norm": 0.08457014709711075, "learning_rate": 2.8518232274243046e-05, "loss": 0.0049, "step": 65930 }, { "grad_norm": 0.0868084654211998, "learning_rate": 2.8503302575815638e-05, "loss": 0.0041, "step": 65940 }, { "grad_norm": 0.09230548143386841, "learning_rate": 2.8488375228234716e-05, "loss": 0.0054, "step": 65950 }, { "grad_norm": 0.11606144160032272, "learning_rate": 2.8473450233132738e-05, "loss": 0.0041, "step": 65960 }, { "grad_norm": 0.1314060091972351, "learning_rate": 2.8458527592141898e-05, "loss": 0.0054, "step": 65970 }, { "grad_norm": 0.09619592130184174, "learning_rate": 2.8443607306894073e-05, "loss": 0.0064, "step": 65980 }, { "grad_norm": 0.08521215617656708, "learning_rate": 2.8428689379020946e-05, "loss": 0.0041, "step": 65990 }, { "grad_norm": 0.10821696370840073, "learning_rate": 2.841377381015391e-05, "loss": 0.0075, "step": 66000 }, { "grad_norm": 0.0902823805809021, "learning_rate": 2.8398860601924144e-05, "loss": 0.0048, "step": 66010 }, { "grad_norm": 0.08001894503831863, "learning_rate": 2.838394975596248e-05, "loss": 0.0049, "step": 66020 }, { "grad_norm": 0.08033064752817154, "learning_rate": 2.83690412738996e-05, "loss": 0.0043, "step": 66030 }, { "grad_norm": 0.10000525414943695, "learning_rate": 2.835413515736579e-05, "loss": 0.0064, "step": 66040 }, { "grad_norm": 0.11861953884363174, "learning_rate": 2.8339231407991267e-05, "loss": 0.0052, "step": 66050 }, { "grad_norm": 0.10793021321296692, "learning_rate": 2.832433002740581e-05, "loss": 0.0048, "step": 66060 }, { "grad_norm": 0.13355693221092224, "learning_rate": 2.8309431017239047e-05, "loss": 0.0048, "step": 66070 }, { "grad_norm": 0.0877157598733902, "learning_rate": 2.829453437912028e-05, "loss": 0.0043, "step": 66080 }, { "grad_norm": 0.1342114359140396, "learning_rate": 2.8279640114678617e-05, "loss": 0.005, "step": 66090 }, { "grad_norm": 0.1058364287018776, "learning_rate": 2.8264748225542836e-05, "loss": 0.0048, "step": 66100 }, { "grad_norm": 0.06685709953308105, "learning_rate": 2.8249858713341516e-05, "loss": 0.0055, "step": 66110 }, { "grad_norm": 0.09171430766582489, "learning_rate": 2.823497157970294e-05, "loss": 0.0056, "step": 66120 }, { "grad_norm": 0.11998134106397629, "learning_rate": 2.8220086826255164e-05, "loss": 0.0048, "step": 66130 }, { "grad_norm": 0.18510089814662933, "learning_rate": 2.8205204454625934e-05, "loss": 0.0056, "step": 66140 }, { "grad_norm": 0.10264280438423157, "learning_rate": 2.8190324466442798e-05, "loss": 0.0045, "step": 66150 }, { "grad_norm": 0.09793541580438614, "learning_rate": 2.8175446863332965e-05, "loss": 0.0052, "step": 66160 }, { "grad_norm": 0.07761361449956894, "learning_rate": 2.816057164692345e-05, "loss": 0.006, "step": 66170 }, { "grad_norm": 0.11945349723100662, "learning_rate": 2.8145698818840992e-05, "loss": 0.0046, "step": 66180 }, { "grad_norm": 0.07397449016571045, "learning_rate": 2.8130828380712065e-05, "loss": 0.0039, "step": 66190 }, { "grad_norm": 0.07118941098451614, "learning_rate": 2.811596033416285e-05, "loss": 0.0057, "step": 66200 }, { "grad_norm": 0.06135290861129761, "learning_rate": 2.8101094680819307e-05, "loss": 0.005, "step": 66210 }, { "grad_norm": 0.10343687981367111, "learning_rate": 2.808623142230715e-05, "loss": 0.0048, "step": 66220 }, { "grad_norm": 0.07856860011816025, "learning_rate": 2.8071370560251757e-05, "loss": 0.0042, "step": 66230 }, { "grad_norm": 0.07182443141937256, "learning_rate": 2.805651209627831e-05, "loss": 0.0055, "step": 66240 }, { "grad_norm": 0.11603537946939468, "learning_rate": 2.804165603201171e-05, "loss": 0.0052, "step": 66250 }, { "grad_norm": 0.100776806473732, "learning_rate": 2.802680236907661e-05, "loss": 0.0052, "step": 66260 }, { "grad_norm": 0.0754573717713356, "learning_rate": 2.8011951109097344e-05, "loss": 0.0053, "step": 66270 }, { "grad_norm": 0.09600662440061569, "learning_rate": 2.7997102253698072e-05, "loss": 0.0045, "step": 66280 }, { "grad_norm": 0.08957801759243011, "learning_rate": 2.7982255804502588e-05, "loss": 0.0047, "step": 66290 }, { "grad_norm": 0.08943283557891846, "learning_rate": 2.7967411763134516e-05, "loss": 0.0051, "step": 66300 }, { "grad_norm": 0.11989498883485794, "learning_rate": 2.7952570131217166e-05, "loss": 0.0057, "step": 66310 }, { "grad_norm": 0.10677880793809891, "learning_rate": 2.7937730910373618e-05, "loss": 0.0059, "step": 66320 }, { "grad_norm": 0.12649530172348022, "learning_rate": 2.7922894102226628e-05, "loss": 0.0049, "step": 66330 }, { "grad_norm": 0.10205557197332382, "learning_rate": 2.790805970839876e-05, "loss": 0.0053, "step": 66340 }, { "grad_norm": 0.11104179918766022, "learning_rate": 2.789322773051225e-05, "loss": 0.0047, "step": 66350 }, { "grad_norm": 0.09392207860946655, "learning_rate": 2.7878398170189124e-05, "loss": 0.0053, "step": 66360 }, { "grad_norm": 0.0803266316652298, "learning_rate": 2.7863571029051112e-05, "loss": 0.0047, "step": 66370 }, { "grad_norm": 0.10826981067657471, "learning_rate": 2.784874630871971e-05, "loss": 0.0042, "step": 66380 }, { "grad_norm": 0.11922501772642136, "learning_rate": 2.7833924010816086e-05, "loss": 0.005, "step": 66390 }, { "grad_norm": 0.12341512739658356, "learning_rate": 2.7819104136961226e-05, "loss": 0.007, "step": 66400 }, { "grad_norm": 0.09434578567743301, "learning_rate": 2.780428668877577e-05, "loss": 0.0043, "step": 66410 }, { "grad_norm": 0.07414640486240387, "learning_rate": 2.7789471667880152e-05, "loss": 0.0058, "step": 66420 }, { "grad_norm": 0.1298140287399292, "learning_rate": 2.7774659075894523e-05, "loss": 0.0049, "step": 66430 }, { "grad_norm": 0.08122478425502777, "learning_rate": 2.7759848914438778e-05, "loss": 0.0041, "step": 66440 }, { "grad_norm": 0.14572733640670776, "learning_rate": 2.7745041185132496e-05, "loss": 0.0049, "step": 66450 }, { "grad_norm": 0.0947711169719696, "learning_rate": 2.773023588959506e-05, "loss": 0.0054, "step": 66460 }, { "grad_norm": 0.1020190566778183, "learning_rate": 2.7715433029445555e-05, "loss": 0.0043, "step": 66470 }, { "grad_norm": 0.11860573291778564, "learning_rate": 2.7700632606302766e-05, "loss": 0.006, "step": 66480 }, { "grad_norm": 0.15987320244312286, "learning_rate": 2.768583462178528e-05, "loss": 0.0045, "step": 66490 }, { "grad_norm": 0.10381017625331879, "learning_rate": 2.767103907751136e-05, "loss": 0.004, "step": 66500 }, { "grad_norm": 0.10164611786603928, "learning_rate": 2.7656245975099065e-05, "loss": 0.0057, "step": 66510 }, { "grad_norm": 0.09566640108823776, "learning_rate": 2.7641455316166093e-05, "loss": 0.0048, "step": 66520 }, { "grad_norm": 0.12689077854156494, "learning_rate": 2.7626667102329973e-05, "loss": 0.005, "step": 66530 }, { "grad_norm": 0.0931500568985939, "learning_rate": 2.761188133520788e-05, "loss": 0.0047, "step": 66540 }, { "grad_norm": 0.09355924278497696, "learning_rate": 2.7597098016416784e-05, "loss": 0.0051, "step": 66550 }, { "grad_norm": 0.11913242936134338, "learning_rate": 2.7582317147573368e-05, "loss": 0.0062, "step": 66560 }, { "grad_norm": 0.07517435401678085, "learning_rate": 2.7567538730294064e-05, "loss": 0.0049, "step": 66570 }, { "grad_norm": 0.10031232237815857, "learning_rate": 2.7552762766194973e-05, "loss": 0.0059, "step": 66580 }, { "grad_norm": 0.1116834357380867, "learning_rate": 2.753798925689202e-05, "loss": 0.0053, "step": 66590 }, { "grad_norm": 0.10809800773859024, "learning_rate": 2.7523218204000774e-05, "loss": 0.0031, "step": 66600 }, { "grad_norm": 0.06583938002586365, "learning_rate": 2.7508449609136612e-05, "loss": 0.0045, "step": 66610 }, { "grad_norm": 0.10450280457735062, "learning_rate": 2.749368347391454e-05, "loss": 0.0035, "step": 66620 }, { "grad_norm": 0.13220013678073883, "learning_rate": 2.7478919799949443e-05, "loss": 0.0057, "step": 66630 }, { "grad_norm": 0.074183888733387, "learning_rate": 2.7464158588855803e-05, "loss": 0.0054, "step": 66640 }, { "grad_norm": 0.08183370530605316, "learning_rate": 2.7449399842247918e-05, "loss": 0.0051, "step": 66650 }, { "grad_norm": 0.09800392389297485, "learning_rate": 2.7434643561739736e-05, "loss": 0.005, "step": 66660 }, { "grad_norm": 0.0985320508480072, "learning_rate": 2.7419889748945016e-05, "loss": 0.0046, "step": 66670 }, { "grad_norm": 0.07626178860664368, "learning_rate": 2.7405138405477203e-05, "loss": 0.0033, "step": 66680 }, { "grad_norm": 0.11415302008390427, "learning_rate": 2.73903895329495e-05, "loss": 0.0051, "step": 66690 }, { "grad_norm": 0.05420643091201782, "learning_rate": 2.737564313297478e-05, "loss": 0.0048, "step": 66700 }, { "grad_norm": 0.09188124537467957, "learning_rate": 2.736089920716572e-05, "loss": 0.0047, "step": 66710 }, { "grad_norm": 0.07951468974351883, "learning_rate": 2.7346157757134706e-05, "loss": 0.0054, "step": 66720 }, { "grad_norm": 0.10474236309528351, "learning_rate": 2.733141878449379e-05, "loss": 0.0061, "step": 66730 }, { "grad_norm": 0.09476592391729355, "learning_rate": 2.7316682290854844e-05, "loss": 0.0048, "step": 66740 }, { "grad_norm": 0.0870848074555397, "learning_rate": 2.7301948277829414e-05, "loss": 0.0048, "step": 66750 }, { "grad_norm": 0.09546808898448944, "learning_rate": 2.7287216747028814e-05, "loss": 0.0048, "step": 66760 }, { "grad_norm": 0.07333610951900482, "learning_rate": 2.7272487700064025e-05, "loss": 0.004, "step": 66770 }, { "grad_norm": 0.13189835846424103, "learning_rate": 2.725776113854583e-05, "loss": 0.0052, "step": 66780 }, { "grad_norm": 0.09312258660793304, "learning_rate": 2.7243037064084664e-05, "loss": 0.0042, "step": 66790 }, { "grad_norm": 0.11751100420951843, "learning_rate": 2.7228315478290766e-05, "loss": 0.0046, "step": 66800 }, { "grad_norm": 0.07736528664827347, "learning_rate": 2.721359638277402e-05, "loss": 0.0047, "step": 66810 }, { "grad_norm": 0.10249056667089462, "learning_rate": 2.7198879779144148e-05, "loss": 0.0049, "step": 66820 }, { "grad_norm": 0.08733994513750076, "learning_rate": 2.7184165669010485e-05, "loss": 0.0052, "step": 66830 }, { "grad_norm": 0.10625004023313522, "learning_rate": 2.716945405398218e-05, "loss": 0.0068, "step": 66840 }, { "grad_norm": 0.10298176854848862, "learning_rate": 2.715474493566803e-05, "loss": 0.0043, "step": 66850 }, { "grad_norm": 0.08717308938503265, "learning_rate": 2.7140038315676652e-05, "loss": 0.005, "step": 66860 }, { "grad_norm": 0.07568353414535522, "learning_rate": 2.7125334195616275e-05, "loss": 0.0044, "step": 66870 }, { "grad_norm": 0.11311917752027512, "learning_rate": 2.7110632577094997e-05, "loss": 0.0053, "step": 66880 }, { "grad_norm": 0.12677761912345886, "learning_rate": 2.709593346172051e-05, "loss": 0.0057, "step": 66890 }, { "grad_norm": 0.09479221701622009, "learning_rate": 2.708123685110032e-05, "loss": 0.0042, "step": 66900 }, { "grad_norm": 0.11186680942773819, "learning_rate": 2.7066542746841588e-05, "loss": 0.0047, "step": 66910 }, { "grad_norm": 0.12490895390510559, "learning_rate": 2.7051851150551267e-05, "loss": 0.0054, "step": 66920 }, { "grad_norm": 0.1382511854171753, "learning_rate": 2.7037162063835996e-05, "loss": 0.0056, "step": 66930 }, { "grad_norm": 0.14415906369686127, "learning_rate": 2.702247548830218e-05, "loss": 0.0049, "step": 66940 }, { "grad_norm": 0.10010868310928345, "learning_rate": 2.7007791425555883e-05, "loss": 0.004, "step": 66950 }, { "grad_norm": 0.09398628026247025, "learning_rate": 2.6993109877202945e-05, "loss": 0.0045, "step": 66960 }, { "grad_norm": 0.12751981616020203, "learning_rate": 2.6978430844848935e-05, "loss": 0.0047, "step": 66970 }, { "grad_norm": 0.08489103615283966, "learning_rate": 2.69637543300991e-05, "loss": 0.0061, "step": 66980 }, { "grad_norm": 0.09472517669200897, "learning_rate": 2.694908033455848e-05, "loss": 0.0051, "step": 66990 }, { "grad_norm": 0.10923012346029282, "learning_rate": 2.6934408859831733e-05, "loss": 0.0058, "step": 67000 }, { "grad_norm": 0.09726374596357346, "learning_rate": 2.69197399075234e-05, "loss": 0.0039, "step": 67010 }, { "grad_norm": 0.10306684672832489, "learning_rate": 2.6905073479237584e-05, "loss": 0.0039, "step": 67020 }, { "grad_norm": 0.08121533691883087, "learning_rate": 2.6890409576578235e-05, "loss": 0.0033, "step": 67030 }, { "grad_norm": 0.09839366376399994, "learning_rate": 2.687574820114892e-05, "loss": 0.0048, "step": 67040 }, { "grad_norm": 0.09399323165416718, "learning_rate": 2.686108935455305e-05, "loss": 0.0046, "step": 67050 }, { "grad_norm": 0.11982637643814087, "learning_rate": 2.6846433038393603e-05, "loss": 0.0046, "step": 67060 }, { "grad_norm": 0.2281108945608139, "learning_rate": 2.6831779254273482e-05, "loss": 0.0055, "step": 67070 }, { "grad_norm": 0.07060013711452484, "learning_rate": 2.6817128003795122e-05, "loss": 0.0041, "step": 67080 }, { "grad_norm": 0.07637694478034973, "learning_rate": 2.680247928856081e-05, "loss": 0.0041, "step": 67090 }, { "grad_norm": 0.0765548124909401, "learning_rate": 2.6787833110172466e-05, "loss": 0.0043, "step": 67100 }, { "grad_norm": 0.1086326539516449, "learning_rate": 2.6773189470231807e-05, "loss": 0.0058, "step": 67110 }, { "grad_norm": 0.10444798320531845, "learning_rate": 2.675854837034022e-05, "loss": 0.0041, "step": 67120 }, { "grad_norm": 0.1065308228135109, "learning_rate": 2.6743909812098833e-05, "loss": 0.0051, "step": 67130 }, { "grad_norm": 0.1238999217748642, "learning_rate": 2.67292737971085e-05, "loss": 0.0052, "step": 67140 }, { "grad_norm": 0.10013492405414581, "learning_rate": 2.671464032696982e-05, "loss": 0.0049, "step": 67150 }, { "grad_norm": 0.08450648188591003, "learning_rate": 2.6700009403283055e-05, "loss": 0.0054, "step": 67160 }, { "grad_norm": 0.06895928084850311, "learning_rate": 2.6685381027648214e-05, "loss": 0.0067, "step": 67170 }, { "grad_norm": 0.0968245416879654, "learning_rate": 2.6670755201665086e-05, "loss": 0.0046, "step": 67180 }, { "grad_norm": 0.1177113875746727, "learning_rate": 2.6656131926933066e-05, "loss": 0.0053, "step": 67190 }, { "grad_norm": 0.0713159590959549, "learning_rate": 2.664151120505136e-05, "loss": 0.0042, "step": 67200 }, { "grad_norm": 0.1469849795103073, "learning_rate": 2.6626893037618873e-05, "loss": 0.0046, "step": 67210 }, { "grad_norm": 0.08994774520397186, "learning_rate": 2.6612277426234243e-05, "loss": 0.0061, "step": 67220 }, { "grad_norm": 0.12240509688854218, "learning_rate": 2.659766437249577e-05, "loss": 0.0044, "step": 67230 }, { "grad_norm": 0.0835346132516861, "learning_rate": 2.6583053878001557e-05, "loss": 0.0051, "step": 67240 }, { "grad_norm": 0.11417347192764282, "learning_rate": 2.6568445944349323e-05, "loss": 0.0048, "step": 67250 }, { "grad_norm": 0.08491148799657822, "learning_rate": 2.6553840573136657e-05, "loss": 0.0044, "step": 67260 }, { "grad_norm": 0.0678926631808281, "learning_rate": 2.6539237765960706e-05, "loss": 0.0046, "step": 67270 }, { "grad_norm": 0.09534827619791031, "learning_rate": 2.6524637524418465e-05, "loss": 0.0057, "step": 67280 }, { "grad_norm": 0.0949302539229393, "learning_rate": 2.651003985010655e-05, "loss": 0.0045, "step": 67290 }, { "grad_norm": 0.10639575123786926, "learning_rate": 2.649544474462138e-05, "loss": 0.0051, "step": 67300 }, { "grad_norm": 0.08158640563488007, "learning_rate": 2.6480852209559015e-05, "loss": 0.0045, "step": 67310 }, { "grad_norm": 0.12632028758525848, "learning_rate": 2.6466262246515296e-05, "loss": 0.0056, "step": 67320 }, { "grad_norm": 0.12710876762866974, "learning_rate": 2.6451674857085746e-05, "loss": 0.0048, "step": 67330 }, { "grad_norm": 0.09783072024583817, "learning_rate": 2.6437090042865655e-05, "loss": 0.005, "step": 67340 }, { "grad_norm": 0.06971550732851028, "learning_rate": 2.6422507805449947e-05, "loss": 0.005, "step": 67350 }, { "grad_norm": 0.08940466493368149, "learning_rate": 2.6407928146433357e-05, "loss": 0.0044, "step": 67360 }, { "grad_norm": 0.08760768175125122, "learning_rate": 2.639335106741026e-05, "loss": 0.0034, "step": 67370 }, { "grad_norm": 0.11436456441879272, "learning_rate": 2.6378776569974796e-05, "loss": 0.0058, "step": 67380 }, { "grad_norm": 0.07536433637142181, "learning_rate": 2.6364204655720814e-05, "loss": 0.0051, "step": 67390 }, { "grad_norm": 0.07891402393579483, "learning_rate": 2.6349635326241894e-05, "loss": 0.0046, "step": 67400 }, { "grad_norm": 0.08915268629789352, "learning_rate": 2.633506858313127e-05, "loss": 0.004, "step": 67410 }, { "grad_norm": 0.1283470243215561, "learning_rate": 2.6320504427981975e-05, "loss": 0.0063, "step": 67420 }, { "grad_norm": 0.10444795340299606, "learning_rate": 2.6305942862386734e-05, "loss": 0.0059, "step": 67430 }, { "grad_norm": 0.08915167301893234, "learning_rate": 2.629138388793793e-05, "loss": 0.0052, "step": 67440 }, { "grad_norm": 0.09757548570632935, "learning_rate": 2.6276827506227743e-05, "loss": 0.0053, "step": 67450 }, { "grad_norm": 0.12385875731706619, "learning_rate": 2.6262273718848037e-05, "loss": 0.0055, "step": 67460 }, { "grad_norm": 0.08061225712299347, "learning_rate": 2.6247722527390407e-05, "loss": 0.005, "step": 67470 }, { "grad_norm": 0.07998025417327881, "learning_rate": 2.6233173933446104e-05, "loss": 0.0042, "step": 67480 }, { "grad_norm": 0.06532354652881622, "learning_rate": 2.6218627938606188e-05, "loss": 0.0043, "step": 67490 }, { "grad_norm": 0.10429532080888748, "learning_rate": 2.620408454446135e-05, "loss": 0.0051, "step": 67500 }, { "grad_norm": 0.10327699780464172, "learning_rate": 2.6189543752602046e-05, "loss": 0.0055, "step": 67510 }, { "grad_norm": 0.09175742417573929, "learning_rate": 2.6175005564618444e-05, "loss": 0.006, "step": 67520 }, { "grad_norm": 0.11210986971855164, "learning_rate": 2.6160469982100428e-05, "loss": 0.005, "step": 67530 }, { "grad_norm": 0.08380530029535294, "learning_rate": 2.6145937006637555e-05, "loss": 0.0047, "step": 67540 }, { "grad_norm": 0.15748624503612518, "learning_rate": 2.6131406639819166e-05, "loss": 0.0056, "step": 67550 }, { "grad_norm": 0.11241015046834946, "learning_rate": 2.611687888323424e-05, "loss": 0.0044, "step": 67560 }, { "grad_norm": 0.10262670367956161, "learning_rate": 2.610235373847153e-05, "loss": 0.004, "step": 67570 }, { "grad_norm": 0.09001893550157547, "learning_rate": 2.6087831207119494e-05, "loss": 0.0055, "step": 67580 }, { "grad_norm": 0.07258393615484238, "learning_rate": 2.60733112907663e-05, "loss": 0.0044, "step": 67590 }, { "grad_norm": 0.08610209077596664, "learning_rate": 2.6058793990999786e-05, "loss": 0.0042, "step": 67600 }, { "grad_norm": 0.11059325188398361, "learning_rate": 2.6044279309407593e-05, "loss": 0.0053, "step": 67610 }, { "grad_norm": 0.15179017186164856, "learning_rate": 2.6029767247576976e-05, "loss": 0.0058, "step": 67620 }, { "grad_norm": 0.07916733622550964, "learning_rate": 2.601525780709497e-05, "loss": 0.0047, "step": 67630 }, { "grad_norm": 0.09275417029857635, "learning_rate": 2.6000750989548318e-05, "loss": 0.0044, "step": 67640 }, { "grad_norm": 0.07851304113864899, "learning_rate": 2.5986246796523462e-05, "loss": 0.0045, "step": 67650 }, { "grad_norm": 0.08363531529903412, "learning_rate": 2.5971745229606546e-05, "loss": 0.005, "step": 67660 }, { "grad_norm": 0.11485539376735687, "learning_rate": 2.5957246290383442e-05, "loss": 0.0042, "step": 67670 }, { "grad_norm": 0.11310196667909622, "learning_rate": 2.594274998043975e-05, "loss": 0.0061, "step": 67680 }, { "grad_norm": 0.162906676530838, "learning_rate": 2.5928256301360733e-05, "loss": 0.0041, "step": 67690 }, { "grad_norm": 0.11301463842391968, "learning_rate": 2.591376525473142e-05, "loss": 0.005, "step": 67700 }, { "grad_norm": 0.09814959019422531, "learning_rate": 2.5899276842136523e-05, "loss": 0.0057, "step": 67710 }, { "grad_norm": 0.081241175532341, "learning_rate": 2.5884791065160495e-05, "loss": 0.0045, "step": 67720 }, { "grad_norm": 0.12637363374233246, "learning_rate": 2.5870307925387436e-05, "loss": 0.0046, "step": 67730 }, { "grad_norm": 0.13217930495738983, "learning_rate": 2.585582742440125e-05, "loss": 0.0058, "step": 67740 }, { "grad_norm": 0.12641900777816772, "learning_rate": 2.5841349563785455e-05, "loss": 0.0054, "step": 67750 }, { "grad_norm": 0.09819933027029037, "learning_rate": 2.5826874345123355e-05, "loss": 0.0046, "step": 67760 }, { "grad_norm": 0.08314433693885803, "learning_rate": 2.581240176999792e-05, "loss": 0.0045, "step": 67770 }, { "grad_norm": 0.10236753523349762, "learning_rate": 2.579793183999189e-05, "loss": 0.0044, "step": 67780 }, { "grad_norm": 0.08408879488706589, "learning_rate": 2.5783464556687627e-05, "loss": 0.0044, "step": 67790 }, { "grad_norm": 0.08499147742986679, "learning_rate": 2.576899992166729e-05, "loss": 0.0056, "step": 67800 }, { "grad_norm": 0.09838172048330307, "learning_rate": 2.5754537936512668e-05, "loss": 0.0045, "step": 67810 }, { "grad_norm": 0.09663784503936768, "learning_rate": 2.574007860280535e-05, "loss": 0.0055, "step": 67820 }, { "grad_norm": 0.08584654331207275, "learning_rate": 2.5725621922126518e-05, "loss": 0.0046, "step": 67830 }, { "grad_norm": 0.11129152774810791, "learning_rate": 2.571116789605722e-05, "loss": 0.0052, "step": 67840 }, { "grad_norm": 0.09090495854616165, "learning_rate": 2.5696716526178054e-05, "loss": 0.004, "step": 67850 }, { "grad_norm": 0.14757205545902252, "learning_rate": 2.568226781406946e-05, "loss": 0.0056, "step": 67860 }, { "grad_norm": 0.08942075073719025, "learning_rate": 2.5667821761311474e-05, "loss": 0.0045, "step": 67870 }, { "grad_norm": 0.08426379412412643, "learning_rate": 2.5653378369483917e-05, "loss": 0.0041, "step": 67880 }, { "grad_norm": 0.09235217422246933, "learning_rate": 2.5638937640166292e-05, "loss": 0.0059, "step": 67890 }, { "grad_norm": 0.07589922845363617, "learning_rate": 2.5624499574937842e-05, "loss": 0.0056, "step": 67900 }, { "grad_norm": 0.09026651829481125, "learning_rate": 2.5610064175377456e-05, "loss": 0.0045, "step": 67910 }, { "grad_norm": 0.08921296894550323, "learning_rate": 2.5595631443063782e-05, "loss": 0.0056, "step": 67920 }, { "grad_norm": 0.09032171964645386, "learning_rate": 2.5581201379575177e-05, "loss": 0.0069, "step": 67930 }, { "grad_norm": 0.10849671065807343, "learning_rate": 2.5566773986489655e-05, "loss": 0.007, "step": 67940 }, { "grad_norm": 0.10094346106052399, "learning_rate": 2.5552349265385e-05, "loss": 0.0043, "step": 67950 }, { "grad_norm": 0.10769080370664597, "learning_rate": 2.5537927217838675e-05, "loss": 0.0061, "step": 67960 }, { "grad_norm": 0.11177406460046768, "learning_rate": 2.5523507845427864e-05, "loss": 0.0042, "step": 67970 }, { "grad_norm": 0.07461186498403549, "learning_rate": 2.5509091149729413e-05, "loss": 0.0046, "step": 67980 }, { "grad_norm": 0.12151596695184708, "learning_rate": 2.5494677132319962e-05, "loss": 0.0054, "step": 67990 }, { "grad_norm": 0.08475412428379059, "learning_rate": 2.548026579477575e-05, "loss": 0.0047, "step": 68000 }, { "grad_norm": 0.09455665946006775, "learning_rate": 2.5465857138672834e-05, "loss": 0.0049, "step": 68010 }, { "grad_norm": 0.10467153787612915, "learning_rate": 2.5451451165586848e-05, "loss": 0.0049, "step": 68020 }, { "grad_norm": 0.07607178390026093, "learning_rate": 2.543704787709329e-05, "loss": 0.004, "step": 68030 }, { "grad_norm": 0.10068974643945694, "learning_rate": 2.542264727476723e-05, "loss": 0.0045, "step": 68040 }, { "grad_norm": 0.09444135427474976, "learning_rate": 2.5408249360183535e-05, "loss": 0.0048, "step": 68050 }, { "grad_norm": 0.08599283546209335, "learning_rate": 2.5393854134916694e-05, "loss": 0.0069, "step": 68060 }, { "grad_norm": 0.08749951422214508, "learning_rate": 2.5379461600540993e-05, "loss": 0.0048, "step": 68070 }, { "grad_norm": 0.07966301590204239, "learning_rate": 2.5365071758630317e-05, "loss": 0.004, "step": 68080 }, { "grad_norm": 0.09143434464931488, "learning_rate": 2.535068461075839e-05, "loss": 0.0043, "step": 68090 }, { "grad_norm": 0.11284508556127548, "learning_rate": 2.5336300158498516e-05, "loss": 0.0077, "step": 68100 }, { "grad_norm": 0.08966857194900513, "learning_rate": 2.5321918403423793e-05, "loss": 0.006, "step": 68110 }, { "grad_norm": 0.08176808804273605, "learning_rate": 2.5307539347106957e-05, "loss": 0.0053, "step": 68120 }, { "grad_norm": 0.09978803992271423, "learning_rate": 2.529316299112049e-05, "loss": 0.0058, "step": 68130 }, { "grad_norm": 0.09728537499904633, "learning_rate": 2.5278789337036567e-05, "loss": 0.0042, "step": 68140 }, { "grad_norm": 0.08524754643440247, "learning_rate": 2.52644183864271e-05, "loss": 0.0041, "step": 68150 }, { "grad_norm": 0.10865573585033417, "learning_rate": 2.525005014086363e-05, "loss": 0.0062, "step": 68160 }, { "grad_norm": 0.08076240867376328, "learning_rate": 2.5235684601917465e-05, "loss": 0.006, "step": 68170 }, { "grad_norm": 0.11935889720916748, "learning_rate": 2.5221321771159618e-05, "loss": 0.0069, "step": 68180 }, { "grad_norm": 0.09034457802772522, "learning_rate": 2.5206961650160743e-05, "loss": 0.0053, "step": 68190 }, { "grad_norm": 0.08614625781774521, "learning_rate": 2.519260424049129e-05, "loss": 0.0059, "step": 68200 }, { "grad_norm": 0.11124823242425919, "learning_rate": 2.5178249543721304e-05, "loss": 0.0051, "step": 68210 }, { "grad_norm": 0.07680539786815643, "learning_rate": 2.5163897561420668e-05, "loss": 0.0039, "step": 68220 }, { "grad_norm": 0.07114569842815399, "learning_rate": 2.514954829515883e-05, "loss": 0.0045, "step": 68230 }, { "grad_norm": 0.08369407802820206, "learning_rate": 2.513520174650505e-05, "loss": 0.0039, "step": 68240 }, { "grad_norm": 0.0821036547422409, "learning_rate": 2.5120857917028207e-05, "loss": 0.0044, "step": 68250 }, { "grad_norm": 0.08810825645923615, "learning_rate": 2.5106516808296953e-05, "loss": 0.0044, "step": 68260 }, { "grad_norm": 0.08927358686923981, "learning_rate": 2.5092178421879552e-05, "loss": 0.0048, "step": 68270 }, { "grad_norm": 0.10204208642244339, "learning_rate": 2.5077842759344116e-05, "loss": 0.0047, "step": 68280 }, { "grad_norm": 0.07429880648851395, "learning_rate": 2.5063509822258304e-05, "loss": 0.0043, "step": 68290 }, { "grad_norm": 0.10432550311088562, "learning_rate": 2.5049179612189576e-05, "loss": 0.0058, "step": 68300 }, { "grad_norm": 0.10921239107847214, "learning_rate": 2.5034852130705043e-05, "loss": 0.0049, "step": 68310 }, { "grad_norm": 0.08608651906251907, "learning_rate": 2.5020527379371572e-05, "loss": 0.0047, "step": 68320 }, { "grad_norm": 0.13846783339977264, "learning_rate": 2.500620535975564e-05, "loss": 0.0067, "step": 68330 }, { "grad_norm": 0.12519851326942444, "learning_rate": 2.4991886073423515e-05, "loss": 0.0039, "step": 68340 }, { "grad_norm": 0.07565342634916306, "learning_rate": 2.4977569521941134e-05, "loss": 0.005, "step": 68350 }, { "grad_norm": 0.08147823810577393, "learning_rate": 2.4963255706874143e-05, "loss": 0.0045, "step": 68360 }, { "grad_norm": 0.0940353125333786, "learning_rate": 2.4948944629787853e-05, "loss": 0.0042, "step": 68370 }, { "grad_norm": 0.11580944061279297, "learning_rate": 2.493463629224731e-05, "loss": 0.0052, "step": 68380 }, { "grad_norm": 0.08310884237289429, "learning_rate": 2.492033069581728e-05, "loss": 0.0039, "step": 68390 }, { "grad_norm": 0.08693011105060577, "learning_rate": 2.4906027842062137e-05, "loss": 0.0042, "step": 68400 }, { "grad_norm": 0.08662471920251846, "learning_rate": 2.48917277325461e-05, "loss": 0.0045, "step": 68410 }, { "grad_norm": 0.1063520684838295, "learning_rate": 2.487743036883295e-05, "loss": 0.0059, "step": 68420 }, { "grad_norm": 0.10017713159322739, "learning_rate": 2.486313575248626e-05, "loss": 0.0053, "step": 68430 }, { "grad_norm": 0.08204812556505203, "learning_rate": 2.4848843885069235e-05, "loss": 0.0049, "step": 68440 }, { "grad_norm": 0.06324068456888199, "learning_rate": 2.483455476814484e-05, "loss": 0.0047, "step": 68450 }, { "grad_norm": 0.12169446796178818, "learning_rate": 2.4820268403275672e-05, "loss": 0.005, "step": 68460 }, { "grad_norm": 0.09819345921278, "learning_rate": 2.480598479202413e-05, "loss": 0.0042, "step": 68470 }, { "grad_norm": 0.14160972833633423, "learning_rate": 2.4791703935952193e-05, "loss": 0.0049, "step": 68480 }, { "grad_norm": 0.057654399424791336, "learning_rate": 2.477742583662163e-05, "loss": 0.0041, "step": 68490 }, { "grad_norm": 0.1515672653913498, "learning_rate": 2.476315049559384e-05, "loss": 0.0049, "step": 68500 }, { "grad_norm": 0.08405417203903198, "learning_rate": 2.4748877914429987e-05, "loss": 0.0049, "step": 68510 }, { "grad_norm": 0.09645441174507141, "learning_rate": 2.4734608094690857e-05, "loss": 0.0056, "step": 68520 }, { "grad_norm": 0.05655932426452637, "learning_rate": 2.4720341037937006e-05, "loss": 0.0051, "step": 68530 }, { "grad_norm": 0.08172344416379929, "learning_rate": 2.4706076745728645e-05, "loss": 0.0047, "step": 68540 }, { "grad_norm": 0.08948298543691635, "learning_rate": 2.469181521962573e-05, "loss": 0.0043, "step": 68550 }, { "grad_norm": 0.11962348222732544, "learning_rate": 2.467755646118783e-05, "loss": 0.0038, "step": 68560 }, { "grad_norm": 0.13006143271923065, "learning_rate": 2.46633004719743e-05, "loss": 0.0047, "step": 68570 }, { "grad_norm": 0.09653276205062866, "learning_rate": 2.4649047253544123e-05, "loss": 0.0045, "step": 68580 }, { "grad_norm": 0.08126359432935715, "learning_rate": 2.4634796807456016e-05, "loss": 0.0063, "step": 68590 }, { "grad_norm": 0.10461877286434174, "learning_rate": 2.46205491352684e-05, "loss": 0.0059, "step": 68600 }, { "grad_norm": 0.11802169680595398, "learning_rate": 2.4606304238539375e-05, "loss": 0.0068, "step": 68610 }, { "grad_norm": 0.10430028289556503, "learning_rate": 2.4592062118826753e-05, "loss": 0.0048, "step": 68620 }, { "grad_norm": 0.0859459638595581, "learning_rate": 2.4577822777688004e-05, "loss": 0.004, "step": 68630 }, { "grad_norm": 0.08893893659114838, "learning_rate": 2.4563586216680347e-05, "loss": 0.0047, "step": 68640 }, { "grad_norm": 0.08758720010519028, "learning_rate": 2.4549352437360622e-05, "loss": 0.006, "step": 68650 }, { "grad_norm": 0.15327326953411102, "learning_rate": 2.4535121441285493e-05, "loss": 0.0049, "step": 68660 }, { "grad_norm": 0.065967857837677, "learning_rate": 2.4520893230011172e-05, "loss": 0.0051, "step": 68670 }, { "grad_norm": 0.0844891294836998, "learning_rate": 2.450666780509368e-05, "loss": 0.0038, "step": 68680 }, { "grad_norm": 0.08483294397592545, "learning_rate": 2.4492445168088656e-05, "loss": 0.0047, "step": 68690 }, { "grad_norm": 0.10790292173624039, "learning_rate": 2.447822532055149e-05, "loss": 0.0046, "step": 68700 }, { "grad_norm": 0.0729367807507515, "learning_rate": 2.4464008264037212e-05, "loss": 0.0039, "step": 68710 }, { "grad_norm": 0.08568164706230164, "learning_rate": 2.4449794000100605e-05, "loss": 0.0056, "step": 68720 }, { "grad_norm": 0.08301640301942825, "learning_rate": 2.4435582530296108e-05, "loss": 0.0056, "step": 68730 }, { "grad_norm": 0.10678987950086594, "learning_rate": 2.4421373856177887e-05, "loss": 0.0041, "step": 68740 }, { "grad_norm": 0.12387432903051376, "learning_rate": 2.440716797929975e-05, "loss": 0.0044, "step": 68750 }, { "grad_norm": 0.12799128890037537, "learning_rate": 2.439296490121526e-05, "loss": 0.005, "step": 68760 }, { "grad_norm": 0.09969525039196014, "learning_rate": 2.4378764623477608e-05, "loss": 0.0052, "step": 68770 }, { "grad_norm": 0.09813813120126724, "learning_rate": 2.436456714763974e-05, "loss": 0.0047, "step": 68780 }, { "grad_norm": 0.10291548818349838, "learning_rate": 2.4350372475254262e-05, "loss": 0.0051, "step": 68790 }, { "grad_norm": 0.1044069305062294, "learning_rate": 2.433618060787351e-05, "loss": 0.0053, "step": 68800 }, { "grad_norm": 0.08777055144309998, "learning_rate": 2.432199154704944e-05, "loss": 0.0053, "step": 68810 }, { "grad_norm": 0.09310944378376007, "learning_rate": 2.430780529433377e-05, "loss": 0.0046, "step": 68820 }, { "grad_norm": 0.1100979670882225, "learning_rate": 2.42936218512779e-05, "loss": 0.0051, "step": 68830 }, { "grad_norm": 0.10010778158903122, "learning_rate": 2.427944121943288e-05, "loss": 0.0043, "step": 68840 }, { "grad_norm": 0.09204923361539841, "learning_rate": 2.4265263400349497e-05, "loss": 0.0058, "step": 68850 }, { "grad_norm": 0.07107318937778473, "learning_rate": 2.4251088395578213e-05, "loss": 0.005, "step": 68860 }, { "grad_norm": 0.08668546378612518, "learning_rate": 2.4236916206669214e-05, "loss": 0.0045, "step": 68870 }, { "grad_norm": 0.07641501724720001, "learning_rate": 2.4222746835172304e-05, "loss": 0.0055, "step": 68880 }, { "grad_norm": 0.12217344343662262, "learning_rate": 2.4208580282637065e-05, "loss": 0.0039, "step": 68890 }, { "grad_norm": 0.10235679149627686, "learning_rate": 2.4194416550612687e-05, "loss": 0.0054, "step": 68900 }, { "grad_norm": 0.1089763417840004, "learning_rate": 2.418025564064812e-05, "loss": 0.0056, "step": 68910 }, { "grad_norm": 0.0800071582198143, "learning_rate": 2.4166097554291978e-05, "loss": 0.0043, "step": 68920 }, { "grad_norm": 0.11441252380609512, "learning_rate": 2.4151942293092583e-05, "loss": 0.0048, "step": 68930 }, { "grad_norm": 0.13350041210651398, "learning_rate": 2.41377898585979e-05, "loss": 0.0055, "step": 68940 }, { "grad_norm": 0.07950355857610703, "learning_rate": 2.4123640252355652e-05, "loss": 0.0044, "step": 68950 }, { "grad_norm": 0.0774020105600357, "learning_rate": 2.410949347591318e-05, "loss": 0.0061, "step": 68960 }, { "grad_norm": 0.0660601407289505, "learning_rate": 2.4095349530817578e-05, "loss": 0.004, "step": 68970 }, { "grad_norm": 0.08191036432981491, "learning_rate": 2.4081208418615603e-05, "loss": 0.0043, "step": 68980 }, { "grad_norm": 0.11136860400438309, "learning_rate": 2.4067070140853725e-05, "loss": 0.0066, "step": 68990 }, { "grad_norm": 0.12555193901062012, "learning_rate": 2.405293469907805e-05, "loss": 0.005, "step": 69000 }, { "grad_norm": 0.1246752217411995, "learning_rate": 2.4038802094834446e-05, "loss": 0.0063, "step": 69010 }, { "grad_norm": 0.1200701892375946, "learning_rate": 2.4024672329668395e-05, "loss": 0.0045, "step": 69020 }, { "grad_norm": 0.0976206511259079, "learning_rate": 2.4010545405125125e-05, "loss": 0.0046, "step": 69030 }, { "grad_norm": 0.10399601608514786, "learning_rate": 2.3996421322749528e-05, "loss": 0.0066, "step": 69040 }, { "grad_norm": 0.07039480656385422, "learning_rate": 2.3982300084086224e-05, "loss": 0.0042, "step": 69050 }, { "grad_norm": 0.11196252703666687, "learning_rate": 2.396818169067944e-05, "loss": 0.0078, "step": 69060 }, { "grad_norm": 0.10782202333211899, "learning_rate": 2.3954066144073178e-05, "loss": 0.0051, "step": 69070 }, { "grad_norm": 0.0979488343000412, "learning_rate": 2.39399534458111e-05, "loss": 0.004, "step": 69080 }, { "grad_norm": 0.06733447313308716, "learning_rate": 2.392584359743651e-05, "loss": 0.0038, "step": 69090 }, { "grad_norm": 0.12631122767925262, "learning_rate": 2.3911736600492463e-05, "loss": 0.006, "step": 69100 }, { "grad_norm": 0.11613089591264725, "learning_rate": 2.3897632456521672e-05, "loss": 0.0042, "step": 69110 }, { "grad_norm": 0.15226447582244873, "learning_rate": 2.388353116706658e-05, "loss": 0.0045, "step": 69120 }, { "grad_norm": 0.09952570497989655, "learning_rate": 2.3869432733669232e-05, "loss": 0.0066, "step": 69130 }, { "grad_norm": 0.1071367934346199, "learning_rate": 2.385533715787145e-05, "loss": 0.0046, "step": 69140 }, { "grad_norm": 0.09661617875099182, "learning_rate": 2.3841244441214673e-05, "loss": 0.0052, "step": 69150 }, { "grad_norm": 0.1287042498588562, "learning_rate": 2.3827154585240073e-05, "loss": 0.0058, "step": 69160 }, { "grad_norm": 0.1450715810060501, "learning_rate": 2.3813067591488498e-05, "loss": 0.0062, "step": 69170 }, { "grad_norm": 0.10292787104845047, "learning_rate": 2.3798983461500494e-05, "loss": 0.0056, "step": 69180 }, { "grad_norm": 0.14286449551582336, "learning_rate": 2.3784902196816256e-05, "loss": 0.0054, "step": 69190 }, { "grad_norm": 0.10964709520339966, "learning_rate": 2.3770823798975716e-05, "loss": 0.004, "step": 69200 }, { "grad_norm": 0.10683911293745041, "learning_rate": 2.3756748269518424e-05, "loss": 0.0045, "step": 69210 }, { "grad_norm": 0.11538712680339813, "learning_rate": 2.374267560998371e-05, "loss": 0.0058, "step": 69220 }, { "grad_norm": 0.10125377774238586, "learning_rate": 2.3728605821910477e-05, "loss": 0.0052, "step": 69230 }, { "grad_norm": 0.10445237159729004, "learning_rate": 2.3714538906837453e-05, "loss": 0.0051, "step": 69240 }, { "grad_norm": 0.09202481806278229, "learning_rate": 2.3700474866302913e-05, "loss": 0.0052, "step": 69250 }, { "grad_norm": 0.12667793035507202, "learning_rate": 2.3686413701844917e-05, "loss": 0.0046, "step": 69260 }, { "grad_norm": 0.08363362401723862, "learning_rate": 2.367235541500114e-05, "loss": 0.0051, "step": 69270 }, { "grad_norm": 0.10763859003782272, "learning_rate": 2.3658300007308993e-05, "loss": 0.0045, "step": 69280 }, { "grad_norm": 0.11784294247627258, "learning_rate": 2.364424748030555e-05, "loss": 0.0057, "step": 69290 }, { "grad_norm": 0.10395275801420212, "learning_rate": 2.3630197835527595e-05, "loss": 0.0036, "step": 69300 }, { "grad_norm": 0.10817301273345947, "learning_rate": 2.3616151074511546e-05, "loss": 0.0041, "step": 69310 }, { "grad_norm": 0.12164762616157532, "learning_rate": 2.3602107198793545e-05, "loss": 0.0047, "step": 69320 }, { "grad_norm": 0.07922053337097168, "learning_rate": 2.3588066209909427e-05, "loss": 0.0043, "step": 69330 }, { "grad_norm": 0.11745491623878479, "learning_rate": 2.3574028109394662e-05, "loss": 0.0082, "step": 69340 }, { "grad_norm": 0.12242481857538223, "learning_rate": 2.3559992898784454e-05, "loss": 0.0044, "step": 69350 }, { "grad_norm": 0.09723532199859619, "learning_rate": 2.354596057961367e-05, "loss": 0.0051, "step": 69360 }, { "grad_norm": 0.09633292257785797, "learning_rate": 2.353193115341688e-05, "loss": 0.0054, "step": 69370 }, { "grad_norm": 0.08348283916711807, "learning_rate": 2.3517904621728294e-05, "loss": 0.0048, "step": 69380 }, { "grad_norm": 0.0674317255616188, "learning_rate": 2.3503880986081855e-05, "loss": 0.0039, "step": 69390 }, { "grad_norm": 0.07920102775096893, "learning_rate": 2.3489860248011136e-05, "loss": 0.005, "step": 69400 }, { "grad_norm": 0.13135941326618195, "learning_rate": 2.3475842409049464e-05, "loss": 0.0042, "step": 69410 }, { "grad_norm": 0.08323384076356888, "learning_rate": 2.346182747072975e-05, "loss": 0.005, "step": 69420 }, { "grad_norm": 0.09859203547239304, "learning_rate": 2.3447815434584718e-05, "loss": 0.0056, "step": 69430 }, { "grad_norm": 0.10778512805700302, "learning_rate": 2.3433806302146656e-05, "loss": 0.0045, "step": 69440 }, { "grad_norm": 0.08592332154512405, "learning_rate": 2.3419800074947612e-05, "loss": 0.0047, "step": 69450 }, { "grad_norm": 0.1435687094926834, "learning_rate": 2.3405796754519248e-05, "loss": 0.0046, "step": 69460 }, { "grad_norm": 0.11083235591650009, "learning_rate": 2.3391796342392992e-05, "loss": 0.0046, "step": 69470 }, { "grad_norm": 0.09452378004789352, "learning_rate": 2.337779884009984e-05, "loss": 0.0045, "step": 69480 }, { "grad_norm": 0.11456867307424545, "learning_rate": 2.3363804249170624e-05, "loss": 0.0051, "step": 69490 }, { "grad_norm": 0.1036602333188057, "learning_rate": 2.3349812571135703e-05, "loss": 0.0047, "step": 69500 }, { "grad_norm": 0.08356144279241562, "learning_rate": 2.333582380752523e-05, "loss": 0.0059, "step": 69510 }, { "grad_norm": 0.2784084379673004, "learning_rate": 2.3321837959868963e-05, "loss": 0.0051, "step": 69520 }, { "grad_norm": 0.06515055894851685, "learning_rate": 2.3307855029696383e-05, "loss": 0.0046, "step": 69530 }, { "grad_norm": 0.091393381357193, "learning_rate": 2.3293875018536648e-05, "loss": 0.0044, "step": 69540 }, { "grad_norm": 0.13817153871059418, "learning_rate": 2.327989792791861e-05, "loss": 0.0062, "step": 69550 }, { "grad_norm": 0.1326133906841278, "learning_rate": 2.326592375937074e-05, "loss": 0.005, "step": 69560 }, { "grad_norm": 0.11597209423780441, "learning_rate": 2.3251952514421248e-05, "loss": 0.0048, "step": 69570 }, { "grad_norm": 0.12434261292219162, "learning_rate": 2.323798419459804e-05, "loss": 0.0051, "step": 69580 }, { "grad_norm": 0.09565339237451553, "learning_rate": 2.3224018801428625e-05, "loss": 0.0043, "step": 69590 }, { "grad_norm": 0.08349014818668365, "learning_rate": 2.3210056336440273e-05, "loss": 0.0045, "step": 69600 }, { "grad_norm": 0.0922996997833252, "learning_rate": 2.3196096801159844e-05, "loss": 0.0061, "step": 69610 }, { "grad_norm": 0.08015062659978867, "learning_rate": 2.3182140197114012e-05, "loss": 0.005, "step": 69620 }, { "grad_norm": 0.07671041041612625, "learning_rate": 2.3168186525828985e-05, "loss": 0.0037, "step": 69630 }, { "grad_norm": 0.13493357598781586, "learning_rate": 2.3154235788830757e-05, "loss": 0.0048, "step": 69640 }, { "grad_norm": 0.08333555608987808, "learning_rate": 2.3140287987644927e-05, "loss": 0.0048, "step": 69650 }, { "grad_norm": 0.08989030122756958, "learning_rate": 2.3126343123796833e-05, "loss": 0.0047, "step": 69660 }, { "grad_norm": 0.10788286477327347, "learning_rate": 2.3112401198811418e-05, "loss": 0.0036, "step": 69670 }, { "grad_norm": 0.09029155224561691, "learning_rate": 2.3098462214213418e-05, "loss": 0.0036, "step": 69680 }, { "grad_norm": 0.11234693974256516, "learning_rate": 2.308452617152712e-05, "loss": 0.0045, "step": 69690 }, { "grad_norm": 0.11809076368808746, "learning_rate": 2.3070593072276603e-05, "loss": 0.0053, "step": 69700 }, { "grad_norm": 0.14291885495185852, "learning_rate": 2.3056662917985518e-05, "loss": 0.0042, "step": 69710 }, { "grad_norm": 0.09157346934080124, "learning_rate": 2.3042735710177283e-05, "loss": 0.0059, "step": 69720 }, { "grad_norm": 0.09787211567163467, "learning_rate": 2.302881145037492e-05, "loss": 0.0046, "step": 69730 }, { "grad_norm": 0.09397575259208679, "learning_rate": 2.3014890140101192e-05, "loss": 0.004, "step": 69740 }, { "grad_norm": 0.07282618433237076, "learning_rate": 2.30009717808785e-05, "loss": 0.0054, "step": 69750 }, { "grad_norm": 0.09601973742246628, "learning_rate": 2.2987056374228967e-05, "loss": 0.0041, "step": 69760 }, { "grad_norm": 0.1086965948343277, "learning_rate": 2.297314392167431e-05, "loss": 0.0037, "step": 69770 }, { "grad_norm": 0.07606326788663864, "learning_rate": 2.295923442473601e-05, "loss": 0.0037, "step": 69780 }, { "grad_norm": 0.08770782500505447, "learning_rate": 2.294532788493519e-05, "loss": 0.0048, "step": 69790 }, { "grad_norm": 0.10196381062269211, "learning_rate": 2.293142430379262e-05, "loss": 0.0044, "step": 69800 }, { "grad_norm": 0.09993256628513336, "learning_rate": 2.291752368282879e-05, "loss": 0.0041, "step": 69810 }, { "grad_norm": 0.09247679263353348, "learning_rate": 2.290362602356385e-05, "loss": 0.005, "step": 69820 }, { "grad_norm": 0.09985306859016418, "learning_rate": 2.288973132751765e-05, "loss": 0.0056, "step": 69830 }, { "grad_norm": 0.08426962792873383, "learning_rate": 2.287583959620965e-05, "loss": 0.0045, "step": 69840 }, { "grad_norm": 0.10306230932474136, "learning_rate": 2.2861950831159074e-05, "loss": 0.0044, "step": 69850 }, { "grad_norm": 0.12001305818557739, "learning_rate": 2.284806503388472e-05, "loss": 0.0054, "step": 69860 }, { "grad_norm": 0.13625970482826233, "learning_rate": 2.283418220590518e-05, "loss": 0.005, "step": 69870 }, { "grad_norm": 0.12012579292058945, "learning_rate": 2.282030234873862e-05, "loss": 0.0045, "step": 69880 }, { "grad_norm": 0.12384737282991409, "learning_rate": 2.2806425463902948e-05, "loss": 0.0046, "step": 69890 }, { "grad_norm": 0.12161214649677277, "learning_rate": 2.279255155291568e-05, "loss": 0.0044, "step": 69900 }, { "grad_norm": 0.08481495082378387, "learning_rate": 2.277868061729409e-05, "loss": 0.0048, "step": 69910 }, { "grad_norm": 0.09861017763614655, "learning_rate": 2.276481265855504e-05, "loss": 0.0056, "step": 69920 }, { "grad_norm": 0.08804459124803543, "learning_rate": 2.2750947678215128e-05, "loss": 0.0054, "step": 69930 }, { "grad_norm": 0.07716041058301926, "learning_rate": 2.2737085677790616e-05, "loss": 0.004, "step": 69940 }, { "grad_norm": 0.08019058406352997, "learning_rate": 2.2723226658797443e-05, "loss": 0.0047, "step": 69950 }, { "grad_norm": 0.10852444916963577, "learning_rate": 2.270937062275117e-05, "loss": 0.0045, "step": 69960 }, { "grad_norm": 0.10663316398859024, "learning_rate": 2.2695517571167113e-05, "loss": 0.004, "step": 69970 }, { "grad_norm": 0.10741426795721054, "learning_rate": 2.268166750556019e-05, "loss": 0.0056, "step": 69980 }, { "grad_norm": 0.10860413312911987, "learning_rate": 2.266782042744503e-05, "loss": 0.0054, "step": 69990 }, { "grad_norm": 0.13731390237808228, "learning_rate": 2.2653976338335935e-05, "loss": 0.005, "step": 70000 }, { "grad_norm": 0.06677502393722534, "learning_rate": 2.2640135239746895e-05, "loss": 0.0044, "step": 70010 }, { "grad_norm": 0.1059170812368393, "learning_rate": 2.2626297133191515e-05, "loss": 0.004, "step": 70020 }, { "grad_norm": 0.11250566691160202, "learning_rate": 2.261246202018312e-05, "loss": 0.0056, "step": 70030 }, { "grad_norm": 0.0822145864367485, "learning_rate": 2.2598629902234718e-05, "loss": 0.0039, "step": 70040 }, { "grad_norm": 0.07178038358688354, "learning_rate": 2.258480078085894e-05, "loss": 0.004, "step": 70050 }, { "grad_norm": 0.1036999300122261, "learning_rate": 2.2570974657568126e-05, "loss": 0.0044, "step": 70060 }, { "grad_norm": 0.1028108298778534, "learning_rate": 2.2557151533874286e-05, "loss": 0.0055, "step": 70070 }, { "grad_norm": 0.0693674311041832, "learning_rate": 2.2543331411289116e-05, "loss": 0.0036, "step": 70080 }, { "grad_norm": 0.06494753807783127, "learning_rate": 2.252951429132392e-05, "loss": 0.0043, "step": 70090 }, { "grad_norm": 0.12728242576122284, "learning_rate": 2.2515700175489762e-05, "loss": 0.0049, "step": 70100 }, { "grad_norm": 0.0921732634305954, "learning_rate": 2.2501889065297288e-05, "loss": 0.0036, "step": 70110 }, { "grad_norm": 0.11119425296783447, "learning_rate": 2.2488080962256887e-05, "loss": 0.005, "step": 70120 }, { "grad_norm": 0.10284118354320526, "learning_rate": 2.2474275867878585e-05, "loss": 0.0038, "step": 70130 }, { "grad_norm": 0.10504812002182007, "learning_rate": 2.246047378367211e-05, "loss": 0.0041, "step": 70140 }, { "grad_norm": 0.10813745111227036, "learning_rate": 2.2446674711146798e-05, "loss": 0.0068, "step": 70150 }, { "grad_norm": 0.18445280194282532, "learning_rate": 2.2432878651811734e-05, "loss": 0.0067, "step": 70160 }, { "grad_norm": 0.10050738602876663, "learning_rate": 2.2419085607175594e-05, "loss": 0.0046, "step": 70170 }, { "grad_norm": 0.11395587772130966, "learning_rate": 2.240529557874678e-05, "loss": 0.0055, "step": 70180 }, { "grad_norm": 0.11227195709943771, "learning_rate": 2.2391508568033358e-05, "loss": 0.0043, "step": 70190 }, { "grad_norm": 0.21654947102069855, "learning_rate": 2.237772457654307e-05, "loss": 0.0033, "step": 70200 }, { "grad_norm": 0.0740555003285408, "learning_rate": 2.236394360578327e-05, "loss": 0.0057, "step": 70210 }, { "grad_norm": 0.09436158090829849, "learning_rate": 2.2350165657261073e-05, "loss": 0.0049, "step": 70220 }, { "grad_norm": 0.10423152893781662, "learning_rate": 2.2336390732483164e-05, "loss": 0.0063, "step": 70230 }, { "grad_norm": 0.10399322211742401, "learning_rate": 2.2322618832955976e-05, "loss": 0.0061, "step": 70240 }, { "grad_norm": 0.09418358653783798, "learning_rate": 2.2308849960185578e-05, "loss": 0.0035, "step": 70250 }, { "grad_norm": 0.08277469873428345, "learning_rate": 2.229508411567774e-05, "loss": 0.0041, "step": 70260 }, { "grad_norm": 0.10870333760976791, "learning_rate": 2.2281321300937823e-05, "loss": 0.0045, "step": 70270 }, { "grad_norm": 0.0982934758067131, "learning_rate": 2.2267561517470937e-05, "loss": 0.0047, "step": 70280 }, { "grad_norm": 0.07740792632102966, "learning_rate": 2.2253804766781845e-05, "loss": 0.0049, "step": 70290 }, { "grad_norm": 0.09791765362024307, "learning_rate": 2.2240051050374933e-05, "loss": 0.0042, "step": 70300 }, { "grad_norm": 0.13633675873279572, "learning_rate": 2.22263003697543e-05, "loss": 0.0056, "step": 70310 }, { "grad_norm": 0.09305690973997116, "learning_rate": 2.2212552726423702e-05, "loss": 0.0044, "step": 70320 }, { "grad_norm": 0.0900753065943718, "learning_rate": 2.2198808121886578e-05, "loss": 0.0057, "step": 70330 }, { "grad_norm": 0.08509325236082077, "learning_rate": 2.218506655764598e-05, "loss": 0.0037, "step": 70340 }, { "grad_norm": 0.08936042338609695, "learning_rate": 2.2171328035204704e-05, "loss": 0.0042, "step": 70350 }, { "grad_norm": 0.08820836991071701, "learning_rate": 2.2157592556065133e-05, "loss": 0.0041, "step": 70360 }, { "grad_norm": 0.09409570693969727, "learning_rate": 2.2143860121729382e-05, "loss": 0.0042, "step": 70370 }, { "grad_norm": 0.10827106237411499, "learning_rate": 2.2130130733699206e-05, "loss": 0.0043, "step": 70380 }, { "grad_norm": 0.08417564630508423, "learning_rate": 2.2116404393476052e-05, "loss": 0.0043, "step": 70390 }, { "grad_norm": 0.0733344629406929, "learning_rate": 2.2102681102560967e-05, "loss": 0.0055, "step": 70400 }, { "grad_norm": 0.10596372187137604, "learning_rate": 2.208896086245476e-05, "loss": 0.004, "step": 70410 }, { "grad_norm": 0.09450290352106094, "learning_rate": 2.2075243674657808e-05, "loss": 0.0039, "step": 70420 }, { "grad_norm": 0.098329558968544, "learning_rate": 2.206152954067024e-05, "loss": 0.0047, "step": 70430 }, { "grad_norm": 0.10196533054113388, "learning_rate": 2.2047818461991758e-05, "loss": 0.0069, "step": 70440 }, { "grad_norm": 0.07320927083492279, "learning_rate": 2.2034110440121864e-05, "loss": 0.0044, "step": 70450 }, { "grad_norm": 0.07354462146759033, "learning_rate": 2.2020405476559585e-05, "loss": 0.0053, "step": 70460 }, { "grad_norm": 0.08781851828098297, "learning_rate": 2.2006703572803722e-05, "loss": 0.006, "step": 70470 }, { "grad_norm": 0.08875878155231476, "learning_rate": 2.1993004730352646e-05, "loss": 0.0053, "step": 70480 }, { "grad_norm": 0.10429637879133224, "learning_rate": 2.1979308950704464e-05, "loss": 0.0038, "step": 70490 }, { "grad_norm": 0.09910040348768234, "learning_rate": 2.1965616235356923e-05, "loss": 0.0042, "step": 70500 }, { "grad_norm": 0.09929545223712921, "learning_rate": 2.1951926585807465e-05, "loss": 0.0041, "step": 70510 }, { "grad_norm": 0.10671210289001465, "learning_rate": 2.193824000355313e-05, "loss": 0.0033, "step": 70520 }, { "grad_norm": 0.0959111824631691, "learning_rate": 2.192455649009067e-05, "loss": 0.0057, "step": 70530 }, { "grad_norm": 0.11213543266057968, "learning_rate": 2.1910876046916523e-05, "loss": 0.0044, "step": 70540 }, { "grad_norm": 0.07328993827104568, "learning_rate": 2.1897198675526725e-05, "loss": 0.004, "step": 70550 }, { "grad_norm": 0.10816267877817154, "learning_rate": 2.1883524377417024e-05, "loss": 0.0053, "step": 70560 }, { "grad_norm": 0.0882125049829483, "learning_rate": 2.1869853154082825e-05, "loss": 0.004, "step": 70570 }, { "grad_norm": 0.09319769591093063, "learning_rate": 2.1856185007019215e-05, "loss": 0.0038, "step": 70580 }, { "grad_norm": 0.10889515280723572, "learning_rate": 2.1842519937720874e-05, "loss": 0.0059, "step": 70590 }, { "grad_norm": 0.06294591724872589, "learning_rate": 2.182885794768224e-05, "loss": 0.0043, "step": 70600 }, { "grad_norm": 0.10254374891519547, "learning_rate": 2.1815199038397326e-05, "loss": 0.0048, "step": 70610 }, { "grad_norm": 0.10357743501663208, "learning_rate": 2.1801543211359893e-05, "loss": 0.0046, "step": 70620 }, { "grad_norm": 0.08124114573001862, "learning_rate": 2.178789046806326e-05, "loss": 0.0041, "step": 70630 }, { "grad_norm": 0.09280645102262497, "learning_rate": 2.1774240810000547e-05, "loss": 0.0043, "step": 70640 }, { "grad_norm": 0.09398391842842102, "learning_rate": 2.17605942386644e-05, "loss": 0.0039, "step": 70650 }, { "grad_norm": 0.12466003000736237, "learning_rate": 2.174695075554723e-05, "loss": 0.0053, "step": 70660 }, { "grad_norm": 0.16137397289276123, "learning_rate": 2.173331036214103e-05, "loss": 0.0049, "step": 70670 }, { "grad_norm": 0.14022225141525269, "learning_rate": 2.171967305993753e-05, "loss": 0.0045, "step": 70680 }, { "grad_norm": 0.14493460953235626, "learning_rate": 2.1706038850428024e-05, "loss": 0.0043, "step": 70690 }, { "grad_norm": 0.10035204887390137, "learning_rate": 2.169240773510361e-05, "loss": 0.004, "step": 70700 }, { "grad_norm": 0.09700331091880798, "learning_rate": 2.1678779715454906e-05, "loss": 0.0044, "step": 70710 }, { "grad_norm": 0.10613216459751129, "learning_rate": 2.166515479297229e-05, "loss": 0.0062, "step": 70720 }, { "grad_norm": 0.0863301083445549, "learning_rate": 2.1651532969145732e-05, "loss": 0.0044, "step": 70730 }, { "grad_norm": 0.12256637215614319, "learning_rate": 2.1637914245464903e-05, "loss": 0.0042, "step": 70740 }, { "grad_norm": 0.07087250798940659, "learning_rate": 2.1624298623419133e-05, "loss": 0.0037, "step": 70750 }, { "grad_norm": 0.10118553042411804, "learning_rate": 2.1610686104497413e-05, "loss": 0.0044, "step": 70760 }, { "grad_norm": 0.07040338218212128, "learning_rate": 2.1597076690188366e-05, "loss": 0.0043, "step": 70770 }, { "grad_norm": 0.07036466151475906, "learning_rate": 2.1583470381980315e-05, "loss": 0.005, "step": 70780 }, { "grad_norm": 0.14005926251411438, "learning_rate": 2.156986718136123e-05, "loss": 0.0051, "step": 70790 }, { "grad_norm": 0.1125299483537674, "learning_rate": 2.155626708981871e-05, "loss": 0.0046, "step": 70800 }, { "grad_norm": 0.07951019704341888, "learning_rate": 2.1542670108840075e-05, "loss": 0.0049, "step": 70810 }, { "grad_norm": 0.08481071889400482, "learning_rate": 2.152907623991221e-05, "loss": 0.0038, "step": 70820 }, { "grad_norm": 0.08542224764823914, "learning_rate": 2.1515485484521802e-05, "loss": 0.0044, "step": 70830 }, { "grad_norm": 0.10831228643655777, "learning_rate": 2.150189784415506e-05, "loss": 0.0045, "step": 70840 }, { "grad_norm": 0.10302004963159561, "learning_rate": 2.148831332029794e-05, "loss": 0.004, "step": 70850 }, { "grad_norm": 0.04982871189713478, "learning_rate": 2.147473191443599e-05, "loss": 0.0039, "step": 70860 }, { "grad_norm": 0.0675809234380722, "learning_rate": 2.1461153628054492e-05, "loss": 0.0057, "step": 70870 }, { "grad_norm": 0.09736508131027222, "learning_rate": 2.1447578462638284e-05, "loss": 0.0047, "step": 70880 }, { "grad_norm": 0.09468915313482285, "learning_rate": 2.143400641967201e-05, "loss": 0.0039, "step": 70890 }, { "grad_norm": 0.08625539392232895, "learning_rate": 2.1420437500639823e-05, "loss": 0.0038, "step": 70900 }, { "grad_norm": 0.09111239016056061, "learning_rate": 2.1406871707025646e-05, "loss": 0.0049, "step": 70910 }, { "grad_norm": 0.09761685878038406, "learning_rate": 2.1393309040312963e-05, "loss": 0.0049, "step": 70920 }, { "grad_norm": 0.11594667285680771, "learning_rate": 2.137974950198502e-05, "loss": 0.0045, "step": 70930 }, { "grad_norm": 0.06820476800203323, "learning_rate": 2.1366193093524617e-05, "loss": 0.0043, "step": 70940 }, { "grad_norm": 0.08691661059856415, "learning_rate": 2.1352639816414288e-05, "loss": 0.0043, "step": 70950 }, { "grad_norm": 0.10804492980241776, "learning_rate": 2.1339089672136202e-05, "loss": 0.0062, "step": 70960 }, { "grad_norm": 0.10111045837402344, "learning_rate": 2.1325542662172193e-05, "loss": 0.0044, "step": 70970 }, { "grad_norm": 0.099835105240345, "learning_rate": 2.1311998788003712e-05, "loss": 0.0039, "step": 70980 }, { "grad_norm": 0.08532677590847015, "learning_rate": 2.1298458051111907e-05, "loss": 0.0038, "step": 70990 }, { "grad_norm": 0.07675503939390182, "learning_rate": 2.1284920452977602e-05, "loss": 0.0041, "step": 71000 }, { "grad_norm": 0.08278652280569077, "learning_rate": 2.12713859950812e-05, "loss": 0.0041, "step": 71010 }, { "grad_norm": 0.07999724894762039, "learning_rate": 2.1257854678902832e-05, "loss": 0.0034, "step": 71020 }, { "grad_norm": 0.09915410727262497, "learning_rate": 2.124432650592227e-05, "loss": 0.0043, "step": 71030 }, { "grad_norm": 0.06800558418035507, "learning_rate": 2.1230801477618937e-05, "loss": 0.0043, "step": 71040 }, { "grad_norm": 0.10043510049581528, "learning_rate": 2.121727959547189e-05, "loss": 0.0041, "step": 71050 }, { "grad_norm": 0.08578255027532578, "learning_rate": 2.1203760860959886e-05, "loss": 0.005, "step": 71060 }, { "grad_norm": 0.08147171884775162, "learning_rate": 2.119024527556127e-05, "loss": 0.0047, "step": 71070 }, { "grad_norm": 0.10760039836168289, "learning_rate": 2.117673284075415e-05, "loss": 0.0045, "step": 71080 }, { "grad_norm": 0.13278141617774963, "learning_rate": 2.1163223558016165e-05, "loss": 0.0053, "step": 71090 }, { "grad_norm": 0.11139386147260666, "learning_rate": 2.1149717428824722e-05, "loss": 0.0058, "step": 71100 }, { "grad_norm": 0.11311261355876923, "learning_rate": 2.1136214454656784e-05, "loss": 0.0054, "step": 71110 }, { "grad_norm": 0.09118767082691193, "learning_rate": 2.1122714636989056e-05, "loss": 0.0035, "step": 71120 }, { "grad_norm": 0.1294955462217331, "learning_rate": 2.1109217977297823e-05, "loss": 0.0051, "step": 71130 }, { "grad_norm": 0.08288717269897461, "learning_rate": 2.1095724477059077e-05, "loss": 0.0046, "step": 71140 }, { "grad_norm": 0.07218637317419052, "learning_rate": 2.1082234137748434e-05, "loss": 0.0038, "step": 71150 }, { "grad_norm": 0.09821029752492905, "learning_rate": 2.106874696084122e-05, "loss": 0.0039, "step": 71160 }, { "grad_norm": 0.08875329792499542, "learning_rate": 2.1055262947812315e-05, "loss": 0.0055, "step": 71170 }, { "grad_norm": 0.11515993624925613, "learning_rate": 2.104178210013636e-05, "loss": 0.0059, "step": 71180 }, { "grad_norm": 0.10638127475976944, "learning_rate": 2.1028304419287554e-05, "loss": 0.0049, "step": 71190 }, { "grad_norm": 0.12061542272567749, "learning_rate": 2.101482990673982e-05, "loss": 0.0046, "step": 71200 }, { "grad_norm": 0.08534588664770126, "learning_rate": 2.1001358563966707e-05, "loss": 0.004, "step": 71210 }, { "grad_norm": 0.08857712149620056, "learning_rate": 2.098789039244144e-05, "loss": 0.0045, "step": 71220 }, { "grad_norm": 0.05951235815882683, "learning_rate": 2.097442539363685e-05, "loss": 0.0046, "step": 71230 }, { "grad_norm": 0.10334403812885284, "learning_rate": 2.0960963569025448e-05, "loss": 0.0039, "step": 71240 }, { "grad_norm": 0.0938333198428154, "learning_rate": 2.0947504920079435e-05, "loss": 0.0044, "step": 71250 }, { "grad_norm": 0.09804055839776993, "learning_rate": 2.093404944827059e-05, "loss": 0.0049, "step": 71260 }, { "grad_norm": 0.07592467963695526, "learning_rate": 2.0920597155070394e-05, "loss": 0.0039, "step": 71270 }, { "grad_norm": 0.08430413156747818, "learning_rate": 2.090714804194997e-05, "loss": 0.0034, "step": 71280 }, { "grad_norm": 0.08601117134094238, "learning_rate": 2.0893702110380114e-05, "loss": 0.0054, "step": 71290 }, { "grad_norm": 0.07635900378227234, "learning_rate": 2.0880259361831216e-05, "loss": 0.0039, "step": 71300 }, { "grad_norm": 0.08966459333896637, "learning_rate": 2.0866819797773397e-05, "loss": 0.004, "step": 71310 }, { "grad_norm": 0.07050157338380814, "learning_rate": 2.085338341967634e-05, "loss": 0.0042, "step": 71320 }, { "grad_norm": 0.0927976667881012, "learning_rate": 2.083995022900946e-05, "loss": 0.004, "step": 71330 }, { "grad_norm": 0.12557440996170044, "learning_rate": 2.0826520227241768e-05, "loss": 0.0046, "step": 71340 }, { "grad_norm": 0.08385782688856125, "learning_rate": 2.081309341584199e-05, "loss": 0.0045, "step": 71350 }, { "grad_norm": 0.10705190896987915, "learning_rate": 2.0799669796278413e-05, "loss": 0.004, "step": 71360 }, { "grad_norm": 0.12937578558921814, "learning_rate": 2.078624937001906e-05, "loss": 0.0049, "step": 71370 }, { "grad_norm": 0.05613931640982628, "learning_rate": 2.0772832138531544e-05, "loss": 0.0065, "step": 71380 }, { "grad_norm": 0.08300302922725677, "learning_rate": 2.0759418103283158e-05, "loss": 0.0039, "step": 71390 }, { "grad_norm": 0.10066601634025574, "learning_rate": 2.0746007265740846e-05, "loss": 0.0049, "step": 71400 }, { "grad_norm": 0.08872409909963608, "learning_rate": 2.0732599627371215e-05, "loss": 0.0055, "step": 71410 }, { "grad_norm": 0.10391521453857422, "learning_rate": 2.0719195189640468e-05, "loss": 0.0044, "step": 71420 }, { "grad_norm": 0.11564406752586365, "learning_rate": 2.0705793954014528e-05, "loss": 0.0045, "step": 71430 }, { "grad_norm": 0.2351330816745758, "learning_rate": 2.06923959219589e-05, "loss": 0.0032, "step": 71440 }, { "grad_norm": 0.07181227207183838, "learning_rate": 2.0679001094938793e-05, "loss": 0.0045, "step": 71450 }, { "grad_norm": 0.1243915930390358, "learning_rate": 2.066560947441904e-05, "loss": 0.0054, "step": 71460 }, { "grad_norm": 0.10311000794172287, "learning_rate": 2.0652221061864152e-05, "loss": 0.0041, "step": 71470 }, { "grad_norm": 0.06950153410434723, "learning_rate": 2.0638835858738226e-05, "loss": 0.0042, "step": 71480 }, { "grad_norm": 0.08316893130540848, "learning_rate": 2.0625453866505068e-05, "loss": 0.0063, "step": 71490 }, { "grad_norm": 0.08614913374185562, "learning_rate": 2.0612075086628123e-05, "loss": 0.0032, "step": 71500 }, { "grad_norm": 0.10876154899597168, "learning_rate": 2.059869952057045e-05, "loss": 0.0065, "step": 71510 }, { "grad_norm": 0.07270019501447678, "learning_rate": 2.0585327169794793e-05, "loss": 0.0036, "step": 71520 }, { "grad_norm": 0.06993550062179565, "learning_rate": 2.0571958035763534e-05, "loss": 0.0042, "step": 71530 }, { "grad_norm": 0.08339273929595947, "learning_rate": 2.0558592119938715e-05, "loss": 0.005, "step": 71540 }, { "grad_norm": 0.10584250092506409, "learning_rate": 2.054522942378198e-05, "loss": 0.0047, "step": 71550 }, { "grad_norm": 0.08748111128807068, "learning_rate": 2.053186994875469e-05, "loss": 0.0045, "step": 71560 }, { "grad_norm": 0.0691889226436615, "learning_rate": 2.0518513696317775e-05, "loss": 0.0034, "step": 71570 }, { "grad_norm": 0.1371753066778183, "learning_rate": 2.050516066793188e-05, "loss": 0.0048, "step": 71580 }, { "grad_norm": 0.07791350781917572, "learning_rate": 2.0491810865057276e-05, "loss": 0.004, "step": 71590 }, { "grad_norm": 0.06943652778863907, "learning_rate": 2.0478464289153882e-05, "loss": 0.0036, "step": 71600 }, { "grad_norm": 0.060898084193468094, "learning_rate": 2.0465120941681232e-05, "loss": 0.0033, "step": 71610 }, { "grad_norm": 0.07144532352685928, "learning_rate": 2.0451780824098566e-05, "loss": 0.0038, "step": 71620 }, { "grad_norm": 0.11907369643449783, "learning_rate": 2.043844393786471e-05, "loss": 0.0051, "step": 71630 }, { "grad_norm": 0.09038542956113815, "learning_rate": 2.0425110284438193e-05, "loss": 0.0043, "step": 71640 }, { "grad_norm": 0.05372690409421921, "learning_rate": 2.041177986527712e-05, "loss": 0.0037, "step": 71650 }, { "grad_norm": 0.09352009743452072, "learning_rate": 2.0398452681839347e-05, "loss": 0.005, "step": 71660 }, { "grad_norm": 0.07722107321023941, "learning_rate": 2.0385128735582272e-05, "loss": 0.0037, "step": 71670 }, { "grad_norm": 0.05986359342932701, "learning_rate": 2.0371808027963003e-05, "loss": 0.0042, "step": 71680 }, { "grad_norm": 0.08523496240377426, "learning_rate": 2.035849056043825e-05, "loss": 0.0039, "step": 71690 }, { "grad_norm": 0.07478390634059906, "learning_rate": 2.03451763344644e-05, "loss": 0.0037, "step": 71700 }, { "grad_norm": 0.09932120889425278, "learning_rate": 2.0331865351497482e-05, "loss": 0.0051, "step": 71710 }, { "grad_norm": 0.10727892816066742, "learning_rate": 2.0318557612993185e-05, "loss": 0.0043, "step": 71720 }, { "grad_norm": 0.08911755681037903, "learning_rate": 2.0305253120406782e-05, "loss": 0.0051, "step": 71730 }, { "grad_norm": 0.10066209733486176, "learning_rate": 2.029195187519325e-05, "loss": 0.004, "step": 71740 }, { "grad_norm": 0.06372150033712387, "learning_rate": 2.0278653878807217e-05, "loss": 0.0037, "step": 71750 }, { "grad_norm": 0.2058730274438858, "learning_rate": 2.0265359132702892e-05, "loss": 0.0039, "step": 71760 }, { "grad_norm": 0.10335276275873184, "learning_rate": 2.0252067638334182e-05, "loss": 0.0042, "step": 71770 }, { "grad_norm": 0.08846868574619293, "learning_rate": 2.023877939715464e-05, "loss": 0.0059, "step": 71780 }, { "grad_norm": 0.07428035140037537, "learning_rate": 2.0225494410617448e-05, "loss": 0.0041, "step": 71790 }, { "grad_norm": 0.0738501325249672, "learning_rate": 2.02122126801754e-05, "loss": 0.0036, "step": 71800 }, { "grad_norm": 0.08676058053970337, "learning_rate": 2.0198934207281018e-05, "loss": 0.0037, "step": 71810 }, { "grad_norm": 0.14399108290672302, "learning_rate": 2.0185658993386357e-05, "loss": 0.0044, "step": 71820 }, { "grad_norm": 0.08049171417951584, "learning_rate": 2.0172387039943226e-05, "loss": 0.0035, "step": 71830 }, { "grad_norm": 0.05478488281369209, "learning_rate": 2.0159118348402968e-05, "loss": 0.0063, "step": 71840 }, { "grad_norm": 0.12270252406597137, "learning_rate": 2.0145852920216697e-05, "loss": 0.0047, "step": 71850 }, { "grad_norm": 0.08943402022123337, "learning_rate": 2.0132590756835045e-05, "loss": 0.0047, "step": 71860 }, { "grad_norm": 0.0891333594918251, "learning_rate": 2.0119331859708386e-05, "loss": 0.0038, "step": 71870 }, { "grad_norm": 0.05185883119702339, "learning_rate": 2.0106076230286654e-05, "loss": 0.0037, "step": 71880 }, { "grad_norm": 0.11954598128795624, "learning_rate": 2.0092823870019488e-05, "loss": 0.0038, "step": 71890 }, { "grad_norm": 0.06876859813928604, "learning_rate": 2.0079574780356115e-05, "loss": 0.0032, "step": 71900 }, { "grad_norm": 0.1302282214164734, "learning_rate": 2.0066328962745486e-05, "loss": 0.0039, "step": 71910 }, { "grad_norm": 0.14411379396915436, "learning_rate": 2.00530864186361e-05, "loss": 0.0039, "step": 71920 }, { "grad_norm": 0.08043494075536728, "learning_rate": 2.003984714947618e-05, "loss": 0.0042, "step": 71930 }, { "grad_norm": 0.08200064301490784, "learning_rate": 2.0026611156713515e-05, "loss": 0.0037, "step": 71940 }, { "grad_norm": 0.09784597903490067, "learning_rate": 2.0013378441795584e-05, "loss": 0.0039, "step": 71950 }, { "grad_norm": 0.06684480607509613, "learning_rate": 2.0000149006169506e-05, "loss": 0.0045, "step": 71960 }, { "grad_norm": 0.09728351980447769, "learning_rate": 1.9986922851282048e-05, "loss": 0.0074, "step": 71970 }, { "grad_norm": 0.06782060861587524, "learning_rate": 1.997369997857956e-05, "loss": 0.0044, "step": 71980 }, { "grad_norm": 0.08048918843269348, "learning_rate": 1.9960480389508103e-05, "loss": 0.0046, "step": 71990 }, { "grad_norm": 0.07448653131723404, "learning_rate": 1.9947264085513363e-05, "loss": 0.0049, "step": 72000 }, { "grad_norm": 0.0956600084900856, "learning_rate": 1.993405106804062e-05, "loss": 0.005, "step": 72010 }, { "grad_norm": 0.11294283717870712, "learning_rate": 1.9920841338534867e-05, "loss": 0.0051, "step": 72020 }, { "grad_norm": 0.08388303220272064, "learning_rate": 1.9907634898440646e-05, "loss": 0.004, "step": 72030 }, { "grad_norm": 0.22340263426303864, "learning_rate": 1.989443174920227e-05, "loss": 0.0035, "step": 72040 }, { "grad_norm": 0.1336401402950287, "learning_rate": 1.988123189226355e-05, "loss": 0.0041, "step": 72050 }, { "grad_norm": 0.08031731098890305, "learning_rate": 1.986803532906805e-05, "loss": 0.0047, "step": 72060 }, { "grad_norm": 0.08222374320030212, "learning_rate": 1.985484206105889e-05, "loss": 0.0042, "step": 72070 }, { "grad_norm": 0.10288432240486145, "learning_rate": 1.9841652089678887e-05, "loss": 0.0048, "step": 72080 }, { "grad_norm": 0.09690307825803757, "learning_rate": 1.9828465416370435e-05, "loss": 0.0039, "step": 72090 }, { "grad_norm": 0.09070012718439102, "learning_rate": 1.9815282042575685e-05, "loss": 0.004, "step": 72100 }, { "grad_norm": 0.0957537516951561, "learning_rate": 1.9802101969736285e-05, "loss": 0.0048, "step": 72110 }, { "grad_norm": 0.08385298401117325, "learning_rate": 1.9788925199293633e-05, "loss": 0.0059, "step": 72120 }, { "grad_norm": 0.09144825488328934, "learning_rate": 1.977575173268868e-05, "loss": 0.0041, "step": 72130 }, { "grad_norm": 0.09489544481039047, "learning_rate": 1.9762581571362094e-05, "loss": 0.004, "step": 72140 }, { "grad_norm": 0.09938894212245941, "learning_rate": 1.9749414716754112e-05, "loss": 0.0041, "step": 72150 }, { "grad_norm": 0.10957366228103638, "learning_rate": 1.973625117030465e-05, "loss": 0.0042, "step": 72160 }, { "grad_norm": 0.09148549288511276, "learning_rate": 1.972309093345326e-05, "loss": 0.004, "step": 72170 }, { "grad_norm": 0.09316886961460114, "learning_rate": 1.970993400763915e-05, "loss": 0.0035, "step": 72180 }, { "grad_norm": 0.07621559500694275, "learning_rate": 1.969678039430109e-05, "loss": 0.0037, "step": 72190 }, { "grad_norm": 0.11345816403627396, "learning_rate": 1.9683630094877564e-05, "loss": 0.0042, "step": 72200 }, { "grad_norm": 0.09746101498603821, "learning_rate": 1.9670483110806696e-05, "loss": 0.0055, "step": 72210 }, { "grad_norm": 0.07916540652513504, "learning_rate": 1.9657339443526172e-05, "loss": 0.0042, "step": 72220 }, { "grad_norm": 0.10048588365316391, "learning_rate": 1.964419909447339e-05, "loss": 0.0039, "step": 72230 }, { "grad_norm": 0.10368873924016953, "learning_rate": 1.963106206508536e-05, "loss": 0.0056, "step": 72240 }, { "grad_norm": 0.11094163358211517, "learning_rate": 1.9617928356798738e-05, "loss": 0.0063, "step": 72250 }, { "grad_norm": 0.10886197537183762, "learning_rate": 1.960479797104977e-05, "loss": 0.0046, "step": 72260 }, { "grad_norm": 0.06605766713619232, "learning_rate": 1.9591670909274418e-05, "loss": 0.0044, "step": 72270 }, { "grad_norm": 0.09335552155971527, "learning_rate": 1.9578547172908184e-05, "loss": 0.0041, "step": 72280 }, { "grad_norm": 0.09231524169445038, "learning_rate": 1.9565426763386325e-05, "loss": 0.0032, "step": 72290 }, { "grad_norm": 0.08926551043987274, "learning_rate": 1.9552309682143622e-05, "loss": 0.0039, "step": 72300 }, { "grad_norm": 0.09034629166126251, "learning_rate": 1.9539195930614574e-05, "loss": 0.0056, "step": 72310 }, { "grad_norm": 0.13008235394954681, "learning_rate": 1.952608551023324e-05, "loss": 0.0048, "step": 72320 }, { "grad_norm": 0.15419048070907593, "learning_rate": 1.9512978422433397e-05, "loss": 0.0059, "step": 72330 }, { "grad_norm": 0.14445823431015015, "learning_rate": 1.9499874668648376e-05, "loss": 0.0036, "step": 72340 }, { "grad_norm": 0.10965510457754135, "learning_rate": 1.9486774250311213e-05, "loss": 0.0042, "step": 72350 }, { "grad_norm": 0.08409935235977173, "learning_rate": 1.9473677168854537e-05, "loss": 0.0037, "step": 72360 }, { "grad_norm": 0.13964419066905975, "learning_rate": 1.946058342571065e-05, "loss": 0.0048, "step": 72370 }, { "grad_norm": 0.09391111135482788, "learning_rate": 1.9447493022311424e-05, "loss": 0.0053, "step": 72380 }, { "grad_norm": 0.12006262689828873, "learning_rate": 1.9434405960088442e-05, "loss": 0.0058, "step": 72390 }, { "grad_norm": 0.09570534527301788, "learning_rate": 1.9421322240472857e-05, "loss": 0.004, "step": 72400 }, { "grad_norm": 0.1027933806180954, "learning_rate": 1.9408241864895494e-05, "loss": 0.0041, "step": 72410 }, { "grad_norm": 0.1093338131904602, "learning_rate": 1.939516483478681e-05, "loss": 0.0045, "step": 72420 }, { "grad_norm": 0.13070183992385864, "learning_rate": 1.9382091151576902e-05, "loss": 0.0045, "step": 72430 }, { "grad_norm": 0.09550726413726807, "learning_rate": 1.9369020816695454e-05, "loss": 0.0041, "step": 72440 }, { "grad_norm": 0.05649649724364281, "learning_rate": 1.935595383157184e-05, "loss": 0.0038, "step": 72450 }, { "grad_norm": 0.1476614773273468, "learning_rate": 1.9342890197635057e-05, "loss": 0.0049, "step": 72460 }, { "grad_norm": 0.12556062638759613, "learning_rate": 1.9329829916313685e-05, "loss": 0.0042, "step": 72470 }, { "grad_norm": 0.09692128002643585, "learning_rate": 1.9316772989036013e-05, "loss": 0.0043, "step": 72480 }, { "grad_norm": 0.0888805091381073, "learning_rate": 1.930371941722991e-05, "loss": 0.0036, "step": 72490 }, { "grad_norm": 0.09193422645330429, "learning_rate": 1.929066920232292e-05, "loss": 0.0046, "step": 72500 }, { "grad_norm": 0.04759465530514717, "learning_rate": 1.9277622345742152e-05, "loss": 0.0033, "step": 72510 }, { "grad_norm": 0.11408901959657669, "learning_rate": 1.926457884891444e-05, "loss": 0.0058, "step": 72520 }, { "grad_norm": 0.09697134792804718, "learning_rate": 1.9251538713266147e-05, "loss": 0.0036, "step": 72530 }, { "grad_norm": 0.07436862587928772, "learning_rate": 1.9238501940223358e-05, "loss": 0.0043, "step": 72540 }, { "grad_norm": 0.10382243990898132, "learning_rate": 1.922546853121175e-05, "loss": 0.0042, "step": 72550 }, { "grad_norm": 0.08830488473176956, "learning_rate": 1.9212438487656647e-05, "loss": 0.0042, "step": 72560 }, { "grad_norm": 0.0999143049120903, "learning_rate": 1.9199411810982965e-05, "loss": 0.0034, "step": 72570 }, { "grad_norm": 0.07623124122619629, "learning_rate": 1.918638850261532e-05, "loss": 0.0037, "step": 72580 }, { "grad_norm": 0.08068294823169708, "learning_rate": 1.9173368563977885e-05, "loss": 0.0044, "step": 72590 }, { "grad_norm": 0.12549763917922974, "learning_rate": 1.9160351996494513e-05, "loss": 0.006, "step": 72600 }, { "grad_norm": 0.11814084649085999, "learning_rate": 1.9147338801588677e-05, "loss": 0.0047, "step": 72610 }, { "grad_norm": 0.07388264685869217, "learning_rate": 1.913432898068351e-05, "loss": 0.0039, "step": 72620 }, { "grad_norm": 0.08522019535303116, "learning_rate": 1.912132253520169e-05, "loss": 0.0041, "step": 72630 }, { "grad_norm": 0.14345301687717438, "learning_rate": 1.910831946656564e-05, "loss": 0.005, "step": 72640 }, { "grad_norm": 0.10671425610780716, "learning_rate": 1.909531977619731e-05, "loss": 0.004, "step": 72650 }, { "grad_norm": 0.13218790292739868, "learning_rate": 1.908232346551834e-05, "loss": 0.0046, "step": 72660 }, { "grad_norm": 0.09060925245285034, "learning_rate": 1.9069330535949993e-05, "loss": 0.0051, "step": 72670 }, { "grad_norm": 0.137075737118721, "learning_rate": 1.9056340988913167e-05, "loss": 0.0036, "step": 72680 }, { "grad_norm": 0.08564580231904984, "learning_rate": 1.9043354825828357e-05, "loss": 0.0049, "step": 72690 }, { "grad_norm": 0.07579869776964188, "learning_rate": 1.9030372048115724e-05, "loss": 0.0042, "step": 72700 }, { "grad_norm": 0.08580876141786575, "learning_rate": 1.9017392657195056e-05, "loss": 0.0035, "step": 72710 }, { "grad_norm": 0.0914488434791565, "learning_rate": 1.9004416654485724e-05, "loss": 0.0058, "step": 72720 }, { "grad_norm": 0.09128756821155548, "learning_rate": 1.8991444041406785e-05, "loss": 0.0037, "step": 72730 }, { "grad_norm": 0.06405311077833176, "learning_rate": 1.8978474819376912e-05, "loss": 0.0047, "step": 72740 }, { "grad_norm": 0.08098158985376358, "learning_rate": 1.8965508989814406e-05, "loss": 0.0053, "step": 72750 }, { "grad_norm": 0.052654191851615906, "learning_rate": 1.8952546554137158e-05, "loss": 0.0033, "step": 72760 }, { "grad_norm": 0.09565553814172745, "learning_rate": 1.8939587513762757e-05, "loss": 0.0042, "step": 72770 }, { "grad_norm": 0.09108462929725647, "learning_rate": 1.892663187010835e-05, "loss": 0.0048, "step": 72780 }, { "grad_norm": 0.09152951836585999, "learning_rate": 1.891367962459077e-05, "loss": 0.005, "step": 72790 }, { "grad_norm": 0.09253538399934769, "learning_rate": 1.890073077862644e-05, "loss": 0.0048, "step": 72800 }, { "grad_norm": 0.09670846909284592, "learning_rate": 1.888778533363146e-05, "loss": 0.0041, "step": 72810 }, { "grad_norm": 0.09503272920846939, "learning_rate": 1.8874843291021478e-05, "loss": 0.0041, "step": 72820 }, { "grad_norm": 0.08006210625171661, "learning_rate": 1.8861904652211853e-05, "loss": 0.004, "step": 72830 }, { "grad_norm": 0.07746962457895279, "learning_rate": 1.8848969418617507e-05, "loss": 0.0036, "step": 72840 }, { "grad_norm": 0.07601004838943481, "learning_rate": 1.8836037591653044e-05, "loss": 0.0036, "step": 72850 }, { "grad_norm": 0.09071499854326248, "learning_rate": 1.882310917273262e-05, "loss": 0.0046, "step": 72860 }, { "grad_norm": 0.10403266549110413, "learning_rate": 1.881018416327014e-05, "loss": 0.0046, "step": 72870 }, { "grad_norm": 0.06859157234430313, "learning_rate": 1.8797262564679014e-05, "loss": 0.0035, "step": 72880 }, { "grad_norm": 0.09201256185770035, "learning_rate": 1.8784344378372355e-05, "loss": 0.0046, "step": 72890 }, { "grad_norm": 0.07985049486160278, "learning_rate": 1.8771429605762843e-05, "loss": 0.0038, "step": 72900 }, { "grad_norm": 0.09586656093597412, "learning_rate": 1.8758518248262835e-05, "loss": 0.0037, "step": 72910 }, { "grad_norm": 0.11852745711803436, "learning_rate": 1.8745610307284305e-05, "loss": 0.0054, "step": 72920 }, { "grad_norm": 0.10413024574518204, "learning_rate": 1.8732705784238862e-05, "loss": 0.0046, "step": 72930 }, { "grad_norm": 0.09104154258966446, "learning_rate": 1.8719804680537682e-05, "loss": 0.0038, "step": 72940 }, { "grad_norm": 0.09436988830566406, "learning_rate": 1.870690699759164e-05, "loss": 0.0035, "step": 72950 }, { "grad_norm": 0.1088562086224556, "learning_rate": 1.8694012736811213e-05, "loss": 0.0053, "step": 72960 }, { "grad_norm": 0.08721918612718582, "learning_rate": 1.8681121899606463e-05, "loss": 0.0042, "step": 72970 }, { "grad_norm": 0.09314946830272675, "learning_rate": 1.8668234487387144e-05, "loss": 0.0046, "step": 72980 }, { "grad_norm": 0.09300267696380615, "learning_rate": 1.8655350501562592e-05, "loss": 0.0047, "step": 72990 }, { "grad_norm": 0.08557920902967453, "learning_rate": 1.8642469943541796e-05, "loss": 0.0036, "step": 73000 }, { "grad_norm": 0.061918605118989944, "learning_rate": 1.8629592814733328e-05, "loss": 0.0043, "step": 73010 }, { "grad_norm": 0.11992672085762024, "learning_rate": 1.8616719116545438e-05, "loss": 0.004, "step": 73020 }, { "grad_norm": 0.11012531071901321, "learning_rate": 1.8603848850385952e-05, "loss": 0.0038, "step": 73030 }, { "grad_norm": 0.0833653062582016, "learning_rate": 1.859098201766236e-05, "loss": 0.0053, "step": 73040 }, { "grad_norm": 0.08178528398275375, "learning_rate": 1.8578118619781726e-05, "loss": 0.0034, "step": 73050 }, { "grad_norm": 0.08444992452859879, "learning_rate": 1.856525865815083e-05, "loss": 0.0054, "step": 73060 }, { "grad_norm": 0.06580507755279541, "learning_rate": 1.8552402134175966e-05, "loss": 0.0045, "step": 73070 }, { "grad_norm": 0.1387774795293808, "learning_rate": 1.8539549049263143e-05, "loss": 0.0047, "step": 73080 }, { "grad_norm": 0.13817238807678223, "learning_rate": 1.8526699404817917e-05, "loss": 0.0048, "step": 73090 }, { "grad_norm": 0.11151276528835297, "learning_rate": 1.8513853202245546e-05, "loss": 0.0038, "step": 73100 }, { "grad_norm": 0.10652732104063034, "learning_rate": 1.850101044295081e-05, "loss": 0.0047, "step": 73110 }, { "grad_norm": 0.0683908686041832, "learning_rate": 1.848817112833825e-05, "loss": 0.0081, "step": 73120 }, { "grad_norm": 0.08849947154521942, "learning_rate": 1.8475335259811905e-05, "loss": 0.0041, "step": 73130 }, { "grad_norm": 0.10858073830604553, "learning_rate": 1.8462502838775508e-05, "loss": 0.005, "step": 73140 }, { "grad_norm": 0.08445944637060165, "learning_rate": 1.8449673866632378e-05, "loss": 0.0038, "step": 73150 }, { "grad_norm": 0.10092717409133911, "learning_rate": 1.8436848344785473e-05, "loss": 0.0037, "step": 73160 }, { "grad_norm": 0.12551656365394592, "learning_rate": 1.8424026274637378e-05, "loss": 0.006, "step": 73170 }, { "grad_norm": 0.07443474978208542, "learning_rate": 1.8411207657590312e-05, "loss": 0.0036, "step": 73180 }, { "grad_norm": 0.08621365576982498, "learning_rate": 1.839839249504607e-05, "loss": 0.0039, "step": 73190 }, { "grad_norm": 0.06416714936494827, "learning_rate": 1.8385580788406108e-05, "loss": 0.0048, "step": 73200 }, { "grad_norm": 0.1387181282043457, "learning_rate": 1.837277253907152e-05, "loss": 0.0043, "step": 73210 }, { "grad_norm": 0.11524401605129242, "learning_rate": 1.8359967748442957e-05, "loss": 0.0068, "step": 73220 }, { "grad_norm": 0.13595090806484222, "learning_rate": 1.8347166417920774e-05, "loss": 0.0047, "step": 73230 }, { "grad_norm": 0.08538541942834854, "learning_rate": 1.833436854890484e-05, "loss": 0.0036, "step": 73240 }, { "grad_norm": 0.06694868952035904, "learning_rate": 1.8321574142794796e-05, "loss": 0.0036, "step": 73250 }, { "grad_norm": 0.09993670135736465, "learning_rate": 1.8308783200989748e-05, "loss": 0.0033, "step": 73260 }, { "grad_norm": 0.09845127910375595, "learning_rate": 1.829599572488855e-05, "loss": 0.0048, "step": 73270 }, { "grad_norm": 0.09287488460540771, "learning_rate": 1.828321171588957e-05, "loss": 0.005, "step": 73280 }, { "grad_norm": 0.07541246712207794, "learning_rate": 1.827043117539089e-05, "loss": 0.0034, "step": 73290 }, { "grad_norm": 0.0716419667005539, "learning_rate": 1.8257654104790116e-05, "loss": 0.004, "step": 73300 }, { "grad_norm": 0.17630042135715485, "learning_rate": 1.82448805054846e-05, "loss": 0.004, "step": 73310 }, { "grad_norm": 0.10564440488815308, "learning_rate": 1.823211037887119e-05, "loss": 0.0043, "step": 73320 }, { "grad_norm": 0.1209222599864006, "learning_rate": 1.821934372634645e-05, "loss": 0.0038, "step": 73330 }, { "grad_norm": 0.09271222352981567, "learning_rate": 1.8206580549306472e-05, "loss": 0.0042, "step": 73340 }, { "grad_norm": 0.10694368183612823, "learning_rate": 1.819382084914707e-05, "loss": 0.0036, "step": 73350 }, { "grad_norm": 0.10631919652223587, "learning_rate": 1.8181064627263577e-05, "loss": 0.0041, "step": 73360 }, { "grad_norm": 0.07063278555870056, "learning_rate": 1.8168311885051022e-05, "loss": 0.0041, "step": 73370 }, { "grad_norm": 0.07483809441328049, "learning_rate": 1.815556262390402e-05, "loss": 0.0044, "step": 73380 }, { "grad_norm": 0.06880585849285126, "learning_rate": 1.8142816845216826e-05, "loss": 0.0042, "step": 73390 }, { "grad_norm": 0.09121301025152206, "learning_rate": 1.8130074550383276e-05, "loss": 0.0054, "step": 73400 }, { "grad_norm": 0.0870843306183815, "learning_rate": 1.8117335740796848e-05, "loss": 0.0041, "step": 73410 }, { "grad_norm": 0.11115220934152603, "learning_rate": 1.810460041785067e-05, "loss": 0.0036, "step": 73420 }, { "grad_norm": 0.10390132665634155, "learning_rate": 1.809186858293742e-05, "loss": 0.005, "step": 73430 }, { "grad_norm": 0.1305094212293625, "learning_rate": 1.8079140237449448e-05, "loss": 0.0046, "step": 73440 }, { "grad_norm": 0.07792217284440994, "learning_rate": 1.8066415382778707e-05, "loss": 0.0039, "step": 73450 }, { "grad_norm": 0.07641995698213577, "learning_rate": 1.8053694020316786e-05, "loss": 0.0052, "step": 73460 }, { "grad_norm": 0.08388577401638031, "learning_rate": 1.8040976151454837e-05, "loss": 0.0036, "step": 73470 }, { "grad_norm": 0.08165444433689117, "learning_rate": 1.8028261777583704e-05, "loss": 0.0048, "step": 73480 }, { "grad_norm": 0.0937076136469841, "learning_rate": 1.8015550900093758e-05, "loss": 0.0044, "step": 73490 }, { "grad_norm": 0.1663283407688141, "learning_rate": 1.800284352037512e-05, "loss": 0.0043, "step": 73500 }, { "grad_norm": 0.05790480971336365, "learning_rate": 1.799013963981738e-05, "loss": 0.005, "step": 73510 }, { "grad_norm": 0.08491842448711395, "learning_rate": 1.797743925980987e-05, "loss": 0.0047, "step": 73520 }, { "grad_norm": 0.09465336054563522, "learning_rate": 1.7964742381741435e-05, "loss": 0.0042, "step": 73530 }, { "grad_norm": 0.05607306957244873, "learning_rate": 1.7952049007000627e-05, "loss": 0.0031, "step": 73540 }, { "grad_norm": 0.08657871186733246, "learning_rate": 1.7939359136975535e-05, "loss": 0.0034, "step": 73550 }, { "grad_norm": 0.08558613061904907, "learning_rate": 1.7926672773053932e-05, "loss": 0.004, "step": 73560 }, { "grad_norm": 0.08190926164388657, "learning_rate": 1.791398991662317e-05, "loss": 0.0045, "step": 73570 }, { "grad_norm": 0.11365645378828049, "learning_rate": 1.7901310569070244e-05, "loss": 0.0042, "step": 73580 }, { "grad_norm": 0.11416956782341003, "learning_rate": 1.7888634731781724e-05, "loss": 0.0038, "step": 73590 }, { "grad_norm": 0.09568667411804199, "learning_rate": 1.7875962406143844e-05, "loss": 0.0067, "step": 73600 }, { "grad_norm": 0.10003498941659927, "learning_rate": 1.7863293593542403e-05, "loss": 0.0055, "step": 73610 }, { "grad_norm": 0.10357929021120071, "learning_rate": 1.7850628295362853e-05, "loss": 0.004, "step": 73620 }, { "grad_norm": 0.07390741258859634, "learning_rate": 1.7837966512990263e-05, "loss": 0.0053, "step": 73630 }, { "grad_norm": 0.08292216062545776, "learning_rate": 1.782530824780931e-05, "loss": 0.0055, "step": 73640 }, { "grad_norm": 0.0947984978556633, "learning_rate": 1.7812653501204263e-05, "loss": 0.0048, "step": 73650 }, { "grad_norm": 0.11907323449850082, "learning_rate": 1.780000227455903e-05, "loss": 0.0048, "step": 73660 }, { "grad_norm": 0.09411172568798065, "learning_rate": 1.778735456925716e-05, "loss": 0.0042, "step": 73670 }, { "grad_norm": 0.1256149262189865, "learning_rate": 1.7774710386681737e-05, "loss": 0.0054, "step": 73680 }, { "grad_norm": 0.09470720589160919, "learning_rate": 1.7762069728215545e-05, "loss": 0.0037, "step": 73690 }, { "grad_norm": 0.0834605023264885, "learning_rate": 1.7749432595240932e-05, "loss": 0.0041, "step": 73700 }, { "grad_norm": 0.10855352878570557, "learning_rate": 1.773679898913991e-05, "loss": 0.0043, "step": 73710 }, { "grad_norm": 0.08341524004936218, "learning_rate": 1.7724168911294015e-05, "loss": 0.0039, "step": 73720 }, { "grad_norm": 0.06625061482191086, "learning_rate": 1.7711542363084504e-05, "loss": 0.0043, "step": 73730 }, { "grad_norm": 0.12672159075737, "learning_rate": 1.7698919345892158e-05, "loss": 0.0052, "step": 73740 }, { "grad_norm": 0.08033467084169388, "learning_rate": 1.768629986109743e-05, "loss": 0.0041, "step": 73750 }, { "grad_norm": 0.06200210377573967, "learning_rate": 1.7673683910080357e-05, "loss": 0.0029, "step": 73760 }, { "grad_norm": 0.10646557062864304, "learning_rate": 1.7661071494220638e-05, "loss": 0.0043, "step": 73770 }, { "grad_norm": 0.11669143289327621, "learning_rate": 1.7648462614897498e-05, "loss": 0.0058, "step": 73780 }, { "grad_norm": 0.12892155349254608, "learning_rate": 1.7635857273489863e-05, "loss": 0.0046, "step": 73790 }, { "grad_norm": 0.09387756884098053, "learning_rate": 1.7623255471376198e-05, "loss": 0.0041, "step": 73800 }, { "grad_norm": 0.0870465561747551, "learning_rate": 1.7610657209934646e-05, "loss": 0.0034, "step": 73810 }, { "grad_norm": 0.0815238431096077, "learning_rate": 1.7598062490542915e-05, "loss": 0.0034, "step": 73820 }, { "grad_norm": 0.08503144979476929, "learning_rate": 1.7585471314578377e-05, "loss": 0.0049, "step": 73830 }, { "grad_norm": 0.11725099384784698, "learning_rate": 1.7572883683417945e-05, "loss": 0.0033, "step": 73840 }, { "grad_norm": 0.08313582092523575, "learning_rate": 1.7560299598438212e-05, "loss": 0.0034, "step": 73850 }, { "grad_norm": 0.08252666145563126, "learning_rate": 1.7547719061015334e-05, "loss": 0.0032, "step": 73860 }, { "grad_norm": 0.0643668994307518, "learning_rate": 1.75351420725251e-05, "loss": 0.0032, "step": 73870 }, { "grad_norm": 0.06600276380777359, "learning_rate": 1.7522568634342923e-05, "loss": 0.0038, "step": 73880 }, { "grad_norm": 0.059242572635412216, "learning_rate": 1.7509998747843825e-05, "loss": 0.0046, "step": 73890 }, { "grad_norm": 0.08190914243459702, "learning_rate": 1.7497432414402403e-05, "loss": 0.0044, "step": 73900 }, { "grad_norm": 0.08394084870815277, "learning_rate": 1.7484869635392903e-05, "loss": 0.0036, "step": 73910 }, { "grad_norm": 0.09587205946445465, "learning_rate": 1.747231041218919e-05, "loss": 0.004, "step": 73920 }, { "grad_norm": 0.05466922000050545, "learning_rate": 1.7459754746164692e-05, "loss": 0.0029, "step": 73930 }, { "grad_norm": 0.07399352639913559, "learning_rate": 1.744720263869249e-05, "loss": 0.0031, "step": 73940 }, { "grad_norm": 0.06270074844360352, "learning_rate": 1.7434654091145263e-05, "loss": 0.004, "step": 73950 }, { "grad_norm": 0.1047489270567894, "learning_rate": 1.742210910489532e-05, "loss": 0.0042, "step": 73960 }, { "grad_norm": 0.10634936392307281, "learning_rate": 1.7409567681314532e-05, "loss": 0.0049, "step": 73970 }, { "grad_norm": 0.06633172929286957, "learning_rate": 1.7397029821774434e-05, "loss": 0.0053, "step": 73980 }, { "grad_norm": 0.23987750709056854, "learning_rate": 1.7384495527646126e-05, "loss": 0.0042, "step": 73990 }, { "grad_norm": 0.07676448673009872, "learning_rate": 1.737196480030035e-05, "loss": 0.0048, "step": 74000 }, { "grad_norm": 0.20300541818141937, "learning_rate": 1.7359437641107447e-05, "loss": 0.0049, "step": 74010 }, { "grad_norm": 0.13990755379199982, "learning_rate": 1.7346914051437387e-05, "loss": 0.0067, "step": 74020 }, { "grad_norm": 0.08498521149158478, "learning_rate": 1.733439403265969e-05, "loss": 0.0037, "step": 74030 }, { "grad_norm": 0.05780738592147827, "learning_rate": 1.732187758614357e-05, "loss": 0.0038, "step": 74040 }, { "grad_norm": 0.07425753027200699, "learning_rate": 1.730936471325777e-05, "loss": 0.0061, "step": 74050 }, { "grad_norm": 0.07053973525762558, "learning_rate": 1.7296855415370718e-05, "loss": 0.0034, "step": 74060 }, { "grad_norm": 0.13778086006641388, "learning_rate": 1.728434969385035e-05, "loss": 0.0051, "step": 74070 }, { "grad_norm": 0.07632992416620255, "learning_rate": 1.7271847550064346e-05, "loss": 0.0035, "step": 74080 }, { "grad_norm": 0.08384261280298233, "learning_rate": 1.725934898537987e-05, "loss": 0.0031, "step": 74090 }, { "grad_norm": 0.07612665742635727, "learning_rate": 1.7246854001163783e-05, "loss": 0.005, "step": 74100 }, { "grad_norm": 0.0930727943778038, "learning_rate": 1.7234362598782476e-05, "loss": 0.0044, "step": 74110 }, { "grad_norm": 0.1021028608083725, "learning_rate": 1.722187477960202e-05, "loss": 0.004, "step": 74120 }, { "grad_norm": 0.10220087319612503, "learning_rate": 1.7209390544988057e-05, "loss": 0.0035, "step": 74130 }, { "grad_norm": 0.08241923898458481, "learning_rate": 1.7196909896305856e-05, "loss": 0.0035, "step": 74140 }, { "grad_norm": 0.12991085648536682, "learning_rate": 1.718443283492026e-05, "loss": 0.006, "step": 74150 }, { "grad_norm": 0.08509991317987442, "learning_rate": 1.7171959362195746e-05, "loss": 0.0043, "step": 74160 }, { "grad_norm": 0.0722183808684349, "learning_rate": 1.715948947949642e-05, "loss": 0.0039, "step": 74170 }, { "grad_norm": 0.06190512701869011, "learning_rate": 1.714702318818593e-05, "loss": 0.0038, "step": 74180 }, { "grad_norm": 0.10477323830127716, "learning_rate": 1.7134560489627592e-05, "loss": 0.0043, "step": 74190 }, { "grad_norm": 0.07363594323396683, "learning_rate": 1.712210138518431e-05, "loss": 0.0037, "step": 74200 }, { "grad_norm": 0.09585743397474289, "learning_rate": 1.71096458762186e-05, "loss": 0.0042, "step": 74210 }, { "grad_norm": 0.08742061257362366, "learning_rate": 1.7097193964092552e-05, "loss": 0.0036, "step": 74220 }, { "grad_norm": 0.10544972866773605, "learning_rate": 1.708474565016792e-05, "loss": 0.0066, "step": 74230 }, { "grad_norm": 0.0854722335934639, "learning_rate": 1.7072300935806e-05, "loss": 0.0041, "step": 74240 }, { "grad_norm": 0.07035553455352783, "learning_rate": 1.705985982236776e-05, "loss": 0.0042, "step": 74250 }, { "grad_norm": 0.06356660276651382, "learning_rate": 1.7047422311213697e-05, "loss": 0.0031, "step": 74260 }, { "grad_norm": 0.07157214730978012, "learning_rate": 1.703498840370401e-05, "loss": 0.0044, "step": 74270 }, { "grad_norm": 0.07786555588245392, "learning_rate": 1.702255810119842e-05, "loss": 0.0034, "step": 74280 }, { "grad_norm": 0.13639052212238312, "learning_rate": 1.7010131405056305e-05, "loss": 0.0038, "step": 74290 }, { "grad_norm": 0.08003468066453934, "learning_rate": 1.6997708316636613e-05, "loss": 0.0047, "step": 74300 }, { "grad_norm": 0.0770358219742775, "learning_rate": 1.6985288837297932e-05, "loss": 0.0038, "step": 74310 }, { "grad_norm": 0.09291994571685791, "learning_rate": 1.6972872968398394e-05, "loss": 0.0038, "step": 74320 }, { "grad_norm": 0.04576832801103592, "learning_rate": 1.6960460711295845e-05, "loss": 0.004, "step": 74330 }, { "grad_norm": 0.08369195461273193, "learning_rate": 1.6948052067347623e-05, "loss": 0.0037, "step": 74340 }, { "grad_norm": 0.15703226625919342, "learning_rate": 1.693564703791075e-05, "loss": 0.0049, "step": 74350 }, { "grad_norm": 0.12387488782405853, "learning_rate": 1.6923245624341793e-05, "loss": 0.0039, "step": 74360 }, { "grad_norm": 0.0960649847984314, "learning_rate": 1.6910847827996962e-05, "loss": 0.0035, "step": 74370 }, { "grad_norm": 0.0820353701710701, "learning_rate": 1.689845365023206e-05, "loss": 0.0035, "step": 74380 }, { "grad_norm": 0.1145479679107666, "learning_rate": 1.6886063092402505e-05, "loss": 0.0048, "step": 74390 }, { "grad_norm": 0.13919948041439056, "learning_rate": 1.6873676155863317e-05, "loss": 0.0035, "step": 74400 }, { "grad_norm": 0.10275726020336151, "learning_rate": 1.6861292841969085e-05, "loss": 0.0059, "step": 74410 }, { "grad_norm": 0.0939289927482605, "learning_rate": 1.684891315207406e-05, "loss": 0.0036, "step": 74420 }, { "grad_norm": 0.09836678951978683, "learning_rate": 1.6836537087532034e-05, "loss": 0.0043, "step": 74430 }, { "grad_norm": 0.1030036136507988, "learning_rate": 1.6824164649696465e-05, "loss": 0.0041, "step": 74440 }, { "grad_norm": 0.09214820712804794, "learning_rate": 1.681179583992034e-05, "loss": 0.0039, "step": 74450 }, { "grad_norm": 0.11504528671503067, "learning_rate": 1.679943065955636e-05, "loss": 0.0038, "step": 74460 }, { "grad_norm": 0.07486294955015182, "learning_rate": 1.678706910995671e-05, "loss": 0.0047, "step": 74470 }, { "grad_norm": 0.07760608196258545, "learning_rate": 1.6774711192473262e-05, "loss": 0.0039, "step": 74480 }, { "grad_norm": 0.10825468599796295, "learning_rate": 1.676235690845743e-05, "loss": 0.004, "step": 74490 }, { "grad_norm": 0.10304217040538788, "learning_rate": 1.6750006259260285e-05, "loss": 0.0039, "step": 74500 }, { "grad_norm": 0.08711985498666763, "learning_rate": 1.673765924623244e-05, "loss": 0.0039, "step": 74510 }, { "grad_norm": 0.10602077841758728, "learning_rate": 1.67253158707242e-05, "loss": 0.0057, "step": 74520 }, { "grad_norm": 0.09209448844194412, "learning_rate": 1.671297613408537e-05, "loss": 0.0048, "step": 74530 }, { "grad_norm": 0.26384103298187256, "learning_rate": 1.6700640037665444e-05, "loss": 0.0042, "step": 74540 }, { "grad_norm": 0.08577477186918259, "learning_rate": 1.668830758281344e-05, "loss": 0.0042, "step": 74550 }, { "grad_norm": 0.09583044052124023, "learning_rate": 1.6675978770878052e-05, "loss": 0.0045, "step": 74560 }, { "grad_norm": 0.09148003906011581, "learning_rate": 1.6663653603207506e-05, "loss": 0.0037, "step": 74570 }, { "grad_norm": 0.09657197445631027, "learning_rate": 1.6651332081149685e-05, "loss": 0.0039, "step": 74580 }, { "grad_norm": 0.13060759007930756, "learning_rate": 1.6639014206052046e-05, "loss": 0.0052, "step": 74590 }, { "grad_norm": 0.11673180013895035, "learning_rate": 1.6626699979261657e-05, "loss": 0.0049, "step": 74600 }, { "grad_norm": 0.10139971971511841, "learning_rate": 1.6614389402125197e-05, "loss": 0.0058, "step": 74610 }, { "grad_norm": 0.06904182583093643, "learning_rate": 1.66020824759889e-05, "loss": 0.003, "step": 74620 }, { "grad_norm": 0.10930177569389343, "learning_rate": 1.6589779202198675e-05, "loss": 0.0038, "step": 74630 }, { "grad_norm": 0.10570472478866577, "learning_rate": 1.6577479582099924e-05, "loss": 0.0031, "step": 74640 }, { "grad_norm": 0.08764465898275375, "learning_rate": 1.6565183617037788e-05, "loss": 0.0058, "step": 74650 }, { "grad_norm": 0.0737922340631485, "learning_rate": 1.6552891308356894e-05, "loss": 0.0049, "step": 74660 }, { "grad_norm": 0.12247269600629807, "learning_rate": 1.6540602657401526e-05, "loss": 0.0036, "step": 74670 }, { "grad_norm": 0.08406893163919449, "learning_rate": 1.6528317665515537e-05, "loss": 0.0048, "step": 74680 }, { "grad_norm": 0.1328982710838318, "learning_rate": 1.6516036334042416e-05, "loss": 0.0042, "step": 74690 }, { "grad_norm": 0.07454192638397217, "learning_rate": 1.650375866432519e-05, "loss": 0.0053, "step": 74700 }, { "grad_norm": 0.1059776172041893, "learning_rate": 1.649148465770659e-05, "loss": 0.0034, "step": 74710 }, { "grad_norm": 0.13035792112350464, "learning_rate": 1.647921431552883e-05, "loss": 0.0052, "step": 74720 }, { "grad_norm": 0.07823336124420166, "learning_rate": 1.646694763913381e-05, "loss": 0.0048, "step": 74730 }, { "grad_norm": 0.07856540381908417, "learning_rate": 1.6454684629862966e-05, "loss": 0.0047, "step": 74740 }, { "grad_norm": 0.10932469367980957, "learning_rate": 1.6442425289057388e-05, "loss": 0.0033, "step": 74750 }, { "grad_norm": 0.08539479225873947, "learning_rate": 1.6430169618057718e-05, "loss": 0.0037, "step": 74760 }, { "grad_norm": 0.11448782682418823, "learning_rate": 1.641791761820422e-05, "loss": 0.0031, "step": 74770 }, { "grad_norm": 0.08354150503873825, "learning_rate": 1.640566929083676e-05, "loss": 0.005, "step": 74780 }, { "grad_norm": 0.08525244891643524, "learning_rate": 1.6393424637294823e-05, "loss": 0.0031, "step": 74790 }, { "grad_norm": 0.1023387536406517, "learning_rate": 1.638118365891742e-05, "loss": 0.0041, "step": 74800 }, { "grad_norm": 0.08867467939853668, "learning_rate": 1.636894635704323e-05, "loss": 0.003, "step": 74810 }, { "grad_norm": 0.12435808777809143, "learning_rate": 1.6356712733010517e-05, "loss": 0.004, "step": 74820 }, { "grad_norm": 0.07441357523202896, "learning_rate": 1.6344482788157096e-05, "loss": 0.0046, "step": 74830 }, { "grad_norm": 0.1306850165128708, "learning_rate": 1.6332256523820444e-05, "loss": 0.0042, "step": 74840 }, { "grad_norm": 0.13701923191547394, "learning_rate": 1.6320033941337597e-05, "loss": 0.0035, "step": 74850 }, { "grad_norm": 0.12526945769786835, "learning_rate": 1.6307815042045216e-05, "loss": 0.0037, "step": 74860 }, { "grad_norm": 0.12254996597766876, "learning_rate": 1.6295599827279513e-05, "loss": 0.0031, "step": 74870 }, { "grad_norm": 0.13033942878246307, "learning_rate": 1.6283388298376356e-05, "loss": 0.0032, "step": 74880 }, { "grad_norm": 0.10933616012334824, "learning_rate": 1.6271180456671125e-05, "loss": 0.0044, "step": 74890 }, { "grad_norm": 0.06954660266637802, "learning_rate": 1.625897630349893e-05, "loss": 0.0046, "step": 74900 }, { "grad_norm": 0.11195121705532074, "learning_rate": 1.6246775840194335e-05, "loss": 0.0048, "step": 74910 }, { "grad_norm": 0.06291883438825607, "learning_rate": 1.6234579068091604e-05, "loss": 0.004, "step": 74920 }, { "grad_norm": 0.14981171488761902, "learning_rate": 1.6222385988524524e-05, "loss": 0.0045, "step": 74930 }, { "grad_norm": 0.10161922127008438, "learning_rate": 1.6210196602826544e-05, "loss": 0.0042, "step": 74940 }, { "grad_norm": 0.09911494702100754, "learning_rate": 1.6198010912330646e-05, "loss": 0.0035, "step": 74950 }, { "grad_norm": 0.0682196393609047, "learning_rate": 1.6185828918369456e-05, "loss": 0.0036, "step": 74960 }, { "grad_norm": 0.08187162131071091, "learning_rate": 1.617365062227517e-05, "loss": 0.003, "step": 74970 }, { "grad_norm": 0.0908324345946312, "learning_rate": 1.616147602537962e-05, "loss": 0.0043, "step": 74980 }, { "grad_norm": 0.0849454402923584, "learning_rate": 1.614930512901415e-05, "loss": 0.0037, "step": 74990 }, { "grad_norm": 0.07598445564508438, "learning_rate": 1.6137137934509804e-05, "loss": 0.0042, "step": 75000 }, { "grad_norm": 0.10887032002210617, "learning_rate": 1.6124974443197123e-05, "loss": 0.003, "step": 75010 }, { "grad_norm": 0.07087692618370056, "learning_rate": 1.6112814656406306e-05, "loss": 0.004, "step": 75020 }, { "grad_norm": 0.10354835540056229, "learning_rate": 1.6100658575467125e-05, "loss": 0.0041, "step": 75030 }, { "grad_norm": 0.07497715950012207, "learning_rate": 1.6088506201708974e-05, "loss": 0.0041, "step": 75040 }, { "grad_norm": 0.08052562177181244, "learning_rate": 1.6076357536460785e-05, "loss": 0.0035, "step": 75050 }, { "grad_norm": 0.06567918509244919, "learning_rate": 1.606421258105113e-05, "loss": 0.0039, "step": 75060 }, { "grad_norm": 0.10584881156682968, "learning_rate": 1.6052071336808183e-05, "loss": 0.0035, "step": 75070 }, { "grad_norm": 0.07072820514440536, "learning_rate": 1.6039933805059658e-05, "loss": 0.0042, "step": 75080 }, { "grad_norm": 0.08566533029079437, "learning_rate": 1.6027799987132915e-05, "loss": 0.0039, "step": 75090 }, { "grad_norm": 0.09387246519327164, "learning_rate": 1.6015669884354888e-05, "loss": 0.0052, "step": 75100 }, { "grad_norm": 0.10962404310703278, "learning_rate": 1.6003543498052125e-05, "loss": 0.0056, "step": 75110 }, { "grad_norm": 0.08870203047990799, "learning_rate": 1.5991420829550715e-05, "loss": 0.0035, "step": 75120 }, { "grad_norm": 0.07156961411237717, "learning_rate": 1.5979301880176405e-05, "loss": 0.0025, "step": 75130 }, { "grad_norm": 0.07201570272445679, "learning_rate": 1.5967186651254485e-05, "loss": 0.0042, "step": 75140 }, { "grad_norm": 0.09348928183317184, "learning_rate": 1.5955075144109853e-05, "loss": 0.0047, "step": 75150 }, { "grad_norm": 0.14849166572093964, "learning_rate": 1.594296736006703e-05, "loss": 0.0035, "step": 75160 }, { "grad_norm": 0.09716303646564484, "learning_rate": 1.59308633004501e-05, "loss": 0.0054, "step": 75170 }, { "grad_norm": 0.09480267018079758, "learning_rate": 1.5918762966582724e-05, "loss": 0.0034, "step": 75180 }, { "grad_norm": 0.09408286213874817, "learning_rate": 1.5906666359788203e-05, "loss": 0.0032, "step": 75190 }, { "grad_norm": 0.05458597093820572, "learning_rate": 1.589457348138938e-05, "loss": 0.0036, "step": 75200 }, { "grad_norm": 0.07550226151943207, "learning_rate": 1.588248433270872e-05, "loss": 0.0028, "step": 75210 }, { "grad_norm": 0.07749452441930771, "learning_rate": 1.587039891506828e-05, "loss": 0.004, "step": 75220 }, { "grad_norm": 0.07233327627182007, "learning_rate": 1.5858317229789715e-05, "loss": 0.0033, "step": 75230 }, { "grad_norm": 0.08258826285600662, "learning_rate": 1.584623927819423e-05, "loss": 0.004, "step": 75240 }, { "grad_norm": 0.09964066743850708, "learning_rate": 1.583416506160269e-05, "loss": 0.0041, "step": 75250 }, { "grad_norm": 0.08226446062326431, "learning_rate": 1.5822094581335472e-05, "loss": 0.0038, "step": 75260 }, { "grad_norm": 0.07471546530723572, "learning_rate": 1.5810027838712604e-05, "loss": 0.0037, "step": 75270 }, { "grad_norm": 0.11146917194128036, "learning_rate": 1.5797964835053687e-05, "loss": 0.0038, "step": 75280 }, { "grad_norm": 0.07509792596101761, "learning_rate": 1.5785905571677933e-05, "loss": 0.0037, "step": 75290 }, { "grad_norm": 0.12077835202217102, "learning_rate": 1.5773850049904086e-05, "loss": 0.0036, "step": 75300 }, { "grad_norm": 0.12417177855968475, "learning_rate": 1.5761798271050543e-05, "loss": 0.0048, "step": 75310 }, { "grad_norm": 0.08447650820016861, "learning_rate": 1.5749750236435277e-05, "loss": 0.0033, "step": 75320 }, { "grad_norm": 0.08292137831449509, "learning_rate": 1.5737705947375815e-05, "loss": 0.0039, "step": 75330 }, { "grad_norm": 0.07324644178152084, "learning_rate": 1.5725665405189322e-05, "loss": 0.0031, "step": 75340 }, { "grad_norm": 0.09534351527690887, "learning_rate": 1.571362861119253e-05, "loss": 0.0046, "step": 75350 }, { "grad_norm": 0.08545180410146713, "learning_rate": 1.570159556670177e-05, "loss": 0.003, "step": 75360 }, { "grad_norm": 0.06546241790056229, "learning_rate": 1.5689566273032946e-05, "loss": 0.004, "step": 75370 }, { "grad_norm": 0.08003544807434082, "learning_rate": 1.5677540731501583e-05, "loss": 0.0033, "step": 75380 }, { "grad_norm": 0.09967289119958878, "learning_rate": 1.5665518943422746e-05, "loss": 0.0035, "step": 75390 }, { "grad_norm": 0.05972093716263771, "learning_rate": 1.5653500910111134e-05, "loss": 0.0038, "step": 75400 }, { "grad_norm": 0.08742337673902512, "learning_rate": 1.5641486632881023e-05, "loss": 0.004, "step": 75410 }, { "grad_norm": 0.07940012961626053, "learning_rate": 1.56294761130463e-05, "loss": 0.0044, "step": 75420 }, { "grad_norm": 0.09697091579437256, "learning_rate": 1.5617469351920373e-05, "loss": 0.0038, "step": 75430 }, { "grad_norm": 0.09589193761348724, "learning_rate": 1.5605466350816315e-05, "loss": 0.0045, "step": 75440 }, { "grad_norm": 0.08639448881149292, "learning_rate": 1.559346711104674e-05, "loss": 0.0043, "step": 75450 }, { "grad_norm": 0.09016495198011398, "learning_rate": 1.5581471633923883e-05, "loss": 0.0036, "step": 75460 }, { "grad_norm": 0.10334931313991547, "learning_rate": 1.5569479920759505e-05, "loss": 0.0036, "step": 75470 }, { "grad_norm": 0.094644695520401, "learning_rate": 1.5557491972865072e-05, "loss": 0.0036, "step": 75480 }, { "grad_norm": 0.10436419397592545, "learning_rate": 1.554550779155152e-05, "loss": 0.0062, "step": 75490 }, { "grad_norm": 0.12595851719379425, "learning_rate": 1.5533527378129453e-05, "loss": 0.0034, "step": 75500 }, { "grad_norm": 0.07067041844129562, "learning_rate": 1.552155073390899e-05, "loss": 0.0048, "step": 75510 }, { "grad_norm": 0.06641734391450882, "learning_rate": 1.5509577860199903e-05, "loss": 0.0045, "step": 75520 }, { "grad_norm": 0.0817393884062767, "learning_rate": 1.5497608758311525e-05, "loss": 0.0045, "step": 75530 }, { "grad_norm": 0.09927919507026672, "learning_rate": 1.5485643429552797e-05, "loss": 0.0038, "step": 75540 }, { "grad_norm": 0.13080763816833496, "learning_rate": 1.5473681875232197e-05, "loss": 0.0034, "step": 75550 }, { "grad_norm": 0.07849564403295517, "learning_rate": 1.5461724096657837e-05, "loss": 0.0034, "step": 75560 }, { "grad_norm": 0.07932254672050476, "learning_rate": 1.5449770095137416e-05, "loss": 0.0051, "step": 75570 }, { "grad_norm": 0.12688304483890533, "learning_rate": 1.5437819871978178e-05, "loss": 0.0036, "step": 75580 }, { "grad_norm": 0.08937404304742813, "learning_rate": 1.5425873428486992e-05, "loss": 0.0034, "step": 75590 }, { "grad_norm": 0.0868377834558487, "learning_rate": 1.5413930765970312e-05, "loss": 0.0038, "step": 75600 }, { "grad_norm": 0.07438509911298752, "learning_rate": 1.5401991885734175e-05, "loss": 0.0049, "step": 75610 }, { "grad_norm": 0.10238975286483765, "learning_rate": 1.5390056789084174e-05, "loss": 0.0045, "step": 75620 }, { "grad_norm": 0.054363735020160675, "learning_rate": 1.5378125477325545e-05, "loss": 0.0051, "step": 75630 }, { "grad_norm": 0.06830035150051117, "learning_rate": 1.5366197951763046e-05, "loss": 0.0033, "step": 75640 }, { "grad_norm": 0.09968536347150803, "learning_rate": 1.5354274213701075e-05, "loss": 0.0043, "step": 75650 }, { "grad_norm": 0.06430695205926895, "learning_rate": 1.5342354264443566e-05, "loss": 0.0053, "step": 75660 }, { "grad_norm": 0.10638687759637833, "learning_rate": 1.533043810529411e-05, "loss": 0.0044, "step": 75670 }, { "grad_norm": 0.08320403099060059, "learning_rate": 1.5318525737555807e-05, "loss": 0.0045, "step": 75680 }, { "grad_norm": 0.12377570569515228, "learning_rate": 1.53066171625314e-05, "loss": 0.0037, "step": 75690 }, { "grad_norm": 0.07683790475130081, "learning_rate": 1.529471238152317e-05, "loss": 0.0032, "step": 75700 }, { "grad_norm": 0.09979744255542755, "learning_rate": 1.5282811395833023e-05, "loss": 0.0049, "step": 75710 }, { "grad_norm": 0.0899023786187172, "learning_rate": 1.5270914206762397e-05, "loss": 0.0031, "step": 75720 }, { "grad_norm": 0.07942233234643936, "learning_rate": 1.525902081561241e-05, "loss": 0.0041, "step": 75730 }, { "grad_norm": 0.08038000762462616, "learning_rate": 1.5247131223683664e-05, "loss": 0.0034, "step": 75740 }, { "grad_norm": 0.11664841324090958, "learning_rate": 1.5235245432276408e-05, "loss": 0.005, "step": 75750 }, { "grad_norm": 0.10105055570602417, "learning_rate": 1.5223363442690424e-05, "loss": 0.0042, "step": 75760 }, { "grad_norm": 0.08567377179861069, "learning_rate": 1.5211485256225133e-05, "loss": 0.004, "step": 75770 }, { "grad_norm": 0.11381026357412338, "learning_rate": 1.519961087417951e-05, "loss": 0.0034, "step": 75780 }, { "grad_norm": 0.06989770382642746, "learning_rate": 1.5187740297852137e-05, "loss": 0.004, "step": 75790 }, { "grad_norm": 0.11183518171310425, "learning_rate": 1.517587352854113e-05, "loss": 0.0038, "step": 75800 }, { "grad_norm": 0.09213664382696152, "learning_rate": 1.516401056754423e-05, "loss": 0.0038, "step": 75810 }, { "grad_norm": 0.08582372218370438, "learning_rate": 1.5152151416158783e-05, "loss": 0.0042, "step": 75820 }, { "grad_norm": 0.10278643667697906, "learning_rate": 1.5140296075681642e-05, "loss": 0.0031, "step": 75830 }, { "grad_norm": 0.10967820882797241, "learning_rate": 1.5128444547409331e-05, "loss": 0.005, "step": 75840 }, { "grad_norm": 0.06742650270462036, "learning_rate": 1.5116596832637864e-05, "loss": 0.0046, "step": 75850 }, { "grad_norm": 0.07971023768186569, "learning_rate": 1.5104752932662952e-05, "loss": 0.0034, "step": 75860 }, { "grad_norm": 0.06710043549537659, "learning_rate": 1.5092912848779778e-05, "loss": 0.0029, "step": 75870 }, { "grad_norm": 0.0666341781616211, "learning_rate": 1.5081076582283194e-05, "loss": 0.0031, "step": 75880 }, { "grad_norm": 0.0806630477309227, "learning_rate": 1.5069244134467553e-05, "loss": 0.0032, "step": 75890 }, { "grad_norm": 0.09724368900060654, "learning_rate": 1.5057415506626876e-05, "loss": 0.0038, "step": 75900 }, { "grad_norm": 0.12639573216438293, "learning_rate": 1.5045590700054668e-05, "loss": 0.0039, "step": 75910 }, { "grad_norm": 0.08370373398065567, "learning_rate": 1.5033769716044144e-05, "loss": 0.0037, "step": 75920 }, { "grad_norm": 0.07339468598365784, "learning_rate": 1.5021952555887975e-05, "loss": 0.0041, "step": 75930 }, { "grad_norm": 0.12147942930459976, "learning_rate": 1.50101392208785e-05, "loss": 0.0036, "step": 75940 }, { "grad_norm": 0.1295509934425354, "learning_rate": 1.4998329712307579e-05, "loss": 0.0043, "step": 75950 }, { "grad_norm": 0.14574140310287476, "learning_rate": 1.4986524031466708e-05, "loss": 0.0039, "step": 75960 }, { "grad_norm": 0.09650319069623947, "learning_rate": 1.4974722179646916e-05, "loss": 0.0039, "step": 75970 }, { "grad_norm": 0.05854326859116554, "learning_rate": 1.4962924158138847e-05, "loss": 0.0071, "step": 75980 }, { "grad_norm": 0.07920211553573608, "learning_rate": 1.4951129968232714e-05, "loss": 0.0053, "step": 75990 }, { "grad_norm": 0.07536260038614273, "learning_rate": 1.4939339611218329e-05, "loss": 0.0044, "step": 76000 }, { "grad_norm": 0.14687886834144592, "learning_rate": 1.4927553088385032e-05, "loss": 0.0033, "step": 76010 }, { "grad_norm": 0.09726665914058685, "learning_rate": 1.4915770401021801e-05, "loss": 0.005, "step": 76020 }, { "grad_norm": 0.08796721696853638, "learning_rate": 1.4903991550417185e-05, "loss": 0.0033, "step": 76030 }, { "grad_norm": 0.05299026891589165, "learning_rate": 1.4892216537859277e-05, "loss": 0.0041, "step": 76040 }, { "grad_norm": 0.06117803975939751, "learning_rate": 1.4880445364635775e-05, "loss": 0.0052, "step": 76050 }, { "grad_norm": 0.14899958670139313, "learning_rate": 1.4868678032033973e-05, "loss": 0.0045, "step": 76060 }, { "grad_norm": 0.0722431167960167, "learning_rate": 1.4856914541340739e-05, "loss": 0.0038, "step": 76070 }, { "grad_norm": 0.10244603455066681, "learning_rate": 1.4845154893842472e-05, "loss": 0.0039, "step": 76080 }, { "grad_norm": 0.08908258378505707, "learning_rate": 1.4833399090825234e-05, "loss": 0.007, "step": 76090 }, { "grad_norm": 0.07517910748720169, "learning_rate": 1.4821647133574562e-05, "loss": 0.0039, "step": 76100 }, { "grad_norm": 0.0814637616276741, "learning_rate": 1.4809899023375706e-05, "loss": 0.005, "step": 76110 }, { "grad_norm": 0.09690143167972565, "learning_rate": 1.4798154761513366e-05, "loss": 0.0038, "step": 76120 }, { "grad_norm": 0.11202854663133621, "learning_rate": 1.4786414349271916e-05, "loss": 0.0059, "step": 76130 }, { "grad_norm": 0.09286761283874512, "learning_rate": 1.4774677787935226e-05, "loss": 0.0038, "step": 76140 }, { "grad_norm": 0.08247334510087967, "learning_rate": 1.4762945078786838e-05, "loss": 0.004, "step": 76150 }, { "grad_norm": 0.0968092530965805, "learning_rate": 1.4751216223109777e-05, "loss": 0.0042, "step": 76160 }, { "grad_norm": 0.10214747488498688, "learning_rate": 1.4739491222186714e-05, "loss": 0.0033, "step": 76170 }, { "grad_norm": 0.1481703668832779, "learning_rate": 1.4727770077299879e-05, "loss": 0.0035, "step": 76180 }, { "grad_norm": 0.09488129615783691, "learning_rate": 1.4716052789731094e-05, "loss": 0.0054, "step": 76190 }, { "grad_norm": 0.08912067860364914, "learning_rate": 1.470433936076171e-05, "loss": 0.0044, "step": 76200 }, { "grad_norm": 0.06159006804227829, "learning_rate": 1.4692629791672719e-05, "loss": 0.0034, "step": 76210 }, { "grad_norm": 0.08525238186120987, "learning_rate": 1.468092408374464e-05, "loss": 0.0038, "step": 76220 }, { "grad_norm": 0.09388454258441925, "learning_rate": 1.4669222238257602e-05, "loss": 0.0035, "step": 76230 }, { "grad_norm": 0.06650884449481964, "learning_rate": 1.4657524256491301e-05, "loss": 0.0048, "step": 76240 }, { "grad_norm": 0.08411058038473129, "learning_rate": 1.464583013972502e-05, "loss": 0.0038, "step": 76250 }, { "grad_norm": 0.06557498872280121, "learning_rate": 1.4634139889237591e-05, "loss": 0.0033, "step": 76260 }, { "grad_norm": 0.14858968555927277, "learning_rate": 1.4622453506307448e-05, "loss": 0.0039, "step": 76270 }, { "grad_norm": 0.0943608209490776, "learning_rate": 1.461077099221262e-05, "loss": 0.0038, "step": 76280 }, { "grad_norm": 0.08395905047655106, "learning_rate": 1.4599092348230652e-05, "loss": 0.0052, "step": 76290 }, { "grad_norm": 0.06475745141506195, "learning_rate": 1.4587417575638724e-05, "loss": 0.003, "step": 76300 }, { "grad_norm": 0.07692896574735641, "learning_rate": 1.4575746675713563e-05, "loss": 0.0031, "step": 76310 }, { "grad_norm": 0.08992653340101242, "learning_rate": 1.4564079649731504e-05, "loss": 0.0033, "step": 76320 }, { "grad_norm": 0.15307478606700897, "learning_rate": 1.4552416498968402e-05, "loss": 0.0044, "step": 76330 }, { "grad_norm": 0.06760987639427185, "learning_rate": 1.4540757224699747e-05, "loss": 0.0032, "step": 76340 }, { "grad_norm": 0.09493957459926605, "learning_rate": 1.452910182820056e-05, "loss": 0.0042, "step": 76350 }, { "grad_norm": 0.08051423728466034, "learning_rate": 1.4517450310745468e-05, "loss": 0.0033, "step": 76360 }, { "grad_norm": 0.08290135860443115, "learning_rate": 1.450580267360866e-05, "loss": 0.0033, "step": 76370 }, { "grad_norm": 0.1154274120926857, "learning_rate": 1.4494158918063922e-05, "loss": 0.0038, "step": 76380 }, { "grad_norm": 0.08807558566331863, "learning_rate": 1.4482519045384568e-05, "loss": 0.0039, "step": 76390 }, { "grad_norm": 0.11068574339151382, "learning_rate": 1.4470883056843548e-05, "loss": 0.0054, "step": 76400 }, { "grad_norm": 0.0888887569308281, "learning_rate": 1.4459250953713322e-05, "loss": 0.0031, "step": 76410 }, { "grad_norm": 0.07482584565877914, "learning_rate": 1.444762273726598e-05, "loss": 0.0032, "step": 76420 }, { "grad_norm": 0.07352583110332489, "learning_rate": 1.443599840877316e-05, "loss": 0.0036, "step": 76430 }, { "grad_norm": 0.09574636816978455, "learning_rate": 1.4424377969506103e-05, "loss": 0.0039, "step": 76440 }, { "grad_norm": 0.10303032398223877, "learning_rate": 1.4412761420735565e-05, "loss": 0.004, "step": 76450 }, { "grad_norm": 0.1073305532336235, "learning_rate": 1.4401148763731953e-05, "loss": 0.0031, "step": 76460 }, { "grad_norm": 0.09632402658462524, "learning_rate": 1.4389539999765173e-05, "loss": 0.0032, "step": 76470 }, { "grad_norm": 0.06472611427307129, "learning_rate": 1.4377935130104758e-05, "loss": 0.0036, "step": 76480 }, { "grad_norm": 0.08297200500965118, "learning_rate": 1.4366334156019801e-05, "loss": 0.004, "step": 76490 }, { "grad_norm": 0.08865805715322495, "learning_rate": 1.435473707877898e-05, "loss": 0.0054, "step": 76500 }, { "grad_norm": 0.06791476905345917, "learning_rate": 1.4343143899650508e-05, "loss": 0.0058, "step": 76510 }, { "grad_norm": 0.16398632526397705, "learning_rate": 1.4331554619902204e-05, "loss": 0.0035, "step": 76520 }, { "grad_norm": 0.07683821767568588, "learning_rate": 1.4319969240801484e-05, "loss": 0.0056, "step": 76530 }, { "grad_norm": 0.06875310838222504, "learning_rate": 1.430838776361526e-05, "loss": 0.0038, "step": 76540 }, { "grad_norm": 0.0844353437423706, "learning_rate": 1.4296810189610093e-05, "loss": 0.0057, "step": 76550 }, { "grad_norm": 0.14038118720054626, "learning_rate": 1.4285236520052087e-05, "loss": 0.003, "step": 76560 }, { "grad_norm": 0.07157158851623535, "learning_rate": 1.4273666756206933e-05, "loss": 0.0035, "step": 76570 }, { "grad_norm": 0.08373922854661942, "learning_rate": 1.4262100899339848e-05, "loss": 0.0038, "step": 76580 }, { "grad_norm": 0.12646587193012238, "learning_rate": 1.4250538950715696e-05, "loss": 0.0039, "step": 76590 }, { "grad_norm": 0.08664733916521072, "learning_rate": 1.4238980911598842e-05, "loss": 0.0048, "step": 76600 }, { "grad_norm": 0.06418031454086304, "learning_rate": 1.4227426783253267e-05, "loss": 0.0033, "step": 76610 }, { "grad_norm": 0.10705012828111649, "learning_rate": 1.4215876566942521e-05, "loss": 0.0046, "step": 76620 }, { "grad_norm": 0.10255160182714462, "learning_rate": 1.4204330263929721e-05, "loss": 0.0036, "step": 76630 }, { "grad_norm": 0.09156080335378647, "learning_rate": 1.4192787875477537e-05, "loss": 0.004, "step": 76640 }, { "grad_norm": 0.07588981091976166, "learning_rate": 1.4181249402848246e-05, "loss": 0.0034, "step": 76650 }, { "grad_norm": 0.08592043071985245, "learning_rate": 1.4169714847303656e-05, "loss": 0.0035, "step": 76660 }, { "grad_norm": 0.09336446225643158, "learning_rate": 1.4158184210105197e-05, "loss": 0.0035, "step": 76670 }, { "grad_norm": 0.10606637597084045, "learning_rate": 1.4146657492513787e-05, "loss": 0.0042, "step": 76680 }, { "grad_norm": 0.08683699369430542, "learning_rate": 1.4135134695790047e-05, "loss": 0.003, "step": 76690 }, { "grad_norm": 0.08855399489402771, "learning_rate": 1.4123615821194037e-05, "loss": 0.0043, "step": 76700 }, { "grad_norm": 0.10365787893533707, "learning_rate": 1.4112100869985473e-05, "loss": 0.0043, "step": 76710 }, { "grad_norm": 0.10025659948587418, "learning_rate": 1.4100589843423589e-05, "loss": 0.0035, "step": 76720 }, { "grad_norm": 0.10468318313360214, "learning_rate": 1.408908274276723e-05, "loss": 0.0034, "step": 76730 }, { "grad_norm": 0.09599392116069794, "learning_rate": 1.4077579569274786e-05, "loss": 0.0047, "step": 76740 }, { "grad_norm": 0.06268290430307388, "learning_rate": 1.4066080324204245e-05, "loss": 0.0035, "step": 76750 }, { "grad_norm": 0.0896308645606041, "learning_rate": 1.405458500881312e-05, "loss": 0.0032, "step": 76760 }, { "grad_norm": 0.07976260781288147, "learning_rate": 1.4043093624358538e-05, "loss": 0.0035, "step": 76770 }, { "grad_norm": 0.08914472907781601, "learning_rate": 1.4031606172097189e-05, "loss": 0.0035, "step": 76780 }, { "grad_norm": 0.09613005071878433, "learning_rate": 1.40201226532853e-05, "loss": 0.0044, "step": 76790 }, { "grad_norm": 0.08232519030570984, "learning_rate": 1.4008643069178701e-05, "loss": 0.0036, "step": 76800 }, { "grad_norm": 0.07397227734327316, "learning_rate": 1.399716742103278e-05, "loss": 0.0036, "step": 76810 }, { "grad_norm": 0.09387676417827606, "learning_rate": 1.398569571010252e-05, "loss": 0.0037, "step": 76820 }, { "grad_norm": 0.05096126347780228, "learning_rate": 1.3974227937642414e-05, "loss": 0.0032, "step": 76830 }, { "grad_norm": 0.09411301463842392, "learning_rate": 1.3962764104906595e-05, "loss": 0.0035, "step": 76840 }, { "grad_norm": 0.09602412581443787, "learning_rate": 1.3951304213148696e-05, "loss": 0.0033, "step": 76850 }, { "grad_norm": 0.08772836625576019, "learning_rate": 1.3939848263621979e-05, "loss": 0.0037, "step": 76860 }, { "grad_norm": 0.08380322903394699, "learning_rate": 1.3928396257579219e-05, "loss": 0.0032, "step": 76870 }, { "grad_norm": 0.09089363366365433, "learning_rate": 1.3916948196272833e-05, "loss": 0.0031, "step": 76880 }, { "grad_norm": 0.11511080712080002, "learning_rate": 1.3905504080954734e-05, "loss": 0.0035, "step": 76890 }, { "grad_norm": 0.06141526252031326, "learning_rate": 1.3894063912876454e-05, "loss": 0.0037, "step": 76900 }, { "grad_norm": 0.1085771694779396, "learning_rate": 1.3882627693289046e-05, "loss": 0.0046, "step": 76910 }, { "grad_norm": 0.06437285989522934, "learning_rate": 1.3871195423443183e-05, "loss": 0.0039, "step": 76920 }, { "grad_norm": 0.06155576556921005, "learning_rate": 1.3859767104589044e-05, "loss": 0.0037, "step": 76930 }, { "grad_norm": 0.06230445206165314, "learning_rate": 1.3848342737976467e-05, "loss": 0.0031, "step": 76940 }, { "grad_norm": 0.08467863500118256, "learning_rate": 1.3836922324854757e-05, "loss": 0.0041, "step": 76950 }, { "grad_norm": 0.10598880797624588, "learning_rate": 1.382550586647287e-05, "loss": 0.0034, "step": 76960 }, { "grad_norm": 0.22134985029697418, "learning_rate": 1.3814093364079257e-05, "loss": 0.0041, "step": 76970 }, { "grad_norm": 0.09272579848766327, "learning_rate": 1.3802684818921991e-05, "loss": 0.0047, "step": 76980 }, { "grad_norm": 0.07451020926237106, "learning_rate": 1.3791280232248687e-05, "loss": 0.0044, "step": 76990 }, { "grad_norm": 0.10952143371105194, "learning_rate": 1.3779879605306555e-05, "loss": 0.0044, "step": 77000 }, { "grad_norm": 0.08865270018577576, "learning_rate": 1.376848293934232e-05, "loss": 0.0038, "step": 77010 }, { "grad_norm": 0.08130916208028793, "learning_rate": 1.3757090235602316e-05, "loss": 0.0041, "step": 77020 }, { "grad_norm": 0.10090994834899902, "learning_rate": 1.3745701495332447e-05, "loss": 0.0036, "step": 77030 }, { "grad_norm": 0.07848317921161652, "learning_rate": 1.3734316719778145e-05, "loss": 0.0065, "step": 77040 }, { "grad_norm": 0.07079247385263443, "learning_rate": 1.3722935910184454e-05, "loss": 0.0029, "step": 77050 }, { "grad_norm": 0.16046380996704102, "learning_rate": 1.3711559067795921e-05, "loss": 0.0045, "step": 77060 }, { "grad_norm": 0.09461274743080139, "learning_rate": 1.3700186193856756e-05, "loss": 0.0047, "step": 77070 }, { "grad_norm": 0.1158602237701416, "learning_rate": 1.368881728961064e-05, "loss": 0.0044, "step": 77080 }, { "grad_norm": 0.0733305811882019, "learning_rate": 1.3677452356300885e-05, "loss": 0.0044, "step": 77090 }, { "grad_norm": 0.08005056530237198, "learning_rate": 1.3666091395170321e-05, "loss": 0.0034, "step": 77100 }, { "grad_norm": 0.07862215489149094, "learning_rate": 1.3654734407461384e-05, "loss": 0.0031, "step": 77110 }, { "grad_norm": 0.07460544258356094, "learning_rate": 1.3643381394416022e-05, "loss": 0.0048, "step": 77120 }, { "grad_norm": 0.06735408306121826, "learning_rate": 1.3632032357275838e-05, "loss": 0.004, "step": 77130 }, { "grad_norm": 0.12168878316879272, "learning_rate": 1.3620687297281898e-05, "loss": 0.0049, "step": 77140 }, { "grad_norm": 0.0983135998249054, "learning_rate": 1.3609346215674917e-05, "loss": 0.0031, "step": 77150 }, { "grad_norm": 0.17521825432777405, "learning_rate": 1.3598009113695109e-05, "loss": 0.0039, "step": 77160 }, { "grad_norm": 0.0969722718000412, "learning_rate": 1.358667599258231e-05, "loss": 0.0037, "step": 77170 }, { "grad_norm": 0.24646227061748505, "learning_rate": 1.357534685357586e-05, "loss": 0.0031, "step": 77180 }, { "grad_norm": 0.13628309965133667, "learning_rate": 1.3564021697914715e-05, "loss": 0.0039, "step": 77190 }, { "grad_norm": 0.09251923859119415, "learning_rate": 1.3552700526837381e-05, "loss": 0.0038, "step": 77200 }, { "grad_norm": 0.08642642945051193, "learning_rate": 1.3541383341581925e-05, "loss": 0.0046, "step": 77210 }, { "grad_norm": 0.08908838778734207, "learning_rate": 1.3530070143385965e-05, "loss": 0.0039, "step": 77220 }, { "grad_norm": 0.1121513620018959, "learning_rate": 1.3518760933486702e-05, "loss": 0.0061, "step": 77230 }, { "grad_norm": 0.09312468767166138, "learning_rate": 1.3507455713120908e-05, "loss": 0.0037, "step": 77240 }, { "grad_norm": 0.09160041809082031, "learning_rate": 1.3496154483524876e-05, "loss": 0.0034, "step": 77250 }, { "grad_norm": 0.10119910538196564, "learning_rate": 1.3484857245934506e-05, "loss": 0.003, "step": 77260 }, { "grad_norm": 0.07964535802602768, "learning_rate": 1.3473564001585248e-05, "loss": 0.0055, "step": 77270 }, { "grad_norm": 0.08412052690982819, "learning_rate": 1.3462274751712129e-05, "loss": 0.0041, "step": 77280 }, { "grad_norm": 0.11902894824743271, "learning_rate": 1.3450989497549698e-05, "loss": 0.0038, "step": 77290 }, { "grad_norm": 0.11645087599754333, "learning_rate": 1.3439708240332115e-05, "loss": 0.0034, "step": 77300 }, { "grad_norm": 0.06202711910009384, "learning_rate": 1.3428430981293043e-05, "loss": 0.0047, "step": 77310 }, { "grad_norm": 0.05284819006919861, "learning_rate": 1.3417157721665808e-05, "loss": 0.003, "step": 77320 }, { "grad_norm": 0.19214177131652832, "learning_rate": 1.3405888462683186e-05, "loss": 0.0036, "step": 77330 }, { "grad_norm": 0.06913647800683975, "learning_rate": 1.3394623205577595e-05, "loss": 0.0033, "step": 77340 }, { "grad_norm": 0.08507570624351501, "learning_rate": 1.3383361951580963e-05, "loss": 0.0031, "step": 77350 }, { "grad_norm": 0.0774884894490242, "learning_rate": 1.337210470192483e-05, "loss": 0.004, "step": 77360 }, { "grad_norm": 0.06815006583929062, "learning_rate": 1.3360851457840245e-05, "loss": 0.0039, "step": 77370 }, { "grad_norm": 0.10673201084136963, "learning_rate": 1.3349602220557856e-05, "loss": 0.0036, "step": 77380 }, { "grad_norm": 0.1026737242937088, "learning_rate": 1.3338356991307865e-05, "loss": 0.0047, "step": 77390 }, { "grad_norm": 0.07205124944448471, "learning_rate": 1.3327115771320048e-05, "loss": 0.003, "step": 77400 }, { "grad_norm": 0.06430581212043762, "learning_rate": 1.3315878561823697e-05, "loss": 0.0046, "step": 77410 }, { "grad_norm": 0.06711568683385849, "learning_rate": 1.330464536404773e-05, "loss": 0.0038, "step": 77420 }, { "grad_norm": 0.08306621015071869, "learning_rate": 1.3293416179220563e-05, "loss": 0.004, "step": 77430 }, { "grad_norm": 0.10038697719573975, "learning_rate": 1.328219100857021e-05, "loss": 0.0037, "step": 77440 }, { "grad_norm": 0.1529398262500763, "learning_rate": 1.327096985332425e-05, "loss": 0.0036, "step": 77450 }, { "grad_norm": 0.08811940997838974, "learning_rate": 1.3259752714709816e-05, "loss": 0.0036, "step": 77460 }, { "grad_norm": 0.07417918741703033, "learning_rate": 1.3248539593953569e-05, "loss": 0.0032, "step": 77470 }, { "grad_norm": 0.11810282617807388, "learning_rate": 1.323733049228178e-05, "loss": 0.0034, "step": 77480 }, { "grad_norm": 0.07364539802074432, "learning_rate": 1.3226125410920276e-05, "loss": 0.0035, "step": 77490 }, { "grad_norm": 0.08300763368606567, "learning_rate": 1.3214924351094388e-05, "loss": 0.004, "step": 77500 }, { "grad_norm": 0.07786466926336288, "learning_rate": 1.3203727314029069e-05, "loss": 0.0032, "step": 77510 }, { "grad_norm": 0.1300908923149109, "learning_rate": 1.3192534300948811e-05, "loss": 0.004, "step": 77520 }, { "grad_norm": 0.060262806713581085, "learning_rate": 1.3181345313077675e-05, "loss": 0.0037, "step": 77530 }, { "grad_norm": 0.09059752523899078, "learning_rate": 1.3170160351639243e-05, "loss": 0.004, "step": 77540 }, { "grad_norm": 0.10550235956907272, "learning_rate": 1.3158979417856721e-05, "loss": 0.0047, "step": 77550 }, { "grad_norm": 0.17204122245311737, "learning_rate": 1.31478025129528e-05, "loss": 0.0038, "step": 77560 }, { "grad_norm": 0.1302105188369751, "learning_rate": 1.3136629638149794e-05, "loss": 0.0044, "step": 77570 }, { "grad_norm": 0.07575961202383041, "learning_rate": 1.3125460794669547e-05, "loss": 0.0042, "step": 77580 }, { "grad_norm": 0.07213679701089859, "learning_rate": 1.3114295983733483e-05, "loss": 0.003, "step": 77590 }, { "grad_norm": 0.10817721486091614, "learning_rate": 1.3103135206562534e-05, "loss": 0.0044, "step": 77600 }, { "grad_norm": 0.06813829392194748, "learning_rate": 1.3091978464377264e-05, "loss": 0.0033, "step": 77610 }, { "grad_norm": 0.09468355029821396, "learning_rate": 1.3080825758397724e-05, "loss": 0.0029, "step": 77620 }, { "grad_norm": 0.08023549616336823, "learning_rate": 1.3069677089843568e-05, "loss": 0.0031, "step": 77630 }, { "grad_norm": 0.05852340906858444, "learning_rate": 1.305853245993401e-05, "loss": 0.0035, "step": 77640 }, { "grad_norm": 0.05317038297653198, "learning_rate": 1.3047391869887809e-05, "loss": 0.0037, "step": 77650 }, { "grad_norm": 0.19882594048976898, "learning_rate": 1.3036255320923263e-05, "loss": 0.0054, "step": 77660 }, { "grad_norm": 0.07701679319143295, "learning_rate": 1.3025122814258273e-05, "loss": 0.0037, "step": 77670 }, { "grad_norm": 0.09755776822566986, "learning_rate": 1.3013994351110249e-05, "loss": 0.0034, "step": 77680 }, { "grad_norm": 0.10497850924730301, "learning_rate": 1.3002869932696194e-05, "loss": 0.0042, "step": 77690 }, { "grad_norm": 0.09607261419296265, "learning_rate": 1.2991749560232664e-05, "loss": 0.0034, "step": 77700 }, { "grad_norm": 0.058873988687992096, "learning_rate": 1.2980633234935768e-05, "loss": 0.0038, "step": 77710 }, { "grad_norm": 0.10953681170940399, "learning_rate": 1.2969520958021153e-05, "loss": 0.0035, "step": 77720 }, { "grad_norm": 0.12571105360984802, "learning_rate": 1.295841273070405e-05, "loss": 0.0034, "step": 77730 }, { "grad_norm": 0.10894237458705902, "learning_rate": 1.2947308554199255e-05, "loss": 0.0028, "step": 77740 }, { "grad_norm": 0.08843690901994705, "learning_rate": 1.2936208429721075e-05, "loss": 0.0036, "step": 77750 }, { "grad_norm": 0.07168927788734436, "learning_rate": 1.2925112358483416e-05, "loss": 0.0035, "step": 77760 }, { "grad_norm": 0.07651229947805405, "learning_rate": 1.2914020341699728e-05, "loss": 0.0045, "step": 77770 }, { "grad_norm": 0.09188258647918701, "learning_rate": 1.2902932380583032e-05, "loss": 0.0032, "step": 77780 }, { "grad_norm": 0.06731946021318436, "learning_rate": 1.2891848476345863e-05, "loss": 0.0053, "step": 77790 }, { "grad_norm": 0.05377225577831268, "learning_rate": 1.2880768630200368e-05, "loss": 0.0037, "step": 77800 }, { "grad_norm": 0.08090825378894806, "learning_rate": 1.286969284335819e-05, "loss": 0.0039, "step": 77810 }, { "grad_norm": 0.07245393097400665, "learning_rate": 1.2858621117030578e-05, "loss": 0.0031, "step": 77820 }, { "grad_norm": 0.11064308881759644, "learning_rate": 1.2847553452428324e-05, "loss": 0.0035, "step": 77830 }, { "grad_norm": 0.1228935644030571, "learning_rate": 1.283648985076178e-05, "loss": 0.004, "step": 77840 }, { "grad_norm": 0.0926259309053421, "learning_rate": 1.2825430313240817e-05, "loss": 0.0036, "step": 77850 }, { "grad_norm": 0.0948309600353241, "learning_rate": 1.281437484107491e-05, "loss": 0.0036, "step": 77860 }, { "grad_norm": 0.09647192060947418, "learning_rate": 1.2803323435473058e-05, "loss": 0.0033, "step": 77870 }, { "grad_norm": 0.11539565771818161, "learning_rate": 1.2792276097643841e-05, "loss": 0.0034, "step": 77880 }, { "grad_norm": 0.09979253262281418, "learning_rate": 1.2781232828795342e-05, "loss": 0.0034, "step": 77890 }, { "grad_norm": 0.07070386409759521, "learning_rate": 1.2770193630135292e-05, "loss": 0.0026, "step": 77900 }, { "grad_norm": 0.09403391927480698, "learning_rate": 1.2759158502870877e-05, "loss": 0.0038, "step": 77910 }, { "grad_norm": 0.04839037358760834, "learning_rate": 1.2748127448208913e-05, "loss": 0.0039, "step": 77920 }, { "grad_norm": 0.05836589261889458, "learning_rate": 1.2737100467355706e-05, "loss": 0.0038, "step": 77930 }, { "grad_norm": 0.07056877017021179, "learning_rate": 1.2726077561517169e-05, "loss": 0.0044, "step": 77940 }, { "grad_norm": 0.08617164939641953, "learning_rate": 1.2715058731898755e-05, "loss": 0.0037, "step": 77950 }, { "grad_norm": 0.11393999308347702, "learning_rate": 1.2704043979705471e-05, "loss": 0.0045, "step": 77960 }, { "grad_norm": 0.1411018967628479, "learning_rate": 1.269303330614185e-05, "loss": 0.0042, "step": 77970 }, { "grad_norm": 0.056801553815603256, "learning_rate": 1.2682026712412015e-05, "loss": 0.005, "step": 77980 }, { "grad_norm": 0.11905582249164581, "learning_rate": 1.2671024199719644e-05, "loss": 0.0038, "step": 77990 }, { "grad_norm": 0.062694251537323, "learning_rate": 1.2660025769267936e-05, "loss": 0.0035, "step": 78000 }, { "grad_norm": 0.10106100887060165, "learning_rate": 1.2649031422259667e-05, "loss": 0.0068, "step": 78010 }, { "grad_norm": 0.08983711898326874, "learning_rate": 1.2638041159897163e-05, "loss": 0.0038, "step": 78020 }, { "grad_norm": 0.08873433619737625, "learning_rate": 1.262705498338232e-05, "loss": 0.0034, "step": 78030 }, { "grad_norm": 0.05945565551519394, "learning_rate": 1.2616072893916542e-05, "loss": 0.0029, "step": 78040 }, { "grad_norm": 0.08272312581539154, "learning_rate": 1.2605094892700836e-05, "loss": 0.003, "step": 78050 }, { "grad_norm": 0.11050590127706528, "learning_rate": 1.259412098093572e-05, "loss": 0.0036, "step": 78060 }, { "grad_norm": 0.10610956698656082, "learning_rate": 1.258315115982131e-05, "loss": 0.0048, "step": 78070 }, { "grad_norm": 0.08359836786985397, "learning_rate": 1.2572185430557203e-05, "loss": 0.0028, "step": 78080 }, { "grad_norm": 0.056150633841753006, "learning_rate": 1.2561223794342659e-05, "loss": 0.0023, "step": 78090 }, { "grad_norm": 0.07132013142108917, "learning_rate": 1.255026625237637e-05, "loss": 0.0033, "step": 78100 }, { "grad_norm": 0.10839643329381943, "learning_rate": 1.2539312805856678e-05, "loss": 0.0038, "step": 78110 }, { "grad_norm": 0.09789098799228668, "learning_rate": 1.25283634559814e-05, "loss": 0.0041, "step": 78120 }, { "grad_norm": 0.09611628204584122, "learning_rate": 1.2517418203947972e-05, "loss": 0.0046, "step": 78130 }, { "grad_norm": 0.05580734461545944, "learning_rate": 1.2506477050953308e-05, "loss": 0.0051, "step": 78140 }, { "grad_norm": 0.10960230231285095, "learning_rate": 1.2495539998193972e-05, "loss": 0.0041, "step": 78150 }, { "grad_norm": 0.0659448578953743, "learning_rate": 1.2484607046865975e-05, "loss": 0.0051, "step": 78160 }, { "grad_norm": 0.06672448664903641, "learning_rate": 1.2473678198164968e-05, "loss": 0.004, "step": 78170 }, { "grad_norm": 0.06049671396613121, "learning_rate": 1.2462753453286075e-05, "loss": 0.0042, "step": 78180 }, { "grad_norm": 0.08107762038707733, "learning_rate": 1.2451832813424031e-05, "loss": 0.0035, "step": 78190 }, { "grad_norm": 0.07161375135183334, "learning_rate": 1.2440916279773095e-05, "loss": 0.005, "step": 78200 }, { "grad_norm": 0.07272853702306747, "learning_rate": 1.2430003853527101e-05, "loss": 0.0044, "step": 78210 }, { "grad_norm": 0.09253428131341934, "learning_rate": 1.2419095535879383e-05, "loss": 0.0034, "step": 78220 }, { "grad_norm": 0.1046014353632927, "learning_rate": 1.2408191328022878e-05, "loss": 0.0044, "step": 78230 }, { "grad_norm": 0.04987874627113342, "learning_rate": 1.2397291231150066e-05, "loss": 0.0061, "step": 78240 }, { "grad_norm": 0.05081808194518089, "learning_rate": 1.2386395246452937e-05, "loss": 0.0042, "step": 78250 }, { "grad_norm": 0.06744956970214844, "learning_rate": 1.2375503375123083e-05, "loss": 0.0032, "step": 78260 }, { "grad_norm": 0.078075110912323, "learning_rate": 1.2364615618351583e-05, "loss": 0.004, "step": 78270 }, { "grad_norm": 0.0855991467833519, "learning_rate": 1.2353731977329169e-05, "loss": 0.0042, "step": 78280 }, { "grad_norm": 0.09888766705989838, "learning_rate": 1.2342852453246007e-05, "loss": 0.0037, "step": 78290 }, { "grad_norm": 0.07780317217111588, "learning_rate": 1.2331977047291898e-05, "loss": 0.003, "step": 78300 }, { "grad_norm": 0.13981083035469055, "learning_rate": 1.2321105760656137e-05, "loss": 0.0037, "step": 78310 }, { "grad_norm": 0.09457353502511978, "learning_rate": 1.2310238594527613e-05, "loss": 0.0047, "step": 78320 }, { "grad_norm": 0.07808183133602142, "learning_rate": 1.2299375550094693e-05, "loss": 0.0039, "step": 78330 }, { "grad_norm": 0.0980026125907898, "learning_rate": 1.2288516628545421e-05, "loss": 0.0027, "step": 78340 }, { "grad_norm": 0.09174147248268127, "learning_rate": 1.2277661831067255e-05, "loss": 0.004, "step": 78350 }, { "grad_norm": 0.08894490450620651, "learning_rate": 1.2266811158847286e-05, "loss": 0.0045, "step": 78360 }, { "grad_norm": 0.11839884519577026, "learning_rate": 1.2255964613072107e-05, "loss": 0.0045, "step": 78370 }, { "grad_norm": 0.05871120095252991, "learning_rate": 1.2245122194927905e-05, "loss": 0.0028, "step": 78380 }, { "grad_norm": 0.07556015253067017, "learning_rate": 1.2234283905600364e-05, "loss": 0.0032, "step": 78390 }, { "grad_norm": 0.09749800711870193, "learning_rate": 1.222344974627475e-05, "loss": 0.0046, "step": 78400 }, { "grad_norm": 0.1387530267238617, "learning_rate": 1.2212619718135875e-05, "loss": 0.0031, "step": 78410 }, { "grad_norm": 0.09424035251140594, "learning_rate": 1.2201793822368118e-05, "loss": 0.0038, "step": 78420 }, { "grad_norm": 0.10972350090742111, "learning_rate": 1.2190972060155337e-05, "loss": 0.003, "step": 78430 }, { "grad_norm": 0.06996550410985947, "learning_rate": 1.2180154432681013e-05, "loss": 0.0043, "step": 78440 }, { "grad_norm": 0.12470600008964539, "learning_rate": 1.2169340941128148e-05, "loss": 0.0048, "step": 78450 }, { "grad_norm": 0.04788042977452278, "learning_rate": 1.2158531586679267e-05, "loss": 0.0026, "step": 78460 }, { "grad_norm": 0.07404214143753052, "learning_rate": 1.2147726370516477e-05, "loss": 0.0029, "step": 78470 }, { "grad_norm": 0.11349886655807495, "learning_rate": 1.2136925293821422e-05, "loss": 0.0033, "step": 78480 }, { "grad_norm": 0.08772256970405579, "learning_rate": 1.2126128357775306e-05, "loss": 0.0045, "step": 78490 }, { "grad_norm": 0.13134989142417908, "learning_rate": 1.2115335563558838e-05, "loss": 0.0032, "step": 78500 }, { "grad_norm": 0.14101478457450867, "learning_rate": 1.2104546912352327e-05, "loss": 0.0033, "step": 78510 }, { "grad_norm": 0.10065742582082748, "learning_rate": 1.2093762405335557e-05, "loss": 0.0031, "step": 78520 }, { "grad_norm": 0.11045316606760025, "learning_rate": 1.2082982043687974e-05, "loss": 0.0039, "step": 78530 }, { "grad_norm": 0.07003790885210037, "learning_rate": 1.207220582858845e-05, "loss": 0.0044, "step": 78540 }, { "grad_norm": 0.12927275896072388, "learning_rate": 1.206143376121549e-05, "loss": 0.0041, "step": 78550 }, { "grad_norm": 0.08812731504440308, "learning_rate": 1.2050665842747078e-05, "loss": 0.0033, "step": 78560 }, { "grad_norm": 0.08309164643287659, "learning_rate": 1.2039902074360804e-05, "loss": 0.0036, "step": 78570 }, { "grad_norm": 0.07023383677005768, "learning_rate": 1.2029142457233755e-05, "loss": 0.0044, "step": 78580 }, { "grad_norm": 0.1187177449464798, "learning_rate": 1.2018386992542601e-05, "loss": 0.0055, "step": 78590 }, { "grad_norm": 0.11578052490949631, "learning_rate": 1.2007635681463542e-05, "loss": 0.0039, "step": 78600 }, { "grad_norm": 0.07846398651599884, "learning_rate": 1.1996888525172334e-05, "loss": 0.004, "step": 78610 }, { "grad_norm": 0.1535266786813736, "learning_rate": 1.198614552484425e-05, "loss": 0.0034, "step": 78620 }, { "grad_norm": 0.06294115632772446, "learning_rate": 1.197540668165415e-05, "loss": 0.0031, "step": 78630 }, { "grad_norm": 0.08535342663526535, "learning_rate": 1.1964671996776395e-05, "loss": 0.0033, "step": 78640 }, { "grad_norm": 0.07810073345899582, "learning_rate": 1.1953941471384922e-05, "loss": 0.0035, "step": 78650 }, { "grad_norm": 0.08778686076402664, "learning_rate": 1.1943215106653205e-05, "loss": 0.0038, "step": 78660 }, { "grad_norm": 0.06299459934234619, "learning_rate": 1.193249290375429e-05, "loss": 0.0037, "step": 78670 }, { "grad_norm": 0.09946944564580917, "learning_rate": 1.1921774863860696e-05, "loss": 0.0043, "step": 78680 }, { "grad_norm": 0.07184426486492157, "learning_rate": 1.1911060988144556e-05, "loss": 0.0037, "step": 78690 }, { "grad_norm": 0.09314613044261932, "learning_rate": 1.1900351277777533e-05, "loss": 0.0039, "step": 78700 }, { "grad_norm": 0.1476200520992279, "learning_rate": 1.1889645733930798e-05, "loss": 0.0034, "step": 78710 }, { "grad_norm": 0.1290358453989029, "learning_rate": 1.1878944357775107e-05, "loss": 0.0036, "step": 78720 }, { "grad_norm": 0.10807038098573685, "learning_rate": 1.1868247150480749e-05, "loss": 0.0034, "step": 78730 }, { "grad_norm": 0.08970320969820023, "learning_rate": 1.1857554113217567e-05, "loss": 0.0043, "step": 78740 }, { "grad_norm": 0.1377875655889511, "learning_rate": 1.1846865247154903e-05, "loss": 0.004, "step": 78750 }, { "grad_norm": 0.1315641850233078, "learning_rate": 1.1836180553461711e-05, "loss": 0.0044, "step": 78760 }, { "grad_norm": 0.0657372698187828, "learning_rate": 1.1825500033306419e-05, "loss": 0.003, "step": 78770 }, { "grad_norm": 0.10931355506181717, "learning_rate": 1.1814823687857052e-05, "loss": 0.0041, "step": 78780 }, { "grad_norm": 0.15367887914180756, "learning_rate": 1.1804151518281153e-05, "loss": 0.0038, "step": 78790 }, { "grad_norm": 0.08977442979812622, "learning_rate": 1.179348352574583e-05, "loss": 0.0027, "step": 78800 }, { "grad_norm": 0.08230432122945786, "learning_rate": 1.1782819711417697e-05, "loss": 0.0038, "step": 78810 }, { "grad_norm": 0.08662835508584976, "learning_rate": 1.1772160076462951e-05, "loss": 0.0036, "step": 78820 }, { "grad_norm": 0.09931708872318268, "learning_rate": 1.1761504622047293e-05, "loss": 0.0032, "step": 78830 }, { "grad_norm": 0.06897237151861191, "learning_rate": 1.1750853349335999e-05, "loss": 0.0042, "step": 78840 }, { "grad_norm": 0.08158712834119797, "learning_rate": 1.1740206259493874e-05, "loss": 0.0032, "step": 78850 }, { "grad_norm": 0.12458455562591553, "learning_rate": 1.1729563353685286e-05, "loss": 0.0036, "step": 78860 }, { "grad_norm": 0.06254777312278748, "learning_rate": 1.1718924633074091e-05, "loss": 0.0037, "step": 78870 }, { "grad_norm": 0.07100506871938705, "learning_rate": 1.1708290098823765e-05, "loss": 0.0051, "step": 78880 }, { "grad_norm": 0.14501336216926575, "learning_rate": 1.1697659752097245e-05, "loss": 0.0037, "step": 78890 }, { "grad_norm": 0.15603694319725037, "learning_rate": 1.1687033594057068e-05, "loss": 0.0032, "step": 78900 }, { "grad_norm": 0.0803370401263237, "learning_rate": 1.1676411625865297e-05, "loss": 0.0036, "step": 78910 }, { "grad_norm": 0.08293856680393219, "learning_rate": 1.1665793848683542e-05, "loss": 0.0026, "step": 78920 }, { "grad_norm": 0.06087760627269745, "learning_rate": 1.1655180263672927e-05, "loss": 0.0039, "step": 78930 }, { "grad_norm": 0.08230527490377426, "learning_rate": 1.1644570871994142e-05, "loss": 0.0035, "step": 78940 }, { "grad_norm": 0.11819925904273987, "learning_rate": 1.163396567480744e-05, "loss": 0.0035, "step": 78950 }, { "grad_norm": 0.14290976524353027, "learning_rate": 1.1623364673272552e-05, "loss": 0.004, "step": 78960 }, { "grad_norm": 0.08371712267398834, "learning_rate": 1.1612767868548807e-05, "loss": 0.0029, "step": 78970 }, { "grad_norm": 0.08088640868663788, "learning_rate": 1.1602175261795057e-05, "loss": 0.0044, "step": 78980 }, { "grad_norm": 0.09941881895065308, "learning_rate": 1.1591586854169696e-05, "loss": 0.0052, "step": 78990 }, { "grad_norm": 0.06828382611274719, "learning_rate": 1.158100264683064e-05, "loss": 0.0054, "step": 79000 }, { "grad_norm": 0.06536701321601868, "learning_rate": 1.1570422640935386e-05, "loss": 0.0035, "step": 79010 }, { "grad_norm": 0.1229565367102623, "learning_rate": 1.155984683764092e-05, "loss": 0.0037, "step": 79020 }, { "grad_norm": 0.08495086431503296, "learning_rate": 1.1549275238103812e-05, "loss": 0.0031, "step": 79030 }, { "grad_norm": 0.09870882332324982, "learning_rate": 1.1538707843480152e-05, "loss": 0.003, "step": 79040 }, { "grad_norm": 0.09977932274341583, "learning_rate": 1.1528144654925593e-05, "loss": 0.004, "step": 79050 }, { "grad_norm": 0.0852857306599617, "learning_rate": 1.1517585673595277e-05, "loss": 0.0037, "step": 79060 }, { "grad_norm": 0.0746292993426323, "learning_rate": 1.150703090064395e-05, "loss": 0.0033, "step": 79070 }, { "grad_norm": 0.08888320624828339, "learning_rate": 1.1496480337225835e-05, "loss": 0.003, "step": 79080 }, { "grad_norm": 0.07456934452056885, "learning_rate": 1.1485933984494751e-05, "loss": 0.0031, "step": 79090 }, { "grad_norm": 0.07063014805316925, "learning_rate": 1.1475391843604e-05, "loss": 0.0034, "step": 79100 }, { "grad_norm": 0.10111910104751587, "learning_rate": 1.1464853915706503e-05, "loss": 0.0031, "step": 79110 }, { "grad_norm": 0.0635465458035469, "learning_rate": 1.1454320201954626e-05, "loss": 0.0033, "step": 79120 }, { "grad_norm": 0.051375143229961395, "learning_rate": 1.1443790703500356e-05, "loss": 0.0036, "step": 79130 }, { "grad_norm": 0.061348918825387955, "learning_rate": 1.1433265421495154e-05, "loss": 0.0029, "step": 79140 }, { "grad_norm": 0.06891530752182007, "learning_rate": 1.1422744357090065e-05, "loss": 0.004, "step": 79150 }, { "grad_norm": 0.07586221396923065, "learning_rate": 1.1412227511435647e-05, "loss": 0.0034, "step": 79160 }, { "grad_norm": 0.052125923335552216, "learning_rate": 1.1401714885682025e-05, "loss": 0.0033, "step": 79170 }, { "grad_norm": 0.0924774706363678, "learning_rate": 1.1391206480978823e-05, "loss": 0.0038, "step": 79180 }, { "grad_norm": 0.11256413906812668, "learning_rate": 1.1380702298475227e-05, "loss": 0.0039, "step": 79190 }, { "grad_norm": 0.10868555307388306, "learning_rate": 1.1370202339319985e-05, "loss": 0.0044, "step": 79200 }, { "grad_norm": 0.12048996984958649, "learning_rate": 1.1359706604661319e-05, "loss": 0.0027, "step": 79210 }, { "grad_norm": 0.11391960829496384, "learning_rate": 1.1349215095647042e-05, "loss": 0.0031, "step": 79220 }, { "grad_norm": 0.05271729826927185, "learning_rate": 1.1338727813424493e-05, "loss": 0.0028, "step": 79230 }, { "grad_norm": 0.07197800278663635, "learning_rate": 1.132824475914056e-05, "loss": 0.0032, "step": 79240 }, { "grad_norm": 0.0736842006444931, "learning_rate": 1.131776593394162e-05, "loss": 0.0035, "step": 79250 }, { "grad_norm": 0.08333197981119156, "learning_rate": 1.1307291338973657e-05, "loss": 0.0041, "step": 79260 }, { "grad_norm": 0.1071963980793953, "learning_rate": 1.1296820975382121e-05, "loss": 0.0028, "step": 79270 }, { "grad_norm": 0.1135689988732338, "learning_rate": 1.1286354844312074e-05, "loss": 0.0044, "step": 79280 }, { "grad_norm": 0.07076875865459442, "learning_rate": 1.1275892946908023e-05, "loss": 0.0033, "step": 79290 }, { "grad_norm": 0.1492077261209488, "learning_rate": 1.126543528431413e-05, "loss": 0.0045, "step": 79300 }, { "grad_norm": 0.10707747936248779, "learning_rate": 1.125498185767398e-05, "loss": 0.0043, "step": 79310 }, { "grad_norm": 0.07490526884794235, "learning_rate": 1.1244532668130781e-05, "loss": 0.0031, "step": 79320 }, { "grad_norm": 0.07976862788200378, "learning_rate": 1.1234087716827203e-05, "loss": 0.0043, "step": 79330 }, { "grad_norm": 0.08059582859277725, "learning_rate": 1.1223647004905524e-05, "loss": 0.0031, "step": 79340 }, { "grad_norm": 0.0808425098657608, "learning_rate": 1.1213210533507485e-05, "loss": 0.0052, "step": 79350 }, { "grad_norm": 0.1088743507862091, "learning_rate": 1.1202778303774447e-05, "loss": 0.0032, "step": 79360 }, { "grad_norm": 0.07834342867136002, "learning_rate": 1.1192350316847234e-05, "loss": 0.0046, "step": 79370 }, { "grad_norm": 0.0891840010881424, "learning_rate": 1.1181926573866258e-05, "loss": 0.0052, "step": 79380 }, { "grad_norm": 0.09875711053609848, "learning_rate": 1.1171507075971416e-05, "loss": 0.003, "step": 79390 }, { "grad_norm": 0.07221177220344543, "learning_rate": 1.1161091824302177e-05, "loss": 0.0042, "step": 79400 }, { "grad_norm": 0.07653395086526871, "learning_rate": 1.115068081999755e-05, "loss": 0.0032, "step": 79410 }, { "grad_norm": 0.0910440981388092, "learning_rate": 1.1140274064196071e-05, "loss": 0.003, "step": 79420 }, { "grad_norm": 0.06923814862966537, "learning_rate": 1.112987155803578e-05, "loss": 0.0034, "step": 79430 }, { "grad_norm": 0.12757131457328796, "learning_rate": 1.1119473302654298e-05, "loss": 0.0033, "step": 79440 }, { "grad_norm": 0.10154112428426743, "learning_rate": 1.1109079299188769e-05, "loss": 0.0039, "step": 79450 }, { "grad_norm": 0.10392246395349503, "learning_rate": 1.1098689548775847e-05, "loss": 0.0036, "step": 79460 }, { "grad_norm": 0.07991623878479004, "learning_rate": 1.1088304052551762e-05, "loss": 0.0033, "step": 79470 }, { "grad_norm": 0.049957189708948135, "learning_rate": 1.1077922811652215e-05, "loss": 0.0032, "step": 79480 }, { "grad_norm": 0.09079896658658981, "learning_rate": 1.1067545827212534e-05, "loss": 0.0035, "step": 79490 }, { "grad_norm": 0.06776483356952667, "learning_rate": 1.1057173100367496e-05, "loss": 0.0038, "step": 79500 }, { "grad_norm": 0.06341473013162613, "learning_rate": 1.1046804632251474e-05, "loss": 0.0041, "step": 79510 }, { "grad_norm": 0.10492798686027527, "learning_rate": 1.1036440423998318e-05, "loss": 0.0032, "step": 79520 }, { "grad_norm": 0.061040639877319336, "learning_rate": 1.1026080476741469e-05, "loss": 0.0031, "step": 79530 }, { "grad_norm": 0.11345129460096359, "learning_rate": 1.1015724791613836e-05, "loss": 0.0048, "step": 79540 }, { "grad_norm": 0.154547318816185, "learning_rate": 1.1005373369747951e-05, "loss": 0.0049, "step": 79550 }, { "grad_norm": 0.07472431659698486, "learning_rate": 1.09950262122758e-05, "loss": 0.0032, "step": 79560 }, { "grad_norm": 0.09668587893247604, "learning_rate": 1.0984683320328948e-05, "loss": 0.0048, "step": 79570 }, { "grad_norm": 0.061003897339105606, "learning_rate": 1.0974344695038458e-05, "loss": 0.0035, "step": 79580 }, { "grad_norm": 0.0604754239320755, "learning_rate": 1.0964010337534968e-05, "loss": 0.0041, "step": 79590 }, { "grad_norm": 0.12819351255893707, "learning_rate": 1.0953680248948611e-05, "loss": 0.0048, "step": 79600 }, { "grad_norm": 0.09729181230068207, "learning_rate": 1.0943354430409069e-05, "loss": 0.0038, "step": 79610 }, { "grad_norm": 0.12236884981393814, "learning_rate": 1.0933032883045574e-05, "loss": 0.0036, "step": 79620 }, { "grad_norm": 0.12116425484418869, "learning_rate": 1.0922715607986872e-05, "loss": 0.0039, "step": 79630 }, { "grad_norm": 0.09465741366147995, "learning_rate": 1.0912402606361234e-05, "loss": 0.0043, "step": 79640 }, { "grad_norm": 0.11762160062789917, "learning_rate": 1.090209387929647e-05, "loss": 0.003, "step": 79650 }, { "grad_norm": 0.13723085820674896, "learning_rate": 1.0891789427919957e-05, "loss": 0.0039, "step": 79660 }, { "grad_norm": 0.052569661289453506, "learning_rate": 1.088148925335854e-05, "loss": 0.0032, "step": 79670 }, { "grad_norm": 0.10000479221343994, "learning_rate": 1.087119335673864e-05, "loss": 0.0042, "step": 79680 }, { "grad_norm": 0.09252951294183731, "learning_rate": 1.0860901739186208e-05, "loss": 0.0034, "step": 79690 }, { "grad_norm": 0.07535271346569061, "learning_rate": 1.085061440182673e-05, "loss": 0.0036, "step": 79700 }, { "grad_norm": 0.07448267936706543, "learning_rate": 1.0840331345785193e-05, "loss": 0.004, "step": 79710 }, { "grad_norm": 0.09661445021629333, "learning_rate": 1.083005257218615e-05, "loss": 0.0031, "step": 79720 }, { "grad_norm": 0.07660593092441559, "learning_rate": 1.0819778082153642e-05, "loss": 0.0042, "step": 79730 }, { "grad_norm": 0.06315509229898453, "learning_rate": 1.0809507876811325e-05, "loss": 0.0043, "step": 79740 }, { "grad_norm": 0.0750444084405899, "learning_rate": 1.0799241957282292e-05, "loss": 0.0038, "step": 79750 }, { "grad_norm": 0.07026544958353043, "learning_rate": 1.078898032468923e-05, "loss": 0.0026, "step": 79760 }, { "grad_norm": 0.11612297594547272, "learning_rate": 1.0778722980154315e-05, "loss": 0.0044, "step": 79770 }, { "grad_norm": 0.09552060812711716, "learning_rate": 1.0768469924799302e-05, "loss": 0.0027, "step": 79780 }, { "grad_norm": 0.07756765186786652, "learning_rate": 1.0758221159745419e-05, "loss": 0.0026, "step": 79790 }, { "grad_norm": 0.08846350759267807, "learning_rate": 1.074797668611347e-05, "loss": 0.0046, "step": 79800 }, { "grad_norm": 0.08304907381534576, "learning_rate": 1.0737736505023777e-05, "loss": 0.0052, "step": 79810 }, { "grad_norm": 0.08506019413471222, "learning_rate": 1.0727500617596203e-05, "loss": 0.0033, "step": 79820 }, { "grad_norm": 0.0844024047255516, "learning_rate": 1.0717269024950104e-05, "loss": 0.0031, "step": 79830 }, { "grad_norm": 0.08480605483055115, "learning_rate": 1.070704172820442e-05, "loss": 0.0035, "step": 79840 }, { "grad_norm": 0.08200028538703918, "learning_rate": 1.0696818728477559e-05, "loss": 0.0057, "step": 79850 }, { "grad_norm": 0.06377332657575607, "learning_rate": 1.0686600026887511e-05, "loss": 0.005, "step": 79860 }, { "grad_norm": 0.09779617935419083, "learning_rate": 1.0676385624551777e-05, "loss": 0.0027, "step": 79870 }, { "grad_norm": 0.04978339746594429, "learning_rate": 1.0666175522587402e-05, "loss": 0.0051, "step": 79880 }, { "grad_norm": 0.08839166909456253, "learning_rate": 1.0655969722110925e-05, "loss": 0.0057, "step": 79890 }, { "grad_norm": 0.13254623115062714, "learning_rate": 1.0645768224238439e-05, "loss": 0.0031, "step": 79900 }, { "grad_norm": 0.09418173879384995, "learning_rate": 1.063557103008559e-05, "loss": 0.0036, "step": 79910 }, { "grad_norm": 0.1110239252448082, "learning_rate": 1.0625378140767494e-05, "loss": 0.0034, "step": 79920 }, { "grad_norm": 0.08536951243877411, "learning_rate": 1.0615189557398846e-05, "loss": 0.004, "step": 79930 }, { "grad_norm": 0.09708378463983536, "learning_rate": 1.060500528109386e-05, "loss": 0.0053, "step": 79940 }, { "grad_norm": 0.06321890652179718, "learning_rate": 1.0594825312966272e-05, "loss": 0.0023, "step": 79950 }, { "grad_norm": 0.062342483550310135, "learning_rate": 1.0584649654129336e-05, "loss": 0.0039, "step": 79960 }, { "grad_norm": 0.10626669973134995, "learning_rate": 1.0574478305695863e-05, "loss": 0.0042, "step": 79970 }, { "grad_norm": 0.08736484497785568, "learning_rate": 1.0564311268778154e-05, "loss": 0.0032, "step": 79980 }, { "grad_norm": 0.04935245215892792, "learning_rate": 1.0554148544488073e-05, "loss": 0.0041, "step": 79990 }, { "grad_norm": 0.03890278562903404, "learning_rate": 1.0543990133937004e-05, "loss": 0.0033, "step": 80000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 64, "trial_name": null, "trial_params": null }