{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6, "eval_steps": 500, "global_step": 9000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.1528221219778061, "learning_rate": 1.2000000000000002e-06, "loss": 1.2068, "step": 10 }, { "grad_norm": 0.12877728044986725, "learning_rate": 2.5333333333333334e-06, "loss": 1.2134, "step": 20 }, { "grad_norm": 0.17161273956298828, "learning_rate": 3.866666666666667e-06, "loss": 1.206, "step": 30 }, { "grad_norm": 0.2625730037689209, "learning_rate": 5.2e-06, "loss": 1.1817, "step": 40 }, { "grad_norm": 0.2874051034450531, "learning_rate": 6.533333333333333e-06, "loss": 1.1488, "step": 50 }, { "grad_norm": 0.34452179074287415, "learning_rate": 7.866666666666667e-06, "loss": 1.1238, "step": 60 }, { "grad_norm": 0.3187244236469269, "learning_rate": 9.2e-06, "loss": 1.1072, "step": 70 }, { "grad_norm": 0.41171619296073914, "learning_rate": 1.0533333333333335e-05, "loss": 1.0936, "step": 80 }, { "grad_norm": 0.22929424047470093, "learning_rate": 1.1866666666666668e-05, "loss": 1.0933, "step": 90 }, { "grad_norm": 0.5159542560577393, "learning_rate": 1.32e-05, "loss": 1.0919, "step": 100 }, { "grad_norm": 0.4722374975681305, "learning_rate": 1.4533333333333335e-05, "loss": 1.0945, "step": 110 }, { "grad_norm": 0.2221572995185852, "learning_rate": 1.586666666666667e-05, "loss": 1.0911, "step": 120 }, { "grad_norm": 0.5312994122505188, "learning_rate": 1.7199999999999998e-05, "loss": 1.0815, "step": 130 }, { "grad_norm": 0.3242526650428772, "learning_rate": 1.8533333333333334e-05, "loss": 1.0778, "step": 140 }, { "grad_norm": 0.28936564922332764, "learning_rate": 1.9866666666666667e-05, "loss": 1.0505, "step": 150 }, { "grad_norm": 0.3376358151435852, "learning_rate": 2.12e-05, "loss": 1.02, "step": 160 }, { "grad_norm": 0.5211641192436218, "learning_rate": 2.2533333333333333e-05, "loss": 0.9858, "step": 170 }, { "grad_norm": 0.41194382309913635, "learning_rate": 2.3866666666666666e-05, "loss": 0.9434, "step": 180 }, { "grad_norm": 0.5447150468826294, "learning_rate": 2.5200000000000003e-05, "loss": 0.9057, "step": 190 }, { "grad_norm": 0.8995847702026367, "learning_rate": 2.6533333333333332e-05, "loss": 0.8692, "step": 200 }, { "grad_norm": 0.7821958065032959, "learning_rate": 2.786666666666667e-05, "loss": 0.8394, "step": 210 }, { "grad_norm": 1.4184938669204712, "learning_rate": 2.9199999999999998e-05, "loss": 0.8041, "step": 220 }, { "grad_norm": 0.9261109828948975, "learning_rate": 3.0533333333333335e-05, "loss": 0.7759, "step": 230 }, { "grad_norm": 1.2726421356201172, "learning_rate": 3.1866666666666664e-05, "loss": 0.7407, "step": 240 }, { "grad_norm": 1.1038875579833984, "learning_rate": 3.32e-05, "loss": 0.7069, "step": 250 }, { "grad_norm": 1.1673762798309326, "learning_rate": 3.453333333333334e-05, "loss": 0.6797, "step": 260 }, { "grad_norm": 1.0007288455963135, "learning_rate": 3.586666666666667e-05, "loss": 0.6563, "step": 270 }, { "grad_norm": 0.8429192304611206, "learning_rate": 3.72e-05, "loss": 0.6337, "step": 280 }, { "grad_norm": 0.9413348436355591, "learning_rate": 3.853333333333334e-05, "loss": 0.6113, "step": 290 }, { "grad_norm": 0.99355548620224, "learning_rate": 3.986666666666667e-05, "loss": 0.6006, "step": 300 }, { "grad_norm": 1.1705503463745117, "learning_rate": 4.12e-05, "loss": 0.5865, "step": 310 }, { "grad_norm": 1.3885726928710938, "learning_rate": 4.2533333333333335e-05, "loss": 0.5685, "step": 320 }, { "grad_norm": 1.224961280822754, "learning_rate": 4.3866666666666665e-05, "loss": 0.5633, "step": 330 }, { "grad_norm": 1.2367912530899048, "learning_rate": 4.52e-05, "loss": 0.5547, "step": 340 }, { "grad_norm": 1.1392102241516113, "learning_rate": 4.653333333333334e-05, "loss": 0.5403, "step": 350 }, { "grad_norm": 1.213355541229248, "learning_rate": 4.7866666666666674e-05, "loss": 0.5357, "step": 360 }, { "grad_norm": 1.2761033773422241, "learning_rate": 4.92e-05, "loss": 0.5058, "step": 370 }, { "grad_norm": 1.1698124408721924, "learning_rate": 5.053333333333333e-05, "loss": 0.4966, "step": 380 }, { "grad_norm": 1.201072096824646, "learning_rate": 5.1866666666666676e-05, "loss": 0.4828, "step": 390 }, { "grad_norm": 1.0158417224884033, "learning_rate": 5.3200000000000006e-05, "loss": 0.4651, "step": 400 }, { "grad_norm": 1.346627116203308, "learning_rate": 5.4533333333333335e-05, "loss": 0.4594, "step": 410 }, { "grad_norm": 1.0970842838287354, "learning_rate": 5.5866666666666665e-05, "loss": 0.4445, "step": 420 }, { "grad_norm": 1.2192479372024536, "learning_rate": 5.72e-05, "loss": 0.4299, "step": 430 }, { "grad_norm": 1.2104856967926025, "learning_rate": 5.853333333333334e-05, "loss": 0.4117, "step": 440 }, { "grad_norm": 1.2335385084152222, "learning_rate": 5.9866666666666674e-05, "loss": 0.4031, "step": 450 }, { "grad_norm": 0.9111726880073547, "learning_rate": 6.12e-05, "loss": 0.3901, "step": 460 }, { "grad_norm": 1.1724517345428467, "learning_rate": 6.253333333333333e-05, "loss": 0.3773, "step": 470 }, { "grad_norm": 1.2332345247268677, "learning_rate": 6.386666666666667e-05, "loss": 0.3737, "step": 480 }, { "grad_norm": 1.1842234134674072, "learning_rate": 6.52e-05, "loss": 0.3628, "step": 490 }, { "grad_norm": 1.2130239009857178, "learning_rate": 6.653333333333334e-05, "loss": 0.3544, "step": 500 }, { "grad_norm": 1.0411441326141357, "learning_rate": 6.786666666666667e-05, "loss": 0.3326, "step": 510 }, { "grad_norm": 1.2490897178649902, "learning_rate": 6.92e-05, "loss": 0.3279, "step": 520 }, { "grad_norm": 1.33174467086792, "learning_rate": 7.053333333333334e-05, "loss": 0.3177, "step": 530 }, { "grad_norm": 1.1573432683944702, "learning_rate": 7.186666666666667e-05, "loss": 0.3106, "step": 540 }, { "grad_norm": 1.2820322513580322, "learning_rate": 7.32e-05, "loss": 0.2829, "step": 550 }, { "grad_norm": 1.2154183387756348, "learning_rate": 7.453333333333333e-05, "loss": 0.2846, "step": 560 }, { "grad_norm": 1.0811824798583984, "learning_rate": 7.586666666666668e-05, "loss": 0.274, "step": 570 }, { "grad_norm": 1.2624622583389282, "learning_rate": 7.72e-05, "loss": 0.2581, "step": 580 }, { "grad_norm": 1.4608895778656006, "learning_rate": 7.853333333333334e-05, "loss": 0.2436, "step": 590 }, { "grad_norm": 1.3445744514465332, "learning_rate": 7.986666666666667e-05, "loss": 0.2339, "step": 600 }, { "grad_norm": 1.2945458889007568, "learning_rate": 8.120000000000001e-05, "loss": 0.22, "step": 610 }, { "grad_norm": 2.7877917289733887, "learning_rate": 8.253333333333334e-05, "loss": 0.206, "step": 620 }, { "grad_norm": 1.1054668426513672, "learning_rate": 8.386666666666667e-05, "loss": 0.2005, "step": 630 }, { "grad_norm": 1.2020183801651, "learning_rate": 8.52e-05, "loss": 0.1906, "step": 640 }, { "grad_norm": 1.3809555768966675, "learning_rate": 8.653333333333333e-05, "loss": 0.1761, "step": 650 }, { "grad_norm": 1.2968473434448242, "learning_rate": 8.786666666666667e-05, "loss": 0.17, "step": 660 }, { "grad_norm": 1.5829375982284546, "learning_rate": 8.92e-05, "loss": 0.1582, "step": 670 }, { "grad_norm": 1.4462990760803223, "learning_rate": 9.053333333333334e-05, "loss": 0.1532, "step": 680 }, { "grad_norm": 1.4937175512313843, "learning_rate": 9.186666666666667e-05, "loss": 0.1436, "step": 690 }, { "grad_norm": 1.0529409646987915, "learning_rate": 9.320000000000002e-05, "loss": 0.146, "step": 700 }, { "grad_norm": 1.4069396257400513, "learning_rate": 9.453333333333335e-05, "loss": 0.1401, "step": 710 }, { "grad_norm": 1.448398232460022, "learning_rate": 9.586666666666667e-05, "loss": 0.1399, "step": 720 }, { "grad_norm": 1.3302212953567505, "learning_rate": 9.72e-05, "loss": 0.1397, "step": 730 }, { "grad_norm": 1.2735353708267212, "learning_rate": 9.853333333333333e-05, "loss": 0.1411, "step": 740 }, { "grad_norm": 1.265570044517517, "learning_rate": 9.986666666666668e-05, "loss": 0.1434, "step": 750 }, { "grad_norm": 1.3887313604354858, "learning_rate": 9.999990157738453e-05, "loss": 0.1386, "step": 760 }, { "grad_norm": 1.2708746194839478, "learning_rate": 9.999956135155687e-05, "loss": 0.1394, "step": 770 }, { "grad_norm": 1.3188905715942383, "learning_rate": 9.99989781090763e-05, "loss": 0.1362, "step": 780 }, { "grad_norm": 2.1023690700531006, "learning_rate": 9.999815185277755e-05, "loss": 0.132, "step": 790 }, { "grad_norm": 1.0511329174041748, "learning_rate": 9.999708258667652e-05, "loss": 0.1393, "step": 800 }, { "grad_norm": 1.076863169670105, "learning_rate": 9.999577031597029e-05, "loss": 0.1219, "step": 810 }, { "grad_norm": 1.0692609548568726, "learning_rate": 9.999421504703696e-05, "loss": 0.1249, "step": 820 }, { "grad_norm": 1.0652865171432495, "learning_rate": 9.999241678743574e-05, "loss": 0.124, "step": 830 }, { "grad_norm": 1.1164456605911255, "learning_rate": 9.999037554590683e-05, "loss": 0.1301, "step": 840 }, { "grad_norm": 1.4576336145401, "learning_rate": 9.998809133237143e-05, "loss": 0.1263, "step": 850 }, { "grad_norm": 1.0527737140655518, "learning_rate": 9.998556415793169e-05, "loss": 0.1283, "step": 860 }, { "grad_norm": 1.2238246202468872, "learning_rate": 9.998279403487062e-05, "loss": 0.1271, "step": 870 }, { "grad_norm": 1.146423578262329, "learning_rate": 9.997978097665205e-05, "loss": 0.1269, "step": 880 }, { "grad_norm": 1.0656620264053345, "learning_rate": 9.99765249979206e-05, "loss": 0.1239, "step": 890 }, { "grad_norm": 1.1446648836135864, "learning_rate": 9.997302611450154e-05, "loss": 0.1237, "step": 900 }, { "grad_norm": 1.1433049440383911, "learning_rate": 9.996928434340073e-05, "loss": 0.1147, "step": 910 }, { "grad_norm": 0.9724878072738647, "learning_rate": 9.996529970280462e-05, "loss": 0.1113, "step": 920 }, { "grad_norm": 1.074139952659607, "learning_rate": 9.996107221208004e-05, "loss": 0.1198, "step": 930 }, { "grad_norm": 0.9398661851882935, "learning_rate": 9.995660189177419e-05, "loss": 0.1149, "step": 940 }, { "grad_norm": 0.8259074687957764, "learning_rate": 9.995188876361451e-05, "loss": 0.1155, "step": 950 }, { "grad_norm": 1.1398017406463623, "learning_rate": 9.994693285050857e-05, "loss": 0.1192, "step": 960 }, { "grad_norm": 0.9386220574378967, "learning_rate": 9.994173417654395e-05, "loss": 0.1186, "step": 970 }, { "grad_norm": 0.9809224605560303, "learning_rate": 9.993629276698821e-05, "loss": 0.1135, "step": 980 }, { "grad_norm": 1.196757197380066, "learning_rate": 9.993060864828858e-05, "loss": 0.1247, "step": 990 }, { "grad_norm": 0.9535070061683655, "learning_rate": 9.992468184807206e-05, "loss": 0.1098, "step": 1000 }, { "grad_norm": 1.011692762374878, "learning_rate": 9.991851239514511e-05, "loss": 0.1093, "step": 1010 }, { "grad_norm": 0.9049733877182007, "learning_rate": 9.991210031949359e-05, "loss": 0.1171, "step": 1020 }, { "grad_norm": 0.9988471269607544, "learning_rate": 9.990544565228259e-05, "loss": 0.1085, "step": 1030 }, { "grad_norm": 0.9829862713813782, "learning_rate": 9.989854842585631e-05, "loss": 0.1041, "step": 1040 }, { "grad_norm": 0.9640742540359497, "learning_rate": 9.989140867373783e-05, "loss": 0.1053, "step": 1050 }, { "grad_norm": 0.9070046544075012, "learning_rate": 9.988402643062907e-05, "loss": 0.1035, "step": 1060 }, { "grad_norm": 1.0155551433563232, "learning_rate": 9.987640173241046e-05, "loss": 0.1075, "step": 1070 }, { "grad_norm": 1.1102415323257446, "learning_rate": 9.986853461614093e-05, "loss": 0.1132, "step": 1080 }, { "grad_norm": 1.4231481552124023, "learning_rate": 9.986042512005763e-05, "loss": 0.1107, "step": 1090 }, { "grad_norm": 0.9319804310798645, "learning_rate": 9.985207328357573e-05, "loss": 0.1026, "step": 1100 }, { "grad_norm": 1.4864873886108398, "learning_rate": 9.984347914728829e-05, "loss": 0.1011, "step": 1110 }, { "grad_norm": 1.0301569700241089, "learning_rate": 9.983464275296605e-05, "loss": 0.1114, "step": 1120 }, { "grad_norm": 0.9580268859863281, "learning_rate": 9.982556414355724e-05, "loss": 0.1099, "step": 1130 }, { "grad_norm": 1.0523960590362549, "learning_rate": 9.981624336318726e-05, "loss": 0.1053, "step": 1140 }, { "grad_norm": 1.060106635093689, "learning_rate": 9.980668045715864e-05, "loss": 0.1, "step": 1150 }, { "grad_norm": 0.9312297701835632, "learning_rate": 9.979687547195066e-05, "loss": 0.1061, "step": 1160 }, { "grad_norm": 0.9556792974472046, "learning_rate": 9.978682845521927e-05, "loss": 0.107, "step": 1170 }, { "grad_norm": 1.0657126903533936, "learning_rate": 9.977653945579673e-05, "loss": 0.1049, "step": 1180 }, { "grad_norm": 0.8762414455413818, "learning_rate": 9.976600852369144e-05, "loss": 0.1001, "step": 1190 }, { "grad_norm": 0.861569344997406, "learning_rate": 9.975523571008769e-05, "loss": 0.1064, "step": 1200 }, { "grad_norm": 0.9219115376472473, "learning_rate": 9.97442210673454e-05, "loss": 0.0983, "step": 1210 }, { "grad_norm": 1.0116729736328125, "learning_rate": 9.973296464899988e-05, "loss": 0.1018, "step": 1220 }, { "grad_norm": 0.9443844556808472, "learning_rate": 9.972146650976154e-05, "loss": 0.1033, "step": 1230 }, { "grad_norm": 0.9114744067192078, "learning_rate": 9.970972670551566e-05, "loss": 0.1034, "step": 1240 }, { "grad_norm": 0.8281648755073547, "learning_rate": 9.969774529332212e-05, "loss": 0.0995, "step": 1250 }, { "grad_norm": 0.8310883045196533, "learning_rate": 9.968552233141504e-05, "loss": 0.0992, "step": 1260 }, { "grad_norm": 0.948841392993927, "learning_rate": 9.967305787920264e-05, "loss": 0.0985, "step": 1270 }, { "grad_norm": 0.8757680654525757, "learning_rate": 9.966035199726684e-05, "loss": 0.1062, "step": 1280 }, { "grad_norm": 0.7957766056060791, "learning_rate": 9.9647404747363e-05, "loss": 0.0949, "step": 1290 }, { "grad_norm": 0.9588861465454102, "learning_rate": 9.96342161924196e-05, "loss": 0.0952, "step": 1300 }, { "grad_norm": 0.8227360248565674, "learning_rate": 9.962078639653797e-05, "loss": 0.0931, "step": 1310 }, { "grad_norm": 1.0412708520889282, "learning_rate": 9.960711542499202e-05, "loss": 0.0922, "step": 1320 }, { "grad_norm": 0.8027946352958679, "learning_rate": 9.959320334422772e-05, "loss": 0.1075, "step": 1330 }, { "grad_norm": 0.9406609535217285, "learning_rate": 9.957905022186309e-05, "loss": 0.0984, "step": 1340 }, { "grad_norm": 0.8434228301048279, "learning_rate": 9.956465612668757e-05, "loss": 0.0979, "step": 1350 }, { "grad_norm": 0.9157196879386902, "learning_rate": 9.95500211286619e-05, "loss": 0.0941, "step": 1360 }, { "grad_norm": 0.8689035177230835, "learning_rate": 9.953514529891763e-05, "loss": 0.0949, "step": 1370 }, { "grad_norm": 0.7697312831878662, "learning_rate": 9.952002870975693e-05, "loss": 0.0914, "step": 1380 }, { "grad_norm": 0.8707885146141052, "learning_rate": 9.950467143465207e-05, "loss": 0.0926, "step": 1390 }, { "grad_norm": 1.025254487991333, "learning_rate": 9.94890735482452e-05, "loss": 0.0912, "step": 1400 }, { "grad_norm": 0.6527482271194458, "learning_rate": 9.947323512634788e-05, "loss": 0.0873, "step": 1410 }, { "grad_norm": 0.8045205473899841, "learning_rate": 9.945715624594081e-05, "loss": 0.09, "step": 1420 }, { "grad_norm": 0.7984917163848877, "learning_rate": 9.944083698517339e-05, "loss": 0.0907, "step": 1430 }, { "grad_norm": 0.7339720726013184, "learning_rate": 9.942427742336334e-05, "loss": 0.0923, "step": 1440 }, { "grad_norm": 0.8728718757629395, "learning_rate": 9.940747764099638e-05, "loss": 0.0891, "step": 1450 }, { "grad_norm": 0.8001891374588013, "learning_rate": 9.939043771972574e-05, "loss": 0.0999, "step": 1460 }, { "grad_norm": 0.861858606338501, "learning_rate": 9.937315774237186e-05, "loss": 0.0892, "step": 1470 }, { "grad_norm": 0.9286336898803711, "learning_rate": 9.93556377929219e-05, "loss": 0.0946, "step": 1480 }, { "grad_norm": 0.9956574440002441, "learning_rate": 9.933787795652942e-05, "loss": 0.0913, "step": 1490 }, { "grad_norm": 0.8465809226036072, "learning_rate": 9.931987831951386e-05, "loss": 0.0912, "step": 1500 }, { "grad_norm": 0.8478185534477234, "learning_rate": 9.930163896936027e-05, "loss": 0.0958, "step": 1510 }, { "grad_norm": 0.7136925458908081, "learning_rate": 9.92831599947187e-05, "loss": 0.0848, "step": 1520 }, { "grad_norm": 0.8562847375869751, "learning_rate": 9.926444148540393e-05, "loss": 0.0952, "step": 1530 }, { "grad_norm": 0.8773907423019409, "learning_rate": 9.924548353239495e-05, "loss": 0.0972, "step": 1540 }, { "grad_norm": 0.770634114742279, "learning_rate": 9.922628622783451e-05, "loss": 0.0913, "step": 1550 }, { "grad_norm": 0.6977461576461792, "learning_rate": 9.920684966502878e-05, "loss": 0.0914, "step": 1560 }, { "grad_norm": 0.7688239216804504, "learning_rate": 9.918717393844669e-05, "loss": 0.0927, "step": 1570 }, { "grad_norm": 0.7350327372550964, "learning_rate": 9.916725914371969e-05, "loss": 0.0884, "step": 1580 }, { "grad_norm": 0.7852026224136353, "learning_rate": 9.914710537764117e-05, "loss": 0.0869, "step": 1590 }, { "grad_norm": 0.6127405166625977, "learning_rate": 9.912671273816601e-05, "loss": 0.0928, "step": 1600 }, { "grad_norm": 0.8078952431678772, "learning_rate": 9.910608132441008e-05, "loss": 0.0863, "step": 1610 }, { "grad_norm": 0.6758414506912231, "learning_rate": 9.908521123664981e-05, "loss": 0.0797, "step": 1620 }, { "grad_norm": 0.5950756072998047, "learning_rate": 9.906410257632168e-05, "loss": 0.0876, "step": 1630 }, { "grad_norm": 0.9129499793052673, "learning_rate": 9.904275544602169e-05, "loss": 0.0842, "step": 1640 }, { "grad_norm": 0.6972604990005493, "learning_rate": 9.902116994950493e-05, "loss": 0.0842, "step": 1650 }, { "grad_norm": 0.7520588040351868, "learning_rate": 9.899934619168501e-05, "loss": 0.0791, "step": 1660 }, { "grad_norm": 0.7351466417312622, "learning_rate": 9.89772842786336e-05, "loss": 0.0831, "step": 1670 }, { "grad_norm": 0.7723562717437744, "learning_rate": 9.895498431757989e-05, "loss": 0.0868, "step": 1680 }, { "grad_norm": 0.7624236345291138, "learning_rate": 9.893244641691006e-05, "loss": 0.0911, "step": 1690 }, { "grad_norm": 0.6789768934249878, "learning_rate": 9.890967068616677e-05, "loss": 0.0889, "step": 1700 }, { "grad_norm": 0.6409830451011658, "learning_rate": 9.888665723604864e-05, "loss": 0.0853, "step": 1710 }, { "grad_norm": 0.653907835483551, "learning_rate": 9.886340617840968e-05, "loss": 0.0814, "step": 1720 }, { "grad_norm": 0.767878532409668, "learning_rate": 9.883991762625876e-05, "loss": 0.0825, "step": 1730 }, { "grad_norm": 0.7938691973686218, "learning_rate": 9.881619169375908e-05, "loss": 0.0797, "step": 1740 }, { "grad_norm": 0.863423764705658, "learning_rate": 9.879222849622758e-05, "loss": 0.0868, "step": 1750 }, { "grad_norm": 0.7077615261077881, "learning_rate": 9.876802815013439e-05, "loss": 0.083, "step": 1760 }, { "grad_norm": 0.7735344171524048, "learning_rate": 9.87435907731023e-05, "loss": 0.0877, "step": 1770 }, { "grad_norm": 0.7485953569412231, "learning_rate": 9.871891648390614e-05, "loss": 0.0865, "step": 1780 }, { "grad_norm": 0.9190590381622314, "learning_rate": 9.869400540247223e-05, "loss": 0.0893, "step": 1790 }, { "grad_norm": 0.7002912759780884, "learning_rate": 9.866885764987776e-05, "loss": 0.0821, "step": 1800 }, { "grad_norm": 0.7234448790550232, "learning_rate": 9.86434733483503e-05, "loss": 0.0805, "step": 1810 }, { "grad_norm": 0.6578029990196228, "learning_rate": 9.861785262126705e-05, "loss": 0.0853, "step": 1820 }, { "grad_norm": 0.7728158831596375, "learning_rate": 9.85919955931544e-05, "loss": 0.0788, "step": 1830 }, { "grad_norm": 0.6882961392402649, "learning_rate": 9.856590238968721e-05, "loss": 0.0891, "step": 1840 }, { "grad_norm": 0.7271476984024048, "learning_rate": 9.853957313768824e-05, "loss": 0.0859, "step": 1850 }, { "grad_norm": 0.7411231398582458, "learning_rate": 9.851300796512755e-05, "loss": 0.0844, "step": 1860 }, { "grad_norm": 0.7105385661125183, "learning_rate": 9.848620700112188e-05, "loss": 0.078, "step": 1870 }, { "grad_norm": 0.6703158617019653, "learning_rate": 9.845917037593396e-05, "loss": 0.0796, "step": 1880 }, { "grad_norm": 0.6238304376602173, "learning_rate": 9.843189822097196e-05, "loss": 0.0871, "step": 1890 }, { "grad_norm": 0.5811266899108887, "learning_rate": 9.84043906687888e-05, "loss": 0.0779, "step": 1900 }, { "grad_norm": 0.6616613864898682, "learning_rate": 9.837664785308149e-05, "loss": 0.0777, "step": 1910 }, { "grad_norm": 0.7105765342712402, "learning_rate": 9.834866990869059e-05, "loss": 0.0826, "step": 1920 }, { "grad_norm": 0.7052018642425537, "learning_rate": 9.832045697159938e-05, "loss": 0.0817, "step": 1930 }, { "grad_norm": 0.6906279921531677, "learning_rate": 9.829200917893334e-05, "loss": 0.0815, "step": 1940 }, { "grad_norm": 0.7250940203666687, "learning_rate": 9.826332666895944e-05, "loss": 0.0791, "step": 1950 }, { "grad_norm": 0.7513362765312195, "learning_rate": 9.823440958108545e-05, "loss": 0.0793, "step": 1960 }, { "grad_norm": 0.7912622690200806, "learning_rate": 9.820525805585927e-05, "loss": 0.0818, "step": 1970 }, { "grad_norm": 0.8074571490287781, "learning_rate": 9.81758722349683e-05, "loss": 0.0828, "step": 1980 }, { "grad_norm": 0.7632444500923157, "learning_rate": 9.814625226123862e-05, "loss": 0.0869, "step": 1990 }, { "grad_norm": 0.7208681702613831, "learning_rate": 9.811639827863449e-05, "loss": 0.0828, "step": 2000 }, { "grad_norm": 0.763584554195404, "learning_rate": 9.808631043225741e-05, "loss": 0.0807, "step": 2010 }, { "grad_norm": 0.7192198634147644, "learning_rate": 9.805598886834567e-05, "loss": 0.0845, "step": 2020 }, { "grad_norm": 0.6376525163650513, "learning_rate": 9.802543373427344e-05, "loss": 0.0807, "step": 2030 }, { "grad_norm": 0.737170934677124, "learning_rate": 9.799464517855018e-05, "loss": 0.0769, "step": 2040 }, { "grad_norm": 0.6961826086044312, "learning_rate": 9.79636233508198e-05, "loss": 0.082, "step": 2050 }, { "grad_norm": 0.8413031101226807, "learning_rate": 9.793236840186005e-05, "loss": 0.0806, "step": 2060 }, { "grad_norm": 0.7423960566520691, "learning_rate": 9.790088048358175e-05, "loss": 0.0819, "step": 2070 }, { "grad_norm": 0.8054006695747375, "learning_rate": 9.786915974902798e-05, "loss": 0.0733, "step": 2080 }, { "grad_norm": 0.674681544303894, "learning_rate": 9.783720635237343e-05, "loss": 0.0802, "step": 2090 }, { "grad_norm": 0.6383704543113708, "learning_rate": 9.780502044892362e-05, "loss": 0.079, "step": 2100 }, { "grad_norm": 0.6757469773292542, "learning_rate": 9.777260219511415e-05, "loss": 0.076, "step": 2110 }, { "grad_norm": 0.823462724685669, "learning_rate": 9.773995174850989e-05, "loss": 0.0786, "step": 2120 }, { "grad_norm": 0.718417227268219, "learning_rate": 9.770706926780428e-05, "loss": 0.0854, "step": 2130 }, { "grad_norm": 0.7160131931304932, "learning_rate": 9.767395491281855e-05, "loss": 0.0773, "step": 2140 }, { "grad_norm": 0.7102839350700378, "learning_rate": 9.764060884450086e-05, "loss": 0.0741, "step": 2150 }, { "grad_norm": 0.6298885345458984, "learning_rate": 9.76070312249257e-05, "loss": 0.0726, "step": 2160 }, { "grad_norm": 0.7110960483551025, "learning_rate": 9.757322221729283e-05, "loss": 0.0722, "step": 2170 }, { "grad_norm": 0.6902167797088623, "learning_rate": 9.753918198592682e-05, "loss": 0.0737, "step": 2180 }, { "grad_norm": 0.721842885017395, "learning_rate": 9.750491069627593e-05, "loss": 0.0741, "step": 2190 }, { "grad_norm": 0.6490733027458191, "learning_rate": 9.747040851491149e-05, "loss": 0.0869, "step": 2200 }, { "grad_norm": 0.6896973252296448, "learning_rate": 9.743567560952711e-05, "loss": 0.081, "step": 2210 }, { "grad_norm": 0.6303186416625977, "learning_rate": 9.740071214893773e-05, "loss": 0.0791, "step": 2220 }, { "grad_norm": 0.5973021388053894, "learning_rate": 9.736551830307892e-05, "loss": 0.0726, "step": 2230 }, { "grad_norm": 0.6787769198417664, "learning_rate": 9.733009424300597e-05, "loss": 0.0776, "step": 2240 }, { "grad_norm": 0.8184981346130371, "learning_rate": 9.729444014089314e-05, "loss": 0.0771, "step": 2250 }, { "grad_norm": 0.6981377601623535, "learning_rate": 9.725855617003275e-05, "loss": 0.0834, "step": 2260 }, { "grad_norm": 0.6165905594825745, "learning_rate": 9.72224425048344e-05, "loss": 0.0767, "step": 2270 }, { "grad_norm": 0.6394922137260437, "learning_rate": 9.718609932082405e-05, "loss": 0.07, "step": 2280 }, { "grad_norm": 0.6943788528442383, "learning_rate": 9.714952679464323e-05, "loss": 0.0756, "step": 2290 }, { "grad_norm": 0.7439070343971252, "learning_rate": 9.711272510404816e-05, "loss": 0.0762, "step": 2300 }, { "grad_norm": 0.6924915909767151, "learning_rate": 9.70756944279089e-05, "loss": 0.0801, "step": 2310 }, { "grad_norm": 0.6574571132659912, "learning_rate": 9.70384349462084e-05, "loss": 0.0756, "step": 2320 }, { "grad_norm": 0.6405468583106995, "learning_rate": 9.700094684004182e-05, "loss": 0.0827, "step": 2330 }, { "grad_norm": 0.6328576803207397, "learning_rate": 9.696323029161535e-05, "loss": 0.071, "step": 2340 }, { "grad_norm": 0.6095026135444641, "learning_rate": 9.692528548424567e-05, "loss": 0.0788, "step": 2350 }, { "grad_norm": 0.6449398398399353, "learning_rate": 9.688711260235872e-05, "loss": 0.0772, "step": 2360 }, { "grad_norm": 0.6582117676734924, "learning_rate": 9.684871183148912e-05, "loss": 0.0723, "step": 2370 }, { "grad_norm": 0.7571628093719482, "learning_rate": 9.681008335827898e-05, "loss": 0.0766, "step": 2380 }, { "grad_norm": 0.7709692716598511, "learning_rate": 9.677122737047724e-05, "loss": 0.0794, "step": 2390 }, { "grad_norm": 0.6195625066757202, "learning_rate": 9.673214405693857e-05, "loss": 0.0736, "step": 2400 }, { "grad_norm": 0.7350249290466309, "learning_rate": 9.669283360762258e-05, "loss": 0.0753, "step": 2410 }, { "grad_norm": 0.6809076070785522, "learning_rate": 9.66532962135928e-05, "loss": 0.086, "step": 2420 }, { "grad_norm": 0.7334204912185669, "learning_rate": 9.661353206701582e-05, "loss": 0.0782, "step": 2430 }, { "grad_norm": 0.6254956722259521, "learning_rate": 9.657354136116035e-05, "loss": 0.0777, "step": 2440 }, { "grad_norm": 0.9059262871742249, "learning_rate": 9.653332429039625e-05, "loss": 0.0798, "step": 2450 }, { "grad_norm": 0.70732182264328, "learning_rate": 9.649288105019356e-05, "loss": 0.0726, "step": 2460 }, { "grad_norm": 0.7314443588256836, "learning_rate": 9.645221183712165e-05, "loss": 0.0779, "step": 2470 }, { "grad_norm": 0.7098656296730042, "learning_rate": 9.641131684884817e-05, "loss": 0.0743, "step": 2480 }, { "grad_norm": 0.6610057950019836, "learning_rate": 9.637019628413813e-05, "loss": 0.082, "step": 2490 }, { "grad_norm": 0.5614057183265686, "learning_rate": 9.632885034285291e-05, "loss": 0.0767, "step": 2500 }, { "grad_norm": 0.6637583374977112, "learning_rate": 9.628727922594931e-05, "loss": 0.0733, "step": 2510 }, { "grad_norm": 0.595269501209259, "learning_rate": 9.624548313547862e-05, "loss": 0.0764, "step": 2520 }, { "grad_norm": 0.8634958863258362, "learning_rate": 9.620346227458547e-05, "loss": 0.0761, "step": 2530 }, { "grad_norm": 0.8040565252304077, "learning_rate": 9.616121684750712e-05, "loss": 0.0709, "step": 2540 }, { "grad_norm": 0.5635712742805481, "learning_rate": 9.611874705957215e-05, "loss": 0.079, "step": 2550 }, { "grad_norm": 0.6693586707115173, "learning_rate": 9.607605311719972e-05, "loss": 0.0805, "step": 2560 }, { "grad_norm": 0.6621934771537781, "learning_rate": 9.603313522789841e-05, "loss": 0.0706, "step": 2570 }, { "grad_norm": 0.662166953086853, "learning_rate": 9.598999360026529e-05, "loss": 0.0764, "step": 2580 }, { "grad_norm": 0.682948887348175, "learning_rate": 9.59466284439849e-05, "loss": 0.0671, "step": 2590 }, { "grad_norm": 0.6689608097076416, "learning_rate": 9.590303996982815e-05, "loss": 0.0762, "step": 2600 }, { "grad_norm": 0.6426261067390442, "learning_rate": 9.585922838965145e-05, "loss": 0.0774, "step": 2610 }, { "grad_norm": 0.6558887362480164, "learning_rate": 9.581519391639549e-05, "loss": 0.0759, "step": 2620 }, { "grad_norm": 0.6026521921157837, "learning_rate": 9.577093676408439e-05, "loss": 0.0701, "step": 2630 }, { "grad_norm": 0.6800580620765686, "learning_rate": 9.572645714782453e-05, "loss": 0.0748, "step": 2640 }, { "grad_norm": 0.6971288323402405, "learning_rate": 9.568175528380354e-05, "loss": 0.075, "step": 2650 }, { "grad_norm": 0.6518421769142151, "learning_rate": 9.56368313892893e-05, "loss": 0.0752, "step": 2660 }, { "grad_norm": 0.5340580344200134, "learning_rate": 9.55916856826288e-05, "loss": 0.073, "step": 2670 }, { "grad_norm": 0.7508995532989502, "learning_rate": 9.554631838324713e-05, "loss": 0.0757, "step": 2680 }, { "grad_norm": 0.5836358666419983, "learning_rate": 9.55007297116464e-05, "loss": 0.0743, "step": 2690 }, { "grad_norm": 0.7291355133056641, "learning_rate": 9.545491988940472e-05, "loss": 0.0726, "step": 2700 }, { "grad_norm": 0.7412474751472473, "learning_rate": 9.540888913917501e-05, "loss": 0.0692, "step": 2710 }, { "grad_norm": 0.5670576691627502, "learning_rate": 9.536263768468401e-05, "loss": 0.0744, "step": 2720 }, { "grad_norm": 0.8577015995979309, "learning_rate": 9.531616575073117e-05, "loss": 0.0793, "step": 2730 }, { "grad_norm": 0.6365728378295898, "learning_rate": 9.526947356318754e-05, "loss": 0.07, "step": 2740 }, { "grad_norm": 0.5782703757286072, "learning_rate": 9.52225613489947e-05, "loss": 0.0678, "step": 2750 }, { "grad_norm": 0.6387004852294922, "learning_rate": 9.517542933616365e-05, "loss": 0.0686, "step": 2760 }, { "grad_norm": 0.7190784215927124, "learning_rate": 9.512807775377366e-05, "loss": 0.0762, "step": 2770 }, { "grad_norm": 0.679921567440033, "learning_rate": 9.508050683197121e-05, "loss": 0.072, "step": 2780 }, { "grad_norm": 0.5193501114845276, "learning_rate": 9.503271680196888e-05, "loss": 0.0737, "step": 2790 }, { "grad_norm": 0.5788854956626892, "learning_rate": 9.498470789604413e-05, "loss": 0.0784, "step": 2800 }, { "grad_norm": 0.5486730933189392, "learning_rate": 9.49364803475383e-05, "loss": 0.0767, "step": 2810 }, { "grad_norm": 0.5417566299438477, "learning_rate": 9.48880343908554e-05, "loss": 0.0732, "step": 2820 }, { "grad_norm": 0.6178622245788574, "learning_rate": 9.4839370261461e-05, "loss": 0.0714, "step": 2830 }, { "grad_norm": 0.5790642499923706, "learning_rate": 9.479048819588098e-05, "loss": 0.0721, "step": 2840 }, { "grad_norm": 0.5813261270523071, "learning_rate": 9.474138843170063e-05, "loss": 0.0713, "step": 2850 }, { "grad_norm": 0.6615369915962219, "learning_rate": 9.46920712075632e-05, "loss": 0.0811, "step": 2860 }, { "grad_norm": 0.6264895796775818, "learning_rate": 9.464253676316893e-05, "loss": 0.0736, "step": 2870 }, { "grad_norm": 0.6997485160827637, "learning_rate": 9.459278533927384e-05, "loss": 0.0713, "step": 2880 }, { "grad_norm": 0.619152307510376, "learning_rate": 9.454281717768854e-05, "loss": 0.0741, "step": 2890 }, { "grad_norm": 0.6027862429618835, "learning_rate": 9.449263252127708e-05, "loss": 0.0664, "step": 2900 }, { "grad_norm": 0.8105010986328125, "learning_rate": 9.444223161395573e-05, "loss": 0.0671, "step": 2910 }, { "grad_norm": 0.6472406387329102, "learning_rate": 9.439161470069184e-05, "loss": 0.0756, "step": 2920 }, { "grad_norm": 0.6002207398414612, "learning_rate": 9.43407820275026e-05, "loss": 0.0767, "step": 2930 }, { "grad_norm": 0.644810140132904, "learning_rate": 9.428973384145396e-05, "loss": 0.0795, "step": 2940 }, { "grad_norm": 0.6397400498390198, "learning_rate": 9.423847039065922e-05, "loss": 0.0705, "step": 2950 }, { "grad_norm": 0.6125301718711853, "learning_rate": 9.418699192427805e-05, "loss": 0.0722, "step": 2960 }, { "grad_norm": 0.6119952201843262, "learning_rate": 9.41352986925151e-05, "loss": 0.069, "step": 2970 }, { "grad_norm": 0.619590163230896, "learning_rate": 9.408339094661895e-05, "loss": 0.0658, "step": 2980 }, { "grad_norm": 0.6107400059700012, "learning_rate": 9.40312689388807e-05, "loss": 0.0673, "step": 2990 }, { "grad_norm": 0.6754778623580933, "learning_rate": 9.397893292263292e-05, "loss": 0.0681, "step": 3000 }, { "grad_norm": 0.7630986571311951, "learning_rate": 9.392638315224829e-05, "loss": 0.0673, "step": 3010 }, { "grad_norm": 0.6023984551429749, "learning_rate": 9.387361988313846e-05, "loss": 0.0684, "step": 3020 }, { "grad_norm": 0.6075767874717712, "learning_rate": 9.38206433717527e-05, "loss": 0.0649, "step": 3030 }, { "grad_norm": 0.6851189136505127, "learning_rate": 9.376745387557681e-05, "loss": 0.0706, "step": 3040 }, { "grad_norm": 0.5601722598075867, "learning_rate": 9.371405165313169e-05, "loss": 0.0639, "step": 3050 }, { "grad_norm": 0.6507348418235779, "learning_rate": 9.366043696397222e-05, "loss": 0.0676, "step": 3060 }, { "grad_norm": 0.6262578964233398, "learning_rate": 9.360661006868592e-05, "loss": 0.0658, "step": 3070 }, { "grad_norm": 0.5408547520637512, "learning_rate": 9.355257122889173e-05, "loss": 0.0664, "step": 3080 }, { "grad_norm": 0.5795503854751587, "learning_rate": 9.349832070723871e-05, "loss": 0.0758, "step": 3090 }, { "grad_norm": 0.6226188540458679, "learning_rate": 9.34438587674048e-05, "loss": 0.0755, "step": 3100 }, { "grad_norm": 0.6497605443000793, "learning_rate": 9.338918567409545e-05, "loss": 0.0695, "step": 3110 }, { "grad_norm": 0.6185932755470276, "learning_rate": 9.333430169304247e-05, "loss": 0.0707, "step": 3120 }, { "grad_norm": 0.5427372455596924, "learning_rate": 9.327920709100259e-05, "loss": 0.0703, "step": 3130 }, { "grad_norm": 0.5989561080932617, "learning_rate": 9.322390213575631e-05, "loss": 0.0712, "step": 3140 }, { "grad_norm": 0.5449648499488831, "learning_rate": 9.316838709610648e-05, "loss": 0.0711, "step": 3150 }, { "grad_norm": 0.654105007648468, "learning_rate": 9.311266224187706e-05, "loss": 0.0672, "step": 3160 }, { "grad_norm": 0.5506020188331604, "learning_rate": 9.305672784391175e-05, "loss": 0.0672, "step": 3170 }, { "grad_norm": 0.503079354763031, "learning_rate": 9.300058417407276e-05, "loss": 0.0657, "step": 3180 }, { "grad_norm": 0.5641545653343201, "learning_rate": 9.29442315052394e-05, "loss": 0.0663, "step": 3190 }, { "grad_norm": 0.5445094108581543, "learning_rate": 9.288767011130684e-05, "loss": 0.0707, "step": 3200 }, { "grad_norm": 0.6540117263793945, "learning_rate": 9.283090026718466e-05, "loss": 0.068, "step": 3210 }, { "grad_norm": 0.7890759110450745, "learning_rate": 9.277392224879568e-05, "loss": 0.0637, "step": 3220 }, { "grad_norm": 0.5481688976287842, "learning_rate": 9.271673633307445e-05, "loss": 0.0618, "step": 3230 }, { "grad_norm": 0.5165878534317017, "learning_rate": 9.265934279796602e-05, "loss": 0.0686, "step": 3240 }, { "grad_norm": 0.6548216342926025, "learning_rate": 9.260174192242453e-05, "loss": 0.0701, "step": 3250 }, { "grad_norm": 0.8453028798103333, "learning_rate": 9.254393398641185e-05, "loss": 0.075, "step": 3260 }, { "grad_norm": 0.6304769515991211, "learning_rate": 9.248591927089628e-05, "loss": 0.0709, "step": 3270 }, { "grad_norm": 0.6732708215713501, "learning_rate": 9.242769805785115e-05, "loss": 0.0709, "step": 3280 }, { "grad_norm": 0.6626738905906677, "learning_rate": 9.236927063025342e-05, "loss": 0.0838, "step": 3290 }, { "grad_norm": 0.5738134384155273, "learning_rate": 9.231063727208234e-05, "loss": 0.0718, "step": 3300 }, { "grad_norm": 0.5651624202728271, "learning_rate": 9.225179826831807e-05, "loss": 0.0708, "step": 3310 }, { "grad_norm": 0.6571042537689209, "learning_rate": 9.219275390494024e-05, "loss": 0.0752, "step": 3320 }, { "grad_norm": 0.5358458757400513, "learning_rate": 9.213350446892668e-05, "loss": 0.0683, "step": 3330 }, { "grad_norm": 0.6220871806144714, "learning_rate": 9.207405024825186e-05, "loss": 0.0713, "step": 3340 }, { "grad_norm": 0.5692563056945801, "learning_rate": 9.201439153188569e-05, "loss": 0.0698, "step": 3350 }, { "grad_norm": 0.644374430179596, "learning_rate": 9.19545286097919e-05, "loss": 0.0682, "step": 3360 }, { "grad_norm": 0.5985931754112244, "learning_rate": 9.189446177292679e-05, "loss": 0.0744, "step": 3370 }, { "grad_norm": 0.6280204653739929, "learning_rate": 9.183419131323778e-05, "loss": 0.0667, "step": 3380 }, { "grad_norm": 0.6723644137382507, "learning_rate": 9.177371752366191e-05, "loss": 0.0648, "step": 3390 }, { "grad_norm": 0.6714949011802673, "learning_rate": 9.171304069812454e-05, "loss": 0.0767, "step": 3400 }, { "grad_norm": 0.7225590944290161, "learning_rate": 9.165216113153782e-05, "loss": 0.0644, "step": 3410 }, { "grad_norm": 0.5510388612747192, "learning_rate": 9.159107911979936e-05, "loss": 0.0637, "step": 3420 }, { "grad_norm": 0.47715097665786743, "learning_rate": 9.152979495979063e-05, "loss": 0.0694, "step": 3430 }, { "grad_norm": 0.6667302846908569, "learning_rate": 9.146830894937571e-05, "loss": 0.0739, "step": 3440 }, { "grad_norm": 0.5869659781455994, "learning_rate": 9.140662138739969e-05, "loss": 0.0708, "step": 3450 }, { "grad_norm": 0.5377545952796936, "learning_rate": 9.134473257368732e-05, "loss": 0.0744, "step": 3460 }, { "grad_norm": 0.6936116218566895, "learning_rate": 9.128264280904145e-05, "loss": 0.0606, "step": 3470 }, { "grad_norm": 0.5792559385299683, "learning_rate": 9.122035239524169e-05, "loss": 0.0711, "step": 3480 }, { "grad_norm": 0.5261418223381042, "learning_rate": 9.115786163504285e-05, "loss": 0.0628, "step": 3490 }, { "grad_norm": 0.6396857500076294, "learning_rate": 9.10951708321735e-05, "loss": 0.0648, "step": 3500 }, { "grad_norm": 0.5282208919525146, "learning_rate": 9.10322802913345e-05, "loss": 0.0645, "step": 3510 }, { "grad_norm": 0.7117626667022705, "learning_rate": 9.096919031819751e-05, "loss": 0.0702, "step": 3520 }, { "grad_norm": 0.6305633187294006, "learning_rate": 9.090590121940348e-05, "loss": 0.0706, "step": 3530 }, { "grad_norm": 0.5667370557785034, "learning_rate": 9.084241330256121e-05, "loss": 0.0639, "step": 3540 }, { "grad_norm": 0.608974277973175, "learning_rate": 9.077872687624586e-05, "loss": 0.0647, "step": 3550 }, { "grad_norm": 0.6044687032699585, "learning_rate": 9.071484224999735e-05, "loss": 0.067, "step": 3560 }, { "grad_norm": 0.5174039006233215, "learning_rate": 9.0650759734319e-05, "loss": 0.0661, "step": 3570 }, { "grad_norm": 0.6718793511390686, "learning_rate": 9.05864796406759e-05, "loss": 0.0697, "step": 3580 }, { "grad_norm": 0.6483207941055298, "learning_rate": 9.052200228149343e-05, "loss": 0.0715, "step": 3590 }, { "grad_norm": 0.5423234105110168, "learning_rate": 9.04573279701558e-05, "loss": 0.0719, "step": 3600 }, { "grad_norm": 0.5089258551597595, "learning_rate": 9.039245702100448e-05, "loss": 0.064, "step": 3610 }, { "grad_norm": 0.6893640756607056, "learning_rate": 9.032738974933664e-05, "loss": 0.0638, "step": 3620 }, { "grad_norm": 0.8346790671348572, "learning_rate": 9.026212647140365e-05, "loss": 0.0648, "step": 3630 }, { "grad_norm": 0.49496138095855713, "learning_rate": 9.019666750440956e-05, "loss": 0.0652, "step": 3640 }, { "grad_norm": 0.5753434300422668, "learning_rate": 9.013101316650956e-05, "loss": 0.069, "step": 3650 }, { "grad_norm": 0.4407230615615845, "learning_rate": 9.00651637768084e-05, "loss": 0.0636, "step": 3660 }, { "grad_norm": 0.4615444839000702, "learning_rate": 8.999911965535885e-05, "loss": 0.0684, "step": 3670 }, { "grad_norm": 0.4918903708457947, "learning_rate": 8.993288112316014e-05, "loss": 0.0621, "step": 3680 }, { "grad_norm": 0.5506250858306885, "learning_rate": 8.986644850215644e-05, "loss": 0.0671, "step": 3690 }, { "grad_norm": 0.5882449150085449, "learning_rate": 8.979982211523523e-05, "loss": 0.0665, "step": 3700 }, { "grad_norm": 0.5929926633834839, "learning_rate": 8.97330022862258e-05, "loss": 0.0658, "step": 3710 }, { "grad_norm": 0.5656607747077942, "learning_rate": 8.96659893398976e-05, "loss": 0.063, "step": 3720 }, { "grad_norm": 0.5676708817481995, "learning_rate": 8.959878360195876e-05, "loss": 0.0658, "step": 3730 }, { "grad_norm": 0.5738428235054016, "learning_rate": 8.953138539905438e-05, "loss": 0.0648, "step": 3740 }, { "grad_norm": 0.5268324613571167, "learning_rate": 8.946379505876506e-05, "loss": 0.0695, "step": 3750 }, { "grad_norm": 0.7279912233352661, "learning_rate": 8.939601290960527e-05, "loss": 0.0662, "step": 3760 }, { "grad_norm": 0.5215412974357605, "learning_rate": 8.932803928102167e-05, "loss": 0.0679, "step": 3770 }, { "grad_norm": 0.6495506763458252, "learning_rate": 8.925987450339168e-05, "loss": 0.0636, "step": 3780 }, { "grad_norm": 0.6008911728858948, "learning_rate": 8.919151890802172e-05, "loss": 0.0677, "step": 3790 }, { "grad_norm": 0.6184180378913879, "learning_rate": 8.912297282714564e-05, "loss": 0.0704, "step": 3800 }, { "grad_norm": 0.5512374639511108, "learning_rate": 8.905423659392316e-05, "loss": 0.0607, "step": 3810 }, { "grad_norm": 0.5496158599853516, "learning_rate": 8.898531054243822e-05, "loss": 0.0645, "step": 3820 }, { "grad_norm": 0.5281623005867004, "learning_rate": 8.891619500769729e-05, "loss": 0.0663, "step": 3830 }, { "grad_norm": 0.5163659453392029, "learning_rate": 8.884689032562785e-05, "loss": 0.0631, "step": 3840 }, { "grad_norm": 0.61015385389328, "learning_rate": 8.87773968330767e-05, "loss": 0.0621, "step": 3850 }, { "grad_norm": 0.5940917134284973, "learning_rate": 8.870771486780832e-05, "loss": 0.0654, "step": 3860 }, { "grad_norm": 0.4781402051448822, "learning_rate": 8.863784476850322e-05, "loss": 0.0656, "step": 3870 }, { "grad_norm": 0.5629350543022156, "learning_rate": 8.856778687475635e-05, "loss": 0.0631, "step": 3880 }, { "grad_norm": 0.49277231097221375, "learning_rate": 8.849754152707541e-05, "loss": 0.0608, "step": 3890 }, { "grad_norm": 0.4736921191215515, "learning_rate": 8.842710906687916e-05, "loss": 0.0597, "step": 3900 }, { "grad_norm": 0.5930362939834595, "learning_rate": 8.83564898364958e-05, "loss": 0.0694, "step": 3910 }, { "grad_norm": 0.609360933303833, "learning_rate": 8.828568417916136e-05, "loss": 0.072, "step": 3920 }, { "grad_norm": 0.5193405151367188, "learning_rate": 8.821469243901794e-05, "loss": 0.0603, "step": 3930 }, { "grad_norm": 0.6480633020401001, "learning_rate": 8.814351496111201e-05, "loss": 0.0652, "step": 3940 }, { "grad_norm": 0.5353022217750549, "learning_rate": 8.807215209139293e-05, "loss": 0.0618, "step": 3950 }, { "grad_norm": 0.6103038787841797, "learning_rate": 8.8000604176711e-05, "loss": 0.0668, "step": 3960 }, { "grad_norm": 0.5791677832603455, "learning_rate": 8.792887156481598e-05, "loss": 0.0667, "step": 3970 }, { "grad_norm": 0.5314576029777527, "learning_rate": 8.785695460435534e-05, "loss": 0.0739, "step": 3980 }, { "grad_norm": 0.6145054697990417, "learning_rate": 8.778485364487248e-05, "loss": 0.0719, "step": 3990 }, { "grad_norm": 0.528359055519104, "learning_rate": 8.771256903680519e-05, "loss": 0.0662, "step": 4000 }, { "grad_norm": 0.45334354043006897, "learning_rate": 8.764010113148382e-05, "loss": 0.0672, "step": 4010 }, { "grad_norm": 0.543138861656189, "learning_rate": 8.756745028112959e-05, "loss": 0.0631, "step": 4020 }, { "grad_norm": 0.5311511158943176, "learning_rate": 8.749461683885296e-05, "loss": 0.0619, "step": 4030 }, { "grad_norm": 0.5041663646697998, "learning_rate": 8.742160115865179e-05, "loss": 0.0615, "step": 4040 }, { "grad_norm": 0.44580286741256714, "learning_rate": 8.734840359540974e-05, "loss": 0.0559, "step": 4050 }, { "grad_norm": 0.5962316393852234, "learning_rate": 8.727502450489446e-05, "loss": 0.063, "step": 4060 }, { "grad_norm": 0.5856122374534607, "learning_rate": 8.720146424375591e-05, "loss": 0.0605, "step": 4070 }, { "grad_norm": 0.564177930355072, "learning_rate": 8.712772316952458e-05, "loss": 0.0656, "step": 4080 }, { "grad_norm": 0.5987861156463623, "learning_rate": 8.705380164060982e-05, "loss": 0.0637, "step": 4090 }, { "grad_norm": 0.5117592811584473, "learning_rate": 8.697970001629799e-05, "loss": 0.0668, "step": 4100 }, { "grad_norm": 0.5237540602684021, "learning_rate": 8.690541865675084e-05, "loss": 0.0656, "step": 4110 }, { "grad_norm": 0.6092665791511536, "learning_rate": 8.68309579230037e-05, "loss": 0.0647, "step": 4120 }, { "grad_norm": 0.5419392585754395, "learning_rate": 8.675631817696372e-05, "loss": 0.0611, "step": 4130 }, { "grad_norm": 0.4946824014186859, "learning_rate": 8.668149978140808e-05, "loss": 0.0706, "step": 4140 }, { "grad_norm": 0.6564487814903259, "learning_rate": 8.66065030999823e-05, "loss": 0.0596, "step": 4150 }, { "grad_norm": 0.497086763381958, "learning_rate": 8.653132849719845e-05, "loss": 0.0663, "step": 4160 }, { "grad_norm": 0.5801936388015747, "learning_rate": 8.64559763384333e-05, "loss": 0.0622, "step": 4170 }, { "grad_norm": 0.5763471722602844, "learning_rate": 8.638044698992669e-05, "loss": 0.0652, "step": 4180 }, { "grad_norm": 0.5335597395896912, "learning_rate": 8.630474081877959e-05, "loss": 0.0583, "step": 4190 }, { "grad_norm": 0.5130449533462524, "learning_rate": 8.62288581929525e-05, "loss": 0.059, "step": 4200 }, { "grad_norm": 0.6195492744445801, "learning_rate": 8.615279948126343e-05, "loss": 0.0621, "step": 4210 }, { "grad_norm": 0.5328609943389893, "learning_rate": 8.60765650533863e-05, "loss": 0.0595, "step": 4220 }, { "grad_norm": 0.6657204031944275, "learning_rate": 8.60001552798491e-05, "loss": 0.0592, "step": 4230 }, { "grad_norm": 0.5727778077125549, "learning_rate": 8.592357053203202e-05, "loss": 0.0649, "step": 4240 }, { "grad_norm": 0.5315781235694885, "learning_rate": 8.58468111821657e-05, "loss": 0.0662, "step": 4250 }, { "grad_norm": 0.5774106979370117, "learning_rate": 8.576987760332943e-05, "loss": 0.0666, "step": 4260 }, { "grad_norm": 0.5133674144744873, "learning_rate": 8.56927701694493e-05, "loss": 0.0599, "step": 4270 }, { "grad_norm": 0.5090412497520447, "learning_rate": 8.561548925529643e-05, "loss": 0.064, "step": 4280 }, { "grad_norm": 0.5547593235969543, "learning_rate": 8.553803523648506e-05, "loss": 0.0617, "step": 4290 }, { "grad_norm": 0.5150351524353027, "learning_rate": 8.546040848947086e-05, "loss": 0.0643, "step": 4300 }, { "grad_norm": 0.5246621370315552, "learning_rate": 8.538260939154894e-05, "loss": 0.0639, "step": 4310 }, { "grad_norm": 0.5461138486862183, "learning_rate": 8.530463832085218e-05, "loss": 0.0582, "step": 4320 }, { "grad_norm": 0.49885794520378113, "learning_rate": 8.522649565634927e-05, "loss": 0.0618, "step": 4330 }, { "grad_norm": 0.48420804738998413, "learning_rate": 8.51481817778429e-05, "loss": 0.0607, "step": 4340 }, { "grad_norm": 0.5489194393157959, "learning_rate": 8.506969706596797e-05, "loss": 0.0576, "step": 4350 }, { "grad_norm": 0.6276731491088867, "learning_rate": 8.499104190218964e-05, "loss": 0.057, "step": 4360 }, { "grad_norm": 0.4996776878833771, "learning_rate": 8.49122166688016e-05, "loss": 0.0626, "step": 4370 }, { "grad_norm": 0.49057215452194214, "learning_rate": 8.483322174892404e-05, "loss": 0.0658, "step": 4380 }, { "grad_norm": 0.491436630487442, "learning_rate": 8.475405752650199e-05, "loss": 0.0614, "step": 4390 }, { "grad_norm": 0.49525755643844604, "learning_rate": 8.467472438630328e-05, "loss": 0.0658, "step": 4400 }, { "grad_norm": 0.4709268808364868, "learning_rate": 8.459522271391682e-05, "loss": 0.0647, "step": 4410 }, { "grad_norm": 0.5747040510177612, "learning_rate": 8.451555289575057e-05, "loss": 0.0582, "step": 4420 }, { "grad_norm": 0.5364845395088196, "learning_rate": 8.443571531902981e-05, "loss": 0.0668, "step": 4430 }, { "grad_norm": 0.4799308180809021, "learning_rate": 8.435571037179512e-05, "loss": 0.0609, "step": 4440 }, { "grad_norm": 0.525842010974884, "learning_rate": 8.427553844290062e-05, "loss": 0.0624, "step": 4450 }, { "grad_norm": 0.6582574844360352, "learning_rate": 8.419519992201201e-05, "loss": 0.0607, "step": 4460 }, { "grad_norm": 0.4682302176952362, "learning_rate": 8.411469519960469e-05, "loss": 0.0599, "step": 4470 }, { "grad_norm": 0.6320589184761047, "learning_rate": 8.403402466696182e-05, "loss": 0.0641, "step": 4480 }, { "grad_norm": 0.6417620778083801, "learning_rate": 8.395318871617255e-05, "loss": 0.0569, "step": 4490 }, { "grad_norm": 0.6075902581214905, "learning_rate": 8.387218774012992e-05, "loss": 0.0666, "step": 4500 }, { "grad_norm": 0.4782496392726898, "learning_rate": 8.379102213252915e-05, "loss": 0.0556, "step": 4510 }, { "grad_norm": 0.5786132216453552, "learning_rate": 8.370969228786556e-05, "loss": 0.0584, "step": 4520 }, { "grad_norm": 0.502007007598877, "learning_rate": 8.362819860143275e-05, "loss": 0.0592, "step": 4530 }, { "grad_norm": 0.5026768445968628, "learning_rate": 8.354654146932066e-05, "loss": 0.0606, "step": 4540 }, { "grad_norm": 0.6542195677757263, "learning_rate": 8.346472128841364e-05, "loss": 0.0625, "step": 4550 }, { "grad_norm": 0.4475997984409332, "learning_rate": 8.338273845638848e-05, "loss": 0.0655, "step": 4560 }, { "grad_norm": 0.5038512349128723, "learning_rate": 8.330059337171258e-05, "loss": 0.0561, "step": 4570 }, { "grad_norm": 0.5571680068969727, "learning_rate": 8.32182864336419e-05, "loss": 0.064, "step": 4580 }, { "grad_norm": 0.5078977942466736, "learning_rate": 8.313581804221908e-05, "loss": 0.0656, "step": 4590 }, { "grad_norm": 0.5355502963066101, "learning_rate": 8.305318859827147e-05, "loss": 0.0586, "step": 4600 }, { "grad_norm": 0.6439716219902039, "learning_rate": 8.297039850340923e-05, "loss": 0.0638, "step": 4610 }, { "grad_norm": 0.4819484353065491, "learning_rate": 8.288744816002331e-05, "loss": 0.0602, "step": 4620 }, { "grad_norm": 0.5139064788818359, "learning_rate": 8.280433797128357e-05, "loss": 0.0597, "step": 4630 }, { "grad_norm": 0.523800253868103, "learning_rate": 8.272106834113674e-05, "loss": 0.0567, "step": 4640 }, { "grad_norm": 0.5478535890579224, "learning_rate": 8.26376396743045e-05, "loss": 0.0583, "step": 4650 }, { "grad_norm": 0.4601856768131256, "learning_rate": 8.25540523762815e-05, "loss": 0.0567, "step": 4660 }, { "grad_norm": 0.5561063885688782, "learning_rate": 8.247030685333346e-05, "loss": 0.0623, "step": 4670 }, { "grad_norm": 0.5215331315994263, "learning_rate": 8.238640351249503e-05, "loss": 0.0597, "step": 4680 }, { "grad_norm": 0.49608728289604187, "learning_rate": 8.2302342761568e-05, "loss": 0.0649, "step": 4690 }, { "grad_norm": 0.5840057134628296, "learning_rate": 8.221812500911919e-05, "loss": 0.0582, "step": 4700 }, { "grad_norm": 0.485944926738739, "learning_rate": 8.213375066447853e-05, "loss": 0.0588, "step": 4710 }, { "grad_norm": 0.5371300578117371, "learning_rate": 8.204922013773702e-05, "loss": 0.0644, "step": 4720 }, { "grad_norm": 0.5104807615280151, "learning_rate": 8.196453383974478e-05, "loss": 0.0588, "step": 4730 }, { "grad_norm": 0.47362270951271057, "learning_rate": 8.187969218210904e-05, "loss": 0.0596, "step": 4740 }, { "grad_norm": 0.45317813754081726, "learning_rate": 8.179469557719213e-05, "loss": 0.0633, "step": 4750 }, { "grad_norm": 0.5414328575134277, "learning_rate": 8.170954443810948e-05, "loss": 0.0586, "step": 4760 }, { "grad_norm": 0.4773976504802704, "learning_rate": 8.162423917872764e-05, "loss": 0.0632, "step": 4770 }, { "grad_norm": 0.4718090891838074, "learning_rate": 8.153878021366217e-05, "loss": 0.0639, "step": 4780 }, { "grad_norm": 0.5311549305915833, "learning_rate": 8.14531679582758e-05, "loss": 0.0648, "step": 4790 }, { "grad_norm": 0.6562144756317139, "learning_rate": 8.136740282867621e-05, "loss": 0.0566, "step": 4800 }, { "grad_norm": 0.5640397667884827, "learning_rate": 8.128148524171418e-05, "loss": 0.0696, "step": 4810 }, { "grad_norm": 0.5281029343605042, "learning_rate": 8.119541561498146e-05, "loss": 0.0597, "step": 4820 }, { "grad_norm": 0.5679964423179626, "learning_rate": 8.110919436680877e-05, "loss": 0.0605, "step": 4830 }, { "grad_norm": 0.5190762281417847, "learning_rate": 8.102282191626378e-05, "loss": 0.0606, "step": 4840 }, { "grad_norm": 0.5041296482086182, "learning_rate": 8.0936298683149e-05, "loss": 0.056, "step": 4850 }, { "grad_norm": 0.49754947423934937, "learning_rate": 8.084962508799991e-05, "loss": 0.0582, "step": 4860 }, { "grad_norm": 0.45711421966552734, "learning_rate": 8.076280155208273e-05, "loss": 0.0602, "step": 4870 }, { "grad_norm": 0.5648583173751831, "learning_rate": 8.067582849739245e-05, "loss": 0.0578, "step": 4880 }, { "grad_norm": 0.5352514982223511, "learning_rate": 8.058870634665079e-05, "loss": 0.056, "step": 4890 }, { "grad_norm": 0.5403103828430176, "learning_rate": 8.050143552330414e-05, "loss": 0.0566, "step": 4900 }, { "grad_norm": 0.6360242962837219, "learning_rate": 8.041401645152151e-05, "loss": 0.057, "step": 4910 }, { "grad_norm": 0.6194398403167725, "learning_rate": 8.032644955619239e-05, "loss": 0.0674, "step": 4920 }, { "grad_norm": 0.5398883819580078, "learning_rate": 8.023873526292483e-05, "loss": 0.0675, "step": 4930 }, { "grad_norm": 0.44168609380722046, "learning_rate": 8.015087399804322e-05, "loss": 0.0566, "step": 4940 }, { "grad_norm": 0.42431899905204773, "learning_rate": 8.006286618858635e-05, "loss": 0.0656, "step": 4950 }, { "grad_norm": 0.4777362048625946, "learning_rate": 7.99747122623052e-05, "loss": 0.0591, "step": 4960 }, { "grad_norm": 0.5099514722824097, "learning_rate": 7.988641264766097e-05, "loss": 0.0593, "step": 4970 }, { "grad_norm": 0.5469658374786377, "learning_rate": 7.9797967773823e-05, "loss": 0.0621, "step": 4980 }, { "grad_norm": 0.3929344117641449, "learning_rate": 7.970937807066659e-05, "loss": 0.0588, "step": 4990 }, { "grad_norm": 0.433919221162796, "learning_rate": 7.962064396877098e-05, "loss": 0.0615, "step": 5000 }, { "grad_norm": 0.45964470505714417, "learning_rate": 7.953176589941722e-05, "loss": 0.0603, "step": 5010 }, { "grad_norm": 0.4343070089817047, "learning_rate": 7.944274429458614e-05, "loss": 0.0637, "step": 5020 }, { "grad_norm": 0.4875974953174591, "learning_rate": 7.93535795869562e-05, "loss": 0.059, "step": 5030 }, { "grad_norm": 0.4838794767856598, "learning_rate": 7.926427220990134e-05, "loss": 0.0559, "step": 5040 }, { "grad_norm": 0.44954898953437805, "learning_rate": 7.9174822597489e-05, "loss": 0.0572, "step": 5050 }, { "grad_norm": 0.47687119245529175, "learning_rate": 7.908523118447789e-05, "loss": 0.0563, "step": 5060 }, { "grad_norm": 0.5028277635574341, "learning_rate": 7.89954984063159e-05, "loss": 0.0535, "step": 5070 }, { "grad_norm": 0.6060270667076111, "learning_rate": 7.890562469913811e-05, "loss": 0.056, "step": 5080 }, { "grad_norm": 0.48532041907310486, "learning_rate": 7.881561049976447e-05, "loss": 0.0592, "step": 5090 }, { "grad_norm": 0.5645678043365479, "learning_rate": 7.872545624569779e-05, "loss": 0.0606, "step": 5100 }, { "grad_norm": 0.5222822427749634, "learning_rate": 7.863516237512164e-05, "loss": 0.0584, "step": 5110 }, { "grad_norm": 0.4604170024394989, "learning_rate": 7.854472932689815e-05, "loss": 0.0577, "step": 5120 }, { "grad_norm": 0.44878771901130676, "learning_rate": 7.845415754056591e-05, "loss": 0.0559, "step": 5130 }, { "grad_norm": 0.4780261516571045, "learning_rate": 7.836344745633783e-05, "loss": 0.0642, "step": 5140 }, { "grad_norm": 0.4232221245765686, "learning_rate": 7.8272599515099e-05, "loss": 0.0594, "step": 5150 }, { "grad_norm": 0.556639552116394, "learning_rate": 7.818161415840453e-05, "loss": 0.0591, "step": 5160 }, { "grad_norm": 0.508721113204956, "learning_rate": 7.809049182847745e-05, "loss": 0.056, "step": 5170 }, { "grad_norm": 0.5749722123146057, "learning_rate": 7.799923296820653e-05, "loss": 0.0538, "step": 5180 }, { "grad_norm": 0.47284308075904846, "learning_rate": 7.790783802114408e-05, "loss": 0.0604, "step": 5190 }, { "grad_norm": 0.5093716382980347, "learning_rate": 7.781630743150392e-05, "loss": 0.057, "step": 5200 }, { "grad_norm": 0.5395658612251282, "learning_rate": 7.772464164415907e-05, "loss": 0.0585, "step": 5210 }, { "grad_norm": 0.5696783065795898, "learning_rate": 7.763284110463973e-05, "loss": 0.061, "step": 5220 }, { "grad_norm": 0.46128687262535095, "learning_rate": 7.754090625913099e-05, "loss": 0.0577, "step": 5230 }, { "grad_norm": 0.4267308712005615, "learning_rate": 7.744883755447075e-05, "loss": 0.0565, "step": 5240 }, { "grad_norm": 0.5390599966049194, "learning_rate": 7.735663543814749e-05, "loss": 0.0536, "step": 5250 }, { "grad_norm": 0.5402194261550903, "learning_rate": 7.726430035829813e-05, "loss": 0.0609, "step": 5260 }, { "grad_norm": 0.5593826174736023, "learning_rate": 7.717183276370586e-05, "loss": 0.0604, "step": 5270 }, { "grad_norm": 0.48568156361579895, "learning_rate": 7.707923310379794e-05, "loss": 0.056, "step": 5280 }, { "grad_norm": 0.5259646773338318, "learning_rate": 7.698650182864351e-05, "loss": 0.0549, "step": 5290 }, { "grad_norm": 0.4322700798511505, "learning_rate": 7.689363938895138e-05, "loss": 0.0575, "step": 5300 }, { "grad_norm": 0.46357619762420654, "learning_rate": 7.680064623606791e-05, "loss": 0.0535, "step": 5310 }, { "grad_norm": 0.4487699270248413, "learning_rate": 7.670752282197476e-05, "loss": 0.0557, "step": 5320 }, { "grad_norm": 0.5622449517250061, "learning_rate": 7.66142695992867e-05, "loss": 0.0521, "step": 5330 }, { "grad_norm": 0.5403662919998169, "learning_rate": 7.652088702124944e-05, "loss": 0.0548, "step": 5340 }, { "grad_norm": 0.46101441979408264, "learning_rate": 7.64273755417374e-05, "loss": 0.0547, "step": 5350 }, { "grad_norm": 0.48611223697662354, "learning_rate": 7.633373561525148e-05, "loss": 0.0553, "step": 5360 }, { "grad_norm": 0.5463878512382507, "learning_rate": 7.623996769691691e-05, "loss": 0.0547, "step": 5370 }, { "grad_norm": 0.5002387762069702, "learning_rate": 7.614607224248103e-05, "loss": 0.0575, "step": 5380 }, { "grad_norm": 0.4558352530002594, "learning_rate": 7.605204970831096e-05, "loss": 0.0563, "step": 5390 }, { "grad_norm": 0.5165241360664368, "learning_rate": 7.595790055139163e-05, "loss": 0.0544, "step": 5400 }, { "grad_norm": 0.49622392654418945, "learning_rate": 7.586362522932323e-05, "loss": 0.0556, "step": 5410 }, { "grad_norm": 0.5313038229942322, "learning_rate": 7.576922420031929e-05, "loss": 0.0596, "step": 5420 }, { "grad_norm": 0.5192410945892334, "learning_rate": 7.567469792320428e-05, "loss": 0.0536, "step": 5430 }, { "grad_norm": 0.44219714403152466, "learning_rate": 7.558004685741137e-05, "loss": 0.0541, "step": 5440 }, { "grad_norm": 0.552193284034729, "learning_rate": 7.548527146298036e-05, "loss": 0.0525, "step": 5450 }, { "grad_norm": 0.4243243336677551, "learning_rate": 7.539037220055527e-05, "loss": 0.0527, "step": 5460 }, { "grad_norm": 0.5068473219871521, "learning_rate": 7.529534953138213e-05, "loss": 0.058, "step": 5470 }, { "grad_norm": 0.46479758620262146, "learning_rate": 7.520020391730684e-05, "loss": 0.0563, "step": 5480 }, { "grad_norm": 0.4513914883136749, "learning_rate": 7.510493582077281e-05, "loss": 0.0553, "step": 5490 }, { "grad_norm": 0.4353752136230469, "learning_rate": 7.500954570481882e-05, "loss": 0.063, "step": 5500 }, { "grad_norm": 0.4668118953704834, "learning_rate": 7.491403403307662e-05, "loss": 0.0549, "step": 5510 }, { "grad_norm": 0.40962520241737366, "learning_rate": 7.481840126976885e-05, "loss": 0.0556, "step": 5520 }, { "grad_norm": 0.4395747184753418, "learning_rate": 7.472264787970666e-05, "loss": 0.0503, "step": 5530 }, { "grad_norm": 0.5224431157112122, "learning_rate": 7.462677432828751e-05, "loss": 0.058, "step": 5540 }, { "grad_norm": 0.5443442463874817, "learning_rate": 7.453078108149287e-05, "loss": 0.057, "step": 5550 }, { "grad_norm": 0.4638959765434265, "learning_rate": 7.443466860588599e-05, "loss": 0.0592, "step": 5560 }, { "grad_norm": 0.5321382284164429, "learning_rate": 7.43384373686096e-05, "loss": 0.0546, "step": 5570 }, { "grad_norm": 0.44946324825286865, "learning_rate": 7.424208783738367e-05, "loss": 0.0622, "step": 5580 }, { "grad_norm": 0.5622340440750122, "learning_rate": 7.414562048050315e-05, "loss": 0.0574, "step": 5590 }, { "grad_norm": 0.44878771901130676, "learning_rate": 7.404903576683559e-05, "loss": 0.0621, "step": 5600 }, { "grad_norm": 0.4351768493652344, "learning_rate": 7.3952334165819e-05, "loss": 0.0587, "step": 5610 }, { "grad_norm": 0.48724520206451416, "learning_rate": 7.385551614745952e-05, "loss": 0.0564, "step": 5620 }, { "grad_norm": 0.41412195563316345, "learning_rate": 7.375858218232905e-05, "loss": 0.0596, "step": 5630 }, { "grad_norm": 0.4886058270931244, "learning_rate": 7.366153274156312e-05, "loss": 0.0542, "step": 5640 }, { "grad_norm": 0.5508632063865662, "learning_rate": 7.356436829685844e-05, "loss": 0.0594, "step": 5650 }, { "grad_norm": 0.48630625009536743, "learning_rate": 7.346708932047074e-05, "loss": 0.0533, "step": 5660 }, { "grad_norm": 0.44918766617774963, "learning_rate": 7.336969628521237e-05, "loss": 0.0618, "step": 5670 }, { "grad_norm": 0.47114333510398865, "learning_rate": 7.32721896644501e-05, "loss": 0.0531, "step": 5680 }, { "grad_norm": 0.48327508568763733, "learning_rate": 7.317456993210272e-05, "loss": 0.0566, "step": 5690 }, { "grad_norm": 0.40669572353363037, "learning_rate": 7.307683756263881e-05, "loss": 0.0528, "step": 5700 }, { "grad_norm": 0.5093392729759216, "learning_rate": 7.297899303107441e-05, "loss": 0.0596, "step": 5710 }, { "grad_norm": 0.3939994275569916, "learning_rate": 7.288103681297068e-05, "loss": 0.0538, "step": 5720 }, { "grad_norm": 0.48547250032424927, "learning_rate": 7.278296938443166e-05, "loss": 0.0639, "step": 5730 }, { "grad_norm": 0.5965085029602051, "learning_rate": 7.26847912221019e-05, "loss": 0.057, "step": 5740 }, { "grad_norm": 0.5538206696510315, "learning_rate": 7.258650280316415e-05, "loss": 0.054, "step": 5750 }, { "grad_norm": 0.47953662276268005, "learning_rate": 7.248810460533706e-05, "loss": 0.0527, "step": 5760 }, { "grad_norm": 0.40855997800827026, "learning_rate": 7.238959710687282e-05, "loss": 0.0546, "step": 5770 }, { "grad_norm": 0.49721816182136536, "learning_rate": 7.229098078655489e-05, "loss": 0.0581, "step": 5780 }, { "grad_norm": 0.4714704155921936, "learning_rate": 7.219225612369565e-05, "loss": 0.0523, "step": 5790 }, { "grad_norm": 0.37301206588745117, "learning_rate": 7.209342359813404e-05, "loss": 0.0558, "step": 5800 }, { "grad_norm": 0.49109941720962524, "learning_rate": 7.199448369023327e-05, "loss": 0.0568, "step": 5810 }, { "grad_norm": 0.5443580150604248, "learning_rate": 7.189543688087845e-05, "loss": 0.0523, "step": 5820 }, { "grad_norm": 0.5693491697311401, "learning_rate": 7.17962836514743e-05, "loss": 0.0525, "step": 5830 }, { "grad_norm": 0.609158992767334, "learning_rate": 7.169702448394279e-05, "loss": 0.0536, "step": 5840 }, { "grad_norm": 0.5174346566200256, "learning_rate": 7.159765986072071e-05, "loss": 0.0555, "step": 5850 }, { "grad_norm": 0.5700979828834534, "learning_rate": 7.149819026475751e-05, "loss": 0.059, "step": 5860 }, { "grad_norm": 0.47033268213272095, "learning_rate": 7.139861617951275e-05, "loss": 0.0541, "step": 5870 }, { "grad_norm": 0.45670151710510254, "learning_rate": 7.129893808895395e-05, "loss": 0.056, "step": 5880 }, { "grad_norm": 0.41775184869766235, "learning_rate": 7.119915647755404e-05, "loss": 0.0504, "step": 5890 }, { "grad_norm": 0.43254759907722473, "learning_rate": 7.109927183028914e-05, "loss": 0.054, "step": 5900 }, { "grad_norm": 0.4849887192249298, "learning_rate": 7.099928463263619e-05, "loss": 0.0561, "step": 5910 }, { "grad_norm": 0.4883417785167694, "learning_rate": 7.08991953705705e-05, "loss": 0.0554, "step": 5920 }, { "grad_norm": 0.4409164488315582, "learning_rate": 7.07990045305635e-05, "loss": 0.0587, "step": 5930 }, { "grad_norm": 0.5324891209602356, "learning_rate": 7.069871259958034e-05, "loss": 0.0629, "step": 5940 }, { "grad_norm": 0.41858968138694763, "learning_rate": 7.059832006507745e-05, "loss": 0.0579, "step": 5950 }, { "grad_norm": 0.490377813577652, "learning_rate": 7.049782741500028e-05, "loss": 0.054, "step": 5960 }, { "grad_norm": 0.4170517921447754, "learning_rate": 7.039723513778087e-05, "loss": 0.0519, "step": 5970 }, { "grad_norm": 0.4487491548061371, "learning_rate": 7.029654372233544e-05, "loss": 0.0543, "step": 5980 }, { "grad_norm": 0.4388402998447418, "learning_rate": 7.019575365806215e-05, "loss": 0.0563, "step": 5990 }, { "grad_norm": 0.43077412247657776, "learning_rate": 7.009486543483858e-05, "loss": 0.0536, "step": 6000 }, { "grad_norm": 0.48064032196998596, "learning_rate": 6.999387954301934e-05, "loss": 0.0594, "step": 6010 }, { "grad_norm": 0.6669104099273682, "learning_rate": 6.989279647343388e-05, "loss": 0.0547, "step": 6020 }, { "grad_norm": 0.4026600122451782, "learning_rate": 6.979161671738382e-05, "loss": 0.0607, "step": 6030 }, { "grad_norm": 0.499647319316864, "learning_rate": 6.969034076664085e-05, "loss": 0.055, "step": 6040 }, { "grad_norm": 0.5666208863258362, "learning_rate": 6.958896911344411e-05, "loss": 0.0526, "step": 6050 }, { "grad_norm": 0.4235285818576813, "learning_rate": 6.948750225049791e-05, "loss": 0.0541, "step": 6060 }, { "grad_norm": 0.49159005284309387, "learning_rate": 6.938594067096936e-05, "loss": 0.0509, "step": 6070 }, { "grad_norm": 0.4927906394004822, "learning_rate": 6.928428486848587e-05, "loss": 0.0588, "step": 6080 }, { "grad_norm": 0.5298722982406616, "learning_rate": 6.918253533713282e-05, "loss": 0.055, "step": 6090 }, { "grad_norm": 0.5602335929870605, "learning_rate": 6.908069257145118e-05, "loss": 0.0541, "step": 6100 }, { "grad_norm": 0.5721182823181152, "learning_rate": 6.897875706643506e-05, "loss": 0.0566, "step": 6110 }, { "grad_norm": 0.5546931028366089, "learning_rate": 6.887672931752927e-05, "loss": 0.0572, "step": 6120 }, { "grad_norm": 0.5690977573394775, "learning_rate": 6.877460982062706e-05, "loss": 0.0548, "step": 6130 }, { "grad_norm": 0.5065085887908936, "learning_rate": 6.86723990720675e-05, "loss": 0.0543, "step": 6140 }, { "grad_norm": 0.5057834982872009, "learning_rate": 6.857009756863326e-05, "loss": 0.0508, "step": 6150 }, { "grad_norm": 0.49150559306144714, "learning_rate": 6.846770580754807e-05, "loss": 0.0515, "step": 6160 }, { "grad_norm": 0.4263674318790436, "learning_rate": 6.836522428647438e-05, "loss": 0.0506, "step": 6170 }, { "grad_norm": 0.48156484961509705, "learning_rate": 6.826265350351083e-05, "loss": 0.052, "step": 6180 }, { "grad_norm": 0.5364832878112793, "learning_rate": 6.815999395719e-05, "loss": 0.0519, "step": 6190 }, { "grad_norm": 0.5239402055740356, "learning_rate": 6.805724614647586e-05, "loss": 0.0548, "step": 6200 }, { "grad_norm": 0.4247373938560486, "learning_rate": 6.795441057076136e-05, "loss": 0.0565, "step": 6210 }, { "grad_norm": 0.4012097120285034, "learning_rate": 6.785148772986603e-05, "loss": 0.0551, "step": 6220 }, { "grad_norm": 0.4227180778980255, "learning_rate": 6.774847812403355e-05, "loss": 0.0501, "step": 6230 }, { "grad_norm": 0.3935389816761017, "learning_rate": 6.76453822539293e-05, "loss": 0.0533, "step": 6240 }, { "grad_norm": 0.4247536361217499, "learning_rate": 6.754220062063793e-05, "loss": 0.0504, "step": 6250 }, { "grad_norm": 0.5370370745658875, "learning_rate": 6.743893372566099e-05, "loss": 0.0504, "step": 6260 }, { "grad_norm": 0.5107715129852295, "learning_rate": 6.733558207091434e-05, "loss": 0.0493, "step": 6270 }, { "grad_norm": 0.48918360471725464, "learning_rate": 6.723214615872585e-05, "loss": 0.0489, "step": 6280 }, { "grad_norm": 0.45437511801719666, "learning_rate": 6.712862649183295e-05, "loss": 0.0543, "step": 6290 }, { "grad_norm": 0.4711165428161621, "learning_rate": 6.70250235733801e-05, "loss": 0.0501, "step": 6300 }, { "grad_norm": 0.42373597621917725, "learning_rate": 6.692133790691639e-05, "loss": 0.0497, "step": 6310 }, { "grad_norm": 0.4745435118675232, "learning_rate": 6.681756999639311e-05, "loss": 0.0517, "step": 6320 }, { "grad_norm": 0.6865962147712708, "learning_rate": 6.671372034616132e-05, "loss": 0.0547, "step": 6330 }, { "grad_norm": 0.4948456883430481, "learning_rate": 6.660978946096933e-05, "loss": 0.0528, "step": 6340 }, { "grad_norm": 0.4967009127140045, "learning_rate": 6.650577784596026e-05, "loss": 0.0601, "step": 6350 }, { "grad_norm": 0.45806553959846497, "learning_rate": 6.640168600666967e-05, "loss": 0.0502, "step": 6360 }, { "grad_norm": 0.49053776264190674, "learning_rate": 6.629751444902299e-05, "loss": 0.0563, "step": 6370 }, { "grad_norm": 0.5414012670516968, "learning_rate": 6.619326367933312e-05, "loss": 0.0502, "step": 6380 }, { "grad_norm": 0.5823125839233398, "learning_rate": 6.608893420429798e-05, "loss": 0.056, "step": 6390 }, { "grad_norm": 0.5064966678619385, "learning_rate": 6.598452653099803e-05, "loss": 0.0567, "step": 6400 }, { "grad_norm": 0.5080152750015259, "learning_rate": 6.588004116689375e-05, "loss": 0.0619, "step": 6410 }, { "grad_norm": 0.5127304792404175, "learning_rate": 6.57754786198233e-05, "loss": 0.0568, "step": 6420 }, { "grad_norm": 0.5565342307090759, "learning_rate": 6.567083939799992e-05, "loss": 0.0516, "step": 6430 }, { "grad_norm": 0.5701274275779724, "learning_rate": 6.556612401000954e-05, "loss": 0.0583, "step": 6440 }, { "grad_norm": 0.4266544580459595, "learning_rate": 6.54613329648083e-05, "loss": 0.0524, "step": 6450 }, { "grad_norm": 0.447275847196579, "learning_rate": 6.535646677172005e-05, "loss": 0.0554, "step": 6460 }, { "grad_norm": 0.4642195999622345, "learning_rate": 6.52515259404339e-05, "loss": 0.0485, "step": 6470 }, { "grad_norm": 0.4528159499168396, "learning_rate": 6.514651098100167e-05, "loss": 0.0529, "step": 6480 }, { "grad_norm": 0.48384183645248413, "learning_rate": 6.504142240383555e-05, "loss": 0.0523, "step": 6490 }, { "grad_norm": 0.46060407161712646, "learning_rate": 6.493626071970549e-05, "loss": 0.0531, "step": 6500 }, { "grad_norm": 0.4837590456008911, "learning_rate": 6.483102643973682e-05, "loss": 0.0557, "step": 6510 }, { "grad_norm": 0.39597758650779724, "learning_rate": 6.472572007540764e-05, "loss": 0.0497, "step": 6520 }, { "grad_norm": 0.467807412147522, "learning_rate": 6.462034213854645e-05, "loss": 0.0528, "step": 6530 }, { "grad_norm": 0.494264155626297, "learning_rate": 6.451489314132962e-05, "loss": 0.0451, "step": 6540 }, { "grad_norm": 0.32682308554649353, "learning_rate": 6.440937359627893e-05, "loss": 0.0527, "step": 6550 }, { "grad_norm": 0.4055556654930115, "learning_rate": 6.430378401625894e-05, "loss": 0.0532, "step": 6560 }, { "grad_norm": 0.39702871441841125, "learning_rate": 6.419812491447472e-05, "loss": 0.0538, "step": 6570 }, { "grad_norm": 0.4437915086746216, "learning_rate": 6.409239680446919e-05, "loss": 0.0469, "step": 6580 }, { "grad_norm": 0.46032238006591797, "learning_rate": 6.398660020012072e-05, "loss": 0.051, "step": 6590 }, { "grad_norm": 0.44362038373947144, "learning_rate": 6.38807356156405e-05, "loss": 0.0532, "step": 6600 }, { "grad_norm": 0.5227200388908386, "learning_rate": 6.377480356557022e-05, "loss": 0.0554, "step": 6610 }, { "grad_norm": 0.5115537047386169, "learning_rate": 6.366880456477942e-05, "loss": 0.0545, "step": 6620 }, { "grad_norm": 0.42375367879867554, "learning_rate": 6.356273912846312e-05, "loss": 0.0565, "step": 6630 }, { "grad_norm": 0.49164649844169617, "learning_rate": 6.34566077721391e-05, "loss": 0.0603, "step": 6640 }, { "grad_norm": 0.5009482502937317, "learning_rate": 6.335041101164569e-05, "loss": 0.0593, "step": 6650 }, { "grad_norm": 0.45363879203796387, "learning_rate": 6.324414936313904e-05, "loss": 0.051, "step": 6660 }, { "grad_norm": 0.43615537881851196, "learning_rate": 6.313782334309066e-05, "loss": 0.0561, "step": 6670 }, { "grad_norm": 0.3973962962627411, "learning_rate": 6.303143346828499e-05, "loss": 0.0477, "step": 6680 }, { "grad_norm": 0.4147135317325592, "learning_rate": 6.292498025581674e-05, "loss": 0.049, "step": 6690 }, { "grad_norm": 0.4847707748413086, "learning_rate": 6.281846422308857e-05, "loss": 0.0484, "step": 6700 }, { "grad_norm": 0.4829198122024536, "learning_rate": 6.271188588780839e-05, "loss": 0.0537, "step": 6710 }, { "grad_norm": 0.3959425687789917, "learning_rate": 6.260524576798694e-05, "loss": 0.0572, "step": 6720 }, { "grad_norm": 0.4396299123764038, "learning_rate": 6.249854438193528e-05, "loss": 0.0535, "step": 6730 }, { "grad_norm": 0.5060888528823853, "learning_rate": 6.239178224826224e-05, "loss": 0.0517, "step": 6740 }, { "grad_norm": 0.39701133966445923, "learning_rate": 6.228495988587188e-05, "loss": 0.0576, "step": 6750 }, { "grad_norm": 0.43098127841949463, "learning_rate": 6.217807781396106e-05, "loss": 0.0532, "step": 6760 }, { "grad_norm": 0.42610257863998413, "learning_rate": 6.207113655201676e-05, "loss": 0.054, "step": 6770 }, { "grad_norm": 0.47635921835899353, "learning_rate": 6.196413661981368e-05, "loss": 0.0563, "step": 6780 }, { "grad_norm": 0.43938231468200684, "learning_rate": 6.185707853741175e-05, "loss": 0.0508, "step": 6790 }, { "grad_norm": 0.3904918134212494, "learning_rate": 6.174996282515344e-05, "loss": 0.0555, "step": 6800 }, { "grad_norm": 0.42746877670288086, "learning_rate": 6.164279000366131e-05, "loss": 0.0526, "step": 6810 }, { "grad_norm": 0.43453386425971985, "learning_rate": 6.153556059383561e-05, "loss": 0.054, "step": 6820 }, { "grad_norm": 0.42769187688827515, "learning_rate": 6.142827511685152e-05, "loss": 0.0531, "step": 6830 }, { "grad_norm": 0.4483128488063812, "learning_rate": 6.132093409415678e-05, "loss": 0.0542, "step": 6840 }, { "grad_norm": 0.4847145974636078, "learning_rate": 6.121353804746907e-05, "loss": 0.0531, "step": 6850 }, { "grad_norm": 0.41745612025260925, "learning_rate": 6.110608749877352e-05, "loss": 0.0528, "step": 6860 }, { "grad_norm": 0.3488568663597107, "learning_rate": 6.0998582970320205e-05, "loss": 0.0547, "step": 6870 }, { "grad_norm": 0.44237440824508667, "learning_rate": 6.0891024984621506e-05, "loss": 0.053, "step": 6880 }, { "grad_norm": 0.3580341339111328, "learning_rate": 6.078341406444961e-05, "loss": 0.0482, "step": 6890 }, { "grad_norm": 0.4013528525829315, "learning_rate": 6.067575073283405e-05, "loss": 0.0523, "step": 6900 }, { "grad_norm": 0.39542070031166077, "learning_rate": 6.0568035513059073e-05, "loss": 0.0548, "step": 6910 }, { "grad_norm": 0.40254467725753784, "learning_rate": 6.046026892866109e-05, "loss": 0.0516, "step": 6920 }, { "grad_norm": 0.3407313823699951, "learning_rate": 6.0352451503426214e-05, "loss": 0.0504, "step": 6930 }, { "grad_norm": 0.3474937379360199, "learning_rate": 6.024458376138762e-05, "loss": 0.0466, "step": 6940 }, { "grad_norm": 0.377045214176178, "learning_rate": 6.013666622682306e-05, "loss": 0.0537, "step": 6950 }, { "grad_norm": 0.37717339396476746, "learning_rate": 6.002869942425231e-05, "loss": 0.0501, "step": 6960 }, { "grad_norm": 0.4188041090965271, "learning_rate": 5.992068387843459e-05, "loss": 0.0467, "step": 6970 }, { "grad_norm": 0.4354075491428375, "learning_rate": 5.981262011436603e-05, "loss": 0.0521, "step": 6980 }, { "grad_norm": 0.4561382830142975, "learning_rate": 5.970450865727712e-05, "loss": 0.0486, "step": 6990 }, { "grad_norm": 0.3954760730266571, "learning_rate": 5.9596350032630156e-05, "loss": 0.0453, "step": 7000 }, { "grad_norm": 0.5127900242805481, "learning_rate": 5.9488144766116714e-05, "loss": 0.0522, "step": 7010 }, { "grad_norm": 0.4630087912082672, "learning_rate": 5.9379893383655006e-05, "loss": 0.0496, "step": 7020 }, { "grad_norm": 0.5121265053749084, "learning_rate": 5.927159641138744e-05, "loss": 0.0497, "step": 7030 }, { "grad_norm": 0.4236236810684204, "learning_rate": 5.916325437567799e-05, "loss": 0.0511, "step": 7040 }, { "grad_norm": 0.4083375334739685, "learning_rate": 5.905486780310966e-05, "loss": 0.0487, "step": 7050 }, { "grad_norm": 0.417698472738266, "learning_rate": 5.8946437220481887e-05, "loss": 0.044, "step": 7060 }, { "grad_norm": 0.5372232794761658, "learning_rate": 5.883796315480805e-05, "loss": 0.0529, "step": 7070 }, { "grad_norm": 0.4315335750579834, "learning_rate": 5.872944613331288e-05, "loss": 0.0485, "step": 7080 }, { "grad_norm": 0.4357188940048218, "learning_rate": 5.862088668342986e-05, "loss": 0.0506, "step": 7090 }, { "grad_norm": 0.47571825981140137, "learning_rate": 5.8512285332798714e-05, "loss": 0.0497, "step": 7100 }, { "grad_norm": 0.37382951378822327, "learning_rate": 5.840364260926277e-05, "loss": 0.0516, "step": 7110 }, { "grad_norm": 0.4397519528865814, "learning_rate": 5.8294959040866505e-05, "loss": 0.0488, "step": 7120 }, { "grad_norm": 0.48862946033477783, "learning_rate": 5.818623515585292e-05, "loss": 0.0513, "step": 7130 }, { "grad_norm": 0.40299516916275024, "learning_rate": 5.8077471482660896e-05, "loss": 0.0471, "step": 7140 }, { "grad_norm": 0.49317651987075806, "learning_rate": 5.796866854992276e-05, "loss": 0.045, "step": 7150 }, { "grad_norm": 0.49936503171920776, "learning_rate": 5.7859826886461676e-05, "loss": 0.0539, "step": 7160 }, { "grad_norm": 0.5492211580276489, "learning_rate": 5.775094702128899e-05, "loss": 0.0564, "step": 7170 }, { "grad_norm": 0.44816526770591736, "learning_rate": 5.7642029483601746e-05, "loss": 0.0503, "step": 7180 }, { "grad_norm": 0.4234413206577301, "learning_rate": 5.753307480278012e-05, "loss": 0.0521, "step": 7190 }, { "grad_norm": 0.49466946721076965, "learning_rate": 5.742408350838478e-05, "loss": 0.0486, "step": 7200 }, { "grad_norm": 0.4848979711532593, "learning_rate": 5.7315056130154374e-05, "loss": 0.0491, "step": 7210 }, { "grad_norm": 0.3802797198295593, "learning_rate": 5.720599319800292e-05, "loss": 0.0446, "step": 7220 }, { "grad_norm": 0.50968998670578, "learning_rate": 5.709689524201722e-05, "loss": 0.0543, "step": 7230 }, { "grad_norm": 0.4330335259437561, "learning_rate": 5.698776279245437e-05, "loss": 0.0535, "step": 7240 }, { "grad_norm": 0.44570741057395935, "learning_rate": 5.6878596379739036e-05, "loss": 0.0478, "step": 7250 }, { "grad_norm": 0.41537046432495117, "learning_rate": 5.676939653446103e-05, "loss": 0.0463, "step": 7260 }, { "grad_norm": 0.43384793400764465, "learning_rate": 5.666016378737261e-05, "loss": 0.0532, "step": 7270 }, { "grad_norm": 0.3811190724372864, "learning_rate": 5.655089866938596e-05, "loss": 0.0522, "step": 7280 }, { "grad_norm": 0.34280845522880554, "learning_rate": 5.6441601711570615e-05, "loss": 0.0503, "step": 7290 }, { "grad_norm": 0.4587463438510895, "learning_rate": 5.633227344515085e-05, "loss": 0.0498, "step": 7300 }, { "grad_norm": 0.40999212861061096, "learning_rate": 5.6222914401503116e-05, "loss": 0.0513, "step": 7310 }, { "grad_norm": 0.4035322368144989, "learning_rate": 5.611352511215343e-05, "loss": 0.0544, "step": 7320 }, { "grad_norm": 0.39643338322639465, "learning_rate": 5.600410610877488e-05, "loss": 0.0485, "step": 7330 }, { "grad_norm": 0.4413648843765259, "learning_rate": 5.58946579231849e-05, "loss": 0.0476, "step": 7340 }, { "grad_norm": 0.4641997218132019, "learning_rate": 5.578518108734279e-05, "loss": 0.0471, "step": 7350 }, { "grad_norm": 0.3835093379020691, "learning_rate": 5.5675676133347096e-05, "loss": 0.0566, "step": 7360 }, { "grad_norm": 0.458680659532547, "learning_rate": 5.556614359343307e-05, "loss": 0.0564, "step": 7370 }, { "grad_norm": 0.43564116954803467, "learning_rate": 5.545658399996999e-05, "loss": 0.0495, "step": 7380 }, { "grad_norm": 0.4177677631378174, "learning_rate": 5.534699788545862e-05, "loss": 0.0557, "step": 7390 }, { "grad_norm": 0.402659147977829, "learning_rate": 5.523738578252867e-05, "loss": 0.0464, "step": 7400 }, { "grad_norm": 0.3850119411945343, "learning_rate": 5.512774822393614e-05, "loss": 0.0466, "step": 7410 }, { "grad_norm": 0.3956775963306427, "learning_rate": 5.5018085742560744e-05, "loss": 0.0472, "step": 7420 }, { "grad_norm": 0.36993297934532166, "learning_rate": 5.4908398871403365e-05, "loss": 0.0472, "step": 7430 }, { "grad_norm": 0.4440155029296875, "learning_rate": 5.4798688143583375e-05, "loss": 0.0479, "step": 7440 }, { "grad_norm": 0.4351479411125183, "learning_rate": 5.468895409233615e-05, "loss": 0.0446, "step": 7450 }, { "grad_norm": 0.391032338142395, "learning_rate": 5.4579197251010414e-05, "loss": 0.0523, "step": 7460 }, { "grad_norm": 0.39971309900283813, "learning_rate": 5.446941815306563e-05, "loss": 0.0496, "step": 7470 }, { "grad_norm": 0.4459120035171509, "learning_rate": 5.435961733206947e-05, "loss": 0.0438, "step": 7480 }, { "grad_norm": 0.35966992378234863, "learning_rate": 5.424979532169516e-05, "loss": 0.0427, "step": 7490 }, { "grad_norm": 0.4109203815460205, "learning_rate": 5.413995265571895e-05, "loss": 0.0506, "step": 7500 }, { "grad_norm": 0.423715204000473, "learning_rate": 5.403008986801746e-05, "loss": 0.0445, "step": 7510 }, { "grad_norm": 0.38528698682785034, "learning_rate": 5.3920207492565114e-05, "loss": 0.0498, "step": 7520 }, { "grad_norm": 0.33950096368789673, "learning_rate": 5.381030606343154e-05, "loss": 0.0479, "step": 7530 }, { "grad_norm": 0.45253995060920715, "learning_rate": 5.370038611477894e-05, "loss": 0.0478, "step": 7540 }, { "grad_norm": 0.4006006121635437, "learning_rate": 5.359044818085963e-05, "loss": 0.0469, "step": 7550 }, { "grad_norm": 0.4765866696834564, "learning_rate": 5.3480492796013214e-05, "loss": 0.0578, "step": 7560 }, { "grad_norm": 0.39808720350265503, "learning_rate": 5.33705204946642e-05, "loss": 0.0498, "step": 7570 }, { "grad_norm": 0.4232408404350281, "learning_rate": 5.326053181131927e-05, "loss": 0.0447, "step": 7580 }, { "grad_norm": 0.35713475942611694, "learning_rate": 5.3150527280564776e-05, "loss": 0.0462, "step": 7590 }, { "grad_norm": 0.3734726011753082, "learning_rate": 5.3040507437064034e-05, "loss": 0.0482, "step": 7600 }, { "grad_norm": 0.43107184767723083, "learning_rate": 5.293047281555482e-05, "loss": 0.0438, "step": 7610 }, { "grad_norm": 0.4572940468788147, "learning_rate": 5.2820423950846765e-05, "loss": 0.0479, "step": 7620 }, { "grad_norm": 0.39231589436531067, "learning_rate": 5.2710361377818696e-05, "loss": 0.0488, "step": 7630 }, { "grad_norm": 0.44023293256759644, "learning_rate": 5.2600285631416026e-05, "loss": 0.0499, "step": 7640 }, { "grad_norm": 0.4163152873516083, "learning_rate": 5.249019724664826e-05, "loss": 0.0486, "step": 7650 }, { "grad_norm": 0.3857567012310028, "learning_rate": 5.2380096758586315e-05, "loss": 0.0566, "step": 7660 }, { "grad_norm": 0.5152963995933533, "learning_rate": 5.226998470235993e-05, "loss": 0.0533, "step": 7670 }, { "grad_norm": 0.42160046100616455, "learning_rate": 5.215986161315507e-05, "loss": 0.0465, "step": 7680 }, { "grad_norm": 0.4126971364021301, "learning_rate": 5.20497280262113e-05, "loss": 0.0463, "step": 7690 }, { "grad_norm": 0.3896881937980652, "learning_rate": 5.193958447681924e-05, "loss": 0.054, "step": 7700 }, { "grad_norm": 0.38263654708862305, "learning_rate": 5.182943150031793e-05, "loss": 0.0472, "step": 7710 }, { "grad_norm": 0.32398107647895813, "learning_rate": 5.1719269632092204e-05, "loss": 0.0433, "step": 7720 }, { "grad_norm": 0.3992784023284912, "learning_rate": 5.160909940757015e-05, "loss": 0.0467, "step": 7730 }, { "grad_norm": 0.3852086663246155, "learning_rate": 5.149892136222043e-05, "loss": 0.0433, "step": 7740 }, { "grad_norm": 0.35278046131134033, "learning_rate": 5.1388736031549744e-05, "loss": 0.0463, "step": 7750 }, { "grad_norm": 0.422016978263855, "learning_rate": 5.127854395110021e-05, "loss": 0.0451, "step": 7760 }, { "grad_norm": 0.43665069341659546, "learning_rate": 5.116834565644671e-05, "loss": 0.0535, "step": 7770 }, { "grad_norm": 0.37852731347084045, "learning_rate": 5.10581416831944e-05, "loss": 0.044, "step": 7780 }, { "grad_norm": 0.36436188220977783, "learning_rate": 5.094793256697593e-05, "loss": 0.0449, "step": 7790 }, { "grad_norm": 0.41151487827301025, "learning_rate": 5.0837718843449075e-05, "loss": 0.0454, "step": 7800 }, { "grad_norm": 0.33720794320106506, "learning_rate": 5.07275010482939e-05, "loss": 0.0442, "step": 7810 }, { "grad_norm": 0.40955567359924316, "learning_rate": 5.061727971721032e-05, "loss": 0.0413, "step": 7820 }, { "grad_norm": 0.5101323127746582, "learning_rate": 5.050705538591538e-05, "loss": 0.0477, "step": 7830 }, { "grad_norm": 0.4389752745628357, "learning_rate": 5.0396828590140785e-05, "loss": 0.0444, "step": 7840 }, { "grad_norm": 0.41371065378189087, "learning_rate": 5.0286599865630157e-05, "loss": 0.0458, "step": 7850 }, { "grad_norm": 0.4115827977657318, "learning_rate": 5.017636974813649e-05, "loss": 0.0509, "step": 7860 }, { "grad_norm": 0.49503251910209656, "learning_rate": 5.006613877341959e-05, "loss": 0.0448, "step": 7870 }, { "grad_norm": 0.4804134964942932, "learning_rate": 4.99559074772434e-05, "loss": 0.0462, "step": 7880 }, { "grad_norm": 0.4580792188644409, "learning_rate": 4.9845676395373455e-05, "loss": 0.0538, "step": 7890 }, { "grad_norm": 0.38742682337760925, "learning_rate": 4.9735446063574184e-05, "loss": 0.0479, "step": 7900 }, { "grad_norm": 0.4620611071586609, "learning_rate": 4.962521701760645e-05, "loss": 0.0498, "step": 7910 }, { "grad_norm": 0.415036678314209, "learning_rate": 4.951498979322482e-05, "loss": 0.0461, "step": 7920 }, { "grad_norm": 0.36234480142593384, "learning_rate": 4.9404764926174996e-05, "loss": 0.0463, "step": 7930 }, { "grad_norm": 0.4542545974254608, "learning_rate": 4.929454295219127e-05, "loss": 0.0524, "step": 7940 }, { "grad_norm": 0.4701482355594635, "learning_rate": 4.9184324406993844e-05, "loss": 0.0464, "step": 7950 }, { "grad_norm": 0.42406633496284485, "learning_rate": 4.907410982628623e-05, "loss": 0.0508, "step": 7960 }, { "grad_norm": 0.35604408383369446, "learning_rate": 4.896389974575273e-05, "loss": 0.0535, "step": 7970 }, { "grad_norm": 0.3244574964046478, "learning_rate": 4.885369470105571e-05, "loss": 0.0533, "step": 7980 }, { "grad_norm": 0.4028531610965729, "learning_rate": 4.874349522783313e-05, "loss": 0.0468, "step": 7990 }, { "grad_norm": 0.3792223334312439, "learning_rate": 4.863330186169581e-05, "loss": 0.0441, "step": 8000 }, { "grad_norm": 0.3794063329696655, "learning_rate": 4.8523115138224885e-05, "loss": 0.0437, "step": 8010 }, { "grad_norm": 0.44544780254364014, "learning_rate": 4.841293559296928e-05, "loss": 0.0465, "step": 8020 }, { "grad_norm": 0.4357159435749054, "learning_rate": 4.830276376144295e-05, "loss": 0.0498, "step": 8030 }, { "grad_norm": 0.39284729957580566, "learning_rate": 4.819260017912237e-05, "loss": 0.0487, "step": 8040 }, { "grad_norm": 0.3869476914405823, "learning_rate": 4.808244538144396e-05, "loss": 0.0463, "step": 8050 }, { "grad_norm": 0.42206189036369324, "learning_rate": 4.797229990380142e-05, "loss": 0.0438, "step": 8060 }, { "grad_norm": 0.40598607063293457, "learning_rate": 4.786216428154317e-05, "loss": 0.0453, "step": 8070 }, { "grad_norm": 0.3976571559906006, "learning_rate": 4.7752039049969685e-05, "loss": 0.0467, "step": 8080 }, { "grad_norm": 0.39836353063583374, "learning_rate": 4.7641924744330956e-05, "loss": 0.0481, "step": 8090 }, { "grad_norm": 0.43296316266059875, "learning_rate": 4.7531821899823925e-05, "loss": 0.0469, "step": 8100 }, { "grad_norm": 0.4745776355266571, "learning_rate": 4.742173105158973e-05, "loss": 0.0431, "step": 8110 }, { "grad_norm": 0.42880779504776, "learning_rate": 4.731165273471129e-05, "loss": 0.0465, "step": 8120 }, { "grad_norm": 0.37811797857284546, "learning_rate": 4.720158748421057e-05, "loss": 0.0479, "step": 8130 }, { "grad_norm": 0.40468043088912964, "learning_rate": 4.709153583504602e-05, "loss": 0.042, "step": 8140 }, { "grad_norm": 0.3899652063846588, "learning_rate": 4.6981498322110027e-05, "loss": 0.0451, "step": 8150 }, { "grad_norm": 0.34649503231048584, "learning_rate": 4.6871475480226256e-05, "loss": 0.046, "step": 8160 }, { "grad_norm": 0.33068931102752686, "learning_rate": 4.6761467844147004e-05, "loss": 0.0482, "step": 8170 }, { "grad_norm": 0.36666229367256165, "learning_rate": 4.665147594855076e-05, "loss": 0.044, "step": 8180 }, { "grad_norm": 0.37364462018013, "learning_rate": 4.654150032803943e-05, "loss": 0.0477, "step": 8190 }, { "grad_norm": 0.48789554834365845, "learning_rate": 4.643154151713588e-05, "loss": 0.046, "step": 8200 }, { "grad_norm": 0.3791208863258362, "learning_rate": 4.6321600050281225e-05, "loss": 0.0442, "step": 8210 }, { "grad_norm": 0.39227837324142456, "learning_rate": 4.6211676461832264e-05, "loss": 0.0417, "step": 8220 }, { "grad_norm": 0.386371374130249, "learning_rate": 4.610177128605899e-05, "loss": 0.043, "step": 8230 }, { "grad_norm": 0.41427555680274963, "learning_rate": 4.599188505714184e-05, "loss": 0.0447, "step": 8240 }, { "grad_norm": 0.4369830787181854, "learning_rate": 4.588201830916912e-05, "loss": 0.0476, "step": 8250 }, { "grad_norm": 0.3863682150840759, "learning_rate": 4.577217157613456e-05, "loss": 0.0452, "step": 8260 }, { "grad_norm": 0.41397857666015625, "learning_rate": 4.566234539193452e-05, "loss": 0.0435, "step": 8270 }, { "grad_norm": 0.3764868378639221, "learning_rate": 4.555254029036555e-05, "loss": 0.0429, "step": 8280 }, { "grad_norm": 0.4344530701637268, "learning_rate": 4.544275680512165e-05, "loss": 0.0472, "step": 8290 }, { "grad_norm": 0.40178748965263367, "learning_rate": 4.5332995469791836e-05, "loss": 0.0439, "step": 8300 }, { "grad_norm": 0.4508951008319855, "learning_rate": 4.522325681785744e-05, "loss": 0.0466, "step": 8310 }, { "grad_norm": 0.44362568855285645, "learning_rate": 4.511354138268952e-05, "loss": 0.0438, "step": 8320 }, { "grad_norm": 0.3914468586444855, "learning_rate": 4.50038496975463e-05, "loss": 0.0455, "step": 8330 }, { "grad_norm": 0.35512328147888184, "learning_rate": 4.489418229557063e-05, "loss": 0.0455, "step": 8340 }, { "grad_norm": 0.39773914217948914, "learning_rate": 4.478453970978722e-05, "loss": 0.0481, "step": 8350 }, { "grad_norm": 0.3480338752269745, "learning_rate": 4.4674922473100286e-05, "loss": 0.042, "step": 8360 }, { "grad_norm": 0.3580087721347809, "learning_rate": 4.4565331118290756e-05, "loss": 0.0384, "step": 8370 }, { "grad_norm": 0.3791510760784149, "learning_rate": 4.4455766178013775e-05, "loss": 0.044, "step": 8380 }, { "grad_norm": 0.4528161585330963, "learning_rate": 4.434622818479615e-05, "loss": 0.0451, "step": 8390 }, { "grad_norm": 0.43347620964050293, "learning_rate": 4.4236717671033646e-05, "loss": 0.0447, "step": 8400 }, { "grad_norm": 0.3572349548339844, "learning_rate": 4.412723516898853e-05, "loss": 0.0479, "step": 8410 }, { "grad_norm": 0.37822389602661133, "learning_rate": 4.40177812107869e-05, "loss": 0.0451, "step": 8420 }, { "grad_norm": 0.32863834500312805, "learning_rate": 4.390835632841606e-05, "loss": 0.0429, "step": 8430 }, { "grad_norm": 0.3779616355895996, "learning_rate": 4.3798961053722115e-05, "loss": 0.0449, "step": 8440 }, { "grad_norm": 0.32004088163375854, "learning_rate": 4.368959591840718e-05, "loss": 0.0544, "step": 8450 }, { "grad_norm": 0.372659832239151, "learning_rate": 4.3580261454026865e-05, "loss": 0.0423, "step": 8460 }, { "grad_norm": 0.2738495469093323, "learning_rate": 4.3470958191987786e-05, "loss": 0.0433, "step": 8470 }, { "grad_norm": 0.3056885600090027, "learning_rate": 4.336168666354484e-05, "loss": 0.0442, "step": 8480 }, { "grad_norm": 0.3805913031101227, "learning_rate": 4.325244739979873e-05, "loss": 0.0454, "step": 8490 }, { "grad_norm": 0.35787200927734375, "learning_rate": 4.314324093169332e-05, "loss": 0.049, "step": 8500 }, { "grad_norm": 0.3060150444507599, "learning_rate": 4.303406779001302e-05, "loss": 0.0457, "step": 8510 }, { "grad_norm": 0.43035343289375305, "learning_rate": 4.292492850538038e-05, "loss": 0.0445, "step": 8520 }, { "grad_norm": 0.4244695007801056, "learning_rate": 4.28158236082533e-05, "loss": 0.0455, "step": 8530 }, { "grad_norm": 0.39108312129974365, "learning_rate": 4.270675362892256e-05, "loss": 0.05, "step": 8540 }, { "grad_norm": 0.33763113617897034, "learning_rate": 4.2597719097509246e-05, "loss": 0.0419, "step": 8550 }, { "grad_norm": 0.4146230220794678, "learning_rate": 4.2488720543962146e-05, "loss": 0.0496, "step": 8560 }, { "grad_norm": 0.38126295804977417, "learning_rate": 4.23797584980552e-05, "loss": 0.0516, "step": 8570 }, { "grad_norm": 0.36038661003112793, "learning_rate": 4.227083348938486e-05, "loss": 0.046, "step": 8580 }, { "grad_norm": 0.33426985144615173, "learning_rate": 4.2161946047367586e-05, "loss": 0.0449, "step": 8590 }, { "grad_norm": 0.3494066596031189, "learning_rate": 4.2053096701237294e-05, "loss": 0.0448, "step": 8600 }, { "grad_norm": 0.30634987354278564, "learning_rate": 4.1944285980042656e-05, "loss": 0.0436, "step": 8610 }, { "grad_norm": 0.37098971009254456, "learning_rate": 4.183551441264469e-05, "loss": 0.0435, "step": 8620 }, { "grad_norm": 0.30043089389801025, "learning_rate": 4.172678252771408e-05, "loss": 0.0476, "step": 8630 }, { "grad_norm": 0.39938807487487793, "learning_rate": 4.16180908537286e-05, "loss": 0.048, "step": 8640 }, { "grad_norm": 0.43720537424087524, "learning_rate": 4.150943991897065e-05, "loss": 0.0487, "step": 8650 }, { "grad_norm": 0.3159247934818268, "learning_rate": 4.1400830251524605e-05, "loss": 0.0424, "step": 8660 }, { "grad_norm": 0.42136457562446594, "learning_rate": 4.1292262379274215e-05, "loss": 0.0494, "step": 8670 }, { "grad_norm": 0.3358195126056671, "learning_rate": 4.118373682990016e-05, "loss": 0.0486, "step": 8680 }, { "grad_norm": 0.3457491397857666, "learning_rate": 4.107525413087737e-05, "loss": 0.0396, "step": 8690 }, { "grad_norm": 0.31798669695854187, "learning_rate": 4.096681480947252e-05, "loss": 0.0379, "step": 8700 }, { "grad_norm": 0.40467530488967896, "learning_rate": 4.085841939274146e-05, "loss": 0.0481, "step": 8710 }, { "grad_norm": 0.38029882311820984, "learning_rate": 4.075006840752662e-05, "loss": 0.0446, "step": 8720 }, { "grad_norm": 0.3821033835411072, "learning_rate": 4.0641762380454515e-05, "loss": 0.0422, "step": 8730 }, { "grad_norm": 0.3535569906234741, "learning_rate": 4.0533501837933134e-05, "loss": 0.0461, "step": 8740 }, { "grad_norm": 0.40044671297073364, "learning_rate": 4.042528730614936e-05, "loss": 0.0399, "step": 8750 }, { "grad_norm": 0.3393961489200592, "learning_rate": 4.0317119311066486e-05, "loss": 0.0414, "step": 8760 }, { "grad_norm": 0.3337998688220978, "learning_rate": 4.02089983784216e-05, "loss": 0.0495, "step": 8770 }, { "grad_norm": 0.3629573583602905, "learning_rate": 4.010092503372309e-05, "loss": 0.0467, "step": 8780 }, { "grad_norm": 0.29001185297966003, "learning_rate": 3.999289980224797e-05, "loss": 0.0469, "step": 8790 }, { "grad_norm": 0.35299980640411377, "learning_rate": 3.9884923209039455e-05, "loss": 0.0376, "step": 8800 }, { "grad_norm": 0.39730244874954224, "learning_rate": 3.977699577890439e-05, "loss": 0.0382, "step": 8810 }, { "grad_norm": 0.3366188704967499, "learning_rate": 3.96691180364106e-05, "loss": 0.0471, "step": 8820 }, { "grad_norm": 0.3361673951148987, "learning_rate": 3.956129050588446e-05, "loss": 0.0461, "step": 8830 }, { "grad_norm": 0.35741665959358215, "learning_rate": 3.9453513711408275e-05, "loss": 0.0479, "step": 8840 }, { "grad_norm": 0.3107794225215912, "learning_rate": 3.934578817681774e-05, "loss": 0.0474, "step": 8850 }, { "grad_norm": 0.444961816072464, "learning_rate": 3.9238114425699465e-05, "loss": 0.0471, "step": 8860 }, { "grad_norm": 0.3996797204017639, "learning_rate": 3.91304929813883e-05, "loss": 0.0465, "step": 8870 }, { "grad_norm": 0.3907962143421173, "learning_rate": 3.902292436696489e-05, "loss": 0.0417, "step": 8880 }, { "grad_norm": 0.4146788418292999, "learning_rate": 3.891540910525316e-05, "loss": 0.0458, "step": 8890 }, { "grad_norm": 0.37497562170028687, "learning_rate": 3.8807947718817624e-05, "loss": 0.0483, "step": 8900 }, { "grad_norm": 0.3485381603240967, "learning_rate": 3.870054072996103e-05, "loss": 0.0446, "step": 8910 }, { "grad_norm": 0.33798184990882874, "learning_rate": 3.859318866072168e-05, "loss": 0.0469, "step": 8920 }, { "grad_norm": 0.3721936345100403, "learning_rate": 3.8485892032870965e-05, "loss": 0.0474, "step": 8930 }, { "grad_norm": 0.35253363847732544, "learning_rate": 3.83786513679108e-05, "loss": 0.0444, "step": 8940 }, { "grad_norm": 0.47586336731910706, "learning_rate": 3.8271467187071134e-05, "loss": 0.043, "step": 8950 }, { "grad_norm": 0.5007453560829163, "learning_rate": 3.816434001130732e-05, "loss": 0.0447, "step": 8960 }, { "grad_norm": 0.4278405010700226, "learning_rate": 3.8057270361297706e-05, "loss": 0.0467, "step": 8970 }, { "grad_norm": 0.3811814785003662, "learning_rate": 3.7950258757440985e-05, "loss": 0.0469, "step": 8980 }, { "grad_norm": 0.44491490721702576, "learning_rate": 3.78433057198538e-05, "loss": 0.0477, "step": 8990 }, { "grad_norm": 0.4218033254146576, "learning_rate": 3.773641176836807e-05, "loss": 0.0475, "step": 9000 } ], "logging_steps": 10, "max_steps": 15000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 3000, "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 }