{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 15000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 2.380291700363159, "learning_rate": 6.000000000000001e-07, "loss": 1.4121, "step": 10 }, { "grad_norm": 2.220470666885376, "learning_rate": 1.2666666666666667e-06, "loss": 1.4105, "step": 20 }, { "grad_norm": 1.567824125289917, "learning_rate": 1.9333333333333336e-06, "loss": 1.3822, "step": 30 }, { "grad_norm": 1.223537564277649, "learning_rate": 2.6e-06, "loss": 1.3275, "step": 40 }, { "grad_norm": 1.1619495153427124, "learning_rate": 3.2666666666666666e-06, "loss": 1.2514, "step": 50 }, { "grad_norm": 1.1553289890289307, "learning_rate": 3.9333333333333335e-06, "loss": 1.1889, "step": 60 }, { "grad_norm": 0.8499078154563904, "learning_rate": 4.6e-06, "loss": 1.1348, "step": 70 }, { "grad_norm": 0.582207977771759, "learning_rate": 5.266666666666667e-06, "loss": 1.0949, "step": 80 }, { "grad_norm": 1.0987365245819092, "learning_rate": 5.933333333333334e-06, "loss": 1.0814, "step": 90 }, { "grad_norm": 0.724940299987793, "learning_rate": 6.6e-06, "loss": 1.0648, "step": 100 }, { "grad_norm": 0.8098756074905396, "learning_rate": 7.266666666666668e-06, "loss": 1.0518, "step": 110 }, { "grad_norm": 1.0196516513824463, "learning_rate": 7.933333333333334e-06, "loss": 1.0482, "step": 120 }, { "grad_norm": 0.8033488392829895, "learning_rate": 8.599999999999999e-06, "loss": 1.0422, "step": 130 }, { "grad_norm": 0.6941110491752625, "learning_rate": 9.266666666666667e-06, "loss": 1.048, "step": 140 }, { "grad_norm": 0.8264422416687012, "learning_rate": 9.933333333333334e-06, "loss": 1.0393, "step": 150 }, { "grad_norm": 0.6281202435493469, "learning_rate": 1.06e-05, "loss": 1.0309, "step": 160 }, { "grad_norm": 0.7770757675170898, "learning_rate": 1.1266666666666667e-05, "loss": 1.0266, "step": 170 }, { "grad_norm": 0.6598097085952759, "learning_rate": 1.1933333333333333e-05, "loss": 1.0223, "step": 180 }, { "grad_norm": 0.4384413957595825, "learning_rate": 1.2600000000000001e-05, "loss": 1.0285, "step": 190 }, { "grad_norm": 0.7572369575500488, "learning_rate": 1.3266666666666666e-05, "loss": 1.0211, "step": 200 }, { "grad_norm": 0.9369884133338928, "learning_rate": 1.3933333333333334e-05, "loss": 1.0098, "step": 210 }, { "grad_norm": 1.0073491334915161, "learning_rate": 1.4599999999999999e-05, "loss": 0.992, "step": 220 }, { "grad_norm": 0.9934375286102295, "learning_rate": 1.5266666666666667e-05, "loss": 0.9797, "step": 230 }, { "grad_norm": 1.2248018980026245, "learning_rate": 1.5933333333333332e-05, "loss": 0.9569, "step": 240 }, { "grad_norm": 1.1634633541107178, "learning_rate": 1.66e-05, "loss": 0.9303, "step": 250 }, { "grad_norm": 1.2508214712142944, "learning_rate": 1.726666666666667e-05, "loss": 0.8967, "step": 260 }, { "grad_norm": 1.2230318784713745, "learning_rate": 1.7933333333333337e-05, "loss": 0.8643, "step": 270 }, { "grad_norm": 1.2104076147079468, "learning_rate": 1.86e-05, "loss": 0.8393, "step": 280 }, { "grad_norm": 1.1637425422668457, "learning_rate": 1.926666666666667e-05, "loss": 0.82, "step": 290 }, { "grad_norm": 1.3498616218566895, "learning_rate": 1.9933333333333334e-05, "loss": 0.7912, "step": 300 }, { "grad_norm": 1.3146412372589111, "learning_rate": 2.06e-05, "loss": 0.7795, "step": 310 }, { "grad_norm": 1.237015962600708, "learning_rate": 2.1266666666666667e-05, "loss": 0.7535, "step": 320 }, { "grad_norm": 1.4436708688735962, "learning_rate": 2.1933333333333332e-05, "loss": 0.7535, "step": 330 }, { "grad_norm": 1.3675235509872437, "learning_rate": 2.26e-05, "loss": 0.7283, "step": 340 }, { "grad_norm": 1.078903079032898, "learning_rate": 2.326666666666667e-05, "loss": 0.7219, "step": 350 }, { "grad_norm": 1.0850610733032227, "learning_rate": 2.3933333333333337e-05, "loss": 0.7087, "step": 360 }, { "grad_norm": 1.4821397066116333, "learning_rate": 2.46e-05, "loss": 0.6923, "step": 370 }, { "grad_norm": 1.3027005195617676, "learning_rate": 2.5266666666666666e-05, "loss": 0.6784, "step": 380 }, { "grad_norm": 1.020719289779663, "learning_rate": 2.5933333333333338e-05, "loss": 0.6641, "step": 390 }, { "grad_norm": 1.2909667491912842, "learning_rate": 2.6600000000000003e-05, "loss": 0.631, "step": 400 }, { "grad_norm": 1.332321047782898, "learning_rate": 2.7266666666666668e-05, "loss": 0.6089, "step": 410 }, { "grad_norm": 1.2372812032699585, "learning_rate": 2.7933333333333332e-05, "loss": 0.5847, "step": 420 }, { "grad_norm": 1.1915992498397827, "learning_rate": 2.86e-05, "loss": 0.5625, "step": 430 }, { "grad_norm": 1.4657704830169678, "learning_rate": 2.926666666666667e-05, "loss": 0.5625, "step": 440 }, { "grad_norm": 1.6470774412155151, "learning_rate": 2.9933333333333337e-05, "loss": 0.5323, "step": 450 }, { "grad_norm": 1.5229383707046509, "learning_rate": 3.06e-05, "loss": 0.5235, "step": 460 }, { "grad_norm": 1.2884926795959473, "learning_rate": 3.126666666666666e-05, "loss": 0.5011, "step": 470 }, { "grad_norm": 1.5131874084472656, "learning_rate": 3.1933333333333335e-05, "loss": 0.4806, "step": 480 }, { "grad_norm": 1.7456845045089722, "learning_rate": 3.26e-05, "loss": 0.4766, "step": 490 }, { "grad_norm": 1.290762186050415, "learning_rate": 3.326666666666667e-05, "loss": 0.4668, "step": 500 }, { "grad_norm": 1.4464728832244873, "learning_rate": 3.3933333333333336e-05, "loss": 0.4455, "step": 510 }, { "grad_norm": 1.547997236251831, "learning_rate": 3.46e-05, "loss": 0.4305, "step": 520 }, { "grad_norm": 1.3252315521240234, "learning_rate": 3.526666666666667e-05, "loss": 0.4239, "step": 530 }, { "grad_norm": 1.1750408411026, "learning_rate": 3.593333333333334e-05, "loss": 0.4065, "step": 540 }, { "grad_norm": 1.3642805814743042, "learning_rate": 3.66e-05, "loss": 0.3878, "step": 550 }, { "grad_norm": 1.362075924873352, "learning_rate": 3.726666666666667e-05, "loss": 0.3763, "step": 560 }, { "grad_norm": 1.200732707977295, "learning_rate": 3.793333333333334e-05, "loss": 0.3648, "step": 570 }, { "grad_norm": 1.3637202978134155, "learning_rate": 3.86e-05, "loss": 0.3589, "step": 580 }, { "grad_norm": 1.5157769918441772, "learning_rate": 3.926666666666667e-05, "loss": 0.3449, "step": 590 }, { "grad_norm": 1.4462758302688599, "learning_rate": 3.993333333333333e-05, "loss": 0.3389, "step": 600 }, { "grad_norm": 1.339345097541809, "learning_rate": 4.0600000000000004e-05, "loss": 0.3238, "step": 610 }, { "grad_norm": 2.016326904296875, "learning_rate": 4.126666666666667e-05, "loss": 0.3169, "step": 620 }, { "grad_norm": 1.6373475790023804, "learning_rate": 4.1933333333333334e-05, "loss": 0.303, "step": 630 }, { "grad_norm": 1.732764482498169, "learning_rate": 4.26e-05, "loss": 0.2915, "step": 640 }, { "grad_norm": 1.7197413444519043, "learning_rate": 4.3266666666666664e-05, "loss": 0.2943, "step": 650 }, { "grad_norm": 1.338645339012146, "learning_rate": 4.3933333333333335e-05, "loss": 0.2631, "step": 660 }, { "grad_norm": 1.491758942604065, "learning_rate": 4.46e-05, "loss": 0.2522, "step": 670 }, { "grad_norm": 1.3662916421890259, "learning_rate": 4.526666666666667e-05, "loss": 0.2552, "step": 680 }, { "grad_norm": 1.3540022373199463, "learning_rate": 4.5933333333333336e-05, "loss": 0.2568, "step": 690 }, { "grad_norm": 1.55057692527771, "learning_rate": 4.660000000000001e-05, "loss": 0.2441, "step": 700 }, { "grad_norm": 1.7688096761703491, "learning_rate": 4.726666666666667e-05, "loss": 0.253, "step": 710 }, { "grad_norm": 1.326645016670227, "learning_rate": 4.793333333333334e-05, "loss": 0.2426, "step": 720 }, { "grad_norm": 1.3348032236099243, "learning_rate": 4.86e-05, "loss": 0.2277, "step": 730 }, { "grad_norm": 1.5159469842910767, "learning_rate": 4.926666666666667e-05, "loss": 0.2176, "step": 740 }, { "grad_norm": 1.377224326133728, "learning_rate": 4.993333333333334e-05, "loss": 0.2212, "step": 750 }, { "grad_norm": 1.3685816526412964, "learning_rate": 5.0600000000000003e-05, "loss": 0.2195, "step": 760 }, { "grad_norm": 1.5215355157852173, "learning_rate": 5.1266666666666675e-05, "loss": 0.2026, "step": 770 }, { "grad_norm": 1.3709287643432617, "learning_rate": 5.193333333333333e-05, "loss": 0.2019, "step": 780 }, { "grad_norm": 1.227051854133606, "learning_rate": 5.2600000000000005e-05, "loss": 0.1873, "step": 790 }, { "grad_norm": 1.3648353815078735, "learning_rate": 5.326666666666666e-05, "loss": 0.2003, "step": 800 }, { "grad_norm": 1.4473228454589844, "learning_rate": 5.3933333333333334e-05, "loss": 0.1889, "step": 810 }, { "grad_norm": 1.3640116453170776, "learning_rate": 5.4600000000000006e-05, "loss": 0.1858, "step": 820 }, { "grad_norm": 1.2531596422195435, "learning_rate": 5.5266666666666664e-05, "loss": 0.1964, "step": 830 }, { "grad_norm": 1.5175758600234985, "learning_rate": 5.5933333333333335e-05, "loss": 0.176, "step": 840 }, { "grad_norm": 1.1909493207931519, "learning_rate": 5.66e-05, "loss": 0.1808, "step": 850 }, { "grad_norm": 1.8343007564544678, "learning_rate": 5.726666666666667e-05, "loss": 0.1863, "step": 860 }, { "grad_norm": 1.43864905834198, "learning_rate": 5.7933333333333337e-05, "loss": 0.1706, "step": 870 }, { "grad_norm": 1.1080561876296997, "learning_rate": 5.86e-05, "loss": 0.162, "step": 880 }, { "grad_norm": 1.1417601108551025, "learning_rate": 5.926666666666667e-05, "loss": 0.1718, "step": 890 }, { "grad_norm": 1.504669427871704, "learning_rate": 5.9933333333333345e-05, "loss": 0.1688, "step": 900 }, { "grad_norm": 1.2780033349990845, "learning_rate": 6.06e-05, "loss": 0.1722, "step": 910 }, { "grad_norm": 1.3772457838058472, "learning_rate": 6.126666666666667e-05, "loss": 0.1594, "step": 920 }, { "grad_norm": 1.2525526285171509, "learning_rate": 6.193333333333333e-05, "loss": 0.1613, "step": 930 }, { "grad_norm": 1.428037405014038, "learning_rate": 6.26e-05, "loss": 0.1546, "step": 940 }, { "grad_norm": 1.4083114862442017, "learning_rate": 6.326666666666667e-05, "loss": 0.1521, "step": 950 }, { "grad_norm": 1.304903268814087, "learning_rate": 6.393333333333333e-05, "loss": 0.1598, "step": 960 }, { "grad_norm": 1.5561751127243042, "learning_rate": 6.460000000000001e-05, "loss": 0.1388, "step": 970 }, { "grad_norm": 1.298207402229309, "learning_rate": 6.526666666666666e-05, "loss": 0.1394, "step": 980 }, { "grad_norm": 0.8742656111717224, "learning_rate": 6.593333333333334e-05, "loss": 0.1296, "step": 990 }, { "grad_norm": 1.232236623764038, "learning_rate": 6.66e-05, "loss": 0.1315, "step": 1000 }, { "grad_norm": 1.330221176147461, "learning_rate": 6.726666666666667e-05, "loss": 0.1283, "step": 1010 }, { "grad_norm": 0.9815651774406433, "learning_rate": 6.793333333333334e-05, "loss": 0.1288, "step": 1020 }, { "grad_norm": 1.4386731386184692, "learning_rate": 6.860000000000001e-05, "loss": 0.1175, "step": 1030 }, { "grad_norm": 1.421669363975525, "learning_rate": 6.926666666666667e-05, "loss": 0.1088, "step": 1040 }, { "grad_norm": 1.409855604171753, "learning_rate": 6.993333333333334e-05, "loss": 0.1167, "step": 1050 }, { "grad_norm": 1.2262802124023438, "learning_rate": 7.06e-05, "loss": 0.108, "step": 1060 }, { "grad_norm": 1.3135770559310913, "learning_rate": 7.126666666666667e-05, "loss": 0.1053, "step": 1070 }, { "grad_norm": 1.0767534971237183, "learning_rate": 7.193333333333334e-05, "loss": 0.1053, "step": 1080 }, { "grad_norm": 1.4627511501312256, "learning_rate": 7.26e-05, "loss": 0.1013, "step": 1090 }, { "grad_norm": 1.1182359457015991, "learning_rate": 7.326666666666667e-05, "loss": 0.0987, "step": 1100 }, { "grad_norm": 1.4287700653076172, "learning_rate": 7.393333333333333e-05, "loss": 0.1016, "step": 1110 }, { "grad_norm": 1.1720677614212036, "learning_rate": 7.46e-05, "loss": 0.0884, "step": 1120 }, { "grad_norm": 2.1559152603149414, "learning_rate": 7.526666666666668e-05, "loss": 0.1021, "step": 1130 }, { "grad_norm": 1.0774930715560913, "learning_rate": 7.593333333333334e-05, "loss": 0.0914, "step": 1140 }, { "grad_norm": 0.9638585448265076, "learning_rate": 7.66e-05, "loss": 0.0762, "step": 1150 }, { "grad_norm": 1.2960810661315918, "learning_rate": 7.726666666666667e-05, "loss": 0.0732, "step": 1160 }, { "grad_norm": 0.9604659676551819, "learning_rate": 7.793333333333333e-05, "loss": 0.0693, "step": 1170 }, { "grad_norm": 1.3396546840667725, "learning_rate": 7.860000000000001e-05, "loss": 0.0859, "step": 1180 }, { "grad_norm": 1.0190072059631348, "learning_rate": 7.926666666666666e-05, "loss": 0.0824, "step": 1190 }, { "grad_norm": 1.1840877532958984, "learning_rate": 7.993333333333334e-05, "loss": 0.078, "step": 1200 }, { "grad_norm": 1.2197439670562744, "learning_rate": 8.060000000000001e-05, "loss": 0.0868, "step": 1210 }, { "grad_norm": 1.2138742208480835, "learning_rate": 8.126666666666667e-05, "loss": 0.082, "step": 1220 }, { "grad_norm": 0.9524631500244141, "learning_rate": 8.193333333333334e-05, "loss": 0.0784, "step": 1230 }, { "grad_norm": 1.0575062036514282, "learning_rate": 8.26e-05, "loss": 0.0745, "step": 1240 }, { "grad_norm": 1.0756539106369019, "learning_rate": 8.326666666666667e-05, "loss": 0.0768, "step": 1250 }, { "grad_norm": 1.0881471633911133, "learning_rate": 8.393333333333335e-05, "loss": 0.0715, "step": 1260 }, { "grad_norm": 1.2430214881896973, "learning_rate": 8.46e-05, "loss": 0.0778, "step": 1270 }, { "grad_norm": 1.212658166885376, "learning_rate": 8.526666666666667e-05, "loss": 0.0769, "step": 1280 }, { "grad_norm": 1.1884628534317017, "learning_rate": 8.593333333333333e-05, "loss": 0.0755, "step": 1290 }, { "grad_norm": 0.9515352845191956, "learning_rate": 8.66e-05, "loss": 0.0674, "step": 1300 }, { "grad_norm": 0.9969962239265442, "learning_rate": 8.726666666666667e-05, "loss": 0.0732, "step": 1310 }, { "grad_norm": 1.1777493953704834, "learning_rate": 8.793333333333333e-05, "loss": 0.0721, "step": 1320 }, { "grad_norm": 0.979246199131012, "learning_rate": 8.86e-05, "loss": 0.074, "step": 1330 }, { "grad_norm": 1.0967730283737183, "learning_rate": 8.926666666666668e-05, "loss": 0.072, "step": 1340 }, { "grad_norm": 1.1403131484985352, "learning_rate": 8.993333333333334e-05, "loss": 0.0664, "step": 1350 }, { "grad_norm": 0.9110641479492188, "learning_rate": 9.06e-05, "loss": 0.0697, "step": 1360 }, { "grad_norm": 0.9065604209899902, "learning_rate": 9.126666666666667e-05, "loss": 0.0776, "step": 1370 }, { "grad_norm": 1.1146701574325562, "learning_rate": 9.193333333333334e-05, "loss": 0.0694, "step": 1380 }, { "grad_norm": 0.9170656204223633, "learning_rate": 9.260000000000001e-05, "loss": 0.0726, "step": 1390 }, { "grad_norm": 1.0038901567459106, "learning_rate": 9.326666666666667e-05, "loss": 0.0691, "step": 1400 }, { "grad_norm": 1.0873206853866577, "learning_rate": 9.393333333333334e-05, "loss": 0.0727, "step": 1410 }, { "grad_norm": 1.0876661539077759, "learning_rate": 9.46e-05, "loss": 0.0636, "step": 1420 }, { "grad_norm": 1.0423922538757324, "learning_rate": 9.526666666666667e-05, "loss": 0.0648, "step": 1430 }, { "grad_norm": 1.0222643613815308, "learning_rate": 9.593333333333334e-05, "loss": 0.0618, "step": 1440 }, { "grad_norm": 0.879751443862915, "learning_rate": 9.66e-05, "loss": 0.0624, "step": 1450 }, { "grad_norm": 0.8119634985923767, "learning_rate": 9.726666666666667e-05, "loss": 0.0524, "step": 1460 }, { "grad_norm": 0.9365761876106262, "learning_rate": 9.793333333333333e-05, "loss": 0.0603, "step": 1470 }, { "grad_norm": 0.8556691408157349, "learning_rate": 9.86e-05, "loss": 0.0559, "step": 1480 }, { "grad_norm": 0.9922760128974915, "learning_rate": 9.926666666666668e-05, "loss": 0.0623, "step": 1490 }, { "grad_norm": 0.9966976046562195, "learning_rate": 9.993333333333334e-05, "loss": 0.0621, "step": 1500 }, { "grad_norm": 1.1150485277175903, "learning_rate": 9.999997539434007e-05, "loss": 0.0682, "step": 1510 }, { "grad_norm": 0.7289571762084961, "learning_rate": 9.999989033776898e-05, "loss": 0.0669, "step": 1520 }, { "grad_norm": 0.9980589747428894, "learning_rate": 9.999974452661641e-05, "loss": 0.0687, "step": 1530 }, { "grad_norm": 0.9087761640548706, "learning_rate": 9.999953796105959e-05, "loss": 0.0656, "step": 1540 }, { "grad_norm": 0.8854475021362305, "learning_rate": 9.999927064134949e-05, "loss": 0.0637, "step": 1550 }, { "grad_norm": 0.8211564421653748, "learning_rate": 9.999894256781095e-05, "loss": 0.0621, "step": 1560 }, { "grad_norm": 0.9164798855781555, "learning_rate": 9.99985537408426e-05, "loss": 0.0682, "step": 1570 }, { "grad_norm": 0.6808768510818481, "learning_rate": 9.999810416091688e-05, "loss": 0.0666, "step": 1580 }, { "grad_norm": 0.8450937867164612, "learning_rate": 9.99975938285801e-05, "loss": 0.0596, "step": 1590 }, { "grad_norm": 0.9516972899436951, "learning_rate": 9.999702274445236e-05, "loss": 0.0749, "step": 1600 }, { "grad_norm": 0.7658268213272095, "learning_rate": 9.999639090922756e-05, "loss": 0.0611, "step": 1610 }, { "grad_norm": 0.8277539610862732, "learning_rate": 9.999569832367346e-05, "loss": 0.0577, "step": 1620 }, { "grad_norm": 0.8471164107322693, "learning_rate": 9.999494498863162e-05, "loss": 0.0588, "step": 1630 }, { "grad_norm": 0.8782401084899902, "learning_rate": 9.99941309050174e-05, "loss": 0.0618, "step": 1640 }, { "grad_norm": 0.9257912635803223, "learning_rate": 9.999325607381999e-05, "loss": 0.0706, "step": 1650 }, { "grad_norm": 0.7844532132148743, "learning_rate": 9.999232049610238e-05, "loss": 0.0583, "step": 1660 }, { "grad_norm": 0.6863495707511902, "learning_rate": 9.999132417300142e-05, "loss": 0.0567, "step": 1670 }, { "grad_norm": 0.8271988034248352, "learning_rate": 9.99902671057277e-05, "loss": 0.0553, "step": 1680 }, { "grad_norm": 1.0046560764312744, "learning_rate": 9.998914929556569e-05, "loss": 0.0527, "step": 1690 }, { "grad_norm": 0.8451791405677795, "learning_rate": 9.998797074387361e-05, "loss": 0.0531, "step": 1700 }, { "grad_norm": 0.8770014047622681, "learning_rate": 9.99867314520835e-05, "loss": 0.058, "step": 1710 }, { "grad_norm": 0.710878849029541, "learning_rate": 9.998543142170126e-05, "loss": 0.0484, "step": 1720 }, { "grad_norm": 0.8275458216667175, "learning_rate": 9.99840706543065e-05, "loss": 0.0506, "step": 1730 }, { "grad_norm": 0.7862157821655273, "learning_rate": 9.998264915155274e-05, "loss": 0.05, "step": 1740 }, { "grad_norm": 0.6714426279067993, "learning_rate": 9.998116691516718e-05, "loss": 0.0524, "step": 1750 }, { "grad_norm": 0.9966657757759094, "learning_rate": 9.997962394695091e-05, "loss": 0.0611, "step": 1760 }, { "grad_norm": 0.7245919704437256, "learning_rate": 9.997802024877875e-05, "loss": 0.0466, "step": 1770 }, { "grad_norm": 0.8870209455490112, "learning_rate": 9.99763558225994e-05, "loss": 0.0536, "step": 1780 }, { "grad_norm": 0.8304436802864075, "learning_rate": 9.997463067043526e-05, "loss": 0.0594, "step": 1790 }, { "grad_norm": 0.9695459604263306, "learning_rate": 9.997284479438253e-05, "loss": 0.0608, "step": 1800 }, { "grad_norm": 0.9971867799758911, "learning_rate": 9.997099819661127e-05, "loss": 0.0533, "step": 1810 }, { "grad_norm": 0.6639770269393921, "learning_rate": 9.996909087936524e-05, "loss": 0.0538, "step": 1820 }, { "grad_norm": 0.7827335000038147, "learning_rate": 9.996712284496201e-05, "loss": 0.0518, "step": 1830 }, { "grad_norm": 0.8153265118598938, "learning_rate": 9.996509409579293e-05, "loss": 0.058, "step": 1840 }, { "grad_norm": 0.6077260375022888, "learning_rate": 9.996300463432312e-05, "loss": 0.05, "step": 1850 }, { "grad_norm": 0.7871149182319641, "learning_rate": 9.996085446309148e-05, "loss": 0.0462, "step": 1860 }, { "grad_norm": 0.7612835764884949, "learning_rate": 9.995864358471066e-05, "loss": 0.0527, "step": 1870 }, { "grad_norm": 0.8559739589691162, "learning_rate": 9.99563720018671e-05, "loss": 0.0534, "step": 1880 }, { "grad_norm": 0.8805252313613892, "learning_rate": 9.995403971732098e-05, "loss": 0.0529, "step": 1890 }, { "grad_norm": 0.8515594005584717, "learning_rate": 9.995164673390625e-05, "loss": 0.0488, "step": 1900 }, { "grad_norm": 0.8210208415985107, "learning_rate": 9.994919305453059e-05, "loss": 0.0581, "step": 1910 }, { "grad_norm": 0.7993607521057129, "learning_rate": 9.994667868217548e-05, "loss": 0.0533, "step": 1920 }, { "grad_norm": 0.7146683931350708, "learning_rate": 9.99441036198961e-05, "loss": 0.0492, "step": 1930 }, { "grad_norm": 0.7025575041770935, "learning_rate": 9.99414678708214e-05, "loss": 0.0456, "step": 1940 }, { "grad_norm": 0.5726608037948608, "learning_rate": 9.993877143815407e-05, "loss": 0.046, "step": 1950 }, { "grad_norm": 0.6866472959518433, "learning_rate": 9.993601432517053e-05, "loss": 0.0434, "step": 1960 }, { "grad_norm": 0.7527312636375427, "learning_rate": 9.993319653522091e-05, "loss": 0.0443, "step": 1970 }, { "grad_norm": 0.9628674983978271, "learning_rate": 9.993031807172911e-05, "loss": 0.0508, "step": 1980 }, { "grad_norm": 0.746136486530304, "learning_rate": 9.992737893819273e-05, "loss": 0.0483, "step": 1990 }, { "grad_norm": 0.9710631966590881, "learning_rate": 9.992437913818312e-05, "loss": 0.0542, "step": 2000 }, { "grad_norm": 0.7973635196685791, "learning_rate": 9.992131867534526e-05, "loss": 0.0465, "step": 2010 }, { "grad_norm": 0.8818680047988892, "learning_rate": 9.991819755339796e-05, "loss": 0.0515, "step": 2020 }, { "grad_norm": 0.9944676160812378, "learning_rate": 9.991501577613365e-05, "loss": 0.0781, "step": 2030 }, { "grad_norm": 0.7986236810684204, "learning_rate": 9.99117733474185e-05, "loss": 0.0671, "step": 2040 }, { "grad_norm": 0.8032518029212952, "learning_rate": 9.990847027119234e-05, "loss": 0.0582, "step": 2050 }, { "grad_norm": 0.6630721688270569, "learning_rate": 9.990510655146877e-05, "loss": 0.057, "step": 2060 }, { "grad_norm": 0.7591819763183594, "learning_rate": 9.990168219233496e-05, "loss": 0.0518, "step": 2070 }, { "grad_norm": 0.8523391485214233, "learning_rate": 9.989819719795188e-05, "loss": 0.0529, "step": 2080 }, { "grad_norm": 0.5780413150787354, "learning_rate": 9.989465157255412e-05, "loss": 0.0453, "step": 2090 }, { "grad_norm": 0.7530731558799744, "learning_rate": 9.989104532044994e-05, "loss": 0.0529, "step": 2100 }, { "grad_norm": 0.779677152633667, "learning_rate": 9.988737844602128e-05, "loss": 0.0555, "step": 2110 }, { "grad_norm": 0.8249015212059021, "learning_rate": 9.988365095372372e-05, "loss": 0.053, "step": 2120 }, { "grad_norm": 0.5712619423866272, "learning_rate": 9.987986284808654e-05, "loss": 0.0458, "step": 2130 }, { "grad_norm": 0.6727279424667358, "learning_rate": 9.987601413371264e-05, "loss": 0.0417, "step": 2140 }, { "grad_norm": 0.64397794008255, "learning_rate": 9.987210481527855e-05, "loss": 0.0429, "step": 2150 }, { "grad_norm": 0.7556050419807434, "learning_rate": 9.98681348975345e-05, "loss": 0.0523, "step": 2160 }, { "grad_norm": 0.808230459690094, "learning_rate": 9.986410438530427e-05, "loss": 0.0545, "step": 2170 }, { "grad_norm": 0.7168055772781372, "learning_rate": 9.986001328348534e-05, "loss": 0.0482, "step": 2180 }, { "grad_norm": 0.7944677472114563, "learning_rate": 9.985586159704878e-05, "loss": 0.0451, "step": 2190 }, { "grad_norm": 0.6046855449676514, "learning_rate": 9.985164933103929e-05, "loss": 0.0443, "step": 2200 }, { "grad_norm": 0.6509060859680176, "learning_rate": 9.984737649057513e-05, "loss": 0.0416, "step": 2210 }, { "grad_norm": 0.497623473405838, "learning_rate": 9.984304308084827e-05, "loss": 0.0394, "step": 2220 }, { "grad_norm": 0.6639490723609924, "learning_rate": 9.983864910712416e-05, "loss": 0.0405, "step": 2230 }, { "grad_norm": 0.6353468298912048, "learning_rate": 9.98341945747419e-05, "loss": 0.0402, "step": 2240 }, { "grad_norm": 0.5243775248527527, "learning_rate": 9.98296794891142e-05, "loss": 0.0408, "step": 2250 }, { "grad_norm": 0.7712529301643372, "learning_rate": 9.982510385572725e-05, "loss": 0.0385, "step": 2260 }, { "grad_norm": 0.6628357768058777, "learning_rate": 9.982046768014094e-05, "loss": 0.0434, "step": 2270 }, { "grad_norm": 0.7228386998176575, "learning_rate": 9.981577096798863e-05, "loss": 0.0419, "step": 2280 }, { "grad_norm": 0.7136985659599304, "learning_rate": 9.981101372497727e-05, "loss": 0.0536, "step": 2290 }, { "grad_norm": 0.6510632634162903, "learning_rate": 9.980619595688737e-05, "loss": 0.0398, "step": 2300 }, { "grad_norm": 0.5880825519561768, "learning_rate": 9.980131766957295e-05, "loss": 0.0437, "step": 2310 }, { "grad_norm": 0.6415541768074036, "learning_rate": 9.979637886896163e-05, "loss": 0.0375, "step": 2320 }, { "grad_norm": 0.6707743406295776, "learning_rate": 9.979137956105447e-05, "loss": 0.0465, "step": 2330 }, { "grad_norm": 0.7156596779823303, "learning_rate": 9.978631975192613e-05, "loss": 0.0441, "step": 2340 }, { "grad_norm": 0.6110060214996338, "learning_rate": 9.978119944772475e-05, "loss": 0.0436, "step": 2350 }, { "grad_norm": 0.6668084859848022, "learning_rate": 9.977601865467197e-05, "loss": 0.0388, "step": 2360 }, { "grad_norm": 0.5930837392807007, "learning_rate": 9.977077737906297e-05, "loss": 0.042, "step": 2370 }, { "grad_norm": 0.73235684633255, "learning_rate": 9.976547562726636e-05, "loss": 0.0411, "step": 2380 }, { "grad_norm": 0.7019467353820801, "learning_rate": 9.976011340572429e-05, "loss": 0.0385, "step": 2390 }, { "grad_norm": 0.736674427986145, "learning_rate": 9.975469072095237e-05, "loss": 0.045, "step": 2400 }, { "grad_norm": 0.770882785320282, "learning_rate": 9.974920757953965e-05, "loss": 0.0443, "step": 2410 }, { "grad_norm": 0.6460158824920654, "learning_rate": 9.97436639881487e-05, "loss": 0.0429, "step": 2420 }, { "grad_norm": 0.5886706113815308, "learning_rate": 9.973805995351545e-05, "loss": 0.0467, "step": 2430 }, { "grad_norm": 0.7373203635215759, "learning_rate": 9.973239548244939e-05, "loss": 0.0485, "step": 2440 }, { "grad_norm": 0.6167523264884949, "learning_rate": 9.972667058183333e-05, "loss": 0.0379, "step": 2450 }, { "grad_norm": 0.6162266731262207, "learning_rate": 9.972088525862362e-05, "loss": 0.047, "step": 2460 }, { "grad_norm": 0.5308279395103455, "learning_rate": 9.971503951984995e-05, "loss": 0.0414, "step": 2470 }, { "grad_norm": 0.6375431418418884, "learning_rate": 9.970913337261543e-05, "loss": 0.0471, "step": 2480 }, { "grad_norm": 0.7652308344841003, "learning_rate": 9.97031668240966e-05, "loss": 0.0468, "step": 2490 }, { "grad_norm": 0.8610413670539856, "learning_rate": 9.969713988154339e-05, "loss": 0.0425, "step": 2500 }, { "grad_norm": 0.8011583685874939, "learning_rate": 9.969105255227906e-05, "loss": 0.0436, "step": 2510 }, { "grad_norm": 0.5567802786827087, "learning_rate": 9.968490484370035e-05, "loss": 0.0399, "step": 2520 }, { "grad_norm": 0.5165965557098389, "learning_rate": 9.967869676327726e-05, "loss": 0.0417, "step": 2530 }, { "grad_norm": 0.5604172348976135, "learning_rate": 9.96724283185532e-05, "loss": 0.0385, "step": 2540 }, { "grad_norm": 0.6940800547599792, "learning_rate": 9.966609951714494e-05, "loss": 0.0415, "step": 2550 }, { "grad_norm": 0.8311217427253723, "learning_rate": 9.965971036674255e-05, "loss": 0.0432, "step": 2560 }, { "grad_norm": 0.6736733913421631, "learning_rate": 9.965326087510947e-05, "loss": 0.0385, "step": 2570 }, { "grad_norm": 0.5931285619735718, "learning_rate": 9.964675105008244e-05, "loss": 0.0403, "step": 2580 }, { "grad_norm": 0.6911582350730896, "learning_rate": 9.964018089957147e-05, "loss": 0.0362, "step": 2590 }, { "grad_norm": 0.5567492842674255, "learning_rate": 9.963355043155996e-05, "loss": 0.0418, "step": 2600 }, { "grad_norm": 0.44991132616996765, "learning_rate": 9.962685965410455e-05, "loss": 0.0374, "step": 2610 }, { "grad_norm": 0.7982749938964844, "learning_rate": 9.962010857533514e-05, "loss": 0.0398, "step": 2620 }, { "grad_norm": 0.5765446424484253, "learning_rate": 9.961329720345493e-05, "loss": 0.0445, "step": 2630 }, { "grad_norm": 0.5843945145606995, "learning_rate": 9.96064255467404e-05, "loss": 0.0384, "step": 2640 }, { "grad_norm": 0.5728034377098083, "learning_rate": 9.959949361354126e-05, "loss": 0.0395, "step": 2650 }, { "grad_norm": 0.6547420620918274, "learning_rate": 9.959250141228045e-05, "loss": 0.0382, "step": 2660 }, { "grad_norm": 0.5894904732704163, "learning_rate": 9.958544895145414e-05, "loss": 0.0417, "step": 2670 }, { "grad_norm": 0.7364691495895386, "learning_rate": 9.957833623963177e-05, "loss": 0.0405, "step": 2680 }, { "grad_norm": 0.6655886173248291, "learning_rate": 9.957116328545593e-05, "loss": 0.0388, "step": 2690 }, { "grad_norm": 0.4601716697216034, "learning_rate": 9.956393009764244e-05, "loss": 0.035, "step": 2700 }, { "grad_norm": 0.7445112466812134, "learning_rate": 9.955663668498032e-05, "loss": 0.0403, "step": 2710 }, { "grad_norm": 0.6575505137443542, "learning_rate": 9.954928305633173e-05, "loss": 0.0461, "step": 2720 }, { "grad_norm": 0.5855975151062012, "learning_rate": 9.954186922063204e-05, "loss": 0.0411, "step": 2730 }, { "grad_norm": 0.48124492168426514, "learning_rate": 9.953439518688974e-05, "loss": 0.0353, "step": 2740 }, { "grad_norm": 0.5937489867210388, "learning_rate": 9.952686096418652e-05, "loss": 0.0394, "step": 2750 }, { "grad_norm": 0.515085756778717, "learning_rate": 9.951926656167714e-05, "loss": 0.0384, "step": 2760 }, { "grad_norm": 0.5337117910385132, "learning_rate": 9.951161198858953e-05, "loss": 0.0345, "step": 2770 }, { "grad_norm": 0.5596720576286316, "learning_rate": 9.950389725422471e-05, "loss": 0.0359, "step": 2780 }, { "grad_norm": 0.5601538419723511, "learning_rate": 9.949612236795682e-05, "loss": 0.037, "step": 2790 }, { "grad_norm": 0.6343134641647339, "learning_rate": 9.948828733923305e-05, "loss": 0.0339, "step": 2800 }, { "grad_norm": 0.5116497278213501, "learning_rate": 9.948039217757374e-05, "loss": 0.038, "step": 2810 }, { "grad_norm": 0.5677695870399475, "learning_rate": 9.947243689257225e-05, "loss": 0.031, "step": 2820 }, { "grad_norm": 0.7071871161460876, "learning_rate": 9.946442149389497e-05, "loss": 0.0469, "step": 2830 }, { "grad_norm": 0.6054574251174927, "learning_rate": 9.945634599128139e-05, "loss": 0.0424, "step": 2840 }, { "grad_norm": 0.5489897727966309, "learning_rate": 9.944821039454402e-05, "loss": 0.0379, "step": 2850 }, { "grad_norm": 0.6514557600021362, "learning_rate": 9.944001471356835e-05, "loss": 0.0443, "step": 2860 }, { "grad_norm": 0.6226890683174133, "learning_rate": 9.94317589583129e-05, "loss": 0.0372, "step": 2870 }, { "grad_norm": 0.663642406463623, "learning_rate": 9.942344313880922e-05, "loss": 0.0416, "step": 2880 }, { "grad_norm": 0.7010625004768372, "learning_rate": 9.941506726516179e-05, "loss": 0.0399, "step": 2890 }, { "grad_norm": 0.5899003744125366, "learning_rate": 9.94066313475481e-05, "loss": 0.0367, "step": 2900 }, { "grad_norm": 0.5501528978347778, "learning_rate": 9.939813539621857e-05, "loss": 0.0358, "step": 2910 }, { "grad_norm": 0.7116836905479431, "learning_rate": 9.93895794214966e-05, "loss": 0.0393, "step": 2920 }, { "grad_norm": 0.58006751537323, "learning_rate": 9.938096343377852e-05, "loss": 0.032, "step": 2930 }, { "grad_norm": 0.5966971516609192, "learning_rate": 9.937228744353353e-05, "loss": 0.0359, "step": 2940 }, { "grad_norm": 0.585063099861145, "learning_rate": 9.936355146130379e-05, "loss": 0.0356, "step": 2950 }, { "grad_norm": 0.6770583391189575, "learning_rate": 9.935475549770435e-05, "loss": 0.0358, "step": 2960 }, { "grad_norm": 0.5591983199119568, "learning_rate": 9.934589956342315e-05, "loss": 0.0372, "step": 2970 }, { "grad_norm": 0.632769763469696, "learning_rate": 9.933698366922093e-05, "loss": 0.0333, "step": 2980 }, { "grad_norm": 0.588104784488678, "learning_rate": 9.93280078259314e-05, "loss": 0.036, "step": 2990 }, { "grad_norm": 0.5254931449890137, "learning_rate": 9.931897204446104e-05, "loss": 0.0375, "step": 3000 }, { "grad_norm": 0.5712432861328125, "learning_rate": 9.930987633578915e-05, "loss": 0.0354, "step": 3010 }, { "grad_norm": 0.5211623907089233, "learning_rate": 9.93007207109679e-05, "loss": 0.0329, "step": 3020 }, { "grad_norm": 0.5415829420089722, "learning_rate": 9.929150518112224e-05, "loss": 0.0359, "step": 3030 }, { "grad_norm": 0.705829381942749, "learning_rate": 9.928222975744991e-05, "loss": 0.0356, "step": 3040 }, { "grad_norm": 0.543723464012146, "learning_rate": 9.92728944512214e-05, "loss": 0.0365, "step": 3050 }, { "grad_norm": 0.4913904666900635, "learning_rate": 9.926349927378001e-05, "loss": 0.0348, "step": 3060 }, { "grad_norm": 0.5797699093818665, "learning_rate": 9.925404423654174e-05, "loss": 0.0366, "step": 3070 }, { "grad_norm": 0.4701583981513977, "learning_rate": 9.924452935099537e-05, "loss": 0.0347, "step": 3080 }, { "grad_norm": 0.5725277662277222, "learning_rate": 9.92349546287024e-05, "loss": 0.0352, "step": 3090 }, { "grad_norm": 0.6676095128059387, "learning_rate": 9.9225320081297e-05, "loss": 0.0416, "step": 3100 }, { "grad_norm": 0.5679190754890442, "learning_rate": 9.921562572048606e-05, "loss": 0.0418, "step": 3110 }, { "grad_norm": 0.8147238492965698, "learning_rate": 9.920587155804913e-05, "loss": 0.0361, "step": 3120 }, { "grad_norm": 0.5758320093154907, "learning_rate": 9.919605760583845e-05, "loss": 0.035, "step": 3130 }, { "grad_norm": 0.5047492980957031, "learning_rate": 9.91861838757789e-05, "loss": 0.0368, "step": 3140 }, { "grad_norm": 0.6605943441390991, "learning_rate": 9.917625037986798e-05, "loss": 0.0372, "step": 3150 }, { "grad_norm": 0.5971742868423462, "learning_rate": 9.916625713017583e-05, "loss": 0.0381, "step": 3160 }, { "grad_norm": 0.5885633230209351, "learning_rate": 9.915620413884519e-05, "loss": 0.0381, "step": 3170 }, { "grad_norm": 0.6142313480377197, "learning_rate": 9.914609141809139e-05, "loss": 0.0344, "step": 3180 }, { "grad_norm": 0.4758969843387604, "learning_rate": 9.913591898020235e-05, "loss": 0.0311, "step": 3190 }, { "grad_norm": 0.6770864129066467, "learning_rate": 9.912568683753853e-05, "loss": 0.0371, "step": 3200 }, { "grad_norm": 0.4663007855415344, "learning_rate": 9.911539500253295e-05, "loss": 0.0321, "step": 3210 }, { "grad_norm": 0.5785437226295471, "learning_rate": 9.910504348769118e-05, "loss": 0.0405, "step": 3220 }, { "grad_norm": 0.5366180539131165, "learning_rate": 9.909463230559127e-05, "loss": 0.0347, "step": 3230 }, { "grad_norm": 0.4831130802631378, "learning_rate": 9.908416146888376e-05, "loss": 0.0328, "step": 3240 }, { "grad_norm": 0.4668837785720825, "learning_rate": 9.907363099029175e-05, "loss": 0.0337, "step": 3250 }, { "grad_norm": 0.537135660648346, "learning_rate": 9.906304088261073e-05, "loss": 0.0408, "step": 3260 }, { "grad_norm": 0.7323784232139587, "learning_rate": 9.905239115870872e-05, "loss": 0.0413, "step": 3270 }, { "grad_norm": 0.46636152267456055, "learning_rate": 9.90416818315261e-05, "loss": 0.0335, "step": 3280 }, { "grad_norm": 0.4953058063983917, "learning_rate": 9.903091291407573e-05, "loss": 0.0338, "step": 3290 }, { "grad_norm": 0.4598117470741272, "learning_rate": 9.902008441944286e-05, "loss": 0.0338, "step": 3300 }, { "grad_norm": 0.8054019808769226, "learning_rate": 9.900919636078512e-05, "loss": 0.0373, "step": 3310 }, { "grad_norm": 0.5435279607772827, "learning_rate": 9.899824875133255e-05, "loss": 0.039, "step": 3320 }, { "grad_norm": 0.5638160705566406, "learning_rate": 9.898724160438749e-05, "loss": 0.0341, "step": 3330 }, { "grad_norm": 0.6308696269989014, "learning_rate": 9.89761749333247e-05, "loss": 0.0385, "step": 3340 }, { "grad_norm": 0.6113548874855042, "learning_rate": 9.896504875159122e-05, "loss": 0.0365, "step": 3350 }, { "grad_norm": 0.5968590974807739, "learning_rate": 9.89538630727064e-05, "loss": 0.0349, "step": 3360 }, { "grad_norm": 0.6125024557113647, "learning_rate": 9.894261791026189e-05, "loss": 0.0321, "step": 3370 }, { "grad_norm": 0.7245036363601685, "learning_rate": 9.893131327792165e-05, "loss": 0.0367, "step": 3380 }, { "grad_norm": 0.530841052532196, "learning_rate": 9.891994918942182e-05, "loss": 0.0382, "step": 3390 }, { "grad_norm": 0.499847412109375, "learning_rate": 9.890852565857092e-05, "loss": 0.0305, "step": 3400 }, { "grad_norm": 0.667802095413208, "learning_rate": 9.889704269924954e-05, "loss": 0.0432, "step": 3410 }, { "grad_norm": 0.5614846348762512, "learning_rate": 9.888550032541059e-05, "loss": 0.0352, "step": 3420 }, { "grad_norm": 0.4696511924266815, "learning_rate": 9.887389855107916e-05, "loss": 0.0341, "step": 3430 }, { "grad_norm": 0.5708081126213074, "learning_rate": 9.886223739035248e-05, "loss": 0.0336, "step": 3440 }, { "grad_norm": 0.45526570081710815, "learning_rate": 9.885051685739997e-05, "loss": 0.0295, "step": 3450 }, { "grad_norm": 0.6285563111305237, "learning_rate": 9.883873696646316e-05, "loss": 0.0275, "step": 3460 }, { "grad_norm": 0.5453596711158752, "learning_rate": 9.882689773185575e-05, "loss": 0.0342, "step": 3470 }, { "grad_norm": 0.5551162958145142, "learning_rate": 9.881499916796353e-05, "loss": 0.0344, "step": 3480 }, { "grad_norm": 0.6420366764068604, "learning_rate": 9.880304128924434e-05, "loss": 0.0346, "step": 3490 }, { "grad_norm": 0.6295766234397888, "learning_rate": 9.879102411022817e-05, "loss": 0.0316, "step": 3500 }, { "grad_norm": 0.6235019564628601, "learning_rate": 9.877894764551703e-05, "loss": 0.0325, "step": 3510 }, { "grad_norm": 0.5211902260780334, "learning_rate": 9.876681190978494e-05, "loss": 0.0329, "step": 3520 }, { "grad_norm": 0.6439358592033386, "learning_rate": 9.875461691777797e-05, "loss": 0.0327, "step": 3530 }, { "grad_norm": 0.4277244806289673, "learning_rate": 9.874236268431417e-05, "loss": 0.0303, "step": 3540 }, { "grad_norm": 0.5787872672080994, "learning_rate": 9.873004922428361e-05, "loss": 0.031, "step": 3550 }, { "grad_norm": 0.6227014660835266, "learning_rate": 9.871767655264829e-05, "loss": 0.0328, "step": 3560 }, { "grad_norm": 0.47548621892929077, "learning_rate": 9.87052446844422e-05, "loss": 0.036, "step": 3570 }, { "grad_norm": 0.5219529867172241, "learning_rate": 9.869275363477122e-05, "loss": 0.0341, "step": 3580 }, { "grad_norm": 0.5856068730354309, "learning_rate": 9.868020341881312e-05, "loss": 0.0393, "step": 3590 }, { "grad_norm": 0.4280667006969452, "learning_rate": 9.866759405181765e-05, "loss": 0.0349, "step": 3600 }, { "grad_norm": 0.6325046420097351, "learning_rate": 9.865492554910633e-05, "loss": 0.0399, "step": 3610 }, { "grad_norm": 0.47228944301605225, "learning_rate": 9.864219792607262e-05, "loss": 0.0325, "step": 3620 }, { "grad_norm": 0.5280877351760864, "learning_rate": 9.862941119818177e-05, "loss": 0.0392, "step": 3630 }, { "grad_norm": 0.6615012288093567, "learning_rate": 9.861656538097086e-05, "loss": 0.0372, "step": 3640 }, { "grad_norm": 0.6585820913314819, "learning_rate": 9.860366049004877e-05, "loss": 0.0325, "step": 3650 }, { "grad_norm": 0.5302186012268066, "learning_rate": 9.859069654109615e-05, "loss": 0.032, "step": 3660 }, { "grad_norm": 0.5717152953147888, "learning_rate": 9.857767354986545e-05, "loss": 0.0354, "step": 3670 }, { "grad_norm": 0.49941059947013855, "learning_rate": 9.856459153218078e-05, "loss": 0.03, "step": 3680 }, { "grad_norm": 0.551819384098053, "learning_rate": 9.855145050393808e-05, "loss": 0.0367, "step": 3690 }, { "grad_norm": 0.443668395280838, "learning_rate": 9.85382504811049e-05, "loss": 0.0312, "step": 3700 }, { "grad_norm": 0.4996035695075989, "learning_rate": 9.852499147972054e-05, "loss": 0.0324, "step": 3710 }, { "grad_norm": 0.5133060812950134, "learning_rate": 9.851167351589592e-05, "loss": 0.0325, "step": 3720 }, { "grad_norm": 0.5872260332107544, "learning_rate": 9.849829660581363e-05, "loss": 0.04, "step": 3730 }, { "grad_norm": 0.5641419291496277, "learning_rate": 9.848486076572787e-05, "loss": 0.0345, "step": 3740 }, { "grad_norm": 0.5718939304351807, "learning_rate": 9.847136601196446e-05, "loss": 0.0324, "step": 3750 }, { "grad_norm": 0.4982635974884033, "learning_rate": 9.845781236092078e-05, "loss": 0.0325, "step": 3760 }, { "grad_norm": 0.5953066945075989, "learning_rate": 9.844419982906583e-05, "loss": 0.0366, "step": 3770 }, { "grad_norm": 0.5456868410110474, "learning_rate": 9.843052843294008e-05, "loss": 0.0357, "step": 3780 }, { "grad_norm": 0.4179612696170807, "learning_rate": 9.841679818915559e-05, "loss": 0.0354, "step": 3790 }, { "grad_norm": 0.513627290725708, "learning_rate": 9.840300911439591e-05, "loss": 0.0336, "step": 3800 }, { "grad_norm": 0.4293102025985718, "learning_rate": 9.838916122541603e-05, "loss": 0.0325, "step": 3810 }, { "grad_norm": 0.5595964789390564, "learning_rate": 9.837525453904246e-05, "loss": 0.0353, "step": 3820 }, { "grad_norm": 0.5100940465927124, "learning_rate": 9.836128907217314e-05, "loss": 0.0334, "step": 3830 }, { "grad_norm": 0.5661239624023438, "learning_rate": 9.834726484177743e-05, "loss": 0.0366, "step": 3840 }, { "grad_norm": 0.46519017219543457, "learning_rate": 9.833318186489609e-05, "loss": 0.0317, "step": 3850 }, { "grad_norm": 0.49180611968040466, "learning_rate": 9.831904015864126e-05, "loss": 0.0338, "step": 3860 }, { "grad_norm": 0.6366244554519653, "learning_rate": 9.830483974019645e-05, "loss": 0.0351, "step": 3870 }, { "grad_norm": 0.5490753054618835, "learning_rate": 9.82905806268165e-05, "loss": 0.0356, "step": 3880 }, { "grad_norm": 0.5102167725563049, "learning_rate": 9.82762628358276e-05, "loss": 0.0339, "step": 3890 }, { "grad_norm": 0.42410367727279663, "learning_rate": 9.826188638462718e-05, "loss": 0.0317, "step": 3900 }, { "grad_norm": 0.45005306601524353, "learning_rate": 9.824745129068402e-05, "loss": 0.0308, "step": 3910 }, { "grad_norm": 0.4748956859111786, "learning_rate": 9.82329575715381e-05, "loss": 0.0321, "step": 3920 }, { "grad_norm": 0.45965179800987244, "learning_rate": 9.821840524480066e-05, "loss": 0.0361, "step": 3930 }, { "grad_norm": 0.43910783529281616, "learning_rate": 9.820379432815414e-05, "loss": 0.0332, "step": 3940 }, { "grad_norm": 0.5560693144798279, "learning_rate": 9.81891248393522e-05, "loss": 0.0303, "step": 3950 }, { "grad_norm": 0.5932253003120422, "learning_rate": 9.817439679621963e-05, "loss": 0.0421, "step": 3960 }, { "grad_norm": 0.5707384347915649, "learning_rate": 9.815961021665243e-05, "loss": 0.0321, "step": 3970 }, { "grad_norm": 0.4231276214122772, "learning_rate": 9.814476511861763e-05, "loss": 0.0314, "step": 3980 }, { "grad_norm": 0.472631573677063, "learning_rate": 9.812986152015348e-05, "loss": 0.0368, "step": 3990 }, { "grad_norm": 0.5836194753646851, "learning_rate": 9.811489943936922e-05, "loss": 0.0335, "step": 4000 }, { "grad_norm": 0.48724037408828735, "learning_rate": 9.809987889444522e-05, "loss": 0.0332, "step": 4010 }, { "grad_norm": 0.5766056776046753, "learning_rate": 9.808479990363282e-05, "loss": 0.0381, "step": 4020 }, { "grad_norm": 0.5269554853439331, "learning_rate": 9.806966248525445e-05, "loss": 0.0344, "step": 4030 }, { "grad_norm": 0.48379433155059814, "learning_rate": 9.805446665770348e-05, "loss": 0.029, "step": 4040 }, { "grad_norm": 0.4997735321521759, "learning_rate": 9.803921243944429e-05, "loss": 0.0324, "step": 4050 }, { "grad_norm": 0.5773942470550537, "learning_rate": 9.802389984901218e-05, "loss": 0.0335, "step": 4060 }, { "grad_norm": 0.5910668969154358, "learning_rate": 9.80085289050134e-05, "loss": 0.0344, "step": 4070 }, { "grad_norm": 0.6216025352478027, "learning_rate": 9.799309962612508e-05, "loss": 0.0337, "step": 4080 }, { "grad_norm": 0.5212052464485168, "learning_rate": 9.797761203109527e-05, "loss": 0.0324, "step": 4090 }, { "grad_norm": 0.475735068321228, "learning_rate": 9.796206613874283e-05, "loss": 0.0327, "step": 4100 }, { "grad_norm": 0.6021344065666199, "learning_rate": 9.794646196795754e-05, "loss": 0.0311, "step": 4110 }, { "grad_norm": 0.48761188983917236, "learning_rate": 9.793079953769987e-05, "loss": 0.0385, "step": 4120 }, { "grad_norm": 0.6330643892288208, "learning_rate": 9.79150788670012e-05, "loss": 0.0329, "step": 4130 }, { "grad_norm": 0.3942389190196991, "learning_rate": 9.78992999749636e-05, "loss": 0.0307, "step": 4140 }, { "grad_norm": 0.4972276985645294, "learning_rate": 9.788346288075994e-05, "loss": 0.0301, "step": 4150 }, { "grad_norm": 0.49850016832351685, "learning_rate": 9.786756760363373e-05, "loss": 0.0324, "step": 4160 }, { "grad_norm": 0.4252046048641205, "learning_rate": 9.78516141628993e-05, "loss": 0.0308, "step": 4170 }, { "grad_norm": 0.571590006351471, "learning_rate": 9.783560257794154e-05, "loss": 0.0294, "step": 4180 }, { "grad_norm": 0.5250931978225708, "learning_rate": 9.781953286821603e-05, "loss": 0.0297, "step": 4190 }, { "grad_norm": 0.5842002630233765, "learning_rate": 9.780340505324901e-05, "loss": 0.0316, "step": 4200 }, { "grad_norm": 0.6337473392486572, "learning_rate": 9.778721915263727e-05, "loss": 0.0338, "step": 4210 }, { "grad_norm": 0.48300299048423767, "learning_rate": 9.777097518604824e-05, "loss": 0.0303, "step": 4220 }, { "grad_norm": 0.5463433861732483, "learning_rate": 9.775467317321984e-05, "loss": 0.029, "step": 4230 }, { "grad_norm": 0.5539782643318176, "learning_rate": 9.773831313396055e-05, "loss": 0.0307, "step": 4240 }, { "grad_norm": 0.6409865021705627, "learning_rate": 9.77218950881494e-05, "loss": 0.0309, "step": 4250 }, { "grad_norm": 0.5464199185371399, "learning_rate": 9.770541905573583e-05, "loss": 0.0284, "step": 4260 }, { "grad_norm": 0.5354518890380859, "learning_rate": 9.768888505673976e-05, "loss": 0.0311, "step": 4270 }, { "grad_norm": 0.4698927402496338, "learning_rate": 9.767229311125162e-05, "loss": 0.0277, "step": 4280 }, { "grad_norm": 0.5004948377609253, "learning_rate": 9.765564323943211e-05, "loss": 0.028, "step": 4290 }, { "grad_norm": 0.6428542137145996, "learning_rate": 9.763893546151244e-05, "loss": 0.0326, "step": 4300 }, { "grad_norm": 0.567100465297699, "learning_rate": 9.762216979779412e-05, "loss": 0.0337, "step": 4310 }, { "grad_norm": 0.3555885851383209, "learning_rate": 9.760534626864902e-05, "loss": 0.0283, "step": 4320 }, { "grad_norm": 0.6055477857589722, "learning_rate": 9.758846489451931e-05, "loss": 0.0355, "step": 4330 }, { "grad_norm": 0.4653003513813019, "learning_rate": 9.757152569591748e-05, "loss": 0.0346, "step": 4340 }, { "grad_norm": 0.5382339358329773, "learning_rate": 9.75545286934262e-05, "loss": 0.0287, "step": 4350 }, { "grad_norm": 0.46080365777015686, "learning_rate": 9.753747390769847e-05, "loss": 0.0301, "step": 4360 }, { "grad_norm": 0.43235480785369873, "learning_rate": 9.752036135945744e-05, "loss": 0.0343, "step": 4370 }, { "grad_norm": 0.5093517303466797, "learning_rate": 9.750319106949649e-05, "loss": 0.0321, "step": 4380 }, { "grad_norm": 0.5459606647491455, "learning_rate": 9.748596305867913e-05, "loss": 0.0296, "step": 4390 }, { "grad_norm": 0.5006504058837891, "learning_rate": 9.746867734793903e-05, "loss": 0.0319, "step": 4400 }, { "grad_norm": 0.46831148862838745, "learning_rate": 9.745133395827993e-05, "loss": 0.0314, "step": 4410 }, { "grad_norm": 0.38328561186790466, "learning_rate": 9.743393291077572e-05, "loss": 0.033, "step": 4420 }, { "grad_norm": 0.48275116086006165, "learning_rate": 9.741647422657028e-05, "loss": 0.0249, "step": 4430 }, { "grad_norm": 0.4385939836502075, "learning_rate": 9.739895792687758e-05, "loss": 0.0299, "step": 4440 }, { "grad_norm": 0.5937483906745911, "learning_rate": 9.738138403298157e-05, "loss": 0.0337, "step": 4450 }, { "grad_norm": 0.4840114712715149, "learning_rate": 9.736375256623619e-05, "loss": 0.0335, "step": 4460 }, { "grad_norm": 0.5124753713607788, "learning_rate": 9.734606354806533e-05, "loss": 0.0349, "step": 4470 }, { "grad_norm": 0.43529045581817627, "learning_rate": 9.73283169999628e-05, "loss": 0.0272, "step": 4480 }, { "grad_norm": 0.5181006193161011, "learning_rate": 9.731051294349238e-05, "loss": 0.0264, "step": 4490 }, { "grad_norm": 0.4760429263114929, "learning_rate": 9.729265140028762e-05, "loss": 0.0268, "step": 4500 }, { "grad_norm": 0.5206795930862427, "learning_rate": 9.727473239205201e-05, "loss": 0.0267, "step": 4510 }, { "grad_norm": 0.5616780519485474, "learning_rate": 9.725675594055883e-05, "loss": 0.0315, "step": 4520 }, { "grad_norm": 0.6338405013084412, "learning_rate": 9.723872206765116e-05, "loss": 0.0316, "step": 4530 }, { "grad_norm": 0.4781239628791809, "learning_rate": 9.722063079524185e-05, "loss": 0.0348, "step": 4540 }, { "grad_norm": 0.4985649585723877, "learning_rate": 9.720248214531351e-05, "loss": 0.0301, "step": 4550 }, { "grad_norm": 0.4828687012195587, "learning_rate": 9.718427613991848e-05, "loss": 0.0269, "step": 4560 }, { "grad_norm": 0.43408864736557007, "learning_rate": 9.716601280117873e-05, "loss": 0.027, "step": 4570 }, { "grad_norm": 0.48410576581954956, "learning_rate": 9.714769215128596e-05, "loss": 0.0289, "step": 4580 }, { "grad_norm": 0.4228646457195282, "learning_rate": 9.712931421250152e-05, "loss": 0.0285, "step": 4590 }, { "grad_norm": 0.5390908718109131, "learning_rate": 9.711087900715627e-05, "loss": 0.0324, "step": 4600 }, { "grad_norm": 0.5309498310089111, "learning_rate": 9.709238655765078e-05, "loss": 0.0294, "step": 4610 }, { "grad_norm": 0.4404495060443878, "learning_rate": 9.707383688645511e-05, "loss": 0.027, "step": 4620 }, { "grad_norm": 0.45291653275489807, "learning_rate": 9.705523001610883e-05, "loss": 0.0283, "step": 4630 }, { "grad_norm": 0.4844144284725189, "learning_rate": 9.703656596922107e-05, "loss": 0.0279, "step": 4640 }, { "grad_norm": 0.5359585881233215, "learning_rate": 9.70178447684704e-05, "loss": 0.0267, "step": 4650 }, { "grad_norm": 0.513267457485199, "learning_rate": 9.699906643660483e-05, "loss": 0.0279, "step": 4660 }, { "grad_norm": 0.5462450981140137, "learning_rate": 9.698023099644185e-05, "loss": 0.0286, "step": 4670 }, { "grad_norm": 0.6027638912200928, "learning_rate": 9.696133847086823e-05, "loss": 0.0308, "step": 4680 }, { "grad_norm": 0.5719568729400635, "learning_rate": 9.694238888284022e-05, "loss": 0.0399, "step": 4690 }, { "grad_norm": 0.4713907539844513, "learning_rate": 9.692338225538333e-05, "loss": 0.0366, "step": 4700 }, { "grad_norm": 0.4537694752216339, "learning_rate": 9.690431861159241e-05, "loss": 0.0345, "step": 4710 }, { "grad_norm": 0.48868870735168457, "learning_rate": 9.688519797463161e-05, "loss": 0.0351, "step": 4720 }, { "grad_norm": 0.4144360423088074, "learning_rate": 9.686602036773426e-05, "loss": 0.0331, "step": 4730 }, { "grad_norm": 0.5407605767250061, "learning_rate": 9.684678581420302e-05, "loss": 0.0349, "step": 4740 }, { "grad_norm": 0.4362596571445465, "learning_rate": 9.682749433740962e-05, "loss": 0.0306, "step": 4750 }, { "grad_norm": 0.47596433758735657, "learning_rate": 9.680814596079507e-05, "loss": 0.0292, "step": 4760 }, { "grad_norm": 0.48152321577072144, "learning_rate": 9.678874070786945e-05, "loss": 0.0282, "step": 4770 }, { "grad_norm": 0.43656474351882935, "learning_rate": 9.676927860221199e-05, "loss": 0.0283, "step": 4780 }, { "grad_norm": 0.40873607993125916, "learning_rate": 9.674975966747097e-05, "loss": 0.0302, "step": 4790 }, { "grad_norm": 0.5664829611778259, "learning_rate": 9.673018392736374e-05, "loss": 0.0315, "step": 4800 }, { "grad_norm": 0.4111019968986511, "learning_rate": 9.671055140567667e-05, "loss": 0.035, "step": 4810 }, { "grad_norm": 0.42563799023628235, "learning_rate": 9.669086212626511e-05, "loss": 0.0306, "step": 4820 }, { "grad_norm": 0.5026954412460327, "learning_rate": 9.667111611305341e-05, "loss": 0.0304, "step": 4830 }, { "grad_norm": 0.42993810772895813, "learning_rate": 9.665131339003486e-05, "loss": 0.0289, "step": 4840 }, { "grad_norm": 0.4692381024360657, "learning_rate": 9.663145398127158e-05, "loss": 0.0296, "step": 4850 }, { "grad_norm": 0.5437437891960144, "learning_rate": 9.661153791089467e-05, "loss": 0.0257, "step": 4860 }, { "grad_norm": 0.6344571709632874, "learning_rate": 9.659156520310402e-05, "loss": 0.0295, "step": 4870 }, { "grad_norm": 0.41922223567962646, "learning_rate": 9.657153588216835e-05, "loss": 0.0281, "step": 4880 }, { "grad_norm": 0.5096099972724915, "learning_rate": 9.655144997242516e-05, "loss": 0.0297, "step": 4890 }, { "grad_norm": 0.49703431129455566, "learning_rate": 9.653130749828075e-05, "loss": 0.0269, "step": 4900 }, { "grad_norm": 0.5714578032493591, "learning_rate": 9.65111084842101e-05, "loss": 0.031, "step": 4910 }, { "grad_norm": 0.5207062363624573, "learning_rate": 9.649085295475695e-05, "loss": 0.0316, "step": 4920 }, { "grad_norm": 0.4940756559371948, "learning_rate": 9.647054093453365e-05, "loss": 0.0296, "step": 4930 }, { "grad_norm": 0.4935055077075958, "learning_rate": 9.645017244822123e-05, "loss": 0.0312, "step": 4940 }, { "grad_norm": 0.4722101092338562, "learning_rate": 9.642974752056931e-05, "loss": 0.03, "step": 4950 }, { "grad_norm": 0.5050386786460876, "learning_rate": 9.640926617639613e-05, "loss": 0.0309, "step": 4960 }, { "grad_norm": 0.42670542001724243, "learning_rate": 9.638872844058843e-05, "loss": 0.0382, "step": 4970 }, { "grad_norm": 0.4971314072608948, "learning_rate": 9.63681343381015e-05, "loss": 0.0291, "step": 4980 }, { "grad_norm": 0.44043266773223877, "learning_rate": 9.634748389395914e-05, "loss": 0.0253, "step": 4990 }, { "grad_norm": 0.4813491106033325, "learning_rate": 9.632677713325353e-05, "loss": 0.0324, "step": 5000 }, { "grad_norm": 0.43432343006134033, "learning_rate": 9.63060140811454e-05, "loss": 0.0368, "step": 5010 }, { "grad_norm": 0.4000500440597534, "learning_rate": 9.628519476286379e-05, "loss": 0.0279, "step": 5020 }, { "grad_norm": 0.48640185594558716, "learning_rate": 9.626431920370612e-05, "loss": 0.0271, "step": 5030 }, { "grad_norm": 0.4676959812641144, "learning_rate": 9.624338742903819e-05, "loss": 0.0303, "step": 5040 }, { "grad_norm": 0.5149692893028259, "learning_rate": 9.622239946429406e-05, "loss": 0.0258, "step": 5050 }, { "grad_norm": 0.3987928628921509, "learning_rate": 9.620135533497609e-05, "loss": 0.0281, "step": 5060 }, { "grad_norm": 0.4288138151168823, "learning_rate": 9.61802550666549e-05, "loss": 0.0298, "step": 5070 }, { "grad_norm": 0.5656899809837341, "learning_rate": 9.615909868496928e-05, "loss": 0.0292, "step": 5080 }, { "grad_norm": 0.5501643419265747, "learning_rate": 9.613788621562622e-05, "loss": 0.0308, "step": 5090 }, { "grad_norm": 0.5483107566833496, "learning_rate": 9.611661768440093e-05, "loss": 0.0293, "step": 5100 }, { "grad_norm": 0.404590904712677, "learning_rate": 9.609529311713661e-05, "loss": 0.0278, "step": 5110 }, { "grad_norm": 0.5149969458580017, "learning_rate": 9.607391253974466e-05, "loss": 0.0302, "step": 5120 }, { "grad_norm": 0.5531370639801025, "learning_rate": 9.605247597820448e-05, "loss": 0.0262, "step": 5130 }, { "grad_norm": 0.42889198660850525, "learning_rate": 9.603098345856354e-05, "loss": 0.0248, "step": 5140 }, { "grad_norm": 0.5078679323196411, "learning_rate": 9.600943500693725e-05, "loss": 0.0287, "step": 5150 }, { "grad_norm": 0.45680946111679077, "learning_rate": 9.598783064950902e-05, "loss": 0.0393, "step": 5160 }, { "grad_norm": 0.41632264852523804, "learning_rate": 9.596617041253018e-05, "loss": 0.0293, "step": 5170 }, { "grad_norm": 0.3672720491886139, "learning_rate": 9.594445432231996e-05, "loss": 0.0242, "step": 5180 }, { "grad_norm": 0.40611353516578674, "learning_rate": 9.592268240526547e-05, "loss": 0.0298, "step": 5190 }, { "grad_norm": 0.5475040078163147, "learning_rate": 9.590085468782162e-05, "loss": 0.0289, "step": 5200 }, { "grad_norm": 0.381839781999588, "learning_rate": 9.587897119651116e-05, "loss": 0.0275, "step": 5210 }, { "grad_norm": 0.4522201716899872, "learning_rate": 9.585703195792459e-05, "loss": 0.0293, "step": 5220 }, { "grad_norm": 0.3578600287437439, "learning_rate": 9.583503699872016e-05, "loss": 0.0255, "step": 5230 }, { "grad_norm": 0.4494266211986542, "learning_rate": 9.581298634562381e-05, "loss": 0.0279, "step": 5240 }, { "grad_norm": 0.4995555877685547, "learning_rate": 9.579088002542917e-05, "loss": 0.0259, "step": 5250 }, { "grad_norm": 0.5074251890182495, "learning_rate": 9.57687180649975e-05, "loss": 0.0263, "step": 5260 }, { "grad_norm": 0.42040592432022095, "learning_rate": 9.574650049125768e-05, "loss": 0.0265, "step": 5270 }, { "grad_norm": 0.5675180554389954, "learning_rate": 9.572422733120614e-05, "loss": 0.0288, "step": 5280 }, { "grad_norm": 0.41257384419441223, "learning_rate": 9.570189861190689e-05, "loss": 0.0248, "step": 5290 }, { "grad_norm": 0.4741426408290863, "learning_rate": 9.56795143604914e-05, "loss": 0.0332, "step": 5300 }, { "grad_norm": 0.47105324268341064, "learning_rate": 9.565707460415869e-05, "loss": 0.033, "step": 5310 }, { "grad_norm": 0.4463244080543518, "learning_rate": 9.563457937017515e-05, "loss": 0.0268, "step": 5320 }, { "grad_norm": 0.42428091168403625, "learning_rate": 9.56120286858746e-05, "loss": 0.0266, "step": 5330 }, { "grad_norm": 0.4468759298324585, "learning_rate": 9.558942257865829e-05, "loss": 0.0249, "step": 5340 }, { "grad_norm": 0.4177563786506653, "learning_rate": 9.556676107599472e-05, "loss": 0.0279, "step": 5350 }, { "grad_norm": 0.42361655831336975, "learning_rate": 9.554404420541978e-05, "loss": 0.0291, "step": 5360 }, { "grad_norm": 0.4569939374923706, "learning_rate": 9.55212719945366e-05, "loss": 0.0256, "step": 5370 }, { "grad_norm": 0.49631205201148987, "learning_rate": 9.549844447101559e-05, "loss": 0.0279, "step": 5380 }, { "grad_norm": 0.4115065932273865, "learning_rate": 9.547556166259433e-05, "loss": 0.0272, "step": 5390 }, { "grad_norm": 0.45480725169181824, "learning_rate": 9.545262359707756e-05, "loss": 0.0258, "step": 5400 }, { "grad_norm": 0.34017133712768555, "learning_rate": 9.542963030233724e-05, "loss": 0.0254, "step": 5410 }, { "grad_norm": 0.3817330300807953, "learning_rate": 9.540658180631237e-05, "loss": 0.0253, "step": 5420 }, { "grad_norm": 0.5424914360046387, "learning_rate": 9.538347813700904e-05, "loss": 0.0278, "step": 5430 }, { "grad_norm": 0.45396095514297485, "learning_rate": 9.536031932250036e-05, "loss": 0.0279, "step": 5440 }, { "grad_norm": 0.39200082421302795, "learning_rate": 9.533710539092653e-05, "loss": 0.0256, "step": 5450 }, { "grad_norm": 0.5826816558837891, "learning_rate": 9.531383637049464e-05, "loss": 0.0346, "step": 5460 }, { "grad_norm": 0.4850524365901947, "learning_rate": 9.529051228947875e-05, "loss": 0.0252, "step": 5470 }, { "grad_norm": 0.4608486294746399, "learning_rate": 9.52671331762198e-05, "loss": 0.0265, "step": 5480 }, { "grad_norm": 0.4006120264530182, "learning_rate": 9.524369905912565e-05, "loss": 0.0239, "step": 5490 }, { "grad_norm": 0.6033313870429993, "learning_rate": 9.522020996667092e-05, "loss": 0.0276, "step": 5500 }, { "grad_norm": 0.44466960430145264, "learning_rate": 9.519666592739709e-05, "loss": 0.0239, "step": 5510 }, { "grad_norm": 0.5001692175865173, "learning_rate": 9.517306696991241e-05, "loss": 0.024, "step": 5520 }, { "grad_norm": 0.44814756512641907, "learning_rate": 9.51494131228918e-05, "loss": 0.0282, "step": 5530 }, { "grad_norm": 0.5809621810913086, "learning_rate": 9.512570441507695e-05, "loss": 0.0311, "step": 5540 }, { "grad_norm": 0.38321542739868164, "learning_rate": 9.510194087527615e-05, "loss": 0.0292, "step": 5550 }, { "grad_norm": 0.42842525243759155, "learning_rate": 9.507812253236435e-05, "loss": 0.0287, "step": 5560 }, { "grad_norm": 0.4932766854763031, "learning_rate": 9.505424941528309e-05, "loss": 0.0295, "step": 5570 }, { "grad_norm": 0.4855293333530426, "learning_rate": 9.503032155304046e-05, "loss": 0.0261, "step": 5580 }, { "grad_norm": 0.2983034551143646, "learning_rate": 9.500633897471106e-05, "loss": 0.0254, "step": 5590 }, { "grad_norm": 0.4807620644569397, "learning_rate": 9.498230170943596e-05, "loss": 0.0237, "step": 5600 }, { "grad_norm": 0.46898818016052246, "learning_rate": 9.495820978642275e-05, "loss": 0.029, "step": 5610 }, { "grad_norm": 0.3995807468891144, "learning_rate": 9.493406323494535e-05, "loss": 0.0296, "step": 5620 }, { "grad_norm": 0.4525352418422699, "learning_rate": 9.490986208434413e-05, "loss": 0.0276, "step": 5630 }, { "grad_norm": 0.5368863940238953, "learning_rate": 9.488560636402577e-05, "loss": 0.0285, "step": 5640 }, { "grad_norm": 0.5784607529640198, "learning_rate": 9.486129610346321e-05, "loss": 0.0263, "step": 5650 }, { "grad_norm": 0.4409428536891937, "learning_rate": 9.483693133219576e-05, "loss": 0.0275, "step": 5660 }, { "grad_norm": 0.42036542296409607, "learning_rate": 9.481251207982888e-05, "loss": 0.0272, "step": 5670 }, { "grad_norm": 0.398827463388443, "learning_rate": 9.47880383760343e-05, "loss": 0.026, "step": 5680 }, { "grad_norm": 0.4404725134372711, "learning_rate": 9.476351025054983e-05, "loss": 0.0265, "step": 5690 }, { "grad_norm": 0.41246429085731506, "learning_rate": 9.473892773317952e-05, "loss": 0.0262, "step": 5700 }, { "grad_norm": 0.6557384133338928, "learning_rate": 9.471429085379338e-05, "loss": 0.0294, "step": 5710 }, { "grad_norm": 0.4278976023197174, "learning_rate": 9.468959964232757e-05, "loss": 0.0286, "step": 5720 }, { "grad_norm": 0.5034335255622864, "learning_rate": 9.466485412878425e-05, "loss": 0.0266, "step": 5730 }, { "grad_norm": 0.39809751510620117, "learning_rate": 9.464005434323154e-05, "loss": 0.026, "step": 5740 }, { "grad_norm": 0.4159664511680603, "learning_rate": 9.461520031580352e-05, "loss": 0.0227, "step": 5750 }, { "grad_norm": 0.44587886333465576, "learning_rate": 9.459029207670019e-05, "loss": 0.0251, "step": 5760 }, { "grad_norm": 0.4395875930786133, "learning_rate": 9.456532965618737e-05, "loss": 0.0268, "step": 5770 }, { "grad_norm": 0.4429076015949249, "learning_rate": 9.454031308459681e-05, "loss": 0.0227, "step": 5780 }, { "grad_norm": 0.4217134416103363, "learning_rate": 9.451524239232595e-05, "loss": 0.0234, "step": 5790 }, { "grad_norm": 0.4322492778301239, "learning_rate": 9.449011760983809e-05, "loss": 0.0242, "step": 5800 }, { "grad_norm": 0.4535260498523712, "learning_rate": 9.446493876766218e-05, "loss": 0.0312, "step": 5810 }, { "grad_norm": 0.44206225872039795, "learning_rate": 9.44397058963929e-05, "loss": 0.0231, "step": 5820 }, { "grad_norm": 0.5544426441192627, "learning_rate": 9.441441902669056e-05, "loss": 0.0271, "step": 5830 }, { "grad_norm": 0.3378109931945801, "learning_rate": 9.43890781892811e-05, "loss": 0.0213, "step": 5840 }, { "grad_norm": 0.39126208424568176, "learning_rate": 9.436368341495603e-05, "loss": 0.0238, "step": 5850 }, { "grad_norm": 0.4364147484302521, "learning_rate": 9.43382347345724e-05, "loss": 0.0222, "step": 5860 }, { "grad_norm": 0.5010242462158203, "learning_rate": 9.431273217905273e-05, "loss": 0.0237, "step": 5870 }, { "grad_norm": 0.37392815947532654, "learning_rate": 9.428717577938504e-05, "loss": 0.0241, "step": 5880 }, { "grad_norm": 0.4261113107204437, "learning_rate": 9.426156556662276e-05, "loss": 0.0241, "step": 5890 }, { "grad_norm": 0.47384363412857056, "learning_rate": 9.423590157188474e-05, "loss": 0.0296, "step": 5900 }, { "grad_norm": 0.5442717671394348, "learning_rate": 9.421018382635513e-05, "loss": 0.0231, "step": 5910 }, { "grad_norm": 0.3940904140472412, "learning_rate": 9.418441236128343e-05, "loss": 0.0211, "step": 5920 }, { "grad_norm": 0.4920433461666107, "learning_rate": 9.41585872079844e-05, "loss": 0.0226, "step": 5930 }, { "grad_norm": 0.43605488538742065, "learning_rate": 9.413270839783802e-05, "loss": 0.0211, "step": 5940 }, { "grad_norm": 0.3884357810020447, "learning_rate": 9.41067759622895e-05, "loss": 0.0233, "step": 5950 }, { "grad_norm": 0.36957186460494995, "learning_rate": 9.408078993284917e-05, "loss": 0.0261, "step": 5960 }, { "grad_norm": 0.4391566216945648, "learning_rate": 9.405475034109254e-05, "loss": 0.0247, "step": 5970 }, { "grad_norm": 0.49858593940734863, "learning_rate": 9.402865721866015e-05, "loss": 0.0294, "step": 5980 }, { "grad_norm": 0.5844445824623108, "learning_rate": 9.400251059725762e-05, "loss": 0.0238, "step": 5990 }, { "grad_norm": 0.41200026869773865, "learning_rate": 9.397631050865554e-05, "loss": 0.0237, "step": 6000 }, { "grad_norm": 0.3171989321708679, "learning_rate": 9.395005698468949e-05, "loss": 0.0244, "step": 6010 }, { "grad_norm": 0.34872493147850037, "learning_rate": 9.392375005726e-05, "loss": 0.0277, "step": 6020 }, { "grad_norm": 0.411639928817749, "learning_rate": 9.389738975833242e-05, "loss": 0.0211, "step": 6030 }, { "grad_norm": 0.48083916306495667, "learning_rate": 9.387097611993707e-05, "loss": 0.0267, "step": 6040 }, { "grad_norm": 0.523038923740387, "learning_rate": 9.384450917416894e-05, "loss": 0.0229, "step": 6050 }, { "grad_norm": 0.46667537093162537, "learning_rate": 9.381798895318792e-05, "loss": 0.0243, "step": 6060 }, { "grad_norm": 0.47458550333976746, "learning_rate": 9.379141548921853e-05, "loss": 0.0278, "step": 6070 }, { "grad_norm": 0.37740984559059143, "learning_rate": 9.376478881455009e-05, "loss": 0.0211, "step": 6080 }, { "grad_norm": 0.4149538576602936, "learning_rate": 9.373810896153647e-05, "loss": 0.0255, "step": 6090 }, { "grad_norm": 0.44610509276390076, "learning_rate": 9.371137596259623e-05, "loss": 0.0279, "step": 6100 }, { "grad_norm": 0.5716355443000793, "learning_rate": 9.368458985021248e-05, "loss": 0.0291, "step": 6110 }, { "grad_norm": 0.43165573477745056, "learning_rate": 9.365775065693287e-05, "loss": 0.0265, "step": 6120 }, { "grad_norm": 0.44173577427864075, "learning_rate": 9.363085841536957e-05, "loss": 0.0224, "step": 6130 }, { "grad_norm": 0.45178642868995667, "learning_rate": 9.360391315819917e-05, "loss": 0.0218, "step": 6140 }, { "grad_norm": 0.3972025215625763, "learning_rate": 9.357691491816269e-05, "loss": 0.0246, "step": 6150 }, { "grad_norm": 0.6218813061714172, "learning_rate": 9.354986372806557e-05, "loss": 0.0238, "step": 6160 }, { "grad_norm": 0.4248044192790985, "learning_rate": 9.352275962077752e-05, "loss": 0.0242, "step": 6170 }, { "grad_norm": 0.4957454800605774, "learning_rate": 9.349560262923262e-05, "loss": 0.0266, "step": 6180 }, { "grad_norm": 0.43156319856643677, "learning_rate": 9.346839278642913e-05, "loss": 0.0228, "step": 6190 }, { "grad_norm": 0.4218955636024475, "learning_rate": 9.344113012542963e-05, "loss": 0.0221, "step": 6200 }, { "grad_norm": 0.3826098144054413, "learning_rate": 9.341381467936079e-05, "loss": 0.0243, "step": 6210 }, { "grad_norm": 0.656653642654419, "learning_rate": 9.338644648141346e-05, "loss": 0.0247, "step": 6220 }, { "grad_norm": 0.5594580173492432, "learning_rate": 9.335902556484257e-05, "loss": 0.0265, "step": 6230 }, { "grad_norm": 0.473480224609375, "learning_rate": 9.333155196296712e-05, "loss": 0.0312, "step": 6240 }, { "grad_norm": 0.36913251876831055, "learning_rate": 9.330402570917016e-05, "loss": 0.0233, "step": 6250 }, { "grad_norm": 0.4519546926021576, "learning_rate": 9.327644683689865e-05, "loss": 0.0222, "step": 6260 }, { "grad_norm": 0.3610144555568695, "learning_rate": 9.324881537966354e-05, "loss": 0.0254, "step": 6270 }, { "grad_norm": 0.5942192673683167, "learning_rate": 9.322113137103964e-05, "loss": 0.0288, "step": 6280 }, { "grad_norm": 0.41067391633987427, "learning_rate": 9.319339484466564e-05, "loss": 0.025, "step": 6290 }, { "grad_norm": 0.507523775100708, "learning_rate": 9.316560583424403e-05, "loss": 0.0226, "step": 6300 }, { "grad_norm": 0.40283846855163574, "learning_rate": 9.313776437354108e-05, "loss": 0.0233, "step": 6310 }, { "grad_norm": 0.42097848653793335, "learning_rate": 9.31098704963868e-05, "loss": 0.0238, "step": 6320 }, { "grad_norm": 0.49295732378959656, "learning_rate": 9.308192423667485e-05, "loss": 0.022, "step": 6330 }, { "grad_norm": 0.366346150636673, "learning_rate": 9.305392562836262e-05, "loss": 0.02, "step": 6340 }, { "grad_norm": 0.45008862018585205, "learning_rate": 9.3025874705471e-05, "loss": 0.0202, "step": 6350 }, { "grad_norm": 0.45476216077804565, "learning_rate": 9.299777150208457e-05, "loss": 0.0189, "step": 6360 }, { "grad_norm": 0.39774391055107117, "learning_rate": 9.296961605235133e-05, "loss": 0.019, "step": 6370 }, { "grad_norm": 0.45898446440696716, "learning_rate": 9.29414083904828e-05, "loss": 0.0231, "step": 6380 }, { "grad_norm": 0.41240614652633667, "learning_rate": 9.291314855075397e-05, "loss": 0.022, "step": 6390 }, { "grad_norm": 0.33521097898483276, "learning_rate": 9.288483656750322e-05, "loss": 0.0213, "step": 6400 }, { "grad_norm": 0.37693679332733154, "learning_rate": 9.285647247513225e-05, "loss": 0.0177, "step": 6410 }, { "grad_norm": 0.4334852695465088, "learning_rate": 9.282805630810614e-05, "loss": 0.0224, "step": 6420 }, { "grad_norm": 0.3814341127872467, "learning_rate": 9.279958810095317e-05, "loss": 0.0185, "step": 6430 }, { "grad_norm": 0.4334353506565094, "learning_rate": 9.277106788826494e-05, "loss": 0.0225, "step": 6440 }, { "grad_norm": 0.35524722933769226, "learning_rate": 9.274249570469618e-05, "loss": 0.0178, "step": 6450 }, { "grad_norm": 0.40956011414527893, "learning_rate": 9.271387158496476e-05, "loss": 0.0195, "step": 6460 }, { "grad_norm": 0.4977434575557709, "learning_rate": 9.268519556385173e-05, "loss": 0.0203, "step": 6470 }, { "grad_norm": 0.5096681118011475, "learning_rate": 9.265646767620113e-05, "loss": 0.0237, "step": 6480 }, { "grad_norm": 0.5548952221870422, "learning_rate": 9.262768795692006e-05, "loss": 0.0209, "step": 6490 }, { "grad_norm": 0.38688987493515015, "learning_rate": 9.259885644097861e-05, "loss": 0.0223, "step": 6500 }, { "grad_norm": 0.3244021534919739, "learning_rate": 9.256997316340976e-05, "loss": 0.0165, "step": 6510 }, { "grad_norm": 0.3968213200569153, "learning_rate": 9.254103815930943e-05, "loss": 0.0191, "step": 6520 }, { "grad_norm": 0.4290993809700012, "learning_rate": 9.251205146383636e-05, "loss": 0.0192, "step": 6530 }, { "grad_norm": 0.41794154047966003, "learning_rate": 9.248301311221216e-05, "loss": 0.0196, "step": 6540 }, { "grad_norm": 0.36172324419021606, "learning_rate": 9.245392313972115e-05, "loss": 0.0179, "step": 6550 }, { "grad_norm": 0.465830534696579, "learning_rate": 9.242478158171037e-05, "loss": 0.0171, "step": 6560 }, { "grad_norm": 0.4272666275501251, "learning_rate": 9.239558847358957e-05, "loss": 0.0244, "step": 6570 }, { "grad_norm": 0.30127090215682983, "learning_rate": 9.236634385083114e-05, "loss": 0.0209, "step": 6580 }, { "grad_norm": 0.544417142868042, "learning_rate": 9.233704774897005e-05, "loss": 0.0205, "step": 6590 }, { "grad_norm": 0.3972528874874115, "learning_rate": 9.230770020360382e-05, "loss": 0.0217, "step": 6600 }, { "grad_norm": 0.5284895896911621, "learning_rate": 9.227830125039247e-05, "loss": 0.0221, "step": 6610 }, { "grad_norm": 0.42657437920570374, "learning_rate": 9.224885092505853e-05, "loss": 0.0205, "step": 6620 }, { "grad_norm": 0.41983553767204285, "learning_rate": 9.221934926338691e-05, "loss": 0.017, "step": 6630 }, { "grad_norm": 0.6930293440818787, "learning_rate": 9.21897963012249e-05, "loss": 0.0218, "step": 6640 }, { "grad_norm": 0.49144089221954346, "learning_rate": 9.216019207448217e-05, "loss": 0.017, "step": 6650 }, { "grad_norm": 0.5103893876075745, "learning_rate": 9.21305366191306e-05, "loss": 0.0198, "step": 6660 }, { "grad_norm": 0.48660317063331604, "learning_rate": 9.21008299712044e-05, "loss": 0.0212, "step": 6670 }, { "grad_norm": 0.4938410520553589, "learning_rate": 9.207107216679993e-05, "loss": 0.0208, "step": 6680 }, { "grad_norm": 0.48082706332206726, "learning_rate": 9.204126324207574e-05, "loss": 0.0178, "step": 6690 }, { "grad_norm": 0.4938015043735504, "learning_rate": 9.201140323325248e-05, "loss": 0.0238, "step": 6700 }, { "grad_norm": 0.4831843376159668, "learning_rate": 9.198149217661287e-05, "loss": 0.0182, "step": 6710 }, { "grad_norm": 0.411458283662796, "learning_rate": 9.195153010850167e-05, "loss": 0.0179, "step": 6720 }, { "grad_norm": 0.5406591296195984, "learning_rate": 9.192151706532562e-05, "loss": 0.0188, "step": 6730 }, { "grad_norm": 0.4220775067806244, "learning_rate": 9.189145308355337e-05, "loss": 0.0167, "step": 6740 }, { "grad_norm": 0.47835198044776917, "learning_rate": 9.186133819971555e-05, "loss": 0.017, "step": 6750 }, { "grad_norm": 0.37527114152908325, "learning_rate": 9.183117245040455e-05, "loss": 0.0185, "step": 6760 }, { "grad_norm": 0.4802888035774231, "learning_rate": 9.180095587227458e-05, "loss": 0.0164, "step": 6770 }, { "grad_norm": 0.4387568533420563, "learning_rate": 9.177068850204167e-05, "loss": 0.0154, "step": 6780 }, { "grad_norm": 0.3502100706100464, "learning_rate": 9.174037037648351e-05, "loss": 0.0199, "step": 6790 }, { "grad_norm": 0.42201828956604004, "learning_rate": 9.171000153243947e-05, "loss": 0.0186, "step": 6800 }, { "grad_norm": 0.4460437297821045, "learning_rate": 9.167958200681057e-05, "loss": 0.0181, "step": 6810 }, { "grad_norm": 0.4730928838253021, "learning_rate": 9.164911183655943e-05, "loss": 0.0177, "step": 6820 }, { "grad_norm": 0.49859949946403503, "learning_rate": 9.161859105871013e-05, "loss": 0.0172, "step": 6830 }, { "grad_norm": 0.3294234871864319, "learning_rate": 9.158801971034832e-05, "loss": 0.0182, "step": 6840 }, { "grad_norm": 0.4640909433364868, "learning_rate": 9.155739782862107e-05, "loss": 0.0178, "step": 6850 }, { "grad_norm": 0.5015811920166016, "learning_rate": 9.152672545073687e-05, "loss": 0.0164, "step": 6860 }, { "grad_norm": 0.474317729473114, "learning_rate": 9.149600261396551e-05, "loss": 0.0163, "step": 6870 }, { "grad_norm": 0.41502758860588074, "learning_rate": 9.146522935563817e-05, "loss": 0.0136, "step": 6880 }, { "grad_norm": 0.4295918643474579, "learning_rate": 9.143440571314723e-05, "loss": 0.0155, "step": 6890 }, { "grad_norm": 0.428204745054245, "learning_rate": 9.140353172394636e-05, "loss": 0.0185, "step": 6900 }, { "grad_norm": 0.4600503444671631, "learning_rate": 9.137260742555032e-05, "loss": 0.0181, "step": 6910 }, { "grad_norm": 0.45985567569732666, "learning_rate": 9.13416328555351e-05, "loss": 0.0166, "step": 6920 }, { "grad_norm": 0.3849891424179077, "learning_rate": 9.131060805153769e-05, "loss": 0.0171, "step": 6930 }, { "grad_norm": 0.5211637020111084, "learning_rate": 9.127953305125618e-05, "loss": 0.0175, "step": 6940 }, { "grad_norm": 0.4012833535671234, "learning_rate": 9.124840789244958e-05, "loss": 0.0186, "step": 6950 }, { "grad_norm": 0.43856969475746155, "learning_rate": 9.121723261293794e-05, "loss": 0.0229, "step": 6960 }, { "grad_norm": 0.5875193476676941, "learning_rate": 9.118600725060214e-05, "loss": 0.018, "step": 6970 }, { "grad_norm": 0.3514130413532257, "learning_rate": 9.115473184338392e-05, "loss": 0.0182, "step": 6980 }, { "grad_norm": 0.42321932315826416, "learning_rate": 9.11234064292859e-05, "loss": 0.0184, "step": 6990 }, { "grad_norm": 0.4420258700847626, "learning_rate": 9.109203104637137e-05, "loss": 0.0186, "step": 7000 }, { "grad_norm": 0.48635879158973694, "learning_rate": 9.10606057327644e-05, "loss": 0.024, "step": 7010 }, { "grad_norm": 0.400319367647171, "learning_rate": 9.10291305266497e-05, "loss": 0.0239, "step": 7020 }, { "grad_norm": 0.5533655285835266, "learning_rate": 9.099760546627261e-05, "loss": 0.0182, "step": 7030 }, { "grad_norm": 0.5002222657203674, "learning_rate": 9.096603058993907e-05, "loss": 0.0182, "step": 7040 }, { "grad_norm": 0.4045437276363373, "learning_rate": 9.093440593601552e-05, "loss": 0.0178, "step": 7050 }, { "grad_norm": 0.39968350529670715, "learning_rate": 9.090273154292889e-05, "loss": 0.0175, "step": 7060 }, { "grad_norm": 0.3576194643974304, "learning_rate": 9.087100744916656e-05, "loss": 0.0213, "step": 7070 }, { "grad_norm": 0.4119541049003601, "learning_rate": 9.08392336932763e-05, "loss": 0.0185, "step": 7080 }, { "grad_norm": 0.5328904986381531, "learning_rate": 9.080741031386619e-05, "loss": 0.0167, "step": 7090 }, { "grad_norm": 0.43807947635650635, "learning_rate": 9.077553734960468e-05, "loss": 0.0193, "step": 7100 }, { "grad_norm": 0.46781161427497864, "learning_rate": 9.074361483922041e-05, "loss": 0.0165, "step": 7110 }, { "grad_norm": 0.40142184495925903, "learning_rate": 9.071164282150224e-05, "loss": 0.0187, "step": 7120 }, { "grad_norm": 0.3468552529811859, "learning_rate": 9.067962133529919e-05, "loss": 0.0143, "step": 7130 }, { "grad_norm": 0.48054182529449463, "learning_rate": 9.064755041952036e-05, "loss": 0.0175, "step": 7140 }, { "grad_norm": 0.35339662432670593, "learning_rate": 9.061543011313497e-05, "loss": 0.0171, "step": 7150 }, { "grad_norm": 0.36724406480789185, "learning_rate": 9.05832604551722e-05, "loss": 0.0192, "step": 7160 }, { "grad_norm": 0.30261459946632385, "learning_rate": 9.055104148472124e-05, "loss": 0.017, "step": 7170 }, { "grad_norm": 0.44181281328201294, "learning_rate": 9.051877324093114e-05, "loss": 0.0171, "step": 7180 }, { "grad_norm": 0.4808070659637451, "learning_rate": 9.04864557630109e-05, "loss": 0.0213, "step": 7190 }, { "grad_norm": 0.3853558301925659, "learning_rate": 9.045408909022928e-05, "loss": 0.0171, "step": 7200 }, { "grad_norm": 0.3714452087879181, "learning_rate": 9.042167326191484e-05, "loss": 0.0157, "step": 7210 }, { "grad_norm": 0.4355736970901489, "learning_rate": 9.038920831745587e-05, "loss": 0.014, "step": 7220 }, { "grad_norm": 0.43134063482284546, "learning_rate": 9.035669429630036e-05, "loss": 0.0148, "step": 7230 }, { "grad_norm": 0.3626321256160736, "learning_rate": 9.032413123795588e-05, "loss": 0.0167, "step": 7240 }, { "grad_norm": 0.4554193615913391, "learning_rate": 9.029151918198962e-05, "loss": 0.0174, "step": 7250 }, { "grad_norm": 0.43733537197113037, "learning_rate": 9.025885816802833e-05, "loss": 0.0166, "step": 7260 }, { "grad_norm": 0.46259748935699463, "learning_rate": 9.02261482357582e-05, "loss": 0.0176, "step": 7270 }, { "grad_norm": 0.3950237035751343, "learning_rate": 9.019338942492485e-05, "loss": 0.0137, "step": 7280 }, { "grad_norm": 0.43521520495414734, "learning_rate": 9.01605817753334e-05, "loss": 0.0156, "step": 7290 }, { "grad_norm": 0.4026656746864319, "learning_rate": 9.012772532684818e-05, "loss": 0.0177, "step": 7300 }, { "grad_norm": 0.5009707808494568, "learning_rate": 9.00948201193929e-05, "loss": 0.0161, "step": 7310 }, { "grad_norm": 0.5638161897659302, "learning_rate": 9.006186619295047e-05, "loss": 0.0198, "step": 7320 }, { "grad_norm": 0.3717515468597412, "learning_rate": 9.002886358756305e-05, "loss": 0.0209, "step": 7330 }, { "grad_norm": 0.4115362763404846, "learning_rate": 8.999581234333188e-05, "loss": 0.0238, "step": 7340 }, { "grad_norm": 0.4044981002807617, "learning_rate": 8.996271250041734e-05, "loss": 0.0171, "step": 7350 }, { "grad_norm": 0.39848950505256653, "learning_rate": 8.99295640990389e-05, "loss": 0.0156, "step": 7360 }, { "grad_norm": 0.3901309072971344, "learning_rate": 8.989636717947494e-05, "loss": 0.0175, "step": 7370 }, { "grad_norm": 0.46779510378837585, "learning_rate": 8.986312178206291e-05, "loss": 0.0226, "step": 7380 }, { "grad_norm": 0.3916587829589844, "learning_rate": 8.982982794719904e-05, "loss": 0.0155, "step": 7390 }, { "grad_norm": 0.4136561453342438, "learning_rate": 8.979648571533852e-05, "loss": 0.013, "step": 7400 }, { "grad_norm": 0.41917094588279724, "learning_rate": 8.976309512699528e-05, "loss": 0.0179, "step": 7410 }, { "grad_norm": 0.4538816511631012, "learning_rate": 8.972965622274205e-05, "loss": 0.0185, "step": 7420 }, { "grad_norm": 0.44878342747688293, "learning_rate": 8.969616904321026e-05, "loss": 0.0152, "step": 7430 }, { "grad_norm": 0.3817518353462219, "learning_rate": 8.966263362908998e-05, "loss": 0.0183, "step": 7440 }, { "grad_norm": 0.4719610810279846, "learning_rate": 8.962905002112989e-05, "loss": 0.0156, "step": 7450 }, { "grad_norm": 0.39098188281059265, "learning_rate": 8.959541826013724e-05, "loss": 0.017, "step": 7460 }, { "grad_norm": 0.4522194266319275, "learning_rate": 8.956173838697781e-05, "loss": 0.0158, "step": 7470 }, { "grad_norm": 0.43008607625961304, "learning_rate": 8.952801044257581e-05, "loss": 0.0156, "step": 7480 }, { "grad_norm": 0.498237282037735, "learning_rate": 8.949423446791387e-05, "loss": 0.0132, "step": 7490 }, { "grad_norm": 0.3895258903503418, "learning_rate": 8.9460410504033e-05, "loss": 0.0163, "step": 7500 }, { "grad_norm": 0.42185407876968384, "learning_rate": 8.942653859203248e-05, "loss": 0.018, "step": 7510 }, { "grad_norm": 0.43893375992774963, "learning_rate": 8.93926187730699e-05, "loss": 0.0251, "step": 7520 }, { "grad_norm": 0.395225465297699, "learning_rate": 8.935865108836102e-05, "loss": 0.0169, "step": 7530 }, { "grad_norm": 0.36842837929725647, "learning_rate": 8.932463557917981e-05, "loss": 0.0189, "step": 7540 }, { "grad_norm": 0.34908992052078247, "learning_rate": 8.929057228685829e-05, "loss": 0.0137, "step": 7550 }, { "grad_norm": 0.3974362015724182, "learning_rate": 8.925646125278658e-05, "loss": 0.0135, "step": 7560 }, { "grad_norm": 0.3804769217967987, "learning_rate": 8.92223025184128e-05, "loss": 0.0163, "step": 7570 }, { "grad_norm": 0.4336641728878021, "learning_rate": 8.918809612524303e-05, "loss": 0.0177, "step": 7580 }, { "grad_norm": 0.41866055130958557, "learning_rate": 8.91538421148413e-05, "loss": 0.0161, "step": 7590 }, { "grad_norm": 0.5138710141181946, "learning_rate": 8.91195405288294e-05, "loss": 0.0171, "step": 7600 }, { "grad_norm": 0.4078211486339569, "learning_rate": 8.908519140888705e-05, "loss": 0.0177, "step": 7610 }, { "grad_norm": 0.5114139914512634, "learning_rate": 8.905079479675164e-05, "loss": 0.0232, "step": 7620 }, { "grad_norm": 0.5226527452468872, "learning_rate": 8.901635073421831e-05, "loss": 0.0192, "step": 7630 }, { "grad_norm": 0.4252995252609253, "learning_rate": 8.898185926313983e-05, "loss": 0.0168, "step": 7640 }, { "grad_norm": 0.48279619216918945, "learning_rate": 8.894732042542659e-05, "loss": 0.0165, "step": 7650 }, { "grad_norm": 0.39267364144325256, "learning_rate": 8.891273426304654e-05, "loss": 0.0183, "step": 7660 }, { "grad_norm": 0.4174197018146515, "learning_rate": 8.887810081802515e-05, "loss": 0.0234, "step": 7670 }, { "grad_norm": 0.4622577428817749, "learning_rate": 8.884342013244529e-05, "loss": 0.0174, "step": 7680 }, { "grad_norm": 0.31801873445510864, "learning_rate": 8.880869224844725e-05, "loss": 0.0143, "step": 7690 }, { "grad_norm": 0.47394227981567383, "learning_rate": 8.877391720822875e-05, "loss": 0.0202, "step": 7700 }, { "grad_norm": 0.36641019582748413, "learning_rate": 8.873909505404468e-05, "loss": 0.0174, "step": 7710 }, { "grad_norm": 0.31967946887016296, "learning_rate": 8.870422582820726e-05, "loss": 0.0152, "step": 7720 }, { "grad_norm": 0.3392026126384735, "learning_rate": 8.866930957308588e-05, "loss": 0.0137, "step": 7730 }, { "grad_norm": 0.4585781693458557, "learning_rate": 8.86343463311071e-05, "loss": 0.0146, "step": 7740 }, { "grad_norm": 0.4606214761734009, "learning_rate": 8.859933614475452e-05, "loss": 0.0151, "step": 7750 }, { "grad_norm": 0.3915558457374573, "learning_rate": 8.85642790565689e-05, "loss": 0.0149, "step": 7760 }, { "grad_norm": 0.3262130320072174, "learning_rate": 8.852917510914782e-05, "loss": 0.0121, "step": 7770 }, { "grad_norm": 0.45582315325737, "learning_rate": 8.84940243451459e-05, "loss": 0.0168, "step": 7780 }, { "grad_norm": 0.369716078042984, "learning_rate": 8.845882680727469e-05, "loss": 0.016, "step": 7790 }, { "grad_norm": 0.36071422696113586, "learning_rate": 8.842358253830246e-05, "loss": 0.0161, "step": 7800 }, { "grad_norm": 0.29536616802215576, "learning_rate": 8.838829158105434e-05, "loss": 0.0143, "step": 7810 }, { "grad_norm": 0.3977418839931488, "learning_rate": 8.835295397841217e-05, "loss": 0.016, "step": 7820 }, { "grad_norm": 0.3673246502876282, "learning_rate": 8.831756977331447e-05, "loss": 0.0157, "step": 7830 }, { "grad_norm": 0.3096219599246979, "learning_rate": 8.828213900875638e-05, "loss": 0.0124, "step": 7840 }, { "grad_norm": 0.33677202463150024, "learning_rate": 8.824666172778964e-05, "loss": 0.0125, "step": 7850 }, { "grad_norm": 0.44679221510887146, "learning_rate": 8.821113797352245e-05, "loss": 0.0141, "step": 7860 }, { "grad_norm": 0.4535770118236542, "learning_rate": 8.817556778911956e-05, "loss": 0.0145, "step": 7870 }, { "grad_norm": 0.36395806074142456, "learning_rate": 8.81399512178021e-05, "loss": 0.0172, "step": 7880 }, { "grad_norm": 0.3134922981262207, "learning_rate": 8.810428830284752e-05, "loss": 0.0134, "step": 7890 }, { "grad_norm": 0.4432116150856018, "learning_rate": 8.806857908758967e-05, "loss": 0.0121, "step": 7900 }, { "grad_norm": 0.42387428879737854, "learning_rate": 8.80328236154186e-05, "loss": 0.0154, "step": 7910 }, { "grad_norm": 0.3786136507987976, "learning_rate": 8.799702192978056e-05, "loss": 0.0142, "step": 7920 }, { "grad_norm": 0.3870513439178467, "learning_rate": 8.7961174074178e-05, "loss": 0.0177, "step": 7930 }, { "grad_norm": 0.4015447199344635, "learning_rate": 8.792528009216942e-05, "loss": 0.0148, "step": 7940 }, { "grad_norm": 0.3563585877418518, "learning_rate": 8.788934002736943e-05, "loss": 0.0113, "step": 7950 }, { "grad_norm": 0.40029364824295044, "learning_rate": 8.785335392344857e-05, "loss": 0.0186, "step": 7960 }, { "grad_norm": 0.38724836707115173, "learning_rate": 8.781732182413336e-05, "loss": 0.0128, "step": 7970 }, { "grad_norm": 0.42172935605049133, "learning_rate": 8.778124377320618e-05, "loss": 0.0152, "step": 7980 }, { "grad_norm": 0.3535851538181305, "learning_rate": 8.774511981450529e-05, "loss": 0.0163, "step": 7990 }, { "grad_norm": 0.6055047512054443, "learning_rate": 8.770894999192467e-05, "loss": 0.0133, "step": 8000 }, { "grad_norm": 0.43954846262931824, "learning_rate": 8.767273434941412e-05, "loss": 0.0165, "step": 8010 }, { "grad_norm": 0.4815202057361603, "learning_rate": 8.7636472930979e-05, "loss": 0.0197, "step": 8020 }, { "grad_norm": 0.44193658232688904, "learning_rate": 8.76001657806804e-05, "loss": 0.0184, "step": 8030 }, { "grad_norm": 0.4318656027317047, "learning_rate": 8.756381294263491e-05, "loss": 0.015, "step": 8040 }, { "grad_norm": 0.310991108417511, "learning_rate": 8.752741446101464e-05, "loss": 0.0127, "step": 8050 }, { "grad_norm": 0.3603052794933319, "learning_rate": 8.749097038004722e-05, "loss": 0.0122, "step": 8060 }, { "grad_norm": 0.361881285905838, "learning_rate": 8.745448074401561e-05, "loss": 0.012, "step": 8070 }, { "grad_norm": 0.3434341847896576, "learning_rate": 8.741794559725818e-05, "loss": 0.0174, "step": 8080 }, { "grad_norm": 0.34320276975631714, "learning_rate": 8.738136498416856e-05, "loss": 0.0189, "step": 8090 }, { "grad_norm": 0.40803414583206177, "learning_rate": 8.734473894919564e-05, "loss": 0.0146, "step": 8100 }, { "grad_norm": 0.3302277624607086, "learning_rate": 8.730806753684353e-05, "loss": 0.0141, "step": 8110 }, { "grad_norm": 0.2673576772212982, "learning_rate": 8.727135079167144e-05, "loss": 0.0136, "step": 8120 }, { "grad_norm": 0.35611194372177124, "learning_rate": 8.723458875829368e-05, "loss": 0.0123, "step": 8130 }, { "grad_norm": 0.31662964820861816, "learning_rate": 8.719778148137958e-05, "loss": 0.0135, "step": 8140 }, { "grad_norm": 0.304119348526001, "learning_rate": 8.716092900565346e-05, "loss": 0.0138, "step": 8150 }, { "grad_norm": 0.432702898979187, "learning_rate": 8.712403137589456e-05, "loss": 0.0142, "step": 8160 }, { "grad_norm": 0.33718615770339966, "learning_rate": 8.708708863693697e-05, "loss": 0.0135, "step": 8170 }, { "grad_norm": 0.37949517369270325, "learning_rate": 8.70501008336696e-05, "loss": 0.0148, "step": 8180 }, { "grad_norm": 0.4232802093029022, "learning_rate": 8.701306801103611e-05, "loss": 0.0128, "step": 8190 }, { "grad_norm": 0.35351768136024475, "learning_rate": 8.697599021403489e-05, "loss": 0.0138, "step": 8200 }, { "grad_norm": 0.3234887421131134, "learning_rate": 8.693886748771896e-05, "loss": 0.0122, "step": 8210 }, { "grad_norm": 0.3727528154850006, "learning_rate": 8.690169987719593e-05, "loss": 0.0125, "step": 8220 }, { "grad_norm": 0.3550198972225189, "learning_rate": 8.686448742762791e-05, "loss": 0.0155, "step": 8230 }, { "grad_norm": 0.40745657682418823, "learning_rate": 8.682723018423161e-05, "loss": 0.0142, "step": 8240 }, { "grad_norm": 0.287275493144989, "learning_rate": 8.678992819227804e-05, "loss": 0.0134, "step": 8250 }, { "grad_norm": 0.5498289465904236, "learning_rate": 8.675258149709265e-05, "loss": 0.0131, "step": 8260 }, { "grad_norm": 0.32954439520835876, "learning_rate": 8.67151901440552e-05, "loss": 0.0133, "step": 8270 }, { "grad_norm": 0.39001691341400146, "learning_rate": 8.667775417859971e-05, "loss": 0.0115, "step": 8280 }, { "grad_norm": 0.36302289366722107, "learning_rate": 8.664027364621441e-05, "loss": 0.0116, "step": 8290 }, { "grad_norm": 0.3959016799926758, "learning_rate": 8.660274859244168e-05, "loss": 0.012, "step": 8300 }, { "grad_norm": 0.4407598674297333, "learning_rate": 8.656517906287799e-05, "loss": 0.0124, "step": 8310 }, { "grad_norm": 0.3001246154308319, "learning_rate": 8.652756510317387e-05, "loss": 0.0135, "step": 8320 }, { "grad_norm": 0.4005204141139984, "learning_rate": 8.648990675903382e-05, "loss": 0.0142, "step": 8330 }, { "grad_norm": 0.4269216060638428, "learning_rate": 8.64522040762163e-05, "loss": 0.0168, "step": 8340 }, { "grad_norm": 0.39072999358177185, "learning_rate": 8.64144571005336e-05, "loss": 0.023, "step": 8350 }, { "grad_norm": 0.4220088720321655, "learning_rate": 8.637666587785184e-05, "loss": 0.0128, "step": 8360 }, { "grad_norm": 0.6169581413269043, "learning_rate": 8.633883045409096e-05, "loss": 0.0169, "step": 8370 }, { "grad_norm": 0.3867091238498688, "learning_rate": 8.630095087522457e-05, "loss": 0.0129, "step": 8380 }, { "grad_norm": 0.406647652387619, "learning_rate": 8.62630271872799e-05, "loss": 0.0133, "step": 8390 }, { "grad_norm": 0.3207695782184601, "learning_rate": 8.62250594363378e-05, "loss": 0.0145, "step": 8400 }, { "grad_norm": 0.3849739134311676, "learning_rate": 8.61870476685327e-05, "loss": 0.0135, "step": 8410 }, { "grad_norm": 0.39395564794540405, "learning_rate": 8.614899193005248e-05, "loss": 0.0133, "step": 8420 }, { "grad_norm": 0.4138476550579071, "learning_rate": 8.611089226713842e-05, "loss": 0.0127, "step": 8430 }, { "grad_norm": 0.5089305639266968, "learning_rate": 8.607274872608521e-05, "loss": 0.016, "step": 8440 }, { "grad_norm": 0.3301273286342621, "learning_rate": 8.603456135324088e-05, "loss": 0.012, "step": 8450 }, { "grad_norm": 0.3202885091304779, "learning_rate": 8.599633019500665e-05, "loss": 0.0137, "step": 8460 }, { "grad_norm": 0.41656526923179626, "learning_rate": 8.595805529783702e-05, "loss": 0.0144, "step": 8470 }, { "grad_norm": 0.3133356273174286, "learning_rate": 8.591973670823958e-05, "loss": 0.0168, "step": 8480 }, { "grad_norm": 0.32465919852256775, "learning_rate": 8.588137447277503e-05, "loss": 0.0201, "step": 8490 }, { "grad_norm": 0.36198797821998596, "learning_rate": 8.584296863805709e-05, "loss": 0.0146, "step": 8500 }, { "grad_norm": 0.32784873247146606, "learning_rate": 8.580451925075248e-05, "loss": 0.0178, "step": 8510 }, { "grad_norm": 0.3644600808620453, "learning_rate": 8.576602635758085e-05, "loss": 0.0168, "step": 8520 }, { "grad_norm": 0.3618288040161133, "learning_rate": 8.572749000531469e-05, "loss": 0.0156, "step": 8530 }, { "grad_norm": 0.43166637420654297, "learning_rate": 8.568891024077926e-05, "loss": 0.0139, "step": 8540 }, { "grad_norm": 0.4783613085746765, "learning_rate": 8.565028711085265e-05, "loss": 0.0148, "step": 8550 }, { "grad_norm": 0.3569784462451935, "learning_rate": 8.561162066246561e-05, "loss": 0.0149, "step": 8560 }, { "grad_norm": 0.3447094261646271, "learning_rate": 8.557291094260152e-05, "loss": 0.0127, "step": 8570 }, { "grad_norm": 0.4028734266757965, "learning_rate": 8.553415799829629e-05, "loss": 0.0099, "step": 8580 }, { "grad_norm": 0.41679948568344116, "learning_rate": 8.549536187663847e-05, "loss": 0.0102, "step": 8590 }, { "grad_norm": 0.5448471903800964, "learning_rate": 8.545652262476897e-05, "loss": 0.014, "step": 8600 }, { "grad_norm": 0.415286123752594, "learning_rate": 8.541764028988115e-05, "loss": 0.0108, "step": 8610 }, { "grad_norm": 0.27212387323379517, "learning_rate": 8.537871491922071e-05, "loss": 0.0113, "step": 8620 }, { "grad_norm": 0.2904397249221802, "learning_rate": 8.533974656008566e-05, "loss": 0.0101, "step": 8630 }, { "grad_norm": 0.35041171312332153, "learning_rate": 8.530073525982621e-05, "loss": 0.0107, "step": 8640 }, { "grad_norm": 0.32075655460357666, "learning_rate": 8.526168106584476e-05, "loss": 0.0131, "step": 8650 }, { "grad_norm": 0.442484974861145, "learning_rate": 8.522258402559586e-05, "loss": 0.0129, "step": 8660 }, { "grad_norm": 0.29626667499542236, "learning_rate": 8.518344418658613e-05, "loss": 0.0103, "step": 8670 }, { "grad_norm": 0.4161774516105652, "learning_rate": 8.51442615963741e-05, "loss": 0.0119, "step": 8680 }, { "grad_norm": 0.5368458032608032, "learning_rate": 8.510503630257034e-05, "loss": 0.0134, "step": 8690 }, { "grad_norm": 0.3856920003890991, "learning_rate": 8.50657683528373e-05, "loss": 0.0135, "step": 8700 }, { "grad_norm": 0.352224737405777, "learning_rate": 8.502645779488922e-05, "loss": 0.0127, "step": 8710 }, { "grad_norm": 0.3420656621456146, "learning_rate": 8.498710467649213e-05, "loss": 0.0129, "step": 8720 }, { "grad_norm": 0.41603636741638184, "learning_rate": 8.494770904546381e-05, "loss": 0.0135, "step": 8730 }, { "grad_norm": 0.4066091477870941, "learning_rate": 8.490827094967363e-05, "loss": 0.0131, "step": 8740 }, { "grad_norm": 0.3570132255554199, "learning_rate": 8.486879043704263e-05, "loss": 0.0112, "step": 8750 }, { "grad_norm": 0.30130037665367126, "learning_rate": 8.482926755554332e-05, "loss": 0.0138, "step": 8760 }, { "grad_norm": 0.27718329429626465, "learning_rate": 8.478970235319975e-05, "loss": 0.0106, "step": 8770 }, { "grad_norm": 0.33127468824386597, "learning_rate": 8.475009487808737e-05, "loss": 0.014, "step": 8780 }, { "grad_norm": 0.3403458595275879, "learning_rate": 8.471044517833298e-05, "loss": 0.0108, "step": 8790 }, { "grad_norm": 0.28461599349975586, "learning_rate": 8.467075330211474e-05, "loss": 0.0103, "step": 8800 }, { "grad_norm": 0.357270210981369, "learning_rate": 8.463101929766198e-05, "loss": 0.0107, "step": 8810 }, { "grad_norm": 0.3840462267398834, "learning_rate": 8.459124321325528e-05, "loss": 0.0143, "step": 8820 }, { "grad_norm": 0.3820178508758545, "learning_rate": 8.455142509722634e-05, "loss": 0.0159, "step": 8830 }, { "grad_norm": 0.420579195022583, "learning_rate": 8.451156499795792e-05, "loss": 0.0131, "step": 8840 }, { "grad_norm": 0.3685382902622223, "learning_rate": 8.447166296388381e-05, "loss": 0.0115, "step": 8850 }, { "grad_norm": 0.32259470224380493, "learning_rate": 8.443171904348872e-05, "loss": 0.0151, "step": 8860 }, { "grad_norm": 0.3867305517196655, "learning_rate": 8.43917332853083e-05, "loss": 0.013, "step": 8870 }, { "grad_norm": 0.3915140628814697, "learning_rate": 8.435170573792901e-05, "loss": 0.012, "step": 8880 }, { "grad_norm": 0.36622855067253113, "learning_rate": 8.431163644998808e-05, "loss": 0.0127, "step": 8890 }, { "grad_norm": 0.3774382174015045, "learning_rate": 8.42715254701735e-05, "loss": 0.0134, "step": 8900 }, { "grad_norm": 0.4195428490638733, "learning_rate": 8.42313728472239e-05, "loss": 0.0115, "step": 8910 }, { "grad_norm": 0.31390440464019775, "learning_rate": 8.419117862992844e-05, "loss": 0.0111, "step": 8920 }, { "grad_norm": 0.33561190962791443, "learning_rate": 8.415094286712694e-05, "loss": 0.0095, "step": 8930 }, { "grad_norm": 0.33464881777763367, "learning_rate": 8.411066560770965e-05, "loss": 0.0126, "step": 8940 }, { "grad_norm": 0.37910646200180054, "learning_rate": 8.407034690061721e-05, "loss": 0.0139, "step": 8950 }, { "grad_norm": 0.2993716597557068, "learning_rate": 8.402998679484067e-05, "loss": 0.011, "step": 8960 }, { "grad_norm": 0.4253138303756714, "learning_rate": 8.398958533942135e-05, "loss": 0.0115, "step": 8970 }, { "grad_norm": 0.2937369644641876, "learning_rate": 8.394914258345083e-05, "loss": 0.0119, "step": 8980 }, { "grad_norm": 0.31378665566444397, "learning_rate": 8.390865857607089e-05, "loss": 0.0097, "step": 8990 }, { "grad_norm": 0.22702014446258545, "learning_rate": 8.386813336647339e-05, "loss": 0.0103, "step": 9000 }, { "grad_norm": 0.33617955446243286, "learning_rate": 8.38275670039003e-05, "loss": 0.0109, "step": 9010 }, { "grad_norm": 0.3161669075489044, "learning_rate": 8.378695953764356e-05, "loss": 0.0097, "step": 9020 }, { "grad_norm": 0.3696955740451813, "learning_rate": 8.374631101704509e-05, "loss": 0.0103, "step": 9030 }, { "grad_norm": 0.32343894243240356, "learning_rate": 8.370562149149665e-05, "loss": 0.0095, "step": 9040 }, { "grad_norm": 0.3361647427082062, "learning_rate": 8.366489101043989e-05, "loss": 0.0165, "step": 9050 }, { "grad_norm": 0.3349011242389679, "learning_rate": 8.362411962336614e-05, "loss": 0.0133, "step": 9060 }, { "grad_norm": 0.35757875442504883, "learning_rate": 8.358330737981651e-05, "loss": 0.0123, "step": 9070 }, { "grad_norm": 0.32888340950012207, "learning_rate": 8.35424543293817e-05, "loss": 0.0121, "step": 9080 }, { "grad_norm": 0.3183457851409912, "learning_rate": 8.350156052170206e-05, "loss": 0.0109, "step": 9090 }, { "grad_norm": 0.31766706705093384, "learning_rate": 8.346062600646738e-05, "loss": 0.0096, "step": 9100 }, { "grad_norm": 0.37983378767967224, "learning_rate": 8.341965083341697e-05, "loss": 0.0107, "step": 9110 }, { "grad_norm": 0.3250914514064789, "learning_rate": 8.337863505233953e-05, "loss": 0.0111, "step": 9120 }, { "grad_norm": 0.36802181601524353, "learning_rate": 8.33375787130731e-05, "loss": 0.0107, "step": 9130 }, { "grad_norm": 0.4129602313041687, "learning_rate": 8.3296481865505e-05, "loss": 0.0131, "step": 9140 }, { "grad_norm": 0.33686500787734985, "learning_rate": 8.325534455957181e-05, "loss": 0.0106, "step": 9150 }, { "grad_norm": 0.30343806743621826, "learning_rate": 8.321416684525917e-05, "loss": 0.01, "step": 9160 }, { "grad_norm": 0.35890650749206543, "learning_rate": 8.317294877260193e-05, "loss": 0.0131, "step": 9170 }, { "grad_norm": 0.29076987504959106, "learning_rate": 8.313169039168394e-05, "loss": 0.0103, "step": 9180 }, { "grad_norm": 0.40628916025161743, "learning_rate": 8.3090391752638e-05, "loss": 0.0152, "step": 9190 }, { "grad_norm": 0.33720865845680237, "learning_rate": 8.304905290564586e-05, "loss": 0.0114, "step": 9200 }, { "grad_norm": 0.31990286707878113, "learning_rate": 8.300767390093813e-05, "loss": 0.0115, "step": 9210 }, { "grad_norm": 0.33707889914512634, "learning_rate": 8.296625478879417e-05, "loss": 0.0105, "step": 9220 }, { "grad_norm": 0.3644105792045593, "learning_rate": 8.292479561954214e-05, "loss": 0.014, "step": 9230 }, { "grad_norm": 0.3047928512096405, "learning_rate": 8.288329644355883e-05, "loss": 0.0111, "step": 9240 }, { "grad_norm": 0.29711413383483887, "learning_rate": 8.284175731126965e-05, "loss": 0.0101, "step": 9250 }, { "grad_norm": 0.40911582112312317, "learning_rate": 8.280017827314854e-05, "loss": 0.0158, "step": 9260 }, { "grad_norm": 0.3769160509109497, "learning_rate": 8.2758559379718e-05, "loss": 0.0111, "step": 9270 }, { "grad_norm": 0.3644101321697235, "learning_rate": 8.271690068154887e-05, "loss": 0.0127, "step": 9280 }, { "grad_norm": 0.39113473892211914, "learning_rate": 8.26752022292604e-05, "loss": 0.0089, "step": 9290 }, { "grad_norm": 0.4182140827178955, "learning_rate": 8.263346407352018e-05, "loss": 0.011, "step": 9300 }, { "grad_norm": 0.3185349404811859, "learning_rate": 8.259168626504395e-05, "loss": 0.0121, "step": 9310 }, { "grad_norm": 0.42274507880210876, "learning_rate": 8.254986885459572e-05, "loss": 0.0116, "step": 9320 }, { "grad_norm": 0.3910265862941742, "learning_rate": 8.250801189298757e-05, "loss": 0.014, "step": 9330 }, { "grad_norm": 0.42234018445014954, "learning_rate": 8.246611543107967e-05, "loss": 0.012, "step": 9340 }, { "grad_norm": 0.36422139406204224, "learning_rate": 8.242417951978018e-05, "loss": 0.0115, "step": 9350 }, { "grad_norm": 0.3053525388240814, "learning_rate": 8.238220421004517e-05, "loss": 0.0135, "step": 9360 }, { "grad_norm": 0.27072787284851074, "learning_rate": 8.234018955287859e-05, "loss": 0.0127, "step": 9370 }, { "grad_norm": 0.37784138321876526, "learning_rate": 8.229813559933224e-05, "loss": 0.0125, "step": 9380 }, { "grad_norm": 0.362018883228302, "learning_rate": 8.22560424005056e-05, "loss": 0.0166, "step": 9390 }, { "grad_norm": 0.2741208076477051, "learning_rate": 8.22139100075459e-05, "loss": 0.0102, "step": 9400 }, { "grad_norm": 0.4052687883377075, "learning_rate": 8.217173847164799e-05, "loss": 0.0134, "step": 9410 }, { "grad_norm": 0.3573543429374695, "learning_rate": 8.212952784405422e-05, "loss": 0.0125, "step": 9420 }, { "grad_norm": 0.2816816568374634, "learning_rate": 8.208727817605451e-05, "loss": 0.0094, "step": 9430 }, { "grad_norm": 0.3260761499404907, "learning_rate": 8.204498951898618e-05, "loss": 0.0137, "step": 9440 }, { "grad_norm": 0.39164671301841736, "learning_rate": 8.200266192423395e-05, "loss": 0.0141, "step": 9450 }, { "grad_norm": 0.3788048326969147, "learning_rate": 8.196029544322981e-05, "loss": 0.0105, "step": 9460 }, { "grad_norm": 0.31721818447113037, "learning_rate": 8.191789012745305e-05, "loss": 0.0109, "step": 9470 }, { "grad_norm": 0.2978890836238861, "learning_rate": 8.187544602843014e-05, "loss": 0.0116, "step": 9480 }, { "grad_norm": 0.41037440299987793, "learning_rate": 8.183296319773466e-05, "loss": 0.0101, "step": 9490 }, { "grad_norm": 0.31313514709472656, "learning_rate": 8.179044168698721e-05, "loss": 0.0151, "step": 9500 }, { "grad_norm": 0.43483027815818787, "learning_rate": 8.174788154785548e-05, "loss": 0.0112, "step": 9510 }, { "grad_norm": 0.33034300804138184, "learning_rate": 8.170528283205403e-05, "loss": 0.0105, "step": 9520 }, { "grad_norm": 0.41514691710472107, "learning_rate": 8.166264559134435e-05, "loss": 0.0098, "step": 9530 }, { "grad_norm": 0.2780282497406006, "learning_rate": 8.161996987753467e-05, "loss": 0.0119, "step": 9540 }, { "grad_norm": 0.32566148042678833, "learning_rate": 8.157725574248e-05, "loss": 0.0132, "step": 9550 }, { "grad_norm": 0.3275966942310333, "learning_rate": 8.153450323808209e-05, "loss": 0.0122, "step": 9560 }, { "grad_norm": 0.27491697669029236, "learning_rate": 8.149171241628923e-05, "loss": 0.0094, "step": 9570 }, { "grad_norm": 0.27083325386047363, "learning_rate": 8.144888332909631e-05, "loss": 0.0127, "step": 9580 }, { "grad_norm": 0.3441878855228424, "learning_rate": 8.14060160285447e-05, "loss": 0.012, "step": 9590 }, { "grad_norm": 0.3230050206184387, "learning_rate": 8.136311056672223e-05, "loss": 0.0141, "step": 9600 }, { "grad_norm": 0.2915610671043396, "learning_rate": 8.132016699576308e-05, "loss": 0.0107, "step": 9610 }, { "grad_norm": 0.31620246171951294, "learning_rate": 8.127718536784771e-05, "loss": 0.0083, "step": 9620 }, { "grad_norm": 0.34850651025772095, "learning_rate": 8.12341657352029e-05, "loss": 0.0115, "step": 9630 }, { "grad_norm": 0.26506200432777405, "learning_rate": 8.119110815010151e-05, "loss": 0.0089, "step": 9640 }, { "grad_norm": 0.2858802080154419, "learning_rate": 8.114801266486261e-05, "loss": 0.0108, "step": 9650 }, { "grad_norm": 0.35112622380256653, "learning_rate": 8.110487933185122e-05, "loss": 0.0123, "step": 9660 }, { "grad_norm": 0.37826859951019287, "learning_rate": 8.106170820347847e-05, "loss": 0.0098, "step": 9670 }, { "grad_norm": 0.3300885558128357, "learning_rate": 8.101849933220132e-05, "loss": 0.0133, "step": 9680 }, { "grad_norm": 0.373850017786026, "learning_rate": 8.097525277052264e-05, "loss": 0.0101, "step": 9690 }, { "grad_norm": 0.3447759449481964, "learning_rate": 8.093196857099106e-05, "loss": 0.0111, "step": 9700 }, { "grad_norm": 0.4454047977924347, "learning_rate": 8.088864678620096e-05, "loss": 0.0139, "step": 9710 }, { "grad_norm": 0.38471096754074097, "learning_rate": 8.084528746879243e-05, "loss": 0.0136, "step": 9720 }, { "grad_norm": 0.4203638732433319, "learning_rate": 8.080189067145108e-05, "loss": 0.0115, "step": 9730 }, { "grad_norm": 0.2767387926578522, "learning_rate": 8.075845644690813e-05, "loss": 0.0086, "step": 9740 }, { "grad_norm": 0.2493705302476883, "learning_rate": 8.07149848479403e-05, "loss": 0.0112, "step": 9750 }, { "grad_norm": 0.5050793290138245, "learning_rate": 8.067147592736962e-05, "loss": 0.0161, "step": 9760 }, { "grad_norm": 0.37258148193359375, "learning_rate": 8.062792973806357e-05, "loss": 0.0158, "step": 9770 }, { "grad_norm": 0.27847206592559814, "learning_rate": 8.058434633293485e-05, "loss": 0.0143, "step": 9780 }, { "grad_norm": 0.36785247921943665, "learning_rate": 8.054072576494143e-05, "loss": 0.0161, "step": 9790 }, { "grad_norm": 0.3211102783679962, "learning_rate": 8.04970680870864e-05, "loss": 0.012, "step": 9800 }, { "grad_norm": 0.3535841405391693, "learning_rate": 8.045337335241793e-05, "loss": 0.0158, "step": 9810 }, { "grad_norm": 0.31418851017951965, "learning_rate": 8.040964161402932e-05, "loss": 0.012, "step": 9820 }, { "grad_norm": 0.4190634489059448, "learning_rate": 8.036587292505869e-05, "loss": 0.0133, "step": 9830 }, { "grad_norm": 0.30274930596351624, "learning_rate": 8.032206733868912e-05, "loss": 0.0144, "step": 9840 }, { "grad_norm": 0.3628658950328827, "learning_rate": 8.027822490814858e-05, "loss": 0.0122, "step": 9850 }, { "grad_norm": 0.3321343958377838, "learning_rate": 8.023434568670972e-05, "loss": 0.0119, "step": 9860 }, { "grad_norm": 0.45050016045570374, "learning_rate": 8.019042972768992e-05, "loss": 0.0117, "step": 9870 }, { "grad_norm": 0.29774221777915955, "learning_rate": 8.014647708445124e-05, "loss": 0.0117, "step": 9880 }, { "grad_norm": 0.3112192153930664, "learning_rate": 8.010248781040027e-05, "loss": 0.01, "step": 9890 }, { "grad_norm": 0.38309863209724426, "learning_rate": 8.005846195898814e-05, "loss": 0.0096, "step": 9900 }, { "grad_norm": 0.26620104908943176, "learning_rate": 8.00143995837104e-05, "loss": 0.0108, "step": 9910 }, { "grad_norm": 0.3309386074542999, "learning_rate": 7.997030073810698e-05, "loss": 0.0127, "step": 9920 }, { "grad_norm": 0.39696383476257324, "learning_rate": 7.992616547576217e-05, "loss": 0.0111, "step": 9930 }, { "grad_norm": 0.33653682470321655, "learning_rate": 7.988199385030445e-05, "loss": 0.0113, "step": 9940 }, { "grad_norm": 0.36454081535339355, "learning_rate": 7.98377859154065e-05, "loss": 0.0119, "step": 9950 }, { "grad_norm": 0.4529114365577698, "learning_rate": 7.979354172478516e-05, "loss": 0.0121, "step": 9960 }, { "grad_norm": 0.36055809259414673, "learning_rate": 7.974926133220127e-05, "loss": 0.0104, "step": 9970 }, { "grad_norm": 0.30062335729599, "learning_rate": 7.970494479145967e-05, "loss": 0.0095, "step": 9980 }, { "grad_norm": 0.2544906437397003, "learning_rate": 7.966059215640919e-05, "loss": 0.0178, "step": 9990 }, { "grad_norm": 0.33067888021469116, "learning_rate": 7.961620348094241e-05, "loss": 0.0124, "step": 10000 }, { "grad_norm": 0.38539981842041016, "learning_rate": 7.957177881899577e-05, "loss": 0.0144, "step": 10010 }, { "grad_norm": 0.3524967432022095, "learning_rate": 7.952731822454944e-05, "loss": 0.0121, "step": 10020 }, { "grad_norm": 0.38559991121292114, "learning_rate": 7.948282175162722e-05, "loss": 0.0108, "step": 10030 }, { "grad_norm": 0.30049654841423035, "learning_rate": 7.943828945429652e-05, "loss": 0.0102, "step": 10040 }, { "grad_norm": 0.38695329427719116, "learning_rate": 7.939372138666827e-05, "loss": 0.0126, "step": 10050 }, { "grad_norm": 0.30858898162841797, "learning_rate": 7.934911760289692e-05, "loss": 0.0103, "step": 10060 }, { "grad_norm": 0.38485032320022583, "learning_rate": 7.930447815718022e-05, "loss": 0.0152, "step": 10070 }, { "grad_norm": 0.41837865114212036, "learning_rate": 7.925980310375934e-05, "loss": 0.0107, "step": 10080 }, { "grad_norm": 0.4553411900997162, "learning_rate": 7.921509249691865e-05, "loss": 0.013, "step": 10090 }, { "grad_norm": 0.40696337819099426, "learning_rate": 7.917034639098579e-05, "loss": 0.0131, "step": 10100 }, { "grad_norm": 0.30605393648147583, "learning_rate": 7.912556484033146e-05, "loss": 0.0107, "step": 10110 }, { "grad_norm": 0.3726651966571808, "learning_rate": 7.908074789936952e-05, "loss": 0.013, "step": 10120 }, { "grad_norm": 0.3277242183685303, "learning_rate": 7.903589562255673e-05, "loss": 0.0103, "step": 10130 }, { "grad_norm": 0.3119547963142395, "learning_rate": 7.899100806439285e-05, "loss": 0.0172, "step": 10140 }, { "grad_norm": 0.4841569662094116, "learning_rate": 7.894608527942049e-05, "loss": 0.0145, "step": 10150 }, { "grad_norm": 0.4619734585285187, "learning_rate": 7.89011273222251e-05, "loss": 0.0129, "step": 10160 }, { "grad_norm": 0.2540816068649292, "learning_rate": 7.885613424743482e-05, "loss": 0.0144, "step": 10170 }, { "grad_norm": 0.3454608917236328, "learning_rate": 7.881110610972044e-05, "loss": 0.0135, "step": 10180 }, { "grad_norm": 0.27251675724983215, "learning_rate": 7.876604296379544e-05, "loss": 0.0118, "step": 10190 }, { "grad_norm": 0.36014431715011597, "learning_rate": 7.872094486441579e-05, "loss": 0.0126, "step": 10200 }, { "grad_norm": 0.3755357265472412, "learning_rate": 7.867581186637991e-05, "loss": 0.0124, "step": 10210 }, { "grad_norm": 0.35041287541389465, "learning_rate": 7.863064402452866e-05, "loss": 0.014, "step": 10220 }, { "grad_norm": 0.30046433210372925, "learning_rate": 7.858544139374524e-05, "loss": 0.0098, "step": 10230 }, { "grad_norm": 0.27960342168807983, "learning_rate": 7.854020402895508e-05, "loss": 0.0112, "step": 10240 }, { "grad_norm": 0.3508724570274353, "learning_rate": 7.849493198512587e-05, "loss": 0.0131, "step": 10250 }, { "grad_norm": 0.3428138494491577, "learning_rate": 7.84496253172674e-05, "loss": 0.0137, "step": 10260 }, { "grad_norm": 0.2897298038005829, "learning_rate": 7.840428408043155e-05, "loss": 0.0132, "step": 10270 }, { "grad_norm": 0.3261246383190155, "learning_rate": 7.835890832971218e-05, "loss": 0.0105, "step": 10280 }, { "grad_norm": 0.2967917323112488, "learning_rate": 7.831349812024512e-05, "loss": 0.0106, "step": 10290 }, { "grad_norm": 0.37736576795578003, "learning_rate": 7.826805350720807e-05, "loss": 0.0126, "step": 10300 }, { "grad_norm": 0.38696154952049255, "learning_rate": 7.822257454582049e-05, "loss": 0.0105, "step": 10310 }, { "grad_norm": 0.38143306970596313, "learning_rate": 7.817706129134363e-05, "loss": 0.0108, "step": 10320 }, { "grad_norm": 0.33168354630470276, "learning_rate": 7.813151379908036e-05, "loss": 0.0111, "step": 10330 }, { "grad_norm": 0.223832368850708, "learning_rate": 7.808593212437523e-05, "loss": 0.0095, "step": 10340 }, { "grad_norm": 0.34434786438941956, "learning_rate": 7.804031632261421e-05, "loss": 0.0099, "step": 10350 }, { "grad_norm": 0.3805299997329712, "learning_rate": 7.799466644922484e-05, "loss": 0.0145, "step": 10360 }, { "grad_norm": 0.41314366459846497, "learning_rate": 7.794898255967602e-05, "loss": 0.0117, "step": 10370 }, { "grad_norm": 0.3865651786327362, "learning_rate": 7.790326470947795e-05, "loss": 0.0123, "step": 10380 }, { "grad_norm": 0.27097195386886597, "learning_rate": 7.785751295418217e-05, "loss": 0.0097, "step": 10390 }, { "grad_norm": 0.3243426978588104, "learning_rate": 7.781172734938136e-05, "loss": 0.0105, "step": 10400 }, { "grad_norm": 0.3390544056892395, "learning_rate": 7.776590795070933e-05, "loss": 0.0091, "step": 10410 }, { "grad_norm": 0.3892500400543213, "learning_rate": 7.772005481384099e-05, "loss": 0.0096, "step": 10420 }, { "grad_norm": 0.3469524681568146, "learning_rate": 7.767416799449223e-05, "loss": 0.0103, "step": 10430 }, { "grad_norm": 0.3976801931858063, "learning_rate": 7.762824754841984e-05, "loss": 0.0125, "step": 10440 }, { "grad_norm": 0.3225308656692505, "learning_rate": 7.758229353142152e-05, "loss": 0.0094, "step": 10450 }, { "grad_norm": 0.3060131072998047, "learning_rate": 7.753630599933572e-05, "loss": 0.0102, "step": 10460 }, { "grad_norm": 0.2835567891597748, "learning_rate": 7.749028500804161e-05, "loss": 0.0106, "step": 10470 }, { "grad_norm": 0.33139705657958984, "learning_rate": 7.744423061345906e-05, "loss": 0.0133, "step": 10480 }, { "grad_norm": 0.33680808544158936, "learning_rate": 7.73981428715485e-05, "loss": 0.011, "step": 10490 }, { "grad_norm": 0.44460445642471313, "learning_rate": 7.735202183831085e-05, "loss": 0.0129, "step": 10500 }, { "grad_norm": 0.36355099081993103, "learning_rate": 7.730586756978757e-05, "loss": 0.0201, "step": 10510 }, { "grad_norm": 0.3042161464691162, "learning_rate": 7.72596801220604e-05, "loss": 0.0124, "step": 10520 }, { "grad_norm": 0.362052321434021, "learning_rate": 7.721345955125147e-05, "loss": 0.0143, "step": 10530 }, { "grad_norm": 0.2747877538204193, "learning_rate": 7.71672059135231e-05, "loss": 0.0104, "step": 10540 }, { "grad_norm": 0.34670934081077576, "learning_rate": 7.712091926507787e-05, "loss": 0.0142, "step": 10550 }, { "grad_norm": 0.3347124457359314, "learning_rate": 7.70745996621584e-05, "loss": 0.0128, "step": 10560 }, { "grad_norm": 0.3514021337032318, "learning_rate": 7.702824716104735e-05, "loss": 0.0124, "step": 10570 }, { "grad_norm": 0.23168914020061493, "learning_rate": 7.698186181806743e-05, "loss": 0.0158, "step": 10580 }, { "grad_norm": 0.24014471471309662, "learning_rate": 7.693544368958116e-05, "loss": 0.0091, "step": 10590 }, { "grad_norm": 0.4759672284126282, "learning_rate": 7.688899283199096e-05, "loss": 0.0139, "step": 10600 }, { "grad_norm": 0.3577236831188202, "learning_rate": 7.684250930173901e-05, "loss": 0.0129, "step": 10610 }, { "grad_norm": 0.35711148381233215, "learning_rate": 7.679599315530718e-05, "loss": 0.014, "step": 10620 }, { "grad_norm": 0.30543196201324463, "learning_rate": 7.674944444921695e-05, "loss": 0.0089, "step": 10630 }, { "grad_norm": 0.3276115357875824, "learning_rate": 7.670286324002944e-05, "loss": 0.0102, "step": 10640 }, { "grad_norm": 0.3749338388442993, "learning_rate": 7.665624958434514e-05, "loss": 0.0136, "step": 10650 }, { "grad_norm": 0.3328280448913574, "learning_rate": 7.66096035388041e-05, "loss": 0.0125, "step": 10660 }, { "grad_norm": 0.3168521821498871, "learning_rate": 7.656292516008563e-05, "loss": 0.0116, "step": 10670 }, { "grad_norm": 0.40176042914390564, "learning_rate": 7.651621450490837e-05, "loss": 0.0126, "step": 10680 }, { "grad_norm": 0.36332520842552185, "learning_rate": 7.646947163003017e-05, "loss": 0.0117, "step": 10690 }, { "grad_norm": 0.3413226902484894, "learning_rate": 7.642269659224803e-05, "loss": 0.0098, "step": 10700 }, { "grad_norm": 0.31223681569099426, "learning_rate": 7.637588944839803e-05, "loss": 0.0131, "step": 10710 }, { "grad_norm": 0.25687792897224426, "learning_rate": 7.632905025535528e-05, "loss": 0.0091, "step": 10720 }, { "grad_norm": 0.30029481649398804, "learning_rate": 7.628217907003378e-05, "loss": 0.0115, "step": 10730 }, { "grad_norm": 0.34949514269828796, "learning_rate": 7.623527594938649e-05, "loss": 0.0114, "step": 10740 }, { "grad_norm": 0.4217263460159302, "learning_rate": 7.618834095040509e-05, "loss": 0.0108, "step": 10750 }, { "grad_norm": 0.30035609006881714, "learning_rate": 7.614137413012e-05, "loss": 0.0094, "step": 10760 }, { "grad_norm": 0.2800735533237457, "learning_rate": 7.609437554560042e-05, "loss": 0.008, "step": 10770 }, { "grad_norm": 0.29203349351882935, "learning_rate": 7.604734525395398e-05, "loss": 0.0107, "step": 10780 }, { "grad_norm": 0.20077933371067047, "learning_rate": 7.600028331232699e-05, "loss": 0.0083, "step": 10790 }, { "grad_norm": 0.25970229506492615, "learning_rate": 7.595318977790408e-05, "loss": 0.0087, "step": 10800 }, { "grad_norm": 0.307812362909317, "learning_rate": 7.590606470790836e-05, "loss": 0.0083, "step": 10810 }, { "grad_norm": 0.32687681913375854, "learning_rate": 7.585890815960125e-05, "loss": 0.0092, "step": 10820 }, { "grad_norm": 0.29951491951942444, "learning_rate": 7.581172019028238e-05, "loss": 0.0089, "step": 10830 }, { "grad_norm": 0.26941248774528503, "learning_rate": 7.576450085728959e-05, "loss": 0.0084, "step": 10840 }, { "grad_norm": 0.32266977429389954, "learning_rate": 7.571725021799885e-05, "loss": 0.011, "step": 10850 }, { "grad_norm": 0.28205350041389465, "learning_rate": 7.566996832982409e-05, "loss": 0.0098, "step": 10860 }, { "grad_norm": 0.3159908652305603, "learning_rate": 7.56226552502173e-05, "loss": 0.0091, "step": 10870 }, { "grad_norm": 0.31130290031433105, "learning_rate": 7.557531103666833e-05, "loss": 0.0102, "step": 10880 }, { "grad_norm": 0.29061397910118103, "learning_rate": 7.552793574670485e-05, "loss": 0.012, "step": 10890 }, { "grad_norm": 0.3921622633934021, "learning_rate": 7.54805294378923e-05, "loss": 0.0088, "step": 10900 }, { "grad_norm": 0.43555957078933716, "learning_rate": 7.543309216783384e-05, "loss": 0.0117, "step": 10910 }, { "grad_norm": 0.32273146510124207, "learning_rate": 7.53856239941702e-05, "loss": 0.0089, "step": 10920 }, { "grad_norm": 0.29576656222343445, "learning_rate": 7.533812497457972e-05, "loss": 0.0084, "step": 10930 }, { "grad_norm": 0.2760366201400757, "learning_rate": 7.529059516677814e-05, "loss": 0.0077, "step": 10940 }, { "grad_norm": 0.36985543370246887, "learning_rate": 7.524303462851872e-05, "loss": 0.0131, "step": 10950 }, { "grad_norm": 0.4275459349155426, "learning_rate": 7.519544341759192e-05, "loss": 0.0125, "step": 10960 }, { "grad_norm": 0.36508968472480774, "learning_rate": 7.514782159182562e-05, "loss": 0.0138, "step": 10970 }, { "grad_norm": 0.324375718832016, "learning_rate": 7.510016920908481e-05, "loss": 0.0116, "step": 10980 }, { "grad_norm": 0.2863047420978546, "learning_rate": 7.505248632727158e-05, "loss": 0.0085, "step": 10990 }, { "grad_norm": 0.4040570557117462, "learning_rate": 7.50047730043252e-05, "loss": 0.0104, "step": 11000 }, { "grad_norm": 0.2557476758956909, "learning_rate": 7.495702929822183e-05, "loss": 0.0121, "step": 11010 }, { "grad_norm": 0.27350640296936035, "learning_rate": 7.490925526697455e-05, "loss": 0.0083, "step": 11020 }, { "grad_norm": 0.3028150796890259, "learning_rate": 7.486145096863334e-05, "loss": 0.0079, "step": 11030 }, { "grad_norm": 0.3183582127094269, "learning_rate": 7.481361646128491e-05, "loss": 0.0091, "step": 11040 }, { "grad_norm": 0.22178129851818085, "learning_rate": 7.476575180305271e-05, "loss": 0.0069, "step": 11050 }, { "grad_norm": 0.3955713212490082, "learning_rate": 7.471785705209682e-05, "loss": 0.0089, "step": 11060 }, { "grad_norm": 0.34152182936668396, "learning_rate": 7.466993226661387e-05, "loss": 0.0111, "step": 11070 }, { "grad_norm": 0.41312381625175476, "learning_rate": 7.462197750483699e-05, "loss": 0.0091, "step": 11080 }, { "grad_norm": 0.39213383197784424, "learning_rate": 7.457399282503574e-05, "loss": 0.0127, "step": 11090 }, { "grad_norm": 0.31024205684661865, "learning_rate": 7.452597828551604e-05, "loss": 0.0113, "step": 11100 }, { "grad_norm": 0.20336121320724487, "learning_rate": 7.447793394462006e-05, "loss": 0.0102, "step": 11110 }, { "grad_norm": 0.2708738446235657, "learning_rate": 7.442985986072624e-05, "loss": 0.0083, "step": 11120 }, { "grad_norm": 0.33833274245262146, "learning_rate": 7.438175609224908e-05, "loss": 0.0119, "step": 11130 }, { "grad_norm": 0.27847418189048767, "learning_rate": 7.433362269763924e-05, "loss": 0.0095, "step": 11140 }, { "grad_norm": 0.22850477695465088, "learning_rate": 7.428545973538329e-05, "loss": 0.0091, "step": 11150 }, { "grad_norm": 0.3139864504337311, "learning_rate": 7.42372672640038e-05, "loss": 0.0089, "step": 11160 }, { "grad_norm": 0.2883455455303192, "learning_rate": 7.418904534205917e-05, "loss": 0.0121, "step": 11170 }, { "grad_norm": 0.2504955232143402, "learning_rate": 7.414079402814356e-05, "loss": 0.01, "step": 11180 }, { "grad_norm": 0.2814556062221527, "learning_rate": 7.409251338088687e-05, "loss": 0.0099, "step": 11190 }, { "grad_norm": 0.3406390845775604, "learning_rate": 7.404420345895467e-05, "loss": 0.0085, "step": 11200 }, { "grad_norm": 0.2974289059638977, "learning_rate": 7.399586432104804e-05, "loss": 0.009, "step": 11210 }, { "grad_norm": 0.23555757105350494, "learning_rate": 7.394749602590358e-05, "loss": 0.0136, "step": 11220 }, { "grad_norm": 0.2215532809495926, "learning_rate": 7.389909863229336e-05, "loss": 0.0108, "step": 11230 }, { "grad_norm": 0.3040720820426941, "learning_rate": 7.385067219902477e-05, "loss": 0.0101, "step": 11240 }, { "grad_norm": 0.27627989649772644, "learning_rate": 7.380221678494049e-05, "loss": 0.0082, "step": 11250 }, { "grad_norm": 0.23108987510204315, "learning_rate": 7.37537324489184e-05, "loss": 0.0079, "step": 11260 }, { "grad_norm": 0.341304212808609, "learning_rate": 7.370521924987155e-05, "loss": 0.0105, "step": 11270 }, { "grad_norm": 0.24577492475509644, "learning_rate": 7.365667724674806e-05, "loss": 0.0082, "step": 11280 }, { "grad_norm": 0.4778881072998047, "learning_rate": 7.360810649853105e-05, "loss": 0.0104, "step": 11290 }, { "grad_norm": 0.31984004378318787, "learning_rate": 7.355950706423853e-05, "loss": 0.009, "step": 11300 }, { "grad_norm": 0.31845182180404663, "learning_rate": 7.351087900292342e-05, "loss": 0.0086, "step": 11310 }, { "grad_norm": 0.3482874631881714, "learning_rate": 7.346222237367339e-05, "loss": 0.0093, "step": 11320 }, { "grad_norm": 0.3342769145965576, "learning_rate": 7.341353723561081e-05, "loss": 0.0091, "step": 11330 }, { "grad_norm": 0.31442058086395264, "learning_rate": 7.336482364789277e-05, "loss": 0.0092, "step": 11340 }, { "grad_norm": 0.30094844102859497, "learning_rate": 7.331608166971082e-05, "loss": 0.0084, "step": 11350 }, { "grad_norm": 0.2978874444961548, "learning_rate": 7.326731136029108e-05, "loss": 0.0093, "step": 11360 }, { "grad_norm": 0.3218439817428589, "learning_rate": 7.321851277889408e-05, "loss": 0.0119, "step": 11370 }, { "grad_norm": 0.28479912877082825, "learning_rate": 7.316968598481469e-05, "loss": 0.0087, "step": 11380 }, { "grad_norm": 0.25186410546302795, "learning_rate": 7.312083103738206e-05, "loss": 0.0126, "step": 11390 }, { "grad_norm": 0.2577053904533386, "learning_rate": 7.307194799595958e-05, "loss": 0.0083, "step": 11400 }, { "grad_norm": 0.30730319023132324, "learning_rate": 7.302303691994473e-05, "loss": 0.0086, "step": 11410 }, { "grad_norm": 0.3237866461277008, "learning_rate": 7.29740978687691e-05, "loss": 0.0083, "step": 11420 }, { "grad_norm": 0.3452749252319336, "learning_rate": 7.292513090189825e-05, "loss": 0.0126, "step": 11430 }, { "grad_norm": 0.30742043256759644, "learning_rate": 7.287613607883163e-05, "loss": 0.0086, "step": 11440 }, { "grad_norm": 0.2395840436220169, "learning_rate": 7.282711345910263e-05, "loss": 0.0099, "step": 11450 }, { "grad_norm": 0.3597647249698639, "learning_rate": 7.27780631022783e-05, "loss": 0.0093, "step": 11460 }, { "grad_norm": 0.24060678482055664, "learning_rate": 7.272898506795948e-05, "loss": 0.0082, "step": 11470 }, { "grad_norm": 0.3263556659221649, "learning_rate": 7.267987941578058e-05, "loss": 0.0095, "step": 11480 }, { "grad_norm": 0.28339502215385437, "learning_rate": 7.263074620540962e-05, "loss": 0.0096, "step": 11490 }, { "grad_norm": 0.34400632977485657, "learning_rate": 7.25815854965481e-05, "loss": 0.0079, "step": 11500 }, { "grad_norm": 0.29296886920928955, "learning_rate": 7.253239734893089e-05, "loss": 0.0114, "step": 11510 }, { "grad_norm": 0.3033179044723511, "learning_rate": 7.248318182232623e-05, "loss": 0.009, "step": 11520 }, { "grad_norm": 0.283232182264328, "learning_rate": 7.243393897653565e-05, "loss": 0.0085, "step": 11530 }, { "grad_norm": 0.2822434902191162, "learning_rate": 7.23846688713938e-05, "loss": 0.0111, "step": 11540 }, { "grad_norm": 0.3397856056690216, "learning_rate": 7.233537156676854e-05, "loss": 0.0103, "step": 11550 }, { "grad_norm": 0.27789750695228577, "learning_rate": 7.228604712256076e-05, "loss": 0.01, "step": 11560 }, { "grad_norm": 0.254805326461792, "learning_rate": 7.223669559870426e-05, "loss": 0.0102, "step": 11570 }, { "grad_norm": 0.2866707146167755, "learning_rate": 7.218731705516585e-05, "loss": 0.0069, "step": 11580 }, { "grad_norm": 0.3048359751701355, "learning_rate": 7.21379115519451e-05, "loss": 0.0103, "step": 11590 }, { "grad_norm": 0.32242247462272644, "learning_rate": 7.20884791490743e-05, "loss": 0.0078, "step": 11600 }, { "grad_norm": 0.30326324701309204, "learning_rate": 7.203901990661857e-05, "loss": 0.009, "step": 11610 }, { "grad_norm": 0.43640846014022827, "learning_rate": 7.198953388467549e-05, "loss": 0.0096, "step": 11620 }, { "grad_norm": 0.3773597180843353, "learning_rate": 7.194002114337526e-05, "loss": 0.0077, "step": 11630 }, { "grad_norm": 0.2945236265659332, "learning_rate": 7.189048174288054e-05, "loss": 0.0077, "step": 11640 }, { "grad_norm": 0.23524796962738037, "learning_rate": 7.184091574338636e-05, "loss": 0.0085, "step": 11650 }, { "grad_norm": 0.2731389105319977, "learning_rate": 7.179132320512009e-05, "loss": 0.0082, "step": 11660 }, { "grad_norm": 0.32267874479293823, "learning_rate": 7.174170418834134e-05, "loss": 0.0082, "step": 11670 }, { "grad_norm": 0.25170040130615234, "learning_rate": 7.169205875334189e-05, "loss": 0.0086, "step": 11680 }, { "grad_norm": 0.29991281032562256, "learning_rate": 7.164238696044562e-05, "loss": 0.0078, "step": 11690 }, { "grad_norm": 0.21632200479507446, "learning_rate": 7.159268887000846e-05, "loss": 0.0081, "step": 11700 }, { "grad_norm": 0.27993524074554443, "learning_rate": 7.154296454241827e-05, "loss": 0.0094, "step": 11710 }, { "grad_norm": 0.3422974944114685, "learning_rate": 7.149321403809479e-05, "loss": 0.01, "step": 11720 }, { "grad_norm": 0.4339045286178589, "learning_rate": 7.144343741748956e-05, "loss": 0.0137, "step": 11730 }, { "grad_norm": 0.33812734484672546, "learning_rate": 7.139363474108589e-05, "loss": 0.0102, "step": 11740 }, { "grad_norm": 0.29606929421424866, "learning_rate": 7.134380606939874e-05, "loss": 0.0082, "step": 11750 }, { "grad_norm": 0.42082029581069946, "learning_rate": 7.12939514629746e-05, "loss": 0.0116, "step": 11760 }, { "grad_norm": 0.2852288782596588, "learning_rate": 7.124407098239155e-05, "loss": 0.0083, "step": 11770 }, { "grad_norm": 0.3096076250076294, "learning_rate": 7.119416468825908e-05, "loss": 0.008, "step": 11780 }, { "grad_norm": 0.2118469923734665, "learning_rate": 7.114423264121804e-05, "loss": 0.0094, "step": 11790 }, { "grad_norm": 0.282511442899704, "learning_rate": 7.109427490194056e-05, "loss": 0.0074, "step": 11800 }, { "grad_norm": 0.26091334223747253, "learning_rate": 7.104429153113001e-05, "loss": 0.0103, "step": 11810 }, { "grad_norm": 0.2846382260322571, "learning_rate": 7.099428258952092e-05, "loss": 0.0081, "step": 11820 }, { "grad_norm": 0.24236689507961273, "learning_rate": 7.094424813787883e-05, "loss": 0.0087, "step": 11830 }, { "grad_norm": 0.38290461897850037, "learning_rate": 7.089418823700034e-05, "loss": 0.0092, "step": 11840 }, { "grad_norm": 0.40860337018966675, "learning_rate": 7.084410294771298e-05, "loss": 0.0129, "step": 11850 }, { "grad_norm": 0.2941567897796631, "learning_rate": 7.079399233087504e-05, "loss": 0.0092, "step": 11860 }, { "grad_norm": 0.347569078207016, "learning_rate": 7.074385644737568e-05, "loss": 0.0085, "step": 11870 }, { "grad_norm": 0.2763011157512665, "learning_rate": 7.069369535813473e-05, "loss": 0.0123, "step": 11880 }, { "grad_norm": 0.32152655720710754, "learning_rate": 7.06435091241026e-05, "loss": 0.0153, "step": 11890 }, { "grad_norm": 0.35805922746658325, "learning_rate": 7.059329780626034e-05, "loss": 0.0114, "step": 11900 }, { "grad_norm": 0.44301244616508484, "learning_rate": 7.054306146561944e-05, "loss": 0.0132, "step": 11910 }, { "grad_norm": 0.5094828009605408, "learning_rate": 7.049280016322178e-05, "loss": 0.0119, "step": 11920 }, { "grad_norm": 0.29898038506507874, "learning_rate": 7.044251396013957e-05, "loss": 0.0098, "step": 11930 }, { "grad_norm": 0.2888360023498535, "learning_rate": 7.039220291747528e-05, "loss": 0.0144, "step": 11940 }, { "grad_norm": 0.45656365156173706, "learning_rate": 7.034186709636159e-05, "loss": 0.0142, "step": 11950 }, { "grad_norm": 0.3023098111152649, "learning_rate": 7.029150655796129e-05, "loss": 0.0128, "step": 11960 }, { "grad_norm": 0.28289538621902466, "learning_rate": 7.024112136346712e-05, "loss": 0.0097, "step": 11970 }, { "grad_norm": 0.37852033972740173, "learning_rate": 7.01907115741019e-05, "loss": 0.0119, "step": 11980 }, { "grad_norm": 0.38586193323135376, "learning_rate": 7.014027725111825e-05, "loss": 0.0085, "step": 11990 }, { "grad_norm": 0.3567219078540802, "learning_rate": 7.008981845579864e-05, "loss": 0.0147, "step": 12000 }, { "grad_norm": 0.24086809158325195, "learning_rate": 7.003933524945528e-05, "loss": 0.0094, "step": 12010 }, { "grad_norm": 0.362815260887146, "learning_rate": 6.998882769342998e-05, "loss": 0.0087, "step": 12020 }, { "grad_norm": 0.30917811393737793, "learning_rate": 6.993829584909423e-05, "loss": 0.0082, "step": 12030 }, { "grad_norm": 0.21194805204868317, "learning_rate": 6.988773977784895e-05, "loss": 0.008, "step": 12040 }, { "grad_norm": 0.3287794291973114, "learning_rate": 6.983715954112454e-05, "loss": 0.0115, "step": 12050 }, { "grad_norm": 0.33956584334373474, "learning_rate": 6.978655520038079e-05, "loss": 0.0075, "step": 12060 }, { "grad_norm": 0.30232930183410645, "learning_rate": 6.97359268171067e-05, "loss": 0.0096, "step": 12070 }, { "grad_norm": 0.24686980247497559, "learning_rate": 6.968527445282056e-05, "loss": 0.0101, "step": 12080 }, { "grad_norm": 0.22633866965770721, "learning_rate": 6.963459816906977e-05, "loss": 0.0089, "step": 12090 }, { "grad_norm": 0.3177076578140259, "learning_rate": 6.958389802743077e-05, "loss": 0.0107, "step": 12100 }, { "grad_norm": 0.3093923330307007, "learning_rate": 6.953317408950903e-05, "loss": 0.0092, "step": 12110 }, { "grad_norm": 0.2933741509914398, "learning_rate": 6.94824264169389e-05, "loss": 0.009, "step": 12120 }, { "grad_norm": 0.24352285265922546, "learning_rate": 6.94316550713836e-05, "loss": 0.0102, "step": 12130 }, { "grad_norm": 0.2738446891307831, "learning_rate": 6.938086011453513e-05, "loss": 0.0071, "step": 12140 }, { "grad_norm": 0.36744293570518494, "learning_rate": 6.93300416081141e-05, "loss": 0.008, "step": 12150 }, { "grad_norm": 0.3359645903110504, "learning_rate": 6.927919961386984e-05, "loss": 0.0078, "step": 12160 }, { "grad_norm": 0.2611351013183594, "learning_rate": 6.922833419358013e-05, "loss": 0.0091, "step": 12170 }, { "grad_norm": 0.2735004425048828, "learning_rate": 6.917744540905124e-05, "loss": 0.0086, "step": 12180 }, { "grad_norm": 0.3218027651309967, "learning_rate": 6.912653332211787e-05, "loss": 0.0094, "step": 12190 }, { "grad_norm": 0.2841850817203522, "learning_rate": 6.9075597994643e-05, "loss": 0.0088, "step": 12200 }, { "grad_norm": 0.3550814390182495, "learning_rate": 6.902463948851787e-05, "loss": 0.0098, "step": 12210 }, { "grad_norm": 0.3593651354312897, "learning_rate": 6.897365786566184e-05, "loss": 0.0111, "step": 12220 }, { "grad_norm": 0.2855260372161865, "learning_rate": 6.892265318802242e-05, "loss": 0.013, "step": 12230 }, { "grad_norm": 0.2511592209339142, "learning_rate": 6.887162551757507e-05, "loss": 0.0118, "step": 12240 }, { "grad_norm": 0.21981532871723175, "learning_rate": 6.882057491632326e-05, "loss": 0.0069, "step": 12250 }, { "grad_norm": 0.2968486249446869, "learning_rate": 6.876950144629824e-05, "loss": 0.0077, "step": 12260 }, { "grad_norm": 0.28601324558258057, "learning_rate": 6.87184051695591e-05, "loss": 0.0109, "step": 12270 }, { "grad_norm": 0.24612201750278473, "learning_rate": 6.866728614819268e-05, "loss": 0.009, "step": 12280 }, { "grad_norm": 0.28446415066719055, "learning_rate": 6.861614444431337e-05, "loss": 0.0114, "step": 12290 }, { "grad_norm": 0.3010433316230774, "learning_rate": 6.856498012006318e-05, "loss": 0.0107, "step": 12300 }, { "grad_norm": 0.20353767275810242, "learning_rate": 6.851379323761157e-05, "loss": 0.008, "step": 12310 }, { "grad_norm": 0.26758721470832825, "learning_rate": 6.846258385915545e-05, "loss": 0.0098, "step": 12320 }, { "grad_norm": 0.3392726182937622, "learning_rate": 6.841135204691902e-05, "loss": 0.0095, "step": 12330 }, { "grad_norm": 0.316550076007843, "learning_rate": 6.836009786315377e-05, "loss": 0.0095, "step": 12340 }, { "grad_norm": 0.3038289248943329, "learning_rate": 6.830882137013839e-05, "loss": 0.007, "step": 12350 }, { "grad_norm": 0.3878706991672516, "learning_rate": 6.825752263017863e-05, "loss": 0.0104, "step": 12360 }, { "grad_norm": 0.28244853019714355, "learning_rate": 6.820620170560731e-05, "loss": 0.0084, "step": 12370 }, { "grad_norm": 0.35005033016204834, "learning_rate": 6.815485865878418e-05, "loss": 0.0092, "step": 12380 }, { "grad_norm": 0.2636955976486206, "learning_rate": 6.810349355209588e-05, "loss": 0.008, "step": 12390 }, { "grad_norm": 0.3803461790084839, "learning_rate": 6.805210644795588e-05, "loss": 0.0093, "step": 12400 }, { "grad_norm": 0.24555176496505737, "learning_rate": 6.800069740880433e-05, "loss": 0.0127, "step": 12410 }, { "grad_norm": 0.27945923805236816, "learning_rate": 6.794926649710807e-05, "loss": 0.0078, "step": 12420 }, { "grad_norm": 0.31572386622428894, "learning_rate": 6.789781377536052e-05, "loss": 0.0074, "step": 12430 }, { "grad_norm": 0.23393285274505615, "learning_rate": 6.784633930608158e-05, "loss": 0.0076, "step": 12440 }, { "grad_norm": 0.2894907593727112, "learning_rate": 6.779484315181759e-05, "loss": 0.0103, "step": 12450 }, { "grad_norm": 0.27777159214019775, "learning_rate": 6.774332537514122e-05, "loss": 0.0089, "step": 12460 }, { "grad_norm": 0.40001124143600464, "learning_rate": 6.769178603865143e-05, "loss": 0.0103, "step": 12470 }, { "grad_norm": 0.3237916827201843, "learning_rate": 6.764022520497337e-05, "loss": 0.0112, "step": 12480 }, { "grad_norm": 0.3098049759864807, "learning_rate": 6.758864293675833e-05, "loss": 0.013, "step": 12490 }, { "grad_norm": 0.3279193341732025, "learning_rate": 6.753703929668362e-05, "loss": 0.0085, "step": 12500 }, { "grad_norm": 0.28273627161979675, "learning_rate": 6.748541434745253e-05, "loss": 0.0088, "step": 12510 }, { "grad_norm": 0.24343833327293396, "learning_rate": 6.743376815179424e-05, "loss": 0.0096, "step": 12520 }, { "grad_norm": 0.30315065383911133, "learning_rate": 6.738210077246376e-05, "loss": 0.0082, "step": 12530 }, { "grad_norm": 0.2952639162540436, "learning_rate": 6.733041227224181e-05, "loss": 0.0089, "step": 12540 }, { "grad_norm": 0.25487539172172546, "learning_rate": 6.72787027139348e-05, "loss": 0.0082, "step": 12550 }, { "grad_norm": 0.30707040429115295, "learning_rate": 6.72269721603747e-05, "loss": 0.0076, "step": 12560 }, { "grad_norm": 0.2823537290096283, "learning_rate": 6.717522067441904e-05, "loss": 0.0072, "step": 12570 }, { "grad_norm": 0.31651195883750916, "learning_rate": 6.712344831895074e-05, "loss": 0.01, "step": 12580 }, { "grad_norm": 0.23865506052970886, "learning_rate": 6.707165515687811e-05, "loss": 0.0076, "step": 12590 }, { "grad_norm": 0.31185057759284973, "learning_rate": 6.70198412511347e-05, "loss": 0.0086, "step": 12600 }, { "grad_norm": 0.28266116976737976, "learning_rate": 6.69680066646793e-05, "loss": 0.011, "step": 12610 }, { "grad_norm": 0.3373165726661682, "learning_rate": 6.691615146049583e-05, "loss": 0.0082, "step": 12620 }, { "grad_norm": 0.25965192914009094, "learning_rate": 6.686427570159324e-05, "loss": 0.0077, "step": 12630 }, { "grad_norm": 0.18935152888298035, "learning_rate": 6.681237945100548e-05, "loss": 0.0118, "step": 12640 }, { "grad_norm": 0.30266645550727844, "learning_rate": 6.676046277179139e-05, "loss": 0.0107, "step": 12650 }, { "grad_norm": 0.29515695571899414, "learning_rate": 6.670852572703462e-05, "loss": 0.0115, "step": 12660 }, { "grad_norm": 0.3234124779701233, "learning_rate": 6.665656837984359e-05, "loss": 0.008, "step": 12670 }, { "grad_norm": 0.28369423747062683, "learning_rate": 6.660459079335135e-05, "loss": 0.0088, "step": 12680 }, { "grad_norm": 0.26119452714920044, "learning_rate": 6.655259303071558e-05, "loss": 0.0094, "step": 12690 }, { "grad_norm": 0.3171786069869995, "learning_rate": 6.650057515511848e-05, "loss": 0.0135, "step": 12700 }, { "grad_norm": 0.2840951681137085, "learning_rate": 6.644853722976667e-05, "loss": 0.007, "step": 12710 }, { "grad_norm": 0.2470470368862152, "learning_rate": 6.639647931789113e-05, "loss": 0.0074, "step": 12720 }, { "grad_norm": 0.28765252232551575, "learning_rate": 6.634440148274713e-05, "loss": 0.0096, "step": 12730 }, { "grad_norm": 0.278496116399765, "learning_rate": 6.629230378761415e-05, "loss": 0.0072, "step": 12740 }, { "grad_norm": 0.3206707835197449, "learning_rate": 6.624018629579581e-05, "loss": 0.0086, "step": 12750 }, { "grad_norm": 0.2829134464263916, "learning_rate": 6.618804907061976e-05, "loss": 0.0148, "step": 12760 }, { "grad_norm": 0.24961838126182556, "learning_rate": 6.613589217543766e-05, "loss": 0.0074, "step": 12770 }, { "grad_norm": 0.30270251631736755, "learning_rate": 6.608371567362504e-05, "loss": 0.0082, "step": 12780 }, { "grad_norm": 0.2802150845527649, "learning_rate": 6.60315196285813e-05, "loss": 0.008, "step": 12790 }, { "grad_norm": 0.19882707297801971, "learning_rate": 6.597930410372954e-05, "loss": 0.0071, "step": 12800 }, { "grad_norm": 0.34931039810180664, "learning_rate": 6.592706916251653e-05, "loss": 0.0092, "step": 12810 }, { "grad_norm": 0.28804290294647217, "learning_rate": 6.587481486841267e-05, "loss": 0.0107, "step": 12820 }, { "grad_norm": 0.2726111114025116, "learning_rate": 6.582254128491184e-05, "loss": 0.0103, "step": 12830 }, { "grad_norm": 0.2878613770008087, "learning_rate": 6.577024847553139e-05, "loss": 0.007, "step": 12840 }, { "grad_norm": 0.2776271402835846, "learning_rate": 6.571793650381201e-05, "loss": 0.0084, "step": 12850 }, { "grad_norm": 0.34781646728515625, "learning_rate": 6.566560543331766e-05, "loss": 0.0081, "step": 12860 }, { "grad_norm": 0.33001643419265747, "learning_rate": 6.561325532763554e-05, "loss": 0.012, "step": 12870 }, { "grad_norm": 0.20931610465049744, "learning_rate": 6.556088625037597e-05, "loss": 0.0075, "step": 12880 }, { "grad_norm": 0.24424143135547638, "learning_rate": 6.55084982651723e-05, "loss": 0.007, "step": 12890 }, { "grad_norm": 0.32834357023239136, "learning_rate": 6.54560914356809e-05, "loss": 0.0084, "step": 12900 }, { "grad_norm": 0.30544838309288025, "learning_rate": 6.540366582558097e-05, "loss": 0.0084, "step": 12910 }, { "grad_norm": 0.24209871888160706, "learning_rate": 6.53512214985746e-05, "loss": 0.0085, "step": 12920 }, { "grad_norm": 0.29550281167030334, "learning_rate": 6.529875851838658e-05, "loss": 0.0098, "step": 12930 }, { "grad_norm": 0.29326924681663513, "learning_rate": 6.52462769487644e-05, "loss": 0.0073, "step": 12940 }, { "grad_norm": 0.3320079743862152, "learning_rate": 6.519377685347807e-05, "loss": 0.0084, "step": 12950 }, { "grad_norm": 0.3211483657360077, "learning_rate": 6.514125829632021e-05, "loss": 0.0093, "step": 12960 }, { "grad_norm": 0.2319023609161377, "learning_rate": 6.508872134110577e-05, "loss": 0.015, "step": 12970 }, { "grad_norm": 0.25545960664749146, "learning_rate": 6.503616605167213e-05, "loss": 0.0082, "step": 12980 }, { "grad_norm": 0.2906900942325592, "learning_rate": 6.498359249187893e-05, "loss": 0.0115, "step": 12990 }, { "grad_norm": 0.24324578046798706, "learning_rate": 6.493100072560799e-05, "loss": 0.0096, "step": 13000 }, { "grad_norm": 0.2783457338809967, "learning_rate": 6.487839081676327e-05, "loss": 0.0074, "step": 13010 }, { "grad_norm": 0.34545478224754333, "learning_rate": 6.482576282927076e-05, "loss": 0.0089, "step": 13020 }, { "grad_norm": 0.27582794427871704, "learning_rate": 6.477311682707845e-05, "loss": 0.0096, "step": 13030 }, { "grad_norm": 0.2538251280784607, "learning_rate": 6.472045287415616e-05, "loss": 0.0097, "step": 13040 }, { "grad_norm": 0.2904890775680542, "learning_rate": 6.466777103449558e-05, "loss": 0.0106, "step": 13050 }, { "grad_norm": 0.28921595215797424, "learning_rate": 6.461507137211012e-05, "loss": 0.009, "step": 13060 }, { "grad_norm": 0.3478390574455261, "learning_rate": 6.456235395103483e-05, "loss": 0.0094, "step": 13070 }, { "grad_norm": 0.29303067922592163, "learning_rate": 6.450961883532634e-05, "loss": 0.0081, "step": 13080 }, { "grad_norm": 0.2686502933502197, "learning_rate": 6.445686608906283e-05, "loss": 0.0099, "step": 13090 }, { "grad_norm": 0.3172362446784973, "learning_rate": 6.44040957763438e-05, "loss": 0.0089, "step": 13100 }, { "grad_norm": 0.27351462841033936, "learning_rate": 6.435130796129018e-05, "loss": 0.0073, "step": 13110 }, { "grad_norm": 0.233645960688591, "learning_rate": 6.429850270804416e-05, "loss": 0.0071, "step": 13120 }, { "grad_norm": 0.2655302584171295, "learning_rate": 6.424568008076909e-05, "loss": 0.0073, "step": 13130 }, { "grad_norm": 0.3362323045730591, "learning_rate": 6.419284014364943e-05, "loss": 0.0077, "step": 13140 }, { "grad_norm": 0.31689149141311646, "learning_rate": 6.41399829608907e-05, "loss": 0.0064, "step": 13150 }, { "grad_norm": 0.1684710830450058, "learning_rate": 6.408710859671938e-05, "loss": 0.0069, "step": 13160 }, { "grad_norm": 0.3330713212490082, "learning_rate": 6.403421711538278e-05, "loss": 0.0077, "step": 13170 }, { "grad_norm": 0.23600472509860992, "learning_rate": 6.398130858114903e-05, "loss": 0.0059, "step": 13180 }, { "grad_norm": 0.21627266705036163, "learning_rate": 6.392838305830701e-05, "loss": 0.0063, "step": 13190 }, { "grad_norm": 0.3084125816822052, "learning_rate": 6.387544061116621e-05, "loss": 0.0086, "step": 13200 }, { "grad_norm": 0.24651078879833221, "learning_rate": 6.382248130405671e-05, "loss": 0.008, "step": 13210 }, { "grad_norm": 0.28310707211494446, "learning_rate": 6.376950520132905e-05, "loss": 0.0094, "step": 13220 }, { "grad_norm": 0.4586986303329468, "learning_rate": 6.371651236735418e-05, "loss": 0.0107, "step": 13230 }, { "grad_norm": 0.33672696352005005, "learning_rate": 6.366350286652341e-05, "loss": 0.0083, "step": 13240 }, { "grad_norm": 0.30393362045288086, "learning_rate": 6.361047676324827e-05, "loss": 0.0096, "step": 13250 }, { "grad_norm": 0.2600557208061218, "learning_rate": 6.355743412196047e-05, "loss": 0.0094, "step": 13260 }, { "grad_norm": 0.2457929253578186, "learning_rate": 6.350437500711184e-05, "loss": 0.0074, "step": 13270 }, { "grad_norm": 0.24182464182376862, "learning_rate": 6.345129948317418e-05, "loss": 0.0064, "step": 13280 }, { "grad_norm": 0.3336026966571808, "learning_rate": 6.33982076146393e-05, "loss": 0.0085, "step": 13290 }, { "grad_norm": 0.24540174007415771, "learning_rate": 6.334509946601879e-05, "loss": 0.0083, "step": 13300 }, { "grad_norm": 0.3822924792766571, "learning_rate": 6.329197510184406e-05, "loss": 0.0106, "step": 13310 }, { "grad_norm": 0.3230099081993103, "learning_rate": 6.323883458666624e-05, "loss": 0.0093, "step": 13320 }, { "grad_norm": 0.2810288369655609, "learning_rate": 6.318567798505605e-05, "loss": 0.0063, "step": 13330 }, { "grad_norm": 0.30097776651382446, "learning_rate": 6.313250536160378e-05, "loss": 0.0077, "step": 13340 }, { "grad_norm": 0.3012586534023285, "learning_rate": 6.307931678091917e-05, "loss": 0.0074, "step": 13350 }, { "grad_norm": 0.2588575780391693, "learning_rate": 6.302611230763138e-05, "loss": 0.0066, "step": 13360 }, { "grad_norm": 0.2735314965248108, "learning_rate": 6.297289200638887e-05, "loss": 0.0063, "step": 13370 }, { "grad_norm": 0.21499907970428467, "learning_rate": 6.29196559418593e-05, "loss": 0.006, "step": 13380 }, { "grad_norm": 0.30908864736557007, "learning_rate": 6.286640417872951e-05, "loss": 0.009, "step": 13390 }, { "grad_norm": 0.24079449474811554, "learning_rate": 6.281313678170542e-05, "loss": 0.0056, "step": 13400 }, { "grad_norm": 0.3604385256767273, "learning_rate": 6.275985381551194e-05, "loss": 0.0109, "step": 13410 }, { "grad_norm": 0.278324693441391, "learning_rate": 6.270655534489292e-05, "loss": 0.0074, "step": 13420 }, { "grad_norm": 0.240886852145195, "learning_rate": 6.265324143461098e-05, "loss": 0.0071, "step": 13430 }, { "grad_norm": 0.28037452697753906, "learning_rate": 6.259991214944758e-05, "loss": 0.0074, "step": 13440 }, { "grad_norm": 0.30272406339645386, "learning_rate": 6.254656755420283e-05, "loss": 0.0077, "step": 13450 }, { "grad_norm": 0.2854255139827728, "learning_rate": 6.249320771369545e-05, "loss": 0.0071, "step": 13460 }, { "grad_norm": 0.34970709681510925, "learning_rate": 6.243983269276263e-05, "loss": 0.0091, "step": 13470 }, { "grad_norm": 0.24836835265159607, "learning_rate": 6.238644255626012e-05, "loss": 0.0068, "step": 13480 }, { "grad_norm": 0.26397332549095154, "learning_rate": 6.233303736906193e-05, "loss": 0.0078, "step": 13490 }, { "grad_norm": 0.20913416147232056, "learning_rate": 6.227961719606039e-05, "loss": 0.0062, "step": 13500 }, { "grad_norm": 0.2880506217479706, "learning_rate": 6.222618210216608e-05, "loss": 0.009, "step": 13510 }, { "grad_norm": 0.2685801088809967, "learning_rate": 6.217273215230767e-05, "loss": 0.0094, "step": 13520 }, { "grad_norm": 0.30441778898239136, "learning_rate": 6.211926741143188e-05, "loss": 0.0079, "step": 13530 }, { "grad_norm": 0.29919877648353577, "learning_rate": 6.20657879445034e-05, "loss": 0.0103, "step": 13540 }, { "grad_norm": 0.2957618832588196, "learning_rate": 6.201229381650485e-05, "loss": 0.0103, "step": 13550 }, { "grad_norm": 0.23793800175189972, "learning_rate": 6.195878509243661e-05, "loss": 0.0073, "step": 13560 }, { "grad_norm": 0.26748737692832947, "learning_rate": 6.190526183731685e-05, "loss": 0.0077, "step": 13570 }, { "grad_norm": 0.23259417712688446, "learning_rate": 6.185172411618138e-05, "loss": 0.007, "step": 13580 }, { "grad_norm": 0.2970590591430664, "learning_rate": 6.179817199408355e-05, "loss": 0.0087, "step": 13590 }, { "grad_norm": 0.23800526559352875, "learning_rate": 6.174460553609426e-05, "loss": 0.0071, "step": 13600 }, { "grad_norm": 0.3064543902873993, "learning_rate": 6.16910248073018e-05, "loss": 0.0081, "step": 13610 }, { "grad_norm": 0.28656941652297974, "learning_rate": 6.16374298728118e-05, "loss": 0.0075, "step": 13620 }, { "grad_norm": 0.23932425677776337, "learning_rate": 6.158382079774716e-05, "loss": 0.0117, "step": 13630 }, { "grad_norm": 0.30223730206489563, "learning_rate": 6.153019764724799e-05, "loss": 0.01, "step": 13640 }, { "grad_norm": 0.3376515805721283, "learning_rate": 6.147656048647144e-05, "loss": 0.0112, "step": 13650 }, { "grad_norm": 0.28413742780685425, "learning_rate": 6.142290938059173e-05, "loss": 0.0115, "step": 13660 }, { "grad_norm": 0.2288467139005661, "learning_rate": 6.136924439480001e-05, "loss": 0.0076, "step": 13670 }, { "grad_norm": 0.3216090798377991, "learning_rate": 6.13155655943043e-05, "loss": 0.0084, "step": 13680 }, { "grad_norm": 0.2861740291118622, "learning_rate": 6.126187304432941e-05, "loss": 0.0088, "step": 13690 }, { "grad_norm": 0.2751719355583191, "learning_rate": 6.120816681011684e-05, "loss": 0.0073, "step": 13700 }, { "grad_norm": 0.2503930628299713, "learning_rate": 6.115444695692474e-05, "loss": 0.0061, "step": 13710 }, { "grad_norm": 0.32750606536865234, "learning_rate": 6.110071355002779e-05, "loss": 0.0079, "step": 13720 }, { "grad_norm": 0.2416163980960846, "learning_rate": 6.104696665471714e-05, "loss": 0.0085, "step": 13730 }, { "grad_norm": 0.2940443754196167, "learning_rate": 6.099320633630036e-05, "loss": 0.0091, "step": 13740 }, { "grad_norm": 0.29082244634628296, "learning_rate": 6.093943266010128e-05, "loss": 0.0074, "step": 13750 }, { "grad_norm": 0.2597466707229614, "learning_rate": 6.088564569146e-05, "loss": 0.0068, "step": 13760 }, { "grad_norm": 0.2474258691072464, "learning_rate": 6.083184549573274e-05, "loss": 0.0074, "step": 13770 }, { "grad_norm": 0.2937285304069519, "learning_rate": 6.077803213829184e-05, "loss": 0.0087, "step": 13780 }, { "grad_norm": 0.33072125911712646, "learning_rate": 6.0724205684525594e-05, "loss": 0.0072, "step": 13790 }, { "grad_norm": 0.20828744769096375, "learning_rate": 6.0670366199838215e-05, "loss": 0.0058, "step": 13800 }, { "grad_norm": 0.24300971627235413, "learning_rate": 6.061651374964974e-05, "loss": 0.0069, "step": 13810 }, { "grad_norm": 0.2769477963447571, "learning_rate": 6.0562648399396004e-05, "loss": 0.007, "step": 13820 }, { "grad_norm": 0.25297120213508606, "learning_rate": 6.050877021452845e-05, "loss": 0.0076, "step": 13830 }, { "grad_norm": 0.30732452869415283, "learning_rate": 6.04548792605142e-05, "loss": 0.0096, "step": 13840 }, { "grad_norm": 0.22222889959812164, "learning_rate": 6.04009756028358e-05, "loss": 0.009, "step": 13850 }, { "grad_norm": 0.30012136697769165, "learning_rate": 6.0347059306991294e-05, "loss": 0.0083, "step": 13860 }, { "grad_norm": 0.34804004430770874, "learning_rate": 6.029313043849407e-05, "loss": 0.0116, "step": 13870 }, { "grad_norm": 0.25725889205932617, "learning_rate": 6.023918906287279e-05, "loss": 0.009, "step": 13880 }, { "grad_norm": 0.20896264910697937, "learning_rate": 6.01852352456713e-05, "loss": 0.0088, "step": 13890 }, { "grad_norm": 0.279345840215683, "learning_rate": 6.013126905244858e-05, "loss": 0.0077, "step": 13900 }, { "grad_norm": 0.204055517911911, "learning_rate": 6.007729054877864e-05, "loss": 0.0081, "step": 13910 }, { "grad_norm": 0.38018664717674255, "learning_rate": 6.0023299800250466e-05, "loss": 0.0081, "step": 13920 }, { "grad_norm": 0.2822417914867401, "learning_rate": 5.9969296872467894e-05, "loss": 0.0089, "step": 13930 }, { "grad_norm": 0.24374572932720184, "learning_rate": 5.9915281831049585e-05, "loss": 0.005, "step": 13940 }, { "grad_norm": 0.1957499384880066, "learning_rate": 5.98612547416289e-05, "loss": 0.0058, "step": 13950 }, { "grad_norm": 0.2626834213733673, "learning_rate": 5.9807215669853855e-05, "loss": 0.0089, "step": 13960 }, { "grad_norm": 0.4027170240879059, "learning_rate": 5.9753164681387e-05, "loss": 0.0079, "step": 13970 }, { "grad_norm": 0.27507102489471436, "learning_rate": 5.969910184190539e-05, "loss": 0.0076, "step": 13980 }, { "grad_norm": 0.26521947979927063, "learning_rate": 5.9645027217100477e-05, "loss": 0.0087, "step": 13990 }, { "grad_norm": 0.3366831839084625, "learning_rate": 5.959094087267804e-05, "loss": 0.0114, "step": 14000 }, { "grad_norm": 0.30101752281188965, "learning_rate": 5.9536842874358065e-05, "loss": 0.0071, "step": 14010 }, { "grad_norm": 0.21811336278915405, "learning_rate": 5.9482733287874734e-05, "loss": 0.0074, "step": 14020 }, { "grad_norm": 0.2502201199531555, "learning_rate": 5.942861217897631e-05, "loss": 0.0068, "step": 14030 }, { "grad_norm": 0.2900085747241974, "learning_rate": 5.9374479613425e-05, "loss": 0.0059, "step": 14040 }, { "grad_norm": 0.2607484459877014, "learning_rate": 5.932033565699704e-05, "loss": 0.0074, "step": 14050 }, { "grad_norm": 0.30430784821510315, "learning_rate": 5.926618037548237e-05, "loss": 0.0071, "step": 14060 }, { "grad_norm": 0.25428256392478943, "learning_rate": 5.9212013834684824e-05, "loss": 0.0058, "step": 14070 }, { "grad_norm": 0.2547631859779358, "learning_rate": 5.9157836100421824e-05, "loss": 0.0102, "step": 14080 }, { "grad_norm": 0.21145999431610107, "learning_rate": 5.910364723852444e-05, "loss": 0.0071, "step": 14090 }, { "grad_norm": 0.2810400724411011, "learning_rate": 5.904944731483723e-05, "loss": 0.0077, "step": 14100 }, { "grad_norm": 0.26572683453559875, "learning_rate": 5.899523639521825e-05, "loss": 0.0086, "step": 14110 }, { "grad_norm": 0.2648226320743561, "learning_rate": 5.894101454553883e-05, "loss": 0.0116, "step": 14120 }, { "grad_norm": 0.2946101129055023, "learning_rate": 5.888678183168368e-05, "loss": 0.0095, "step": 14130 }, { "grad_norm": 0.2995717525482178, "learning_rate": 5.883253831955061e-05, "loss": 0.0095, "step": 14140 }, { "grad_norm": 0.45960453152656555, "learning_rate": 5.8778284075050625e-05, "loss": 0.011, "step": 14150 }, { "grad_norm": 0.4380562901496887, "learning_rate": 5.872401916410777e-05, "loss": 0.0128, "step": 14160 }, { "grad_norm": 0.26367753744125366, "learning_rate": 5.8669743652659014e-05, "loss": 0.0079, "step": 14170 }, { "grad_norm": 0.3691392242908478, "learning_rate": 5.86154576066542e-05, "loss": 0.0121, "step": 14180 }, { "grad_norm": 0.24430038034915924, "learning_rate": 5.856116109205602e-05, "loss": 0.0082, "step": 14190 }, { "grad_norm": 0.2859657406806946, "learning_rate": 5.850685417483983e-05, "loss": 0.0073, "step": 14200 }, { "grad_norm": 0.22655045986175537, "learning_rate": 5.8452536920993684e-05, "loss": 0.0077, "step": 14210 }, { "grad_norm": 0.28834012150764465, "learning_rate": 5.839820939651817e-05, "loss": 0.0099, "step": 14220 }, { "grad_norm": 0.2714940309524536, "learning_rate": 5.834387166742632e-05, "loss": 0.0074, "step": 14230 }, { "grad_norm": 0.2178003340959549, "learning_rate": 5.828952379974364e-05, "loss": 0.0058, "step": 14240 }, { "grad_norm": 0.18576142191886902, "learning_rate": 5.8235165859507864e-05, "loss": 0.0075, "step": 14250 }, { "grad_norm": 0.30261069536209106, "learning_rate": 5.818079791276907e-05, "loss": 0.0075, "step": 14260 }, { "grad_norm": 0.248159721493721, "learning_rate": 5.8126420025589415e-05, "loss": 0.0094, "step": 14270 }, { "grad_norm": 0.21652115881443024, "learning_rate": 5.807203226404313e-05, "loss": 0.0094, "step": 14280 }, { "grad_norm": 0.21458818018436432, "learning_rate": 5.801763469421652e-05, "loss": 0.0064, "step": 14290 }, { "grad_norm": 0.33186450600624084, "learning_rate": 5.796322738220774e-05, "loss": 0.0074, "step": 14300 }, { "grad_norm": 0.27560651302337646, "learning_rate": 5.79088103941268e-05, "loss": 0.0075, "step": 14310 }, { "grad_norm": 0.23280736804008484, "learning_rate": 5.785438379609549e-05, "loss": 0.0086, "step": 14320 }, { "grad_norm": 0.20840275287628174, "learning_rate": 5.779994765424724e-05, "loss": 0.0087, "step": 14330 }, { "grad_norm": 0.35814759135246277, "learning_rate": 5.7745502034727114e-05, "loss": 0.0098, "step": 14340 }, { "grad_norm": 0.3019750118255615, "learning_rate": 5.769104700369164e-05, "loss": 0.0087, "step": 14350 }, { "grad_norm": 0.2558797299861908, "learning_rate": 5.7636582627308854e-05, "loss": 0.0098, "step": 14360 }, { "grad_norm": 0.17432241141796112, "learning_rate": 5.7582108971758095e-05, "loss": 0.0077, "step": 14370 }, { "grad_norm": 0.2660498023033142, "learning_rate": 5.752762610322999e-05, "loss": 0.0078, "step": 14380 }, { "grad_norm": 0.2865384519100189, "learning_rate": 5.747313408792636e-05, "loss": 0.0137, "step": 14390 }, { "grad_norm": 0.23268406093120575, "learning_rate": 5.741863299206014e-05, "loss": 0.0093, "step": 14400 }, { "grad_norm": 0.32585275173187256, "learning_rate": 5.73641228818553e-05, "loss": 0.012, "step": 14410 }, { "grad_norm": 0.19283880293369293, "learning_rate": 5.730960382354676e-05, "loss": 0.0062, "step": 14420 }, { "grad_norm": 0.6283600330352783, "learning_rate": 5.725507588338035e-05, "loss": 0.0289, "step": 14430 }, { "grad_norm": 3.705643653869629, "learning_rate": 5.7200539127612604e-05, "loss": 0.0252, "step": 14440 }, { "grad_norm": 0.3745173215866089, "learning_rate": 5.714599362251087e-05, "loss": 0.0215, "step": 14450 }, { "grad_norm": 0.3757733404636383, "learning_rate": 5.709143943435307e-05, "loss": 0.0141, "step": 14460 }, { "grad_norm": 0.28835970163345337, "learning_rate": 5.7036876629427646e-05, "loss": 0.0105, "step": 14470 }, { "grad_norm": 0.26223933696746826, "learning_rate": 5.698230527403361e-05, "loss": 0.0116, "step": 14480 }, { "grad_norm": 0.23664352297782898, "learning_rate": 5.6927725434480274e-05, "loss": 0.01, "step": 14490 }, { "grad_norm": 0.32759523391723633, "learning_rate": 5.687313717708728e-05, "loss": 0.0096, "step": 14500 }, { "grad_norm": 0.24322587251663208, "learning_rate": 5.681854056818453e-05, "loss": 0.0084, "step": 14510 }, { "grad_norm": 0.24972772598266602, "learning_rate": 5.676393567411205e-05, "loss": 0.0095, "step": 14520 }, { "grad_norm": 0.28814369440078735, "learning_rate": 5.670932256121991e-05, "loss": 0.0096, "step": 14530 }, { "grad_norm": 0.286001980304718, "learning_rate": 5.6654701295868216e-05, "loss": 0.0082, "step": 14540 }, { "grad_norm": 0.2960240840911865, "learning_rate": 5.660007194442697e-05, "loss": 0.0117, "step": 14550 }, { "grad_norm": 0.2852255403995514, "learning_rate": 5.6545434573275946e-05, "loss": 0.0073, "step": 14560 }, { "grad_norm": 0.25397762656211853, "learning_rate": 5.649078924880472e-05, "loss": 0.0104, "step": 14570 }, { "grad_norm": 0.2655363976955414, "learning_rate": 5.643613603741252e-05, "loss": 0.0063, "step": 14580 }, { "grad_norm": 0.22415034472942352, "learning_rate": 5.6381475005508154e-05, "loss": 0.0095, "step": 14590 }, { "grad_norm": 0.26667696237564087, "learning_rate": 5.632680621950992e-05, "loss": 0.0087, "step": 14600 }, { "grad_norm": 0.2740703225135803, "learning_rate": 5.627212974584555e-05, "loss": 0.0072, "step": 14610 }, { "grad_norm": 0.22848278284072876, "learning_rate": 5.62174456509521e-05, "loss": 0.007, "step": 14620 }, { "grad_norm": 0.251438707113266, "learning_rate": 5.616275400127594e-05, "loss": 0.0075, "step": 14630 }, { "grad_norm": 0.2876316010951996, "learning_rate": 5.610805486327254e-05, "loss": 0.0073, "step": 14640 }, { "grad_norm": 0.3225939869880676, "learning_rate": 5.6053348303406536e-05, "loss": 0.0071, "step": 14650 }, { "grad_norm": 0.2733113765716553, "learning_rate": 5.599863438815156e-05, "loss": 0.011, "step": 14660 }, { "grad_norm": 0.28402066230773926, "learning_rate": 5.594391318399017e-05, "loss": 0.0082, "step": 14670 }, { "grad_norm": 0.3579670786857605, "learning_rate": 5.588918475741377e-05, "loss": 0.0075, "step": 14680 }, { "grad_norm": 0.32270121574401855, "learning_rate": 5.5834449174922586e-05, "loss": 0.0064, "step": 14690 }, { "grad_norm": 0.2798893451690674, "learning_rate": 5.57797065030255e-05, "loss": 0.0074, "step": 14700 }, { "grad_norm": 0.19676008820533752, "learning_rate": 5.5724956808240016e-05, "loss": 0.0072, "step": 14710 }, { "grad_norm": 0.23480835556983948, "learning_rate": 5.567020015709219e-05, "loss": 0.0095, "step": 14720 }, { "grad_norm": 0.36148419976234436, "learning_rate": 5.561543661611649e-05, "loss": 0.0109, "step": 14730 }, { "grad_norm": 0.3079904019832611, "learning_rate": 5.556066625185583e-05, "loss": 0.007, "step": 14740 }, { "grad_norm": 0.18189512193202972, "learning_rate": 5.550588913086131e-05, "loss": 0.0091, "step": 14750 }, { "grad_norm": 0.28694841265678406, "learning_rate": 5.545110531969234e-05, "loss": 0.0078, "step": 14760 }, { "grad_norm": 0.24311190843582153, "learning_rate": 5.539631488491641e-05, "loss": 0.0074, "step": 14770 }, { "grad_norm": 0.23684364557266235, "learning_rate": 5.534151789310904e-05, "loss": 0.0087, "step": 14780 }, { "grad_norm": 0.21150478720664978, "learning_rate": 5.528671441085376e-05, "loss": 0.0074, "step": 14790 }, { "grad_norm": 0.2600259482860565, "learning_rate": 5.523190450474197e-05, "loss": 0.0077, "step": 14800 }, { "grad_norm": 0.2417551726102829, "learning_rate": 5.5177088241372865e-05, "loss": 0.0065, "step": 14810 }, { "grad_norm": 0.3067433834075928, "learning_rate": 5.512226568735338e-05, "loss": 0.0073, "step": 14820 }, { "grad_norm": 0.24862614274024963, "learning_rate": 5.506743690929809e-05, "loss": 0.0081, "step": 14830 }, { "grad_norm": 0.24575498700141907, "learning_rate": 5.501260197382912e-05, "loss": 0.0075, "step": 14840 }, { "grad_norm": 0.19753462076187134, "learning_rate": 5.49577609475761e-05, "loss": 0.0079, "step": 14850 }, { "grad_norm": 0.2014467865228653, "learning_rate": 5.490291389717603e-05, "loss": 0.006, "step": 14860 }, { "grad_norm": 0.31658995151519775, "learning_rate": 5.484806088927329e-05, "loss": 0.0081, "step": 14870 }, { "grad_norm": 0.29588502645492554, "learning_rate": 5.479320199051942e-05, "loss": 0.0089, "step": 14880 }, { "grad_norm": 0.19209228456020355, "learning_rate": 5.473833726757314e-05, "loss": 0.007, "step": 14890 }, { "grad_norm": 0.2832617461681366, "learning_rate": 5.4683466787100304e-05, "loss": 0.0055, "step": 14900 }, { "grad_norm": 0.25133511424064636, "learning_rate": 5.462859061577369e-05, "loss": 0.0067, "step": 14910 }, { "grad_norm": 0.21724727749824524, "learning_rate": 5.4573708820273026e-05, "loss": 0.0061, "step": 14920 }, { "grad_norm": 0.2744887173175812, "learning_rate": 5.451882146728488e-05, "loss": 0.0062, "step": 14930 }, { "grad_norm": 0.23522581160068512, "learning_rate": 5.446392862350255e-05, "loss": 0.006, "step": 14940 }, { "grad_norm": 0.19483914971351624, "learning_rate": 5.440903035562603e-05, "loss": 0.0069, "step": 14950 }, { "grad_norm": 0.2794587314128876, "learning_rate": 5.435412673036188e-05, "loss": 0.0076, "step": 14960 }, { "grad_norm": 0.31146112084388733, "learning_rate": 5.429921781442318e-05, "loss": 0.0064, "step": 14970 }, { "grad_norm": 0.25068211555480957, "learning_rate": 5.424430367452945e-05, "loss": 0.0061, "step": 14980 }, { "grad_norm": 0.2248210459947586, "learning_rate": 5.418938437740655e-05, "loss": 0.0064, "step": 14990 }, { "grad_norm": 0.24893495440483093, "learning_rate": 5.413445998978658e-05, "loss": 0.0061, "step": 15000 } ], "logging_steps": 10, "max_steps": 30000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 5000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 30, "trial_name": null, "trial_params": null }