{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.8, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 2.1501004695892334, "learning_rate": 3.6e-07, "loss": 1.6156, "step": 10 }, { "grad_norm": 2.037454843521118, "learning_rate": 7.6e-07, "loss": 1.6199, "step": 20 }, { "grad_norm": 1.6998989582061768, "learning_rate": 1.16e-06, "loss": 1.6185, "step": 30 }, { "grad_norm": 1.802830696105957, "learning_rate": 1.56e-06, "loss": 1.5984, "step": 40 }, { "grad_norm": 2.088773012161255, "learning_rate": 1.96e-06, "loss": 1.5726, "step": 50 }, { "grad_norm": 1.5814186334609985, "learning_rate": 2.36e-06, "loss": 1.5443, "step": 60 }, { "grad_norm": 1.2159417867660522, "learning_rate": 2.7600000000000003e-06, "loss": 1.4961, "step": 70 }, { "grad_norm": 0.8089196085929871, "learning_rate": 3.1600000000000007e-06, "loss": 1.4597, "step": 80 }, { "grad_norm": 0.7398391366004944, "learning_rate": 3.5600000000000002e-06, "loss": 1.3899, "step": 90 }, { "grad_norm": 0.9069815278053284, "learning_rate": 3.96e-06, "loss": 1.3905, "step": 100 }, { "grad_norm": 0.8984573483467102, "learning_rate": 4.360000000000001e-06, "loss": 1.2886, "step": 110 }, { "grad_norm": 1.1786785125732422, "learning_rate": 4.76e-06, "loss": 1.2248, "step": 120 }, { "grad_norm": 1.7223951816558838, "learning_rate": 5.1600000000000006e-06, "loss": 1.1632, "step": 130 }, { "grad_norm": 1.024466633796692, "learning_rate": 5.56e-06, "loss": 1.1409, "step": 140 }, { "grad_norm": 1.5982815027236938, "learning_rate": 5.9600000000000005e-06, "loss": 1.1076, "step": 150 }, { "grad_norm": 0.9150848388671875, "learning_rate": 6.360000000000001e-06, "loss": 1.0961, "step": 160 }, { "grad_norm": 0.9500464200973511, "learning_rate": 6.76e-06, "loss": 1.0748, "step": 170 }, { "grad_norm": 1.1501988172531128, "learning_rate": 7.16e-06, "loss": 1.0714, "step": 180 }, { "grad_norm": 0.7099096179008484, "learning_rate": 7.5600000000000005e-06, "loss": 1.0582, "step": 190 }, { "grad_norm": 1.0111870765686035, "learning_rate": 7.96e-06, "loss": 1.0545, "step": 200 }, { "grad_norm": 0.9172439575195312, "learning_rate": 8.36e-06, "loss": 1.0538, "step": 210 }, { "grad_norm": 0.7410154342651367, "learning_rate": 8.76e-06, "loss": 1.0622, "step": 220 }, { "grad_norm": 0.8835603594779968, "learning_rate": 9.16e-06, "loss": 1.0477, "step": 230 }, { "grad_norm": 0.8534172773361206, "learning_rate": 9.560000000000002e-06, "loss": 1.0443, "step": 240 }, { "grad_norm": 0.5909944772720337, "learning_rate": 9.96e-06, "loss": 1.0373, "step": 250 }, { "grad_norm": 0.9745896458625793, "learning_rate": 1.036e-05, "loss": 1.0435, "step": 260 }, { "grad_norm": 1.1140064001083374, "learning_rate": 1.076e-05, "loss": 1.0373, "step": 270 }, { "grad_norm": 0.6682170629501343, "learning_rate": 1.1160000000000002e-05, "loss": 1.05, "step": 280 }, { "grad_norm": 0.5360814332962036, "learning_rate": 1.156e-05, "loss": 1.0335, "step": 290 }, { "grad_norm": 0.6506698727607727, "learning_rate": 1.196e-05, "loss": 1.0332, "step": 300 }, { "grad_norm": 0.6195168495178223, "learning_rate": 1.236e-05, "loss": 1.0298, "step": 310 }, { "grad_norm": 0.9853718876838684, "learning_rate": 1.276e-05, "loss": 1.0361, "step": 320 }, { "grad_norm": 0.5408571362495422, "learning_rate": 1.316e-05, "loss": 1.0241, "step": 330 }, { "grad_norm": 0.5152010321617126, "learning_rate": 1.356e-05, "loss": 1.0181, "step": 340 }, { "grad_norm": 0.7603380680084229, "learning_rate": 1.396e-05, "loss": 1.0187, "step": 350 }, { "grad_norm": 1.3632287979125977, "learning_rate": 1.4360000000000001e-05, "loss": 0.9931, "step": 360 }, { "grad_norm": 1.0293426513671875, "learning_rate": 1.4760000000000001e-05, "loss": 0.9713, "step": 370 }, { "grad_norm": 0.861702024936676, "learning_rate": 1.5160000000000002e-05, "loss": 0.9583, "step": 380 }, { "grad_norm": 1.3634897470474243, "learning_rate": 1.556e-05, "loss": 0.9423, "step": 390 }, { "grad_norm": 1.1756104230880737, "learning_rate": 1.596e-05, "loss": 0.9257, "step": 400 }, { "grad_norm": 0.8505347967147827, "learning_rate": 1.636e-05, "loss": 0.9055, "step": 410 }, { "grad_norm": 0.7484205961227417, "learning_rate": 1.6760000000000002e-05, "loss": 0.8776, "step": 420 }, { "grad_norm": 1.0077836513519287, "learning_rate": 1.7160000000000002e-05, "loss": 0.8667, "step": 430 }, { "grad_norm": 1.268897533416748, "learning_rate": 1.756e-05, "loss": 0.829, "step": 440 }, { "grad_norm": 0.9711230397224426, "learning_rate": 1.796e-05, "loss": 0.8001, "step": 450 }, { "grad_norm": 1.1854374408721924, "learning_rate": 1.8360000000000004e-05, "loss": 0.7524, "step": 460 }, { "grad_norm": 1.108564019203186, "learning_rate": 1.876e-05, "loss": 0.7289, "step": 470 }, { "grad_norm": 1.3794487714767456, "learning_rate": 1.916e-05, "loss": 0.6997, "step": 480 }, { "grad_norm": 1.4291104078292847, "learning_rate": 1.956e-05, "loss": 0.6628, "step": 490 }, { "grad_norm": 1.6365247964859009, "learning_rate": 1.9960000000000002e-05, "loss": 0.6305, "step": 500 }, { "grad_norm": 1.6701785326004028, "learning_rate": 2.036e-05, "loss": 0.5791, "step": 510 }, { "grad_norm": 1.725713849067688, "learning_rate": 2.076e-05, "loss": 0.5468, "step": 520 }, { "grad_norm": 1.6206556558609009, "learning_rate": 2.116e-05, "loss": 0.5068, "step": 530 }, { "grad_norm": 1.4598190784454346, "learning_rate": 2.1560000000000004e-05, "loss": 0.4818, "step": 540 }, { "grad_norm": 1.7376526594161987, "learning_rate": 2.196e-05, "loss": 0.4407, "step": 550 }, { "grad_norm": 1.5926108360290527, "learning_rate": 2.236e-05, "loss": 0.4108, "step": 560 }, { "grad_norm": 2.2575418949127197, "learning_rate": 2.2760000000000002e-05, "loss": 0.379, "step": 570 }, { "grad_norm": 2.1090729236602783, "learning_rate": 2.3160000000000002e-05, "loss": 0.3477, "step": 580 }, { "grad_norm": 2.2720255851745605, "learning_rate": 2.356e-05, "loss": 0.3161, "step": 590 }, { "grad_norm": 1.775956153869629, "learning_rate": 2.396e-05, "loss": 0.311, "step": 600 }, { "grad_norm": 1.905683159828186, "learning_rate": 2.4360000000000004e-05, "loss": 0.2801, "step": 610 }, { "grad_norm": 1.627582311630249, "learning_rate": 2.476e-05, "loss": 0.28, "step": 620 }, { "grad_norm": 2.1697802543640137, "learning_rate": 2.516e-05, "loss": 0.2608, "step": 630 }, { "grad_norm": 1.66875422000885, "learning_rate": 2.556e-05, "loss": 0.2439, "step": 640 }, { "grad_norm": 1.894242763519287, "learning_rate": 2.5960000000000002e-05, "loss": 0.2259, "step": 650 }, { "grad_norm": 1.5882703065872192, "learning_rate": 2.6360000000000002e-05, "loss": 0.2082, "step": 660 }, { "grad_norm": 1.8057795763015747, "learning_rate": 2.676e-05, "loss": 0.1834, "step": 670 }, { "grad_norm": 2.050615072250366, "learning_rate": 2.716e-05, "loss": 0.1605, "step": 680 }, { "grad_norm": 2.1150660514831543, "learning_rate": 2.7560000000000004e-05, "loss": 0.1529, "step": 690 }, { "grad_norm": 1.8872995376586914, "learning_rate": 2.7960000000000003e-05, "loss": 0.1404, "step": 700 }, { "grad_norm": 1.9293936491012573, "learning_rate": 2.8360000000000003e-05, "loss": 0.155, "step": 710 }, { "grad_norm": 1.442170262336731, "learning_rate": 2.8760000000000002e-05, "loss": 0.1362, "step": 720 }, { "grad_norm": 1.78200101852417, "learning_rate": 2.9160000000000005e-05, "loss": 0.1194, "step": 730 }, { "grad_norm": 1.5347511768341064, "learning_rate": 2.9559999999999998e-05, "loss": 0.1296, "step": 740 }, { "grad_norm": 2.017202377319336, "learning_rate": 2.9959999999999998e-05, "loss": 0.1298, "step": 750 }, { "grad_norm": 1.8874598741531372, "learning_rate": 3.036e-05, "loss": 0.1306, "step": 760 }, { "grad_norm": 1.5480735301971436, "learning_rate": 3.076e-05, "loss": 0.1241, "step": 770 }, { "grad_norm": 1.3693205118179321, "learning_rate": 3.116e-05, "loss": 0.1223, "step": 780 }, { "grad_norm": 1.288280963897705, "learning_rate": 3.156e-05, "loss": 0.1213, "step": 790 }, { "grad_norm": 1.49338960647583, "learning_rate": 3.196e-05, "loss": 0.1172, "step": 800 }, { "grad_norm": 1.6602270603179932, "learning_rate": 3.236e-05, "loss": 0.115, "step": 810 }, { "grad_norm": 1.9156038761138916, "learning_rate": 3.2760000000000005e-05, "loss": 0.1179, "step": 820 }, { "grad_norm": 1.6822177171707153, "learning_rate": 3.316e-05, "loss": 0.1149, "step": 830 }, { "grad_norm": 1.0263917446136475, "learning_rate": 3.3560000000000004e-05, "loss": 0.1041, "step": 840 }, { "grad_norm": 1.9228767156600952, "learning_rate": 3.396e-05, "loss": 0.1101, "step": 850 }, { "grad_norm": 1.6136722564697266, "learning_rate": 3.436e-05, "loss": 0.1168, "step": 860 }, { "grad_norm": 1.2443492412567139, "learning_rate": 3.4760000000000006e-05, "loss": 0.1035, "step": 870 }, { "grad_norm": 1.2465876340866089, "learning_rate": 3.516e-05, "loss": 0.1137, "step": 880 }, { "grad_norm": 1.4049838781356812, "learning_rate": 3.5560000000000005e-05, "loss": 0.1005, "step": 890 }, { "grad_norm": 1.3370460271835327, "learning_rate": 3.596e-05, "loss": 0.0945, "step": 900 }, { "grad_norm": 1.1820286512374878, "learning_rate": 3.636e-05, "loss": 0.0967, "step": 910 }, { "grad_norm": 1.5359930992126465, "learning_rate": 3.676e-05, "loss": 0.0925, "step": 920 }, { "grad_norm": 1.699334740638733, "learning_rate": 3.716e-05, "loss": 0.0953, "step": 930 }, { "grad_norm": 1.7425698041915894, "learning_rate": 3.756e-05, "loss": 0.0943, "step": 940 }, { "grad_norm": 1.3553595542907715, "learning_rate": 3.796e-05, "loss": 0.1014, "step": 950 }, { "grad_norm": 1.3133673667907715, "learning_rate": 3.836e-05, "loss": 0.0953, "step": 960 }, { "grad_norm": 1.4967938661575317, "learning_rate": 3.876e-05, "loss": 0.098, "step": 970 }, { "grad_norm": 1.3395284414291382, "learning_rate": 3.9160000000000005e-05, "loss": 0.0987, "step": 980 }, { "grad_norm": 1.2718907594680786, "learning_rate": 3.956e-05, "loss": 0.0956, "step": 990 }, { "grad_norm": 1.6551650762557983, "learning_rate": 3.9960000000000004e-05, "loss": 0.0987, "step": 1000 }, { "grad_norm": 1.275582194328308, "learning_rate": 4.0360000000000007e-05, "loss": 0.0904, "step": 1010 }, { "grad_norm": 1.1563726663589478, "learning_rate": 4.076e-05, "loss": 0.0928, "step": 1020 }, { "grad_norm": 1.2716867923736572, "learning_rate": 4.1160000000000006e-05, "loss": 0.0973, "step": 1030 }, { "grad_norm": 1.5117614269256592, "learning_rate": 4.156e-05, "loss": 0.0935, "step": 1040 }, { "grad_norm": 0.9536880850791931, "learning_rate": 4.196e-05, "loss": 0.0885, "step": 1050 }, { "grad_norm": 1.4288803339004517, "learning_rate": 4.236e-05, "loss": 0.0938, "step": 1060 }, { "grad_norm": 1.193376064300537, "learning_rate": 4.276e-05, "loss": 0.0901, "step": 1070 }, { "grad_norm": 1.1308783292770386, "learning_rate": 4.316e-05, "loss": 0.0846, "step": 1080 }, { "grad_norm": 1.2259303331375122, "learning_rate": 4.356e-05, "loss": 0.0924, "step": 1090 }, { "grad_norm": 1.1975528001785278, "learning_rate": 4.396e-05, "loss": 0.093, "step": 1100 }, { "grad_norm": 1.0837070941925049, "learning_rate": 4.436e-05, "loss": 0.0818, "step": 1110 }, { "grad_norm": 1.032973051071167, "learning_rate": 4.4760000000000005e-05, "loss": 0.089, "step": 1120 }, { "grad_norm": 1.1557743549346924, "learning_rate": 4.516e-05, "loss": 0.089, "step": 1130 }, { "grad_norm": 0.9965649843215942, "learning_rate": 4.5560000000000004e-05, "loss": 0.0856, "step": 1140 }, { "grad_norm": 1.1014719009399414, "learning_rate": 4.596e-05, "loss": 0.0897, "step": 1150 }, { "grad_norm": 0.9658842086791992, "learning_rate": 4.636e-05, "loss": 0.0828, "step": 1160 }, { "grad_norm": 1.3758227825164795, "learning_rate": 4.6760000000000006e-05, "loss": 0.0826, "step": 1170 }, { "grad_norm": 1.154292345046997, "learning_rate": 4.716e-05, "loss": 0.0874, "step": 1180 }, { "grad_norm": 1.21751868724823, "learning_rate": 4.7560000000000005e-05, "loss": 0.0801, "step": 1190 }, { "grad_norm": 1.3155711889266968, "learning_rate": 4.796e-05, "loss": 0.0803, "step": 1200 }, { "grad_norm": 0.7955689430236816, "learning_rate": 4.836e-05, "loss": 0.0774, "step": 1210 }, { "grad_norm": 1.1190001964569092, "learning_rate": 4.876e-05, "loss": 0.0742, "step": 1220 }, { "grad_norm": 0.9592888355255127, "learning_rate": 4.9160000000000004e-05, "loss": 0.0743, "step": 1230 }, { "grad_norm": 1.082243800163269, "learning_rate": 4.956e-05, "loss": 0.0827, "step": 1240 }, { "grad_norm": 1.1212068796157837, "learning_rate": 4.996e-05, "loss": 0.0776, "step": 1250 }, { "grad_norm": 1.1639736890792847, "learning_rate": 5.0360000000000006e-05, "loss": 0.0664, "step": 1260 }, { "grad_norm": 1.062734603881836, "learning_rate": 5.076000000000001e-05, "loss": 0.0865, "step": 1270 }, { "grad_norm": 1.0804176330566406, "learning_rate": 5.1160000000000005e-05, "loss": 0.0794, "step": 1280 }, { "grad_norm": 1.193481683731079, "learning_rate": 5.1559999999999994e-05, "loss": 0.0754, "step": 1290 }, { "grad_norm": 1.0871570110321045, "learning_rate": 5.196e-05, "loss": 0.0802, "step": 1300 }, { "grad_norm": 1.2221609354019165, "learning_rate": 5.236e-05, "loss": 0.0713, "step": 1310 }, { "grad_norm": 1.0487910509109497, "learning_rate": 5.2759999999999996e-05, "loss": 0.0863, "step": 1320 }, { "grad_norm": 1.1973837614059448, "learning_rate": 5.316e-05, "loss": 0.0721, "step": 1330 }, { "grad_norm": 0.9255446195602417, "learning_rate": 5.356e-05, "loss": 0.0694, "step": 1340 }, { "grad_norm": 0.976715087890625, "learning_rate": 5.396e-05, "loss": 0.0753, "step": 1350 }, { "grad_norm": 1.0323803424835205, "learning_rate": 5.436e-05, "loss": 0.0795, "step": 1360 }, { "grad_norm": 1.21450936794281, "learning_rate": 5.476e-05, "loss": 0.0748, "step": 1370 }, { "grad_norm": 0.9421394467353821, "learning_rate": 5.516e-05, "loss": 0.0724, "step": 1380 }, { "grad_norm": 0.8997113704681396, "learning_rate": 5.556e-05, "loss": 0.0673, "step": 1390 }, { "grad_norm": 1.1133376359939575, "learning_rate": 5.596e-05, "loss": 0.0678, "step": 1400 }, { "grad_norm": 1.0787427425384521, "learning_rate": 5.636e-05, "loss": 0.0687, "step": 1410 }, { "grad_norm": 0.8768687844276428, "learning_rate": 5.6760000000000005e-05, "loss": 0.0679, "step": 1420 }, { "grad_norm": 0.8012335896492004, "learning_rate": 5.716e-05, "loss": 0.0752, "step": 1430 }, { "grad_norm": 0.9766896367073059, "learning_rate": 5.7560000000000005e-05, "loss": 0.0682, "step": 1440 }, { "grad_norm": 1.113669514656067, "learning_rate": 5.796e-05, "loss": 0.0653, "step": 1450 }, { "grad_norm": 0.9250872731208801, "learning_rate": 5.8360000000000004e-05, "loss": 0.0728, "step": 1460 }, { "grad_norm": 0.9723067879676819, "learning_rate": 5.876000000000001e-05, "loss": 0.0628, "step": 1470 }, { "grad_norm": 0.8448038697242737, "learning_rate": 5.916e-05, "loss": 0.0619, "step": 1480 }, { "grad_norm": 0.9655065536499023, "learning_rate": 5.9560000000000006e-05, "loss": 0.0666, "step": 1490 }, { "grad_norm": 0.9177019000053406, "learning_rate": 5.996e-05, "loss": 0.0645, "step": 1500 }, { "grad_norm": 0.7535051703453064, "learning_rate": 6.0360000000000005e-05, "loss": 0.0653, "step": 1510 }, { "grad_norm": 0.6026338934898376, "learning_rate": 6.076000000000001e-05, "loss": 0.0628, "step": 1520 }, { "grad_norm": 1.5951601266860962, "learning_rate": 6.116e-05, "loss": 0.0636, "step": 1530 }, { "grad_norm": 0.8321687579154968, "learning_rate": 6.156e-05, "loss": 0.0672, "step": 1540 }, { "grad_norm": 0.8999581933021545, "learning_rate": 6.196000000000001e-05, "loss": 0.0681, "step": 1550 }, { "grad_norm": 0.8645537495613098, "learning_rate": 6.236e-05, "loss": 0.0665, "step": 1560 }, { "grad_norm": 0.9557397365570068, "learning_rate": 6.276e-05, "loss": 0.0671, "step": 1570 }, { "grad_norm": 0.8237032890319824, "learning_rate": 6.316000000000001e-05, "loss": 0.0638, "step": 1580 }, { "grad_norm": 0.61284339427948, "learning_rate": 6.356000000000001e-05, "loss": 0.0601, "step": 1590 }, { "grad_norm": 0.8453043699264526, "learning_rate": 6.396e-05, "loss": 0.0678, "step": 1600 }, { "grad_norm": 1.0573639869689941, "learning_rate": 6.436e-05, "loss": 0.0658, "step": 1610 }, { "grad_norm": 0.7517316341400146, "learning_rate": 6.476e-05, "loss": 0.0617, "step": 1620 }, { "grad_norm": 0.7313945293426514, "learning_rate": 6.515999999999999e-05, "loss": 0.0623, "step": 1630 }, { "grad_norm": 0.7904332876205444, "learning_rate": 6.556e-05, "loss": 0.0616, "step": 1640 }, { "grad_norm": 0.7817304134368896, "learning_rate": 6.596e-05, "loss": 0.0607, "step": 1650 }, { "grad_norm": 1.031121015548706, "learning_rate": 6.636e-05, "loss": 0.0585, "step": 1660 }, { "grad_norm": 0.9491569995880127, "learning_rate": 6.676e-05, "loss": 0.065, "step": 1670 }, { "grad_norm": 0.8463736176490784, "learning_rate": 6.716e-05, "loss": 0.0619, "step": 1680 }, { "grad_norm": 0.863246738910675, "learning_rate": 6.756e-05, "loss": 0.0608, "step": 1690 }, { "grad_norm": 0.9248002171516418, "learning_rate": 6.796e-05, "loss": 0.0694, "step": 1700 }, { "grad_norm": 0.8047202229499817, "learning_rate": 6.836e-05, "loss": 0.0643, "step": 1710 }, { "grad_norm": 0.9214352965354919, "learning_rate": 6.876e-05, "loss": 0.0673, "step": 1720 }, { "grad_norm": 0.7228535413742065, "learning_rate": 6.916000000000001e-05, "loss": 0.0632, "step": 1730 }, { "grad_norm": 0.7576349377632141, "learning_rate": 6.956e-05, "loss": 0.0576, "step": 1740 }, { "grad_norm": 0.7681323885917664, "learning_rate": 6.996e-05, "loss": 0.0572, "step": 1750 }, { "grad_norm": 0.6262230277061462, "learning_rate": 7.036e-05, "loss": 0.0557, "step": 1760 }, { "grad_norm": 0.8150022029876709, "learning_rate": 7.076000000000001e-05, "loss": 0.0666, "step": 1770 }, { "grad_norm": 0.6360581517219543, "learning_rate": 7.116e-05, "loss": 0.0563, "step": 1780 }, { "grad_norm": 0.6339467763900757, "learning_rate": 7.156e-05, "loss": 0.0607, "step": 1790 }, { "grad_norm": 0.6636990308761597, "learning_rate": 7.196000000000001e-05, "loss": 0.0628, "step": 1800 }, { "grad_norm": 0.7698532342910767, "learning_rate": 7.236e-05, "loss": 0.0572, "step": 1810 }, { "grad_norm": 0.7807392477989197, "learning_rate": 7.276e-05, "loss": 0.0585, "step": 1820 }, { "grad_norm": 0.7151405811309814, "learning_rate": 7.316000000000001e-05, "loss": 0.0606, "step": 1830 }, { "grad_norm": 0.7595418691635132, "learning_rate": 7.356000000000001e-05, "loss": 0.0545, "step": 1840 }, { "grad_norm": 0.572188675403595, "learning_rate": 7.396e-05, "loss": 0.0547, "step": 1850 }, { "grad_norm": 0.7924692630767822, "learning_rate": 7.436000000000001e-05, "loss": 0.0522, "step": 1860 }, { "grad_norm": 0.6961218118667603, "learning_rate": 7.476000000000001e-05, "loss": 0.0559, "step": 1870 }, { "grad_norm": 0.7582851648330688, "learning_rate": 7.516e-05, "loss": 0.0563, "step": 1880 }, { "grad_norm": 0.6650239825248718, "learning_rate": 7.556000000000002e-05, "loss": 0.0543, "step": 1890 }, { "grad_norm": 1.0109456777572632, "learning_rate": 7.596000000000001e-05, "loss": 0.062, "step": 1900 }, { "grad_norm": 0.7167805433273315, "learning_rate": 7.636e-05, "loss": 0.0574, "step": 1910 }, { "grad_norm": 0.8487099409103394, "learning_rate": 7.676e-05, "loss": 0.0513, "step": 1920 }, { "grad_norm": 0.6562585234642029, "learning_rate": 7.716e-05, "loss": 0.0536, "step": 1930 }, { "grad_norm": 0.6097492575645447, "learning_rate": 7.756e-05, "loss": 0.0623, "step": 1940 }, { "grad_norm": 0.6401941776275635, "learning_rate": 7.796e-05, "loss": 0.0562, "step": 1950 }, { "grad_norm": 0.5501066446304321, "learning_rate": 7.836e-05, "loss": 0.0545, "step": 1960 }, { "grad_norm": 0.8239641189575195, "learning_rate": 7.876e-05, "loss": 0.0567, "step": 1970 }, { "grad_norm": 0.8174377083778381, "learning_rate": 7.916e-05, "loss": 0.0603, "step": 1980 }, { "grad_norm": 0.7055029273033142, "learning_rate": 7.956e-05, "loss": 0.0539, "step": 1990 }, { "grad_norm": 0.692757785320282, "learning_rate": 7.996e-05, "loss": 0.055, "step": 2000 }, { "grad_norm": 0.6794295310974121, "learning_rate": 8.036e-05, "loss": 0.0543, "step": 2010 }, { "grad_norm": 0.612269401550293, "learning_rate": 8.076e-05, "loss": 0.0556, "step": 2020 }, { "grad_norm": 0.534481406211853, "learning_rate": 8.116e-05, "loss": 0.0524, "step": 2030 }, { "grad_norm": 0.9395762085914612, "learning_rate": 8.156e-05, "loss": 0.0559, "step": 2040 }, { "grad_norm": 0.891002893447876, "learning_rate": 8.196000000000001e-05, "loss": 0.0557, "step": 2050 }, { "grad_norm": 0.6305738687515259, "learning_rate": 8.236e-05, "loss": 0.0564, "step": 2060 }, { "grad_norm": 0.6099532246589661, "learning_rate": 8.276e-05, "loss": 0.0535, "step": 2070 }, { "grad_norm": 0.5499309301376343, "learning_rate": 8.316000000000001e-05, "loss": 0.0555, "step": 2080 }, { "grad_norm": 0.6668773293495178, "learning_rate": 8.356e-05, "loss": 0.0549, "step": 2090 }, { "grad_norm": 0.6636919379234314, "learning_rate": 8.396e-05, "loss": 0.0515, "step": 2100 }, { "grad_norm": 0.7716858983039856, "learning_rate": 8.436000000000001e-05, "loss": 0.0538, "step": 2110 }, { "grad_norm": 0.6929726004600525, "learning_rate": 8.476000000000001e-05, "loss": 0.0519, "step": 2120 }, { "grad_norm": 0.8341317176818848, "learning_rate": 8.516e-05, "loss": 0.0571, "step": 2130 }, { "grad_norm": 0.7699419260025024, "learning_rate": 8.556e-05, "loss": 0.0549, "step": 2140 }, { "grad_norm": 0.7539913654327393, "learning_rate": 8.596000000000001e-05, "loss": 0.0521, "step": 2150 }, { "grad_norm": 0.7590199112892151, "learning_rate": 8.636e-05, "loss": 0.0487, "step": 2160 }, { "grad_norm": 0.7493287920951843, "learning_rate": 8.676e-05, "loss": 0.0532, "step": 2170 }, { "grad_norm": 0.6083490252494812, "learning_rate": 8.716000000000001e-05, "loss": 0.0555, "step": 2180 }, { "grad_norm": 0.5034286379814148, "learning_rate": 8.756000000000001e-05, "loss": 0.0547, "step": 2190 }, { "grad_norm": 0.8416244983673096, "learning_rate": 8.796e-05, "loss": 0.0494, "step": 2200 }, { "grad_norm": 0.6405155062675476, "learning_rate": 8.836000000000001e-05, "loss": 0.0522, "step": 2210 }, { "grad_norm": 0.640658438205719, "learning_rate": 8.876e-05, "loss": 0.0522, "step": 2220 }, { "grad_norm": 0.7200348973274231, "learning_rate": 8.916e-05, "loss": 0.0522, "step": 2230 }, { "grad_norm": 0.5867870450019836, "learning_rate": 8.956e-05, "loss": 0.0523, "step": 2240 }, { "grad_norm": 0.7984917759895325, "learning_rate": 8.996e-05, "loss": 0.059, "step": 2250 }, { "grad_norm": 0.9847565293312073, "learning_rate": 9.036e-05, "loss": 0.0575, "step": 2260 }, { "grad_norm": 0.6851131319999695, "learning_rate": 9.076e-05, "loss": 0.0591, "step": 2270 }, { "grad_norm": 0.5983564257621765, "learning_rate": 9.116e-05, "loss": 0.0523, "step": 2280 }, { "grad_norm": 0.5568845272064209, "learning_rate": 9.156e-05, "loss": 0.0475, "step": 2290 }, { "grad_norm": 0.7043612599372864, "learning_rate": 9.196000000000001e-05, "loss": 0.0541, "step": 2300 }, { "grad_norm": 0.8094208240509033, "learning_rate": 9.236e-05, "loss": 0.0534, "step": 2310 }, { "grad_norm": 0.5728733539581299, "learning_rate": 9.276e-05, "loss": 0.0506, "step": 2320 }, { "grad_norm": 0.6831350326538086, "learning_rate": 9.316000000000001e-05, "loss": 0.0571, "step": 2330 }, { "grad_norm": 0.6958439350128174, "learning_rate": 9.356e-05, "loss": 0.0537, "step": 2340 }, { "grad_norm": 0.6414887309074402, "learning_rate": 9.396e-05, "loss": 0.0505, "step": 2350 }, { "grad_norm": 0.7128125429153442, "learning_rate": 9.436e-05, "loss": 0.0537, "step": 2360 }, { "grad_norm": 0.5700227618217468, "learning_rate": 9.476000000000001e-05, "loss": 0.0543, "step": 2370 }, { "grad_norm": 0.6090320944786072, "learning_rate": 9.516e-05, "loss": 0.0535, "step": 2380 }, { "grad_norm": 0.6238791346549988, "learning_rate": 9.556e-05, "loss": 0.0498, "step": 2390 }, { "grad_norm": 0.7641566395759583, "learning_rate": 9.596000000000001e-05, "loss": 0.0517, "step": 2400 }, { "grad_norm": 0.4163866937160492, "learning_rate": 9.636e-05, "loss": 0.0462, "step": 2410 }, { "grad_norm": 0.6031354069709778, "learning_rate": 9.676e-05, "loss": 0.0515, "step": 2420 }, { "grad_norm": 0.6536726355552673, "learning_rate": 9.716000000000001e-05, "loss": 0.0519, "step": 2430 }, { "grad_norm": 0.5245140194892883, "learning_rate": 9.756000000000001e-05, "loss": 0.0508, "step": 2440 }, { "grad_norm": 0.4661409556865692, "learning_rate": 9.796e-05, "loss": 0.0468, "step": 2450 }, { "grad_norm": 0.7656366229057312, "learning_rate": 9.836000000000001e-05, "loss": 0.0561, "step": 2460 }, { "grad_norm": 0.7389382123947144, "learning_rate": 9.876000000000001e-05, "loss": 0.0498, "step": 2470 }, { "grad_norm": 0.5763475894927979, "learning_rate": 9.916e-05, "loss": 0.0553, "step": 2480 }, { "grad_norm": 0.6870394945144653, "learning_rate": 9.956e-05, "loss": 0.05, "step": 2490 }, { "grad_norm": 0.6258082985877991, "learning_rate": 9.996000000000001e-05, "loss": 0.0524, "step": 2500 }, { "grad_norm": 0.666607677936554, "learning_rate": 9.999999114196196e-05, "loss": 0.053, "step": 2510 }, { "grad_norm": 0.6217517852783203, "learning_rate": 9.99999605215876e-05, "loss": 0.0468, "step": 2520 }, { "grad_norm": 0.6555525660514832, "learning_rate": 9.999990802953179e-05, "loss": 0.0474, "step": 2530 }, { "grad_norm": 0.46769359707832336, "learning_rate": 9.99998336658175e-05, "loss": 0.0466, "step": 2540 }, { "grad_norm": 0.6297439336776733, "learning_rate": 9.999973743047727e-05, "loss": 0.0481, "step": 2550 }, { "grad_norm": 0.6413213014602661, "learning_rate": 9.999961932355319e-05, "loss": 0.0424, "step": 2560 }, { "grad_norm": 0.47048550844192505, "learning_rate": 9.999947934509693e-05, "loss": 0.045, "step": 2570 }, { "grad_norm": 0.46701544523239136, "learning_rate": 9.999931749516971e-05, "loss": 0.0518, "step": 2580 }, { "grad_norm": 0.5390322208404541, "learning_rate": 9.999913377384233e-05, "loss": 0.0462, "step": 2590 }, { "grad_norm": 0.6083547472953796, "learning_rate": 9.999892818119517e-05, "loss": 0.0475, "step": 2600 }, { "grad_norm": 0.5764678120613098, "learning_rate": 9.999870071731814e-05, "loss": 0.048, "step": 2610 }, { "grad_norm": 0.7845823764801025, "learning_rate": 9.999845138231076e-05, "loss": 0.0484, "step": 2620 }, { "grad_norm": 0.7288833260536194, "learning_rate": 9.999818017628208e-05, "loss": 0.0504, "step": 2630 }, { "grad_norm": 0.7943646311759949, "learning_rate": 9.999788709935078e-05, "loss": 0.0527, "step": 2640 }, { "grad_norm": 0.5263519287109375, "learning_rate": 9.9997572151645e-05, "loss": 0.05, "step": 2650 }, { "grad_norm": 4.813421726226807, "learning_rate": 9.999723533330254e-05, "loss": 0.0502, "step": 2660 }, { "grad_norm": 0.5246879458427429, "learning_rate": 9.999687664447074e-05, "loss": 0.0502, "step": 2670 }, { "grad_norm": 0.6712841391563416, "learning_rate": 9.99964960853065e-05, "loss": 0.0498, "step": 2680 }, { "grad_norm": 0.48209425806999207, "learning_rate": 9.999609365597627e-05, "loss": 0.051, "step": 2690 }, { "grad_norm": 0.4784863591194153, "learning_rate": 9.99956693566561e-05, "loss": 0.0505, "step": 2700 }, { "grad_norm": 0.6906788945198059, "learning_rate": 9.99952231875316e-05, "loss": 0.0499, "step": 2710 }, { "grad_norm": 0.70460444688797, "learning_rate": 9.999475514879795e-05, "loss": 0.048, "step": 2720 }, { "grad_norm": 0.42499294877052307, "learning_rate": 9.999426524065984e-05, "loss": 0.0456, "step": 2730 }, { "grad_norm": 0.5923506021499634, "learning_rate": 9.999375346333162e-05, "loss": 0.0471, "step": 2740 }, { "grad_norm": 0.5809186697006226, "learning_rate": 9.999321981703715e-05, "loss": 0.0445, "step": 2750 }, { "grad_norm": 0.6539596319198608, "learning_rate": 9.999266430200985e-05, "loss": 0.0461, "step": 2760 }, { "grad_norm": 0.510947585105896, "learning_rate": 9.999208691849271e-05, "loss": 0.0401, "step": 2770 }, { "grad_norm": 0.6728382706642151, "learning_rate": 9.999148766673832e-05, "loss": 0.0474, "step": 2780 }, { "grad_norm": 0.5005466938018799, "learning_rate": 9.999086654700881e-05, "loss": 0.0488, "step": 2790 }, { "grad_norm": 0.4430524706840515, "learning_rate": 9.999022355957588e-05, "loss": 0.0424, "step": 2800 }, { "grad_norm": 0.547695517539978, "learning_rate": 9.998955870472079e-05, "loss": 0.0433, "step": 2810 }, { "grad_norm": 0.48438259959220886, "learning_rate": 9.998887198273437e-05, "loss": 0.0417, "step": 2820 }, { "grad_norm": 0.5398034453392029, "learning_rate": 9.998816339391701e-05, "loss": 0.0528, "step": 2830 }, { "grad_norm": 0.46309730410575867, "learning_rate": 9.998743293857868e-05, "loss": 0.0445, "step": 2840 }, { "grad_norm": 0.512346625328064, "learning_rate": 9.998668061703891e-05, "loss": 0.0401, "step": 2850 }, { "grad_norm": 0.5769650340080261, "learning_rate": 9.998590642962679e-05, "loss": 0.0421, "step": 2860 }, { "grad_norm": 0.39933204650878906, "learning_rate": 9.998511037668095e-05, "loss": 0.0446, "step": 2870 }, { "grad_norm": 0.4191771447658539, "learning_rate": 9.998429245854964e-05, "loss": 0.0446, "step": 2880 }, { "grad_norm": 0.45886674523353577, "learning_rate": 9.998345267559064e-05, "loss": 0.0435, "step": 2890 }, { "grad_norm": 0.4590684771537781, "learning_rate": 9.998259102817129e-05, "loss": 0.0446, "step": 2900 }, { "grad_norm": 0.4770794212818146, "learning_rate": 9.99817075166685e-05, "loss": 0.0403, "step": 2910 }, { "grad_norm": 0.484482079744339, "learning_rate": 9.998080214146878e-05, "loss": 0.04, "step": 2920 }, { "grad_norm": 0.5152581930160522, "learning_rate": 9.997987490296813e-05, "loss": 0.0473, "step": 2930 }, { "grad_norm": 0.5127137899398804, "learning_rate": 9.99789258015722e-05, "loss": 0.0457, "step": 2940 }, { "grad_norm": 0.4280722141265869, "learning_rate": 9.997795483769611e-05, "loss": 0.0467, "step": 2950 }, { "grad_norm": 0.4805239140987396, "learning_rate": 9.997696201176462e-05, "loss": 0.0462, "step": 2960 }, { "grad_norm": 1.1579376459121704, "learning_rate": 9.997594732421203e-05, "loss": 0.0443, "step": 2970 }, { "grad_norm": 0.5262162089347839, "learning_rate": 9.997491077548217e-05, "loss": 0.0423, "step": 2980 }, { "grad_norm": 0.5261365175247192, "learning_rate": 9.997385236602851e-05, "loss": 0.0399, "step": 2990 }, { "grad_norm": 0.45414742827415466, "learning_rate": 9.997277209631399e-05, "loss": 0.0417, "step": 3000 }, { "grad_norm": 0.4128323495388031, "learning_rate": 9.997166996681118e-05, "loss": 0.0457, "step": 3010 }, { "grad_norm": 0.3921716511249542, "learning_rate": 9.997054597800218e-05, "loss": 0.0399, "step": 3020 }, { "grad_norm": 0.41121605038642883, "learning_rate": 9.996940013037866e-05, "loss": 0.0396, "step": 3030 }, { "grad_norm": 0.5086228847503662, "learning_rate": 9.996823242444186e-05, "loss": 0.0393, "step": 3040 }, { "grad_norm": 0.462390273809433, "learning_rate": 9.996704286070258e-05, "loss": 0.0403, "step": 3050 }, { "grad_norm": 0.5634774565696716, "learning_rate": 9.996583143968115e-05, "loss": 0.0421, "step": 3060 }, { "grad_norm": 0.5618262887001038, "learning_rate": 9.99645981619075e-05, "loss": 0.0441, "step": 3070 }, { "grad_norm": 0.5766653418540955, "learning_rate": 9.996334302792114e-05, "loss": 0.0439, "step": 3080 }, { "grad_norm": 0.5606702566146851, "learning_rate": 9.996206603827105e-05, "loss": 0.0421, "step": 3090 }, { "grad_norm": 0.9824246764183044, "learning_rate": 9.996076719351587e-05, "loss": 0.0455, "step": 3100 }, { "grad_norm": 0.5920242071151733, "learning_rate": 9.995944649422374e-05, "loss": 0.0489, "step": 3110 }, { "grad_norm": 0.334346204996109, "learning_rate": 9.995810394097239e-05, "loss": 0.04, "step": 3120 }, { "grad_norm": 0.8098675012588501, "learning_rate": 9.995673953434909e-05, "loss": 0.0516, "step": 3130 }, { "grad_norm": 0.5074200630187988, "learning_rate": 9.995535327495068e-05, "loss": 0.0412, "step": 3140 }, { "grad_norm": 0.5035353899002075, "learning_rate": 9.995394516338355e-05, "loss": 0.0442, "step": 3150 }, { "grad_norm": 0.6545321941375732, "learning_rate": 9.995251520026367e-05, "loss": 0.0385, "step": 3160 }, { "grad_norm": 0.5443585515022278, "learning_rate": 9.995106338621656e-05, "loss": 0.0434, "step": 3170 }, { "grad_norm": 0.499393492937088, "learning_rate": 9.994958972187726e-05, "loss": 0.0419, "step": 3180 }, { "grad_norm": 1.082845687866211, "learning_rate": 9.994809420789044e-05, "loss": 0.0436, "step": 3190 }, { "grad_norm": 0.41878601908683777, "learning_rate": 9.994657684491027e-05, "loss": 0.0412, "step": 3200 }, { "grad_norm": 0.4529288113117218, "learning_rate": 9.994503763360048e-05, "loss": 0.0383, "step": 3210 }, { "grad_norm": 0.6763885617256165, "learning_rate": 9.99434765746344e-05, "loss": 0.0433, "step": 3220 }, { "grad_norm": 0.4430515468120575, "learning_rate": 9.994189366869488e-05, "loss": 0.0484, "step": 3230 }, { "grad_norm": 0.41141071915626526, "learning_rate": 9.994028891647433e-05, "loss": 0.0409, "step": 3240 }, { "grad_norm": 0.4580423831939697, "learning_rate": 9.993866231867475e-05, "loss": 0.0406, "step": 3250 }, { "grad_norm": 0.5903521180152893, "learning_rate": 9.993701387600762e-05, "loss": 0.0388, "step": 3260 }, { "grad_norm": 0.6016672849655151, "learning_rate": 9.993534358919408e-05, "loss": 0.0408, "step": 3270 }, { "grad_norm": 0.4283336400985718, "learning_rate": 9.993365145896473e-05, "loss": 0.0372, "step": 3280 }, { "grad_norm": 0.3902299404144287, "learning_rate": 9.993193748605977e-05, "loss": 0.0406, "step": 3290 }, { "grad_norm": 0.6298147439956665, "learning_rate": 9.993020167122898e-05, "loss": 0.042, "step": 3300 }, { "grad_norm": 0.8748878240585327, "learning_rate": 9.992844401523164e-05, "loss": 0.0448, "step": 3310 }, { "grad_norm": 1.135528326034546, "learning_rate": 9.992666451883661e-05, "loss": 0.0443, "step": 3320 }, { "grad_norm": 0.5113537311553955, "learning_rate": 9.99248631828223e-05, "loss": 0.0418, "step": 3330 }, { "grad_norm": 0.44346487522125244, "learning_rate": 9.99230400079767e-05, "loss": 0.0405, "step": 3340 }, { "grad_norm": 0.5657051205635071, "learning_rate": 9.992119499509728e-05, "loss": 0.0388, "step": 3350 }, { "grad_norm": 0.8705090284347534, "learning_rate": 9.991932814499114e-05, "loss": 0.041, "step": 3360 }, { "grad_norm": 0.38579612970352173, "learning_rate": 9.991743945847493e-05, "loss": 0.0421, "step": 3370 }, { "grad_norm": 0.4091121554374695, "learning_rate": 9.991552893637478e-05, "loss": 0.0386, "step": 3380 }, { "grad_norm": 0.5346468687057495, "learning_rate": 9.991359657952644e-05, "loss": 0.0405, "step": 3390 }, { "grad_norm": 0.48790979385375977, "learning_rate": 9.991164238877519e-05, "loss": 0.0406, "step": 3400 }, { "grad_norm": 0.5100205540657043, "learning_rate": 9.990966636497585e-05, "loss": 0.0426, "step": 3410 }, { "grad_norm": 0.395818829536438, "learning_rate": 9.99076685089928e-05, "loss": 0.0404, "step": 3420 }, { "grad_norm": 0.36916905641555786, "learning_rate": 9.990564882169998e-05, "loss": 0.0363, "step": 3430 }, { "grad_norm": 0.7100426554679871, "learning_rate": 9.990360730398088e-05, "loss": 0.0414, "step": 3440 }, { "grad_norm": 0.45728713274002075, "learning_rate": 9.990154395672849e-05, "loss": 0.0367, "step": 3450 }, { "grad_norm": 0.699738085269928, "learning_rate": 9.989945878084541e-05, "loss": 0.0396, "step": 3460 }, { "grad_norm": 0.53385329246521, "learning_rate": 9.989735177724378e-05, "loss": 0.045, "step": 3470 }, { "grad_norm": 0.6428130865097046, "learning_rate": 9.989522294684526e-05, "loss": 0.0411, "step": 3480 }, { "grad_norm": 0.46374809741973877, "learning_rate": 9.989307229058107e-05, "loss": 0.0402, "step": 3490 }, { "grad_norm": 0.3530658483505249, "learning_rate": 9.989089980939202e-05, "loss": 0.0384, "step": 3500 }, { "grad_norm": 0.42293572425842285, "learning_rate": 9.988870550422835e-05, "loss": 0.0408, "step": 3510 }, { "grad_norm": 0.45932698249816895, "learning_rate": 9.988648937604999e-05, "loss": 0.0385, "step": 3520 }, { "grad_norm": 0.45267292857170105, "learning_rate": 9.988425142582632e-05, "loss": 0.0429, "step": 3530 }, { "grad_norm": 0.39966803789138794, "learning_rate": 9.98819916545363e-05, "loss": 0.0417, "step": 3540 }, { "grad_norm": 0.47320666909217834, "learning_rate": 9.987971006316844e-05, "loss": 0.0404, "step": 3550 }, { "grad_norm": 0.5351072549819946, "learning_rate": 9.987740665272077e-05, "loss": 0.0426, "step": 3560 }, { "grad_norm": 1.395601749420166, "learning_rate": 9.98750814242009e-05, "loss": 0.0411, "step": 3570 }, { "grad_norm": 0.6042975783348083, "learning_rate": 9.987273437862594e-05, "loss": 0.0449, "step": 3580 }, { "grad_norm": 0.5850953459739685, "learning_rate": 9.987036551702259e-05, "loss": 0.0389, "step": 3590 }, { "grad_norm": 1.0517977476119995, "learning_rate": 9.986797484042706e-05, "loss": 0.0385, "step": 3600 }, { "grad_norm": 0.4840013086795807, "learning_rate": 9.986556234988512e-05, "loss": 0.037, "step": 3610 }, { "grad_norm": 0.5090408325195312, "learning_rate": 9.986312804645205e-05, "loss": 0.0335, "step": 3620 }, { "grad_norm": 0.5738756060600281, "learning_rate": 9.986067193119273e-05, "loss": 0.0411, "step": 3630 }, { "grad_norm": 0.45173144340515137, "learning_rate": 9.985819400518153e-05, "loss": 0.0376, "step": 3640 }, { "grad_norm": 0.38132426142692566, "learning_rate": 9.985569426950239e-05, "loss": 0.0352, "step": 3650 }, { "grad_norm": 0.4510959982872009, "learning_rate": 9.985317272524876e-05, "loss": 0.0341, "step": 3660 }, { "grad_norm": 0.9105472564697266, "learning_rate": 9.985062937352366e-05, "loss": 0.0444, "step": 3670 }, { "grad_norm": 0.42007577419281006, "learning_rate": 9.984806421543966e-05, "loss": 0.0422, "step": 3680 }, { "grad_norm": 0.39510488510131836, "learning_rate": 9.984547725211881e-05, "loss": 0.0432, "step": 3690 }, { "grad_norm": 0.5066283941268921, "learning_rate": 9.984286848469276e-05, "loss": 0.0356, "step": 3700 }, { "grad_norm": 0.389982134103775, "learning_rate": 9.984023791430266e-05, "loss": 0.0399, "step": 3710 }, { "grad_norm": 0.4358687400817871, "learning_rate": 9.983758554209924e-05, "loss": 0.0386, "step": 3720 }, { "grad_norm": 0.5040084719657898, "learning_rate": 9.983491136924268e-05, "loss": 0.0416, "step": 3730 }, { "grad_norm": 0.3358801305294037, "learning_rate": 9.983221539690282e-05, "loss": 0.0394, "step": 3740 }, { "grad_norm": 0.366743266582489, "learning_rate": 9.982949762625892e-05, "loss": 0.0369, "step": 3750 }, { "grad_norm": 0.46874749660491943, "learning_rate": 9.982675805849986e-05, "loss": 0.0387, "step": 3760 }, { "grad_norm": 0.32310232520103455, "learning_rate": 9.982399669482399e-05, "loss": 0.0313, "step": 3770 }, { "grad_norm": 0.3226299583911896, "learning_rate": 9.982121353643924e-05, "loss": 0.0394, "step": 3780 }, { "grad_norm": 0.4854668378829956, "learning_rate": 9.981840858456306e-05, "loss": 0.0359, "step": 3790 }, { "grad_norm": 0.5385686159133911, "learning_rate": 9.981558184042243e-05, "loss": 0.0384, "step": 3800 }, { "grad_norm": 0.39116692543029785, "learning_rate": 9.981273330525387e-05, "loss": 0.0425, "step": 3810 }, { "grad_norm": 0.4370647668838501, "learning_rate": 9.980986298030341e-05, "loss": 0.0434, "step": 3820 }, { "grad_norm": 0.3880935311317444, "learning_rate": 9.980697086682662e-05, "loss": 0.0379, "step": 3830 }, { "grad_norm": 1.0312505960464478, "learning_rate": 9.980405696608866e-05, "loss": 0.0369, "step": 3840 }, { "grad_norm": 0.5134274363517761, "learning_rate": 9.980112127936409e-05, "loss": 0.041, "step": 3850 }, { "grad_norm": 0.7203365564346313, "learning_rate": 9.979816380793717e-05, "loss": 0.0394, "step": 3860 }, { "grad_norm": 0.511918306350708, "learning_rate": 9.979518455310151e-05, "loss": 0.0373, "step": 3870 }, { "grad_norm": 0.5424188375473022, "learning_rate": 9.97921835161604e-05, "loss": 0.0365, "step": 3880 }, { "grad_norm": 0.42352986335754395, "learning_rate": 9.978916069842656e-05, "loss": 0.0371, "step": 3890 }, { "grad_norm": 0.43428394198417664, "learning_rate": 9.97861161012223e-05, "loss": 0.0415, "step": 3900 }, { "grad_norm": 0.38414478302001953, "learning_rate": 9.978304972587942e-05, "loss": 0.0363, "step": 3910 }, { "grad_norm": 0.43759217858314514, "learning_rate": 9.977996157373925e-05, "loss": 0.0397, "step": 3920 }, { "grad_norm": 0.4303003251552582, "learning_rate": 9.977685164615265e-05, "loss": 0.0347, "step": 3930 }, { "grad_norm": 0.5542426705360413, "learning_rate": 9.977371994448002e-05, "loss": 0.0375, "step": 3940 }, { "grad_norm": 1.0767014026641846, "learning_rate": 9.977056647009127e-05, "loss": 0.0367, "step": 3950 }, { "grad_norm": 0.5764644742012024, "learning_rate": 9.976739122436582e-05, "loss": 0.0377, "step": 3960 }, { "grad_norm": 0.6226363182067871, "learning_rate": 9.976419420869265e-05, "loss": 0.0354, "step": 3970 }, { "grad_norm": 0.46360599994659424, "learning_rate": 9.976097542447025e-05, "loss": 0.0333, "step": 3980 }, { "grad_norm": 0.5346417427062988, "learning_rate": 9.97577348731066e-05, "loss": 0.0358, "step": 3990 }, { "grad_norm": 0.3435487151145935, "learning_rate": 9.975447255601927e-05, "loss": 0.039, "step": 4000 }, { "grad_norm": 0.3446663022041321, "learning_rate": 9.975118847463525e-05, "loss": 0.0322, "step": 4010 }, { "grad_norm": 0.6067005395889282, "learning_rate": 9.974788263039114e-05, "loss": 0.0325, "step": 4020 }, { "grad_norm": 0.3900807797908783, "learning_rate": 9.974455502473303e-05, "loss": 0.0358, "step": 4030 }, { "grad_norm": 0.8659411668777466, "learning_rate": 9.974120565911652e-05, "loss": 0.034, "step": 4040 }, { "grad_norm": 0.7347794771194458, "learning_rate": 9.973783453500674e-05, "loss": 0.0364, "step": 4050 }, { "grad_norm": 0.34908613562583923, "learning_rate": 9.973444165387835e-05, "loss": 0.0371, "step": 4060 }, { "grad_norm": 0.38037434220314026, "learning_rate": 9.973102701721549e-05, "loss": 0.0318, "step": 4070 }, { "grad_norm": 0.44941502809524536, "learning_rate": 9.972759062651184e-05, "loss": 0.0327, "step": 4080 }, { "grad_norm": 0.40823835134506226, "learning_rate": 9.972413248327059e-05, "loss": 0.0345, "step": 4090 }, { "grad_norm": 0.4431025981903076, "learning_rate": 9.972065258900447e-05, "loss": 0.0337, "step": 4100 }, { "grad_norm": 0.3263889253139496, "learning_rate": 9.971715094523569e-05, "loss": 0.0368, "step": 4110 }, { "grad_norm": 0.4087066352367401, "learning_rate": 9.971362755349598e-05, "loss": 0.0368, "step": 4120 }, { "grad_norm": 0.49816811084747314, "learning_rate": 9.971008241532662e-05, "loss": 0.0384, "step": 4130 }, { "grad_norm": 0.44905272126197815, "learning_rate": 9.970651553227835e-05, "loss": 0.0385, "step": 4140 }, { "grad_norm": 0.4345332980155945, "learning_rate": 9.970292690591143e-05, "loss": 0.038, "step": 4150 }, { "grad_norm": 0.5800269246101379, "learning_rate": 9.969931653779569e-05, "loss": 0.0355, "step": 4160 }, { "grad_norm": 0.4070797562599182, "learning_rate": 9.969568442951038e-05, "loss": 0.0359, "step": 4170 }, { "grad_norm": 1.970528483390808, "learning_rate": 9.969203058264436e-05, "loss": 0.0361, "step": 4180 }, { "grad_norm": 0.5391925573348999, "learning_rate": 9.96883549987959e-05, "loss": 0.0398, "step": 4190 }, { "grad_norm": 0.4633467197418213, "learning_rate": 9.968465767957287e-05, "loss": 0.0309, "step": 4200 }, { "grad_norm": 0.41182956099510193, "learning_rate": 9.968093862659256e-05, "loss": 0.033, "step": 4210 }, { "grad_norm": 0.6671795845031738, "learning_rate": 9.967719784148182e-05, "loss": 0.0337, "step": 4220 }, { "grad_norm": 0.5389276146888733, "learning_rate": 9.967343532587702e-05, "loss": 0.0353, "step": 4230 }, { "grad_norm": 0.42016416788101196, "learning_rate": 9.966965108142399e-05, "loss": 0.0317, "step": 4240 }, { "grad_norm": 0.42216530442237854, "learning_rate": 9.96658451097781e-05, "loss": 0.0351, "step": 4250 }, { "grad_norm": 0.6135571002960205, "learning_rate": 9.966201741260419e-05, "loss": 0.0364, "step": 4260 }, { "grad_norm": 0.3013056218624115, "learning_rate": 9.965816799157665e-05, "loss": 0.0318, "step": 4270 }, { "grad_norm": 0.33364763855934143, "learning_rate": 9.965429684837935e-05, "loss": 0.0359, "step": 4280 }, { "grad_norm": 0.37691277265548706, "learning_rate": 9.965040398470562e-05, "loss": 0.0336, "step": 4290 }, { "grad_norm": 0.407987117767334, "learning_rate": 9.964648940225838e-05, "loss": 0.0321, "step": 4300 }, { "grad_norm": 0.39250004291534424, "learning_rate": 9.964255310274997e-05, "loss": 0.0328, "step": 4310 }, { "grad_norm": 0.3716888725757599, "learning_rate": 9.963859508790228e-05, "loss": 0.0355, "step": 4320 }, { "grad_norm": 0.5078905820846558, "learning_rate": 9.963461535944664e-05, "loss": 0.0343, "step": 4330 }, { "grad_norm": 0.38464808464050293, "learning_rate": 9.963061391912399e-05, "loss": 0.0363, "step": 4340 }, { "grad_norm": 0.4379752576351166, "learning_rate": 9.962659076868463e-05, "loss": 0.0378, "step": 4350 }, { "grad_norm": 1.0404220819473267, "learning_rate": 9.962254590988846e-05, "loss": 0.033, "step": 4360 }, { "grad_norm": 0.3971032500267029, "learning_rate": 9.961847934450481e-05, "loss": 0.0327, "step": 4370 }, { "grad_norm": 3.0986177921295166, "learning_rate": 9.961439107431257e-05, "loss": 0.0368, "step": 4380 }, { "grad_norm": 0.4399210512638092, "learning_rate": 9.961028110110006e-05, "loss": 0.0418, "step": 4390 }, { "grad_norm": 0.35769566893577576, "learning_rate": 9.960614942666513e-05, "loss": 0.0348, "step": 4400 }, { "grad_norm": 0.5083092451095581, "learning_rate": 9.960199605281511e-05, "loss": 0.0324, "step": 4410 }, { "grad_norm": 0.4591376781463623, "learning_rate": 9.959782098136683e-05, "loss": 0.0341, "step": 4420 }, { "grad_norm": 0.3579214811325073, "learning_rate": 9.959362421414662e-05, "loss": 0.032, "step": 4430 }, { "grad_norm": 0.3768116533756256, "learning_rate": 9.958940575299027e-05, "loss": 0.033, "step": 4440 }, { "grad_norm": 0.5879555344581604, "learning_rate": 9.95851655997431e-05, "loss": 0.0341, "step": 4450 }, { "grad_norm": 1.0553815364837646, "learning_rate": 9.958090375625986e-05, "loss": 0.0339, "step": 4460 }, { "grad_norm": 0.36107879877090454, "learning_rate": 9.957662022440486e-05, "loss": 0.0345, "step": 4470 }, { "grad_norm": 0.41857221722602844, "learning_rate": 9.957231500605187e-05, "loss": 0.0335, "step": 4480 }, { "grad_norm": 0.5479002594947815, "learning_rate": 9.95679881030841e-05, "loss": 0.0394, "step": 4490 }, { "grad_norm": 0.2871420979499817, "learning_rate": 9.95636395173943e-05, "loss": 0.0333, "step": 4500 }, { "grad_norm": 0.3548661768436432, "learning_rate": 9.95592692508847e-05, "loss": 0.0301, "step": 4510 }, { "grad_norm": 0.31555038690567017, "learning_rate": 9.9554877305467e-05, "loss": 0.0321, "step": 4520 }, { "grad_norm": 0.3924688398838043, "learning_rate": 9.955046368306237e-05, "loss": 0.0335, "step": 4530 }, { "grad_norm": 0.600656270980835, "learning_rate": 9.954602838560153e-05, "loss": 0.0352, "step": 4540 }, { "grad_norm": 0.3331814408302307, "learning_rate": 9.954157141502456e-05, "loss": 0.0314, "step": 4550 }, { "grad_norm": 0.4658789038658142, "learning_rate": 9.953709277328112e-05, "loss": 0.0341, "step": 4560 }, { "grad_norm": 0.38683637976646423, "learning_rate": 9.953259246233032e-05, "loss": 0.0303, "step": 4570 }, { "grad_norm": 0.45174992084503174, "learning_rate": 9.952807048414077e-05, "loss": 0.0366, "step": 4580 }, { "grad_norm": 0.4037674069404602, "learning_rate": 9.95235268406905e-05, "loss": 0.0366, "step": 4590 }, { "grad_norm": 0.7414112091064453, "learning_rate": 9.951896153396708e-05, "loss": 0.0323, "step": 4600 }, { "grad_norm": 0.5202960968017578, "learning_rate": 9.95143745659675e-05, "loss": 0.037, "step": 4610 }, { "grad_norm": 0.36746180057525635, "learning_rate": 9.95097659386983e-05, "loss": 0.0343, "step": 4620 }, { "grad_norm": 0.4569452106952667, "learning_rate": 9.950513565417542e-05, "loss": 0.036, "step": 4630 }, { "grad_norm": 0.396663635969162, "learning_rate": 9.95004837144243e-05, "loss": 0.0348, "step": 4640 }, { "grad_norm": 0.41087377071380615, "learning_rate": 9.949581012147988e-05, "loss": 0.0327, "step": 4650 }, { "grad_norm": 0.3959002196788788, "learning_rate": 9.949111487738653e-05, "loss": 0.0345, "step": 4660 }, { "grad_norm": 0.47387731075286865, "learning_rate": 9.948639798419813e-05, "loss": 0.0338, "step": 4670 }, { "grad_norm": 0.44305166602134705, "learning_rate": 9.948165944397799e-05, "loss": 0.0306, "step": 4680 }, { "grad_norm": 0.40762627124786377, "learning_rate": 9.94768992587989e-05, "loss": 0.0308, "step": 4690 }, { "grad_norm": 0.39705151319503784, "learning_rate": 9.947211743074313e-05, "loss": 0.0403, "step": 4700 }, { "grad_norm": 0.4475289583206177, "learning_rate": 9.946731396190246e-05, "loss": 0.0354, "step": 4710 }, { "grad_norm": 0.32610854506492615, "learning_rate": 9.946248885437803e-05, "loss": 0.0307, "step": 4720 }, { "grad_norm": 0.6671746373176575, "learning_rate": 9.945764211028053e-05, "loss": 0.0323, "step": 4730 }, { "grad_norm": 0.38858261704444885, "learning_rate": 9.94527737317301e-05, "loss": 0.0357, "step": 4740 }, { "grad_norm": 1.1713411808013916, "learning_rate": 9.944788372085631e-05, "loss": 0.0321, "step": 4750 }, { "grad_norm": 0.37097835540771484, "learning_rate": 9.944297207979825e-05, "loss": 0.0375, "step": 4760 }, { "grad_norm": 0.4110960364341736, "learning_rate": 9.943803881070441e-05, "loss": 0.0401, "step": 4770 }, { "grad_norm": 0.3460695743560791, "learning_rate": 9.943308391573278e-05, "loss": 0.0363, "step": 4780 }, { "grad_norm": 0.21837690472602844, "learning_rate": 9.942810739705079e-05, "loss": 0.0352, "step": 4790 }, { "grad_norm": 0.3516673445701599, "learning_rate": 9.942310925683538e-05, "loss": 0.0331, "step": 4800 }, { "grad_norm": 0.8097186088562012, "learning_rate": 9.941808949727285e-05, "loss": 0.0291, "step": 4810 }, { "grad_norm": 0.4418794512748718, "learning_rate": 9.941304812055903e-05, "loss": 0.0324, "step": 4820 }, { "grad_norm": 0.33339765667915344, "learning_rate": 9.940798512889921e-05, "loss": 0.0316, "step": 4830 }, { "grad_norm": 0.4170093834400177, "learning_rate": 9.94029005245081e-05, "loss": 0.0316, "step": 4840 }, { "grad_norm": 0.40603700280189514, "learning_rate": 9.939779430960988e-05, "loss": 0.0352, "step": 4850 }, { "grad_norm": 1.0618352890014648, "learning_rate": 9.939266648643817e-05, "loss": 0.0325, "step": 4860 }, { "grad_norm": 0.29618754982948303, "learning_rate": 9.938751705723607e-05, "loss": 0.0353, "step": 4870 }, { "grad_norm": 0.4639550447463989, "learning_rate": 9.938234602425613e-05, "loss": 0.0313, "step": 4880 }, { "grad_norm": 0.531430184841156, "learning_rate": 9.93771533897603e-05, "loss": 0.0381, "step": 4890 }, { "grad_norm": 0.4420164227485657, "learning_rate": 9.937193915602004e-05, "loss": 0.0358, "step": 4900 }, { "grad_norm": 0.3825632333755493, "learning_rate": 9.936670332531621e-05, "loss": 0.0308, "step": 4910 }, { "grad_norm": 0.3748462498188019, "learning_rate": 9.936144589993916e-05, "loss": 0.0328, "step": 4920 }, { "grad_norm": 0.6685574650764465, "learning_rate": 9.935616688218867e-05, "loss": 0.0341, "step": 4930 }, { "grad_norm": 0.3396020531654358, "learning_rate": 9.935086627437395e-05, "loss": 0.0313, "step": 4940 }, { "grad_norm": 0.32433152198791504, "learning_rate": 9.934554407881366e-05, "loss": 0.0345, "step": 4950 }, { "grad_norm": 0.41786524653434753, "learning_rate": 9.934020029783593e-05, "loss": 0.0341, "step": 4960 }, { "grad_norm": 0.4014684855937958, "learning_rate": 9.933483493377829e-05, "loss": 0.0291, "step": 4970 }, { "grad_norm": 0.3791406452655792, "learning_rate": 9.932944798898774e-05, "loss": 0.0351, "step": 4980 }, { "grad_norm": 0.5014349222183228, "learning_rate": 9.932403946582072e-05, "loss": 0.0355, "step": 4990 }, { "grad_norm": 0.3663159906864166, "learning_rate": 9.93186093666431e-05, "loss": 0.0321, "step": 5000 }, { "grad_norm": 0.3378576934337616, "learning_rate": 9.931315769383018e-05, "loss": 0.0339, "step": 5010 }, { "grad_norm": 0.38576239347457886, "learning_rate": 9.930768444976672e-05, "loss": 0.0324, "step": 5020 }, { "grad_norm": 2.0507187843322754, "learning_rate": 9.93021896368469e-05, "loss": 0.0322, "step": 5030 }, { "grad_norm": 0.3642105162143707, "learning_rate": 9.929667325747431e-05, "loss": 0.029, "step": 5040 }, { "grad_norm": 0.3559354543685913, "learning_rate": 9.929113531406205e-05, "loss": 0.032, "step": 5050 }, { "grad_norm": 0.45237860083580017, "learning_rate": 9.928557580903257e-05, "loss": 0.0341, "step": 5060 }, { "grad_norm": 0.4858781695365906, "learning_rate": 9.927999474481779e-05, "loss": 0.034, "step": 5070 }, { "grad_norm": 0.5151824355125427, "learning_rate": 9.927439212385907e-05, "loss": 0.0378, "step": 5080 }, { "grad_norm": 0.4060826301574707, "learning_rate": 9.926876794860718e-05, "loss": 0.0374, "step": 5090 }, { "grad_norm": 0.38261517882347107, "learning_rate": 9.926312222152235e-05, "loss": 0.0356, "step": 5100 }, { "grad_norm": 0.4158531427383423, "learning_rate": 9.925745494507414e-05, "loss": 0.0328, "step": 5110 }, { "grad_norm": 0.4933883547782898, "learning_rate": 9.925176612174169e-05, "loss": 0.0301, "step": 5120 }, { "grad_norm": 0.34467199444770813, "learning_rate": 9.924605575401346e-05, "loss": 0.0324, "step": 5130 }, { "grad_norm": 0.543665885925293, "learning_rate": 9.924032384438733e-05, "loss": 0.0323, "step": 5140 }, { "grad_norm": 0.3693884015083313, "learning_rate": 9.923457039537066e-05, "loss": 0.0343, "step": 5150 }, { "grad_norm": 0.2968311607837677, "learning_rate": 9.92287954094802e-05, "loss": 0.0312, "step": 5160 }, { "grad_norm": 0.32283371686935425, "learning_rate": 9.922299888924212e-05, "loss": 0.0297, "step": 5170 }, { "grad_norm": 0.5568734407424927, "learning_rate": 9.921718083719203e-05, "loss": 0.0292, "step": 5180 }, { "grad_norm": 0.3897462487220764, "learning_rate": 9.921134125587491e-05, "loss": 0.0323, "step": 5190 }, { "grad_norm": 0.3496854305267334, "learning_rate": 9.920548014784523e-05, "loss": 0.0312, "step": 5200 }, { "grad_norm": 0.4032914638519287, "learning_rate": 9.919959751566681e-05, "loss": 0.0353, "step": 5210 }, { "grad_norm": 0.3909056782722473, "learning_rate": 9.919369336191291e-05, "loss": 0.0376, "step": 5220 }, { "grad_norm": 0.4068413972854614, "learning_rate": 9.918776768916625e-05, "loss": 0.0334, "step": 5230 }, { "grad_norm": 0.297577828168869, "learning_rate": 9.918182050001888e-05, "loss": 0.0281, "step": 5240 }, { "grad_norm": 0.2914874255657196, "learning_rate": 9.91758517970723e-05, "loss": 0.0292, "step": 5250 }, { "grad_norm": 0.3246360421180725, "learning_rate": 9.916986158293744e-05, "loss": 0.0283, "step": 5260 }, { "grad_norm": 0.43877124786376953, "learning_rate": 9.916384986023463e-05, "loss": 0.0295, "step": 5270 }, { "grad_norm": 0.34605568647384644, "learning_rate": 9.915781663159359e-05, "loss": 0.0312, "step": 5280 }, { "grad_norm": 0.4374711513519287, "learning_rate": 9.915176189965346e-05, "loss": 0.0339, "step": 5290 }, { "grad_norm": 0.3922402262687683, "learning_rate": 9.914568566706279e-05, "loss": 0.0348, "step": 5300 }, { "grad_norm": 0.4346157908439636, "learning_rate": 9.913958793647953e-05, "loss": 0.0388, "step": 5310 }, { "grad_norm": 0.3218871057033539, "learning_rate": 9.913346871057102e-05, "loss": 0.0381, "step": 5320 }, { "grad_norm": 0.32593461871147156, "learning_rate": 9.912732799201403e-05, "loss": 0.0307, "step": 5330 }, { "grad_norm": 0.32591351866722107, "learning_rate": 9.912116578349474e-05, "loss": 0.0336, "step": 5340 }, { "grad_norm": 0.5876274108886719, "learning_rate": 9.911498208770867e-05, "loss": 0.032, "step": 5350 }, { "grad_norm": 0.41508254408836365, "learning_rate": 9.910877690736078e-05, "loss": 0.0292, "step": 5360 }, { "grad_norm": 0.44344279170036316, "learning_rate": 9.910255024516546e-05, "loss": 0.0303, "step": 5370 }, { "grad_norm": 0.40735259652137756, "learning_rate": 9.909630210384644e-05, "loss": 0.0294, "step": 5380 }, { "grad_norm": 0.4121004045009613, "learning_rate": 9.909003248613688e-05, "loss": 0.0303, "step": 5390 }, { "grad_norm": 0.5060335397720337, "learning_rate": 9.908374139477934e-05, "loss": 0.03, "step": 5400 }, { "grad_norm": 0.3351652920246124, "learning_rate": 9.907742883252571e-05, "loss": 0.0282, "step": 5410 }, { "grad_norm": 0.40230220556259155, "learning_rate": 9.907109480213736e-05, "loss": 0.0309, "step": 5420 }, { "grad_norm": 0.30982571840286255, "learning_rate": 9.906473930638498e-05, "loss": 0.033, "step": 5430 }, { "grad_norm": 0.37225309014320374, "learning_rate": 9.90583623480487e-05, "loss": 0.0287, "step": 5440 }, { "grad_norm": 0.4235728085041046, "learning_rate": 9.905196392991802e-05, "loss": 0.0294, "step": 5450 }, { "grad_norm": 4.565465927124023, "learning_rate": 9.904554405479183e-05, "loss": 0.0317, "step": 5460 }, { "grad_norm": 0.33063018321990967, "learning_rate": 9.903910272547838e-05, "loss": 0.0311, "step": 5470 }, { "grad_norm": 0.33484265208244324, "learning_rate": 9.903263994479533e-05, "loss": 0.0317, "step": 5480 }, { "grad_norm": 0.37283769249916077, "learning_rate": 9.902615571556972e-05, "loss": 0.0287, "step": 5490 }, { "grad_norm": 0.44840914011001587, "learning_rate": 9.9019650040638e-05, "loss": 0.029, "step": 5500 }, { "grad_norm": 0.585968554019928, "learning_rate": 9.901312292284594e-05, "loss": 0.0325, "step": 5510 }, { "grad_norm": 0.7802647352218628, "learning_rate": 9.900657436504875e-05, "loss": 0.0333, "step": 5520 }, { "grad_norm": 0.3089160919189453, "learning_rate": 9.900000437011097e-05, "loss": 0.0322, "step": 5530 }, { "grad_norm": 0.3504023849964142, "learning_rate": 9.899341294090653e-05, "loss": 0.0311, "step": 5540 }, { "grad_norm": 0.40245211124420166, "learning_rate": 9.898680008031877e-05, "loss": 0.0302, "step": 5550 }, { "grad_norm": 0.35685279965400696, "learning_rate": 9.898016579124037e-05, "loss": 0.0306, "step": 5560 }, { "grad_norm": 0.6866702437400818, "learning_rate": 9.89735100765734e-05, "loss": 0.0314, "step": 5570 }, { "grad_norm": 0.38346123695373535, "learning_rate": 9.89668329392293e-05, "loss": 0.0288, "step": 5580 }, { "grad_norm": 0.36244997382164, "learning_rate": 9.896013438212885e-05, "loss": 0.0311, "step": 5590 }, { "grad_norm": 0.2966912090778351, "learning_rate": 9.895341440820225e-05, "loss": 0.0285, "step": 5600 }, { "grad_norm": 0.3952410817146301, "learning_rate": 9.894667302038902e-05, "loss": 0.027, "step": 5610 }, { "grad_norm": 0.44906139373779297, "learning_rate": 9.893991022163812e-05, "loss": 0.0286, "step": 5620 }, { "grad_norm": 0.3440074324607849, "learning_rate": 9.893312601490779e-05, "loss": 0.0275, "step": 5630 }, { "grad_norm": 0.569525957107544, "learning_rate": 9.892632040316568e-05, "loss": 0.0294, "step": 5640 }, { "grad_norm": 0.47425317764282227, "learning_rate": 9.891949338938877e-05, "loss": 0.029, "step": 5650 }, { "grad_norm": 0.3293353021144867, "learning_rate": 9.89126449765635e-05, "loss": 0.0354, "step": 5660 }, { "grad_norm": 0.46563905477523804, "learning_rate": 9.890577516768551e-05, "loss": 0.0268, "step": 5670 }, { "grad_norm": 0.2639774680137634, "learning_rate": 9.889888396575995e-05, "loss": 0.0278, "step": 5680 }, { "grad_norm": 0.3479812741279602, "learning_rate": 9.889197137380125e-05, "loss": 0.026, "step": 5690 }, { "grad_norm": 0.4367600977420807, "learning_rate": 9.888503739483321e-05, "loss": 0.0299, "step": 5700 }, { "grad_norm": 0.49057966470718384, "learning_rate": 9.887808203188897e-05, "loss": 0.0308, "step": 5710 }, { "grad_norm": 0.4176672399044037, "learning_rate": 9.887110528801108e-05, "loss": 0.0321, "step": 5720 }, { "grad_norm": 0.3067350387573242, "learning_rate": 9.886410716625137e-05, "loss": 0.0301, "step": 5730 }, { "grad_norm": 0.3619944155216217, "learning_rate": 9.885708766967108e-05, "loss": 0.0312, "step": 5740 }, { "grad_norm": 0.558761477470398, "learning_rate": 9.885004680134076e-05, "loss": 0.029, "step": 5750 }, { "grad_norm": 0.4245290160179138, "learning_rate": 9.884298456434033e-05, "loss": 0.0277, "step": 5760 }, { "grad_norm": 0.5609451532363892, "learning_rate": 9.883590096175905e-05, "loss": 0.0322, "step": 5770 }, { "grad_norm": 0.4590107798576355, "learning_rate": 9.882879599669554e-05, "loss": 0.0282, "step": 5780 }, { "grad_norm": 0.33045536279678345, "learning_rate": 9.882166967225775e-05, "loss": 0.0314, "step": 5790 }, { "grad_norm": 0.31212830543518066, "learning_rate": 9.881452199156296e-05, "loss": 0.0299, "step": 5800 }, { "grad_norm": 0.3400100767612457, "learning_rate": 9.880735295773782e-05, "loss": 0.0288, "step": 5810 }, { "grad_norm": 0.46589699387550354, "learning_rate": 9.88001625739183e-05, "loss": 0.0312, "step": 5820 }, { "grad_norm": 0.5094183683395386, "learning_rate": 9.879295084324976e-05, "loss": 0.0323, "step": 5830 }, { "grad_norm": 0.3440849781036377, "learning_rate": 9.87857177688868e-05, "loss": 0.034, "step": 5840 }, { "grad_norm": 0.29448580741882324, "learning_rate": 9.877846335399343e-05, "loss": 0.0314, "step": 5850 }, { "grad_norm": 0.5158384442329407, "learning_rate": 9.8771187601743e-05, "loss": 0.0299, "step": 5860 }, { "grad_norm": 0.6959230303764343, "learning_rate": 9.876389051531813e-05, "loss": 0.0333, "step": 5870 }, { "grad_norm": 0.4650104343891144, "learning_rate": 9.875657209791088e-05, "loss": 0.031, "step": 5880 }, { "grad_norm": 0.41697824001312256, "learning_rate": 9.874923235272248e-05, "loss": 0.0274, "step": 5890 }, { "grad_norm": 0.3609347641468048, "learning_rate": 9.874187128296369e-05, "loss": 0.0295, "step": 5900 }, { "grad_norm": 0.4162229299545288, "learning_rate": 9.873448889185439e-05, "loss": 0.0311, "step": 5910 }, { "grad_norm": 0.3960859179496765, "learning_rate": 9.872708518262397e-05, "loss": 0.0307, "step": 5920 }, { "grad_norm": 0.501110315322876, "learning_rate": 9.871966015851101e-05, "loss": 0.0281, "step": 5930 }, { "grad_norm": 0.468349814414978, "learning_rate": 9.87122138227635e-05, "loss": 0.0329, "step": 5940 }, { "grad_norm": 0.3948225677013397, "learning_rate": 9.87047461786387e-05, "loss": 0.0297, "step": 5950 }, { "grad_norm": 0.41104939579963684, "learning_rate": 9.869725722940323e-05, "loss": 0.0289, "step": 5960 }, { "grad_norm": 0.723045289516449, "learning_rate": 9.868974697833299e-05, "loss": 0.0318, "step": 5970 }, { "grad_norm": 0.2704564929008484, "learning_rate": 9.868221542871324e-05, "loss": 0.0313, "step": 5980 }, { "grad_norm": 0.32170990109443665, "learning_rate": 9.867466258383853e-05, "loss": 0.0321, "step": 5990 }, { "grad_norm": 0.4128175377845764, "learning_rate": 9.866708844701272e-05, "loss": 0.0317, "step": 6000 }, { "grad_norm": 0.6472228169441223, "learning_rate": 9.865949302154899e-05, "loss": 0.0287, "step": 6010 }, { "grad_norm": 0.5971506834030151, "learning_rate": 9.865187631076987e-05, "loss": 0.0291, "step": 6020 }, { "grad_norm": 0.4337780475616455, "learning_rate": 9.864423831800712e-05, "loss": 0.0304, "step": 6030 }, { "grad_norm": 0.32098057866096497, "learning_rate": 9.863657904660191e-05, "loss": 0.0314, "step": 6040 }, { "grad_norm": 0.4252239763736725, "learning_rate": 9.862889849990462e-05, "loss": 0.0305, "step": 6050 }, { "grad_norm": 0.3157152235507965, "learning_rate": 9.8621196681275e-05, "loss": 0.0277, "step": 6060 }, { "grad_norm": 0.31386443972587585, "learning_rate": 9.861347359408211e-05, "loss": 0.0283, "step": 6070 }, { "grad_norm": 0.3792746663093567, "learning_rate": 9.860572924170426e-05, "loss": 0.0311, "step": 6080 }, { "grad_norm": 0.6001821756362915, "learning_rate": 9.85979636275291e-05, "loss": 0.0259, "step": 6090 }, { "grad_norm": 0.7105381488800049, "learning_rate": 9.859017675495357e-05, "loss": 0.0325, "step": 6100 }, { "grad_norm": 0.28057196736335754, "learning_rate": 9.858236862738392e-05, "loss": 0.0323, "step": 6110 }, { "grad_norm": 0.5621033310890198, "learning_rate": 9.85745392482357e-05, "loss": 0.0368, "step": 6120 }, { "grad_norm": 0.39973336458206177, "learning_rate": 9.856668862093372e-05, "loss": 0.0313, "step": 6130 }, { "grad_norm": 0.6232780814170837, "learning_rate": 9.855881674891214e-05, "loss": 0.0281, "step": 6140 }, { "grad_norm": 0.32140159606933594, "learning_rate": 9.855092363561437e-05, "loss": 0.0327, "step": 6150 }, { "grad_norm": 0.40313154458999634, "learning_rate": 9.854300928449312e-05, "loss": 0.0292, "step": 6160 }, { "grad_norm": 0.32284438610076904, "learning_rate": 9.85350736990104e-05, "loss": 0.0325, "step": 6170 }, { "grad_norm": 0.2919681966304779, "learning_rate": 9.852711688263752e-05, "loss": 0.028, "step": 6180 }, { "grad_norm": 0.3345210552215576, "learning_rate": 9.851913883885503e-05, "loss": 0.0266, "step": 6190 }, { "grad_norm": 0.3730688989162445, "learning_rate": 9.851113957115283e-05, "loss": 0.0305, "step": 6200 }, { "grad_norm": 0.3398101031780243, "learning_rate": 9.850311908303005e-05, "loss": 0.0286, "step": 6210 }, { "grad_norm": 0.4316284954547882, "learning_rate": 9.849507737799514e-05, "loss": 0.0305, "step": 6220 }, { "grad_norm": 0.30353784561157227, "learning_rate": 9.84870144595658e-05, "loss": 0.0294, "step": 6230 }, { "grad_norm": 0.34797149896621704, "learning_rate": 9.847893033126903e-05, "loss": 0.0293, "step": 6240 }, { "grad_norm": 0.2913651466369629, "learning_rate": 9.847082499664111e-05, "loss": 0.033, "step": 6250 }, { "grad_norm": 0.5929544568061829, "learning_rate": 9.846269845922758e-05, "loss": 0.0325, "step": 6260 }, { "grad_norm": 0.2913326025009155, "learning_rate": 9.845455072258326e-05, "loss": 0.0277, "step": 6270 }, { "grad_norm": 0.30201467871665955, "learning_rate": 9.844638179027226e-05, "loss": 0.0305, "step": 6280 }, { "grad_norm": 0.2926842272281647, "learning_rate": 9.843819166586795e-05, "loss": 0.0279, "step": 6290 }, { "grad_norm": 0.39250245690345764, "learning_rate": 9.842998035295294e-05, "loss": 0.0352, "step": 6300 }, { "grad_norm": 0.43702009320259094, "learning_rate": 9.842174785511918e-05, "loss": 0.0335, "step": 6310 }, { "grad_norm": 0.4270804226398468, "learning_rate": 9.841349417596779e-05, "loss": 0.0325, "step": 6320 }, { "grad_norm": 0.28717851638793945, "learning_rate": 9.840521931910926e-05, "loss": 0.0304, "step": 6330 }, { "grad_norm": 0.3839479386806488, "learning_rate": 9.839692328816327e-05, "loss": 0.0348, "step": 6340 }, { "grad_norm": 0.34978100657463074, "learning_rate": 9.838860608675879e-05, "loss": 0.033, "step": 6350 }, { "grad_norm": 0.29833024740219116, "learning_rate": 9.838026771853406e-05, "loss": 0.031, "step": 6360 }, { "grad_norm": 0.3311006724834442, "learning_rate": 9.837190818713655e-05, "loss": 0.0335, "step": 6370 }, { "grad_norm": 0.3934766948223114, "learning_rate": 9.836352749622299e-05, "loss": 0.0313, "step": 6380 }, { "grad_norm": 0.5272948145866394, "learning_rate": 9.835512564945941e-05, "loss": 0.03, "step": 6390 }, { "grad_norm": 0.3657473027706146, "learning_rate": 9.834670265052104e-05, "loss": 0.027, "step": 6400 }, { "grad_norm": 0.2723575234413147, "learning_rate": 9.833825850309239e-05, "loss": 0.0297, "step": 6410 }, { "grad_norm": 0.2959764897823334, "learning_rate": 9.832979321086723e-05, "loss": 0.0282, "step": 6420 }, { "grad_norm": 0.3099808096885681, "learning_rate": 9.832130677754854e-05, "loss": 0.0282, "step": 6430 }, { "grad_norm": 0.722625732421875, "learning_rate": 9.831279920684861e-05, "loss": 0.0293, "step": 6440 }, { "grad_norm": 0.3194388449192047, "learning_rate": 9.830427050248891e-05, "loss": 0.0287, "step": 6450 }, { "grad_norm": 0.3973899185657501, "learning_rate": 9.829572066820021e-05, "loss": 0.0267, "step": 6460 }, { "grad_norm": 0.7925723195075989, "learning_rate": 9.828714970772247e-05, "loss": 0.0266, "step": 6470 }, { "grad_norm": 0.35020995140075684, "learning_rate": 9.827855762480493e-05, "loss": 0.0309, "step": 6480 }, { "grad_norm": 0.2612389028072357, "learning_rate": 9.826994442320608e-05, "loss": 0.0297, "step": 6490 }, { "grad_norm": 0.34016871452331543, "learning_rate": 9.82613101066936e-05, "loss": 0.0266, "step": 6500 }, { "grad_norm": 2.17692232131958, "learning_rate": 9.825265467904445e-05, "loss": 0.0287, "step": 6510 }, { "grad_norm": 0.40704673528671265, "learning_rate": 9.82439781440448e-05, "loss": 0.0284, "step": 6520 }, { "grad_norm": 1.9488784074783325, "learning_rate": 9.823528050549006e-05, "loss": 0.0315, "step": 6530 }, { "grad_norm": 0.3399517834186554, "learning_rate": 9.822656176718487e-05, "loss": 0.0314, "step": 6540 }, { "grad_norm": 0.4494970738887787, "learning_rate": 9.82178219329431e-05, "loss": 0.0303, "step": 6550 }, { "grad_norm": 0.2654111087322235, "learning_rate": 9.820906100658789e-05, "loss": 0.0306, "step": 6560 }, { "grad_norm": 0.5970868468284607, "learning_rate": 9.82002789919515e-05, "loss": 0.0317, "step": 6570 }, { "grad_norm": 0.274383008480072, "learning_rate": 9.819147589287554e-05, "loss": 0.0273, "step": 6580 }, { "grad_norm": 0.3562997281551361, "learning_rate": 9.818265171321074e-05, "loss": 0.0317, "step": 6590 }, { "grad_norm": 0.632788896560669, "learning_rate": 9.817380645681714e-05, "loss": 0.0349, "step": 6600 }, { "grad_norm": 0.6840938329696655, "learning_rate": 9.816494012756392e-05, "loss": 0.034, "step": 6610 }, { "grad_norm": 0.5959239602088928, "learning_rate": 9.815605272932953e-05, "loss": 0.0283, "step": 6620 }, { "grad_norm": 0.3385678231716156, "learning_rate": 9.814714426600162e-05, "loss": 0.0349, "step": 6630 }, { "grad_norm": 0.45651504397392273, "learning_rate": 9.813821474147704e-05, "loss": 0.0335, "step": 6640 }, { "grad_norm": 0.28130242228507996, "learning_rate": 9.812926415966189e-05, "loss": 0.032, "step": 6650 }, { "grad_norm": 0.3081043064594269, "learning_rate": 9.812029252447143e-05, "loss": 0.0324, "step": 6660 }, { "grad_norm": 0.3511684238910675, "learning_rate": 9.811129983983018e-05, "loss": 0.0306, "step": 6670 }, { "grad_norm": 0.542798638343811, "learning_rate": 9.810228610967185e-05, "loss": 0.0302, "step": 6680 }, { "grad_norm": 0.33467158675193787, "learning_rate": 9.809325133793934e-05, "loss": 0.0282, "step": 6690 }, { "grad_norm": 0.2556343972682953, "learning_rate": 9.808419552858477e-05, "loss": 0.0276, "step": 6700 }, { "grad_norm": 0.40779048204421997, "learning_rate": 9.807511868556944e-05, "loss": 0.0308, "step": 6710 }, { "grad_norm": 0.384034126996994, "learning_rate": 9.806602081286392e-05, "loss": 0.0312, "step": 6720 }, { "grad_norm": 0.31741079688072205, "learning_rate": 9.805690191444785e-05, "loss": 0.0283, "step": 6730 }, { "grad_norm": 0.2726857364177704, "learning_rate": 9.804776199431022e-05, "loss": 0.0289, "step": 6740 }, { "grad_norm": 0.2809201180934906, "learning_rate": 9.80386010564491e-05, "loss": 0.0266, "step": 6750 }, { "grad_norm": 0.3554817736148834, "learning_rate": 9.802941910487183e-05, "loss": 0.0261, "step": 6760 }, { "grad_norm": 0.29199808835983276, "learning_rate": 9.802021614359487e-05, "loss": 0.0248, "step": 6770 }, { "grad_norm": 0.34253254532814026, "learning_rate": 9.801099217664394e-05, "loss": 0.0279, "step": 6780 }, { "grad_norm": 0.37720659375190735, "learning_rate": 9.800174720805388e-05, "loss": 0.0294, "step": 6790 }, { "grad_norm": 0.4577329456806183, "learning_rate": 9.799248124186878e-05, "loss": 0.0288, "step": 6800 }, { "grad_norm": 0.4222000241279602, "learning_rate": 9.798319428214187e-05, "loss": 0.0282, "step": 6810 }, { "grad_norm": 1.3489512205123901, "learning_rate": 9.79738863329356e-05, "loss": 0.0298, "step": 6820 }, { "grad_norm": 0.25480273365974426, "learning_rate": 9.79645573983216e-05, "loss": 0.0258, "step": 6830 }, { "grad_norm": 0.3416576087474823, "learning_rate": 9.795520748238062e-05, "loss": 0.029, "step": 6840 }, { "grad_norm": 0.8087950348854065, "learning_rate": 9.794583658920264e-05, "loss": 0.029, "step": 6850 }, { "grad_norm": 0.3761417865753174, "learning_rate": 9.793644472288683e-05, "loss": 0.0307, "step": 6860 }, { "grad_norm": 0.38922083377838135, "learning_rate": 9.79270318875415e-05, "loss": 0.0244, "step": 6870 }, { "grad_norm": 0.20601119101047516, "learning_rate": 9.791759808728416e-05, "loss": 0.0303, "step": 6880 }, { "grad_norm": 0.6240548491477966, "learning_rate": 9.790814332624143e-05, "loss": 0.026, "step": 6890 }, { "grad_norm": 0.39692965149879456, "learning_rate": 9.789866760854919e-05, "loss": 0.0264, "step": 6900 }, { "grad_norm": 0.2589128613471985, "learning_rate": 9.788917093835242e-05, "loss": 0.0243, "step": 6910 }, { "grad_norm": 0.8900690078735352, "learning_rate": 9.787965331980528e-05, "loss": 0.0251, "step": 6920 }, { "grad_norm": 0.32029980421066284, "learning_rate": 9.787011475707111e-05, "loss": 0.0265, "step": 6930 }, { "grad_norm": 0.3753189742565155, "learning_rate": 9.786055525432241e-05, "loss": 0.0254, "step": 6940 }, { "grad_norm": 0.38387560844421387, "learning_rate": 9.78509748157408e-05, "loss": 0.0271, "step": 6950 }, { "grad_norm": 0.40314963459968567, "learning_rate": 9.784137344551713e-05, "loss": 0.0292, "step": 6960 }, { "grad_norm": 0.28321459889411926, "learning_rate": 9.783175114785134e-05, "loss": 0.0271, "step": 6970 }, { "grad_norm": 0.2791863977909088, "learning_rate": 9.782210792695254e-05, "loss": 0.0274, "step": 6980 }, { "grad_norm": 0.3297690153121948, "learning_rate": 9.781244378703904e-05, "loss": 0.0265, "step": 6990 }, { "grad_norm": 0.3981226682662964, "learning_rate": 9.780275873233824e-05, "loss": 0.0273, "step": 7000 }, { "grad_norm": 0.3590679168701172, "learning_rate": 9.77930527670867e-05, "loss": 0.0307, "step": 7010 }, { "grad_norm": 0.486357182264328, "learning_rate": 9.778332589553016e-05, "loss": 0.0295, "step": 7020 }, { "grad_norm": 0.327648788690567, "learning_rate": 9.777357812192349e-05, "loss": 0.0266, "step": 7030 }, { "grad_norm": 4.449892044067383, "learning_rate": 9.776380945053066e-05, "loss": 0.029, "step": 7040 }, { "grad_norm": 0.3848857283592224, "learning_rate": 9.775401988562486e-05, "loss": 0.0267, "step": 7050 }, { "grad_norm": 0.29512786865234375, "learning_rate": 9.774420943148835e-05, "loss": 0.0275, "step": 7060 }, { "grad_norm": 0.2890361547470093, "learning_rate": 9.773437809241257e-05, "loss": 0.0268, "step": 7070 }, { "grad_norm": 0.3647485375404358, "learning_rate": 9.772452587269808e-05, "loss": 0.0264, "step": 7080 }, { "grad_norm": 0.28017953038215637, "learning_rate": 9.771465277665457e-05, "loss": 0.0273, "step": 7090 }, { "grad_norm": 0.22404824197292328, "learning_rate": 9.770475880860089e-05, "loss": 0.0234, "step": 7100 }, { "grad_norm": 0.22810809314250946, "learning_rate": 9.769484397286495e-05, "loss": 0.0278, "step": 7110 }, { "grad_norm": 0.30601876974105835, "learning_rate": 9.768490827378388e-05, "loss": 0.0288, "step": 7120 }, { "grad_norm": 0.47001609206199646, "learning_rate": 9.767495171570387e-05, "loss": 0.0275, "step": 7130 }, { "grad_norm": 0.3682372272014618, "learning_rate": 9.766497430298027e-05, "loss": 0.0308, "step": 7140 }, { "grad_norm": 0.2986961901187897, "learning_rate": 9.765497603997754e-05, "loss": 0.0243, "step": 7150 }, { "grad_norm": 0.28462937474250793, "learning_rate": 9.764495693106926e-05, "loss": 0.0271, "step": 7160 }, { "grad_norm": 0.6958901882171631, "learning_rate": 9.76349169806381e-05, "loss": 0.0263, "step": 7170 }, { "grad_norm": 0.2908198833465576, "learning_rate": 9.762485619307591e-05, "loss": 0.0266, "step": 7180 }, { "grad_norm": 0.32135334610939026, "learning_rate": 9.76147745727836e-05, "loss": 0.0259, "step": 7190 }, { "grad_norm": 0.39513376355171204, "learning_rate": 9.760467212417124e-05, "loss": 0.0258, "step": 7200 }, { "grad_norm": 0.3025319278240204, "learning_rate": 9.759454885165796e-05, "loss": 0.0266, "step": 7210 }, { "grad_norm": 0.37333109974861145, "learning_rate": 9.7584404759672e-05, "loss": 0.0283, "step": 7220 }, { "grad_norm": 0.3525192439556122, "learning_rate": 9.75742398526508e-05, "loss": 0.0287, "step": 7230 }, { "grad_norm": 0.30160269141197205, "learning_rate": 9.756405413504077e-05, "loss": 0.028, "step": 7240 }, { "grad_norm": 0.3546399474143982, "learning_rate": 9.755384761129752e-05, "loss": 0.027, "step": 7250 }, { "grad_norm": 0.361761212348938, "learning_rate": 9.754362028588572e-05, "loss": 0.0256, "step": 7260 }, { "grad_norm": 0.2254158854484558, "learning_rate": 9.753337216327917e-05, "loss": 0.0261, "step": 7270 }, { "grad_norm": 0.41591334342956543, "learning_rate": 9.752310324796073e-05, "loss": 0.0245, "step": 7280 }, { "grad_norm": 0.31468239426612854, "learning_rate": 9.751281354442235e-05, "loss": 0.0261, "step": 7290 }, { "grad_norm": 0.2877004146575928, "learning_rate": 9.750250305716513e-05, "loss": 0.0265, "step": 7300 }, { "grad_norm": 1.0791572332382202, "learning_rate": 9.749217179069923e-05, "loss": 0.0266, "step": 7310 }, { "grad_norm": 0.3150568902492523, "learning_rate": 9.748181974954386e-05, "loss": 0.0265, "step": 7320 }, { "grad_norm": 1.2912917137145996, "learning_rate": 9.74714469382274e-05, "loss": 0.0307, "step": 7330 }, { "grad_norm": 0.3437243700027466, "learning_rate": 9.746105336128724e-05, "loss": 0.0319, "step": 7340 }, { "grad_norm": 0.37844324111938477, "learning_rate": 9.745063902326988e-05, "loss": 0.0308, "step": 7350 }, { "grad_norm": 0.28481441736221313, "learning_rate": 9.744020392873093e-05, "loss": 0.0271, "step": 7360 }, { "grad_norm": 0.2598590552806854, "learning_rate": 9.742974808223504e-05, "loss": 0.0264, "step": 7370 }, { "grad_norm": 0.3179748058319092, "learning_rate": 9.741927148835596e-05, "loss": 0.0258, "step": 7380 }, { "grad_norm": 0.3572416305541992, "learning_rate": 9.740877415167648e-05, "loss": 0.0305, "step": 7390 }, { "grad_norm": 0.30111217498779297, "learning_rate": 9.739825607678854e-05, "loss": 0.0268, "step": 7400 }, { "grad_norm": 0.47637739777565, "learning_rate": 9.738771726829308e-05, "loss": 0.0317, "step": 7410 }, { "grad_norm": 0.2977786660194397, "learning_rate": 9.737715773080011e-05, "loss": 0.0262, "step": 7420 }, { "grad_norm": 0.3648424446582794, "learning_rate": 9.736657746892877e-05, "loss": 0.0326, "step": 7430 }, { "grad_norm": 0.4264236390590668, "learning_rate": 9.735597648730718e-05, "loss": 0.0295, "step": 7440 }, { "grad_norm": 0.47657811641693115, "learning_rate": 9.734535479057262e-05, "loss": 0.0304, "step": 7450 }, { "grad_norm": 0.4086917042732239, "learning_rate": 9.733471238337136e-05, "loss": 0.0298, "step": 7460 }, { "grad_norm": 0.5262983441352844, "learning_rate": 9.732404927035871e-05, "loss": 0.0338, "step": 7470 }, { "grad_norm": 0.3462969958782196, "learning_rate": 9.731336545619915e-05, "loss": 0.028, "step": 7480 }, { "grad_norm": 0.3212639093399048, "learning_rate": 9.73026609455661e-05, "loss": 0.0285, "step": 7490 }, { "grad_norm": 1.1465628147125244, "learning_rate": 9.72919357431421e-05, "loss": 0.031, "step": 7500 }, { "grad_norm": 0.3223799765110016, "learning_rate": 9.72811898536187e-05, "loss": 0.0262, "step": 7510 }, { "grad_norm": 0.28923943638801575, "learning_rate": 9.72704232816965e-05, "loss": 0.025, "step": 7520 }, { "grad_norm": 0.5734128952026367, "learning_rate": 9.725963603208519e-05, "loss": 0.0285, "step": 7530 }, { "grad_norm": 0.32473620772361755, "learning_rate": 9.724882810950348e-05, "loss": 0.0288, "step": 7540 }, { "grad_norm": 0.26279324293136597, "learning_rate": 9.723799951867913e-05, "loss": 0.027, "step": 7550 }, { "grad_norm": 0.3156036138534546, "learning_rate": 9.722715026434889e-05, "loss": 0.0279, "step": 7560 }, { "grad_norm": 0.3125784993171692, "learning_rate": 9.721628035125866e-05, "loss": 0.0254, "step": 7570 }, { "grad_norm": 0.3153317868709564, "learning_rate": 9.720538978416325e-05, "loss": 0.0284, "step": 7580 }, { "grad_norm": 0.34464508295059204, "learning_rate": 9.71944785678266e-05, "loss": 0.024, "step": 7590 }, { "grad_norm": 0.3165366053581238, "learning_rate": 9.718354670702161e-05, "loss": 0.0254, "step": 7600 }, { "grad_norm": 0.3801082670688629, "learning_rate": 9.717259420653029e-05, "loss": 0.0268, "step": 7610 }, { "grad_norm": 0.41107961535453796, "learning_rate": 9.716162107114361e-05, "loss": 0.0269, "step": 7620 }, { "grad_norm": 0.3646952211856842, "learning_rate": 9.71506273056616e-05, "loss": 0.0255, "step": 7630 }, { "grad_norm": 0.43806132674217224, "learning_rate": 9.713961291489331e-05, "loss": 0.0327, "step": 7640 }, { "grad_norm": 0.3187218904495239, "learning_rate": 9.71285779036568e-05, "loss": 0.0328, "step": 7650 }, { "grad_norm": 0.25959110260009766, "learning_rate": 9.711752227677916e-05, "loss": 0.0243, "step": 7660 }, { "grad_norm": 0.2806679904460907, "learning_rate": 9.710644603909649e-05, "loss": 0.0262, "step": 7670 }, { "grad_norm": 0.3043450117111206, "learning_rate": 9.709534919545393e-05, "loss": 0.0274, "step": 7680 }, { "grad_norm": 0.3009388744831085, "learning_rate": 9.708423175070563e-05, "loss": 0.0282, "step": 7690 }, { "grad_norm": 0.3829980194568634, "learning_rate": 9.707309370971468e-05, "loss": 0.0266, "step": 7700 }, { "grad_norm": 0.514723539352417, "learning_rate": 9.70619350773533e-05, "loss": 0.03, "step": 7710 }, { "grad_norm": 0.5019829273223877, "learning_rate": 9.70507558585026e-05, "loss": 0.0262, "step": 7720 }, { "grad_norm": 0.34418514370918274, "learning_rate": 9.703955605805281e-05, "loss": 0.029, "step": 7730 }, { "grad_norm": 0.3408375382423401, "learning_rate": 9.702833568090306e-05, "loss": 0.028, "step": 7740 }, { "grad_norm": 0.7375903725624084, "learning_rate": 9.701709473196154e-05, "loss": 0.0276, "step": 7750 }, { "grad_norm": 0.30899640917778015, "learning_rate": 9.700583321614541e-05, "loss": 0.0287, "step": 7760 }, { "grad_norm": 0.25521737337112427, "learning_rate": 9.699455113838085e-05, "loss": 0.0277, "step": 7770 }, { "grad_norm": 0.4358314573764801, "learning_rate": 9.698324850360303e-05, "loss": 0.0277, "step": 7780 }, { "grad_norm": 0.4392614960670471, "learning_rate": 9.69719253167561e-05, "loss": 0.0254, "step": 7790 }, { "grad_norm": 0.3147609233856201, "learning_rate": 9.69605815827932e-05, "loss": 0.0276, "step": 7800 }, { "grad_norm": 0.3881124258041382, "learning_rate": 9.694921730667647e-05, "loss": 0.0284, "step": 7810 }, { "grad_norm": 0.29466116428375244, "learning_rate": 9.693783249337705e-05, "loss": 0.0267, "step": 7820 }, { "grad_norm": 0.3546999990940094, "learning_rate": 9.692642714787501e-05, "loss": 0.0256, "step": 7830 }, { "grad_norm": 0.3742556571960449, "learning_rate": 9.691500127515945e-05, "loss": 0.0268, "step": 7840 }, { "grad_norm": 0.32023707032203674, "learning_rate": 9.690355488022844e-05, "loss": 0.0235, "step": 7850 }, { "grad_norm": 0.3688245713710785, "learning_rate": 9.689208796808902e-05, "loss": 0.0269, "step": 7860 }, { "grad_norm": 0.22729666531085968, "learning_rate": 9.68806005437572e-05, "loss": 0.0251, "step": 7870 }, { "grad_norm": 0.8114178776741028, "learning_rate": 9.686909261225796e-05, "loss": 0.0306, "step": 7880 }, { "grad_norm": 0.4933836758136749, "learning_rate": 9.685756417862531e-05, "loss": 0.0245, "step": 7890 }, { "grad_norm": 0.3316948711872101, "learning_rate": 9.684601524790212e-05, "loss": 0.0277, "step": 7900 }, { "grad_norm": 0.2638958692550659, "learning_rate": 9.68344458251403e-05, "loss": 0.03, "step": 7910 }, { "grad_norm": 0.3740740716457367, "learning_rate": 9.682285591540072e-05, "loss": 0.0289, "step": 7920 }, { "grad_norm": 0.29287540912628174, "learning_rate": 9.681124552375322e-05, "loss": 0.0301, "step": 7930 }, { "grad_norm": 0.4897215664386749, "learning_rate": 9.679961465527654e-05, "loss": 0.0288, "step": 7940 }, { "grad_norm": 0.4613178074359894, "learning_rate": 9.678796331505843e-05, "loss": 0.0243, "step": 7950 }, { "grad_norm": 0.3214024603366852, "learning_rate": 9.677629150819558e-05, "loss": 0.0264, "step": 7960 }, { "grad_norm": 0.3443026542663574, "learning_rate": 9.676459923979366e-05, "loss": 0.0271, "step": 7970 }, { "grad_norm": 0.2979810833930969, "learning_rate": 9.675288651496723e-05, "loss": 0.0279, "step": 7980 }, { "grad_norm": 0.30150821805000305, "learning_rate": 9.674115333883986e-05, "loss": 0.0233, "step": 7990 }, { "grad_norm": 0.25883013010025024, "learning_rate": 9.672939971654402e-05, "loss": 0.0234, "step": 8000 }, { "grad_norm": 0.3627030849456787, "learning_rate": 9.671762565322117e-05, "loss": 0.0261, "step": 8010 }, { "grad_norm": 0.3884207308292389, "learning_rate": 9.670583115402166e-05, "loss": 0.0307, "step": 8020 }, { "grad_norm": 0.31565743684768677, "learning_rate": 9.669401622410482e-05, "loss": 0.0273, "step": 8030 }, { "grad_norm": 0.5158175230026245, "learning_rate": 9.668218086863887e-05, "loss": 0.0284, "step": 8040 }, { "grad_norm": 0.35528314113616943, "learning_rate": 9.667032509280103e-05, "loss": 0.0245, "step": 8050 }, { "grad_norm": 0.2718878984451294, "learning_rate": 9.665844890177743e-05, "loss": 0.025, "step": 8060 }, { "grad_norm": 0.21688681840896606, "learning_rate": 9.664655230076309e-05, "loss": 0.0252, "step": 8070 }, { "grad_norm": 0.37729015946388245, "learning_rate": 9.663463529496199e-05, "loss": 0.0257, "step": 8080 }, { "grad_norm": 0.43616142868995667, "learning_rate": 9.662269788958705e-05, "loss": 0.0231, "step": 8090 }, { "grad_norm": 0.3529731035232544, "learning_rate": 9.661074008986008e-05, "loss": 0.0255, "step": 8100 }, { "grad_norm": 0.3562738299369812, "learning_rate": 9.659876190101184e-05, "loss": 0.0318, "step": 8110 }, { "grad_norm": 0.8974252939224243, "learning_rate": 9.6586763328282e-05, "loss": 0.0276, "step": 8120 }, { "grad_norm": 0.23673054575920105, "learning_rate": 9.657474437691913e-05, "loss": 0.0266, "step": 8130 }, { "grad_norm": 0.42382895946502686, "learning_rate": 9.656270505218073e-05, "loss": 0.0286, "step": 8140 }, { "grad_norm": 0.2787247896194458, "learning_rate": 9.655064535933321e-05, "loss": 0.0258, "step": 8150 }, { "grad_norm": 0.29845529794692993, "learning_rate": 9.653856530365189e-05, "loss": 0.0267, "step": 8160 }, { "grad_norm": 0.2733931839466095, "learning_rate": 9.6526464890421e-05, "loss": 0.0235, "step": 8170 }, { "grad_norm": 0.3605978786945343, "learning_rate": 9.651434412493367e-05, "loss": 0.0262, "step": 8180 }, { "grad_norm": 0.3166082799434662, "learning_rate": 9.650220301249195e-05, "loss": 0.0255, "step": 8190 }, { "grad_norm": 0.22713831067085266, "learning_rate": 9.649004155840675e-05, "loss": 0.023, "step": 8200 }, { "grad_norm": 0.26046785712242126, "learning_rate": 9.647785976799792e-05, "loss": 0.0234, "step": 8210 }, { "grad_norm": 0.30844685435295105, "learning_rate": 9.646565764659417e-05, "loss": 0.0242, "step": 8220 }, { "grad_norm": 1.8051061630249023, "learning_rate": 9.645343519953314e-05, "loss": 0.0255, "step": 8230 }, { "grad_norm": 0.28844568133354187, "learning_rate": 9.644119243216135e-05, "loss": 0.0228, "step": 8240 }, { "grad_norm": 0.3824547231197357, "learning_rate": 9.642892934983417e-05, "loss": 0.0223, "step": 8250 }, { "grad_norm": 0.3695821762084961, "learning_rate": 9.641664595791593e-05, "loss": 0.0255, "step": 8260 }, { "grad_norm": 0.518327534198761, "learning_rate": 9.640434226177977e-05, "loss": 0.0251, "step": 8270 }, { "grad_norm": 0.3420099914073944, "learning_rate": 9.639201826680777e-05, "loss": 0.0235, "step": 8280 }, { "grad_norm": 0.22544024884700775, "learning_rate": 9.637967397839083e-05, "loss": 0.0232, "step": 8290 }, { "grad_norm": 0.8443505167961121, "learning_rate": 9.63673094019288e-05, "loss": 0.0226, "step": 8300 }, { "grad_norm": 0.8314803838729858, "learning_rate": 9.635492454283035e-05, "loss": 0.0254, "step": 8310 }, { "grad_norm": 0.3869592845439911, "learning_rate": 9.634251940651302e-05, "loss": 0.0253, "step": 8320 }, { "grad_norm": 0.5078569650650024, "learning_rate": 9.633009399840327e-05, "loss": 0.0232, "step": 8330 }, { "grad_norm": 0.3985007703304291, "learning_rate": 9.631764832393639e-05, "loss": 0.0235, "step": 8340 }, { "grad_norm": 0.35007795691490173, "learning_rate": 9.630518238855651e-05, "loss": 0.0267, "step": 8350 }, { "grad_norm": 0.2905629873275757, "learning_rate": 9.629269619771668e-05, "loss": 0.0247, "step": 8360 }, { "grad_norm": 0.39969977736473083, "learning_rate": 9.628018975687878e-05, "loss": 0.0256, "step": 8370 }, { "grad_norm": 0.30943557620048523, "learning_rate": 9.626766307151355e-05, "loss": 0.0259, "step": 8380 }, { "grad_norm": 0.2448265552520752, "learning_rate": 9.62551161471006e-05, "loss": 0.0248, "step": 8390 }, { "grad_norm": 1.9838221073150635, "learning_rate": 9.624254898912837e-05, "loss": 0.0239, "step": 8400 }, { "grad_norm": 0.9082752466201782, "learning_rate": 9.622996160309414e-05, "loss": 0.023, "step": 8410 }, { "grad_norm": 0.26159268617630005, "learning_rate": 9.62173539945041e-05, "loss": 0.0226, "step": 8420 }, { "grad_norm": 0.29464688897132874, "learning_rate": 9.620472616887323e-05, "loss": 0.0218, "step": 8430 }, { "grad_norm": 0.31555652618408203, "learning_rate": 9.619207813172536e-05, "loss": 0.0249, "step": 8440 }, { "grad_norm": 0.3788968622684479, "learning_rate": 9.617940988859318e-05, "loss": 0.0254, "step": 8450 }, { "grad_norm": 0.36058297753334045, "learning_rate": 9.616672144501821e-05, "loss": 0.026, "step": 8460 }, { "grad_norm": 0.3278864622116089, "learning_rate": 9.615401280655081e-05, "loss": 0.0279, "step": 8470 }, { "grad_norm": 0.2710694372653961, "learning_rate": 9.614128397875017e-05, "loss": 0.0269, "step": 8480 }, { "grad_norm": 0.8770906925201416, "learning_rate": 9.612853496718429e-05, "loss": 0.023, "step": 8490 }, { "grad_norm": 0.2977752983570099, "learning_rate": 9.611576577743004e-05, "loss": 0.0243, "step": 8500 }, { "grad_norm": 0.30723100900650024, "learning_rate": 9.610297641507313e-05, "loss": 0.025, "step": 8510 }, { "grad_norm": 0.6131340265274048, "learning_rate": 9.6090166885708e-05, "loss": 0.0256, "step": 8520 }, { "grad_norm": 0.4230344295501709, "learning_rate": 9.607733719493798e-05, "loss": 0.0279, "step": 8530 }, { "grad_norm": 1.272980809211731, "learning_rate": 9.606448734837527e-05, "loss": 0.0287, "step": 8540 }, { "grad_norm": 0.27703163027763367, "learning_rate": 9.605161735164079e-05, "loss": 0.027, "step": 8550 }, { "grad_norm": 0.2963411509990692, "learning_rate": 9.60387272103643e-05, "loss": 0.0268, "step": 8560 }, { "grad_norm": 0.2592790424823761, "learning_rate": 9.60258169301844e-05, "loss": 0.0232, "step": 8570 }, { "grad_norm": 0.2994217276573181, "learning_rate": 9.601288651674851e-05, "loss": 0.026, "step": 8580 }, { "grad_norm": 0.32893311977386475, "learning_rate": 9.59999359757128e-05, "loss": 0.0238, "step": 8590 }, { "grad_norm": 0.2899395227432251, "learning_rate": 9.598696531274227e-05, "loss": 0.0237, "step": 8600 }, { "grad_norm": 0.3548911213874817, "learning_rate": 9.597397453351076e-05, "loss": 0.0229, "step": 8610 }, { "grad_norm": 2.3180437088012695, "learning_rate": 9.596096364370087e-05, "loss": 0.0225, "step": 8620 }, { "grad_norm": 0.487826406955719, "learning_rate": 9.594793264900399e-05, "loss": 0.0203, "step": 8630 }, { "grad_norm": 0.37775319814682007, "learning_rate": 9.593488155512032e-05, "loss": 0.0237, "step": 8640 }, { "grad_norm": 0.27287033200263977, "learning_rate": 9.592181036775886e-05, "loss": 0.0238, "step": 8650 }, { "grad_norm": 0.3498241901397705, "learning_rate": 9.590871909263741e-05, "loss": 0.0246, "step": 8660 }, { "grad_norm": 0.41574785113334656, "learning_rate": 9.589560773548252e-05, "loss": 0.0276, "step": 8670 }, { "grad_norm": 0.3050898313522339, "learning_rate": 9.588247630202952e-05, "loss": 0.0265, "step": 8680 }, { "grad_norm": 0.7887406945228577, "learning_rate": 9.586932479802258e-05, "loss": 0.027, "step": 8690 }, { "grad_norm": 0.6396527290344238, "learning_rate": 9.585615322921462e-05, "loss": 0.0224, "step": 8700 }, { "grad_norm": 0.26331618428230286, "learning_rate": 9.58429616013673e-05, "loss": 0.0224, "step": 8710 }, { "grad_norm": 0.33073657751083374, "learning_rate": 9.58297499202511e-05, "loss": 0.0204, "step": 8720 }, { "grad_norm": 0.6038263440132141, "learning_rate": 9.581651819164525e-05, "loss": 0.0261, "step": 8730 }, { "grad_norm": 1.7499375343322754, "learning_rate": 9.580326642133777e-05, "loss": 0.0252, "step": 8740 }, { "grad_norm": 0.5042064785957336, "learning_rate": 9.578999461512544e-05, "loss": 0.027, "step": 8750 }, { "grad_norm": 0.28125157952308655, "learning_rate": 9.577670277881378e-05, "loss": 0.0248, "step": 8760 }, { "grad_norm": 0.6353233456611633, "learning_rate": 9.57633909182171e-05, "loss": 0.0223, "step": 8770 }, { "grad_norm": 0.3594680428504944, "learning_rate": 9.575005903915847e-05, "loss": 0.0244, "step": 8780 }, { "grad_norm": 0.28598159551620483, "learning_rate": 9.573670714746972e-05, "loss": 0.0213, "step": 8790 }, { "grad_norm": 0.3083154857158661, "learning_rate": 9.572333524899136e-05, "loss": 0.025, "step": 8800 }, { "grad_norm": 0.65378737449646, "learning_rate": 9.570994334957278e-05, "loss": 0.0223, "step": 8810 }, { "grad_norm": 0.34839320182800293, "learning_rate": 9.569653145507201e-05, "loss": 0.0284, "step": 8820 }, { "grad_norm": 0.2451738715171814, "learning_rate": 9.568309957135587e-05, "loss": 0.0265, "step": 8830 }, { "grad_norm": 0.39252644777297974, "learning_rate": 9.566964770429995e-05, "loss": 0.0278, "step": 8840 }, { "grad_norm": 0.3490564227104187, "learning_rate": 9.565617585978853e-05, "loss": 0.0259, "step": 8850 }, { "grad_norm": 0.27216070890426636, "learning_rate": 9.564268404371466e-05, "loss": 0.0236, "step": 8860 }, { "grad_norm": 0.29475799202919006, "learning_rate": 9.562917226198013e-05, "loss": 0.024, "step": 8870 }, { "grad_norm": 0.3094750642776489, "learning_rate": 9.561564052049543e-05, "loss": 0.0255, "step": 8880 }, { "grad_norm": 0.36369749903678894, "learning_rate": 9.560208882517982e-05, "loss": 0.0276, "step": 8890 }, { "grad_norm": 0.31939780712127686, "learning_rate": 9.558851718196127e-05, "loss": 0.0239, "step": 8900 }, { "grad_norm": 0.39295703172683716, "learning_rate": 9.557492559677646e-05, "loss": 0.0235, "step": 8910 }, { "grad_norm": 0.39407169818878174, "learning_rate": 9.556131407557082e-05, "loss": 0.0225, "step": 8920 }, { "grad_norm": 0.34526750445365906, "learning_rate": 9.554768262429853e-05, "loss": 0.021, "step": 8930 }, { "grad_norm": 0.4074964225292206, "learning_rate": 9.553403124892239e-05, "loss": 0.0224, "step": 8940 }, { "grad_norm": 0.24903300404548645, "learning_rate": 9.552035995541402e-05, "loss": 0.0242, "step": 8950 }, { "grad_norm": 0.29749274253845215, "learning_rate": 9.550666874975368e-05, "loss": 0.0268, "step": 8960 }, { "grad_norm": 1.81913161277771, "learning_rate": 9.549295763793038e-05, "loss": 0.0219, "step": 8970 }, { "grad_norm": 0.32365620136260986, "learning_rate": 9.547922662594183e-05, "loss": 0.0246, "step": 8980 }, { "grad_norm": 0.47154563665390015, "learning_rate": 9.546547571979443e-05, "loss": 0.0243, "step": 8990 }, { "grad_norm": 0.26504454016685486, "learning_rate": 9.545170492550331e-05, "loss": 0.0298, "step": 9000 }, { "grad_norm": 0.3070073127746582, "learning_rate": 9.543791424909226e-05, "loss": 0.0289, "step": 9010 }, { "grad_norm": 0.32681146264076233, "learning_rate": 9.542410369659382e-05, "loss": 0.022, "step": 9020 }, { "grad_norm": 0.4826120436191559, "learning_rate": 9.541027327404916e-05, "loss": 0.025, "step": 9030 }, { "grad_norm": 0.2788658142089844, "learning_rate": 9.539642298750821e-05, "loss": 0.026, "step": 9040 }, { "grad_norm": 2.509666919708252, "learning_rate": 9.538255284302954e-05, "loss": 0.0223, "step": 9050 }, { "grad_norm": 0.2827989161014557, "learning_rate": 9.536866284668043e-05, "loss": 0.0235, "step": 9060 }, { "grad_norm": 0.3701671361923218, "learning_rate": 9.53547530045368e-05, "loss": 0.0262, "step": 9070 }, { "grad_norm": 0.46344587206840515, "learning_rate": 9.534082332268335e-05, "loss": 0.0261, "step": 9080 }, { "grad_norm": 0.24721722304821014, "learning_rate": 9.532687380721334e-05, "loss": 0.0273, "step": 9090 }, { "grad_norm": 0.27349853515625, "learning_rate": 9.53129044642288e-05, "loss": 0.0251, "step": 9100 }, { "grad_norm": 0.2571936845779419, "learning_rate": 9.529891529984039e-05, "loss": 0.0263, "step": 9110 }, { "grad_norm": 0.39838773012161255, "learning_rate": 9.528490632016743e-05, "loss": 0.0247, "step": 9120 }, { "grad_norm": 0.25506362318992615, "learning_rate": 9.52708775313379e-05, "loss": 0.0235, "step": 9130 }, { "grad_norm": 0.3207695186138153, "learning_rate": 9.525682893948853e-05, "loss": 0.024, "step": 9140 }, { "grad_norm": 0.3141022026538849, "learning_rate": 9.524276055076463e-05, "loss": 0.0218, "step": 9150 }, { "grad_norm": 0.258967787027359, "learning_rate": 9.522867237132016e-05, "loss": 0.0244, "step": 9160 }, { "grad_norm": 0.32350099086761475, "learning_rate": 9.52145644073178e-05, "loss": 0.0255, "step": 9170 }, { "grad_norm": 1.5343003273010254, "learning_rate": 9.520043666492884e-05, "loss": 0.0226, "step": 9180 }, { "grad_norm": 0.2461271435022354, "learning_rate": 9.518628915033323e-05, "loss": 0.0216, "step": 9190 }, { "grad_norm": 0.39719638228416443, "learning_rate": 9.517212186971957e-05, "loss": 0.0243, "step": 9200 }, { "grad_norm": 0.30831313133239746, "learning_rate": 9.515793482928515e-05, "loss": 0.0249, "step": 9210 }, { "grad_norm": 0.8834030628204346, "learning_rate": 9.514372803523582e-05, "loss": 0.0276, "step": 9220 }, { "grad_norm": 0.3531973659992218, "learning_rate": 9.512950149378613e-05, "loss": 0.0293, "step": 9230 }, { "grad_norm": 0.2850792407989502, "learning_rate": 9.511525521115925e-05, "loss": 0.0252, "step": 9240 }, { "grad_norm": 0.6868327260017395, "learning_rate": 9.510098919358698e-05, "loss": 0.0235, "step": 9250 }, { "grad_norm": 1.1326723098754883, "learning_rate": 9.508670344730979e-05, "loss": 0.024, "step": 9260 }, { "grad_norm": 0.4209141135215759, "learning_rate": 9.507239797857672e-05, "loss": 0.0239, "step": 9270 }, { "grad_norm": 0.26594552397727966, "learning_rate": 9.505807279364548e-05, "loss": 0.0251, "step": 9280 }, { "grad_norm": 0.28670111298561096, "learning_rate": 9.504372789878239e-05, "loss": 0.0228, "step": 9290 }, { "grad_norm": 0.33255669474601746, "learning_rate": 9.502936330026239e-05, "loss": 0.023, "step": 9300 }, { "grad_norm": 0.513319194316864, "learning_rate": 9.501497900436906e-05, "loss": 0.0211, "step": 9310 }, { "grad_norm": 0.43281692266464233, "learning_rate": 9.500057501739457e-05, "loss": 0.0221, "step": 9320 }, { "grad_norm": 0.36335617303848267, "learning_rate": 9.49861513456397e-05, "loss": 0.0226, "step": 9330 }, { "grad_norm": 0.3208250403404236, "learning_rate": 9.497170799541388e-05, "loss": 0.0222, "step": 9340 }, { "grad_norm": 0.29278236627578735, "learning_rate": 9.495724497303508e-05, "loss": 0.0244, "step": 9350 }, { "grad_norm": 1.4067258834838867, "learning_rate": 9.494276228482998e-05, "loss": 0.0238, "step": 9360 }, { "grad_norm": 0.3744891285896301, "learning_rate": 9.492825993713374e-05, "loss": 0.0223, "step": 9370 }, { "grad_norm": 0.22367005050182343, "learning_rate": 9.491373793629023e-05, "loss": 0.024, "step": 9380 }, { "grad_norm": 0.23254112899303436, "learning_rate": 9.489919628865182e-05, "loss": 0.0193, "step": 9390 }, { "grad_norm": 0.23648671805858612, "learning_rate": 9.488463500057955e-05, "loss": 0.0229, "step": 9400 }, { "grad_norm": 0.2883433997631073, "learning_rate": 9.487005407844302e-05, "loss": 0.0208, "step": 9410 }, { "grad_norm": 0.5025551915168762, "learning_rate": 9.485545352862039e-05, "loss": 0.0236, "step": 9420 }, { "grad_norm": 0.31678229570388794, "learning_rate": 9.48408333574985e-05, "loss": 0.0251, "step": 9430 }, { "grad_norm": 0.2648313045501709, "learning_rate": 9.482619357147264e-05, "loss": 0.0235, "step": 9440 }, { "grad_norm": 1.8271492719650269, "learning_rate": 9.481153417694679e-05, "loss": 0.0248, "step": 9450 }, { "grad_norm": 0.39660459756851196, "learning_rate": 9.479685518033347e-05, "loss": 0.0229, "step": 9460 }, { "grad_norm": 0.35442641377449036, "learning_rate": 9.478215658805375e-05, "loss": 0.0239, "step": 9470 }, { "grad_norm": 0.6442456841468811, "learning_rate": 9.47674384065373e-05, "loss": 0.0263, "step": 9480 }, { "grad_norm": 0.46469584107398987, "learning_rate": 9.475270064222237e-05, "loss": 0.025, "step": 9490 }, { "grad_norm": 0.5763760209083557, "learning_rate": 9.473794330155572e-05, "loss": 0.0235, "step": 9500 }, { "grad_norm": 0.2471669763326645, "learning_rate": 9.472316639099276e-05, "loss": 0.0243, "step": 9510 }, { "grad_norm": 0.24884311854839325, "learning_rate": 9.470836991699739e-05, "loss": 0.0203, "step": 9520 }, { "grad_norm": 0.3236096501350403, "learning_rate": 9.469355388604208e-05, "loss": 0.0183, "step": 9530 }, { "grad_norm": 0.3348851501941681, "learning_rate": 9.467871830460787e-05, "loss": 0.0219, "step": 9540 }, { "grad_norm": 0.32609230279922485, "learning_rate": 9.466386317918436e-05, "loss": 0.0254, "step": 9550 }, { "grad_norm": 0.26916563510894775, "learning_rate": 9.464898851626969e-05, "loss": 0.0226, "step": 9560 }, { "grad_norm": 0.2905917763710022, "learning_rate": 9.463409432237051e-05, "loss": 0.0223, "step": 9570 }, { "grad_norm": 0.2934432923793793, "learning_rate": 9.461918060400209e-05, "loss": 0.0247, "step": 9580 }, { "grad_norm": 0.31839415431022644, "learning_rate": 9.460424736768816e-05, "loss": 0.0243, "step": 9590 }, { "grad_norm": 0.749220609664917, "learning_rate": 9.458929461996105e-05, "loss": 0.0207, "step": 9600 }, { "grad_norm": 0.5584128499031067, "learning_rate": 9.457432236736158e-05, "loss": 0.028, "step": 9610 }, { "grad_norm": 0.7085126042366028, "learning_rate": 9.455933061643916e-05, "loss": 0.0212, "step": 9620 }, { "grad_norm": 0.3023313581943512, "learning_rate": 9.454431937375164e-05, "loss": 0.0214, "step": 9630 }, { "grad_norm": 0.6677035093307495, "learning_rate": 9.452928864586547e-05, "loss": 0.0245, "step": 9640 }, { "grad_norm": 0.7836034297943115, "learning_rate": 9.451423843935563e-05, "loss": 0.0209, "step": 9650 }, { "grad_norm": 0.23886284232139587, "learning_rate": 9.449916876080553e-05, "loss": 0.0235, "step": 9660 }, { "grad_norm": 0.23942652344703674, "learning_rate": 9.44840796168072e-05, "loss": 0.0214, "step": 9670 }, { "grad_norm": 0.2622588872909546, "learning_rate": 9.446897101396115e-05, "loss": 0.0205, "step": 9680 }, { "grad_norm": 0.590168833732605, "learning_rate": 9.445384295887638e-05, "loss": 0.0238, "step": 9690 }, { "grad_norm": 0.3515818417072296, "learning_rate": 9.443869545817043e-05, "loss": 0.0227, "step": 9700 }, { "grad_norm": 0.3120192289352417, "learning_rate": 9.442352851846929e-05, "loss": 0.0237, "step": 9710 }, { "grad_norm": 0.3457222580909729, "learning_rate": 9.440834214640755e-05, "loss": 0.0217, "step": 9720 }, { "grad_norm": 0.41705256700515747, "learning_rate": 9.439313634862823e-05, "loss": 0.0213, "step": 9730 }, { "grad_norm": 0.2828817069530487, "learning_rate": 9.437791113178282e-05, "loss": 0.0226, "step": 9740 }, { "grad_norm": 0.3297840356826782, "learning_rate": 9.43626665025314e-05, "loss": 0.0235, "step": 9750 }, { "grad_norm": 0.30532020330429077, "learning_rate": 9.434740246754248e-05, "loss": 0.0215, "step": 9760 }, { "grad_norm": 0.5203633904457092, "learning_rate": 9.433211903349304e-05, "loss": 0.0199, "step": 9770 }, { "grad_norm": 0.26579323410987854, "learning_rate": 9.431681620706858e-05, "loss": 0.0178, "step": 9780 }, { "grad_norm": 0.24679771065711975, "learning_rate": 9.43014939949631e-05, "loss": 0.0203, "step": 9790 }, { "grad_norm": 0.2115209847688675, "learning_rate": 9.428615240387904e-05, "loss": 0.0198, "step": 9800 }, { "grad_norm": 0.2740117013454437, "learning_rate": 9.427079144052732e-05, "loss": 0.0268, "step": 9810 }, { "grad_norm": 0.36923786997795105, "learning_rate": 9.425541111162739e-05, "loss": 0.0245, "step": 9820 }, { "grad_norm": 0.30881398916244507, "learning_rate": 9.424001142390709e-05, "loss": 0.026, "step": 9830 }, { "grad_norm": 0.3089563846588135, "learning_rate": 9.422459238410277e-05, "loss": 0.0245, "step": 9840 }, { "grad_norm": 0.3190114498138428, "learning_rate": 9.420915399895926e-05, "loss": 0.0246, "step": 9850 }, { "grad_norm": 0.22873394191265106, "learning_rate": 9.419369627522981e-05, "loss": 0.0256, "step": 9860 }, { "grad_norm": 0.22152374684810638, "learning_rate": 9.417821921967618e-05, "loss": 0.0215, "step": 9870 }, { "grad_norm": 0.24583159387111664, "learning_rate": 9.416272283906855e-05, "loss": 0.0221, "step": 9880 }, { "grad_norm": 0.22470895946025848, "learning_rate": 9.414720714018554e-05, "loss": 0.0209, "step": 9890 }, { "grad_norm": 0.2679401636123657, "learning_rate": 9.413167212981427e-05, "loss": 0.0208, "step": 9900 }, { "grad_norm": 0.6159454584121704, "learning_rate": 9.41161178147503e-05, "loss": 0.0227, "step": 9910 }, { "grad_norm": 0.33098599314689636, "learning_rate": 9.410054420179755e-05, "loss": 0.0226, "step": 9920 }, { "grad_norm": 0.3727346956729889, "learning_rate": 9.408495129776851e-05, "loss": 0.0204, "step": 9930 }, { "grad_norm": 0.3144613802433014, "learning_rate": 9.406933910948403e-05, "loss": 0.018, "step": 9940 }, { "grad_norm": 0.32740169763565063, "learning_rate": 9.40537076437734e-05, "loss": 0.0242, "step": 9950 }, { "grad_norm": 0.40456509590148926, "learning_rate": 9.403805690747436e-05, "loss": 0.0199, "step": 9960 }, { "grad_norm": 0.29216665029525757, "learning_rate": 9.402238690743308e-05, "loss": 0.0183, "step": 9970 }, { "grad_norm": 0.21952763199806213, "learning_rate": 9.400669765050413e-05, "loss": 0.0169, "step": 9980 }, { "grad_norm": 0.3373316824436188, "learning_rate": 9.399098914355055e-05, "loss": 0.0196, "step": 9990 }, { "grad_norm": 0.29651981592178345, "learning_rate": 9.397526139344378e-05, "loss": 0.0211, "step": 10000 }, { "grad_norm": 0.6384314894676208, "learning_rate": 9.395951440706362e-05, "loss": 0.0216, "step": 10010 }, { "grad_norm": 0.3315469026565552, "learning_rate": 9.394374819129839e-05, "loss": 0.0202, "step": 10020 }, { "grad_norm": 0.2589012682437897, "learning_rate": 9.392796275304474e-05, "loss": 0.0216, "step": 10030 }, { "grad_norm": 0.25446581840515137, "learning_rate": 9.391215809920778e-05, "loss": 0.0214, "step": 10040 }, { "grad_norm": 0.3642868399620056, "learning_rate": 9.389633423670099e-05, "loss": 0.0246, "step": 10050 }, { "grad_norm": 0.23702135682106018, "learning_rate": 9.388049117244626e-05, "loss": 0.0213, "step": 10060 }, { "grad_norm": 0.31641221046447754, "learning_rate": 9.386462891337389e-05, "loss": 0.02, "step": 10070 }, { "grad_norm": 0.19920837879180908, "learning_rate": 9.384874746642257e-05, "loss": 0.0278, "step": 10080 }, { "grad_norm": 0.32323572039604187, "learning_rate": 9.383284683853937e-05, "loss": 0.0204, "step": 10090 }, { "grad_norm": 0.5451815128326416, "learning_rate": 9.381692703667981e-05, "loss": 0.0187, "step": 10100 }, { "grad_norm": 0.19146588444709778, "learning_rate": 9.380098806780771e-05, "loss": 0.0176, "step": 10110 }, { "grad_norm": 0.27831143140792847, "learning_rate": 9.378502993889533e-05, "loss": 0.0203, "step": 10120 }, { "grad_norm": 0.3183557689189911, "learning_rate": 9.376905265692329e-05, "loss": 0.0243, "step": 10130 }, { "grad_norm": 0.29399487376213074, "learning_rate": 9.37530562288806e-05, "loss": 0.0218, "step": 10140 }, { "grad_norm": 0.16677220165729523, "learning_rate": 9.373704066176465e-05, "loss": 0.0202, "step": 10150 }, { "grad_norm": 0.2926161587238312, "learning_rate": 9.372100596258118e-05, "loss": 0.0236, "step": 10160 }, { "grad_norm": 0.2680307924747467, "learning_rate": 9.370495213834433e-05, "loss": 0.0233, "step": 10170 }, { "grad_norm": 0.4250504672527313, "learning_rate": 9.368887919607657e-05, "loss": 0.0233, "step": 10180 }, { "grad_norm": 0.232414111495018, "learning_rate": 9.367278714280876e-05, "loss": 0.0205, "step": 10190 }, { "grad_norm": 0.23694217205047607, "learning_rate": 9.36566759855801e-05, "loss": 0.0192, "step": 10200 }, { "grad_norm": 0.392975777387619, "learning_rate": 9.36405457314382e-05, "loss": 0.0234, "step": 10210 }, { "grad_norm": 0.3411817252635956, "learning_rate": 9.36243963874389e-05, "loss": 0.0243, "step": 10220 }, { "grad_norm": 0.3269507884979248, "learning_rate": 9.360822796064655e-05, "loss": 0.0209, "step": 10230 }, { "grad_norm": 0.27951717376708984, "learning_rate": 9.359204045813372e-05, "loss": 0.0194, "step": 10240 }, { "grad_norm": 0.5409907102584839, "learning_rate": 9.357583388698141e-05, "loss": 0.0198, "step": 10250 }, { "grad_norm": 3.37890887260437, "learning_rate": 9.35596082542789e-05, "loss": 0.0251, "step": 10260 }, { "grad_norm": 0.3329079747200012, "learning_rate": 9.354336356712383e-05, "loss": 0.021, "step": 10270 }, { "grad_norm": 0.24679817259311676, "learning_rate": 9.35270998326222e-05, "loss": 0.0236, "step": 10280 }, { "grad_norm": 0.23722682893276215, "learning_rate": 9.351081705788831e-05, "loss": 0.0186, "step": 10290 }, { "grad_norm": 0.3136445879936218, "learning_rate": 9.349451525004477e-05, "loss": 0.0236, "step": 10300 }, { "grad_norm": 0.40975821018218994, "learning_rate": 9.347819441622261e-05, "loss": 0.0224, "step": 10310 }, { "grad_norm": 0.30926549434661865, "learning_rate": 9.346185456356105e-05, "loss": 0.0206, "step": 10320 }, { "grad_norm": 0.49744611978530884, "learning_rate": 9.344549569920774e-05, "loss": 0.0229, "step": 10330 }, { "grad_norm": 0.33682781457901, "learning_rate": 9.342911783031858e-05, "loss": 0.0206, "step": 10340 }, { "grad_norm": 0.42166391015052795, "learning_rate": 9.341272096405782e-05, "loss": 0.0255, "step": 10350 }, { "grad_norm": 0.3731335401535034, "learning_rate": 9.3396305107598e-05, "loss": 0.0227, "step": 10360 }, { "grad_norm": 0.28083306550979614, "learning_rate": 9.337987026811998e-05, "loss": 0.0195, "step": 10370 }, { "grad_norm": 0.49750226736068726, "learning_rate": 9.33634164528129e-05, "loss": 0.02, "step": 10380 }, { "grad_norm": 0.31327593326568604, "learning_rate": 9.334694366887424e-05, "loss": 0.0221, "step": 10390 }, { "grad_norm": 0.18337106704711914, "learning_rate": 9.333045192350973e-05, "loss": 0.0206, "step": 10400 }, { "grad_norm": 0.281028151512146, "learning_rate": 9.331394122393345e-05, "loss": 0.0191, "step": 10410 }, { "grad_norm": 0.28020918369293213, "learning_rate": 9.329741157736771e-05, "loss": 0.0179, "step": 10420 }, { "grad_norm": 0.23396189510822296, "learning_rate": 9.328086299104317e-05, "loss": 0.0215, "step": 10430 }, { "grad_norm": 2.292588949203491, "learning_rate": 9.326429547219872e-05, "loss": 0.0219, "step": 10440 }, { "grad_norm": 0.23808366060256958, "learning_rate": 9.324770902808155e-05, "loss": 0.0197, "step": 10450 }, { "grad_norm": 0.6423144340515137, "learning_rate": 9.323110366594717e-05, "loss": 0.0201, "step": 10460 }, { "grad_norm": 0.9892075061798096, "learning_rate": 9.32144793930593e-05, "loss": 0.0247, "step": 10470 }, { "grad_norm": 0.3183439373970032, "learning_rate": 9.319783621668996e-05, "loss": 0.0214, "step": 10480 }, { "grad_norm": 0.27389466762542725, "learning_rate": 9.318117414411947e-05, "loss": 0.0207, "step": 10490 }, { "grad_norm": 0.2624729871749878, "learning_rate": 9.316449318263635e-05, "loss": 0.0193, "step": 10500 }, { "grad_norm": 0.3655410706996918, "learning_rate": 9.314779333953744e-05, "loss": 0.0203, "step": 10510 }, { "grad_norm": 0.266464501619339, "learning_rate": 9.313107462212781e-05, "loss": 0.0212, "step": 10520 }, { "grad_norm": 0.2688986659049988, "learning_rate": 9.311433703772082e-05, "loss": 0.0197, "step": 10530 }, { "grad_norm": 0.27783703804016113, "learning_rate": 9.3097580593638e-05, "loss": 0.0197, "step": 10540 }, { "grad_norm": 0.24098612368106842, "learning_rate": 9.308080529720926e-05, "loss": 0.0209, "step": 10550 }, { "grad_norm": 0.23339639604091644, "learning_rate": 9.306401115577264e-05, "loss": 0.0197, "step": 10560 }, { "grad_norm": 0.26668596267700195, "learning_rate": 9.304719817667447e-05, "loss": 0.0209, "step": 10570 }, { "grad_norm": 0.30715110898017883, "learning_rate": 9.303036636726934e-05, "loss": 0.019, "step": 10580 }, { "grad_norm": 0.28553053736686707, "learning_rate": 9.301351573492003e-05, "loss": 0.0218, "step": 10590 }, { "grad_norm": 0.3277141749858856, "learning_rate": 9.299664628699758e-05, "loss": 0.0215, "step": 10600 }, { "grad_norm": 0.2395605742931366, "learning_rate": 9.297975803088129e-05, "loss": 0.0226, "step": 10610 }, { "grad_norm": 0.2421257495880127, "learning_rate": 9.296285097395864e-05, "loss": 0.0245, "step": 10620 }, { "grad_norm": 0.29771488904953003, "learning_rate": 9.294592512362533e-05, "loss": 0.0202, "step": 10630 }, { "grad_norm": 0.31701159477233887, "learning_rate": 9.292898048728533e-05, "loss": 0.0262, "step": 10640 }, { "grad_norm": 0.19613641500473022, "learning_rate": 9.29120170723508e-05, "loss": 0.0222, "step": 10650 }, { "grad_norm": 0.30780717730522156, "learning_rate": 9.28950348862421e-05, "loss": 0.0243, "step": 10660 }, { "grad_norm": 0.2877557873725891, "learning_rate": 9.287803393638781e-05, "loss": 0.0191, "step": 10670 }, { "grad_norm": 0.25818416476249695, "learning_rate": 9.286101423022474e-05, "loss": 0.019, "step": 10680 }, { "grad_norm": 0.279629647731781, "learning_rate": 9.284397577519788e-05, "loss": 0.0177, "step": 10690 }, { "grad_norm": 0.49824798107147217, "learning_rate": 9.282691857876043e-05, "loss": 0.0196, "step": 10700 }, { "grad_norm": 0.29067620635032654, "learning_rate": 9.280984264837377e-05, "loss": 0.017, "step": 10710 }, { "grad_norm": 0.33129310607910156, "learning_rate": 9.279274799150752e-05, "loss": 0.0217, "step": 10720 }, { "grad_norm": 0.26493939757347107, "learning_rate": 9.277563461563945e-05, "loss": 0.0206, "step": 10730 }, { "grad_norm": 0.2780015170574188, "learning_rate": 9.275850252825555e-05, "loss": 0.0185, "step": 10740 }, { "grad_norm": 0.34188729524612427, "learning_rate": 9.274135173684994e-05, "loss": 0.0184, "step": 10750 }, { "grad_norm": 0.3185967206954956, "learning_rate": 9.272418224892498e-05, "loss": 0.0185, "step": 10760 }, { "grad_norm": 0.6467390060424805, "learning_rate": 9.27069940719912e-05, "loss": 0.0211, "step": 10770 }, { "grad_norm": 0.33017098903656006, "learning_rate": 9.268978721356727e-05, "loss": 0.0206, "step": 10780 }, { "grad_norm": 0.8895438313484192, "learning_rate": 9.267256168118008e-05, "loss": 0.0227, "step": 10790 }, { "grad_norm": 0.3023849129676819, "learning_rate": 9.265531748236463e-05, "loss": 0.0209, "step": 10800 }, { "grad_norm": 0.20325274765491486, "learning_rate": 9.263805462466416e-05, "loss": 0.0172, "step": 10810 }, { "grad_norm": 0.32968178391456604, "learning_rate": 9.262077311562998e-05, "loss": 0.0211, "step": 10820 }, { "grad_norm": 0.37901169061660767, "learning_rate": 9.260347296282165e-05, "loss": 0.022, "step": 10830 }, { "grad_norm": 0.5584158301353455, "learning_rate": 9.258615417380683e-05, "loss": 0.0207, "step": 10840 }, { "grad_norm": 0.5716179609298706, "learning_rate": 9.256881675616133e-05, "loss": 0.0188, "step": 10850 }, { "grad_norm": 0.3149513304233551, "learning_rate": 9.255146071746917e-05, "loss": 0.0219, "step": 10860 }, { "grad_norm": 0.23809152841567993, "learning_rate": 9.253408606532241e-05, "loss": 0.0218, "step": 10870 }, { "grad_norm": 0.18771697580814362, "learning_rate": 9.251669280732137e-05, "loss": 0.0192, "step": 10880 }, { "grad_norm": 0.3664298355579376, "learning_rate": 9.249928095107441e-05, "loss": 0.0225, "step": 10890 }, { "grad_norm": 0.2261797934770584, "learning_rate": 9.248185050419811e-05, "loss": 0.0197, "step": 10900 }, { "grad_norm": 0.3617207407951355, "learning_rate": 9.24644014743171e-05, "loss": 0.0209, "step": 10910 }, { "grad_norm": 0.3946956396102905, "learning_rate": 9.24469338690642e-05, "loss": 0.02, "step": 10920 }, { "grad_norm": 0.20211957395076752, "learning_rate": 9.242944769608033e-05, "loss": 0.0195, "step": 10930 }, { "grad_norm": 0.24195511639118195, "learning_rate": 9.241194296301454e-05, "loss": 0.0206, "step": 10940 }, { "grad_norm": 0.4432208836078644, "learning_rate": 9.239441967752397e-05, "loss": 0.0149, "step": 10950 }, { "grad_norm": 0.468880832195282, "learning_rate": 9.237687784727393e-05, "loss": 0.0172, "step": 10960 }, { "grad_norm": 0.23690403997898102, "learning_rate": 9.235931747993781e-05, "loss": 0.018, "step": 10970 }, { "grad_norm": 0.2388639897108078, "learning_rate": 9.234173858319707e-05, "loss": 0.0193, "step": 10980 }, { "grad_norm": 0.24680878221988678, "learning_rate": 9.23241411647414e-05, "loss": 0.0206, "step": 10990 }, { "grad_norm": 1.1903266906738281, "learning_rate": 9.230652523226841e-05, "loss": 0.0214, "step": 11000 }, { "grad_norm": 0.28475770354270935, "learning_rate": 9.2288890793484e-05, "loss": 0.017, "step": 11010 }, { "grad_norm": 0.1900770664215088, "learning_rate": 9.227123785610199e-05, "loss": 0.0173, "step": 11020 }, { "grad_norm": 0.334187775850296, "learning_rate": 9.225356642784443e-05, "loss": 0.0186, "step": 11030 }, { "grad_norm": 0.3313678503036499, "learning_rate": 9.223587651644139e-05, "loss": 0.0162, "step": 11040 }, { "grad_norm": 0.2851036787033081, "learning_rate": 9.221816812963104e-05, "loss": 0.0189, "step": 11050 }, { "grad_norm": 0.2584148943424225, "learning_rate": 9.22004412751596e-05, "loss": 0.0177, "step": 11060 }, { "grad_norm": 0.38020840287208557, "learning_rate": 9.218269596078146e-05, "loss": 0.0185, "step": 11070 }, { "grad_norm": 0.3241646885871887, "learning_rate": 9.216493219425895e-05, "loss": 0.0191, "step": 11080 }, { "grad_norm": 0.2501024603843689, "learning_rate": 9.21471499833626e-05, "loss": 0.0183, "step": 11090 }, { "grad_norm": 0.4410790503025055, "learning_rate": 9.212934933587094e-05, "loss": 0.0173, "step": 11100 }, { "grad_norm": 0.30447620153427124, "learning_rate": 9.211153025957056e-05, "loss": 0.0187, "step": 11110 }, { "grad_norm": 0.24550926685333252, "learning_rate": 9.209369276225614e-05, "loss": 0.0181, "step": 11120 }, { "grad_norm": 0.36070317029953003, "learning_rate": 9.207583685173042e-05, "loss": 0.0195, "step": 11130 }, { "grad_norm": 0.2505704462528229, "learning_rate": 9.205796253580417e-05, "loss": 0.0218, "step": 11140 }, { "grad_norm": 0.695863664150238, "learning_rate": 9.204006982229621e-05, "loss": 0.0243, "step": 11150 }, { "grad_norm": 0.22999463975429535, "learning_rate": 9.202215871903346e-05, "loss": 0.0174, "step": 11160 }, { "grad_norm": 0.7178937792778015, "learning_rate": 9.20042292338508e-05, "loss": 0.0171, "step": 11170 }, { "grad_norm": 0.498617947101593, "learning_rate": 9.198628137459123e-05, "loss": 0.0206, "step": 11180 }, { "grad_norm": 0.27631381154060364, "learning_rate": 9.196831514910572e-05, "loss": 0.0186, "step": 11190 }, { "grad_norm": 0.40561261773109436, "learning_rate": 9.195033056525332e-05, "loss": 0.0196, "step": 11200 }, { "grad_norm": 0.28640738129615784, "learning_rate": 9.193232763090114e-05, "loss": 0.022, "step": 11210 }, { "grad_norm": 0.30384206771850586, "learning_rate": 9.191430635392422e-05, "loss": 0.0197, "step": 11220 }, { "grad_norm": 0.8911752700805664, "learning_rate": 9.18962667422057e-05, "loss": 0.0178, "step": 11230 }, { "grad_norm": 0.24012470245361328, "learning_rate": 9.187820880363671e-05, "loss": 0.0179, "step": 11240 }, { "grad_norm": 0.1924658715724945, "learning_rate": 9.186013254611643e-05, "loss": 0.0172, "step": 11250 }, { "grad_norm": 0.3115248680114746, "learning_rate": 9.1842037977552e-05, "loss": 0.0193, "step": 11260 }, { "grad_norm": 0.2120833843946457, "learning_rate": 9.182392510585862e-05, "loss": 0.0217, "step": 11270 }, { "grad_norm": 0.22704294323921204, "learning_rate": 9.180579393895946e-05, "loss": 0.0189, "step": 11280 }, { "grad_norm": 0.3920730948448181, "learning_rate": 9.178764448478572e-05, "loss": 0.0173, "step": 11290 }, { "grad_norm": 0.38372984528541565, "learning_rate": 9.176947675127658e-05, "loss": 0.0197, "step": 11300 }, { "grad_norm": 0.3175940215587616, "learning_rate": 9.175129074637924e-05, "loss": 0.0198, "step": 11310 }, { "grad_norm": 0.34070655703544617, "learning_rate": 9.173308647804884e-05, "loss": 0.0186, "step": 11320 }, { "grad_norm": 0.24365554749965668, "learning_rate": 9.171486395424859e-05, "loss": 0.0191, "step": 11330 }, { "grad_norm": 0.2758241593837738, "learning_rate": 9.16966231829496e-05, "loss": 0.0203, "step": 11340 }, { "grad_norm": 0.530428946018219, "learning_rate": 9.167836417213105e-05, "loss": 0.0196, "step": 11350 }, { "grad_norm": 0.18636146187782288, "learning_rate": 9.166008692978001e-05, "loss": 0.0155, "step": 11360 }, { "grad_norm": 0.32198426127433777, "learning_rate": 9.164179146389158e-05, "loss": 0.017, "step": 11370 }, { "grad_norm": 0.2712686359882355, "learning_rate": 9.162347778246882e-05, "loss": 0.0197, "step": 11380 }, { "grad_norm": 0.3234510123729706, "learning_rate": 9.160514589352276e-05, "loss": 0.0176, "step": 11390 }, { "grad_norm": 0.3166946470737457, "learning_rate": 9.15867958050724e-05, "loss": 0.0184, "step": 11400 }, { "grad_norm": 0.24377121031284332, "learning_rate": 9.156842752514466e-05, "loss": 0.0191, "step": 11410 }, { "grad_norm": 0.2824077606201172, "learning_rate": 9.155004106177447e-05, "loss": 0.0213, "step": 11420 }, { "grad_norm": 0.39952412247657776, "learning_rate": 9.153163642300471e-05, "loss": 0.0223, "step": 11430 }, { "grad_norm": 0.2462097555398941, "learning_rate": 9.151321361688616e-05, "loss": 0.0212, "step": 11440 }, { "grad_norm": 0.20095816254615784, "learning_rate": 9.149477265147762e-05, "loss": 0.0174, "step": 11450 }, { "grad_norm": 0.2541472911834717, "learning_rate": 9.147631353484574e-05, "loss": 0.0193, "step": 11460 }, { "grad_norm": 0.17514576017856598, "learning_rate": 9.145783627506522e-05, "loss": 0.0189, "step": 11470 }, { "grad_norm": 0.26878446340560913, "learning_rate": 9.143934088021861e-05, "loss": 0.0193, "step": 11480 }, { "grad_norm": 0.25425246357917786, "learning_rate": 9.142082735839645e-05, "loss": 0.0195, "step": 11490 }, { "grad_norm": 0.22442354261875153, "learning_rate": 9.140229571769715e-05, "loss": 0.0214, "step": 11500 }, { "grad_norm": 0.35787639021873474, "learning_rate": 9.138374596622709e-05, "loss": 0.0222, "step": 11510 }, { "grad_norm": 0.31544533371925354, "learning_rate": 9.136517811210059e-05, "loss": 0.0198, "step": 11520 }, { "grad_norm": 0.6875748634338379, "learning_rate": 9.134659216343984e-05, "loss": 0.0197, "step": 11530 }, { "grad_norm": 0.36981165409088135, "learning_rate": 9.132798812837494e-05, "loss": 0.0207, "step": 11540 }, { "grad_norm": 0.3034549057483673, "learning_rate": 9.130936601504396e-05, "loss": 0.0173, "step": 11550 }, { "grad_norm": 0.2657410204410553, "learning_rate": 9.129072583159284e-05, "loss": 0.0206, "step": 11560 }, { "grad_norm": 0.3788413405418396, "learning_rate": 9.127206758617542e-05, "loss": 0.0211, "step": 11570 }, { "grad_norm": 0.24700652062892914, "learning_rate": 9.125339128695346e-05, "loss": 0.0184, "step": 11580 }, { "grad_norm": 0.34084466099739075, "learning_rate": 9.123469694209659e-05, "loss": 0.017, "step": 11590 }, { "grad_norm": 0.40874573588371277, "learning_rate": 9.121598455978239e-05, "loss": 0.0159, "step": 11600 }, { "grad_norm": 0.21119143068790436, "learning_rate": 9.119725414819624e-05, "loss": 0.0183, "step": 11610 }, { "grad_norm": 0.2249925434589386, "learning_rate": 9.117850571553149e-05, "loss": 0.0229, "step": 11620 }, { "grad_norm": 0.2699151337146759, "learning_rate": 9.115973926998935e-05, "loss": 0.0181, "step": 11630 }, { "grad_norm": 0.30375778675079346, "learning_rate": 9.114095481977888e-05, "loss": 0.0166, "step": 11640 }, { "grad_norm": 0.2436141073703766, "learning_rate": 9.112215237311703e-05, "loss": 0.0171, "step": 11650 }, { "grad_norm": 0.2502248287200928, "learning_rate": 9.110333193822867e-05, "loss": 0.0176, "step": 11660 }, { "grad_norm": 0.3211677670478821, "learning_rate": 9.108449352334645e-05, "loss": 0.0151, "step": 11670 }, { "grad_norm": 0.18077018857002258, "learning_rate": 9.106563713671094e-05, "loss": 0.0157, "step": 11680 }, { "grad_norm": 0.3113831877708435, "learning_rate": 9.104676278657061e-05, "loss": 0.0197, "step": 11690 }, { "grad_norm": 0.26857706904411316, "learning_rate": 9.102787048118169e-05, "loss": 0.0211, "step": 11700 }, { "grad_norm": 0.3173653185367584, "learning_rate": 9.100896022880834e-05, "loss": 0.0195, "step": 11710 }, { "grad_norm": 0.24698224663734436, "learning_rate": 9.099003203772254e-05, "loss": 0.0182, "step": 11720 }, { "grad_norm": 0.2873210608959198, "learning_rate": 9.097108591620413e-05, "loss": 0.0169, "step": 11730 }, { "grad_norm": 0.22600260376930237, "learning_rate": 9.095212187254078e-05, "loss": 0.016, "step": 11740 }, { "grad_norm": 0.30411794781684875, "learning_rate": 9.093313991502801e-05, "loss": 0.0172, "step": 11750 }, { "grad_norm": 0.35713693499565125, "learning_rate": 9.091414005196917e-05, "loss": 0.0188, "step": 11760 }, { "grad_norm": 0.24420855939388275, "learning_rate": 9.089512229167545e-05, "loss": 0.0173, "step": 11770 }, { "grad_norm": 0.18916578590869904, "learning_rate": 9.087608664246587e-05, "loss": 0.0201, "step": 11780 }, { "grad_norm": 0.3989925682544708, "learning_rate": 9.085703311266727e-05, "loss": 0.0162, "step": 11790 }, { "grad_norm": 0.2693338692188263, "learning_rate": 9.083796171061429e-05, "loss": 0.0246, "step": 11800 }, { "grad_norm": 0.49343836307525635, "learning_rate": 9.081887244464941e-05, "loss": 0.0197, "step": 11810 }, { "grad_norm": 0.3507305681705475, "learning_rate": 9.079976532312297e-05, "loss": 0.0212, "step": 11820 }, { "grad_norm": 0.24274495244026184, "learning_rate": 9.078064035439301e-05, "loss": 0.0185, "step": 11830 }, { "grad_norm": 0.2872113883495331, "learning_rate": 9.07614975468255e-05, "loss": 0.0171, "step": 11840 }, { "grad_norm": 0.24715767800807953, "learning_rate": 9.074233690879412e-05, "loss": 0.0204, "step": 11850 }, { "grad_norm": 0.21991419792175293, "learning_rate": 9.072315844868038e-05, "loss": 0.0205, "step": 11860 }, { "grad_norm": 0.15792569518089294, "learning_rate": 9.07039621748736e-05, "loss": 0.0216, "step": 11870 }, { "grad_norm": 0.26500311493873596, "learning_rate": 9.06847480957709e-05, "loss": 0.0179, "step": 11880 }, { "grad_norm": 0.5001174211502075, "learning_rate": 9.066551621977713e-05, "loss": 0.0191, "step": 11890 }, { "grad_norm": 0.28008195757865906, "learning_rate": 9.064626655530501e-05, "loss": 0.0256, "step": 11900 }, { "grad_norm": 0.22663821280002594, "learning_rate": 9.062699911077497e-05, "loss": 0.0192, "step": 11910 }, { "grad_norm": 0.2632211446762085, "learning_rate": 9.060771389461524e-05, "loss": 0.0181, "step": 11920 }, { "grad_norm": 0.3565666973590851, "learning_rate": 9.058841091526187e-05, "loss": 0.0163, "step": 11930 }, { "grad_norm": 0.2454863339662552, "learning_rate": 9.056909018115858e-05, "loss": 0.0165, "step": 11940 }, { "grad_norm": 0.2912105619907379, "learning_rate": 9.054975170075697e-05, "loss": 0.0203, "step": 11950 }, { "grad_norm": 0.2914169132709503, "learning_rate": 9.053039548251631e-05, "loss": 0.0186, "step": 11960 }, { "grad_norm": 0.19912081956863403, "learning_rate": 9.051102153490368e-05, "loss": 0.0195, "step": 11970 }, { "grad_norm": 0.45361727476119995, "learning_rate": 9.04916298663939e-05, "loss": 0.0248, "step": 11980 }, { "grad_norm": 0.33830609917640686, "learning_rate": 9.047222048546955e-05, "loss": 0.0218, "step": 11990 }, { "grad_norm": 0.3436487317085266, "learning_rate": 9.045279340062097e-05, "loss": 0.0218, "step": 12000 }, { "grad_norm": 0.2578405737876892, "learning_rate": 9.043334862034618e-05, "loss": 0.0221, "step": 12010 }, { "grad_norm": 0.30038589239120483, "learning_rate": 9.041388615315102e-05, "loss": 0.0201, "step": 12020 }, { "grad_norm": 0.33531704545021057, "learning_rate": 9.039440600754905e-05, "loss": 0.022, "step": 12030 }, { "grad_norm": 0.3082336187362671, "learning_rate": 9.037490819206151e-05, "loss": 0.0223, "step": 12040 }, { "grad_norm": 0.3670685291290283, "learning_rate": 9.035539271521744e-05, "loss": 0.0171, "step": 12050 }, { "grad_norm": 3.1603450775146484, "learning_rate": 9.033585958555356e-05, "loss": 0.0186, "step": 12060 }, { "grad_norm": 0.22855091094970703, "learning_rate": 9.031630881161431e-05, "loss": 0.02, "step": 12070 }, { "grad_norm": 0.33157235383987427, "learning_rate": 9.029674040195186e-05, "loss": 0.0178, "step": 12080 }, { "grad_norm": 0.2333654910326004, "learning_rate": 9.027715436512613e-05, "loss": 0.0168, "step": 12090 }, { "grad_norm": 0.16693559288978577, "learning_rate": 9.02575507097047e-05, "loss": 0.017, "step": 12100 }, { "grad_norm": 0.25290584564208984, "learning_rate": 9.023792944426286e-05, "loss": 0.0171, "step": 12110 }, { "grad_norm": 0.30300942063331604, "learning_rate": 9.021829057738364e-05, "loss": 0.0216, "step": 12120 }, { "grad_norm": 0.4810311794281006, "learning_rate": 9.019863411765775e-05, "loss": 0.0182, "step": 12130 }, { "grad_norm": 0.33339378237724304, "learning_rate": 9.017896007368357e-05, "loss": 0.0135, "step": 12140 }, { "grad_norm": 0.4191296696662903, "learning_rate": 9.015926845406722e-05, "loss": 0.0154, "step": 12150 }, { "grad_norm": 0.24476754665374756, "learning_rate": 9.013955926742245e-05, "loss": 0.0185, "step": 12160 }, { "grad_norm": 0.30692484974861145, "learning_rate": 9.011983252237077e-05, "loss": 0.0189, "step": 12170 }, { "grad_norm": 0.7597121000289917, "learning_rate": 9.01000882275413e-05, "loss": 0.0181, "step": 12180 }, { "grad_norm": 0.18320541083812714, "learning_rate": 9.008032639157088e-05, "loss": 0.0155, "step": 12190 }, { "grad_norm": 0.3306463658809662, "learning_rate": 9.006054702310401e-05, "loss": 0.023, "step": 12200 }, { "grad_norm": 0.24078916013240814, "learning_rate": 9.004075013079283e-05, "loss": 0.0176, "step": 12210 }, { "grad_norm": 0.2869652807712555, "learning_rate": 9.00209357232972e-05, "loss": 0.0185, "step": 12220 }, { "grad_norm": 0.3116816282272339, "learning_rate": 9.000110380928461e-05, "loss": 0.0191, "step": 12230 }, { "grad_norm": 0.23735970258712769, "learning_rate": 8.998125439743021e-05, "loss": 0.0187, "step": 12240 }, { "grad_norm": 0.34359994530677795, "learning_rate": 8.996138749641682e-05, "loss": 0.0214, "step": 12250 }, { "grad_norm": 0.40484124422073364, "learning_rate": 8.994150311493488e-05, "loss": 0.0199, "step": 12260 }, { "grad_norm": 0.35147586464881897, "learning_rate": 8.992160126168247e-05, "loss": 0.0215, "step": 12270 }, { "grad_norm": 0.28303641080856323, "learning_rate": 8.99016819453654e-05, "loss": 0.019, "step": 12280 }, { "grad_norm": 0.5604887008666992, "learning_rate": 8.988174517469702e-05, "loss": 0.0175, "step": 12290 }, { "grad_norm": 0.28186044096946716, "learning_rate": 8.986179095839835e-05, "loss": 0.0187, "step": 12300 }, { "grad_norm": 0.27120956778526306, "learning_rate": 8.984181930519804e-05, "loss": 0.0182, "step": 12310 }, { "grad_norm": 0.31306836009025574, "learning_rate": 8.982183022383237e-05, "loss": 0.0159, "step": 12320 }, { "grad_norm": 0.21397994458675385, "learning_rate": 8.980182372304525e-05, "loss": 0.0185, "step": 12330 }, { "grad_norm": 0.30445465445518494, "learning_rate": 8.97817998115882e-05, "loss": 0.023, "step": 12340 }, { "grad_norm": 0.32514432072639465, "learning_rate": 8.976175849822038e-05, "loss": 0.0186, "step": 12350 }, { "grad_norm": 0.2578253149986267, "learning_rate": 8.97416997917085e-05, "loss": 0.0225, "step": 12360 }, { "grad_norm": 0.4122816026210785, "learning_rate": 8.972162370082695e-05, "loss": 0.0178, "step": 12370 }, { "grad_norm": 0.28473034501075745, "learning_rate": 8.97015302343577e-05, "loss": 0.0182, "step": 12380 }, { "grad_norm": 0.273588627576828, "learning_rate": 8.968141940109027e-05, "loss": 0.0165, "step": 12390 }, { "grad_norm": 0.3063497543334961, "learning_rate": 8.966129120982188e-05, "loss": 0.0184, "step": 12400 }, { "grad_norm": 0.257154643535614, "learning_rate": 8.964114566935724e-05, "loss": 0.0166, "step": 12410 }, { "grad_norm": 0.4417515695095062, "learning_rate": 8.962098278850871e-05, "loss": 0.0156, "step": 12420 }, { "grad_norm": 0.22075635194778442, "learning_rate": 8.960080257609622e-05, "loss": 0.0158, "step": 12430 }, { "grad_norm": 0.1986394226551056, "learning_rate": 8.95806050409473e-05, "loss": 0.0173, "step": 12440 }, { "grad_norm": 0.38554027676582336, "learning_rate": 8.9560390191897e-05, "loss": 0.0227, "step": 12450 }, { "grad_norm": 0.26793062686920166, "learning_rate": 8.9540158037788e-05, "loss": 0.0182, "step": 12460 }, { "grad_norm": 0.3155517876148224, "learning_rate": 8.951990858747054e-05, "loss": 0.0205, "step": 12470 }, { "grad_norm": 0.19566045701503754, "learning_rate": 8.94996418498024e-05, "loss": 0.0151, "step": 12480 }, { "grad_norm": 0.1879924237728119, "learning_rate": 8.947935783364896e-05, "loss": 0.0148, "step": 12490 }, { "grad_norm": 0.2307950109243393, "learning_rate": 8.945905654788311e-05, "loss": 0.0157, "step": 12500 }, { "grad_norm": 0.37804561853408813, "learning_rate": 8.943873800138535e-05, "loss": 0.0171, "step": 12510 }, { "grad_norm": 0.21226295828819275, "learning_rate": 8.94184022030437e-05, "loss": 0.0155, "step": 12520 }, { "grad_norm": 0.43141597509384155, "learning_rate": 8.939804916175372e-05, "loss": 0.0165, "step": 12530 }, { "grad_norm": 0.369667112827301, "learning_rate": 8.93776788864185e-05, "loss": 0.0159, "step": 12540 }, { "grad_norm": 0.24828292429447174, "learning_rate": 8.935729138594873e-05, "loss": 0.0167, "step": 12550 }, { "grad_norm": 0.18590660393238068, "learning_rate": 8.933688666926258e-05, "loss": 0.0143, "step": 12560 }, { "grad_norm": 0.20414933562278748, "learning_rate": 8.931646474528575e-05, "loss": 0.0168, "step": 12570 }, { "grad_norm": 0.29649996757507324, "learning_rate": 8.929602562295151e-05, "loss": 0.0159, "step": 12580 }, { "grad_norm": 0.24468320608139038, "learning_rate": 8.92755693112006e-05, "loss": 0.0158, "step": 12590 }, { "grad_norm": 0.2354905903339386, "learning_rate": 8.925509581898136e-05, "loss": 0.0163, "step": 12600 }, { "grad_norm": 0.2546369135379791, "learning_rate": 8.923460515524951e-05, "loss": 0.016, "step": 12610 }, { "grad_norm": 0.3022312819957733, "learning_rate": 8.921409732896842e-05, "loss": 0.0159, "step": 12620 }, { "grad_norm": 0.37122631072998047, "learning_rate": 8.919357234910887e-05, "loss": 0.0161, "step": 12630 }, { "grad_norm": 0.3364105522632599, "learning_rate": 8.917303022464923e-05, "loss": 0.0179, "step": 12640 }, { "grad_norm": 0.2578449547290802, "learning_rate": 8.915247096457531e-05, "loss": 0.018, "step": 12650 }, { "grad_norm": 0.16283205151557922, "learning_rate": 8.91318945778804e-05, "loss": 0.0151, "step": 12660 }, { "grad_norm": 0.23669902980327606, "learning_rate": 8.911130107356534e-05, "loss": 0.0144, "step": 12670 }, { "grad_norm": 0.2061392217874527, "learning_rate": 8.90906904606384e-05, "loss": 0.0151, "step": 12680 }, { "grad_norm": 0.22428737580776215, "learning_rate": 8.90700627481154e-05, "loss": 0.0173, "step": 12690 }, { "grad_norm": 0.19933205842971802, "learning_rate": 8.904941794501957e-05, "loss": 0.014, "step": 12700 }, { "grad_norm": 0.22328545153141022, "learning_rate": 8.902875606038166e-05, "loss": 0.0149, "step": 12710 }, { "grad_norm": 0.4333564043045044, "learning_rate": 8.900807710323989e-05, "loss": 0.0171, "step": 12720 }, { "grad_norm": 0.22397229075431824, "learning_rate": 8.898738108263993e-05, "loss": 0.0142, "step": 12730 }, { "grad_norm": 0.5831775069236755, "learning_rate": 8.896666800763491e-05, "loss": 0.016, "step": 12740 }, { "grad_norm": 0.21749348938465118, "learning_rate": 8.894593788728546e-05, "loss": 0.0173, "step": 12750 }, { "grad_norm": 1.0668689012527466, "learning_rate": 8.892519073065961e-05, "loss": 0.0185, "step": 12760 }, { "grad_norm": 0.2465914636850357, "learning_rate": 8.89044265468329e-05, "loss": 0.0185, "step": 12770 }, { "grad_norm": 0.2571518123149872, "learning_rate": 8.888364534488827e-05, "loss": 0.0184, "step": 12780 }, { "grad_norm": 0.42684221267700195, "learning_rate": 8.886284713391613e-05, "loss": 0.0177, "step": 12790 }, { "grad_norm": 0.2719726264476776, "learning_rate": 8.884203192301431e-05, "loss": 0.0169, "step": 12800 }, { "grad_norm": 0.2771877646446228, "learning_rate": 8.88211997212881e-05, "loss": 0.0175, "step": 12810 }, { "grad_norm": 0.21071532368659973, "learning_rate": 8.880035053785023e-05, "loss": 0.017, "step": 12820 }, { "grad_norm": 0.1816725879907608, "learning_rate": 8.877948438182083e-05, "loss": 0.0147, "step": 12830 }, { "grad_norm": 3.0101988315582275, "learning_rate": 8.875860126232745e-05, "loss": 0.0166, "step": 12840 }, { "grad_norm": 0.812060534954071, "learning_rate": 8.87377011885051e-05, "loss": 0.017, "step": 12850 }, { "grad_norm": 0.2952303886413574, "learning_rate": 8.871678416949617e-05, "loss": 0.0204, "step": 12860 }, { "grad_norm": 0.2689931094646454, "learning_rate": 8.869585021445046e-05, "loss": 0.0173, "step": 12870 }, { "grad_norm": 0.4891888499259949, "learning_rate": 8.867489933252521e-05, "loss": 0.0171, "step": 12880 }, { "grad_norm": 0.3020201325416565, "learning_rate": 8.865393153288504e-05, "loss": 0.0182, "step": 12890 }, { "grad_norm": 0.26558226346969604, "learning_rate": 8.8632946824702e-05, "loss": 0.016, "step": 12900 }, { "grad_norm": 0.29771173000335693, "learning_rate": 8.86119452171555e-05, "loss": 0.0156, "step": 12910 }, { "grad_norm": 0.4297119975090027, "learning_rate": 8.859092671943234e-05, "loss": 0.0184, "step": 12920 }, { "grad_norm": 0.2872193455696106, "learning_rate": 8.856989134072676e-05, "loss": 0.0181, "step": 12930 }, { "grad_norm": 0.44937947392463684, "learning_rate": 8.85488390902403e-05, "loss": 0.0178, "step": 12940 }, { "grad_norm": 0.23497338593006134, "learning_rate": 8.852776997718199e-05, "loss": 0.0168, "step": 12950 }, { "grad_norm": 1.22459077835083, "learning_rate": 8.850668401076812e-05, "loss": 0.0168, "step": 12960 }, { "grad_norm": 0.25458842515945435, "learning_rate": 8.848558120022246e-05, "loss": 0.0164, "step": 12970 }, { "grad_norm": 0.28635045886039734, "learning_rate": 8.846446155477603e-05, "loss": 0.0181, "step": 12980 }, { "grad_norm": 0.2481595128774643, "learning_rate": 8.844332508366735e-05, "loss": 0.0211, "step": 12990 }, { "grad_norm": 0.8747007250785828, "learning_rate": 8.84221717961422e-05, "loss": 0.0168, "step": 13000 }, { "grad_norm": 1.3228213787078857, "learning_rate": 8.840100170145374e-05, "loss": 0.0187, "step": 13010 }, { "grad_norm": 0.4374644160270691, "learning_rate": 8.837981480886249e-05, "loss": 0.017, "step": 13020 }, { "grad_norm": 0.21548883616924286, "learning_rate": 8.835861112763633e-05, "loss": 0.0171, "step": 13030 }, { "grad_norm": 0.39332127571105957, "learning_rate": 8.833739066705044e-05, "loss": 0.0197, "step": 13040 }, { "grad_norm": 0.2569355070590973, "learning_rate": 8.831615343638742e-05, "loss": 0.018, "step": 13050 }, { "grad_norm": 1.7694684267044067, "learning_rate": 8.829489944493711e-05, "loss": 0.0149, "step": 13060 }, { "grad_norm": 0.2653349041938782, "learning_rate": 8.827362870199675e-05, "loss": 0.0176, "step": 13070 }, { "grad_norm": 0.25563910603523254, "learning_rate": 8.825234121687089e-05, "loss": 0.0187, "step": 13080 }, { "grad_norm": 0.21384982764720917, "learning_rate": 8.823103699887139e-05, "loss": 0.0182, "step": 13090 }, { "grad_norm": 0.2247297465801239, "learning_rate": 8.820971605731745e-05, "loss": 0.0147, "step": 13100 }, { "grad_norm": 0.6651595830917358, "learning_rate": 8.818837840153556e-05, "loss": 0.0163, "step": 13110 }, { "grad_norm": 0.21247904002666473, "learning_rate": 8.816702404085952e-05, "loss": 0.0148, "step": 13120 }, { "grad_norm": 0.33464938402175903, "learning_rate": 8.814565298463048e-05, "loss": 0.0166, "step": 13130 }, { "grad_norm": 0.19604140520095825, "learning_rate": 8.812426524219688e-05, "loss": 0.017, "step": 13140 }, { "grad_norm": 0.24657350778579712, "learning_rate": 8.81028608229144e-05, "loss": 0.0179, "step": 13150 }, { "grad_norm": 0.2835327982902527, "learning_rate": 8.808143973614611e-05, "loss": 0.0164, "step": 13160 }, { "grad_norm": 0.2262912094593048, "learning_rate": 8.806000199126228e-05, "loss": 0.0191, "step": 13170 }, { "grad_norm": 0.2415517270565033, "learning_rate": 8.803854759764052e-05, "loss": 0.0163, "step": 13180 }, { "grad_norm": 0.3438054919242859, "learning_rate": 8.801707656466572e-05, "loss": 0.0192, "step": 13190 }, { "grad_norm": 0.20010779798030853, "learning_rate": 8.799558890173003e-05, "loss": 0.0169, "step": 13200 }, { "grad_norm": 0.2284431755542755, "learning_rate": 8.79740846182329e-05, "loss": 0.0172, "step": 13210 }, { "grad_norm": 0.20473407208919525, "learning_rate": 8.7952563723581e-05, "loss": 0.018, "step": 13220 }, { "grad_norm": 0.2917388081550598, "learning_rate": 8.793102622718834e-05, "loss": 0.0188, "step": 13230 }, { "grad_norm": 0.19651085138320923, "learning_rate": 8.790947213847613e-05, "loss": 0.0207, "step": 13240 }, { "grad_norm": 0.26535019278526306, "learning_rate": 8.788790146687286e-05, "loss": 0.0164, "step": 13250 }, { "grad_norm": 0.2563299536705017, "learning_rate": 8.786631422181429e-05, "loss": 0.0177, "step": 13260 }, { "grad_norm": 2.0024514198303223, "learning_rate": 8.78447104127434e-05, "loss": 0.0171, "step": 13270 }, { "grad_norm": 0.2838561534881592, "learning_rate": 8.782309004911042e-05, "loss": 0.0166, "step": 13280 }, { "grad_norm": 0.25455978512763977, "learning_rate": 8.780145314037286e-05, "loss": 0.0155, "step": 13290 }, { "grad_norm": 0.18592672049999237, "learning_rate": 8.777979969599542e-05, "loss": 0.017, "step": 13300 }, { "grad_norm": 0.3950732350349426, "learning_rate": 8.775812972545006e-05, "loss": 0.0164, "step": 13310 }, { "grad_norm": 0.25979363918304443, "learning_rate": 8.773644323821596e-05, "loss": 0.0174, "step": 13320 }, { "grad_norm": 0.2209811508655548, "learning_rate": 8.771474024377953e-05, "loss": 0.0166, "step": 13330 }, { "grad_norm": 0.2934938967227936, "learning_rate": 8.769302075163438e-05, "loss": 0.0152, "step": 13340 }, { "grad_norm": 0.20089754462242126, "learning_rate": 8.767128477128137e-05, "loss": 0.0138, "step": 13350 }, { "grad_norm": 0.23435324430465698, "learning_rate": 8.764953231222854e-05, "loss": 0.0166, "step": 13360 }, { "grad_norm": 0.25032371282577515, "learning_rate": 8.762776338399119e-05, "loss": 0.0156, "step": 13370 }, { "grad_norm": 0.2851236164569855, "learning_rate": 8.760597799609176e-05, "loss": 0.0148, "step": 13380 }, { "grad_norm": 0.5052134394645691, "learning_rate": 8.758417615805992e-05, "loss": 0.0137, "step": 13390 }, { "grad_norm": 0.34746456146240234, "learning_rate": 8.756235787943254e-05, "loss": 0.0153, "step": 13400 }, { "grad_norm": 3.7923336029052734, "learning_rate": 8.754052316975367e-05, "loss": 0.0169, "step": 13410 }, { "grad_norm": 0.23765864968299866, "learning_rate": 8.751867203857455e-05, "loss": 0.0158, "step": 13420 }, { "grad_norm": 0.2932737469673157, "learning_rate": 8.749680449545363e-05, "loss": 0.0177, "step": 13430 }, { "grad_norm": 0.27026304602622986, "learning_rate": 8.747492054995649e-05, "loss": 0.0226, "step": 13440 }, { "grad_norm": 0.1425122171640396, "learning_rate": 8.745302021165595e-05, "loss": 0.0202, "step": 13450 }, { "grad_norm": 0.2183162420988083, "learning_rate": 8.743110349013192e-05, "loss": 0.0178, "step": 13460 }, { "grad_norm": 0.19107504189014435, "learning_rate": 8.740917039497153e-05, "loss": 0.0217, "step": 13470 }, { "grad_norm": 0.22998477518558502, "learning_rate": 8.738722093576906e-05, "loss": 0.0177, "step": 13480 }, { "grad_norm": 0.26148590445518494, "learning_rate": 8.736525512212597e-05, "loss": 0.015, "step": 13490 }, { "grad_norm": 0.23491457104682922, "learning_rate": 8.734327296365084e-05, "loss": 0.0147, "step": 13500 }, { "grad_norm": 0.6390664577484131, "learning_rate": 8.732127446995939e-05, "loss": 0.0152, "step": 13510 }, { "grad_norm": 0.2236470729112625, "learning_rate": 8.729925965067454e-05, "loss": 0.0138, "step": 13520 }, { "grad_norm": 0.28517478704452515, "learning_rate": 8.72772285154263e-05, "loss": 0.0155, "step": 13530 }, { "grad_norm": 0.36742669343948364, "learning_rate": 8.725518107385187e-05, "loss": 0.0136, "step": 13540 }, { "grad_norm": 0.2602911591529846, "learning_rate": 8.72331173355955e-05, "loss": 0.0153, "step": 13550 }, { "grad_norm": 0.19445912539958954, "learning_rate": 8.721103731030867e-05, "loss": 0.0195, "step": 13560 }, { "grad_norm": 0.29112333059310913, "learning_rate": 8.718894100764989e-05, "loss": 0.0167, "step": 13570 }, { "grad_norm": 0.1926458179950714, "learning_rate": 8.716682843728485e-05, "loss": 0.0159, "step": 13580 }, { "grad_norm": 0.23891200125217438, "learning_rate": 8.714469960888634e-05, "loss": 0.019, "step": 13590 }, { "grad_norm": 0.2599126100540161, "learning_rate": 8.712255453213427e-05, "loss": 0.017, "step": 13600 }, { "grad_norm": 0.26973286271095276, "learning_rate": 8.710039321671563e-05, "loss": 0.0162, "step": 13610 }, { "grad_norm": 0.3210272192955017, "learning_rate": 8.707821567232456e-05, "loss": 0.0137, "step": 13620 }, { "grad_norm": 0.1997053176164627, "learning_rate": 8.705602190866225e-05, "loss": 0.018, "step": 13630 }, { "grad_norm": 0.3689800202846527, "learning_rate": 8.703381193543701e-05, "loss": 0.0165, "step": 13640 }, { "grad_norm": 0.30733442306518555, "learning_rate": 8.701158576236423e-05, "loss": 0.016, "step": 13650 }, { "grad_norm": 0.1996941715478897, "learning_rate": 8.69893433991664e-05, "loss": 0.0144, "step": 13660 }, { "grad_norm": 0.512679398059845, "learning_rate": 8.69670848555731e-05, "loss": 0.0159, "step": 13670 }, { "grad_norm": 0.26480400562286377, "learning_rate": 8.694481014132096e-05, "loss": 0.0146, "step": 13680 }, { "grad_norm": 0.27468734979629517, "learning_rate": 8.69225192661537e-05, "loss": 0.0159, "step": 13690 }, { "grad_norm": 0.4327924847602844, "learning_rate": 8.690021223982208e-05, "loss": 0.0141, "step": 13700 }, { "grad_norm": 0.5110850930213928, "learning_rate": 8.687788907208398e-05, "loss": 0.0175, "step": 13710 }, { "grad_norm": 0.4467664957046509, "learning_rate": 8.685554977270431e-05, "loss": 0.0187, "step": 13720 }, { "grad_norm": 0.3122709095478058, "learning_rate": 8.683319435145503e-05, "loss": 0.0162, "step": 13730 }, { "grad_norm": 0.30163082480430603, "learning_rate": 8.681082281811517e-05, "loss": 0.0169, "step": 13740 }, { "grad_norm": 0.2059383988380432, "learning_rate": 8.67884351824708e-05, "loss": 0.015, "step": 13750 }, { "grad_norm": 0.18613728880882263, "learning_rate": 8.676603145431501e-05, "loss": 0.0162, "step": 13760 }, { "grad_norm": 0.2706719636917114, "learning_rate": 8.674361164344799e-05, "loss": 0.0172, "step": 13770 }, { "grad_norm": 0.588044285774231, "learning_rate": 8.672117575967688e-05, "loss": 0.017, "step": 13780 }, { "grad_norm": 0.21504820883274078, "learning_rate": 8.669872381281595e-05, "loss": 0.0165, "step": 13790 }, { "grad_norm": 0.2183818221092224, "learning_rate": 8.667625581268639e-05, "loss": 0.0157, "step": 13800 }, { "grad_norm": 0.23879095911979675, "learning_rate": 8.665377176911651e-05, "loss": 0.0143, "step": 13810 }, { "grad_norm": 0.3287128806114197, "learning_rate": 8.663127169194159e-05, "loss": 0.0165, "step": 13820 }, { "grad_norm": 0.27966660261154175, "learning_rate": 8.660875559100389e-05, "loss": 0.0152, "step": 13830 }, { "grad_norm": 0.2970609962940216, "learning_rate": 8.658622347615274e-05, "loss": 0.0143, "step": 13840 }, { "grad_norm": 0.26034536957740784, "learning_rate": 8.656367535724448e-05, "loss": 0.0137, "step": 13850 }, { "grad_norm": 0.5653989911079407, "learning_rate": 8.65411112441424e-05, "loss": 0.0161, "step": 13860 }, { "grad_norm": 0.2138514369726181, "learning_rate": 8.651853114671679e-05, "loss": 0.0186, "step": 13870 }, { "grad_norm": 0.2682693600654602, "learning_rate": 8.649593507484499e-05, "loss": 0.0164, "step": 13880 }, { "grad_norm": 1.1546870470046997, "learning_rate": 8.647332303841126e-05, "loss": 0.0138, "step": 13890 }, { "grad_norm": 0.24963362514972687, "learning_rate": 8.645069504730689e-05, "loss": 0.0163, "step": 13900 }, { "grad_norm": 0.26529935002326965, "learning_rate": 8.64280511114301e-05, "loss": 0.0144, "step": 13910 }, { "grad_norm": 0.3891887068748474, "learning_rate": 8.640539124068617e-05, "loss": 0.0131, "step": 13920 }, { "grad_norm": 0.2650001347064972, "learning_rate": 8.638271544498727e-05, "loss": 0.0145, "step": 13930 }, { "grad_norm": 0.32188016176223755, "learning_rate": 8.636002373425257e-05, "loss": 0.0141, "step": 13940 }, { "grad_norm": 0.31203070282936096, "learning_rate": 8.633731611840817e-05, "loss": 0.0177, "step": 13950 }, { "grad_norm": 0.2147238850593567, "learning_rate": 8.631459260738717e-05, "loss": 0.0145, "step": 13960 }, { "grad_norm": 0.24997706711292267, "learning_rate": 8.62918532111296e-05, "loss": 0.0142, "step": 13970 }, { "grad_norm": 0.26796984672546387, "learning_rate": 8.626909793958248e-05, "loss": 0.0162, "step": 13980 }, { "grad_norm": 0.3352988064289093, "learning_rate": 8.624632680269969e-05, "loss": 0.0192, "step": 13990 }, { "grad_norm": 0.32552406191825867, "learning_rate": 8.622353981044212e-05, "loss": 0.0217, "step": 14000 }, { "grad_norm": 0.28546544909477234, "learning_rate": 8.620073697277757e-05, "loss": 0.0152, "step": 14010 }, { "grad_norm": 0.210813507437706, "learning_rate": 8.617791829968079e-05, "loss": 0.0166, "step": 14020 }, { "grad_norm": 0.20588700473308563, "learning_rate": 8.615508380113344e-05, "loss": 0.0144, "step": 14030 }, { "grad_norm": 0.30532148480415344, "learning_rate": 8.613223348712408e-05, "loss": 0.0162, "step": 14040 }, { "grad_norm": 0.2636193335056305, "learning_rate": 8.610936736764824e-05, "loss": 0.0164, "step": 14050 }, { "grad_norm": 0.22242724895477295, "learning_rate": 8.608648545270833e-05, "loss": 0.0152, "step": 14060 }, { "grad_norm": 0.6561760306358337, "learning_rate": 8.606358775231366e-05, "loss": 0.0157, "step": 14070 }, { "grad_norm": 0.27286091446876526, "learning_rate": 8.60406742764805e-05, "loss": 0.0141, "step": 14080 }, { "grad_norm": 0.2519565522670746, "learning_rate": 8.601774503523195e-05, "loss": 0.0152, "step": 14090 }, { "grad_norm": 0.20200775563716888, "learning_rate": 8.599480003859805e-05, "loss": 0.0152, "step": 14100 }, { "grad_norm": 0.3013986349105835, "learning_rate": 8.597183929661573e-05, "loss": 0.0145, "step": 14110 }, { "grad_norm": 0.3438897132873535, "learning_rate": 8.594886281932879e-05, "loss": 0.0164, "step": 14120 }, { "grad_norm": 0.4901350140571594, "learning_rate": 8.59258706167879e-05, "loss": 0.0159, "step": 14130 }, { "grad_norm": 1.015467882156372, "learning_rate": 8.590286269905068e-05, "loss": 0.0146, "step": 14140 }, { "grad_norm": 0.3490464687347412, "learning_rate": 8.587983907618154e-05, "loss": 0.0163, "step": 14150 }, { "grad_norm": 0.29932326078414917, "learning_rate": 8.585679975825178e-05, "loss": 0.0181, "step": 14160 }, { "grad_norm": 0.564354658126831, "learning_rate": 8.583374475533962e-05, "loss": 0.017, "step": 14170 }, { "grad_norm": 0.42130544781684875, "learning_rate": 8.581067407753009e-05, "loss": 0.0208, "step": 14180 }, { "grad_norm": 0.17321977019309998, "learning_rate": 8.578758773491507e-05, "loss": 0.0169, "step": 14190 }, { "grad_norm": 0.16946826875209808, "learning_rate": 8.576448573759332e-05, "loss": 0.0192, "step": 14200 }, { "grad_norm": 0.25037261843681335, "learning_rate": 8.574136809567044e-05, "loss": 0.015, "step": 14210 }, { "grad_norm": 0.2640696167945862, "learning_rate": 8.57182348192589e-05, "loss": 0.0141, "step": 14220 }, { "grad_norm": 0.25656646490097046, "learning_rate": 8.569508591847792e-05, "loss": 0.0162, "step": 14230 }, { "grad_norm": 0.37539976835250854, "learning_rate": 8.567192140345367e-05, "loss": 0.0173, "step": 14240 }, { "grad_norm": 1.2279260158538818, "learning_rate": 8.564874128431906e-05, "loss": 0.0168, "step": 14250 }, { "grad_norm": 0.3041207790374756, "learning_rate": 8.562554557121389e-05, "loss": 0.0153, "step": 14260 }, { "grad_norm": 0.22864951193332672, "learning_rate": 8.560233427428475e-05, "loss": 0.0162, "step": 14270 }, { "grad_norm": 0.26611679792404175, "learning_rate": 8.557910740368503e-05, "loss": 0.0152, "step": 14280 }, { "grad_norm": 0.27714893221855164, "learning_rate": 8.555586496957495e-05, "loss": 0.0165, "step": 14290 }, { "grad_norm": 0.46806225180625916, "learning_rate": 8.553260698212155e-05, "loss": 0.019, "step": 14300 }, { "grad_norm": 0.24236001074314117, "learning_rate": 8.550933345149868e-05, "loss": 0.0164, "step": 14310 }, { "grad_norm": 0.2561631500720978, "learning_rate": 8.548604438788696e-05, "loss": 0.0171, "step": 14320 }, { "grad_norm": 0.26481932401657104, "learning_rate": 8.546273980147383e-05, "loss": 0.0161, "step": 14330 }, { "grad_norm": 0.2167641520500183, "learning_rate": 8.543941970245348e-05, "loss": 0.0132, "step": 14340 }, { "grad_norm": 0.2720724046230316, "learning_rate": 8.541608410102693e-05, "loss": 0.0155, "step": 14350 }, { "grad_norm": 0.22843924164772034, "learning_rate": 8.539273300740195e-05, "loss": 0.0163, "step": 14360 }, { "grad_norm": 0.25120678544044495, "learning_rate": 8.536936643179313e-05, "loss": 0.0145, "step": 14370 }, { "grad_norm": 0.1924671083688736, "learning_rate": 8.534598438442179e-05, "loss": 0.015, "step": 14380 }, { "grad_norm": 0.2159418761730194, "learning_rate": 8.532258687551603e-05, "loss": 0.0162, "step": 14390 }, { "grad_norm": 0.3208827078342438, "learning_rate": 8.529917391531071e-05, "loss": 0.0146, "step": 14400 }, { "grad_norm": 0.24112196266651154, "learning_rate": 8.527574551404747e-05, "loss": 0.0152, "step": 14410 }, { "grad_norm": 0.2439950853586197, "learning_rate": 8.525230168197468e-05, "loss": 0.0161, "step": 14420 }, { "grad_norm": 0.34736165404319763, "learning_rate": 8.522884242934745e-05, "loss": 0.0142, "step": 14430 }, { "grad_norm": 0.2644268274307251, "learning_rate": 8.520536776642768e-05, "loss": 0.0134, "step": 14440 }, { "grad_norm": 0.45394498109817505, "learning_rate": 8.5181877703484e-05, "loss": 0.0164, "step": 14450 }, { "grad_norm": 0.26712939143180847, "learning_rate": 8.51583722507917e-05, "loss": 0.0151, "step": 14460 }, { "grad_norm": 0.22238142788410187, "learning_rate": 8.513485141863293e-05, "loss": 0.0158, "step": 14470 }, { "grad_norm": 0.18520036339759827, "learning_rate": 8.511131521729647e-05, "loss": 0.014, "step": 14480 }, { "grad_norm": 0.19444407522678375, "learning_rate": 8.508776365707787e-05, "loss": 0.0157, "step": 14490 }, { "grad_norm": 0.26748204231262207, "learning_rate": 8.506419674827934e-05, "loss": 0.014, "step": 14500 }, { "grad_norm": 0.314954936504364, "learning_rate": 8.50406145012099e-05, "loss": 0.0146, "step": 14510 }, { "grad_norm": 0.4542345702648163, "learning_rate": 8.501701692618519e-05, "loss": 0.0147, "step": 14520 }, { "grad_norm": 0.26666706800460815, "learning_rate": 8.499340403352761e-05, "loss": 0.0155, "step": 14530 }, { "grad_norm": 0.2742269039154053, "learning_rate": 8.496977583356623e-05, "loss": 0.017, "step": 14540 }, { "grad_norm": 0.236363023519516, "learning_rate": 8.494613233663684e-05, "loss": 0.0144, "step": 14550 }, { "grad_norm": 0.6403789520263672, "learning_rate": 8.492247355308189e-05, "loss": 0.0155, "step": 14560 }, { "grad_norm": 0.2572008967399597, "learning_rate": 8.489879949325056e-05, "loss": 0.0172, "step": 14570 }, { "grad_norm": 0.29427382349967957, "learning_rate": 8.487511016749868e-05, "loss": 0.0158, "step": 14580 }, { "grad_norm": 0.7201873064041138, "learning_rate": 8.485140558618874e-05, "loss": 0.0149, "step": 14590 }, { "grad_norm": 0.17938712239265442, "learning_rate": 8.482768575968995e-05, "loss": 0.0131, "step": 14600 }, { "grad_norm": 0.32430800795555115, "learning_rate": 8.480395069837818e-05, "loss": 0.0153, "step": 14610 }, { "grad_norm": 0.3177424967288971, "learning_rate": 8.478020041263595e-05, "loss": 0.016, "step": 14620 }, { "grad_norm": 0.22683826088905334, "learning_rate": 8.475643491285242e-05, "loss": 0.0156, "step": 14630 }, { "grad_norm": 1.6611593961715698, "learning_rate": 8.473265420942345e-05, "loss": 0.0154, "step": 14640 }, { "grad_norm": 0.2710971534252167, "learning_rate": 8.470885831275151e-05, "loss": 0.0141, "step": 14650 }, { "grad_norm": 0.41855013370513916, "learning_rate": 8.468504723324574e-05, "loss": 0.0187, "step": 14660 }, { "grad_norm": 0.25674232840538025, "learning_rate": 8.466122098132193e-05, "loss": 0.0156, "step": 14670 }, { "grad_norm": 0.24721424281597137, "learning_rate": 8.463737956740245e-05, "loss": 0.0164, "step": 14680 }, { "grad_norm": 0.2346373051404953, "learning_rate": 8.461352300191639e-05, "loss": 0.0145, "step": 14690 }, { "grad_norm": 0.24168945848941803, "learning_rate": 8.45896512952994e-05, "loss": 0.0148, "step": 14700 }, { "grad_norm": 1.0817463397979736, "learning_rate": 8.456576445799377e-05, "loss": 0.0128, "step": 14710 }, { "grad_norm": 0.21370740234851837, "learning_rate": 8.454186250044844e-05, "loss": 0.0158, "step": 14720 }, { "grad_norm": 0.370244562625885, "learning_rate": 8.451794543311892e-05, "loss": 0.0158, "step": 14730 }, { "grad_norm": 0.22015926241874695, "learning_rate": 8.449401326646736e-05, "loss": 0.0147, "step": 14740 }, { "grad_norm": 0.2713366746902466, "learning_rate": 8.447006601096248e-05, "loss": 0.0156, "step": 14750 }, { "grad_norm": 0.22527511417865753, "learning_rate": 8.444610367707964e-05, "loss": 0.0137, "step": 14760 }, { "grad_norm": 0.2482970803976059, "learning_rate": 8.442212627530078e-05, "loss": 0.0137, "step": 14770 }, { "grad_norm": 0.2523469626903534, "learning_rate": 8.439813381611441e-05, "loss": 0.0127, "step": 14780 }, { "grad_norm": 0.21975170075893402, "learning_rate": 8.437412631001567e-05, "loss": 0.0144, "step": 14790 }, { "grad_norm": 0.8437670469284058, "learning_rate": 8.435010376750626e-05, "loss": 0.0147, "step": 14800 }, { "grad_norm": 0.26741793751716614, "learning_rate": 8.432606619909442e-05, "loss": 0.0162, "step": 14810 }, { "grad_norm": 0.23208749294281006, "learning_rate": 8.430201361529506e-05, "loss": 0.0162, "step": 14820 }, { "grad_norm": 0.22325587272644043, "learning_rate": 8.427794602662954e-05, "loss": 0.0151, "step": 14830 }, { "grad_norm": 0.20991641283035278, "learning_rate": 8.425386344362586e-05, "loss": 0.016, "step": 14840 }, { "grad_norm": 0.2802136540412903, "learning_rate": 8.422976587681859e-05, "loss": 0.0144, "step": 14850 }, { "grad_norm": 0.3262398838996887, "learning_rate": 8.42056533367488e-05, "loss": 0.0148, "step": 14860 }, { "grad_norm": 0.648945689201355, "learning_rate": 8.41815258339641e-05, "loss": 0.0143, "step": 14870 }, { "grad_norm": 0.21587830781936646, "learning_rate": 8.415738337901874e-05, "loss": 0.0156, "step": 14880 }, { "grad_norm": 0.7933651804924011, "learning_rate": 8.413322598247342e-05, "loss": 0.0155, "step": 14890 }, { "grad_norm": 0.3324432671070099, "learning_rate": 8.41090536548954e-05, "loss": 0.0136, "step": 14900 }, { "grad_norm": 0.21906043589115143, "learning_rate": 8.408486640685849e-05, "loss": 0.0123, "step": 14910 }, { "grad_norm": 0.17746290564537048, "learning_rate": 8.4060664248943e-05, "loss": 0.0163, "step": 14920 }, { "grad_norm": 0.25670838356018066, "learning_rate": 8.40364471917358e-05, "loss": 0.0165, "step": 14930 }, { "grad_norm": 0.23095059394836426, "learning_rate": 8.401221524583024e-05, "loss": 0.0158, "step": 14940 }, { "grad_norm": 2.4265835285186768, "learning_rate": 8.398796842182619e-05, "loss": 0.0144, "step": 14950 }, { "grad_norm": 0.20330680906772614, "learning_rate": 8.396370673033006e-05, "loss": 0.014, "step": 14960 }, { "grad_norm": 0.20975932478904724, "learning_rate": 8.39394301819547e-05, "loss": 0.0187, "step": 14970 }, { "grad_norm": 1.9589660167694092, "learning_rate": 8.391513878731949e-05, "loss": 0.0131, "step": 14980 }, { "grad_norm": 0.26502153277397156, "learning_rate": 8.389083255705037e-05, "loss": 0.014, "step": 14990 }, { "grad_norm": 0.2522209584712982, "learning_rate": 8.386651150177968e-05, "loss": 0.0155, "step": 15000 }, { "grad_norm": 0.30301955342292786, "learning_rate": 8.384217563214627e-05, "loss": 0.0143, "step": 15010 }, { "grad_norm": 0.1934015154838562, "learning_rate": 8.381782495879546e-05, "loss": 0.016, "step": 15020 }, { "grad_norm": 0.18001475930213928, "learning_rate": 8.37934594923791e-05, "loss": 0.0125, "step": 15030 }, { "grad_norm": 0.172197163105011, "learning_rate": 8.376907924355546e-05, "loss": 0.0122, "step": 15040 }, { "grad_norm": 0.18958939611911774, "learning_rate": 8.374468422298926e-05, "loss": 0.0136, "step": 15050 }, { "grad_norm": 0.27531835436820984, "learning_rate": 8.372027444135176e-05, "loss": 0.0132, "step": 15060 }, { "grad_norm": 0.189304918050766, "learning_rate": 8.36958499093206e-05, "loss": 0.0123, "step": 15070 }, { "grad_norm": 0.19254307448863983, "learning_rate": 8.367141063757988e-05, "loss": 0.0149, "step": 15080 }, { "grad_norm": 0.22923047840595245, "learning_rate": 8.364695663682022e-05, "loss": 0.0123, "step": 15090 }, { "grad_norm": 0.2546452581882477, "learning_rate": 8.362248791773858e-05, "loss": 0.0134, "step": 15100 }, { "grad_norm": 0.25443100929260254, "learning_rate": 8.359800449103842e-05, "loss": 0.014, "step": 15110 }, { "grad_norm": 0.7621694803237915, "learning_rate": 8.357350636742967e-05, "loss": 0.0151, "step": 15120 }, { "grad_norm": 0.2605980932712555, "learning_rate": 8.354899355762858e-05, "loss": 0.0137, "step": 15130 }, { "grad_norm": 0.46013426780700684, "learning_rate": 8.352446607235791e-05, "loss": 0.0215, "step": 15140 }, { "grad_norm": 0.3051428496837616, "learning_rate": 8.349992392234683e-05, "loss": 0.016, "step": 15150 }, { "grad_norm": 0.21129076182842255, "learning_rate": 8.347536711833088e-05, "loss": 0.0152, "step": 15160 }, { "grad_norm": 0.26733607053756714, "learning_rate": 8.345079567105206e-05, "loss": 0.0149, "step": 15170 }, { "grad_norm": 0.22271153330802917, "learning_rate": 8.342620959125875e-05, "loss": 0.0149, "step": 15180 }, { "grad_norm": 0.2886846959590912, "learning_rate": 8.340160888970574e-05, "loss": 0.0159, "step": 15190 }, { "grad_norm": 0.2611342668533325, "learning_rate": 8.337699357715422e-05, "loss": 0.0159, "step": 15200 }, { "grad_norm": 0.26617926359176636, "learning_rate": 8.335236366437174e-05, "loss": 0.0161, "step": 15210 }, { "grad_norm": 0.39779889583587646, "learning_rate": 8.332771916213227e-05, "loss": 0.0167, "step": 15220 }, { "grad_norm": 0.2534269392490387, "learning_rate": 8.330306008121618e-05, "loss": 0.0155, "step": 15230 }, { "grad_norm": 0.18373052775859833, "learning_rate": 8.327838643241014e-05, "loss": 0.0163, "step": 15240 }, { "grad_norm": 0.21589304506778717, "learning_rate": 8.32536982265073e-05, "loss": 0.0132, "step": 15250 }, { "grad_norm": 0.21374471485614777, "learning_rate": 8.32289954743071e-05, "loss": 0.0144, "step": 15260 }, { "grad_norm": 0.31849199533462524, "learning_rate": 8.320427818661532e-05, "loss": 0.014, "step": 15270 }, { "grad_norm": 0.3398205637931824, "learning_rate": 8.317954637424419e-05, "loss": 0.0154, "step": 15280 }, { "grad_norm": 0.26796165108680725, "learning_rate": 8.315480004801221e-05, "loss": 0.0137, "step": 15290 }, { "grad_norm": 0.2543047368526459, "learning_rate": 8.31300392187443e-05, "loss": 0.014, "step": 15300 }, { "grad_norm": 0.16873469948768616, "learning_rate": 8.31052638972717e-05, "loss": 0.0119, "step": 15310 }, { "grad_norm": 0.28705886006355286, "learning_rate": 8.30804740944319e-05, "loss": 0.0155, "step": 15320 }, { "grad_norm": 0.6057817339897156, "learning_rate": 8.305566982106887e-05, "loss": 0.014, "step": 15330 }, { "grad_norm": 0.2520909011363983, "learning_rate": 8.303085108803282e-05, "loss": 0.0146, "step": 15340 }, { "grad_norm": 0.23942095041275024, "learning_rate": 8.300601790618032e-05, "loss": 0.0157, "step": 15350 }, { "grad_norm": 0.30642321705818176, "learning_rate": 8.298117028637423e-05, "loss": 0.0148, "step": 15360 }, { "grad_norm": 0.7142057418823242, "learning_rate": 8.295630823948375e-05, "loss": 0.0171, "step": 15370 }, { "grad_norm": 0.25708070397377014, "learning_rate": 8.293143177638438e-05, "loss": 0.017, "step": 15380 }, { "grad_norm": 0.4836636185646057, "learning_rate": 8.290654090795795e-05, "loss": 0.0147, "step": 15390 }, { "grad_norm": 0.2788259983062744, "learning_rate": 8.288163564509253e-05, "loss": 0.0161, "step": 15400 }, { "grad_norm": 0.23238541185855865, "learning_rate": 8.285671599868257e-05, "loss": 0.0154, "step": 15410 }, { "grad_norm": 0.21630816161632538, "learning_rate": 8.283178197962875e-05, "loss": 0.0131, "step": 15420 }, { "grad_norm": 0.21022281050682068, "learning_rate": 8.280683359883808e-05, "loss": 0.0153, "step": 15430 }, { "grad_norm": 0.3239908218383789, "learning_rate": 8.278187086722377e-05, "loss": 0.0169, "step": 15440 }, { "grad_norm": 0.3130163252353668, "learning_rate": 8.275689379570544e-05, "loss": 0.0131, "step": 15450 }, { "grad_norm": 0.244358092546463, "learning_rate": 8.273190239520887e-05, "loss": 0.0153, "step": 15460 }, { "grad_norm": 0.29825982451438904, "learning_rate": 8.270689667666616e-05, "loss": 0.0149, "step": 15470 }, { "grad_norm": 0.15614750981330872, "learning_rate": 8.268187665101563e-05, "loss": 0.015, "step": 15480 }, { "grad_norm": 0.22576303780078888, "learning_rate": 8.265684232920194e-05, "loss": 0.0195, "step": 15490 }, { "grad_norm": 0.39342328906059265, "learning_rate": 8.263179372217592e-05, "loss": 0.0183, "step": 15500 }, { "grad_norm": 0.28408318758010864, "learning_rate": 8.260673084089468e-05, "loss": 0.0172, "step": 15510 }, { "grad_norm": 0.23279941082000732, "learning_rate": 8.25816536963216e-05, "loss": 0.0166, "step": 15520 }, { "grad_norm": 0.20681865513324738, "learning_rate": 8.255656229942626e-05, "loss": 0.0158, "step": 15530 }, { "grad_norm": 0.27353304624557495, "learning_rate": 8.253145666118447e-05, "loss": 0.0149, "step": 15540 }, { "grad_norm": 0.17273548245429993, "learning_rate": 8.250633679257834e-05, "loss": 0.0145, "step": 15550 }, { "grad_norm": 0.22548295557498932, "learning_rate": 8.24812027045961e-05, "loss": 0.0122, "step": 15560 }, { "grad_norm": 0.27944058179855347, "learning_rate": 8.245605440823226e-05, "loss": 0.013, "step": 15570 }, { "grad_norm": 0.3386555314064026, "learning_rate": 8.243089191448759e-05, "loss": 0.0149, "step": 15580 }, { "grad_norm": 0.22927449643611908, "learning_rate": 8.240571523436896e-05, "loss": 0.0151, "step": 15590 }, { "grad_norm": 0.18811948597431183, "learning_rate": 8.238052437888955e-05, "loss": 0.0153, "step": 15600 }, { "grad_norm": 0.2538744807243347, "learning_rate": 8.235531935906866e-05, "loss": 0.0146, "step": 15610 }, { "grad_norm": 0.16339746117591858, "learning_rate": 8.233010018593185e-05, "loss": 0.0134, "step": 15620 }, { "grad_norm": 0.35437628626823425, "learning_rate": 8.230486687051082e-05, "loss": 0.0147, "step": 15630 }, { "grad_norm": 0.22945618629455566, "learning_rate": 8.22796194238435e-05, "loss": 0.0144, "step": 15640 }, { "grad_norm": 2.3840491771698, "learning_rate": 8.225435785697397e-05, "loss": 0.0164, "step": 15650 }, { "grad_norm": 0.30322861671447754, "learning_rate": 8.222908218095252e-05, "loss": 0.0178, "step": 15660 }, { "grad_norm": 0.15330541133880615, "learning_rate": 8.220379240683553e-05, "loss": 0.0155, "step": 15670 }, { "grad_norm": 0.18439507484436035, "learning_rate": 8.217848854568566e-05, "loss": 0.0163, "step": 15680 }, { "grad_norm": 0.8677042126655579, "learning_rate": 8.215317060857167e-05, "loss": 0.0135, "step": 15690 }, { "grad_norm": 0.2688998878002167, "learning_rate": 8.212783860656847e-05, "loss": 0.0152, "step": 15700 }, { "grad_norm": 0.22155793011188507, "learning_rate": 8.210249255075717e-05, "loss": 0.0133, "step": 15710 }, { "grad_norm": 0.19878748059272766, "learning_rate": 8.207713245222496e-05, "loss": 0.014, "step": 15720 }, { "grad_norm": 0.485564261674881, "learning_rate": 8.205175832206522e-05, "loss": 0.0162, "step": 15730 }, { "grad_norm": 0.27191802859306335, "learning_rate": 8.202637017137747e-05, "loss": 0.0144, "step": 15740 }, { "grad_norm": 0.159694105386734, "learning_rate": 8.200096801126734e-05, "loss": 0.0151, "step": 15750 }, { "grad_norm": 0.2702558934688568, "learning_rate": 8.197555185284658e-05, "loss": 0.017, "step": 15760 }, { "grad_norm": 0.3195769190788269, "learning_rate": 8.19501217072331e-05, "loss": 0.017, "step": 15770 }, { "grad_norm": 0.20758956670761108, "learning_rate": 8.192467758555091e-05, "loss": 0.0157, "step": 15780 }, { "grad_norm": 0.6951310634613037, "learning_rate": 8.189921949893014e-05, "loss": 0.0147, "step": 15790 }, { "grad_norm": 0.6190140843391418, "learning_rate": 8.187374745850702e-05, "loss": 0.017, "step": 15800 }, { "grad_norm": 0.38556501269340515, "learning_rate": 8.184826147542387e-05, "loss": 0.0152, "step": 15810 }, { "grad_norm": 0.26138952374458313, "learning_rate": 8.18227615608291e-05, "loss": 0.0191, "step": 15820 }, { "grad_norm": 0.20185944437980652, "learning_rate": 8.179724772587732e-05, "loss": 0.0132, "step": 15830 }, { "grad_norm": 0.18494455516338348, "learning_rate": 8.177171998172908e-05, "loss": 0.0151, "step": 15840 }, { "grad_norm": 2.570355176925659, "learning_rate": 8.174617833955112e-05, "loss": 0.0152, "step": 15850 }, { "grad_norm": 0.23446570336818695, "learning_rate": 8.172062281051618e-05, "loss": 0.0172, "step": 15860 }, { "grad_norm": 0.22117318212985992, "learning_rate": 8.169505340580316e-05, "loss": 0.0151, "step": 15870 }, { "grad_norm": 0.19939295947551727, "learning_rate": 8.166947013659696e-05, "loss": 0.0176, "step": 15880 }, { "grad_norm": 0.3673568665981293, "learning_rate": 8.164387301408858e-05, "loss": 0.0138, "step": 15890 }, { "grad_norm": 0.3084507882595062, "learning_rate": 8.161826204947507e-05, "loss": 0.0154, "step": 15900 }, { "grad_norm": 0.19308854639530182, "learning_rate": 8.159263725395954e-05, "loss": 0.0152, "step": 15910 }, { "grad_norm": 0.2948331832885742, "learning_rate": 8.156699863875115e-05, "loss": 0.0195, "step": 15920 }, { "grad_norm": 0.2642195224761963, "learning_rate": 8.154134621506508e-05, "loss": 0.0148, "step": 15930 }, { "grad_norm": 0.21338285505771637, "learning_rate": 8.15156799941226e-05, "loss": 0.0138, "step": 15940 }, { "grad_norm": 0.21785786747932434, "learning_rate": 8.148999998715096e-05, "loss": 0.0165, "step": 15950 }, { "grad_norm": 0.18063075840473175, "learning_rate": 8.146430620538348e-05, "loss": 0.0133, "step": 15960 }, { "grad_norm": 2.5658085346221924, "learning_rate": 8.143859866005949e-05, "loss": 0.016, "step": 15970 }, { "grad_norm": 0.4401431083679199, "learning_rate": 8.141287736242438e-05, "loss": 0.0146, "step": 15980 }, { "grad_norm": 1.3378782272338867, "learning_rate": 8.138714232372945e-05, "loss": 0.0144, "step": 15990 }, { "grad_norm": 0.3800533413887024, "learning_rate": 8.136139355523213e-05, "loss": 0.0165, "step": 16000 }, { "grad_norm": 0.9126681685447693, "learning_rate": 8.13356310681958e-05, "loss": 0.0165, "step": 16010 }, { "grad_norm": 0.145780548453331, "learning_rate": 8.130985487388982e-05, "loss": 0.0167, "step": 16020 }, { "grad_norm": 1.6188442707061768, "learning_rate": 8.12840649835896e-05, "loss": 0.0133, "step": 16030 }, { "grad_norm": 0.25000497698783875, "learning_rate": 8.125826140857653e-05, "loss": 0.0127, "step": 16040 }, { "grad_norm": 0.18671564757823944, "learning_rate": 8.123244416013792e-05, "loss": 0.0135, "step": 16050 }, { "grad_norm": 0.4375709295272827, "learning_rate": 8.120661324956716e-05, "loss": 0.016, "step": 16060 }, { "grad_norm": 0.35133257508277893, "learning_rate": 8.118076868816353e-05, "loss": 0.0135, "step": 16070 }, { "grad_norm": 0.2773558795452118, "learning_rate": 8.115491048723234e-05, "loss": 0.0134, "step": 16080 }, { "grad_norm": 0.23076681792736053, "learning_rate": 8.112903865808486e-05, "loss": 0.0128, "step": 16090 }, { "grad_norm": 0.21265143156051636, "learning_rate": 8.110315321203826e-05, "loss": 0.0135, "step": 16100 }, { "grad_norm": 0.25703659653663635, "learning_rate": 8.107725416041575e-05, "loss": 0.0147, "step": 16110 }, { "grad_norm": 0.2694800794124603, "learning_rate": 8.105134151454645e-05, "loss": 0.0149, "step": 16120 }, { "grad_norm": 0.26382219791412354, "learning_rate": 8.102541528576544e-05, "loss": 0.0145, "step": 16130 }, { "grad_norm": 0.4188840985298157, "learning_rate": 8.099947548541371e-05, "loss": 0.0128, "step": 16140 }, { "grad_norm": 0.43424659967422485, "learning_rate": 8.097352212483824e-05, "loss": 0.0139, "step": 16150 }, { "grad_norm": 0.5251836776733398, "learning_rate": 8.094755521539187e-05, "loss": 0.0147, "step": 16160 }, { "grad_norm": 0.205794095993042, "learning_rate": 8.092157476843346e-05, "loss": 0.013, "step": 16170 }, { "grad_norm": 0.275945782661438, "learning_rate": 8.089558079532769e-05, "loss": 0.0143, "step": 16180 }, { "grad_norm": 0.26324063539505005, "learning_rate": 8.086957330744522e-05, "loss": 0.0126, "step": 16190 }, { "grad_norm": 0.21128211915493011, "learning_rate": 8.084355231616265e-05, "loss": 0.0135, "step": 16200 }, { "grad_norm": 0.2565315067768097, "learning_rate": 8.08175178328624e-05, "loss": 0.0124, "step": 16210 }, { "grad_norm": 0.3252500593662262, "learning_rate": 8.079146986893284e-05, "loss": 0.0134, "step": 16220 }, { "grad_norm": 0.3654058575630188, "learning_rate": 8.076540843576826e-05, "loss": 0.0132, "step": 16230 }, { "grad_norm": 0.2385328859090805, "learning_rate": 8.073933354476879e-05, "loss": 0.0138, "step": 16240 }, { "grad_norm": 0.30839869379997253, "learning_rate": 8.07132452073405e-05, "loss": 0.0131, "step": 16250 }, { "grad_norm": 0.29536476731300354, "learning_rate": 8.068714343489529e-05, "loss": 0.0192, "step": 16260 }, { "grad_norm": 0.3682538568973541, "learning_rate": 8.066102823885097e-05, "loss": 0.016, "step": 16270 }, { "grad_norm": 0.2414243221282959, "learning_rate": 8.063489963063125e-05, "loss": 0.018, "step": 16280 }, { "grad_norm": 0.8114544153213501, "learning_rate": 8.06087576216656e-05, "loss": 0.0144, "step": 16290 }, { "grad_norm": 0.26610320806503296, "learning_rate": 8.058260222338946e-05, "loss": 0.015, "step": 16300 }, { "grad_norm": 0.16240692138671875, "learning_rate": 8.05564334472441e-05, "loss": 0.0149, "step": 16310 }, { "grad_norm": 0.1890617311000824, "learning_rate": 8.053025130467664e-05, "loss": 0.0147, "step": 16320 }, { "grad_norm": 0.19430309534072876, "learning_rate": 8.050405580714e-05, "loss": 0.0124, "step": 16330 }, { "grad_norm": 0.1766025424003601, "learning_rate": 8.047784696609301e-05, "loss": 0.0122, "step": 16340 }, { "grad_norm": 0.27392256259918213, "learning_rate": 8.045162479300029e-05, "loss": 0.0177, "step": 16350 }, { "grad_norm": 0.2204681783914566, "learning_rate": 8.042538929933232e-05, "loss": 0.0172, "step": 16360 }, { "grad_norm": 0.18500956892967224, "learning_rate": 8.039914049656539e-05, "loss": 0.0137, "step": 16370 }, { "grad_norm": 0.23197679221630096, "learning_rate": 8.037287839618163e-05, "loss": 0.0141, "step": 16380 }, { "grad_norm": 0.2531525194644928, "learning_rate": 8.034660300966898e-05, "loss": 0.0129, "step": 16390 }, { "grad_norm": 0.23026536405086517, "learning_rate": 8.032031434852115e-05, "loss": 0.0141, "step": 16400 }, { "grad_norm": 0.24281765520572662, "learning_rate": 8.029401242423773e-05, "loss": 0.0148, "step": 16410 }, { "grad_norm": 0.18256793916225433, "learning_rate": 8.026769724832406e-05, "loss": 0.0157, "step": 16420 }, { "grad_norm": 0.2309461236000061, "learning_rate": 8.02413688322913e-05, "loss": 0.0136, "step": 16430 }, { "grad_norm": 0.22398611903190613, "learning_rate": 8.021502718765639e-05, "loss": 0.013, "step": 16440 }, { "grad_norm": 0.24732019007205963, "learning_rate": 8.018867232594204e-05, "loss": 0.0126, "step": 16450 }, { "grad_norm": 0.2690795063972473, "learning_rate": 8.016230425867679e-05, "loss": 0.0149, "step": 16460 }, { "grad_norm": 0.2940412759780884, "learning_rate": 8.013592299739494e-05, "loss": 0.0118, "step": 16470 }, { "grad_norm": 0.17929218709468842, "learning_rate": 8.01095285536365e-05, "loss": 0.0145, "step": 16480 }, { "grad_norm": 0.2195579558610916, "learning_rate": 8.008312093894735e-05, "loss": 0.0121, "step": 16490 }, { "grad_norm": 0.27761560678482056, "learning_rate": 8.005670016487903e-05, "loss": 0.014, "step": 16500 }, { "grad_norm": 0.47778162360191345, "learning_rate": 8.003026624298892e-05, "loss": 0.0143, "step": 16510 }, { "grad_norm": 0.24035248160362244, "learning_rate": 8.00038191848401e-05, "loss": 0.0151, "step": 16520 }, { "grad_norm": 0.1772141307592392, "learning_rate": 7.997735900200142e-05, "loss": 0.0145, "step": 16530 }, { "grad_norm": 0.3267374634742737, "learning_rate": 7.995088570604745e-05, "loss": 0.0127, "step": 16540 }, { "grad_norm": 0.12948235869407654, "learning_rate": 7.992439930855852e-05, "loss": 0.0139, "step": 16550 }, { "grad_norm": 0.20682181417942047, "learning_rate": 7.989789982112067e-05, "loss": 0.0141, "step": 16560 }, { "grad_norm": 0.2904006838798523, "learning_rate": 7.987138725532572e-05, "loss": 0.0167, "step": 16570 }, { "grad_norm": 0.20261622965335846, "learning_rate": 7.98448616227711e-05, "loss": 0.0125, "step": 16580 }, { "grad_norm": 0.17545409500598907, "learning_rate": 7.981832293506004e-05, "loss": 0.0126, "step": 16590 }, { "grad_norm": 0.7193784713745117, "learning_rate": 7.97917712038015e-05, "loss": 0.0144, "step": 16600 }, { "grad_norm": 0.1985563039779663, "learning_rate": 7.976520644061007e-05, "loss": 0.0147, "step": 16610 }, { "grad_norm": 0.21790193021297455, "learning_rate": 7.973862865710611e-05, "loss": 0.0151, "step": 16620 }, { "grad_norm": 0.16051943600177765, "learning_rate": 7.971203786491563e-05, "loss": 0.0176, "step": 16630 }, { "grad_norm": 0.23623022437095642, "learning_rate": 7.968543407567034e-05, "loss": 0.0157, "step": 16640 }, { "grad_norm": 0.29690787196159363, "learning_rate": 7.965881730100765e-05, "loss": 0.0184, "step": 16650 }, { "grad_norm": 0.2927573025226593, "learning_rate": 7.963218755257064e-05, "loss": 0.0119, "step": 16660 }, { "grad_norm": 0.2614576816558838, "learning_rate": 7.960554484200807e-05, "loss": 0.0176, "step": 16670 }, { "grad_norm": 1.7634576559066772, "learning_rate": 7.957888918097436e-05, "loss": 0.0136, "step": 16680 }, { "grad_norm": 0.18837422132492065, "learning_rate": 7.955222058112962e-05, "loss": 0.0119, "step": 16690 }, { "grad_norm": 0.1433788239955902, "learning_rate": 7.952553905413955e-05, "loss": 0.0112, "step": 16700 }, { "grad_norm": 0.2061106562614441, "learning_rate": 7.949884461167563e-05, "loss": 0.0145, "step": 16710 }, { "grad_norm": 0.14836804568767548, "learning_rate": 7.947213726541484e-05, "loss": 0.0112, "step": 16720 }, { "grad_norm": 0.2949449419975281, "learning_rate": 7.944541702703996e-05, "loss": 0.0147, "step": 16730 }, { "grad_norm": 0.21550557017326355, "learning_rate": 7.941868390823928e-05, "loss": 0.0161, "step": 16740 }, { "grad_norm": 0.26455366611480713, "learning_rate": 7.939193792070677e-05, "loss": 0.0137, "step": 16750 }, { "grad_norm": 0.2873455882072449, "learning_rate": 7.936517907614207e-05, "loss": 0.0141, "step": 16760 }, { "grad_norm": 0.20563481748104095, "learning_rate": 7.933840738625036e-05, "loss": 0.0152, "step": 16770 }, { "grad_norm": 0.5943930745124817, "learning_rate": 7.931162286274252e-05, "loss": 0.0139, "step": 16780 }, { "grad_norm": 0.21151027083396912, "learning_rate": 7.928482551733503e-05, "loss": 0.0131, "step": 16790 }, { "grad_norm": 0.5535314679145813, "learning_rate": 7.925801536174991e-05, "loss": 0.0149, "step": 16800 }, { "grad_norm": 0.29273521900177, "learning_rate": 7.923119240771484e-05, "loss": 0.0133, "step": 16810 }, { "grad_norm": 0.24817530810832977, "learning_rate": 7.920435666696313e-05, "loss": 0.0147, "step": 16820 }, { "grad_norm": 0.23931419849395752, "learning_rate": 7.917750815123362e-05, "loss": 0.0144, "step": 16830 }, { "grad_norm": 0.4438270628452301, "learning_rate": 7.915064687227075e-05, "loss": 0.0134, "step": 16840 }, { "grad_norm": 1.1073830127716064, "learning_rate": 7.912377284182458e-05, "loss": 0.0161, "step": 16850 }, { "grad_norm": 0.3234042227268219, "learning_rate": 7.90968860716507e-05, "loss": 0.0133, "step": 16860 }, { "grad_norm": 0.21637988090515137, "learning_rate": 7.906998657351033e-05, "loss": 0.0142, "step": 16870 }, { "grad_norm": 0.23475578427314758, "learning_rate": 7.90430743591702e-05, "loss": 0.0131, "step": 16880 }, { "grad_norm": 2.1066839694976807, "learning_rate": 7.901614944040263e-05, "loss": 0.015, "step": 16890 }, { "grad_norm": 0.1794329434633255, "learning_rate": 7.89892118289855e-05, "loss": 0.0145, "step": 16900 }, { "grad_norm": 0.22811663150787354, "learning_rate": 7.896226153670223e-05, "loss": 0.0123, "step": 16910 }, { "grad_norm": 0.28763189911842346, "learning_rate": 7.89352985753418e-05, "loss": 0.0171, "step": 16920 }, { "grad_norm": 0.23626373708248138, "learning_rate": 7.890832295669871e-05, "loss": 0.0141, "step": 16930 }, { "grad_norm": 0.36302676796913147, "learning_rate": 7.888133469257303e-05, "loss": 0.015, "step": 16940 }, { "grad_norm": 0.326611191034317, "learning_rate": 7.885433379477034e-05, "loss": 0.0139, "step": 16950 }, { "grad_norm": 0.4638749063014984, "learning_rate": 7.882732027510174e-05, "loss": 0.0183, "step": 16960 }, { "grad_norm": 0.19826887547969818, "learning_rate": 7.880029414538386e-05, "loss": 0.0153, "step": 16970 }, { "grad_norm": 0.5460270643234253, "learning_rate": 7.877325541743888e-05, "loss": 0.0144, "step": 16980 }, { "grad_norm": 0.22974586486816406, "learning_rate": 7.874620410309443e-05, "loss": 0.0128, "step": 16990 }, { "grad_norm": 0.22772178053855896, "learning_rate": 7.871914021418368e-05, "loss": 0.0133, "step": 17000 }, { "grad_norm": 0.24444766342639923, "learning_rate": 7.869206376254528e-05, "loss": 0.0171, "step": 17010 }, { "grad_norm": 0.24114921689033508, "learning_rate": 7.86649747600234e-05, "loss": 0.0145, "step": 17020 }, { "grad_norm": 0.2324948012828827, "learning_rate": 7.86378732184677e-05, "loss": 0.0131, "step": 17030 }, { "grad_norm": 0.184087336063385, "learning_rate": 7.86107591497333e-05, "loss": 0.0158, "step": 17040 }, { "grad_norm": 0.21147339046001434, "learning_rate": 7.858363256568083e-05, "loss": 0.0146, "step": 17050 }, { "grad_norm": 0.2380647361278534, "learning_rate": 7.855649347817637e-05, "loss": 0.0146, "step": 17060 }, { "grad_norm": 0.22794963419437408, "learning_rate": 7.85293418990915e-05, "loss": 0.0131, "step": 17070 }, { "grad_norm": 0.2655601501464844, "learning_rate": 7.85021778403032e-05, "loss": 0.0133, "step": 17080 }, { "grad_norm": 6.171209335327148, "learning_rate": 7.847500131369397e-05, "loss": 0.017, "step": 17090 }, { "grad_norm": 0.22942042350769043, "learning_rate": 7.844781233115176e-05, "loss": 0.0125, "step": 17100 }, { "grad_norm": 0.23645706474781036, "learning_rate": 7.842061090456996e-05, "loss": 0.014, "step": 17110 }, { "grad_norm": 0.1937960833311081, "learning_rate": 7.839339704584736e-05, "loss": 0.0132, "step": 17120 }, { "grad_norm": 0.24224403500556946, "learning_rate": 7.836617076688828e-05, "loss": 0.0136, "step": 17130 }, { "grad_norm": 0.20471452176570892, "learning_rate": 7.833893207960236e-05, "loss": 0.0143, "step": 17140 }, { "grad_norm": 0.26517874002456665, "learning_rate": 7.831168099590478e-05, "loss": 0.0141, "step": 17150 }, { "grad_norm": 0.18970276415348053, "learning_rate": 7.828441752771607e-05, "loss": 0.014, "step": 17160 }, { "grad_norm": 0.3095982074737549, "learning_rate": 7.825714168696219e-05, "loss": 0.0138, "step": 17170 }, { "grad_norm": 0.24828112125396729, "learning_rate": 7.82298534855745e-05, "loss": 0.0123, "step": 17180 }, { "grad_norm": 0.311027854681015, "learning_rate": 7.820255293548984e-05, "loss": 0.015, "step": 17190 }, { "grad_norm": 0.3316735029220581, "learning_rate": 7.817524004865035e-05, "loss": 0.0113, "step": 17200 }, { "grad_norm": 0.17961573600769043, "learning_rate": 7.814791483700366e-05, "loss": 0.0116, "step": 17210 }, { "grad_norm": 0.2277907133102417, "learning_rate": 7.81205773125027e-05, "loss": 0.0115, "step": 17220 }, { "grad_norm": 0.23378507792949677, "learning_rate": 7.809322748710585e-05, "loss": 0.016, "step": 17230 }, { "grad_norm": 0.20131871104240417, "learning_rate": 7.806586537277685e-05, "loss": 0.0139, "step": 17240 }, { "grad_norm": 0.18728123605251312, "learning_rate": 7.803849098148484e-05, "loss": 0.0129, "step": 17250 }, { "grad_norm": 0.20408929884433746, "learning_rate": 7.801110432520427e-05, "loss": 0.0139, "step": 17260 }, { "grad_norm": 0.2013038694858551, "learning_rate": 7.798370541591504e-05, "loss": 0.0129, "step": 17270 }, { "grad_norm": 0.32801327109336853, "learning_rate": 7.795629426560232e-05, "loss": 0.0118, "step": 17280 }, { "grad_norm": 0.22721821069717407, "learning_rate": 7.792887088625671e-05, "loss": 0.0124, "step": 17290 }, { "grad_norm": 0.6209973096847534, "learning_rate": 7.790143528987412e-05, "loss": 0.0136, "step": 17300 }, { "grad_norm": 0.18426942825317383, "learning_rate": 7.787398748845578e-05, "loss": 0.0124, "step": 17310 }, { "grad_norm": 0.28078263998031616, "learning_rate": 7.784652749400834e-05, "loss": 0.0122, "step": 17320 }, { "grad_norm": 0.2017505019903183, "learning_rate": 7.78190553185437e-05, "loss": 0.0143, "step": 17330 }, { "grad_norm": 0.2738192677497864, "learning_rate": 7.779157097407915e-05, "loss": 0.013, "step": 17340 }, { "grad_norm": 0.515657365322113, "learning_rate": 7.776407447263725e-05, "loss": 0.0155, "step": 17350 }, { "grad_norm": 0.16482336819171906, "learning_rate": 7.773656582624593e-05, "loss": 0.0128, "step": 17360 }, { "grad_norm": 0.2931860685348511, "learning_rate": 7.770904504693837e-05, "loss": 0.0136, "step": 17370 }, { "grad_norm": 0.25362417101860046, "learning_rate": 7.768151214675314e-05, "loss": 0.0152, "step": 17380 }, { "grad_norm": 0.3685845732688904, "learning_rate": 7.765396713773402e-05, "loss": 0.0142, "step": 17390 }, { "grad_norm": 0.2747783958911896, "learning_rate": 7.762641003193017e-05, "loss": 0.0138, "step": 17400 }, { "grad_norm": 0.13534510135650635, "learning_rate": 7.759884084139597e-05, "loss": 0.013, "step": 17410 }, { "grad_norm": 0.1892116367816925, "learning_rate": 7.757125957819116e-05, "loss": 0.012, "step": 17420 }, { "grad_norm": 0.30920958518981934, "learning_rate": 7.754366625438069e-05, "loss": 0.0149, "step": 17430 }, { "grad_norm": 0.2004566788673401, "learning_rate": 7.751606088203483e-05, "loss": 0.0146, "step": 17440 }, { "grad_norm": 0.2195606827735901, "learning_rate": 7.74884434732291e-05, "loss": 0.0147, "step": 17450 }, { "grad_norm": 0.21766585111618042, "learning_rate": 7.746081404004431e-05, "loss": 0.0126, "step": 17460 }, { "grad_norm": 0.28007936477661133, "learning_rate": 7.743317259456649e-05, "loss": 0.0148, "step": 17470 }, { "grad_norm": 0.21855023503303528, "learning_rate": 7.740551914888696e-05, "loss": 0.013, "step": 17480 }, { "grad_norm": 0.2965550720691681, "learning_rate": 7.737785371510229e-05, "loss": 0.0123, "step": 17490 }, { "grad_norm": 0.15007171034812927, "learning_rate": 7.735017630531426e-05, "loss": 0.0151, "step": 17500 }, { "grad_norm": 0.19554860889911652, "learning_rate": 7.732248693162991e-05, "loss": 0.0122, "step": 17510 }, { "grad_norm": 0.23457928001880646, "learning_rate": 7.729478560616153e-05, "loss": 0.012, "step": 17520 }, { "grad_norm": 0.1904299557209015, "learning_rate": 7.726707234102659e-05, "loss": 0.0126, "step": 17530 }, { "grad_norm": 0.24660681188106537, "learning_rate": 7.723934714834785e-05, "loss": 0.0114, "step": 17540 }, { "grad_norm": 0.42672181129455566, "learning_rate": 7.721161004025323e-05, "loss": 0.0118, "step": 17550 }, { "grad_norm": 0.4708491861820221, "learning_rate": 7.718386102887588e-05, "loss": 0.0143, "step": 17560 }, { "grad_norm": 0.34524059295654297, "learning_rate": 7.715610012635418e-05, "loss": 0.0143, "step": 17570 }, { "grad_norm": 0.19619229435920715, "learning_rate": 7.712832734483166e-05, "loss": 0.0154, "step": 17580 }, { "grad_norm": 2.084773540496826, "learning_rate": 7.710054269645713e-05, "loss": 0.0163, "step": 17590 }, { "grad_norm": 0.24297933280467987, "learning_rate": 7.707274619338449e-05, "loss": 0.0139, "step": 17600 }, { "grad_norm": 0.21108673512935638, "learning_rate": 7.704493784777289e-05, "loss": 0.0134, "step": 17610 }, { "grad_norm": 0.20236845314502716, "learning_rate": 7.701711767178665e-05, "loss": 0.0134, "step": 17620 }, { "grad_norm": 0.32360509037971497, "learning_rate": 7.698928567759525e-05, "loss": 0.0155, "step": 17630 }, { "grad_norm": 0.23067958652973175, "learning_rate": 7.696144187737335e-05, "loss": 0.0136, "step": 17640 }, { "grad_norm": 0.21187177300453186, "learning_rate": 7.69335862833008e-05, "loss": 0.0123, "step": 17650 }, { "grad_norm": 0.13054008781909943, "learning_rate": 7.690571890756256e-05, "loss": 0.0113, "step": 17660 }, { "grad_norm": 0.19997015595436096, "learning_rate": 7.687783976234877e-05, "loss": 0.0134, "step": 17670 }, { "grad_norm": 0.46384549140930176, "learning_rate": 7.684994885985472e-05, "loss": 0.0134, "step": 17680 }, { "grad_norm": 0.2694009244441986, "learning_rate": 7.682204621228083e-05, "loss": 0.0115, "step": 17690 }, { "grad_norm": 0.25777584314346313, "learning_rate": 7.679413183183267e-05, "loss": 0.0137, "step": 17700 }, { "grad_norm": 0.18580316007137299, "learning_rate": 7.676620573072093e-05, "loss": 0.012, "step": 17710 }, { "grad_norm": 0.18377092480659485, "learning_rate": 7.673826792116145e-05, "loss": 0.0149, "step": 17720 }, { "grad_norm": 0.23218846321105957, "learning_rate": 7.671031841537519e-05, "loss": 0.0138, "step": 17730 }, { "grad_norm": 0.18129608035087585, "learning_rate": 7.668235722558819e-05, "loss": 0.0153, "step": 17740 }, { "grad_norm": 0.23392751812934875, "learning_rate": 7.665438436403161e-05, "loss": 0.012, "step": 17750 }, { "grad_norm": 0.2651253342628479, "learning_rate": 7.662639984294178e-05, "loss": 0.0183, "step": 17760 }, { "grad_norm": 0.28396421670913696, "learning_rate": 7.659840367456002e-05, "loss": 0.0142, "step": 17770 }, { "grad_norm": 0.5187919735908508, "learning_rate": 7.657039587113287e-05, "loss": 0.0128, "step": 17780 }, { "grad_norm": 0.21689864993095398, "learning_rate": 7.654237644491182e-05, "loss": 0.0121, "step": 17790 }, { "grad_norm": 0.2127217948436737, "learning_rate": 7.651434540815358e-05, "loss": 0.0106, "step": 17800 }, { "grad_norm": 0.3473731577396393, "learning_rate": 7.648630277311986e-05, "loss": 0.0151, "step": 17810 }, { "grad_norm": 0.3299456238746643, "learning_rate": 7.645824855207744e-05, "loss": 0.0161, "step": 17820 }, { "grad_norm": 0.22909438610076904, "learning_rate": 7.643018275729821e-05, "loss": 0.0126, "step": 17830 }, { "grad_norm": 0.18651649355888367, "learning_rate": 7.640210540105911e-05, "loss": 0.013, "step": 17840 }, { "grad_norm": 0.1901843547821045, "learning_rate": 7.637401649564213e-05, "loss": 0.0125, "step": 17850 }, { "grad_norm": 0.18884409964084625, "learning_rate": 7.63459160533343e-05, "loss": 0.0131, "step": 17860 }, { "grad_norm": 0.28358787298202515, "learning_rate": 7.631780408642769e-05, "loss": 0.0144, "step": 17870 }, { "grad_norm": 0.2491304874420166, "learning_rate": 7.628968060721946e-05, "loss": 0.0119, "step": 17880 }, { "grad_norm": 0.39514851570129395, "learning_rate": 7.626154562801176e-05, "loss": 0.0132, "step": 17890 }, { "grad_norm": 3.062119960784912, "learning_rate": 7.623339916111177e-05, "loss": 0.0148, "step": 17900 }, { "grad_norm": 0.22772528231143951, "learning_rate": 7.620524121883174e-05, "loss": 0.0137, "step": 17910 }, { "grad_norm": 0.5600250363349915, "learning_rate": 7.617707181348892e-05, "loss": 0.0154, "step": 17920 }, { "grad_norm": 0.29974567890167236, "learning_rate": 7.614889095740554e-05, "loss": 0.0141, "step": 17930 }, { "grad_norm": 0.20292247831821442, "learning_rate": 7.612069866290885e-05, "loss": 0.0156, "step": 17940 }, { "grad_norm": 0.19179771840572357, "learning_rate": 7.609249494233115e-05, "loss": 0.0114, "step": 17950 }, { "grad_norm": 0.76060950756073, "learning_rate": 7.60642798080097e-05, "loss": 0.0177, "step": 17960 }, { "grad_norm": 5.01698112487793, "learning_rate": 7.603605327228673e-05, "loss": 0.0136, "step": 17970 }, { "grad_norm": 1.8463765382766724, "learning_rate": 7.600781534750953e-05, "loss": 0.0118, "step": 17980 }, { "grad_norm": 0.21796409785747528, "learning_rate": 7.597956604603029e-05, "loss": 0.014, "step": 17990 }, { "grad_norm": 0.13731731474399567, "learning_rate": 7.595130538020622e-05, "loss": 0.0117, "step": 18000 }, { "grad_norm": 0.16694067418575287, "learning_rate": 7.592303336239952e-05, "loss": 0.0135, "step": 18010 }, { "grad_norm": 0.2016540914773941, "learning_rate": 7.589475000497733e-05, "loss": 0.0112, "step": 18020 }, { "grad_norm": 0.631733775138855, "learning_rate": 7.586645532031171e-05, "loss": 0.0142, "step": 18030 }, { "grad_norm": 0.37644314765930176, "learning_rate": 7.583814932077975e-05, "loss": 0.0125, "step": 18040 }, { "grad_norm": 0.2992361783981323, "learning_rate": 7.580983201876343e-05, "loss": 0.0147, "step": 18050 }, { "grad_norm": 0.21926000714302063, "learning_rate": 7.578150342664973e-05, "loss": 0.0155, "step": 18060 }, { "grad_norm": 0.2631837725639343, "learning_rate": 7.575316355683055e-05, "loss": 0.0157, "step": 18070 }, { "grad_norm": 0.36011457443237305, "learning_rate": 7.572481242170266e-05, "loss": 0.013, "step": 18080 }, { "grad_norm": 0.21541549265384674, "learning_rate": 7.569645003366786e-05, "loss": 0.0129, "step": 18090 }, { "grad_norm": 0.2121535986661911, "learning_rate": 7.566807640513278e-05, "loss": 0.0134, "step": 18100 }, { "grad_norm": 0.26502278447151184, "learning_rate": 7.563969154850905e-05, "loss": 0.0126, "step": 18110 }, { "grad_norm": 0.27423232793807983, "learning_rate": 7.561129547621314e-05, "loss": 0.015, "step": 18120 }, { "grad_norm": 0.2686270773410797, "learning_rate": 7.558288820066648e-05, "loss": 0.014, "step": 18130 }, { "grad_norm": 0.41307905316352844, "learning_rate": 7.555446973429534e-05, "loss": 0.0118, "step": 18140 }, { "grad_norm": 0.16272589564323425, "learning_rate": 7.552604008953097e-05, "loss": 0.011, "step": 18150 }, { "grad_norm": 0.2224539965391159, "learning_rate": 7.549759927880942e-05, "loss": 0.0131, "step": 18160 }, { "grad_norm": 0.18370601534843445, "learning_rate": 7.54691473145717e-05, "loss": 0.0127, "step": 18170 }, { "grad_norm": 0.1788295954465866, "learning_rate": 7.544068420926365e-05, "loss": 0.0129, "step": 18180 }, { "grad_norm": 0.2167971134185791, "learning_rate": 7.5412209975336e-05, "loss": 0.0131, "step": 18190 }, { "grad_norm": 0.24500015377998352, "learning_rate": 7.538372462524435e-05, "loss": 0.0162, "step": 18200 }, { "grad_norm": 0.2841981053352356, "learning_rate": 7.535522817144915e-05, "loss": 0.0143, "step": 18210 }, { "grad_norm": 0.6874076724052429, "learning_rate": 7.532672062641575e-05, "loss": 0.0141, "step": 18220 }, { "grad_norm": 1.4928176403045654, "learning_rate": 7.529820200261427e-05, "loss": 0.0157, "step": 18230 }, { "grad_norm": 0.20395974814891815, "learning_rate": 7.526967231251977e-05, "loss": 0.0149, "step": 18240 }, { "grad_norm": 0.21017436683177948, "learning_rate": 7.524113156861206e-05, "loss": 0.013, "step": 18250 }, { "grad_norm": 0.3665010929107666, "learning_rate": 7.521257978337587e-05, "loss": 0.0133, "step": 18260 }, { "grad_norm": 0.27346542477607727, "learning_rate": 7.518401696930072e-05, "loss": 0.0134, "step": 18270 }, { "grad_norm": 0.24303574860095978, "learning_rate": 7.515544313888092e-05, "loss": 0.0143, "step": 18280 }, { "grad_norm": 0.22021980583667755, "learning_rate": 7.512685830461568e-05, "loss": 0.0117, "step": 18290 }, { "grad_norm": 0.20015805959701538, "learning_rate": 7.509826247900892e-05, "loss": 0.0113, "step": 18300 }, { "grad_norm": 0.14940227568149567, "learning_rate": 7.50696556745695e-05, "loss": 0.015, "step": 18310 }, { "grad_norm": 0.18773436546325684, "learning_rate": 7.504103790381095e-05, "loss": 0.0118, "step": 18320 }, { "grad_norm": 0.24297849833965302, "learning_rate": 7.501240917925167e-05, "loss": 0.0151, "step": 18330 }, { "grad_norm": 0.23991984128952026, "learning_rate": 7.498376951341483e-05, "loss": 0.0121, "step": 18340 }, { "grad_norm": 0.2056460976600647, "learning_rate": 7.495511891882844e-05, "loss": 0.012, "step": 18350 }, { "grad_norm": 0.2240450531244278, "learning_rate": 7.492645740802517e-05, "loss": 0.0109, "step": 18360 }, { "grad_norm": 0.1479732096195221, "learning_rate": 7.489778499354259e-05, "loss": 0.0117, "step": 18370 }, { "grad_norm": 0.18197548389434814, "learning_rate": 7.486910168792298e-05, "loss": 0.0124, "step": 18380 }, { "grad_norm": 0.23240415751934052, "learning_rate": 7.484040750371337e-05, "loss": 0.0108, "step": 18390 }, { "grad_norm": 0.20853480696678162, "learning_rate": 7.48117024534656e-05, "loss": 0.0112, "step": 18400 }, { "grad_norm": 0.22263826429843903, "learning_rate": 7.478298654973621e-05, "loss": 0.0121, "step": 18410 }, { "grad_norm": 0.25532296299934387, "learning_rate": 7.475425980508654e-05, "loss": 0.0109, "step": 18420 }, { "grad_norm": 0.19394664466381073, "learning_rate": 7.47255222320826e-05, "loss": 0.0132, "step": 18430 }, { "grad_norm": 0.19926249980926514, "learning_rate": 7.469677384329522e-05, "loss": 0.0137, "step": 18440 }, { "grad_norm": 0.19187326729297638, "learning_rate": 7.46680146512999e-05, "loss": 0.0128, "step": 18450 }, { "grad_norm": 0.27529534697532654, "learning_rate": 7.463924466867692e-05, "loss": 0.0127, "step": 18460 }, { "grad_norm": 0.28564098477363586, "learning_rate": 7.461046390801121e-05, "loss": 0.0106, "step": 18470 }, { "grad_norm": 0.19619762897491455, "learning_rate": 7.458167238189248e-05, "loss": 0.0122, "step": 18480 }, { "grad_norm": 2.4100968837738037, "learning_rate": 7.455287010291509e-05, "loss": 0.0149, "step": 18490 }, { "grad_norm": 0.20390376448631287, "learning_rate": 7.452405708367818e-05, "loss": 0.0119, "step": 18500 }, { "grad_norm": 0.4954848885536194, "learning_rate": 7.449523333678549e-05, "loss": 0.0134, "step": 18510 }, { "grad_norm": 0.22242657840251923, "learning_rate": 7.446639887484555e-05, "loss": 0.0137, "step": 18520 }, { "grad_norm": 0.22494059801101685, "learning_rate": 7.443755371047152e-05, "loss": 0.0112, "step": 18530 }, { "grad_norm": 0.32313138246536255, "learning_rate": 7.440869785628124e-05, "loss": 0.0142, "step": 18540 }, { "grad_norm": 0.5924232602119446, "learning_rate": 7.437983132489727e-05, "loss": 0.0135, "step": 18550 }, { "grad_norm": 0.15276563167572021, "learning_rate": 7.435095412894678e-05, "loss": 0.0123, "step": 18560 }, { "grad_norm": 0.14396727085113525, "learning_rate": 7.432206628106165e-05, "loss": 0.0106, "step": 18570 }, { "grad_norm": 0.18040505051612854, "learning_rate": 7.429316779387842e-05, "loss": 0.0108, "step": 18580 }, { "grad_norm": 0.21286213397979736, "learning_rate": 7.426425868003825e-05, "loss": 0.0126, "step": 18590 }, { "grad_norm": 0.16266387701034546, "learning_rate": 7.423533895218698e-05, "loss": 0.0117, "step": 18600 }, { "grad_norm": 0.35866737365722656, "learning_rate": 7.420640862297507e-05, "loss": 0.0147, "step": 18610 }, { "grad_norm": 0.2179044932126999, "learning_rate": 7.417746770505766e-05, "loss": 0.014, "step": 18620 }, { "grad_norm": 0.3606296181678772, "learning_rate": 7.414851621109444e-05, "loss": 0.015, "step": 18630 }, { "grad_norm": 1.323176383972168, "learning_rate": 7.411955415374982e-05, "loss": 0.012, "step": 18640 }, { "grad_norm": 0.15151280164718628, "learning_rate": 7.409058154569278e-05, "loss": 0.0112, "step": 18650 }, { "grad_norm": 0.23248930275440216, "learning_rate": 7.406159839959689e-05, "loss": 0.0115, "step": 18660 }, { "grad_norm": 0.33017590641975403, "learning_rate": 7.403260472814039e-05, "loss": 0.0114, "step": 18670 }, { "grad_norm": 0.24261139333248138, "learning_rate": 7.400360054400611e-05, "loss": 0.0146, "step": 18680 }, { "grad_norm": 0.25295644998550415, "learning_rate": 7.397458585988143e-05, "loss": 0.0133, "step": 18690 }, { "grad_norm": 0.2068299502134323, "learning_rate": 7.394556068845841e-05, "loss": 0.0112, "step": 18700 }, { "grad_norm": 0.29296988248825073, "learning_rate": 7.39165250424336e-05, "loss": 0.0106, "step": 18710 }, { "grad_norm": 0.1717558652162552, "learning_rate": 7.38874789345082e-05, "loss": 0.0112, "step": 18720 }, { "grad_norm": 0.2540827691555023, "learning_rate": 7.385842237738794e-05, "loss": 0.013, "step": 18730 }, { "grad_norm": 0.20191673934459686, "learning_rate": 7.382935538378318e-05, "loss": 0.0133, "step": 18740 }, { "grad_norm": 0.24754025042057037, "learning_rate": 7.38002779664088e-05, "loss": 0.0142, "step": 18750 }, { "grad_norm": 0.1982675939798355, "learning_rate": 7.377119013798422e-05, "loss": 0.0105, "step": 18760 }, { "grad_norm": 0.22366848587989807, "learning_rate": 7.374209191123349e-05, "loss": 0.0113, "step": 18770 }, { "grad_norm": 0.7336628437042236, "learning_rate": 7.371298329888512e-05, "loss": 0.0126, "step": 18780 }, { "grad_norm": 0.3441542088985443, "learning_rate": 7.368386431367223e-05, "loss": 0.011, "step": 18790 }, { "grad_norm": 0.13460887968540192, "learning_rate": 7.365473496833246e-05, "loss": 0.0127, "step": 18800 }, { "grad_norm": 0.15105684101581573, "learning_rate": 7.362559527560797e-05, "loss": 0.012, "step": 18810 }, { "grad_norm": 0.15704640746116638, "learning_rate": 7.359644524824545e-05, "loss": 0.014, "step": 18820 }, { "grad_norm": 1.1784011125564575, "learning_rate": 7.35672848989961e-05, "loss": 0.0132, "step": 18830 }, { "grad_norm": 0.20697961747646332, "learning_rate": 7.353811424061565e-05, "loss": 0.0117, "step": 18840 }, { "grad_norm": 0.19646845757961273, "learning_rate": 7.350893328586435e-05, "loss": 0.0123, "step": 18850 }, { "grad_norm": 0.20115670561790466, "learning_rate": 7.347974204750696e-05, "loss": 0.0093, "step": 18860 }, { "grad_norm": 0.3394930064678192, "learning_rate": 7.345054053831266e-05, "loss": 0.0094, "step": 18870 }, { "grad_norm": 0.2991533577442169, "learning_rate": 7.342132877105523e-05, "loss": 0.0133, "step": 18880 }, { "grad_norm": 0.24124982953071594, "learning_rate": 7.339210675851288e-05, "loss": 0.0139, "step": 18890 }, { "grad_norm": 0.17386437952518463, "learning_rate": 7.336287451346831e-05, "loss": 0.0112, "step": 18900 }, { "grad_norm": 1.4445680379867554, "learning_rate": 7.333363204870868e-05, "loss": 0.0123, "step": 18910 }, { "grad_norm": 0.17921748757362366, "learning_rate": 7.330437937702566e-05, "loss": 0.015, "step": 18920 }, { "grad_norm": 0.2434711754322052, "learning_rate": 7.327511651121534e-05, "loss": 0.0129, "step": 18930 }, { "grad_norm": 0.6418560147285461, "learning_rate": 7.324584346407833e-05, "loss": 0.0127, "step": 18940 }, { "grad_norm": 0.2002808153629303, "learning_rate": 7.321656024841962e-05, "loss": 0.0121, "step": 18950 }, { "grad_norm": 0.2232275754213333, "learning_rate": 7.31872668770487e-05, "loss": 0.0138, "step": 18960 }, { "grad_norm": 0.2631484866142273, "learning_rate": 7.315796336277949e-05, "loss": 0.0128, "step": 18970 }, { "grad_norm": 0.1852889060974121, "learning_rate": 7.312864971843031e-05, "loss": 0.0129, "step": 18980 }, { "grad_norm": 0.19329339265823364, "learning_rate": 7.309932595682398e-05, "loss": 0.013, "step": 18990 }, { "grad_norm": 0.19108638167381287, "learning_rate": 7.30699920907877e-05, "loss": 0.0128, "step": 19000 }, { "grad_norm": 0.1583968549966812, "learning_rate": 7.304064813315308e-05, "loss": 0.0124, "step": 19010 }, { "grad_norm": 0.3417547941207886, "learning_rate": 7.30112940967562e-05, "loss": 0.0124, "step": 19020 }, { "grad_norm": 0.41482797265052795, "learning_rate": 7.298192999443747e-05, "loss": 0.0148, "step": 19030 }, { "grad_norm": 0.224809467792511, "learning_rate": 7.295255583904179e-05, "loss": 0.0109, "step": 19040 }, { "grad_norm": 0.3084113299846649, "learning_rate": 7.29231716434184e-05, "loss": 0.0105, "step": 19050 }, { "grad_norm": 0.20580831170082092, "learning_rate": 7.289377742042092e-05, "loss": 0.0107, "step": 19060 }, { "grad_norm": 0.17934612929821014, "learning_rate": 7.286437318290741e-05, "loss": 0.0115, "step": 19070 }, { "grad_norm": 0.16356879472732544, "learning_rate": 7.283495894374027e-05, "loss": 0.0145, "step": 19080 }, { "grad_norm": 0.19714023172855377, "learning_rate": 7.280553471578629e-05, "loss": 0.0115, "step": 19090 }, { "grad_norm": 2.9522252082824707, "learning_rate": 7.277610051191663e-05, "loss": 0.0129, "step": 19100 }, { "grad_norm": 0.5501106977462769, "learning_rate": 7.27466563450068e-05, "loss": 0.012, "step": 19110 }, { "grad_norm": 0.2870889902114868, "learning_rate": 7.271720222793668e-05, "loss": 0.0121, "step": 19120 }, { "grad_norm": 0.19535689055919647, "learning_rate": 7.268773817359054e-05, "loss": 0.0113, "step": 19130 }, { "grad_norm": 0.3840119540691376, "learning_rate": 7.265826419485689e-05, "loss": 0.0109, "step": 19140 }, { "grad_norm": 0.20144198834896088, "learning_rate": 7.26287803046287e-05, "loss": 0.0122, "step": 19150 }, { "grad_norm": 1.0064362287521362, "learning_rate": 7.259928651580322e-05, "loss": 0.0115, "step": 19160 }, { "grad_norm": 0.2475178837776184, "learning_rate": 7.256978284128202e-05, "loss": 0.0103, "step": 19170 }, { "grad_norm": 0.35478973388671875, "learning_rate": 7.254026929397101e-05, "loss": 0.0141, "step": 19180 }, { "grad_norm": 0.20057463645935059, "learning_rate": 7.251074588678044e-05, "loss": 0.0138, "step": 19190 }, { "grad_norm": 0.36484193801879883, "learning_rate": 7.248121263262482e-05, "loss": 0.0124, "step": 19200 }, { "grad_norm": 0.24637368321418762, "learning_rate": 7.245166954442304e-05, "loss": 0.0112, "step": 19210 }, { "grad_norm": 0.4572760760784149, "learning_rate": 7.24221166350982e-05, "loss": 0.0106, "step": 19220 }, { "grad_norm": 0.19801513850688934, "learning_rate": 7.239255391757777e-05, "loss": 0.0104, "step": 19230 }, { "grad_norm": 0.14121407270431519, "learning_rate": 7.236298140479351e-05, "loss": 0.0107, "step": 19240 }, { "grad_norm": 0.22242765128612518, "learning_rate": 7.233339910968141e-05, "loss": 0.0144, "step": 19250 }, { "grad_norm": 0.30728602409362793, "learning_rate": 7.230380704518178e-05, "loss": 0.0116, "step": 19260 }, { "grad_norm": 0.39500749111175537, "learning_rate": 7.227420522423919e-05, "loss": 0.0108, "step": 19270 }, { "grad_norm": 0.18578003346920013, "learning_rate": 7.224459365980249e-05, "loss": 0.0125, "step": 19280 }, { "grad_norm": 0.16547293961048126, "learning_rate": 7.221497236482479e-05, "loss": 0.011, "step": 19290 }, { "grad_norm": 1.212071418762207, "learning_rate": 7.218534135226345e-05, "loss": 0.0098, "step": 19300 }, { "grad_norm": 0.22955180704593658, "learning_rate": 7.215570063508005e-05, "loss": 0.0101, "step": 19310 }, { "grad_norm": 0.23442214727401733, "learning_rate": 7.21260502262405e-05, "loss": 0.0125, "step": 19320 }, { "grad_norm": 0.20970839262008667, "learning_rate": 7.209639013871487e-05, "loss": 0.0146, "step": 19330 }, { "grad_norm": 0.17063571512699127, "learning_rate": 7.206672038547746e-05, "loss": 0.0133, "step": 19340 }, { "grad_norm": 0.2097594141960144, "learning_rate": 7.203704097950687e-05, "loss": 0.0115, "step": 19350 }, { "grad_norm": 0.19787943363189697, "learning_rate": 7.200735193378587e-05, "loss": 0.0115, "step": 19360 }, { "grad_norm": 0.1863662749528885, "learning_rate": 7.197765326130145e-05, "loss": 0.0095, "step": 19370 }, { "grad_norm": 0.1689102202653885, "learning_rate": 7.194794497504481e-05, "loss": 0.012, "step": 19380 }, { "grad_norm": 0.24828654527664185, "learning_rate": 7.19182270880114e-05, "loss": 0.0143, "step": 19390 }, { "grad_norm": 0.2096397578716278, "learning_rate": 7.188849961320079e-05, "loss": 0.0124, "step": 19400 }, { "grad_norm": 0.4743631184101105, "learning_rate": 7.185876256361681e-05, "loss": 0.0154, "step": 19410 }, { "grad_norm": 0.2184489369392395, "learning_rate": 7.182901595226745e-05, "loss": 0.0115, "step": 19420 }, { "grad_norm": 0.16946129500865936, "learning_rate": 7.179925979216491e-05, "loss": 0.0116, "step": 19430 }, { "grad_norm": 0.20644178986549377, "learning_rate": 7.17694940963255e-05, "loss": 0.0124, "step": 19440 }, { "grad_norm": 0.48488280177116394, "learning_rate": 7.17397188777698e-05, "loss": 0.0112, "step": 19450 }, { "grad_norm": 0.2567298114299774, "learning_rate": 7.170993414952245e-05, "loss": 0.0097, "step": 19460 }, { "grad_norm": 0.3405981659889221, "learning_rate": 7.168013992461233e-05, "loss": 0.012, "step": 19470 }, { "grad_norm": 0.17737053334712982, "learning_rate": 7.165033621607245e-05, "loss": 0.0121, "step": 19480 }, { "grad_norm": 0.18082170188426971, "learning_rate": 7.162052303693995e-05, "loss": 0.01, "step": 19490 }, { "grad_norm": 0.16152925789356232, "learning_rate": 7.159070040025614e-05, "loss": 0.0088, "step": 19500 }, { "grad_norm": 0.2169194519519806, "learning_rate": 7.156086831906646e-05, "loss": 0.0135, "step": 19510 }, { "grad_norm": 0.17824049293994904, "learning_rate": 7.153102680642044e-05, "loss": 0.0122, "step": 19520 }, { "grad_norm": 0.2539312541484833, "learning_rate": 7.150117587537182e-05, "loss": 0.012, "step": 19530 }, { "grad_norm": 0.2428131401538849, "learning_rate": 7.147131553897838e-05, "loss": 0.0102, "step": 19540 }, { "grad_norm": 0.37529513239860535, "learning_rate": 7.144144581030205e-05, "loss": 0.0132, "step": 19550 }, { "grad_norm": 0.6379961371421814, "learning_rate": 7.141156670240889e-05, "loss": 0.0123, "step": 19560 }, { "grad_norm": 0.25380679965019226, "learning_rate": 7.138167822836899e-05, "loss": 0.0124, "step": 19570 }, { "grad_norm": 4.6434326171875, "learning_rate": 7.135178040125661e-05, "loss": 0.0117, "step": 19580 }, { "grad_norm": 0.3250114321708679, "learning_rate": 7.13218732341501e-05, "loss": 0.0141, "step": 19590 }, { "grad_norm": 0.2905246913433075, "learning_rate": 7.129195674013181e-05, "loss": 0.0125, "step": 19600 }, { "grad_norm": 0.16393324732780457, "learning_rate": 7.126203093228829e-05, "loss": 0.0118, "step": 19610 }, { "grad_norm": 0.40121906995773315, "learning_rate": 7.123209582371006e-05, "loss": 0.0126, "step": 19620 }, { "grad_norm": 0.1951582282781601, "learning_rate": 7.120215142749179e-05, "loss": 0.0106, "step": 19630 }, { "grad_norm": 0.15961962938308716, "learning_rate": 7.117219775673218e-05, "loss": 0.0149, "step": 19640 }, { "grad_norm": 0.28389111161231995, "learning_rate": 7.114223482453394e-05, "loss": 0.0128, "step": 19650 }, { "grad_norm": 0.17659379541873932, "learning_rate": 7.11122626440039e-05, "loss": 0.0118, "step": 19660 }, { "grad_norm": 0.22361153364181519, "learning_rate": 7.10822812282529e-05, "loss": 0.0107, "step": 19670 }, { "grad_norm": 0.2065725028514862, "learning_rate": 7.105229059039586e-05, "loss": 0.0122, "step": 19680 }, { "grad_norm": 0.18598298728466034, "learning_rate": 7.102229074355168e-05, "loss": 0.0108, "step": 19690 }, { "grad_norm": 0.17150811851024628, "learning_rate": 7.099228170084332e-05, "loss": 0.0113, "step": 19700 }, { "grad_norm": 0.21879787743091583, "learning_rate": 7.096226347539771e-05, "loss": 0.0119, "step": 19710 }, { "grad_norm": 0.19736602902412415, "learning_rate": 7.093223608034592e-05, "loss": 0.0116, "step": 19720 }, { "grad_norm": 0.21518254280090332, "learning_rate": 7.090219952882291e-05, "loss": 0.0144, "step": 19730 }, { "grad_norm": 0.24657301604747772, "learning_rate": 7.087215383396768e-05, "loss": 0.0132, "step": 19740 }, { "grad_norm": 0.8942198157310486, "learning_rate": 7.084209900892325e-05, "loss": 0.0118, "step": 19750 }, { "grad_norm": 5.229581832885742, "learning_rate": 7.081203506683663e-05, "loss": 0.0101, "step": 19760 }, { "grad_norm": 0.2812049090862274, "learning_rate": 7.078196202085879e-05, "loss": 0.0133, "step": 19770 }, { "grad_norm": 2.2645795345306396, "learning_rate": 7.075187988414473e-05, "loss": 0.0116, "step": 19780 }, { "grad_norm": 0.20434783399105072, "learning_rate": 7.072178866985334e-05, "loss": 0.0125, "step": 19790 }, { "grad_norm": 5.99144983291626, "learning_rate": 7.069168839114764e-05, "loss": 0.0117, "step": 19800 }, { "grad_norm": 0.2067071497440338, "learning_rate": 7.066157906119441e-05, "loss": 0.0128, "step": 19810 }, { "grad_norm": 0.19304728507995605, "learning_rate": 7.063146069316456e-05, "loss": 0.0131, "step": 19820 }, { "grad_norm": 0.21487940847873688, "learning_rate": 7.060133330023287e-05, "loss": 0.0144, "step": 19830 }, { "grad_norm": 0.23050925135612488, "learning_rate": 7.057119689557807e-05, "loss": 0.0121, "step": 19840 }, { "grad_norm": 0.24733059108257294, "learning_rate": 7.054105149238287e-05, "loss": 0.0117, "step": 19850 }, { "grad_norm": 0.33393824100494385, "learning_rate": 7.051089710383387e-05, "loss": 0.0134, "step": 19860 }, { "grad_norm": 0.7142218947410583, "learning_rate": 7.048073374312166e-05, "loss": 0.0132, "step": 19870 }, { "grad_norm": 0.3876741826534271, "learning_rate": 7.045056142344067e-05, "loss": 0.0133, "step": 19880 }, { "grad_norm": 0.23232407867908478, "learning_rate": 7.042038015798934e-05, "loss": 0.0116, "step": 19890 }, { "grad_norm": 0.6504204273223877, "learning_rate": 7.039018995996995e-05, "loss": 0.0121, "step": 19900 }, { "grad_norm": 0.27890124917030334, "learning_rate": 7.035999084258876e-05, "loss": 0.0137, "step": 19910 }, { "grad_norm": 0.19253480434417725, "learning_rate": 7.032978281905583e-05, "loss": 0.0121, "step": 19920 }, { "grad_norm": 0.2228357493877411, "learning_rate": 7.029956590258522e-05, "loss": 0.0115, "step": 19930 }, { "grad_norm": 0.24064505100250244, "learning_rate": 7.026934010639483e-05, "loss": 0.0127, "step": 19940 }, { "grad_norm": 0.19785742461681366, "learning_rate": 7.023910544370645e-05, "loss": 0.0107, "step": 19950 }, { "grad_norm": 0.8777288794517517, "learning_rate": 7.020886192774573e-05, "loss": 0.0127, "step": 19960 }, { "grad_norm": 0.22835463285446167, "learning_rate": 7.017860957174222e-05, "loss": 0.0147, "step": 19970 }, { "grad_norm": 0.2363961935043335, "learning_rate": 7.014834838892934e-05, "loss": 0.0124, "step": 19980 }, { "grad_norm": 0.16196639835834503, "learning_rate": 7.011807839254435e-05, "loss": 0.0128, "step": 19990 }, { "grad_norm": 0.1941680610179901, "learning_rate": 7.008779959582837e-05, "loss": 0.0128, "step": 20000 }, { "grad_norm": 0.1777983456850052, "learning_rate": 7.005751201202639e-05, "loss": 0.0139, "step": 20010 }, { "grad_norm": 0.37433505058288574, "learning_rate": 7.002721565438724e-05, "loss": 0.0114, "step": 20020 }, { "grad_norm": 0.21800187230110168, "learning_rate": 6.999691053616353e-05, "loss": 0.0109, "step": 20030 }, { "grad_norm": 0.1888647824525833, "learning_rate": 6.996659667061178e-05, "loss": 0.0107, "step": 20040 }, { "grad_norm": 0.324258029460907, "learning_rate": 6.993627407099229e-05, "loss": 0.0118, "step": 20050 }, { "grad_norm": 0.1977282613515854, "learning_rate": 6.990594275056921e-05, "loss": 0.0118, "step": 20060 }, { "grad_norm": 0.33000072836875916, "learning_rate": 6.987560272261049e-05, "loss": 0.0114, "step": 20070 }, { "grad_norm": 0.1918318271636963, "learning_rate": 6.984525400038788e-05, "loss": 0.0121, "step": 20080 }, { "grad_norm": 0.23091894388198853, "learning_rate": 6.981489659717693e-05, "loss": 0.0141, "step": 20090 }, { "grad_norm": 0.2987763285636902, "learning_rate": 6.978453052625703e-05, "loss": 0.0154, "step": 20100 }, { "grad_norm": 0.30332648754119873, "learning_rate": 6.975415580091131e-05, "loss": 0.0132, "step": 20110 }, { "grad_norm": 0.23062092065811157, "learning_rate": 6.972377243442672e-05, "loss": 0.014, "step": 20120 }, { "grad_norm": 0.22040955722332, "learning_rate": 6.969338044009396e-05, "loss": 0.0137, "step": 20130 }, { "grad_norm": 0.2321990430355072, "learning_rate": 6.966297983120753e-05, "loss": 0.0129, "step": 20140 }, { "grad_norm": 0.1607012301683426, "learning_rate": 6.963257062106571e-05, "loss": 0.0138, "step": 20150 }, { "grad_norm": 0.3799493610858917, "learning_rate": 6.960215282297048e-05, "loss": 0.0123, "step": 20160 }, { "grad_norm": 0.4230436086654663, "learning_rate": 6.957172645022765e-05, "loss": 0.0112, "step": 20170 }, { "grad_norm": 0.20395983755588531, "learning_rate": 6.954129151614673e-05, "loss": 0.0098, "step": 20180 }, { "grad_norm": 0.1529327780008316, "learning_rate": 6.9510848034041e-05, "loss": 0.0114, "step": 20190 }, { "grad_norm": 0.17856991291046143, "learning_rate": 6.94803960172275e-05, "loss": 0.0102, "step": 20200 }, { "grad_norm": 0.16637016832828522, "learning_rate": 6.944993547902692e-05, "loss": 0.0112, "step": 20210 }, { "grad_norm": 0.30004775524139404, "learning_rate": 6.941946643276379e-05, "loss": 0.013, "step": 20220 }, { "grad_norm": 0.20313893258571625, "learning_rate": 6.938898889176626e-05, "loss": 0.0102, "step": 20230 }, { "grad_norm": 0.2961438298225403, "learning_rate": 6.935850286936627e-05, "loss": 0.0103, "step": 20240 }, { "grad_norm": 1.6041573286056519, "learning_rate": 6.932800837889943e-05, "loss": 0.0122, "step": 20250 }, { "grad_norm": 0.29528290033340454, "learning_rate": 6.92975054337051e-05, "loss": 0.0142, "step": 20260 }, { "grad_norm": 0.23684605956077576, "learning_rate": 6.926699404712627e-05, "loss": 0.0113, "step": 20270 }, { "grad_norm": 0.17637033760547638, "learning_rate": 6.923647423250967e-05, "loss": 0.0109, "step": 20280 }, { "grad_norm": 0.1729443520307541, "learning_rate": 6.920594600320567e-05, "loss": 0.0108, "step": 20290 }, { "grad_norm": 0.5745081305503845, "learning_rate": 6.91754093725684e-05, "loss": 0.011, "step": 20300 }, { "grad_norm": 5.131688594818115, "learning_rate": 6.914486435395561e-05, "loss": 0.0145, "step": 20310 }, { "grad_norm": 0.22235850989818573, "learning_rate": 6.911431096072871e-05, "loss": 0.011, "step": 20320 }, { "grad_norm": 0.5571424961090088, "learning_rate": 6.90837492062528e-05, "loss": 0.0148, "step": 20330 }, { "grad_norm": 0.2942033112049103, "learning_rate": 6.905317910389664e-05, "loss": 0.0119, "step": 20340 }, { "grad_norm": 0.24872547388076782, "learning_rate": 6.902260066703262e-05, "loss": 0.0137, "step": 20350 }, { "grad_norm": 0.20984193682670593, "learning_rate": 6.89920139090368e-05, "loss": 0.0125, "step": 20360 }, { "grad_norm": 0.4254193902015686, "learning_rate": 6.896141884328886e-05, "loss": 0.0115, "step": 20370 }, { "grad_norm": 0.23135335743427277, "learning_rate": 6.893081548317211e-05, "loss": 0.0133, "step": 20380 }, { "grad_norm": 0.1735987663269043, "learning_rate": 6.890020384207353e-05, "loss": 0.0141, "step": 20390 }, { "grad_norm": 0.651439368724823, "learning_rate": 6.886958393338366e-05, "loss": 0.0123, "step": 20400 }, { "grad_norm": 1.7511873245239258, "learning_rate": 6.883895577049668e-05, "loss": 0.0141, "step": 20410 }, { "grad_norm": 0.22616378962993622, "learning_rate": 6.880831936681045e-05, "loss": 0.0109, "step": 20420 }, { "grad_norm": 0.1621994823217392, "learning_rate": 6.87776747357263e-05, "loss": 0.0105, "step": 20430 }, { "grad_norm": 2.0861032009124756, "learning_rate": 6.874702189064927e-05, "loss": 0.0107, "step": 20440 }, { "grad_norm": 0.6084022521972656, "learning_rate": 6.871636084498796e-05, "loss": 0.0129, "step": 20450 }, { "grad_norm": 0.27806201577186584, "learning_rate": 6.868569161215453e-05, "loss": 0.0125, "step": 20460 }, { "grad_norm": 0.19865314662456512, "learning_rate": 6.865501420556477e-05, "loss": 0.011, "step": 20470 }, { "grad_norm": 0.2349516600370407, "learning_rate": 6.862432863863802e-05, "loss": 0.0125, "step": 20480 }, { "grad_norm": 0.1806574910879135, "learning_rate": 6.859363492479716e-05, "loss": 0.0135, "step": 20490 }, { "grad_norm": 0.23531892895698547, "learning_rate": 6.856293307746868e-05, "loss": 0.0123, "step": 20500 }, { "grad_norm": 0.2802538573741913, "learning_rate": 6.853222311008263e-05, "loss": 0.0106, "step": 20510 }, { "grad_norm": 0.24226385354995728, "learning_rate": 6.850150503607256e-05, "loss": 0.0153, "step": 20520 }, { "grad_norm": 0.1874987930059433, "learning_rate": 6.847077886887562e-05, "loss": 0.0146, "step": 20530 }, { "grad_norm": 0.1603793054819107, "learning_rate": 6.844004462193248e-05, "loss": 0.0098, "step": 20540 }, { "grad_norm": 0.21864385902881622, "learning_rate": 6.840930230868734e-05, "loss": 0.0106, "step": 20550 }, { "grad_norm": 0.27241015434265137, "learning_rate": 6.837855194258791e-05, "loss": 0.0152, "step": 20560 }, { "grad_norm": 0.7357659935951233, "learning_rate": 6.834779353708548e-05, "loss": 0.0106, "step": 20570 }, { "grad_norm": 0.4350709021091461, "learning_rate": 6.831702710563478e-05, "loss": 0.0127, "step": 20580 }, { "grad_norm": 1.4209911823272705, "learning_rate": 6.828625266169416e-05, "loss": 0.0117, "step": 20590 }, { "grad_norm": 0.17109188437461853, "learning_rate": 6.825547021872534e-05, "loss": 0.0098, "step": 20600 }, { "grad_norm": 0.18366605043411255, "learning_rate": 6.822467979019365e-05, "loss": 0.0117, "step": 20610 }, { "grad_norm": 0.18968169391155243, "learning_rate": 6.819388138956783e-05, "loss": 0.0104, "step": 20620 }, { "grad_norm": 0.23248016834259033, "learning_rate": 6.816307503032019e-05, "loss": 0.0122, "step": 20630 }, { "grad_norm": 0.17605352401733398, "learning_rate": 6.813226072592645e-05, "loss": 0.0114, "step": 20640 }, { "grad_norm": 0.33349499106407166, "learning_rate": 6.810143848986583e-05, "loss": 0.013, "step": 20650 }, { "grad_norm": 0.16928890347480774, "learning_rate": 6.807060833562106e-05, "loss": 0.0146, "step": 20660 }, { "grad_norm": 0.1994473934173584, "learning_rate": 6.803977027667825e-05, "loss": 0.0128, "step": 20670 }, { "grad_norm": 0.1601887196302414, "learning_rate": 6.800892432652704e-05, "loss": 0.0102, "step": 20680 }, { "grad_norm": 0.20635958015918732, "learning_rate": 6.797807049866049e-05, "loss": 0.0112, "step": 20690 }, { "grad_norm": 0.19400694966316223, "learning_rate": 6.794720880657512e-05, "loss": 0.0113, "step": 20700 }, { "grad_norm": 0.19758723676204681, "learning_rate": 6.791633926377085e-05, "loss": 0.0115, "step": 20710 }, { "grad_norm": 0.5289905667304993, "learning_rate": 6.788546188375114e-05, "loss": 0.0102, "step": 20720 }, { "grad_norm": 0.1793040931224823, "learning_rate": 6.785457668002274e-05, "loss": 0.0123, "step": 20730 }, { "grad_norm": 0.22117726504802704, "learning_rate": 6.78236836660959e-05, "loss": 0.0136, "step": 20740 }, { "grad_norm": 0.21219073235988617, "learning_rate": 6.779278285548427e-05, "loss": 0.0107, "step": 20750 }, { "grad_norm": 0.18574275076389313, "learning_rate": 6.776187426170493e-05, "loss": 0.0123, "step": 20760 }, { "grad_norm": 0.3136957585811615, "learning_rate": 6.773095789827837e-05, "loss": 0.0147, "step": 20770 }, { "grad_norm": 1.4235671758651733, "learning_rate": 6.770003377872841e-05, "loss": 0.0094, "step": 20780 }, { "grad_norm": 0.16758455336093903, "learning_rate": 6.766910191658233e-05, "loss": 0.0106, "step": 20790 }, { "grad_norm": 0.14767402410507202, "learning_rate": 6.763816232537079e-05, "loss": 0.0106, "step": 20800 }, { "grad_norm": 0.18843477964401245, "learning_rate": 6.76072150186278e-05, "loss": 0.0108, "step": 20810 }, { "grad_norm": 0.19497498869895935, "learning_rate": 6.757626000989077e-05, "loss": 0.0111, "step": 20820 }, { "grad_norm": 0.19269883632659912, "learning_rate": 6.754529731270048e-05, "loss": 0.0116, "step": 20830 }, { "grad_norm": 0.18895797431468964, "learning_rate": 6.751432694060105e-05, "loss": 0.0105, "step": 20840 }, { "grad_norm": 0.20461240410804749, "learning_rate": 6.748334890713999e-05, "loss": 0.0103, "step": 20850 }, { "grad_norm": 0.23508486151695251, "learning_rate": 6.745236322586813e-05, "loss": 0.0107, "step": 20860 }, { "grad_norm": 0.18015411496162415, "learning_rate": 6.742136991033967e-05, "loss": 0.0095, "step": 20870 }, { "grad_norm": 0.3513264060020447, "learning_rate": 6.739036897411215e-05, "loss": 0.0104, "step": 20880 }, { "grad_norm": 0.16824714839458466, "learning_rate": 6.73593604307464e-05, "loss": 0.01, "step": 20890 }, { "grad_norm": 0.20410361886024475, "learning_rate": 6.732834429380663e-05, "loss": 0.009, "step": 20900 }, { "grad_norm": 0.15340951085090637, "learning_rate": 6.729732057686036e-05, "loss": 0.0118, "step": 20910 }, { "grad_norm": 0.3644082546234131, "learning_rate": 6.726628929347839e-05, "loss": 0.0109, "step": 20920 }, { "grad_norm": 0.5316066145896912, "learning_rate": 6.723525045723488e-05, "loss": 0.0101, "step": 20930 }, { "grad_norm": 0.5144352912902832, "learning_rate": 6.720420408170728e-05, "loss": 0.0146, "step": 20940 }, { "grad_norm": 0.32217922806739807, "learning_rate": 6.71731501804763e-05, "loss": 0.0111, "step": 20950 }, { "grad_norm": 0.30878889560699463, "learning_rate": 6.714208876712602e-05, "loss": 0.0121, "step": 20960 }, { "grad_norm": 0.19266033172607422, "learning_rate": 6.71110198552437e-05, "loss": 0.0101, "step": 20970 }, { "grad_norm": 0.2169681042432785, "learning_rate": 6.707994345841996e-05, "loss": 0.0122, "step": 20980 }, { "grad_norm": 0.3191181719303131, "learning_rate": 6.70488595902487e-05, "loss": 0.0115, "step": 20990 }, { "grad_norm": 0.8739351034164429, "learning_rate": 6.7017768264327e-05, "loss": 0.0124, "step": 21000 }, { "grad_norm": 0.28421416878700256, "learning_rate": 6.698666949425532e-05, "loss": 0.0124, "step": 21010 }, { "grad_norm": 0.25242820382118225, "learning_rate": 6.69555632936373e-05, "loss": 0.0119, "step": 21020 }, { "grad_norm": 0.17886190116405487, "learning_rate": 6.692444967607983e-05, "loss": 0.0105, "step": 21030 }, { "grad_norm": 0.18163129687309265, "learning_rate": 6.689332865519309e-05, "loss": 0.0116, "step": 21040 }, { "grad_norm": 0.24121111631393433, "learning_rate": 6.686220024459045e-05, "loss": 0.0125, "step": 21050 }, { "grad_norm": 0.195506289601326, "learning_rate": 6.683106445788856e-05, "loss": 0.011, "step": 21060 }, { "grad_norm": 1.2553223371505737, "learning_rate": 6.679992130870723e-05, "loss": 0.015, "step": 21070 }, { "grad_norm": 0.3607770800590515, "learning_rate": 6.676877081066957e-05, "loss": 0.0122, "step": 21080 }, { "grad_norm": 0.24044467508792877, "learning_rate": 6.673761297740186e-05, "loss": 0.012, "step": 21090 }, { "grad_norm": 0.3795176148414612, "learning_rate": 6.670644782253358e-05, "loss": 0.013, "step": 21100 }, { "grad_norm": 2.018319845199585, "learning_rate": 6.667527535969744e-05, "loss": 0.0138, "step": 21110 }, { "grad_norm": 0.24560301005840302, "learning_rate": 6.664409560252933e-05, "loss": 0.0127, "step": 21120 }, { "grad_norm": 0.18316195905208588, "learning_rate": 6.661290856466833e-05, "loss": 0.0111, "step": 21130 }, { "grad_norm": 0.24441339075565338, "learning_rate": 6.658171425975672e-05, "loss": 0.0133, "step": 21140 }, { "grad_norm": 0.18254970014095306, "learning_rate": 6.655051270143994e-05, "loss": 0.0122, "step": 21150 }, { "grad_norm": 0.21908028423786163, "learning_rate": 6.651930390336662e-05, "loss": 0.0118, "step": 21160 }, { "grad_norm": 0.22481977939605713, "learning_rate": 6.648808787918854e-05, "loss": 0.0123, "step": 21170 }, { "grad_norm": 0.8610631227493286, "learning_rate": 6.645686464256065e-05, "loss": 0.0118, "step": 21180 }, { "grad_norm": 0.22974419593811035, "learning_rate": 6.642563420714106e-05, "loss": 0.0129, "step": 21190 }, { "grad_norm": 0.3173369765281677, "learning_rate": 6.639439658659102e-05, "loss": 0.0144, "step": 21200 }, { "grad_norm": 0.25538983941078186, "learning_rate": 6.636315179457492e-05, "loss": 0.0096, "step": 21210 }, { "grad_norm": 0.2543008029460907, "learning_rate": 6.633189984476031e-05, "loss": 0.0122, "step": 21220 }, { "grad_norm": 0.21510539948940277, "learning_rate": 6.630064075081785e-05, "loss": 0.0128, "step": 21230 }, { "grad_norm": 0.5195950269699097, "learning_rate": 6.626937452642132e-05, "loss": 0.0125, "step": 21240 }, { "grad_norm": 0.14714357256889343, "learning_rate": 6.623810118524765e-05, "loss": 0.0106, "step": 21250 }, { "grad_norm": 0.25509047508239746, "learning_rate": 6.620682074097682e-05, "loss": 0.0106, "step": 21260 }, { "grad_norm": 0.2104186862707138, "learning_rate": 6.6175533207292e-05, "loss": 0.0118, "step": 21270 }, { "grad_norm": 0.5611180067062378, "learning_rate": 6.61442385978794e-05, "loss": 0.0104, "step": 21280 }, { "grad_norm": 0.22245410084724426, "learning_rate": 6.611293692642836e-05, "loss": 0.0115, "step": 21290 }, { "grad_norm": 0.4566667079925537, "learning_rate": 6.608162820663128e-05, "loss": 0.0115, "step": 21300 }, { "grad_norm": 0.16811858117580414, "learning_rate": 6.605031245218365e-05, "loss": 0.0107, "step": 21310 }, { "grad_norm": 0.19854339957237244, "learning_rate": 6.601898967678408e-05, "loss": 0.0105, "step": 21320 }, { "grad_norm": 0.7665351033210754, "learning_rate": 6.598765989413419e-05, "loss": 0.0098, "step": 21330 }, { "grad_norm": 0.1949780434370041, "learning_rate": 6.595632311793867e-05, "loss": 0.0128, "step": 21340 }, { "grad_norm": 0.20064480602741241, "learning_rate": 6.592497936190534e-05, "loss": 0.0121, "step": 21350 }, { "grad_norm": 0.21662206947803497, "learning_rate": 6.5893628639745e-05, "loss": 0.0106, "step": 21360 }, { "grad_norm": 0.12926851212978363, "learning_rate": 6.586227096517151e-05, "loss": 0.0114, "step": 21370 }, { "grad_norm": 0.15963371098041534, "learning_rate": 6.58309063519018e-05, "loss": 0.0089, "step": 21380 }, { "grad_norm": 0.29133084416389465, "learning_rate": 6.579953481365582e-05, "loss": 0.0117, "step": 21390 }, { "grad_norm": 0.29398056864738464, "learning_rate": 6.576815636415652e-05, "loss": 0.0123, "step": 21400 }, { "grad_norm": 0.18927304446697235, "learning_rate": 6.573677101712991e-05, "loss": 0.0106, "step": 21410 }, { "grad_norm": 1.5046077966690063, "learning_rate": 6.570537878630502e-05, "loss": 0.0114, "step": 21420 }, { "grad_norm": 0.19987213611602783, "learning_rate": 6.56739796854139e-05, "loss": 0.0117, "step": 21430 }, { "grad_norm": 0.17478060722351074, "learning_rate": 6.564257372819153e-05, "loss": 0.0121, "step": 21440 }, { "grad_norm": 0.18061795830726624, "learning_rate": 6.561116092837596e-05, "loss": 0.0149, "step": 21450 }, { "grad_norm": 0.19423669576644897, "learning_rate": 6.557974129970825e-05, "loss": 0.0118, "step": 21460 }, { "grad_norm": 0.17649099230766296, "learning_rate": 6.554831485593239e-05, "loss": 0.011, "step": 21470 }, { "grad_norm": 0.18013258278369904, "learning_rate": 6.551688161079534e-05, "loss": 0.0104, "step": 21480 }, { "grad_norm": 0.2868078947067261, "learning_rate": 6.548544157804712e-05, "loss": 0.0096, "step": 21490 }, { "grad_norm": 0.12987716495990753, "learning_rate": 6.545399477144065e-05, "loss": 0.0111, "step": 21500 }, { "grad_norm": 0.32770276069641113, "learning_rate": 6.542254120473184e-05, "loss": 0.0096, "step": 21510 }, { "grad_norm": 0.2047567367553711, "learning_rate": 6.539108089167953e-05, "loss": 0.0116, "step": 21520 }, { "grad_norm": 0.1765434741973877, "learning_rate": 6.535961384604554e-05, "loss": 0.0123, "step": 21530 }, { "grad_norm": 0.19138361513614655, "learning_rate": 6.532814008159461e-05, "loss": 0.0113, "step": 21540 }, { "grad_norm": 0.17549113929271698, "learning_rate": 6.529665961209446e-05, "loss": 0.0101, "step": 21550 }, { "grad_norm": 0.31626731157302856, "learning_rate": 6.52651724513157e-05, "loss": 0.0097, "step": 21560 }, { "grad_norm": 0.2528264820575714, "learning_rate": 6.523367861303189e-05, "loss": 0.0105, "step": 21570 }, { "grad_norm": 0.14255593717098236, "learning_rate": 6.520217811101951e-05, "loss": 0.0119, "step": 21580 }, { "grad_norm": 0.21261869370937347, "learning_rate": 6.517067095905793e-05, "loss": 0.0117, "step": 21590 }, { "grad_norm": 0.218321293592453, "learning_rate": 6.513915717092948e-05, "loss": 0.0128, "step": 21600 }, { "grad_norm": 0.20660638809204102, "learning_rate": 6.510763676041935e-05, "loss": 0.0118, "step": 21610 }, { "grad_norm": 0.2033044695854187, "learning_rate": 6.50761097413156e-05, "loss": 0.0103, "step": 21620 }, { "grad_norm": 0.20470041036605835, "learning_rate": 6.504457612740928e-05, "loss": 0.0111, "step": 21630 }, { "grad_norm": 0.1335742324590683, "learning_rate": 6.501303593249423e-05, "loss": 0.0108, "step": 21640 }, { "grad_norm": 0.6427335739135742, "learning_rate": 6.49814891703672e-05, "loss": 0.0086, "step": 21650 }, { "grad_norm": 0.7094137668609619, "learning_rate": 6.494993585482783e-05, "loss": 0.0118, "step": 21660 }, { "grad_norm": 0.21027374267578125, "learning_rate": 6.49183759996786e-05, "loss": 0.011, "step": 21670 }, { "grad_norm": 0.48267731070518494, "learning_rate": 6.488680961872486e-05, "loss": 0.0122, "step": 21680 }, { "grad_norm": 0.27886268496513367, "learning_rate": 6.485523672577485e-05, "loss": 0.0128, "step": 21690 }, { "grad_norm": 0.15376032888889313, "learning_rate": 6.482365733463959e-05, "loss": 0.0113, "step": 21700 }, { "grad_norm": 1.899855613708496, "learning_rate": 6.4792071459133e-05, "loss": 0.011, "step": 21710 }, { "grad_norm": 0.302951842546463, "learning_rate": 6.476047911307179e-05, "loss": 0.0148, "step": 21720 }, { "grad_norm": 0.20329156517982483, "learning_rate": 6.472888031027556e-05, "loss": 0.013, "step": 21730 }, { "grad_norm": 0.24289916455745697, "learning_rate": 6.469727506456666e-05, "loss": 0.0101, "step": 21740 }, { "grad_norm": 0.21325457096099854, "learning_rate": 6.466566338977034e-05, "loss": 0.0114, "step": 21750 }, { "grad_norm": 0.22808949649333954, "learning_rate": 6.46340452997146e-05, "loss": 0.0109, "step": 21760 }, { "grad_norm": 0.16442137956619263, "learning_rate": 6.460242080823025e-05, "loss": 0.014, "step": 21770 }, { "grad_norm": 0.3650147318840027, "learning_rate": 6.457078992915093e-05, "loss": 0.0105, "step": 21780 }, { "grad_norm": 0.4457665979862213, "learning_rate": 6.453915267631308e-05, "loss": 0.0105, "step": 21790 }, { "grad_norm": 0.2614070475101471, "learning_rate": 6.450750906355589e-05, "loss": 0.0109, "step": 21800 }, { "grad_norm": 0.15073329210281372, "learning_rate": 6.447585910472134e-05, "loss": 0.0108, "step": 21810 }, { "grad_norm": 0.17718759179115295, "learning_rate": 6.444420281365424e-05, "loss": 0.0087, "step": 21820 }, { "grad_norm": 0.30388227105140686, "learning_rate": 6.441254020420208e-05, "loss": 0.0111, "step": 21830 }, { "grad_norm": 0.18142040073871613, "learning_rate": 6.438087129021522e-05, "loss": 0.0107, "step": 21840 }, { "grad_norm": 0.3063833713531494, "learning_rate": 6.434919608554663e-05, "loss": 0.0114, "step": 21850 }, { "grad_norm": 0.21649089455604553, "learning_rate": 6.43175146040522e-05, "loss": 0.0104, "step": 21860 }, { "grad_norm": 0.1582442820072174, "learning_rate": 6.428582685959045e-05, "loss": 0.0108, "step": 21870 }, { "grad_norm": 4.338150501251221, "learning_rate": 6.42541328660227e-05, "loss": 0.013, "step": 21880 }, { "grad_norm": 0.212630495429039, "learning_rate": 6.422243263721293e-05, "loss": 0.0104, "step": 21890 }, { "grad_norm": 0.1568959355354309, "learning_rate": 6.419072618702794e-05, "loss": 0.0119, "step": 21900 }, { "grad_norm": 0.1982341855764389, "learning_rate": 6.41590135293372e-05, "loss": 0.0094, "step": 21910 }, { "grad_norm": 0.20034368336200714, "learning_rate": 6.41272946780129e-05, "loss": 0.0085, "step": 21920 }, { "grad_norm": 0.2278696447610855, "learning_rate": 6.409556964692996e-05, "loss": 0.0103, "step": 21930 }, { "grad_norm": 0.16058224439620972, "learning_rate": 6.406383844996595e-05, "loss": 0.0118, "step": 21940 }, { "grad_norm": 0.3192102313041687, "learning_rate": 6.403210110100121e-05, "loss": 0.0099, "step": 21950 }, { "grad_norm": 0.34254953265190125, "learning_rate": 6.400035761391871e-05, "loss": 0.0093, "step": 21960 }, { "grad_norm": 1.2224074602127075, "learning_rate": 6.396860800260416e-05, "loss": 0.0117, "step": 21970 }, { "grad_norm": 0.28976792097091675, "learning_rate": 6.393685228094588e-05, "loss": 0.0113, "step": 21980 }, { "grad_norm": 0.17884276807308197, "learning_rate": 6.390509046283493e-05, "loss": 0.0097, "step": 21990 }, { "grad_norm": 0.17360566556453705, "learning_rate": 6.387332256216498e-05, "loss": 0.01, "step": 22000 }, { "grad_norm": 0.17030391097068787, "learning_rate": 6.384154859283243e-05, "loss": 0.01, "step": 22010 }, { "grad_norm": 0.1614316999912262, "learning_rate": 6.380976856873627e-05, "loss": 0.01, "step": 22020 }, { "grad_norm": 0.16832570731639862, "learning_rate": 6.377798250377815e-05, "loss": 0.0095, "step": 22030 }, { "grad_norm": 0.18803389370441437, "learning_rate": 6.374619041186238e-05, "loss": 0.0129, "step": 22040 }, { "grad_norm": 0.2002236247062683, "learning_rate": 6.371439230689592e-05, "loss": 0.0095, "step": 22050 }, { "grad_norm": 0.20670288801193237, "learning_rate": 6.368258820278833e-05, "loss": 0.0096, "step": 22060 }, { "grad_norm": 0.18384571373462677, "learning_rate": 6.365077811345182e-05, "loss": 0.0095, "step": 22070 }, { "grad_norm": 0.1390002816915512, "learning_rate": 6.361896205280117e-05, "loss": 0.0091, "step": 22080 }, { "grad_norm": 0.39168038964271545, "learning_rate": 6.358714003475384e-05, "loss": 0.0089, "step": 22090 }, { "grad_norm": 0.17104263603687286, "learning_rate": 6.355531207322983e-05, "loss": 0.0125, "step": 22100 }, { "grad_norm": 0.20200836658477783, "learning_rate": 6.35234781821518e-05, "loss": 0.0102, "step": 22110 }, { "grad_norm": 0.1352982074022293, "learning_rate": 6.349163837544497e-05, "loss": 0.0107, "step": 22120 }, { "grad_norm": 0.23813191056251526, "learning_rate": 6.345979266703714e-05, "loss": 0.0126, "step": 22130 }, { "grad_norm": 0.20007853209972382, "learning_rate": 6.342794107085872e-05, "loss": 0.0105, "step": 22140 }, { "grad_norm": 0.16892839968204498, "learning_rate": 6.339608360084265e-05, "loss": 0.0108, "step": 22150 }, { "grad_norm": 0.5139213800430298, "learning_rate": 6.336422027092452e-05, "loss": 0.0095, "step": 22160 }, { "grad_norm": 0.22896748781204224, "learning_rate": 6.333235109504241e-05, "loss": 0.0115, "step": 22170 }, { "grad_norm": 0.21068616211414337, "learning_rate": 6.330047608713695e-05, "loss": 0.0119, "step": 22180 }, { "grad_norm": 0.2622199058532715, "learning_rate": 6.32685952611514e-05, "loss": 0.0115, "step": 22190 }, { "grad_norm": 0.19632381200790405, "learning_rate": 6.323670863103149e-05, "loss": 0.0106, "step": 22200 }, { "grad_norm": 0.15681211650371552, "learning_rate": 6.32048162107255e-05, "loss": 0.0112, "step": 22210 }, { "grad_norm": 0.2939480245113373, "learning_rate": 6.317291801418431e-05, "loss": 0.012, "step": 22220 }, { "grad_norm": 0.2387717217206955, "learning_rate": 6.31410140553612e-05, "loss": 0.0101, "step": 22230 }, { "grad_norm": 0.1631423979997635, "learning_rate": 6.310910434821211e-05, "loss": 0.0097, "step": 22240 }, { "grad_norm": 6.92457389831543, "learning_rate": 6.307718890669541e-05, "loss": 0.0151, "step": 22250 }, { "grad_norm": 0.2382858246564865, "learning_rate": 6.304526774477198e-05, "loss": 0.0126, "step": 22260 }, { "grad_norm": 0.16401971876621246, "learning_rate": 6.301334087640526e-05, "loss": 0.011, "step": 22270 }, { "grad_norm": 0.13619166612625122, "learning_rate": 6.298140831556112e-05, "loss": 0.0097, "step": 22280 }, { "grad_norm": 0.160292387008667, "learning_rate": 6.294947007620794e-05, "loss": 0.0096, "step": 22290 }, { "grad_norm": 0.18508732318878174, "learning_rate": 6.291752617231661e-05, "loss": 0.0099, "step": 22300 }, { "grad_norm": 0.3362847566604614, "learning_rate": 6.288557661786047e-05, "loss": 0.0107, "step": 22310 }, { "grad_norm": 0.1618131697177887, "learning_rate": 6.285362142681534e-05, "loss": 0.0095, "step": 22320 }, { "grad_norm": 0.19082175195217133, "learning_rate": 6.28216606131595e-05, "loss": 0.0117, "step": 22330 }, { "grad_norm": 0.35065993666648865, "learning_rate": 6.27896941908737e-05, "loss": 0.0106, "step": 22340 }, { "grad_norm": 0.1716901957988739, "learning_rate": 6.275772217394115e-05, "loss": 0.0094, "step": 22350 }, { "grad_norm": 0.16537325084209442, "learning_rate": 6.27257445763475e-05, "loss": 0.0114, "step": 22360 }, { "grad_norm": 0.16990266740322113, "learning_rate": 6.269376141208082e-05, "loss": 0.0099, "step": 22370 }, { "grad_norm": 0.2817721962928772, "learning_rate": 6.266177269513162e-05, "loss": 0.0091, "step": 22380 }, { "grad_norm": 0.3521973192691803, "learning_rate": 6.262977843949289e-05, "loss": 0.0119, "step": 22390 }, { "grad_norm": 0.16404277086257935, "learning_rate": 6.259777865915997e-05, "loss": 0.0117, "step": 22400 }, { "grad_norm": 0.15084776282310486, "learning_rate": 6.256577336813069e-05, "loss": 0.01, "step": 22410 }, { "grad_norm": 0.12697792053222656, "learning_rate": 6.253376258040521e-05, "loss": 0.0118, "step": 22420 }, { "grad_norm": 0.23337775468826294, "learning_rate": 6.250174630998616e-05, "loss": 0.0132, "step": 22430 }, { "grad_norm": 0.18455515801906586, "learning_rate": 6.246972457087854e-05, "loss": 0.0107, "step": 22440 }, { "grad_norm": 0.2553003430366516, "learning_rate": 6.243769737708974e-05, "loss": 0.0101, "step": 22450 }, { "grad_norm": 0.25333815813064575, "learning_rate": 6.240566474262955e-05, "loss": 0.0094, "step": 22460 }, { "grad_norm": 0.19594399631023407, "learning_rate": 6.237362668151012e-05, "loss": 0.0092, "step": 22470 }, { "grad_norm": 0.256695419549942, "learning_rate": 6.2341583207746e-05, "loss": 0.0085, "step": 22480 }, { "grad_norm": 0.21744634211063385, "learning_rate": 6.23095343353541e-05, "loss": 0.0097, "step": 22490 }, { "grad_norm": 0.19665947556495667, "learning_rate": 6.227748007835366e-05, "loss": 0.011, "step": 22500 }, { "grad_norm": 0.7127801775932312, "learning_rate": 6.224542045076632e-05, "loss": 0.0102, "step": 22510 }, { "grad_norm": 0.13767512142658234, "learning_rate": 6.221335546661606e-05, "loss": 0.0101, "step": 22520 }, { "grad_norm": 0.32080745697021484, "learning_rate": 6.218128513992917e-05, "loss": 0.0106, "step": 22530 }, { "grad_norm": 0.2224815934896469, "learning_rate": 6.214920948473433e-05, "loss": 0.0108, "step": 22540 }, { "grad_norm": 2.225083589553833, "learning_rate": 6.211712851506249e-05, "loss": 0.0122, "step": 22550 }, { "grad_norm": 0.23771217465400696, "learning_rate": 6.208504224494696e-05, "loss": 0.0117, "step": 22560 }, { "grad_norm": 0.2174556851387024, "learning_rate": 6.205295068842338e-05, "loss": 0.0097, "step": 22570 }, { "grad_norm": 1.6079559326171875, "learning_rate": 6.202085385952968e-05, "loss": 0.0096, "step": 22580 }, { "grad_norm": 2.1215081214904785, "learning_rate": 6.19887517723061e-05, "loss": 0.0098, "step": 22590 }, { "grad_norm": 0.2870337963104248, "learning_rate": 6.19566444407952e-05, "loss": 0.0103, "step": 22600 }, { "grad_norm": 0.1807693988084793, "learning_rate": 6.192453187904178e-05, "loss": 0.0106, "step": 22610 }, { "grad_norm": 0.10826466232538223, "learning_rate": 6.189241410109303e-05, "loss": 0.0078, "step": 22620 }, { "grad_norm": 0.22057375311851501, "learning_rate": 6.186029112099829e-05, "loss": 0.0103, "step": 22630 }, { "grad_norm": 0.3282719552516937, "learning_rate": 6.182816295280926e-05, "loss": 0.01, "step": 22640 }, { "grad_norm": 0.1525106430053711, "learning_rate": 6.179602961057993e-05, "loss": 0.0129, "step": 22650 }, { "grad_norm": 0.21601001918315887, "learning_rate": 6.176389110836647e-05, "loss": 0.0096, "step": 22660 }, { "grad_norm": 0.3090600073337555, "learning_rate": 6.173174746022738e-05, "loss": 0.0127, "step": 22670 }, { "grad_norm": 0.594753086566925, "learning_rate": 6.169959868022338e-05, "loss": 0.0111, "step": 22680 }, { "grad_norm": 0.14294902980327606, "learning_rate": 6.166744478241742e-05, "loss": 0.0111, "step": 22690 }, { "grad_norm": 0.20268921554088593, "learning_rate": 6.163528578087473e-05, "loss": 0.0099, "step": 22700 }, { "grad_norm": 0.11615379899740219, "learning_rate": 6.160312168966274e-05, "loss": 0.0103, "step": 22710 }, { "grad_norm": 0.21075734496116638, "learning_rate": 6.157095252285112e-05, "loss": 0.01, "step": 22720 }, { "grad_norm": 0.5470455288887024, "learning_rate": 6.153877829451174e-05, "loss": 0.0092, "step": 22730 }, { "grad_norm": 0.2306608259677887, "learning_rate": 6.150659901871872e-05, "loss": 0.0102, "step": 22740 }, { "grad_norm": 0.21830575168132782, "learning_rate": 6.147441470954836e-05, "loss": 0.0093, "step": 22750 }, { "grad_norm": 0.7966911196708679, "learning_rate": 6.144222538107918e-05, "loss": 0.0094, "step": 22760 }, { "grad_norm": 0.3955439031124115, "learning_rate": 6.141003104739188e-05, "loss": 0.0118, "step": 22770 }, { "grad_norm": 0.2521211504936218, "learning_rate": 6.137783172256935e-05, "loss": 0.0115, "step": 22780 }, { "grad_norm": 0.19080717861652374, "learning_rate": 6.134562742069669e-05, "loss": 0.0097, "step": 22790 }, { "grad_norm": 0.21954353153705597, "learning_rate": 6.131341815586111e-05, "loss": 0.009, "step": 22800 }, { "grad_norm": 0.32199713587760925, "learning_rate": 6.128120394215209e-05, "loss": 0.0122, "step": 22810 }, { "grad_norm": 1.3092120885849, "learning_rate": 6.124898479366116e-05, "loss": 0.0098, "step": 22820 }, { "grad_norm": 0.2626214027404785, "learning_rate": 6.121676072448214e-05, "loss": 0.0105, "step": 22830 }, { "grad_norm": 0.20028086006641388, "learning_rate": 6.118453174871089e-05, "loss": 0.0108, "step": 22840 }, { "grad_norm": 0.3317696154117584, "learning_rate": 6.115229788044548e-05, "loss": 0.0116, "step": 22850 }, { "grad_norm": 0.19654372334480286, "learning_rate": 6.112005913378607e-05, "loss": 0.0106, "step": 22860 }, { "grad_norm": 0.2662537395954132, "learning_rate": 6.1087815522835e-05, "loss": 0.0104, "step": 22870 }, { "grad_norm": 0.41485264897346497, "learning_rate": 6.105556706169673e-05, "loss": 0.0124, "step": 22880 }, { "grad_norm": 0.1569453626871109, "learning_rate": 6.102331376447782e-05, "loss": 0.0085, "step": 22890 }, { "grad_norm": 0.2150498330593109, "learning_rate": 6.099105564528697e-05, "loss": 0.0108, "step": 22900 }, { "grad_norm": 0.17874419689178467, "learning_rate": 6.095879271823495e-05, "loss": 0.01, "step": 22910 }, { "grad_norm": 0.21450363099575043, "learning_rate": 6.09265249974347e-05, "loss": 0.0096, "step": 22920 }, { "grad_norm": 0.1731007993221283, "learning_rate": 6.089425249700118e-05, "loss": 0.0093, "step": 22930 }, { "grad_norm": 0.19621311128139496, "learning_rate": 6.0861975231051505e-05, "loss": 0.0095, "step": 22940 }, { "grad_norm": 0.3864806294441223, "learning_rate": 6.082969321370484e-05, "loss": 0.0103, "step": 22950 }, { "grad_norm": 0.21273179352283478, "learning_rate": 6.079740645908243e-05, "loss": 0.0112, "step": 22960 }, { "grad_norm": 0.3039936423301697, "learning_rate": 6.07651149813076e-05, "loss": 0.0117, "step": 22970 }, { "grad_norm": 0.18014569580554962, "learning_rate": 6.073281879450574e-05, "loss": 0.01, "step": 22980 }, { "grad_norm": 0.14216934144496918, "learning_rate": 6.070051791280431e-05, "loss": 0.0081, "step": 22990 }, { "grad_norm": 0.6199039816856384, "learning_rate": 6.066821235033281e-05, "loss": 0.0116, "step": 23000 }, { "grad_norm": 0.1451219618320465, "learning_rate": 6.063590212122277e-05, "loss": 0.0103, "step": 23010 }, { "grad_norm": 0.13974125683307648, "learning_rate": 6.0603587239607794e-05, "loss": 0.0098, "step": 23020 }, { "grad_norm": 0.1306496560573578, "learning_rate": 6.057126771962353e-05, "loss": 0.0095, "step": 23030 }, { "grad_norm": 0.6470075249671936, "learning_rate": 6.05389435754076e-05, "loss": 0.0113, "step": 23040 }, { "grad_norm": 0.23407432436943054, "learning_rate": 6.0506614821099705e-05, "loss": 0.0091, "step": 23050 }, { "grad_norm": 0.20707949995994568, "learning_rate": 6.047428147084153e-05, "loss": 0.0106, "step": 23060 }, { "grad_norm": 0.2029084712266922, "learning_rate": 6.044194353877679e-05, "loss": 0.0111, "step": 23070 }, { "grad_norm": 0.15924882888793945, "learning_rate": 6.04096010390512e-05, "loss": 0.0126, "step": 23080 }, { "grad_norm": 1.0185987949371338, "learning_rate": 6.0377253985812445e-05, "loss": 0.0105, "step": 23090 }, { "grad_norm": 0.19661301374435425, "learning_rate": 6.0344902393210235e-05, "loss": 0.012, "step": 23100 }, { "grad_norm": 0.2757243812084198, "learning_rate": 6.031254627539625e-05, "loss": 0.0108, "step": 23110 }, { "grad_norm": 1.8909153938293457, "learning_rate": 6.028018564652418e-05, "loss": 0.0104, "step": 23120 }, { "grad_norm": 0.17092251777648926, "learning_rate": 6.024782052074964e-05, "loss": 0.0102, "step": 23130 }, { "grad_norm": 0.37384024262428284, "learning_rate": 6.021545091223026e-05, "loss": 0.0094, "step": 23140 }, { "grad_norm": 0.23190554976463318, "learning_rate": 6.0183076835125586e-05, "loss": 0.0105, "step": 23150 }, { "grad_norm": 0.1556633859872818, "learning_rate": 6.015069830359714e-05, "loss": 0.0103, "step": 23160 }, { "grad_norm": 0.2434571236371994, "learning_rate": 6.0118315331808384e-05, "loss": 0.01, "step": 23170 }, { "grad_norm": 2.6681172847747803, "learning_rate": 6.0085927933924745e-05, "loss": 0.0101, "step": 23180 }, { "grad_norm": 0.25078022480010986, "learning_rate": 6.0053536124113585e-05, "loss": 0.0129, "step": 23190 }, { "grad_norm": 0.21331433951854706, "learning_rate": 6.0021139916544144e-05, "loss": 0.015, "step": 23200 }, { "grad_norm": 0.23799993097782135, "learning_rate": 5.998873932538764e-05, "loss": 0.0096, "step": 23210 }, { "grad_norm": 1.068767786026001, "learning_rate": 5.995633436481721e-05, "loss": 0.01, "step": 23220 }, { "grad_norm": 1.715530276298523, "learning_rate": 5.9923925049007855e-05, "loss": 0.0101, "step": 23230 }, { "grad_norm": 0.19330526888370514, "learning_rate": 5.989151139213656e-05, "loss": 0.0112, "step": 23240 }, { "grad_norm": 0.17715680599212646, "learning_rate": 5.9859093408382094e-05, "loss": 0.0115, "step": 23250 }, { "grad_norm": 0.14903758466243744, "learning_rate": 5.982667111192523e-05, "loss": 0.0093, "step": 23260 }, { "grad_norm": 0.30895230174064636, "learning_rate": 5.979424451694857e-05, "loss": 0.0099, "step": 23270 }, { "grad_norm": 0.189705491065979, "learning_rate": 5.97618136376366e-05, "loss": 0.0101, "step": 23280 }, { "grad_norm": 0.43011099100112915, "learning_rate": 5.97293784881757e-05, "loss": 0.0093, "step": 23290 }, { "grad_norm": 0.1796337515115738, "learning_rate": 5.969693908275411e-05, "loss": 0.0091, "step": 23300 }, { "grad_norm": 0.22726674377918243, "learning_rate": 5.9664495435561916e-05, "loss": 0.0101, "step": 23310 }, { "grad_norm": 0.14355120062828064, "learning_rate": 5.963204756079108e-05, "loss": 0.0109, "step": 23320 }, { "grad_norm": 0.35256460309028625, "learning_rate": 5.959959547263538e-05, "loss": 0.0099, "step": 23330 }, { "grad_norm": 0.25717318058013916, "learning_rate": 5.95671391852905e-05, "loss": 0.0095, "step": 23340 }, { "grad_norm": 0.214140385389328, "learning_rate": 5.95346787129539e-05, "loss": 0.0096, "step": 23350 }, { "grad_norm": 0.20906834304332733, "learning_rate": 5.950221406982489e-05, "loss": 0.0108, "step": 23360 }, { "grad_norm": 0.16003303229808807, "learning_rate": 5.9469745270104625e-05, "loss": 0.0114, "step": 23370 }, { "grad_norm": 0.24673724174499512, "learning_rate": 5.9437272327996064e-05, "loss": 0.012, "step": 23380 }, { "grad_norm": 0.20382758975028992, "learning_rate": 5.940479525770395e-05, "loss": 0.009, "step": 23390 }, { "grad_norm": 0.278106689453125, "learning_rate": 5.9372314073434876e-05, "loss": 0.0093, "step": 23400 }, { "grad_norm": 0.18261155486106873, "learning_rate": 5.93398287893972e-05, "loss": 0.01, "step": 23410 }, { "grad_norm": 0.19290849566459656, "learning_rate": 5.930733941980111e-05, "loss": 0.0128, "step": 23420 }, { "grad_norm": 1.8868759870529175, "learning_rate": 5.927484597885855e-05, "loss": 0.0128, "step": 23430 }, { "grad_norm": 0.22303809225559235, "learning_rate": 5.924234848078325e-05, "loss": 0.01, "step": 23440 }, { "grad_norm": 0.1478397101163864, "learning_rate": 5.9209846939790715e-05, "loss": 0.0107, "step": 23450 }, { "grad_norm": 0.2926880717277527, "learning_rate": 5.917734137009824e-05, "loss": 0.0112, "step": 23460 }, { "grad_norm": 0.1935516744852066, "learning_rate": 5.914483178592485e-05, "loss": 0.0115, "step": 23470 }, { "grad_norm": 0.14962728321552277, "learning_rate": 5.911231820149136e-05, "loss": 0.0086, "step": 23480 }, { "grad_norm": 1.1678258180618286, "learning_rate": 5.907980063102031e-05, "loss": 0.0078, "step": 23490 }, { "grad_norm": 1.0110116004943848, "learning_rate": 5.904727908873597e-05, "loss": 0.0101, "step": 23500 }, { "grad_norm": 0.19777432084083557, "learning_rate": 5.90147535888644e-05, "loss": 0.0102, "step": 23510 }, { "grad_norm": 0.9514297842979431, "learning_rate": 5.898222414563333e-05, "loss": 0.0129, "step": 23520 }, { "grad_norm": 0.18550172448158264, "learning_rate": 5.8949690773272256e-05, "loss": 0.0111, "step": 23530 }, { "grad_norm": 0.22558069229125977, "learning_rate": 5.891715348601239e-05, "loss": 0.0116, "step": 23540 }, { "grad_norm": 0.1545521467924118, "learning_rate": 5.888461229808663e-05, "loss": 0.0097, "step": 23550 }, { "grad_norm": 0.14468134939670563, "learning_rate": 5.885206722372959e-05, "loss": 0.0097, "step": 23560 }, { "grad_norm": 0.31419095396995544, "learning_rate": 5.881951827717763e-05, "loss": 0.009, "step": 23570 }, { "grad_norm": 0.39454853534698486, "learning_rate": 5.878696547266872e-05, "loss": 0.0125, "step": 23580 }, { "grad_norm": 0.1723177284002304, "learning_rate": 5.875440882444259e-05, "loss": 0.0107, "step": 23590 }, { "grad_norm": 0.15723024308681488, "learning_rate": 5.8721848346740617e-05, "loss": 0.0107, "step": 23600 }, { "grad_norm": 0.16979746520519257, "learning_rate": 5.868928405380585e-05, "loss": 0.0093, "step": 23610 }, { "grad_norm": 0.14721544086933136, "learning_rate": 5.865671595988304e-05, "loss": 0.0088, "step": 23620 }, { "grad_norm": 0.1553332507610321, "learning_rate": 5.862414407921857e-05, "loss": 0.0095, "step": 23630 }, { "grad_norm": 0.5166021585464478, "learning_rate": 5.8591568426060475e-05, "loss": 0.0093, "step": 23640 }, { "grad_norm": 0.12472393363714218, "learning_rate": 5.8558989014658464e-05, "loss": 0.0086, "step": 23650 }, { "grad_norm": 0.21930940449237823, "learning_rate": 5.8526405859263855e-05, "loss": 0.0084, "step": 23660 }, { "grad_norm": 0.21742869913578033, "learning_rate": 5.8493818974129667e-05, "loss": 0.01, "step": 23670 }, { "grad_norm": 0.24750259518623352, "learning_rate": 5.846122837351048e-05, "loss": 0.008, "step": 23680 }, { "grad_norm": 1.0633939504623413, "learning_rate": 5.842863407166255e-05, "loss": 0.0099, "step": 23690 }, { "grad_norm": 0.8240009546279907, "learning_rate": 5.83960360828437e-05, "loss": 0.0092, "step": 23700 }, { "grad_norm": 0.12974737584590912, "learning_rate": 5.836343442131343e-05, "loss": 0.008, "step": 23710 }, { "grad_norm": 0.4182237386703491, "learning_rate": 5.833082910133281e-05, "loss": 0.0114, "step": 23720 }, { "grad_norm": 0.16770495474338531, "learning_rate": 5.82982201371645e-05, "loss": 0.0092, "step": 23730 }, { "grad_norm": 0.16548994183540344, "learning_rate": 5.826560754307279e-05, "loss": 0.011, "step": 23740 }, { "grad_norm": 0.26822224259376526, "learning_rate": 5.8232991333323516e-05, "loss": 0.0094, "step": 23750 }, { "grad_norm": 0.1810222715139389, "learning_rate": 5.8200371522184116e-05, "loss": 0.0093, "step": 23760 }, { "grad_norm": 0.2479417324066162, "learning_rate": 5.816774812392362e-05, "loss": 0.0097, "step": 23770 }, { "grad_norm": 0.25804954767227173, "learning_rate": 5.8135121152812614e-05, "loss": 0.0096, "step": 23780 }, { "grad_norm": 0.1512289196252823, "learning_rate": 5.8102490623123204e-05, "loss": 0.0086, "step": 23790 }, { "grad_norm": 0.19207142293453217, "learning_rate": 5.806985654912914e-05, "loss": 0.0093, "step": 23800 }, { "grad_norm": 0.24797223508358002, "learning_rate": 5.8037218945105644e-05, "loss": 0.0096, "step": 23810 }, { "grad_norm": 0.1895029991865158, "learning_rate": 5.800457782532953e-05, "loss": 0.0107, "step": 23820 }, { "grad_norm": 0.20826302468776703, "learning_rate": 5.797193320407912e-05, "loss": 0.0095, "step": 23830 }, { "grad_norm": 0.30058640241622925, "learning_rate": 5.793928509563431e-05, "loss": 0.0105, "step": 23840 }, { "grad_norm": 0.15729452669620514, "learning_rate": 5.790663351427644e-05, "loss": 0.0098, "step": 23850 }, { "grad_norm": 0.18413738906383514, "learning_rate": 5.7873978474288484e-05, "loss": 0.01, "step": 23860 }, { "grad_norm": 0.14190930128097534, "learning_rate": 5.78413199899548e-05, "loss": 0.0095, "step": 23870 }, { "grad_norm": 0.3930024206638336, "learning_rate": 5.780865807556136e-05, "loss": 0.0086, "step": 23880 }, { "grad_norm": 0.20393230020999908, "learning_rate": 5.777599274539559e-05, "loss": 0.0096, "step": 23890 }, { "grad_norm": 0.20524215698242188, "learning_rate": 5.77433240137464e-05, "loss": 0.0109, "step": 23900 }, { "grad_norm": 0.3412962555885315, "learning_rate": 5.771065189490422e-05, "loss": 0.0093, "step": 23910 }, { "grad_norm": 0.1382097750902176, "learning_rate": 5.767797640316095e-05, "loss": 0.0078, "step": 23920 }, { "grad_norm": 0.3163672089576721, "learning_rate": 5.764529755280994e-05, "loss": 0.0087, "step": 23930 }, { "grad_norm": 0.23066729307174683, "learning_rate": 5.761261535814603e-05, "loss": 0.0091, "step": 23940 }, { "grad_norm": 0.1899222433567047, "learning_rate": 5.757992983346554e-05, "loss": 0.0105, "step": 23950 }, { "grad_norm": 0.16780565679073334, "learning_rate": 5.754724099306621e-05, "loss": 0.0094, "step": 23960 }, { "grad_norm": 0.2588163912296295, "learning_rate": 5.751454885124727e-05, "loss": 0.0077, "step": 23970 }, { "grad_norm": 0.2479761391878128, "learning_rate": 5.748185342230935e-05, "loss": 0.0076, "step": 23980 }, { "grad_norm": 0.8221665024757385, "learning_rate": 5.744915472055456e-05, "loss": 0.0083, "step": 23990 }, { "grad_norm": 0.5029041171073914, "learning_rate": 5.741645276028642e-05, "loss": 0.0082, "step": 24000 }, { "grad_norm": 0.21964219212532043, "learning_rate": 5.7383747555809873e-05, "loss": 0.0097, "step": 24010 }, { "grad_norm": 0.18128837645053864, "learning_rate": 5.7351039121431294e-05, "loss": 0.0102, "step": 24020 }, { "grad_norm": 0.17208044230937958, "learning_rate": 5.7318327471458454e-05, "loss": 0.0104, "step": 24030 }, { "grad_norm": 0.22594547271728516, "learning_rate": 5.728561262020054e-05, "loss": 0.0111, "step": 24040 }, { "grad_norm": 0.18817824125289917, "learning_rate": 5.725289458196814e-05, "loss": 0.0099, "step": 24050 }, { "grad_norm": 0.18550992012023926, "learning_rate": 5.722017337107324e-05, "loss": 0.0111, "step": 24060 }, { "grad_norm": 0.17318086326122284, "learning_rate": 5.7187449001829206e-05, "loss": 0.0089, "step": 24070 }, { "grad_norm": 0.12462971359491348, "learning_rate": 5.715472148855081e-05, "loss": 0.0085, "step": 24080 }, { "grad_norm": 0.214408740401268, "learning_rate": 5.712199084555414e-05, "loss": 0.0114, "step": 24090 }, { "grad_norm": 0.1782691478729248, "learning_rate": 5.7089257087156734e-05, "loss": 0.0095, "step": 24100 }, { "grad_norm": 0.18746784329414368, "learning_rate": 5.70565202276774e-05, "loss": 0.0096, "step": 24110 }, { "grad_norm": 0.16808851063251495, "learning_rate": 5.70237802814364e-05, "loss": 0.0095, "step": 24120 }, { "grad_norm": 0.27288171648979187, "learning_rate": 5.6991037262755277e-05, "loss": 0.0097, "step": 24130 }, { "grad_norm": 0.29405340552330017, "learning_rate": 5.6958291185956924e-05, "loss": 0.0088, "step": 24140 }, { "grad_norm": 4.302324295043945, "learning_rate": 5.692554206536561e-05, "loss": 0.0096, "step": 24150 }, { "grad_norm": 0.15219679474830627, "learning_rate": 5.689278991530692e-05, "loss": 0.0099, "step": 24160 }, { "grad_norm": 0.17067930102348328, "learning_rate": 5.686003475010773e-05, "loss": 0.0088, "step": 24170 }, { "grad_norm": 0.3743131160736084, "learning_rate": 5.682727658409628e-05, "loss": 0.0107, "step": 24180 }, { "grad_norm": 0.14116623997688293, "learning_rate": 5.67945154316021e-05, "loss": 0.0113, "step": 24190 }, { "grad_norm": 0.17576517164707184, "learning_rate": 5.676175130695602e-05, "loss": 0.0084, "step": 24200 }, { "grad_norm": 0.16947348415851593, "learning_rate": 5.672898422449022e-05, "loss": 0.0093, "step": 24210 }, { "grad_norm": 2.5225517749786377, "learning_rate": 5.66962141985381e-05, "loss": 0.0141, "step": 24220 }, { "grad_norm": 0.2453920990228653, "learning_rate": 5.666344124343439e-05, "loss": 0.0108, "step": 24230 }, { "grad_norm": 0.20060761272907257, "learning_rate": 5.66306653735151e-05, "loss": 0.0101, "step": 24240 }, { "grad_norm": 0.20294366776943207, "learning_rate": 5.659788660311749e-05, "loss": 0.0093, "step": 24250 }, { "grad_norm": 0.2906268239021301, "learning_rate": 5.656510494658014e-05, "loss": 0.009, "step": 24260 }, { "grad_norm": 0.35070252418518066, "learning_rate": 5.6532320418242835e-05, "loss": 0.0089, "step": 24270 }, { "grad_norm": 0.2116616666316986, "learning_rate": 5.649953303244665e-05, "loss": 0.0087, "step": 24280 }, { "grad_norm": 0.1883307695388794, "learning_rate": 5.646674280353389e-05, "loss": 0.0114, "step": 24290 }, { "grad_norm": 0.14346863329410553, "learning_rate": 5.6433949745848135e-05, "loss": 0.0101, "step": 24300 }, { "grad_norm": 0.2306625097990036, "learning_rate": 5.6401153873734166e-05, "loss": 0.0111, "step": 24310 }, { "grad_norm": 0.15799637138843536, "learning_rate": 5.6368355201538036e-05, "loss": 0.009, "step": 24320 }, { "grad_norm": 0.5729697346687317, "learning_rate": 5.633555374360697e-05, "loss": 0.0074, "step": 24330 }, { "grad_norm": 0.17695115506649017, "learning_rate": 5.630274951428944e-05, "loss": 0.0097, "step": 24340 }, { "grad_norm": 0.2177169919013977, "learning_rate": 5.626994252793517e-05, "loss": 0.0092, "step": 24350 }, { "grad_norm": 0.2847258150577545, "learning_rate": 5.6237132798895006e-05, "loss": 0.0108, "step": 24360 }, { "grad_norm": 0.17253372073173523, "learning_rate": 5.620432034152106e-05, "loss": 0.0105, "step": 24370 }, { "grad_norm": 0.16194526851177216, "learning_rate": 5.6171505170166604e-05, "loss": 0.0102, "step": 24380 }, { "grad_norm": 0.16315677762031555, "learning_rate": 5.613868729918612e-05, "loss": 0.0101, "step": 24390 }, { "grad_norm": 0.12971077859401703, "learning_rate": 5.6105866742935256e-05, "loss": 0.0086, "step": 24400 }, { "grad_norm": 0.21982833743095398, "learning_rate": 5.607304351577083e-05, "loss": 0.0107, "step": 24410 }, { "grad_norm": 1.212994933128357, "learning_rate": 5.604021763205085e-05, "loss": 0.0096, "step": 24420 }, { "grad_norm": 0.273762971162796, "learning_rate": 5.600738910613445e-05, "loss": 0.0116, "step": 24430 }, { "grad_norm": 0.17004996538162231, "learning_rate": 5.597455795238198e-05, "loss": 0.0097, "step": 24440 }, { "grad_norm": 0.6258017420768738, "learning_rate": 5.5941724185154876e-05, "loss": 0.0102, "step": 24450 }, { "grad_norm": 0.2695268392562866, "learning_rate": 5.5908887818815746e-05, "loss": 0.0088, "step": 24460 }, { "grad_norm": 0.18806156516075134, "learning_rate": 5.587604886772834e-05, "loss": 0.0111, "step": 24470 }, { "grad_norm": 0.129507377743721, "learning_rate": 5.584320734625752e-05, "loss": 0.0089, "step": 24480 }, { "grad_norm": 0.22711153328418732, "learning_rate": 5.581036326876928e-05, "loss": 0.0107, "step": 24490 }, { "grad_norm": 0.3904903829097748, "learning_rate": 5.5777516649630756e-05, "loss": 0.0127, "step": 24500 }, { "grad_norm": 0.13605834543704987, "learning_rate": 5.574466750321017e-05, "loss": 0.0101, "step": 24510 }, { "grad_norm": 0.22648589313030243, "learning_rate": 5.571181584387683e-05, "loss": 0.011, "step": 24520 }, { "grad_norm": 0.4012121558189392, "learning_rate": 5.5678961686001196e-05, "loss": 0.0108, "step": 24530 }, { "grad_norm": 0.20160526037216187, "learning_rate": 5.5646105043954785e-05, "loss": 0.012, "step": 24540 }, { "grad_norm": 0.11945579200983047, "learning_rate": 5.5613245932110205e-05, "loss": 0.0104, "step": 24550 }, { "grad_norm": 0.16591472923755646, "learning_rate": 5.5580384364841157e-05, "loss": 0.009, "step": 24560 }, { "grad_norm": 0.14610417187213898, "learning_rate": 5.554752035652237e-05, "loss": 0.0109, "step": 24570 }, { "grad_norm": 0.1721740961074829, "learning_rate": 5.5514653921529736e-05, "loss": 0.0089, "step": 24580 }, { "grad_norm": 0.20893843472003937, "learning_rate": 5.5481785074240114e-05, "loss": 0.0082, "step": 24590 }, { "grad_norm": 1.1596810817718506, "learning_rate": 5.544891382903146e-05, "loss": 0.0115, "step": 24600 }, { "grad_norm": 0.21993543207645416, "learning_rate": 5.541604020028277e-05, "loss": 0.0092, "step": 24610 }, { "grad_norm": 0.23371246457099915, "learning_rate": 5.5383164202374105e-05, "loss": 0.0117, "step": 24620 }, { "grad_norm": 0.41264739632606506, "learning_rate": 5.5350285849686515e-05, "loss": 0.0109, "step": 24630 }, { "grad_norm": 0.20309358835220337, "learning_rate": 5.531740515660213e-05, "loss": 0.0094, "step": 24640 }, { "grad_norm": 0.1982370913028717, "learning_rate": 5.528452213750408e-05, "loss": 0.0095, "step": 24650 }, { "grad_norm": 0.5641371011734009, "learning_rate": 5.52516368067765e-05, "loss": 0.0099, "step": 24660 }, { "grad_norm": 0.1681140959262848, "learning_rate": 5.5218749178804563e-05, "loss": 0.0103, "step": 24670 }, { "grad_norm": 0.1727553904056549, "learning_rate": 5.518585926797443e-05, "loss": 0.01, "step": 24680 }, { "grad_norm": 0.16714543104171753, "learning_rate": 5.5152967088673265e-05, "loss": 0.0107, "step": 24690 }, { "grad_norm": 0.2530636191368103, "learning_rate": 5.512007265528924e-05, "loss": 0.0085, "step": 24700 }, { "grad_norm": 0.22405752539634705, "learning_rate": 5.508717598221148e-05, "loss": 0.0077, "step": 24710 }, { "grad_norm": 0.16538086533546448, "learning_rate": 5.505427708383012e-05, "loss": 0.0119, "step": 24720 }, { "grad_norm": 0.22856195271015167, "learning_rate": 5.502137597453624e-05, "loss": 0.0089, "step": 24730 }, { "grad_norm": 1.3322291374206543, "learning_rate": 5.4988472668721914e-05, "loss": 0.0114, "step": 24740 }, { "grad_norm": 0.17547908425331116, "learning_rate": 5.4955567180780164e-05, "loss": 0.0094, "step": 24750 }, { "grad_norm": 0.4009079039096832, "learning_rate": 5.492265952510497e-05, "loss": 0.0126, "step": 24760 }, { "grad_norm": 0.20327062904834747, "learning_rate": 5.4889749716091254e-05, "loss": 0.0094, "step": 24770 }, { "grad_norm": 0.1896876096725464, "learning_rate": 5.4856837768134885e-05, "loss": 0.0094, "step": 24780 }, { "grad_norm": 0.1785062700510025, "learning_rate": 5.482392369563267e-05, "loss": 0.0113, "step": 24790 }, { "grad_norm": 0.19842834770679474, "learning_rate": 5.4791007512982354e-05, "loss": 0.0098, "step": 24800 }, { "grad_norm": 0.21179825067520142, "learning_rate": 5.4758089234582566e-05, "loss": 0.0106, "step": 24810 }, { "grad_norm": 0.41178378462791443, "learning_rate": 5.47251688748329e-05, "loss": 0.0092, "step": 24820 }, { "grad_norm": 1.187294363975525, "learning_rate": 5.4692246448133843e-05, "loss": 0.01, "step": 24830 }, { "grad_norm": 0.20133233070373535, "learning_rate": 5.465932196888676e-05, "loss": 0.008, "step": 24840 }, { "grad_norm": 0.19910940527915955, "learning_rate": 5.462639545149395e-05, "loss": 0.0095, "step": 24850 }, { "grad_norm": 0.12440875917673111, "learning_rate": 5.459346691035861e-05, "loss": 0.0087, "step": 24860 }, { "grad_norm": 0.16012442111968994, "learning_rate": 5.4560536359884776e-05, "loss": 0.0078, "step": 24870 }, { "grad_norm": 0.1560838520526886, "learning_rate": 5.45276038144774e-05, "loss": 0.0085, "step": 24880 }, { "grad_norm": 0.23087623715400696, "learning_rate": 5.449466928854229e-05, "loss": 0.0079, "step": 24890 }, { "grad_norm": 0.18871288001537323, "learning_rate": 5.446173279648613e-05, "loss": 0.01, "step": 24900 }, { "grad_norm": 0.22783607244491577, "learning_rate": 5.442879435271646e-05, "loss": 0.0083, "step": 24910 }, { "grad_norm": 0.16978022456169128, "learning_rate": 5.4395853971641675e-05, "loss": 0.0106, "step": 24920 }, { "grad_norm": 1.30705726146698, "learning_rate": 5.4362911667671015e-05, "loss": 0.0099, "step": 24930 }, { "grad_norm": 0.39413508772850037, "learning_rate": 5.432996745521458e-05, "loss": 0.0105, "step": 24940 }, { "grad_norm": 0.1430092751979828, "learning_rate": 5.4297021348683264e-05, "loss": 0.0101, "step": 24950 }, { "grad_norm": 0.17708897590637207, "learning_rate": 5.426407336248882e-05, "loss": 0.0099, "step": 24960 }, { "grad_norm": 0.1563284546136856, "learning_rate": 5.423112351104382e-05, "loss": 0.0089, "step": 24970 }, { "grad_norm": 0.24043186008930206, "learning_rate": 5.4198171808761654e-05, "loss": 0.0102, "step": 24980 }, { "grad_norm": 0.1430344134569168, "learning_rate": 5.41652182700565e-05, "loss": 0.01, "step": 24990 }, { "grad_norm": 0.21501988172531128, "learning_rate": 5.413226290934338e-05, "loss": 0.0091, "step": 25000 }, { "grad_norm": 0.15067799389362335, "learning_rate": 5.4099305741038056e-05, "loss": 0.0081, "step": 25010 }, { "grad_norm": 0.16945186257362366, "learning_rate": 5.406634677955713e-05, "loss": 0.0112, "step": 25020 }, { "grad_norm": 0.4049825668334961, "learning_rate": 5.403338603931798e-05, "loss": 0.0115, "step": 25030 }, { "grad_norm": 0.14165568351745605, "learning_rate": 5.4000423534738745e-05, "loss": 0.0094, "step": 25040 }, { "grad_norm": 1.5055509805679321, "learning_rate": 5.396745928023835e-05, "loss": 0.0106, "step": 25050 }, { "grad_norm": 0.40047550201416016, "learning_rate": 5.3934493290236474e-05, "loss": 0.0103, "step": 25060 }, { "grad_norm": 0.8140053153038025, "learning_rate": 5.390152557915357e-05, "loss": 0.0098, "step": 25070 }, { "grad_norm": 0.27889350056648254, "learning_rate": 5.3868556161410796e-05, "loss": 0.0088, "step": 25080 }, { "grad_norm": 0.2837294340133667, "learning_rate": 5.383558505143015e-05, "loss": 0.0078, "step": 25090 }, { "grad_norm": 0.32466089725494385, "learning_rate": 5.38026122636343e-05, "loss": 0.01, "step": 25100 }, { "grad_norm": 0.2193988859653473, "learning_rate": 5.3769637812446636e-05, "loss": 0.0078, "step": 25110 }, { "grad_norm": 0.14782670140266418, "learning_rate": 5.373666171229133e-05, "loss": 0.0097, "step": 25120 }, { "grad_norm": 0.19771704077720642, "learning_rate": 5.370368397759324e-05, "loss": 0.0094, "step": 25130 }, { "grad_norm": 0.26023849844932556, "learning_rate": 5.3670704622777944e-05, "loss": 0.01, "step": 25140 }, { "grad_norm": 0.2788967788219452, "learning_rate": 5.3637723662271744e-05, "loss": 0.0101, "step": 25150 }, { "grad_norm": 0.6222851872444153, "learning_rate": 5.360474111050162e-05, "loss": 0.0092, "step": 25160 }, { "grad_norm": 0.2765846848487854, "learning_rate": 5.357175698189527e-05, "loss": 0.011, "step": 25170 }, { "grad_norm": 0.15854394435882568, "learning_rate": 5.353877129088107e-05, "loss": 0.009, "step": 25180 }, { "grad_norm": 0.3161512613296509, "learning_rate": 5.350578405188809e-05, "loss": 0.0098, "step": 25190 }, { "grad_norm": 0.18550556898117065, "learning_rate": 5.347279527934603e-05, "loss": 0.0082, "step": 25200 }, { "grad_norm": 0.19044296443462372, "learning_rate": 5.343980498768536e-05, "loss": 0.0075, "step": 25210 }, { "grad_norm": 2.5134243965148926, "learning_rate": 5.340681319133711e-05, "loss": 0.0115, "step": 25220 }, { "grad_norm": 0.1782037317752838, "learning_rate": 5.337381990473305e-05, "loss": 0.0104, "step": 25230 }, { "grad_norm": 0.21125246584415436, "learning_rate": 5.3340825142305504e-05, "loss": 0.0088, "step": 25240 }, { "grad_norm": 0.18794575333595276, "learning_rate": 5.330782891848756e-05, "loss": 0.0079, "step": 25250 }, { "grad_norm": 0.15770764648914337, "learning_rate": 5.327483124771285e-05, "loss": 0.009, "step": 25260 }, { "grad_norm": 0.16159817576408386, "learning_rate": 5.3241832144415706e-05, "loss": 0.0105, "step": 25270 }, { "grad_norm": 0.19771462678909302, "learning_rate": 5.320883162303104e-05, "loss": 0.0096, "step": 25280 }, { "grad_norm": 0.18763746321201324, "learning_rate": 5.3175829697994416e-05, "loss": 0.01, "step": 25290 }, { "grad_norm": 0.3990212380886078, "learning_rate": 5.3142826383741975e-05, "loss": 0.0103, "step": 25300 }, { "grad_norm": 0.23006384074687958, "learning_rate": 5.3109821694710506e-05, "loss": 0.01, "step": 25310 }, { "grad_norm": 0.17094814777374268, "learning_rate": 5.307681564533736e-05, "loss": 0.0098, "step": 25320 }, { "grad_norm": 0.8795074820518494, "learning_rate": 5.3043808250060535e-05, "loss": 0.0085, "step": 25330 }, { "grad_norm": 0.14182110130786896, "learning_rate": 5.3010799523318566e-05, "loss": 0.0087, "step": 25340 }, { "grad_norm": 0.17874225974082947, "learning_rate": 5.29777894795506e-05, "loss": 0.0086, "step": 25350 }, { "grad_norm": 0.32607072591781616, "learning_rate": 5.294477813319634e-05, "loss": 0.0085, "step": 25360 }, { "grad_norm": 2.25864839553833, "learning_rate": 5.291176549869608e-05, "loss": 0.0115, "step": 25370 }, { "grad_norm": 0.22506266832351685, "learning_rate": 5.287875159049065e-05, "loss": 0.0094, "step": 25380 }, { "grad_norm": 0.19653873145580292, "learning_rate": 5.284573642302149e-05, "loss": 0.0109, "step": 25390 }, { "grad_norm": 0.14452503621578217, "learning_rate": 5.2812720010730543e-05, "loss": 0.011, "step": 25400 }, { "grad_norm": 0.15781915187835693, "learning_rate": 5.277970236806029e-05, "loss": 0.012, "step": 25410 }, { "grad_norm": 0.18935585021972656, "learning_rate": 5.2746683509453795e-05, "loss": 0.0113, "step": 25420 }, { "grad_norm": 0.1884695291519165, "learning_rate": 5.271366344935461e-05, "loss": 0.0113, "step": 25430 }, { "grad_norm": 0.16750630736351013, "learning_rate": 5.268064220220683e-05, "loss": 0.0096, "step": 25440 }, { "grad_norm": 0.18887768685817719, "learning_rate": 5.2647619782455094e-05, "loss": 0.0094, "step": 25450 }, { "grad_norm": 0.16872671246528625, "learning_rate": 5.261459620454451e-05, "loss": 0.0107, "step": 25460 }, { "grad_norm": 0.1554439663887024, "learning_rate": 5.258157148292071e-05, "loss": 0.0123, "step": 25470 }, { "grad_norm": 0.18429110944271088, "learning_rate": 5.2548545632029846e-05, "loss": 0.0117, "step": 25480 }, { "grad_norm": 0.3327803611755371, "learning_rate": 5.2515518666318534e-05, "loss": 0.0103, "step": 25490 }, { "grad_norm": 0.23778197169303894, "learning_rate": 5.248249060023389e-05, "loss": 0.0092, "step": 25500 }, { "grad_norm": 0.12726274132728577, "learning_rate": 5.2449461448223517e-05, "loss": 0.0092, "step": 25510 }, { "grad_norm": 0.17695100605487823, "learning_rate": 5.2416431224735486e-05, "loss": 0.0118, "step": 25520 }, { "grad_norm": 0.25061264634132385, "learning_rate": 5.238339994421836e-05, "loss": 0.0105, "step": 25530 }, { "grad_norm": 0.2387041598558426, "learning_rate": 5.235036762112111e-05, "loss": 0.011, "step": 25540 }, { "grad_norm": 0.1612476110458374, "learning_rate": 5.2317334269893206e-05, "loss": 0.0073, "step": 25550 }, { "grad_norm": 0.15078039467334747, "learning_rate": 5.228429990498458e-05, "loss": 0.0085, "step": 25560 }, { "grad_norm": 0.9032996296882629, "learning_rate": 5.225126454084556e-05, "loss": 0.009, "step": 25570 }, { "grad_norm": 0.2335481196641922, "learning_rate": 5.221822819192694e-05, "loss": 0.0107, "step": 25580 }, { "grad_norm": 0.21909572184085846, "learning_rate": 5.218519087267995e-05, "loss": 0.0103, "step": 25590 }, { "grad_norm": 0.18377819657325745, "learning_rate": 5.2152152597556225e-05, "loss": 0.0093, "step": 25600 }, { "grad_norm": 0.5256941318511963, "learning_rate": 5.211911338100784e-05, "loss": 0.0096, "step": 25610 }, { "grad_norm": 0.3081772029399872, "learning_rate": 5.2086073237487264e-05, "loss": 0.0103, "step": 25620 }, { "grad_norm": 0.30731096863746643, "learning_rate": 5.205303218144739e-05, "loss": 0.0104, "step": 25630 }, { "grad_norm": 0.23939327895641327, "learning_rate": 5.2019990227341494e-05, "loss": 0.0089, "step": 25640 }, { "grad_norm": 0.2602151930332184, "learning_rate": 5.198694738962324e-05, "loss": 0.009, "step": 25650 }, { "grad_norm": 0.26548370718955994, "learning_rate": 5.19539036827467e-05, "loss": 0.0094, "step": 25660 }, { "grad_norm": 0.4358850121498108, "learning_rate": 5.192085912116629e-05, "loss": 0.0115, "step": 25670 }, { "grad_norm": 0.16247327625751495, "learning_rate": 5.188781371933685e-05, "loss": 0.0095, "step": 25680 }, { "grad_norm": 0.19934074580669403, "learning_rate": 5.185476749171356e-05, "loss": 0.0067, "step": 25690 }, { "grad_norm": 0.22535622119903564, "learning_rate": 5.1821720452751945e-05, "loss": 0.0108, "step": 25700 }, { "grad_norm": 0.1519203931093216, "learning_rate": 5.178867261690791e-05, "loss": 0.0101, "step": 25710 }, { "grad_norm": 0.16388148069381714, "learning_rate": 5.1755623998637715e-05, "loss": 0.0079, "step": 25720 }, { "grad_norm": 0.15908931195735931, "learning_rate": 5.172257461239794e-05, "loss": 0.0101, "step": 25730 }, { "grad_norm": 0.2533472180366516, "learning_rate": 5.168952447264548e-05, "loss": 0.0101, "step": 25740 }, { "grad_norm": 0.2320566177368164, "learning_rate": 5.165647359383763e-05, "loss": 0.0112, "step": 25750 }, { "grad_norm": 0.2754916548728943, "learning_rate": 5.1623421990431954e-05, "loss": 0.0109, "step": 25760 }, { "grad_norm": 0.7487048506736755, "learning_rate": 5.159036967688635e-05, "loss": 0.0087, "step": 25770 }, { "grad_norm": 0.1771295964717865, "learning_rate": 5.1557316667659015e-05, "loss": 0.0088, "step": 25780 }, { "grad_norm": 0.3011193573474884, "learning_rate": 5.152426297720846e-05, "loss": 0.0094, "step": 25790 }, { "grad_norm": 0.19523242115974426, "learning_rate": 5.1491208619993505e-05, "loss": 0.0089, "step": 25800 }, { "grad_norm": 0.16023270785808563, "learning_rate": 5.145815361047321e-05, "loss": 0.0099, "step": 25810 }, { "grad_norm": 0.2788575291633606, "learning_rate": 5.1425097963107003e-05, "loss": 0.0098, "step": 25820 }, { "grad_norm": 0.09927152842283249, "learning_rate": 5.1392041692354544e-05, "loss": 0.0104, "step": 25830 }, { "grad_norm": 0.20512452721595764, "learning_rate": 5.1358984812675736e-05, "loss": 0.0092, "step": 25840 }, { "grad_norm": 0.1803385466337204, "learning_rate": 5.13259273385308e-05, "loss": 0.0092, "step": 25850 }, { "grad_norm": 0.17077723145484924, "learning_rate": 5.12928692843802e-05, "loss": 0.0107, "step": 25860 }, { "grad_norm": 0.15104302763938904, "learning_rate": 5.125981066468465e-05, "loss": 0.0075, "step": 25870 }, { "grad_norm": 0.19592063128948212, "learning_rate": 5.122675149390514e-05, "loss": 0.0123, "step": 25880 }, { "grad_norm": 0.2689017057418823, "learning_rate": 5.119369178650282e-05, "loss": 0.0103, "step": 25890 }, { "grad_norm": 0.19582970440387726, "learning_rate": 5.1160631556939156e-05, "loss": 0.0083, "step": 25900 }, { "grad_norm": 0.2550731599330902, "learning_rate": 5.112757081967584e-05, "loss": 0.0098, "step": 25910 }, { "grad_norm": 0.21872200071811676, "learning_rate": 5.109450958917472e-05, "loss": 0.0097, "step": 25920 }, { "grad_norm": 0.17812490463256836, "learning_rate": 5.106144787989794e-05, "loss": 0.0093, "step": 25930 }, { "grad_norm": 0.2397008091211319, "learning_rate": 5.102838570630778e-05, "loss": 0.0078, "step": 25940 }, { "grad_norm": 0.5702326893806458, "learning_rate": 5.099532308286678e-05, "loss": 0.0086, "step": 25950 }, { "grad_norm": 0.11352180689573288, "learning_rate": 5.0962260024037645e-05, "loss": 0.0079, "step": 25960 }, { "grad_norm": 2.4711668491363525, "learning_rate": 5.0929196544283276e-05, "loss": 0.0122, "step": 25970 }, { "grad_norm": 0.19683833420276642, "learning_rate": 5.089613265806678e-05, "loss": 0.0099, "step": 25980 }, { "grad_norm": 0.32363906502723694, "learning_rate": 5.086306837985141e-05, "loss": 0.0097, "step": 25990 }, { "grad_norm": 0.17391443252563477, "learning_rate": 5.0830003724100604e-05, "loss": 0.0087, "step": 26000 }, { "grad_norm": 0.23519504070281982, "learning_rate": 5.0796938705277984e-05, "loss": 0.0114, "step": 26010 }, { "grad_norm": 0.16582858562469482, "learning_rate": 5.076387333784729e-05, "loss": 0.0089, "step": 26020 }, { "grad_norm": 0.21240095794200897, "learning_rate": 5.0730807636272445e-05, "loss": 0.0116, "step": 26030 }, { "grad_norm": 0.17670941352844238, "learning_rate": 5.0697741615017524e-05, "loss": 0.0091, "step": 26040 }, { "grad_norm": 3.6802899837493896, "learning_rate": 5.06646752885467e-05, "loss": 0.0075, "step": 26050 }, { "grad_norm": 0.1652572602033615, "learning_rate": 5.0631608671324324e-05, "loss": 0.0097, "step": 26060 }, { "grad_norm": 0.12358053028583527, "learning_rate": 5.059854177781487e-05, "loss": 0.0093, "step": 26070 }, { "grad_norm": 0.14036349952220917, "learning_rate": 5.05654746224829e-05, "loss": 0.0093, "step": 26080 }, { "grad_norm": 0.16357819736003876, "learning_rate": 5.053240721979311e-05, "loss": 0.0096, "step": 26090 }, { "grad_norm": 0.20821885764598846, "learning_rate": 5.049933958421031e-05, "loss": 0.0101, "step": 26100 }, { "grad_norm": 0.21557198464870453, "learning_rate": 5.046627173019943e-05, "loss": 0.0084, "step": 26110 }, { "grad_norm": 0.208607479929924, "learning_rate": 5.043320367222545e-05, "loss": 0.0084, "step": 26120 }, { "grad_norm": 1.3465774059295654, "learning_rate": 5.040013542475346e-05, "loss": 0.0111, "step": 26130 }, { "grad_norm": 0.25698670744895935, "learning_rate": 5.0367067002248646e-05, "loss": 0.01, "step": 26140 }, { "grad_norm": 0.18271754682064056, "learning_rate": 5.0333998419176254e-05, "loss": 0.0108, "step": 26150 }, { "grad_norm": 0.14711090922355652, "learning_rate": 5.030092969000161e-05, "loss": 0.008, "step": 26160 }, { "grad_norm": 0.19178830087184906, "learning_rate": 5.026786082919008e-05, "loss": 0.0114, "step": 26170 }, { "grad_norm": 0.3199276924133301, "learning_rate": 5.0234791851207144e-05, "loss": 0.0074, "step": 26180 }, { "grad_norm": 0.12966424226760864, "learning_rate": 5.0201722770518256e-05, "loss": 0.0078, "step": 26190 }, { "grad_norm": 0.18802864849567413, "learning_rate": 5.0168653601589e-05, "loss": 0.0088, "step": 26200 }, { "grad_norm": 0.3491520583629608, "learning_rate": 5.01355843588849e-05, "loss": 0.0122, "step": 26210 }, { "grad_norm": 0.12394449859857559, "learning_rate": 5.010251505687161e-05, "loss": 0.0093, "step": 26220 }, { "grad_norm": 0.15631575882434845, "learning_rate": 5.006944571001474e-05, "loss": 0.0101, "step": 26230 }, { "grad_norm": 0.31161463260650635, "learning_rate": 5.003637633277997e-05, "loss": 0.0073, "step": 26240 }, { "grad_norm": 0.19869296252727509, "learning_rate": 5.0003306939632946e-05, "loss": 0.0073, "step": 26250 }, { "grad_norm": 0.1027400940656662, "learning_rate": 4.997023754503937e-05, "loss": 0.0094, "step": 26260 }, { "grad_norm": 0.1617780178785324, "learning_rate": 4.9937168163464897e-05, "loss": 0.008, "step": 26270 }, { "grad_norm": 0.20085443556308746, "learning_rate": 4.990409880937519e-05, "loss": 0.008, "step": 26280 }, { "grad_norm": 1.0359225273132324, "learning_rate": 4.9871029497235936e-05, "loss": 0.0075, "step": 26290 }, { "grad_norm": 0.5284215807914734, "learning_rate": 4.9837960241512766e-05, "loss": 0.0103, "step": 26300 }, { "grad_norm": 0.19239340722560883, "learning_rate": 4.9804891056671265e-05, "loss": 0.0075, "step": 26310 }, { "grad_norm": 0.17320473492145538, "learning_rate": 4.9771821957177074e-05, "loss": 0.0082, "step": 26320 }, { "grad_norm": 1.0599334239959717, "learning_rate": 4.9738752957495694e-05, "loss": 0.0096, "step": 26330 }, { "grad_norm": 0.162753164768219, "learning_rate": 4.970568407209265e-05, "loss": 0.0082, "step": 26340 }, { "grad_norm": 0.15343144536018372, "learning_rate": 4.967261531543339e-05, "loss": 0.0115, "step": 26350 }, { "grad_norm": 0.23374496400356293, "learning_rate": 4.963954670198332e-05, "loss": 0.009, "step": 26360 }, { "grad_norm": 0.22962722182273865, "learning_rate": 4.960647824620776e-05, "loss": 0.0092, "step": 26370 }, { "grad_norm": 0.23836679756641388, "learning_rate": 4.9573409962571995e-05, "loss": 0.009, "step": 26380 }, { "grad_norm": 0.20516933500766754, "learning_rate": 4.954034186554121e-05, "loss": 0.01, "step": 26390 }, { "grad_norm": 0.17654217779636383, "learning_rate": 4.950727396958051e-05, "loss": 0.0087, "step": 26400 }, { "grad_norm": 0.29311832785606384, "learning_rate": 4.947420628915489e-05, "loss": 0.012, "step": 26410 }, { "grad_norm": 0.31959521770477295, "learning_rate": 4.944113883872932e-05, "loss": 0.0094, "step": 26420 }, { "grad_norm": 0.11270368099212646, "learning_rate": 4.9408071632768624e-05, "loss": 0.0084, "step": 26430 }, { "grad_norm": 0.15197744965553284, "learning_rate": 4.937500468573747e-05, "loss": 0.0092, "step": 26440 }, { "grad_norm": 0.17671328783035278, "learning_rate": 4.934193801210053e-05, "loss": 0.0074, "step": 26450 }, { "grad_norm": 0.21689759194850922, "learning_rate": 4.930887162632224e-05, "loss": 0.0081, "step": 26460 }, { "grad_norm": 0.17926938831806183, "learning_rate": 4.9275805542866994e-05, "loss": 0.0077, "step": 26470 }, { "grad_norm": 0.15769439935684204, "learning_rate": 4.9242739776199006e-05, "loss": 0.0082, "step": 26480 }, { "grad_norm": 1.3289110660552979, "learning_rate": 4.920967434078237e-05, "loss": 0.0091, "step": 26490 }, { "grad_norm": 0.13386742770671844, "learning_rate": 4.917660925108101e-05, "loss": 0.0076, "step": 26500 }, { "grad_norm": 0.19160041213035583, "learning_rate": 4.914354452155876e-05, "loss": 0.0094, "step": 26510 }, { "grad_norm": 0.14778675138950348, "learning_rate": 4.911048016667923e-05, "loss": 0.0083, "step": 26520 }, { "grad_norm": 0.21750134229660034, "learning_rate": 4.907741620090588e-05, "loss": 0.0093, "step": 26530 }, { "grad_norm": 0.2160332202911377, "learning_rate": 4.9044352638702046e-05, "loss": 0.0099, "step": 26540 }, { "grad_norm": 0.17766164243221283, "learning_rate": 4.9011289494530834e-05, "loss": 0.0097, "step": 26550 }, { "grad_norm": 0.21741336584091187, "learning_rate": 4.897822678285516e-05, "loss": 0.009, "step": 26560 }, { "grad_norm": 3.1959848403930664, "learning_rate": 4.8945164518137835e-05, "loss": 0.0102, "step": 26570 }, { "grad_norm": 0.20955441892147064, "learning_rate": 4.891210271484138e-05, "loss": 0.0117, "step": 26580 }, { "grad_norm": 0.1840115189552307, "learning_rate": 4.8879041387428164e-05, "loss": 0.0081, "step": 26590 }, { "grad_norm": 0.15861079096794128, "learning_rate": 4.8845980550360306e-05, "loss": 0.0075, "step": 26600 }, { "grad_norm": 0.2389969378709793, "learning_rate": 4.8812920218099786e-05, "loss": 0.0105, "step": 26610 }, { "grad_norm": 0.5092324018478394, "learning_rate": 4.8779860405108294e-05, "loss": 0.0109, "step": 26620 }, { "grad_norm": 0.17142723500728607, "learning_rate": 4.8746801125847285e-05, "loss": 0.0087, "step": 26630 }, { "grad_norm": 0.14276424050331116, "learning_rate": 4.8713742394778064e-05, "loss": 0.0085, "step": 26640 }, { "grad_norm": 0.41500264406204224, "learning_rate": 4.868068422636162e-05, "loss": 0.0088, "step": 26650 }, { "grad_norm": 0.9915323257446289, "learning_rate": 4.86476266350587e-05, "loss": 0.0087, "step": 26660 }, { "grad_norm": 0.29861053824424744, "learning_rate": 4.861456963532985e-05, "loss": 0.0082, "step": 26670 }, { "grad_norm": 0.22487720847129822, "learning_rate": 4.8581513241635316e-05, "loss": 0.0106, "step": 26680 }, { "grad_norm": 0.19774140417575836, "learning_rate": 4.8548457468435065e-05, "loss": 0.0077, "step": 26690 }, { "grad_norm": 0.21335691213607788, "learning_rate": 4.851540233018884e-05, "loss": 0.0082, "step": 26700 }, { "grad_norm": 0.16356778144836426, "learning_rate": 4.848234784135608e-05, "loss": 0.0089, "step": 26710 }, { "grad_norm": 0.40612226724624634, "learning_rate": 4.8449294016395916e-05, "loss": 0.0092, "step": 26720 }, { "grad_norm": 2.0979044437408447, "learning_rate": 4.8416240869767246e-05, "loss": 0.0078, "step": 26730 }, { "grad_norm": 0.2162989377975464, "learning_rate": 4.838318841592864e-05, "loss": 0.0071, "step": 26740 }, { "grad_norm": 0.22617091238498688, "learning_rate": 4.835013666933834e-05, "loss": 0.008, "step": 26750 }, { "grad_norm": 0.4672699272632599, "learning_rate": 4.83170856444543e-05, "loss": 0.0089, "step": 26760 }, { "grad_norm": 0.19619598984718323, "learning_rate": 4.828403535573419e-05, "loss": 0.0067, "step": 26770 }, { "grad_norm": 0.24560222029685974, "learning_rate": 4.825098581763532e-05, "loss": 0.0108, "step": 26780 }, { "grad_norm": 0.17815476655960083, "learning_rate": 4.821793704461465e-05, "loss": 0.007, "step": 26790 }, { "grad_norm": 0.1911143809556961, "learning_rate": 4.8184889051128884e-05, "loss": 0.0089, "step": 26800 }, { "grad_norm": 2.0329484939575195, "learning_rate": 4.81518418516343e-05, "loss": 0.0099, "step": 26810 }, { "grad_norm": 0.206064835190773, "learning_rate": 4.811879546058688e-05, "loss": 0.0088, "step": 26820 }, { "grad_norm": 0.1909775584936142, "learning_rate": 4.808574989244224e-05, "loss": 0.011, "step": 26830 }, { "grad_norm": 0.16240034997463226, "learning_rate": 4.8052705161655644e-05, "loss": 0.0095, "step": 26840 }, { "grad_norm": 0.1607295274734497, "learning_rate": 4.8019661282681936e-05, "loss": 0.0073, "step": 26850 }, { "grad_norm": 0.7299208045005798, "learning_rate": 4.79866182699757e-05, "loss": 0.0069, "step": 26860 }, { "grad_norm": 0.9936730861663818, "learning_rate": 4.795357613799103e-05, "loss": 0.0085, "step": 26870 }, { "grad_norm": 0.26085489988327026, "learning_rate": 4.792053490118166e-05, "loss": 0.0085, "step": 26880 }, { "grad_norm": 0.2177339494228363, "learning_rate": 4.788749457400099e-05, "loss": 0.0095, "step": 26890 }, { "grad_norm": 0.2012946605682373, "learning_rate": 4.7854455170901966e-05, "loss": 0.0083, "step": 26900 }, { "grad_norm": 0.2946806848049164, "learning_rate": 4.7821416706337136e-05, "loss": 0.0088, "step": 26910 }, { "grad_norm": 0.12609781324863434, "learning_rate": 4.7788379194758645e-05, "loss": 0.0084, "step": 26920 }, { "grad_norm": 0.9041496515274048, "learning_rate": 4.7755342650618246e-05, "loss": 0.0087, "step": 26930 }, { "grad_norm": 5.205366611480713, "learning_rate": 4.7722307088367235e-05, "loss": 0.0094, "step": 26940 }, { "grad_norm": 0.2245074063539505, "learning_rate": 4.768927252245646e-05, "loss": 0.0098, "step": 26950 }, { "grad_norm": 0.1822582185268402, "learning_rate": 4.765623896733642e-05, "loss": 0.009, "step": 26960 }, { "grad_norm": 0.21877314150333405, "learning_rate": 4.762320643745708e-05, "loss": 0.0089, "step": 26970 }, { "grad_norm": 0.15910395979881287, "learning_rate": 4.7590174947267966e-05, "loss": 0.0084, "step": 26980 }, { "grad_norm": 0.6515395641326904, "learning_rate": 4.755714451121823e-05, "loss": 0.0092, "step": 26990 }, { "grad_norm": 0.3774939179420471, "learning_rate": 4.7524115143756476e-05, "loss": 0.0079, "step": 27000 }, { "grad_norm": 0.184650719165802, "learning_rate": 4.7491086859330854e-05, "loss": 0.0094, "step": 27010 }, { "grad_norm": 0.24284756183624268, "learning_rate": 4.74580596723891e-05, "loss": 0.0073, "step": 27020 }, { "grad_norm": 0.15784186124801636, "learning_rate": 4.742503359737841e-05, "loss": 0.0087, "step": 27030 }, { "grad_norm": 0.1625804305076599, "learning_rate": 4.739200864874547e-05, "loss": 0.0075, "step": 27040 }, { "grad_norm": 0.4501628875732422, "learning_rate": 4.735898484093658e-05, "loss": 0.0069, "step": 27050 }, { "grad_norm": 0.234886035323143, "learning_rate": 4.7325962188397444e-05, "loss": 0.0086, "step": 27060 }, { "grad_norm": 0.11351709812879562, "learning_rate": 4.729294070557329e-05, "loss": 0.0092, "step": 27070 }, { "grad_norm": 0.2561488151550293, "learning_rate": 4.7259920406908833e-05, "loss": 0.01, "step": 27080 }, { "grad_norm": 0.18692831695079803, "learning_rate": 4.7226901306848306e-05, "loss": 0.0074, "step": 27090 }, { "grad_norm": 0.22226756811141968, "learning_rate": 4.719388341983535e-05, "loss": 0.0083, "step": 27100 }, { "grad_norm": 0.11863621324300766, "learning_rate": 4.71608667603131e-05, "loss": 0.0077, "step": 27110 }, { "grad_norm": 0.30457183718681335, "learning_rate": 4.712785134272422e-05, "loss": 0.0096, "step": 27120 }, { "grad_norm": 0.2362470030784607, "learning_rate": 4.7094837181510734e-05, "loss": 0.0088, "step": 27130 }, { "grad_norm": 0.14740106463432312, "learning_rate": 4.706182429111414e-05, "loss": 0.0077, "step": 27140 }, { "grad_norm": 0.17990729212760925, "learning_rate": 4.702881268597545e-05, "loss": 0.0094, "step": 27150 }, { "grad_norm": 0.11196538060903549, "learning_rate": 4.699580238053503e-05, "loss": 0.0085, "step": 27160 }, { "grad_norm": 0.174104243516922, "learning_rate": 4.69627933892327e-05, "loss": 0.0099, "step": 27170 }, { "grad_norm": 0.2046133130788803, "learning_rate": 4.692978572650773e-05, "loss": 0.0084, "step": 27180 }, { "grad_norm": 0.3629465401172638, "learning_rate": 4.689677940679881e-05, "loss": 0.0102, "step": 27190 }, { "grad_norm": 0.39325132966041565, "learning_rate": 4.6863774444543984e-05, "loss": 0.0078, "step": 27200 }, { "grad_norm": 0.3904510736465454, "learning_rate": 4.6830770854180784e-05, "loss": 0.008, "step": 27210 }, { "grad_norm": 0.3439260721206665, "learning_rate": 4.679776865014608e-05, "loss": 0.0098, "step": 27220 }, { "grad_norm": 0.1789885312318802, "learning_rate": 4.676476784687615e-05, "loss": 0.0072, "step": 27230 }, { "grad_norm": 0.20613503456115723, "learning_rate": 4.673176845880669e-05, "loss": 0.007, "step": 27240 }, { "grad_norm": 0.4031970500946045, "learning_rate": 4.669877050037275e-05, "loss": 0.0093, "step": 27250 }, { "grad_norm": 0.22602270543575287, "learning_rate": 4.666577398600874e-05, "loss": 0.0092, "step": 27260 }, { "grad_norm": 0.15382955968379974, "learning_rate": 4.663277893014846e-05, "loss": 0.0085, "step": 27270 }, { "grad_norm": 0.13498422503471375, "learning_rate": 4.659978534722508e-05, "loss": 0.0094, "step": 27280 }, { "grad_norm": 0.14627225697040558, "learning_rate": 4.6566793251671104e-05, "loss": 0.0092, "step": 27290 }, { "grad_norm": 0.16606850922107697, "learning_rate": 4.65338026579184e-05, "loss": 0.0087, "step": 27300 }, { "grad_norm": 0.8600703477859497, "learning_rate": 4.650081358039817e-05, "loss": 0.0087, "step": 27310 }, { "grad_norm": 0.22858192026615143, "learning_rate": 4.646782603354098e-05, "loss": 0.0076, "step": 27320 }, { "grad_norm": 0.18589048087596893, "learning_rate": 4.643484003177666e-05, "loss": 0.007, "step": 27330 }, { "grad_norm": 0.16153503954410553, "learning_rate": 4.6401855589534454e-05, "loss": 0.0081, "step": 27340 }, { "grad_norm": 0.20104308426380157, "learning_rate": 4.636887272124286e-05, "loss": 0.0076, "step": 27350 }, { "grad_norm": 0.24487091600894928, "learning_rate": 4.6335891441329685e-05, "loss": 0.0079, "step": 27360 }, { "grad_norm": 0.15463776886463165, "learning_rate": 4.630291176422209e-05, "loss": 0.009, "step": 27370 }, { "grad_norm": 0.18308596312999725, "learning_rate": 4.62699337043465e-05, "loss": 0.0092, "step": 27380 }, { "grad_norm": 0.13931916654109955, "learning_rate": 4.623695727612861e-05, "loss": 0.0083, "step": 27390 }, { "grad_norm": 0.19634239375591278, "learning_rate": 4.620398249399347e-05, "loss": 0.0082, "step": 27400 }, { "grad_norm": 0.15853926539421082, "learning_rate": 4.617100937236535e-05, "loss": 0.0084, "step": 27410 }, { "grad_norm": 0.1704259067773819, "learning_rate": 4.613803792566781e-05, "loss": 0.0087, "step": 27420 }, { "grad_norm": 0.24570848047733307, "learning_rate": 4.61050681683237e-05, "loss": 0.0102, "step": 27430 }, { "grad_norm": 0.2613743543624878, "learning_rate": 4.6072100114755104e-05, "loss": 0.0079, "step": 27440 }, { "grad_norm": 0.17726248502731323, "learning_rate": 4.603913377938336e-05, "loss": 0.0106, "step": 27450 }, { "grad_norm": 0.15115928649902344, "learning_rate": 4.600616917662905e-05, "loss": 0.0071, "step": 27460 }, { "grad_norm": 1.5061230659484863, "learning_rate": 4.597320632091206e-05, "loss": 0.0084, "step": 27470 }, { "grad_norm": 0.39184215664863586, "learning_rate": 4.594024522665142e-05, "loss": 0.0102, "step": 27480 }, { "grad_norm": 0.16460202634334564, "learning_rate": 4.590728590826543e-05, "loss": 0.008, "step": 27490 }, { "grad_norm": 0.1806839257478714, "learning_rate": 4.587432838017165e-05, "loss": 0.0092, "step": 27500 }, { "grad_norm": 0.1847437024116516, "learning_rate": 4.58413726567868e-05, "loss": 0.0074, "step": 27510 }, { "grad_norm": 0.124883733689785, "learning_rate": 4.5808418752526814e-05, "loss": 0.0082, "step": 27520 }, { "grad_norm": 0.16057895123958588, "learning_rate": 4.5775466681806895e-05, "loss": 0.0072, "step": 27530 }, { "grad_norm": 0.17000889778137207, "learning_rate": 4.574251645904136e-05, "loss": 0.0085, "step": 27540 }, { "grad_norm": 0.2681126594543457, "learning_rate": 4.5709568098643776e-05, "loss": 0.0086, "step": 27550 }, { "grad_norm": 0.17649711668491364, "learning_rate": 4.567662161502687e-05, "loss": 0.0077, "step": 27560 }, { "grad_norm": 0.2058940827846527, "learning_rate": 4.5643677022602554e-05, "loss": 0.0097, "step": 27570 }, { "grad_norm": 0.2766430377960205, "learning_rate": 4.56107343357819e-05, "loss": 0.0086, "step": 27580 }, { "grad_norm": 0.3273946940898895, "learning_rate": 4.557779356897518e-05, "loss": 0.0101, "step": 27590 }, { "grad_norm": 0.22176699340343475, "learning_rate": 4.55448547365918e-05, "loss": 0.008, "step": 27600 }, { "grad_norm": 0.18261216580867767, "learning_rate": 4.5511917853040315e-05, "loss": 0.0083, "step": 27610 }, { "grad_norm": 1.4253618717193604, "learning_rate": 4.547898293272842e-05, "loss": 0.012, "step": 27620 }, { "grad_norm": 0.19233399629592896, "learning_rate": 4.544604999006301e-05, "loss": 0.0088, "step": 27630 }, { "grad_norm": 0.2860043942928314, "learning_rate": 4.541311903945004e-05, "loss": 0.0084, "step": 27640 }, { "grad_norm": 0.12146367877721786, "learning_rate": 4.538019009529463e-05, "loss": 0.0086, "step": 27650 }, { "grad_norm": 1.246623158454895, "learning_rate": 4.5347263172001035e-05, "loss": 0.0087, "step": 27660 }, { "grad_norm": 0.13797426223754883, "learning_rate": 4.531433828397261e-05, "loss": 0.0072, "step": 27670 }, { "grad_norm": 0.1319216936826706, "learning_rate": 4.528141544561179e-05, "loss": 0.0074, "step": 27680 }, { "grad_norm": 1.0313998460769653, "learning_rate": 4.5248494671320186e-05, "loss": 0.0083, "step": 27690 }, { "grad_norm": 0.20900854468345642, "learning_rate": 4.521557597549844e-05, "loss": 0.0089, "step": 27700 }, { "grad_norm": 0.17339371144771576, "learning_rate": 4.5182659372546284e-05, "loss": 0.0082, "step": 27710 }, { "grad_norm": 0.2592509090900421, "learning_rate": 4.514974487686261e-05, "loss": 0.0068, "step": 27720 }, { "grad_norm": 0.17429213225841522, "learning_rate": 4.5116832502845306e-05, "loss": 0.0075, "step": 27730 }, { "grad_norm": 0.4153602719306946, "learning_rate": 4.508392226489136e-05, "loss": 0.0073, "step": 27740 }, { "grad_norm": 0.25489819049835205, "learning_rate": 4.505101417739683e-05, "loss": 0.0081, "step": 27750 }, { "grad_norm": 0.46378204226493835, "learning_rate": 4.501810825475685e-05, "loss": 0.0082, "step": 27760 }, { "grad_norm": 0.18189282715320587, "learning_rate": 4.4985204511365556e-05, "loss": 0.0101, "step": 27770 }, { "grad_norm": 0.15396252274513245, "learning_rate": 4.49523029616162e-05, "loss": 0.0066, "step": 27780 }, { "grad_norm": 0.151885524392128, "learning_rate": 4.491940361990101e-05, "loss": 0.0085, "step": 27790 }, { "grad_norm": 0.1895301789045334, "learning_rate": 4.48865065006113e-05, "loss": 0.0095, "step": 27800 }, { "grad_norm": 0.1706954836845398, "learning_rate": 4.485361161813735e-05, "loss": 0.0084, "step": 27810 }, { "grad_norm": 0.1546109914779663, "learning_rate": 4.482071898686854e-05, "loss": 0.0101, "step": 27820 }, { "grad_norm": 0.2350548803806305, "learning_rate": 4.478782862119321e-05, "loss": 0.0074, "step": 27830 }, { "grad_norm": 0.18095487356185913, "learning_rate": 4.475494053549869e-05, "loss": 0.0077, "step": 27840 }, { "grad_norm": 0.2589525580406189, "learning_rate": 4.4722054744171405e-05, "loss": 0.0074, "step": 27850 }, { "grad_norm": 0.2698777914047241, "learning_rate": 4.468917126159669e-05, "loss": 0.0075, "step": 27860 }, { "grad_norm": 0.1460246592760086, "learning_rate": 4.465629010215887e-05, "loss": 0.0077, "step": 27870 }, { "grad_norm": 11.085430145263672, "learning_rate": 4.462341128024133e-05, "loss": 0.0082, "step": 27880 }, { "grad_norm": 0.42670294642448425, "learning_rate": 4.4590534810226356e-05, "loss": 0.0087, "step": 27890 }, { "grad_norm": 0.31195783615112305, "learning_rate": 4.455766070649523e-05, "loss": 0.0085, "step": 27900 }, { "grad_norm": 0.1760215163230896, "learning_rate": 4.4524788983428215e-05, "loss": 0.0092, "step": 27910 }, { "grad_norm": 1.262623906135559, "learning_rate": 4.4491919655404515e-05, "loss": 0.0083, "step": 27920 }, { "grad_norm": 0.2221306413412094, "learning_rate": 4.445905273680228e-05, "loss": 0.0114, "step": 27930 }, { "grad_norm": 1.2934545278549194, "learning_rate": 4.442618824199865e-05, "loss": 0.0085, "step": 27940 }, { "grad_norm": 0.23268964886665344, "learning_rate": 4.4393326185369635e-05, "loss": 0.0108, "step": 27950 }, { "grad_norm": 0.2664499878883362, "learning_rate": 4.436046658129024e-05, "loss": 0.0085, "step": 27960 }, { "grad_norm": 0.8012252449989319, "learning_rate": 4.432760944413433e-05, "loss": 0.0104, "step": 27970 }, { "grad_norm": 5.206298351287842, "learning_rate": 4.4294754788274796e-05, "loss": 0.0097, "step": 27980 }, { "grad_norm": 0.20199662446975708, "learning_rate": 4.426190262808334e-05, "loss": 0.0089, "step": 27990 }, { "grad_norm": 0.12489937990903854, "learning_rate": 4.42290529779306e-05, "loss": 0.009, "step": 28000 }, { "grad_norm": 0.18481695652008057, "learning_rate": 4.4196205852186176e-05, "loss": 0.0086, "step": 28010 }, { "grad_norm": 0.1645239144563675, "learning_rate": 4.416336126521848e-05, "loss": 0.0079, "step": 28020 }, { "grad_norm": 0.26518112421035767, "learning_rate": 4.4130519231394865e-05, "loss": 0.0077, "step": 28030 }, { "grad_norm": 0.918475866317749, "learning_rate": 4.409767976508154e-05, "loss": 0.0068, "step": 28040 }, { "grad_norm": 2.7012248039245605, "learning_rate": 4.406484288064363e-05, "loss": 0.0098, "step": 28050 }, { "grad_norm": 0.23488563299179077, "learning_rate": 4.4032008592445055e-05, "loss": 0.0081, "step": 28060 }, { "grad_norm": 0.18354444205760956, "learning_rate": 4.3999176914848714e-05, "loss": 0.007, "step": 28070 }, { "grad_norm": 0.27285048365592957, "learning_rate": 4.396634786221625e-05, "loss": 0.0072, "step": 28080 }, { "grad_norm": 0.20533321797847748, "learning_rate": 4.3933521448908203e-05, "loss": 0.0065, "step": 28090 }, { "grad_norm": 0.21803995966911316, "learning_rate": 4.3900697689284e-05, "loss": 0.0107, "step": 28100 }, { "grad_norm": 0.23146836459636688, "learning_rate": 4.386787659770185e-05, "loss": 0.0083, "step": 28110 }, { "grad_norm": 0.2246437966823578, "learning_rate": 4.383505818851878e-05, "loss": 0.0087, "step": 28120 }, { "grad_norm": 0.21095970273017883, "learning_rate": 4.380224247609073e-05, "loss": 0.0075, "step": 28130 }, { "grad_norm": 0.16164982318878174, "learning_rate": 4.3769429474772376e-05, "loss": 0.0079, "step": 28140 }, { "grad_norm": 0.2731838524341583, "learning_rate": 4.373661919891726e-05, "loss": 0.0063, "step": 28150 }, { "grad_norm": 0.1663159281015396, "learning_rate": 4.3703811662877674e-05, "loss": 0.0078, "step": 28160 }, { "grad_norm": 0.12766501307487488, "learning_rate": 4.36710068810048e-05, "loss": 0.0089, "step": 28170 }, { "grad_norm": 0.5785871744155884, "learning_rate": 4.363820486764853e-05, "loss": 0.0069, "step": 28180 }, { "grad_norm": 0.1571747362613678, "learning_rate": 4.360540563715758e-05, "loss": 0.0087, "step": 28190 }, { "grad_norm": 0.22047576308250427, "learning_rate": 4.357260920387946e-05, "loss": 0.0093, "step": 28200 }, { "grad_norm": 0.1731286644935608, "learning_rate": 4.353981558216045e-05, "loss": 0.0084, "step": 28210 }, { "grad_norm": 0.37238115072250366, "learning_rate": 4.3507024786345565e-05, "loss": 0.0084, "step": 28220 }, { "grad_norm": 0.14384567737579346, "learning_rate": 4.3474236830778646e-05, "loss": 0.0074, "step": 28230 }, { "grad_norm": 0.2122192233800888, "learning_rate": 4.344145172980223e-05, "loss": 0.0088, "step": 28240 }, { "grad_norm": 0.4425942897796631, "learning_rate": 4.3408669497757656e-05, "loss": 0.0076, "step": 28250 }, { "grad_norm": 1.7191662788391113, "learning_rate": 4.3375890148984976e-05, "loss": 0.0104, "step": 28260 }, { "grad_norm": 0.16127854585647583, "learning_rate": 4.334311369782301e-05, "loss": 0.0084, "step": 28270 }, { "grad_norm": 0.26909512281417847, "learning_rate": 4.3310340158609244e-05, "loss": 0.0076, "step": 28280 }, { "grad_norm": 0.4445036053657532, "learning_rate": 4.327756954568e-05, "loss": 0.0084, "step": 28290 }, { "grad_norm": 0.12991812825202942, "learning_rate": 4.324480187337021e-05, "loss": 0.0074, "step": 28300 }, { "grad_norm": 0.13900019228458405, "learning_rate": 4.3212037156013596e-05, "loss": 0.0093, "step": 28310 }, { "grad_norm": 0.2357272356748581, "learning_rate": 4.317927540794251e-05, "loss": 0.0077, "step": 28320 }, { "grad_norm": 0.11088847368955612, "learning_rate": 4.314651664348812e-05, "loss": 0.0063, "step": 28330 }, { "grad_norm": 0.162581667304039, "learning_rate": 4.3113760876980195e-05, "loss": 0.0064, "step": 28340 }, { "grad_norm": 0.18735867738723755, "learning_rate": 4.308100812274719e-05, "loss": 0.0065, "step": 28350 }, { "grad_norm": 0.21249881386756897, "learning_rate": 4.304825839511632e-05, "loss": 0.008, "step": 28360 }, { "grad_norm": 0.23851890861988068, "learning_rate": 4.3015511708413395e-05, "loss": 0.0075, "step": 28370 }, { "grad_norm": 0.12690770626068115, "learning_rate": 4.298276807696294e-05, "loss": 0.0084, "step": 28380 }, { "grad_norm": 0.32200804352760315, "learning_rate": 4.2950027515088136e-05, "loss": 0.0086, "step": 28390 }, { "grad_norm": 0.2138897031545639, "learning_rate": 4.291729003711083e-05, "loss": 0.0069, "step": 28400 }, { "grad_norm": 0.16773097217082977, "learning_rate": 4.2884555657351455e-05, "loss": 0.01, "step": 28410 }, { "grad_norm": 1.8370649814605713, "learning_rate": 4.285182439012921e-05, "loss": 0.0082, "step": 28420 }, { "grad_norm": 0.1519298106431961, "learning_rate": 4.281909624976183e-05, "loss": 0.009, "step": 28430 }, { "grad_norm": 0.1428373157978058, "learning_rate": 4.278637125056568e-05, "loss": 0.0089, "step": 28440 }, { "grad_norm": 0.20083899796009064, "learning_rate": 4.2753649406855855e-05, "loss": 0.0075, "step": 28450 }, { "grad_norm": 0.25625693798065186, "learning_rate": 4.2720930732945966e-05, "loss": 0.0089, "step": 28460 }, { "grad_norm": 0.17675966024398804, "learning_rate": 4.268821524314826e-05, "loss": 0.0096, "step": 28470 }, { "grad_norm": 0.1700308620929718, "learning_rate": 4.2655502951773623e-05, "loss": 0.0096, "step": 28480 }, { "grad_norm": 0.2747124135494232, "learning_rate": 4.262279387313152e-05, "loss": 0.008, "step": 28490 }, { "grad_norm": 0.40901634097099304, "learning_rate": 4.259008802153001e-05, "loss": 0.0075, "step": 28500 }, { "grad_norm": 0.3981708288192749, "learning_rate": 4.2557385411275735e-05, "loss": 0.0076, "step": 28510 }, { "grad_norm": 0.14842349290847778, "learning_rate": 4.252468605667395e-05, "loss": 0.0073, "step": 28520 }, { "grad_norm": 0.15653887391090393, "learning_rate": 4.2491989972028456e-05, "loss": 0.0083, "step": 28530 }, { "grad_norm": 0.14155802130699158, "learning_rate": 4.245929717164161e-05, "loss": 0.0078, "step": 28540 }, { "grad_norm": 0.17721381783485413, "learning_rate": 4.242660766981438e-05, "loss": 0.0086, "step": 28550 }, { "grad_norm": 0.21933411061763763, "learning_rate": 4.2393921480846264e-05, "loss": 0.0093, "step": 28560 }, { "grad_norm": 0.15535254776477814, "learning_rate": 4.236123861903528e-05, "loss": 0.0075, "step": 28570 }, { "grad_norm": 0.1609506756067276, "learning_rate": 4.232855909867806e-05, "loss": 0.0074, "step": 28580 }, { "grad_norm": 0.18954993784427643, "learning_rate": 4.229588293406972e-05, "loss": 0.0079, "step": 28590 }, { "grad_norm": 0.14661164581775665, "learning_rate": 4.226321013950392e-05, "loss": 0.0064, "step": 28600 }, { "grad_norm": 0.17912356555461884, "learning_rate": 4.223054072927286e-05, "loss": 0.0069, "step": 28610 }, { "grad_norm": 0.18469814956188202, "learning_rate": 4.219787471766724e-05, "loss": 0.0081, "step": 28620 }, { "grad_norm": 0.2923794984817505, "learning_rate": 4.216521211897629e-05, "loss": 0.0093, "step": 28630 }, { "grad_norm": 0.2073487937450409, "learning_rate": 4.2132552947487734e-05, "loss": 0.0074, "step": 28640 }, { "grad_norm": 0.15747101604938507, "learning_rate": 4.209989721748783e-05, "loss": 0.0078, "step": 28650 }, { "grad_norm": 0.16988207399845123, "learning_rate": 4.206724494326127e-05, "loss": 0.0067, "step": 28660 }, { "grad_norm": 0.17110754549503326, "learning_rate": 4.203459613909126e-05, "loss": 0.0076, "step": 28670 }, { "grad_norm": 0.22912272810935974, "learning_rate": 4.200195081925955e-05, "loss": 0.0087, "step": 28680 }, { "grad_norm": 0.18307851254940033, "learning_rate": 4.196930899804627e-05, "loss": 0.0088, "step": 28690 }, { "grad_norm": 0.2063833326101303, "learning_rate": 4.1936670689730044e-05, "loss": 0.0088, "step": 28700 }, { "grad_norm": 0.13665974140167236, "learning_rate": 4.190403590858804e-05, "loss": 0.0085, "step": 28710 }, { "grad_norm": 0.1571311354637146, "learning_rate": 4.187140466889575e-05, "loss": 0.0066, "step": 28720 }, { "grad_norm": 0.19115063548088074, "learning_rate": 4.183877698492724e-05, "loss": 0.0078, "step": 28730 }, { "grad_norm": 0.1498035043478012, "learning_rate": 4.1806152870954935e-05, "loss": 0.0067, "step": 28740 }, { "grad_norm": 0.1966436207294464, "learning_rate": 4.177353234124976e-05, "loss": 0.0083, "step": 28750 }, { "grad_norm": 0.09478287398815155, "learning_rate": 4.1740915410081e-05, "loss": 0.0077, "step": 28760 }, { "grad_norm": 0.45141714811325073, "learning_rate": 4.170830209171646e-05, "loss": 0.0089, "step": 28770 }, { "grad_norm": 0.3697056174278259, "learning_rate": 4.1675692400422286e-05, "loss": 0.0092, "step": 28780 }, { "grad_norm": 0.19473899900913239, "learning_rate": 4.164308635046304e-05, "loss": 0.0069, "step": 28790 }, { "grad_norm": 0.1719372272491455, "learning_rate": 4.161048395610177e-05, "loss": 0.0079, "step": 28800 }, { "grad_norm": 0.21181802451610565, "learning_rate": 4.157788523159985e-05, "loss": 0.0071, "step": 28810 }, { "grad_norm": 0.129121795296669, "learning_rate": 4.154529019121706e-05, "loss": 0.009, "step": 28820 }, { "grad_norm": 0.20428310334682465, "learning_rate": 4.1512698849211564e-05, "loss": 0.0073, "step": 28830 }, { "grad_norm": 0.24529606103897095, "learning_rate": 4.1480111219839954e-05, "loss": 0.0086, "step": 28840 }, { "grad_norm": 0.24157676100730896, "learning_rate": 4.144752731735714e-05, "loss": 0.0074, "step": 28850 }, { "grad_norm": 0.12080275267362595, "learning_rate": 4.1414947156016446e-05, "loss": 0.0074, "step": 28860 }, { "grad_norm": 0.1856779009103775, "learning_rate": 4.138237075006953e-05, "loss": 0.0075, "step": 28870 }, { "grad_norm": 0.25361308455467224, "learning_rate": 4.134979811376645e-05, "loss": 0.008, "step": 28880 }, { "grad_norm": 0.13154172897338867, "learning_rate": 4.131722926135552e-05, "loss": 0.0065, "step": 28890 }, { "grad_norm": 0.18576553463935852, "learning_rate": 4.1284664207083526e-05, "loss": 0.0079, "step": 28900 }, { "grad_norm": 0.17728561162948608, "learning_rate": 4.125210296519551e-05, "loss": 0.0083, "step": 28910 }, { "grad_norm": 0.15116845071315765, "learning_rate": 4.121954554993484e-05, "loss": 0.0074, "step": 28920 }, { "grad_norm": 0.6298755407333374, "learning_rate": 4.118699197554327e-05, "loss": 0.0091, "step": 28930 }, { "grad_norm": 0.19051717221736908, "learning_rate": 4.1154442256260814e-05, "loss": 0.0067, "step": 28940 }, { "grad_norm": 0.15447354316711426, "learning_rate": 4.1121896406325824e-05, "loss": 0.0083, "step": 28950 }, { "grad_norm": 0.20035788416862488, "learning_rate": 4.108935443997499e-05, "loss": 0.0092, "step": 28960 }, { "grad_norm": 0.20596365630626678, "learning_rate": 4.105681637144325e-05, "loss": 0.0077, "step": 28970 }, { "grad_norm": 0.15674370527267456, "learning_rate": 4.102428221496387e-05, "loss": 0.0073, "step": 28980 }, { "grad_norm": 1.7981736660003662, "learning_rate": 4.099175198476838e-05, "loss": 0.0087, "step": 28990 }, { "grad_norm": 0.5605013966560364, "learning_rate": 4.0959225695086635e-05, "loss": 0.009, "step": 29000 }, { "grad_norm": 0.18034055829048157, "learning_rate": 4.0926703360146724e-05, "loss": 0.0086, "step": 29010 }, { "grad_norm": 0.14541997015476227, "learning_rate": 4.0894184994175e-05, "loss": 0.0071, "step": 29020 }, { "grad_norm": 0.17630474269390106, "learning_rate": 4.086167061139614e-05, "loss": 0.0069, "step": 29030 }, { "grad_norm": 0.22182443737983704, "learning_rate": 4.082916022603303e-05, "loss": 0.0102, "step": 29040 }, { "grad_norm": 0.3403959274291992, "learning_rate": 4.079665385230678e-05, "loss": 0.0069, "step": 29050 }, { "grad_norm": 0.5063358545303345, "learning_rate": 4.076415150443683e-05, "loss": 0.0086, "step": 29060 }, { "grad_norm": 0.1982097625732422, "learning_rate": 4.073165319664079e-05, "loss": 0.0077, "step": 29070 }, { "grad_norm": 0.21237854659557343, "learning_rate": 4.069915894313451e-05, "loss": 0.0076, "step": 29080 }, { "grad_norm": 0.26104041934013367, "learning_rate": 4.066666875813212e-05, "loss": 0.0069, "step": 29090 }, { "grad_norm": 0.15997929871082306, "learning_rate": 4.063418265584589e-05, "loss": 0.0074, "step": 29100 }, { "grad_norm": 2.39125919342041, "learning_rate": 4.060170065048636e-05, "loss": 0.0063, "step": 29110 }, { "grad_norm": 0.29336461424827576, "learning_rate": 4.056922275626227e-05, "loss": 0.007, "step": 29120 }, { "grad_norm": 0.2972928285598755, "learning_rate": 4.053674898738056e-05, "loss": 0.007, "step": 29130 }, { "grad_norm": 0.2345687448978424, "learning_rate": 4.050427935804633e-05, "loss": 0.0096, "step": 29140 }, { "grad_norm": 0.2933153212070465, "learning_rate": 4.047181388246293e-05, "loss": 0.0086, "step": 29150 }, { "grad_norm": 0.20010055601596832, "learning_rate": 4.043935257483185e-05, "loss": 0.0096, "step": 29160 }, { "grad_norm": 0.1969807893037796, "learning_rate": 4.0406895449352775e-05, "loss": 0.0086, "step": 29170 }, { "grad_norm": 0.16416116058826447, "learning_rate": 4.0374442520223525e-05, "loss": 0.0074, "step": 29180 }, { "grad_norm": 1.6677368879318237, "learning_rate": 4.0341993801640155e-05, "loss": 0.0092, "step": 29190 }, { "grad_norm": 0.20496384799480438, "learning_rate": 4.030954930779681e-05, "loss": 0.0093, "step": 29200 }, { "grad_norm": 0.15332649648189545, "learning_rate": 4.027710905288582e-05, "loss": 0.0086, "step": 29210 }, { "grad_norm": 0.14601081609725952, "learning_rate": 4.0244673051097666e-05, "loss": 0.0079, "step": 29220 }, { "grad_norm": 0.2561589479446411, "learning_rate": 4.021224131662095e-05, "loss": 0.0083, "step": 29230 }, { "grad_norm": 0.13163268566131592, "learning_rate": 4.0179813863642396e-05, "loss": 0.0085, "step": 29240 }, { "grad_norm": 0.1906137317419052, "learning_rate": 4.014739070634691e-05, "loss": 0.0081, "step": 29250 }, { "grad_norm": 0.889896810054779, "learning_rate": 4.011497185891746e-05, "loss": 0.0075, "step": 29260 }, { "grad_norm": 0.403948038816452, "learning_rate": 4.008255733553513e-05, "loss": 0.0078, "step": 29270 }, { "grad_norm": 0.2856982946395874, "learning_rate": 4.0050147150379194e-05, "loss": 0.0077, "step": 29280 }, { "grad_norm": 0.19922365248203278, "learning_rate": 4.001774131762692e-05, "loss": 0.0079, "step": 29290 }, { "grad_norm": 0.27301982045173645, "learning_rate": 3.9985339851453705e-05, "loss": 0.0088, "step": 29300 }, { "grad_norm": 0.18864205479621887, "learning_rate": 3.9952942766033114e-05, "loss": 0.0077, "step": 29310 }, { "grad_norm": 0.27125293016433716, "learning_rate": 3.9920550075536686e-05, "loss": 0.0076, "step": 29320 }, { "grad_norm": 0.20983846485614777, "learning_rate": 3.988816179413408e-05, "loss": 0.009, "step": 29330 }, { "grad_norm": 0.23284991085529327, "learning_rate": 3.985577793599307e-05, "loss": 0.0077, "step": 29340 }, { "grad_norm": 0.28384196758270264, "learning_rate": 3.982339851527942e-05, "loss": 0.0084, "step": 29350 }, { "grad_norm": 0.3515392541885376, "learning_rate": 3.979102354615701e-05, "loss": 0.0083, "step": 29360 }, { "grad_norm": 0.14557835459709167, "learning_rate": 3.975865304278772e-05, "loss": 0.0073, "step": 29370 }, { "grad_norm": 0.22076500952243805, "learning_rate": 3.9726287019331546e-05, "loss": 0.0077, "step": 29380 }, { "grad_norm": 0.13729864358901978, "learning_rate": 3.9693925489946474e-05, "loss": 0.0058, "step": 29390 }, { "grad_norm": 0.2542385160923004, "learning_rate": 3.966156846878852e-05, "loss": 0.0068, "step": 29400 }, { "grad_norm": 0.22847381234169006, "learning_rate": 3.962921597001177e-05, "loss": 0.0064, "step": 29410 }, { "grad_norm": 0.14232103526592255, "learning_rate": 3.9596868007768294e-05, "loss": 0.0073, "step": 29420 }, { "grad_norm": 0.22923558950424194, "learning_rate": 3.9564524596208165e-05, "loss": 0.0088, "step": 29430 }, { "grad_norm": 0.2169714868068695, "learning_rate": 3.953218574947954e-05, "loss": 0.0075, "step": 29440 }, { "grad_norm": 0.26472797989845276, "learning_rate": 3.949985148172849e-05, "loss": 0.0067, "step": 29450 }, { "grad_norm": 0.1587640643119812, "learning_rate": 3.946752180709914e-05, "loss": 0.008, "step": 29460 }, { "grad_norm": 1.7275604009628296, "learning_rate": 3.943519673973357e-05, "loss": 0.0078, "step": 29470 }, { "grad_norm": 0.1461465060710907, "learning_rate": 3.940287629377189e-05, "loss": 0.0077, "step": 29480 }, { "grad_norm": 0.9208916425704956, "learning_rate": 3.937056048335214e-05, "loss": 0.008, "step": 29490 }, { "grad_norm": 1.781956434249878, "learning_rate": 3.933824932261037e-05, "loss": 0.0093, "step": 29500 }, { "grad_norm": 0.5191961526870728, "learning_rate": 3.9305942825680576e-05, "loss": 0.0087, "step": 29510 }, { "grad_norm": 0.22414369881153107, "learning_rate": 3.9273641006694706e-05, "loss": 0.0099, "step": 29520 }, { "grad_norm": 0.22398565709590912, "learning_rate": 3.924134387978266e-05, "loss": 0.0068, "step": 29530 }, { "grad_norm": 0.18576066195964813, "learning_rate": 3.920905145907234e-05, "loss": 0.0096, "step": 29540 }, { "grad_norm": 0.7598080635070801, "learning_rate": 3.9176763758689525e-05, "loss": 0.0075, "step": 29550 }, { "grad_norm": 0.2242153286933899, "learning_rate": 3.9144480792757925e-05, "loss": 0.0082, "step": 29560 }, { "grad_norm": 0.16222113370895386, "learning_rate": 3.911220257539926e-05, "loss": 0.0071, "step": 29570 }, { "grad_norm": 0.25328314304351807, "learning_rate": 3.907992912073308e-05, "loss": 0.0089, "step": 29580 }, { "grad_norm": 0.27407100796699524, "learning_rate": 3.90476604428769e-05, "loss": 0.0065, "step": 29590 }, { "grad_norm": 0.21491628885269165, "learning_rate": 3.901539655594614e-05, "loss": 0.0089, "step": 29600 }, { "grad_norm": 0.1592937856912613, "learning_rate": 3.898313747405413e-05, "loss": 0.0091, "step": 29610 }, { "grad_norm": 0.953508734703064, "learning_rate": 3.895088321131205e-05, "loss": 0.0089, "step": 29620 }, { "grad_norm": 0.28885167837142944, "learning_rate": 3.891863378182907e-05, "loss": 0.008, "step": 29630 }, { "grad_norm": 0.21603211760520935, "learning_rate": 3.8886389199712145e-05, "loss": 0.0071, "step": 29640 }, { "grad_norm": 0.111051544547081, "learning_rate": 3.885414947906614e-05, "loss": 0.0079, "step": 29650 }, { "grad_norm": 0.21008946001529694, "learning_rate": 3.882191463399385e-05, "loss": 0.0078, "step": 29660 }, { "grad_norm": 0.15923289954662323, "learning_rate": 3.878968467859585e-05, "loss": 0.0058, "step": 29670 }, { "grad_norm": 0.23186242580413818, "learning_rate": 3.875745962697063e-05, "loss": 0.0075, "step": 29680 }, { "grad_norm": 0.20158438384532928, "learning_rate": 3.872523949321454e-05, "loss": 0.0065, "step": 29690 }, { "grad_norm": 0.5266711711883545, "learning_rate": 3.869302429142173e-05, "loss": 0.008, "step": 29700 }, { "grad_norm": 2.593003034591675, "learning_rate": 3.866081403568425e-05, "loss": 0.008, "step": 29710 }, { "grad_norm": 0.4560745060443878, "learning_rate": 3.8628608740091934e-05, "loss": 0.0083, "step": 29720 }, { "grad_norm": 0.31288954615592957, "learning_rate": 3.85964084187325e-05, "loss": 0.0124, "step": 29730 }, { "grad_norm": 0.1514764428138733, "learning_rate": 3.856421308569146e-05, "loss": 0.0068, "step": 29740 }, { "grad_norm": 0.1586446464061737, "learning_rate": 3.853202275505212e-05, "loss": 0.0082, "step": 29750 }, { "grad_norm": 0.14273624122142792, "learning_rate": 3.849983744089565e-05, "loss": 0.0068, "step": 29760 }, { "grad_norm": 0.2407248318195343, "learning_rate": 3.8467657157301e-05, "loss": 0.0087, "step": 29770 }, { "grad_norm": 0.22131812572479248, "learning_rate": 3.8435481918344884e-05, "loss": 0.0084, "step": 29780 }, { "grad_norm": 0.13935694098472595, "learning_rate": 3.84033117381019e-05, "loss": 0.0069, "step": 29790 }, { "grad_norm": 0.2383270263671875, "learning_rate": 3.8371146630644335e-05, "loss": 0.0082, "step": 29800 }, { "grad_norm": 6.5985918045043945, "learning_rate": 3.833898661004231e-05, "loss": 0.0075, "step": 29810 }, { "grad_norm": 0.12980397045612335, "learning_rate": 3.830683169036372e-05, "loss": 0.0069, "step": 29820 }, { "grad_norm": 0.1874266266822815, "learning_rate": 3.827468188567422e-05, "loss": 0.0078, "step": 29830 }, { "grad_norm": 0.22429928183555603, "learning_rate": 3.82425372100372e-05, "loss": 0.0074, "step": 29840 }, { "grad_norm": 0.23236559331417084, "learning_rate": 3.821039767751387e-05, "loss": 0.0083, "step": 29850 }, { "grad_norm": 0.1351011097431183, "learning_rate": 3.817826330216315e-05, "loss": 0.0097, "step": 29860 }, { "grad_norm": 0.23635078966617584, "learning_rate": 3.814613409804169e-05, "loss": 0.0087, "step": 29870 }, { "grad_norm": 0.22028440237045288, "learning_rate": 3.811401007920388e-05, "loss": 0.0103, "step": 29880 }, { "grad_norm": 0.29102247953414917, "learning_rate": 3.8081891259701895e-05, "loss": 0.0084, "step": 29890 }, { "grad_norm": 0.19887053966522217, "learning_rate": 3.8049777653585584e-05, "loss": 0.009, "step": 29900 }, { "grad_norm": 0.3209555149078369, "learning_rate": 3.80176692749025e-05, "loss": 0.0083, "step": 29910 }, { "grad_norm": 0.16557995975017548, "learning_rate": 3.798556613769799e-05, "loss": 0.0075, "step": 29920 }, { "grad_norm": 1.4715430736541748, "learning_rate": 3.795346825601502e-05, "loss": 0.0106, "step": 29930 }, { "grad_norm": 0.18962311744689941, "learning_rate": 3.7921375643894305e-05, "loss": 0.0078, "step": 29940 }, { "grad_norm": 0.13641640543937683, "learning_rate": 3.788928831537424e-05, "loss": 0.0072, "step": 29950 }, { "grad_norm": 0.1667749285697937, "learning_rate": 3.7857206284490934e-05, "loss": 0.0058, "step": 29960 }, { "grad_norm": 0.12236186861991882, "learning_rate": 3.7825129565278114e-05, "loss": 0.0066, "step": 29970 }, { "grad_norm": 0.10328030586242676, "learning_rate": 3.779305817176728e-05, "loss": 0.0057, "step": 29980 }, { "grad_norm": 0.10793677717447281, "learning_rate": 3.776099211798751e-05, "loss": 0.0063, "step": 29990 }, { "grad_norm": 0.15340065956115723, "learning_rate": 3.7728931417965584e-05, "loss": 0.0066, "step": 30000 }, { "grad_norm": 0.2521892786026001, "learning_rate": 3.769687608572598e-05, "loss": 0.0066, "step": 30010 }, { "grad_norm": 0.2815264165401459, "learning_rate": 3.766482613529076e-05, "loss": 0.009, "step": 30020 }, { "grad_norm": 0.2803257405757904, "learning_rate": 3.763278158067964e-05, "loss": 0.0067, "step": 30030 }, { "grad_norm": 0.09313266724348068, "learning_rate": 3.7600742435910045e-05, "loss": 0.0073, "step": 30040 }, { "grad_norm": 0.13435383141040802, "learning_rate": 3.756870871499696e-05, "loss": 0.0076, "step": 30050 }, { "grad_norm": 0.2813585698604584, "learning_rate": 3.7536680431953025e-05, "loss": 0.0084, "step": 30060 }, { "grad_norm": 0.145328551530838, "learning_rate": 3.7504657600788485e-05, "loss": 0.0077, "step": 30070 }, { "grad_norm": 0.12638525664806366, "learning_rate": 3.747264023551125e-05, "loss": 0.0065, "step": 30080 }, { "grad_norm": 0.4782331883907318, "learning_rate": 3.744062835012678e-05, "loss": 0.0084, "step": 30090 }, { "grad_norm": 5.527360439300537, "learning_rate": 3.740862195863814e-05, "loss": 0.0076, "step": 30100 }, { "grad_norm": 0.29874342679977417, "learning_rate": 3.737662107504606e-05, "loss": 0.0076, "step": 30110 }, { "grad_norm": 0.1758892685174942, "learning_rate": 3.734462571334879e-05, "loss": 0.0064, "step": 30120 }, { "grad_norm": 0.30519813299179077, "learning_rate": 3.731263588754217e-05, "loss": 0.0074, "step": 30130 }, { "grad_norm": 0.3032386004924774, "learning_rate": 3.7280651611619665e-05, "loss": 0.0099, "step": 30140 }, { "grad_norm": 0.23476238548755646, "learning_rate": 3.724867289957229e-05, "loss": 0.0087, "step": 30150 }, { "grad_norm": 0.42118191719055176, "learning_rate": 3.7216699765388576e-05, "loss": 0.0084, "step": 30160 }, { "grad_norm": 0.2660072147846222, "learning_rate": 3.71847322230547e-05, "loss": 0.0105, "step": 30170 }, { "grad_norm": 0.4290156364440918, "learning_rate": 3.7152770286554335e-05, "loss": 0.0075, "step": 30180 }, { "grad_norm": 0.3172979950904846, "learning_rate": 3.712081396986872e-05, "loss": 0.0062, "step": 30190 }, { "grad_norm": 0.1279105842113495, "learning_rate": 3.708886328697663e-05, "loss": 0.0078, "step": 30200 }, { "grad_norm": 0.22145357728004456, "learning_rate": 3.7056918251854395e-05, "loss": 0.0089, "step": 30210 }, { "grad_norm": 0.19342365860939026, "learning_rate": 3.7024978878475846e-05, "loss": 0.0072, "step": 30220 }, { "grad_norm": 0.1906013935804367, "learning_rate": 3.699304518081233e-05, "loss": 0.006, "step": 30230 }, { "grad_norm": 0.19411800801753998, "learning_rate": 3.696111717283277e-05, "loss": 0.0073, "step": 30240 }, { "grad_norm": 0.27388155460357666, "learning_rate": 3.692919486850355e-05, "loss": 0.0113, "step": 30250 }, { "grad_norm": 0.16125933825969696, "learning_rate": 3.689727828178854e-05, "loss": 0.0069, "step": 30260 }, { "grad_norm": 0.15177875757217407, "learning_rate": 3.6865367426649185e-05, "loss": 0.0077, "step": 30270 }, { "grad_norm": 0.21618565917015076, "learning_rate": 3.6833462317044346e-05, "loss": 0.01, "step": 30280 }, { "grad_norm": 0.171136274933815, "learning_rate": 3.6801562966930406e-05, "loss": 0.0066, "step": 30290 }, { "grad_norm": 0.2452395111322403, "learning_rate": 3.676966939026123e-05, "loss": 0.0068, "step": 30300 }, { "grad_norm": 0.14275965094566345, "learning_rate": 3.6737781600988166e-05, "loss": 0.0072, "step": 30310 }, { "grad_norm": 0.17676807940006256, "learning_rate": 3.670589961305998e-05, "loss": 0.0081, "step": 30320 }, { "grad_norm": 1.589040994644165, "learning_rate": 3.667402344042298e-05, "loss": 0.0066, "step": 30330 }, { "grad_norm": 0.17998020350933075, "learning_rate": 3.664215309702086e-05, "loss": 0.0086, "step": 30340 }, { "grad_norm": 0.23998092114925385, "learning_rate": 3.661028859679478e-05, "loss": 0.0086, "step": 30350 }, { "grad_norm": 0.25422361493110657, "learning_rate": 3.657842995368339e-05, "loss": 0.0084, "step": 30360 }, { "grad_norm": 0.17156299948692322, "learning_rate": 3.654657718162273e-05, "loss": 0.0108, "step": 30370 }, { "grad_norm": 0.16936489939689636, "learning_rate": 3.6514730294546275e-05, "loss": 0.007, "step": 30380 }, { "grad_norm": 0.2529628872871399, "learning_rate": 3.648288930638492e-05, "loss": 0.0067, "step": 30390 }, { "grad_norm": 0.13106510043144226, "learning_rate": 3.645105423106704e-05, "loss": 0.0063, "step": 30400 }, { "grad_norm": 0.13550357520580292, "learning_rate": 3.6419225082518336e-05, "loss": 0.0072, "step": 30410 }, { "grad_norm": 2.0444588661193848, "learning_rate": 3.6387401874661975e-05, "loss": 0.0087, "step": 30420 }, { "grad_norm": 0.33080044388771057, "learning_rate": 3.6355584621418516e-05, "loss": 0.0079, "step": 30430 }, { "grad_norm": 0.27561551332473755, "learning_rate": 3.632377333670591e-05, "loss": 0.0072, "step": 30440 }, { "grad_norm": 0.2385382503271103, "learning_rate": 3.629196803443946e-05, "loss": 0.0074, "step": 30450 }, { "grad_norm": 0.11726775765419006, "learning_rate": 3.626016872853193e-05, "loss": 0.0061, "step": 30460 }, { "grad_norm": 0.1505720168352127, "learning_rate": 3.6228375432893405e-05, "loss": 0.0071, "step": 30470 }, { "grad_norm": 0.17424918711185455, "learning_rate": 3.619658816143131e-05, "loss": 0.006, "step": 30480 }, { "grad_norm": 0.1870856136083603, "learning_rate": 3.6164806928050544e-05, "loss": 0.0056, "step": 30490 }, { "grad_norm": 0.2725129723548889, "learning_rate": 3.613303174665327e-05, "loss": 0.0083, "step": 30500 }, { "grad_norm": 0.19922895729541779, "learning_rate": 3.6101262631138996e-05, "loss": 0.0069, "step": 30510 }, { "grad_norm": 0.16408851742744446, "learning_rate": 3.606949959540467e-05, "loss": 0.0061, "step": 30520 }, { "grad_norm": 0.3407064974308014, "learning_rate": 3.603774265334449e-05, "loss": 0.0089, "step": 30530 }, { "grad_norm": 0.24840454757213593, "learning_rate": 3.600599181885003e-05, "loss": 0.0075, "step": 30540 }, { "grad_norm": 0.21652410924434662, "learning_rate": 3.597424710581018e-05, "loss": 0.0075, "step": 30550 }, { "grad_norm": 0.17807860672473907, "learning_rate": 3.5942508528111164e-05, "loss": 0.0078, "step": 30560 }, { "grad_norm": 4.917682647705078, "learning_rate": 3.591077609963651e-05, "loss": 0.0104, "step": 30570 }, { "grad_norm": 0.17281369864940643, "learning_rate": 3.587904983426704e-05, "loss": 0.0077, "step": 30580 }, { "grad_norm": 0.19385550916194916, "learning_rate": 3.584732974588094e-05, "loss": 0.0068, "step": 30590 }, { "grad_norm": 0.1558273583650589, "learning_rate": 3.5815615848353645e-05, "loss": 0.0061, "step": 30600 }, { "grad_norm": 0.17734301090240479, "learning_rate": 3.578390815555784e-05, "loss": 0.0078, "step": 30610 }, { "grad_norm": 3.0951900482177734, "learning_rate": 3.5752206681363626e-05, "loss": 0.0051, "step": 30620 }, { "grad_norm": 0.13683094084262848, "learning_rate": 3.572051143963826e-05, "loss": 0.0076, "step": 30630 }, { "grad_norm": 0.1798025220632553, "learning_rate": 3.5688822444246294e-05, "loss": 0.0073, "step": 30640 }, { "grad_norm": 0.10271057486534119, "learning_rate": 3.565713970904963e-05, "loss": 0.0071, "step": 30650 }, { "grad_norm": 0.2199382185935974, "learning_rate": 3.562546324790733e-05, "loss": 0.0075, "step": 30660 }, { "grad_norm": 2.760183811187744, "learning_rate": 3.559379307467577e-05, "loss": 0.0064, "step": 30670 }, { "grad_norm": 0.17177079617977142, "learning_rate": 3.556212920320856e-05, "loss": 0.0069, "step": 30680 }, { "grad_norm": 0.23246453702449799, "learning_rate": 3.553047164735655e-05, "loss": 0.007, "step": 30690 }, { "grad_norm": 1.5406688451766968, "learning_rate": 3.549882042096783e-05, "loss": 0.0085, "step": 30700 }, { "grad_norm": 0.2831099033355713, "learning_rate": 3.546717553788772e-05, "loss": 0.0081, "step": 30710 }, { "grad_norm": 0.36335551738739014, "learning_rate": 3.543553701195879e-05, "loss": 0.0071, "step": 30720 }, { "grad_norm": 0.90294349193573, "learning_rate": 3.540390485702077e-05, "loss": 0.0076, "step": 30730 }, { "grad_norm": 0.1484231948852539, "learning_rate": 3.5372279086910645e-05, "loss": 0.0082, "step": 30740 }, { "grad_norm": 0.12661628425121307, "learning_rate": 3.5340659715462635e-05, "loss": 0.0054, "step": 30750 }, { "grad_norm": 0.1750911921262741, "learning_rate": 3.530904675650808e-05, "loss": 0.0083, "step": 30760 }, { "grad_norm": 0.12844841182231903, "learning_rate": 3.527744022387559e-05, "loss": 0.0073, "step": 30770 }, { "grad_norm": 0.20263932645320892, "learning_rate": 3.524584013139094e-05, "loss": 0.008, "step": 30780 }, { "grad_norm": 0.20591023564338684, "learning_rate": 3.521424649287708e-05, "loss": 0.0064, "step": 30790 }, { "grad_norm": 0.18505200743675232, "learning_rate": 3.518265932215412e-05, "loss": 0.0081, "step": 30800 }, { "grad_norm": 0.5106185078620911, "learning_rate": 3.51510786330394e-05, "loss": 0.0067, "step": 30810 }, { "grad_norm": 1.1490917205810547, "learning_rate": 3.511950443934736e-05, "loss": 0.0071, "step": 30820 }, { "grad_norm": 0.6424381732940674, "learning_rate": 3.508793675488961e-05, "loss": 0.0086, "step": 30830 }, { "grad_norm": 0.14553464949131012, "learning_rate": 3.505637559347498e-05, "loss": 0.0067, "step": 30840 }, { "grad_norm": 0.8425905108451843, "learning_rate": 3.5024820968909345e-05, "loss": 0.0066, "step": 30850 }, { "grad_norm": 0.15550659596920013, "learning_rate": 3.4993272894995774e-05, "loss": 0.0087, "step": 30860 }, { "grad_norm": 0.37481629848480225, "learning_rate": 3.49617313855345e-05, "loss": 0.0098, "step": 30870 }, { "grad_norm": 0.2712935507297516, "learning_rate": 3.4930196454322834e-05, "loss": 0.0079, "step": 30880 }, { "grad_norm": 0.30631205439567566, "learning_rate": 3.48986681151552e-05, "loss": 0.0086, "step": 30890 }, { "grad_norm": 0.17528316378593445, "learning_rate": 3.4867146381823204e-05, "loss": 0.0073, "step": 30900 }, { "grad_norm": 0.3933018445968628, "learning_rate": 3.48356312681155e-05, "loss": 0.0079, "step": 30910 }, { "grad_norm": 0.11210078001022339, "learning_rate": 3.480412278781788e-05, "loss": 0.0071, "step": 30920 }, { "grad_norm": 0.25320109724998474, "learning_rate": 3.47726209547132e-05, "loss": 0.0074, "step": 30930 }, { "grad_norm": 0.1558327078819275, "learning_rate": 3.474112578258147e-05, "loss": 0.0078, "step": 30940 }, { "grad_norm": 3.014529228210449, "learning_rate": 3.470963728519974e-05, "loss": 0.0069, "step": 30950 }, { "grad_norm": 0.1376621276140213, "learning_rate": 3.467815547634211e-05, "loss": 0.0081, "step": 30960 }, { "grad_norm": 0.27114325761795044, "learning_rate": 3.4646680369779836e-05, "loss": 0.0071, "step": 30970 }, { "grad_norm": 1.768790602684021, "learning_rate": 3.46152119792812e-05, "loss": 0.0066, "step": 30980 }, { "grad_norm": 0.11818020045757294, "learning_rate": 3.45837503186115e-05, "loss": 0.0092, "step": 30990 }, { "grad_norm": 0.2507489323616028, "learning_rate": 3.455229540153318e-05, "loss": 0.0065, "step": 31000 }, { "grad_norm": 0.19208240509033203, "learning_rate": 3.4520847241805665e-05, "loss": 0.0077, "step": 31010 }, { "grad_norm": 0.15859895944595337, "learning_rate": 3.4489405853185465e-05, "loss": 0.0077, "step": 31020 }, { "grad_norm": 0.32577943801879883, "learning_rate": 3.44579712494261e-05, "loss": 0.0073, "step": 31030 }, { "grad_norm": 0.30343955755233765, "learning_rate": 3.442654344427814e-05, "loss": 0.0083, "step": 31040 }, { "grad_norm": 0.2090558260679245, "learning_rate": 3.439512245148916e-05, "loss": 0.0064, "step": 31050 }, { "grad_norm": 0.16837644577026367, "learning_rate": 3.4363708284803796e-05, "loss": 0.0064, "step": 31060 }, { "grad_norm": 0.16884863376617432, "learning_rate": 3.433230095796367e-05, "loss": 0.0062, "step": 31070 }, { "grad_norm": 2.087160587310791, "learning_rate": 3.430090048470739e-05, "loss": 0.0075, "step": 31080 }, { "grad_norm": 1.2404941320419312, "learning_rate": 3.4269506878770585e-05, "loss": 0.0076, "step": 31090 }, { "grad_norm": 0.17987936735153198, "learning_rate": 3.4238120153885925e-05, "loss": 0.0084, "step": 31100 }, { "grad_norm": 0.2056048959493637, "learning_rate": 3.4206740323782994e-05, "loss": 0.0067, "step": 31110 }, { "grad_norm": 0.18325380980968475, "learning_rate": 3.4175367402188395e-05, "loss": 0.0083, "step": 31120 }, { "grad_norm": 0.16447289288043976, "learning_rate": 3.414400140282574e-05, "loss": 0.0068, "step": 31130 }, { "grad_norm": 0.14339514076709747, "learning_rate": 3.411264233941556e-05, "loss": 0.0065, "step": 31140 }, { "grad_norm": 0.1619756519794464, "learning_rate": 3.408129022567537e-05, "loss": 0.0069, "step": 31150 }, { "grad_norm": 0.2310570329427719, "learning_rate": 3.404994507531968e-05, "loss": 0.0075, "step": 31160 }, { "grad_norm": 0.15162742137908936, "learning_rate": 3.40186069020599e-05, "loss": 0.0075, "step": 31170 }, { "grad_norm": 0.15691162645816803, "learning_rate": 3.3987275719604386e-05, "loss": 0.0072, "step": 31180 }, { "grad_norm": 0.1668512225151062, "learning_rate": 3.3955951541658526e-05, "loss": 0.0076, "step": 31190 }, { "grad_norm": 0.3255804181098938, "learning_rate": 3.392463438192454e-05, "loss": 0.0064, "step": 31200 }, { "grad_norm": 0.16985277831554413, "learning_rate": 3.38933242541016e-05, "loss": 0.0069, "step": 31210 }, { "grad_norm": 0.18003863096237183, "learning_rate": 3.386202117188586e-05, "loss": 0.008, "step": 31220 }, { "grad_norm": 0.16954867541790009, "learning_rate": 3.383072514897033e-05, "loss": 0.0064, "step": 31230 }, { "grad_norm": 0.21693618595600128, "learning_rate": 3.379943619904494e-05, "loss": 0.0095, "step": 31240 }, { "grad_norm": 0.33734816312789917, "learning_rate": 3.3768154335796583e-05, "loss": 0.0074, "step": 31250 }, { "grad_norm": 0.16356627643108368, "learning_rate": 3.3736879572908963e-05, "loss": 0.0062, "step": 31260 }, { "grad_norm": 0.2875787019729614, "learning_rate": 3.370561192406276e-05, "loss": 0.0076, "step": 31270 }, { "grad_norm": 0.290874183177948, "learning_rate": 3.367435140293547e-05, "loss": 0.009, "step": 31280 }, { "grad_norm": 0.19031448662281036, "learning_rate": 3.3643098023201545e-05, "loss": 0.0087, "step": 31290 }, { "grad_norm": 0.2315976917743683, "learning_rate": 3.361185179853225e-05, "loss": 0.0093, "step": 31300 }, { "grad_norm": 0.21330250799655914, "learning_rate": 3.358061274259574e-05, "loss": 0.0067, "step": 31310 }, { "grad_norm": 0.2913542687892914, "learning_rate": 3.3549380869057064e-05, "loss": 0.006, "step": 31320 }, { "grad_norm": 0.7288862466812134, "learning_rate": 3.351815619157809e-05, "loss": 0.0084, "step": 31330 }, { "grad_norm": 0.11735580861568451, "learning_rate": 3.348693872381754e-05, "loss": 0.0073, "step": 31340 }, { "grad_norm": 1.9052397012710571, "learning_rate": 3.3455728479431025e-05, "loss": 0.0084, "step": 31350 }, { "grad_norm": 1.6692060232162476, "learning_rate": 3.342452547207094e-05, "loss": 0.0078, "step": 31360 }, { "grad_norm": 0.665456235408783, "learning_rate": 3.3393329715386555e-05, "loss": 0.008, "step": 31370 }, { "grad_norm": 0.13505293428897858, "learning_rate": 3.336214122302395e-05, "loss": 0.0075, "step": 31380 }, { "grad_norm": 0.12738199532032013, "learning_rate": 3.333096000862604e-05, "loss": 0.0085, "step": 31390 }, { "grad_norm": 0.2222694456577301, "learning_rate": 3.3299786085832516e-05, "loss": 0.0068, "step": 31400 }, { "grad_norm": 0.1515454202890396, "learning_rate": 3.3268619468279956e-05, "loss": 0.0086, "step": 31410 }, { "grad_norm": 0.21385721862316132, "learning_rate": 3.323746016960167e-05, "loss": 0.0092, "step": 31420 }, { "grad_norm": 0.40262678265571594, "learning_rate": 3.3206308203427806e-05, "loss": 0.0068, "step": 31430 }, { "grad_norm": 0.25149911642074585, "learning_rate": 3.3175163583385254e-05, "loss": 0.0067, "step": 31440 }, { "grad_norm": 0.11320620030164719, "learning_rate": 3.314402632309779e-05, "loss": 0.0063, "step": 31450 }, { "grad_norm": 1.3413901329040527, "learning_rate": 3.311289643618587e-05, "loss": 0.0098, "step": 31460 }, { "grad_norm": 0.23453277349472046, "learning_rate": 3.308177393626676e-05, "loss": 0.007, "step": 31470 }, { "grad_norm": 0.8302075266838074, "learning_rate": 3.305065883695453e-05, "loss": 0.0062, "step": 31480 }, { "grad_norm": 0.11840437352657318, "learning_rate": 3.301955115185996e-05, "loss": 0.0084, "step": 31490 }, { "grad_norm": 0.2228572815656662, "learning_rate": 3.29884508945906e-05, "loss": 0.0073, "step": 31500 }, { "grad_norm": 0.3397265076637268, "learning_rate": 3.295735807875079e-05, "loss": 0.0067, "step": 31510 }, { "grad_norm": 0.16564728319644928, "learning_rate": 3.292627271794156e-05, "loss": 0.0062, "step": 31520 }, { "grad_norm": 0.2830057740211487, "learning_rate": 3.289519482576069e-05, "loss": 0.0053, "step": 31530 }, { "grad_norm": 0.1528569608926773, "learning_rate": 3.286412441580274e-05, "loss": 0.0078, "step": 31540 }, { "grad_norm": 0.2700439393520355, "learning_rate": 3.283306150165896e-05, "loss": 0.0091, "step": 31550 }, { "grad_norm": 0.45140692591667175, "learning_rate": 3.2802006096917285e-05, "loss": 0.0067, "step": 31560 }, { "grad_norm": 0.23982886970043182, "learning_rate": 3.277095821516245e-05, "loss": 0.006, "step": 31570 }, { "grad_norm": 0.24676188826560974, "learning_rate": 3.273991786997583e-05, "loss": 0.0061, "step": 31580 }, { "grad_norm": 0.1307685822248459, "learning_rate": 3.2708885074935514e-05, "loss": 0.0056, "step": 31590 }, { "grad_norm": 0.2918209135532379, "learning_rate": 3.267785984361633e-05, "loss": 0.007, "step": 31600 }, { "grad_norm": 0.2674775719642639, "learning_rate": 3.2646842189589754e-05, "loss": 0.006, "step": 31610 }, { "grad_norm": 0.15208199620246887, "learning_rate": 3.261583212642397e-05, "loss": 0.0087, "step": 31620 }, { "grad_norm": 0.12689797580242157, "learning_rate": 3.25848296676838e-05, "loss": 0.0053, "step": 31630 }, { "grad_norm": 0.14656901359558105, "learning_rate": 3.255383482693083e-05, "loss": 0.007, "step": 31640 }, { "grad_norm": 0.12692643702030182, "learning_rate": 3.252284761772322e-05, "loss": 0.0073, "step": 31650 }, { "grad_norm": 0.6035916805267334, "learning_rate": 3.249186805361581e-05, "loss": 0.0075, "step": 31660 }, { "grad_norm": 1.8136303424835205, "learning_rate": 3.246089614816016e-05, "loss": 0.0097, "step": 31670 }, { "grad_norm": 1.025485873222351, "learning_rate": 3.242993191490441e-05, "loss": 0.0078, "step": 31680 }, { "grad_norm": 0.154423788189888, "learning_rate": 3.239897536739335e-05, "loss": 0.0064, "step": 31690 }, { "grad_norm": 0.13966244459152222, "learning_rate": 3.236802651916845e-05, "loss": 0.0074, "step": 31700 }, { "grad_norm": 0.1982085406780243, "learning_rate": 3.23370853837678e-05, "loss": 0.0058, "step": 31710 }, { "grad_norm": 0.13489271700382233, "learning_rate": 3.2306151974726056e-05, "loss": 0.0068, "step": 31720 }, { "grad_norm": 0.2787174582481384, "learning_rate": 3.227522630557459e-05, "loss": 0.0072, "step": 31730 }, { "grad_norm": 0.18373186886310577, "learning_rate": 3.224430838984132e-05, "loss": 0.0071, "step": 31740 }, { "grad_norm": 0.16887308657169342, "learning_rate": 3.2213398241050784e-05, "loss": 0.0093, "step": 31750 }, { "grad_norm": 0.14318451285362244, "learning_rate": 3.218249587272415e-05, "loss": 0.0064, "step": 31760 }, { "grad_norm": 0.16707883775234222, "learning_rate": 3.215160129837916e-05, "loss": 0.0068, "step": 31770 }, { "grad_norm": 0.250136137008667, "learning_rate": 3.2120714531530146e-05, "loss": 0.0062, "step": 31780 }, { "grad_norm": 0.1920623481273651, "learning_rate": 3.2089835585688014e-05, "loss": 0.0077, "step": 31790 }, { "grad_norm": 0.14698421955108643, "learning_rate": 3.20589644743603e-05, "loss": 0.0073, "step": 31800 }, { "grad_norm": 0.16916127502918243, "learning_rate": 3.2028101211051054e-05, "loss": 0.0078, "step": 31810 }, { "grad_norm": 0.17510592937469482, "learning_rate": 3.19972458092609e-05, "loss": 0.0071, "step": 31820 }, { "grad_norm": 4.293372631072998, "learning_rate": 3.196639828248707e-05, "loss": 0.0075, "step": 31830 }, { "grad_norm": 7.307094097137451, "learning_rate": 3.193555864422329e-05, "loss": 0.0076, "step": 31840 }, { "grad_norm": 0.12673623859882355, "learning_rate": 3.19047269079599e-05, "loss": 0.007, "step": 31850 }, { "grad_norm": 0.13177178800106049, "learning_rate": 3.1873903087183734e-05, "loss": 0.0069, "step": 31860 }, { "grad_norm": 0.13800063729286194, "learning_rate": 3.1843087195378176e-05, "loss": 0.0072, "step": 31870 }, { "grad_norm": 0.13045300543308258, "learning_rate": 3.181227924602312e-05, "loss": 0.0059, "step": 31880 }, { "grad_norm": 0.061625465750694275, "learning_rate": 3.178147925259506e-05, "loss": 0.0062, "step": 31890 }, { "grad_norm": 0.16871277987957, "learning_rate": 3.175068722856693e-05, "loss": 0.006, "step": 31900 }, { "grad_norm": 0.11610434204339981, "learning_rate": 3.17199031874082e-05, "loss": 0.007, "step": 31910 }, { "grad_norm": 0.4797964096069336, "learning_rate": 3.168912714258487e-05, "loss": 0.0059, "step": 31920 }, { "grad_norm": 0.1450154334306717, "learning_rate": 3.165835910755944e-05, "loss": 0.0064, "step": 31930 }, { "grad_norm": 0.2820737659931183, "learning_rate": 3.1627599095790874e-05, "loss": 0.008, "step": 31940 }, { "grad_norm": 0.3264654278755188, "learning_rate": 3.159684712073462e-05, "loss": 0.0073, "step": 31950 }, { "grad_norm": 0.20148825645446777, "learning_rate": 3.1566103195842695e-05, "loss": 0.0089, "step": 31960 }, { "grad_norm": 0.16985216736793518, "learning_rate": 3.153536733456349e-05, "loss": 0.0075, "step": 31970 }, { "grad_norm": 0.22246409952640533, "learning_rate": 3.150463955034194e-05, "loss": 0.0077, "step": 31980 }, { "grad_norm": 0.22989732027053833, "learning_rate": 3.147391985661942e-05, "loss": 0.0064, "step": 31990 }, { "grad_norm": 0.24279719591140747, "learning_rate": 3.1443208266833755e-05, "loss": 0.0075, "step": 32000 }, { "grad_norm": 0.42186734080314636, "learning_rate": 3.141250479441924e-05, "loss": 0.009, "step": 32010 }, { "grad_norm": 0.14559657871723175, "learning_rate": 3.138180945280664e-05, "loss": 0.0067, "step": 32020 }, { "grad_norm": 0.1649354249238968, "learning_rate": 3.1351122255423116e-05, "loss": 0.007, "step": 32030 }, { "grad_norm": 0.25451165437698364, "learning_rate": 3.132044321569228e-05, "loss": 0.0068, "step": 32040 }, { "grad_norm": 0.27279430627822876, "learning_rate": 3.1289772347034226e-05, "loss": 0.0064, "step": 32050 }, { "grad_norm": 0.2583874762058258, "learning_rate": 3.1259109662865404e-05, "loss": 0.0073, "step": 32060 }, { "grad_norm": 0.15712659060955048, "learning_rate": 3.122845517659871e-05, "loss": 0.0053, "step": 32070 }, { "grad_norm": 0.3096923232078552, "learning_rate": 3.119780890164348e-05, "loss": 0.0078, "step": 32080 }, { "grad_norm": 0.8187267184257507, "learning_rate": 3.116717085140542e-05, "loss": 0.0075, "step": 32090 }, { "grad_norm": 0.32058051228523254, "learning_rate": 3.1136541039286657e-05, "loss": 0.0064, "step": 32100 }, { "grad_norm": 0.12966203689575195, "learning_rate": 3.110591947868571e-05, "loss": 0.0066, "step": 32110 }, { "grad_norm": 0.1742822378873825, "learning_rate": 3.107530618299751e-05, "loss": 0.007, "step": 32120 }, { "grad_norm": 0.18057896196842194, "learning_rate": 3.104470116561333e-05, "loss": 0.0063, "step": 32130 }, { "grad_norm": 0.1756819784641266, "learning_rate": 3.1014104439920835e-05, "loss": 0.0078, "step": 32140 }, { "grad_norm": 1.0965816974639893, "learning_rate": 3.098351601930411e-05, "loss": 0.0082, "step": 32150 }, { "grad_norm": 0.4580520689487457, "learning_rate": 3.0952935917143533e-05, "loss": 0.0065, "step": 32160 }, { "grad_norm": 0.3220163583755493, "learning_rate": 3.092236414681588e-05, "loss": 0.0072, "step": 32170 }, { "grad_norm": 0.19264492392539978, "learning_rate": 3.089180072169431e-05, "loss": 0.0076, "step": 32180 }, { "grad_norm": 0.1401558220386505, "learning_rate": 3.08612456551483e-05, "loss": 0.0055, "step": 32190 }, { "grad_norm": 0.20286284387111664, "learning_rate": 3.083069896054363e-05, "loss": 0.0053, "step": 32200 }, { "grad_norm": 0.17381605505943298, "learning_rate": 3.080016065124253e-05, "loss": 0.0093, "step": 32210 }, { "grad_norm": 0.2124539315700531, "learning_rate": 3.076963074060344e-05, "loss": 0.0069, "step": 32220 }, { "grad_norm": 0.1655552238225937, "learning_rate": 3.07391092419812e-05, "loss": 0.0063, "step": 32230 }, { "grad_norm": 0.1616905927658081, "learning_rate": 3.070859616872695e-05, "loss": 0.0074, "step": 32240 }, { "grad_norm": 0.6986306309700012, "learning_rate": 3.067809153418816e-05, "loss": 0.0064, "step": 32250 }, { "grad_norm": 0.3186802864074707, "learning_rate": 3.0647595351708545e-05, "loss": 0.0087, "step": 32260 }, { "grad_norm": 0.23090390861034393, "learning_rate": 3.061710763462824e-05, "loss": 0.0073, "step": 32270 }, { "grad_norm": 10.194826126098633, "learning_rate": 3.0586628396283564e-05, "loss": 0.0088, "step": 32280 }, { "grad_norm": 0.293033242225647, "learning_rate": 3.0556157650007174e-05, "loss": 0.0081, "step": 32290 }, { "grad_norm": 0.1695576161146164, "learning_rate": 3.052569540912801e-05, "loss": 0.0073, "step": 32300 }, { "grad_norm": 0.4607749283313751, "learning_rate": 3.049524168697131e-05, "loss": 0.0073, "step": 32310 }, { "grad_norm": 0.19135650992393494, "learning_rate": 3.0464796496858538e-05, "loss": 0.0046, "step": 32320 }, { "grad_norm": 0.832122266292572, "learning_rate": 3.0434359852107473e-05, "loss": 0.0066, "step": 32330 }, { "grad_norm": 0.2556922733783722, "learning_rate": 3.0403931766032135e-05, "loss": 0.0067, "step": 32340 }, { "grad_norm": 0.20858637988567352, "learning_rate": 3.0373512251942814e-05, "loss": 0.0062, "step": 32350 }, { "grad_norm": 0.19415225088596344, "learning_rate": 3.0343101323146007e-05, "loss": 0.0071, "step": 32360 }, { "grad_norm": 0.18373402953147888, "learning_rate": 3.0312698992944523e-05, "loss": 0.0072, "step": 32370 }, { "grad_norm": 0.325003981590271, "learning_rate": 3.028230527463736e-05, "loss": 0.0061, "step": 32380 }, { "grad_norm": 0.16487467288970947, "learning_rate": 3.0251920181519746e-05, "loss": 0.0079, "step": 32390 }, { "grad_norm": 0.20693807303905487, "learning_rate": 3.0221543726883183e-05, "loss": 0.0079, "step": 32400 }, { "grad_norm": 0.42895206809043884, "learning_rate": 3.019117592401535e-05, "loss": 0.0083, "step": 32410 }, { "grad_norm": 1.8671127557754517, "learning_rate": 3.0160816786200137e-05, "loss": 0.0077, "step": 32420 }, { "grad_norm": 0.19045144319534302, "learning_rate": 3.01304663267177e-05, "loss": 0.0054, "step": 32430 }, { "grad_norm": 0.16347037255764008, "learning_rate": 3.0100124558844338e-05, "loss": 0.006, "step": 32440 }, { "grad_norm": 0.11777054518461227, "learning_rate": 3.0069791495852562e-05, "loss": 0.006, "step": 32450 }, { "grad_norm": 0.2492164522409439, "learning_rate": 3.0039467151011113e-05, "loss": 0.0065, "step": 32460 }, { "grad_norm": 0.12019333243370056, "learning_rate": 3.000915153758487e-05, "loss": 0.0061, "step": 32470 }, { "grad_norm": 0.6177948713302612, "learning_rate": 2.9978844668834917e-05, "loss": 0.0076, "step": 32480 }, { "grad_norm": 0.7708391547203064, "learning_rate": 2.994854655801849e-05, "loss": 0.0081, "step": 32490 }, { "grad_norm": 0.17368361353874207, "learning_rate": 2.991825721838905e-05, "loss": 0.0052, "step": 32500 }, { "grad_norm": 0.2279869168996811, "learning_rate": 2.9887976663196166e-05, "loss": 0.0065, "step": 32510 }, { "grad_norm": 0.2012099325656891, "learning_rate": 2.985770490568556e-05, "loss": 0.0074, "step": 32520 }, { "grad_norm": 0.18037770688533783, "learning_rate": 2.9827441959099157e-05, "loss": 0.0056, "step": 32530 }, { "grad_norm": 1.1061755418777466, "learning_rate": 2.979718783667499e-05, "loss": 0.0109, "step": 32540 }, { "grad_norm": 0.28860798478126526, "learning_rate": 2.9766942551647215e-05, "loss": 0.006, "step": 32550 }, { "grad_norm": 0.16260944306850433, "learning_rate": 2.9736706117246188e-05, "loss": 0.0066, "step": 32560 }, { "grad_norm": 0.21662898361682892, "learning_rate": 2.9706478546698324e-05, "loss": 0.0076, "step": 32570 }, { "grad_norm": 0.2769913375377655, "learning_rate": 2.9676259853226195e-05, "loss": 0.0078, "step": 32580 }, { "grad_norm": 0.5367249250411987, "learning_rate": 2.964605005004849e-05, "loss": 0.0107, "step": 32590 }, { "grad_norm": 0.7179877161979675, "learning_rate": 2.9615849150380016e-05, "loss": 0.0096, "step": 32600 }, { "grad_norm": 0.1592380255460739, "learning_rate": 2.9585657167431637e-05, "loss": 0.0059, "step": 32610 }, { "grad_norm": 0.1187354028224945, "learning_rate": 2.9555474114410388e-05, "loss": 0.0051, "step": 32620 }, { "grad_norm": 0.13290677964687347, "learning_rate": 2.9525300004519348e-05, "loss": 0.0071, "step": 32630 }, { "grad_norm": 0.1789836436510086, "learning_rate": 2.9495134850957696e-05, "loss": 0.0064, "step": 32640 }, { "grad_norm": 0.41480305790901184, "learning_rate": 2.9464978666920674e-05, "loss": 0.0055, "step": 32650 }, { "grad_norm": 0.2248174250125885, "learning_rate": 2.9434831465599662e-05, "loss": 0.0071, "step": 32660 }, { "grad_norm": 0.13519220054149628, "learning_rate": 2.9404693260182047e-05, "loss": 0.0062, "step": 32670 }, { "grad_norm": 0.2983284294605255, "learning_rate": 2.9374564063851295e-05, "loss": 0.0093, "step": 32680 }, { "grad_norm": 0.19322729110717773, "learning_rate": 2.9344443889786967e-05, "loss": 0.0062, "step": 32690 }, { "grad_norm": 0.23020316660404205, "learning_rate": 2.931433275116463e-05, "loss": 0.0063, "step": 32700 }, { "grad_norm": 0.09800046682357788, "learning_rate": 2.928423066115592e-05, "loss": 0.0057, "step": 32710 }, { "grad_norm": 0.1989578753709793, "learning_rate": 2.9254137632928537e-05, "loss": 0.0064, "step": 32720 }, { "grad_norm": 0.4704994857311249, "learning_rate": 2.9224053679646167e-05, "loss": 0.0067, "step": 32730 }, { "grad_norm": 0.1344849020242691, "learning_rate": 2.9193978814468547e-05, "loss": 0.0047, "step": 32740 }, { "grad_norm": 0.2123449295759201, "learning_rate": 2.9163913050551485e-05, "loss": 0.0066, "step": 32750 }, { "grad_norm": 0.15689769387245178, "learning_rate": 2.9133856401046735e-05, "loss": 0.0067, "step": 32760 }, { "grad_norm": 0.17453089356422424, "learning_rate": 2.910380887910209e-05, "loss": 0.0066, "step": 32770 }, { "grad_norm": 0.33537575602531433, "learning_rate": 2.9073770497861386e-05, "loss": 0.0069, "step": 32780 }, { "grad_norm": 0.2724795341491699, "learning_rate": 2.904374127046442e-05, "loss": 0.0053, "step": 32790 }, { "grad_norm": 0.24841183423995972, "learning_rate": 2.9013721210046973e-05, "loss": 0.007, "step": 32800 }, { "grad_norm": 0.22830012440681458, "learning_rate": 2.8983710329740882e-05, "loss": 0.0055, "step": 32810 }, { "grad_norm": 0.34967780113220215, "learning_rate": 2.8953708642673898e-05, "loss": 0.0064, "step": 32820 }, { "grad_norm": 0.15623393654823303, "learning_rate": 2.8923716161969782e-05, "loss": 0.0077, "step": 32830 }, { "grad_norm": 0.1793881505727768, "learning_rate": 2.8893732900748248e-05, "loss": 0.0067, "step": 32840 }, { "grad_norm": 0.13475681841373444, "learning_rate": 2.8863758872125036e-05, "loss": 0.007, "step": 32850 }, { "grad_norm": 0.30794158577919006, "learning_rate": 2.883379408921175e-05, "loss": 0.0055, "step": 32860 }, { "grad_norm": 0.14806793630123138, "learning_rate": 2.880383856511606e-05, "loss": 0.0049, "step": 32870 }, { "grad_norm": 0.14046335220336914, "learning_rate": 2.8773892312941476e-05, "loss": 0.0079, "step": 32880 }, { "grad_norm": 0.19969777762889862, "learning_rate": 2.8743955345787555e-05, "loss": 0.0059, "step": 32890 }, { "grad_norm": 0.19384445250034332, "learning_rate": 2.87140276767497e-05, "loss": 0.0064, "step": 32900 }, { "grad_norm": 0.26299384236335754, "learning_rate": 2.868410931891934e-05, "loss": 0.0051, "step": 32910 }, { "grad_norm": 0.1429213285446167, "learning_rate": 2.8654200285383748e-05, "loss": 0.0059, "step": 32920 }, { "grad_norm": 0.5205681920051575, "learning_rate": 2.8624300589226126e-05, "loss": 0.0064, "step": 32930 }, { "grad_norm": 0.5833966135978699, "learning_rate": 2.859441024352568e-05, "loss": 0.0082, "step": 32940 }, { "grad_norm": 0.16297376155853271, "learning_rate": 2.8564529261357427e-05, "loss": 0.0067, "step": 32950 }, { "grad_norm": 0.2065500169992447, "learning_rate": 2.8534657655792307e-05, "loss": 0.007, "step": 32960 }, { "grad_norm": 0.35927143692970276, "learning_rate": 2.8504795439897208e-05, "loss": 0.0064, "step": 32970 }, { "grad_norm": 0.1542193740606308, "learning_rate": 2.847494262673487e-05, "loss": 0.0066, "step": 32980 }, { "grad_norm": 0.14014539122581482, "learning_rate": 2.8445099229363925e-05, "loss": 0.0078, "step": 32990 }, { "grad_norm": 0.14325962960720062, "learning_rate": 2.841526526083886e-05, "loss": 0.0069, "step": 33000 }, { "grad_norm": 0.25320670008659363, "learning_rate": 2.838544073421011e-05, "loss": 0.0059, "step": 33010 }, { "grad_norm": 0.3575594425201416, "learning_rate": 2.8355625662523928e-05, "loss": 0.0076, "step": 33020 }, { "grad_norm": 0.138482004404068, "learning_rate": 2.8325820058822407e-05, "loss": 0.0073, "step": 33030 }, { "grad_norm": 0.09702517837285995, "learning_rate": 2.829602393614357e-05, "loss": 0.0071, "step": 33040 }, { "grad_norm": 0.2330627292394638, "learning_rate": 2.8266237307521236e-05, "loss": 0.0072, "step": 33050 }, { "grad_norm": 1.5130599737167358, "learning_rate": 2.8236460185985074e-05, "loss": 0.007, "step": 33060 }, { "grad_norm": 0.11803996562957764, "learning_rate": 2.8206692584560646e-05, "loss": 0.0062, "step": 33070 }, { "grad_norm": 0.14807970821857452, "learning_rate": 2.8176934516269283e-05, "loss": 0.0073, "step": 33080 }, { "grad_norm": 2.943632125854492, "learning_rate": 2.8147185994128167e-05, "loss": 0.0073, "step": 33090 }, { "grad_norm": 0.25615477561950684, "learning_rate": 2.8117447031150346e-05, "loss": 0.0056, "step": 33100 }, { "grad_norm": 0.18241718411445618, "learning_rate": 2.8087717640344618e-05, "loss": 0.0065, "step": 33110 }, { "grad_norm": 0.6880800127983093, "learning_rate": 2.805799783471566e-05, "loss": 0.0068, "step": 33120 }, { "grad_norm": 0.10398607701063156, "learning_rate": 2.8028287627263895e-05, "loss": 0.0056, "step": 33130 }, { "grad_norm": 0.2516706585884094, "learning_rate": 2.7998587030985608e-05, "loss": 0.0085, "step": 33140 }, { "grad_norm": 0.15675389766693115, "learning_rate": 2.7968896058872806e-05, "loss": 0.0069, "step": 33150 }, { "grad_norm": 0.2191963940858841, "learning_rate": 2.7939214723913388e-05, "loss": 0.007, "step": 33160 }, { "grad_norm": 0.1591121256351471, "learning_rate": 2.7909543039090935e-05, "loss": 0.0088, "step": 33170 }, { "grad_norm": 0.6904376149177551, "learning_rate": 2.787988101738487e-05, "loss": 0.0093, "step": 33180 }, { "grad_norm": 0.5697170495986938, "learning_rate": 2.7850228671770334e-05, "loss": 0.006, "step": 33190 }, { "grad_norm": 0.19783098995685577, "learning_rate": 2.782058601521832e-05, "loss": 0.0058, "step": 33200 }, { "grad_norm": 0.17208872735500336, "learning_rate": 2.77909530606955e-05, "loss": 0.0074, "step": 33210 }, { "grad_norm": 1.364640235900879, "learning_rate": 2.7761329821164335e-05, "loss": 0.0071, "step": 33220 }, { "grad_norm": 0.12838852405548096, "learning_rate": 2.7731716309583062e-05, "loss": 0.0076, "step": 33230 }, { "grad_norm": 0.18693789839744568, "learning_rate": 2.7702112538905624e-05, "loss": 0.0065, "step": 33240 }, { "grad_norm": 0.10243342071771622, "learning_rate": 2.7672518522081682e-05, "loss": 0.0055, "step": 33250 }, { "grad_norm": 0.27171963453292847, "learning_rate": 2.7642934272056726e-05, "loss": 0.0081, "step": 33260 }, { "grad_norm": 0.12171158194541931, "learning_rate": 2.7613359801771877e-05, "loss": 0.0086, "step": 33270 }, { "grad_norm": 0.23303444683551788, "learning_rate": 2.7583795124163996e-05, "loss": 0.0077, "step": 33280 }, { "grad_norm": 0.2434859722852707, "learning_rate": 2.755424025216572e-05, "loss": 0.007, "step": 33290 }, { "grad_norm": 0.15332968533039093, "learning_rate": 2.7524695198705334e-05, "loss": 0.0064, "step": 33300 }, { "grad_norm": 0.6266391277313232, "learning_rate": 2.749515997670683e-05, "loss": 0.0072, "step": 33310 }, { "grad_norm": 0.6977930665016174, "learning_rate": 2.746563459908995e-05, "loss": 0.0081, "step": 33320 }, { "grad_norm": 0.41155606508255005, "learning_rate": 2.7436119078770085e-05, "loss": 0.0096, "step": 33330 }, { "grad_norm": 0.12634925544261932, "learning_rate": 2.74066134286583e-05, "loss": 0.0052, "step": 33340 }, { "grad_norm": 0.13249154388904572, "learning_rate": 2.7377117661661412e-05, "loss": 0.0056, "step": 33350 }, { "grad_norm": 0.4639136493206024, "learning_rate": 2.734763179068183e-05, "loss": 0.0073, "step": 33360 }, { "grad_norm": 0.1256348341703415, "learning_rate": 2.7318155828617718e-05, "loss": 0.0057, "step": 33370 }, { "grad_norm": 0.21838311851024628, "learning_rate": 2.7288689788362832e-05, "loss": 0.0075, "step": 33380 }, { "grad_norm": 0.1212487667798996, "learning_rate": 2.725923368280664e-05, "loss": 0.0063, "step": 33390 }, { "grad_norm": 0.1554040163755417, "learning_rate": 2.7229787524834237e-05, "loss": 0.0059, "step": 33400 }, { "grad_norm": 0.18950842320919037, "learning_rate": 2.7200351327326345e-05, "loss": 0.0063, "step": 33410 }, { "grad_norm": 0.15435421466827393, "learning_rate": 2.7170925103159407e-05, "loss": 0.0055, "step": 33420 }, { "grad_norm": 0.20120497047901154, "learning_rate": 2.714150886520542e-05, "loss": 0.0067, "step": 33430 }, { "grad_norm": 1.2087512016296387, "learning_rate": 2.7112102626332038e-05, "loss": 0.0084, "step": 33440 }, { "grad_norm": 0.1275172233581543, "learning_rate": 2.7082706399402574e-05, "loss": 0.0061, "step": 33450 }, { "grad_norm": 0.3134233355522156, "learning_rate": 2.7053320197275934e-05, "loss": 0.0072, "step": 33460 }, { "grad_norm": 0.27617284655570984, "learning_rate": 2.7023944032806614e-05, "loss": 0.0068, "step": 33470 }, { "grad_norm": 0.1950412541627884, "learning_rate": 2.699457791884478e-05, "loss": 0.0074, "step": 33480 }, { "grad_norm": 0.1589265912771225, "learning_rate": 2.6965221868236155e-05, "loss": 0.0059, "step": 33490 }, { "grad_norm": 0.2134278118610382, "learning_rate": 2.6935875893822077e-05, "loss": 0.0066, "step": 33500 }, { "grad_norm": 2.692404270172119, "learning_rate": 2.690654000843944e-05, "loss": 0.0097, "step": 33510 }, { "grad_norm": 0.21636702120304108, "learning_rate": 2.687721422492081e-05, "loss": 0.0061, "step": 33520 }, { "grad_norm": 0.2612169682979584, "learning_rate": 2.6847898556094252e-05, "loss": 0.0073, "step": 33530 }, { "grad_norm": 0.1601058691740036, "learning_rate": 2.6818593014783417e-05, "loss": 0.0057, "step": 33540 }, { "grad_norm": 0.1864699125289917, "learning_rate": 2.6789297613807585e-05, "loss": 0.0072, "step": 33550 }, { "grad_norm": 0.1395406275987625, "learning_rate": 2.676001236598154e-05, "loss": 0.0052, "step": 33560 }, { "grad_norm": 0.490374356508255, "learning_rate": 2.6730737284115626e-05, "loss": 0.0062, "step": 33570 }, { "grad_norm": 0.14290760457515717, "learning_rate": 2.670147238101579e-05, "loss": 0.0061, "step": 33580 }, { "grad_norm": 0.7408372163772583, "learning_rate": 2.6672217669483473e-05, "loss": 0.0072, "step": 33590 }, { "grad_norm": 0.13419277966022491, "learning_rate": 2.6642973162315706e-05, "loss": 0.0053, "step": 33600 }, { "grad_norm": 0.11239402741193771, "learning_rate": 2.6613738872304988e-05, "loss": 0.0085, "step": 33610 }, { "grad_norm": 0.631540834903717, "learning_rate": 2.6584514812239437e-05, "loss": 0.0067, "step": 33620 }, { "grad_norm": 0.2433711439371109, "learning_rate": 2.655530099490261e-05, "loss": 0.0064, "step": 33630 }, { "grad_norm": 0.12421354651451111, "learning_rate": 2.6526097433073653e-05, "loss": 0.0054, "step": 33640 }, { "grad_norm": 0.22765924036502838, "learning_rate": 2.649690413952719e-05, "loss": 0.0076, "step": 33650 }, { "grad_norm": 1.784278392791748, "learning_rate": 2.646772112703334e-05, "loss": 0.0051, "step": 33660 }, { "grad_norm": 0.11195980757474899, "learning_rate": 2.6438548408357768e-05, "loss": 0.0064, "step": 33670 }, { "grad_norm": 0.14400620758533478, "learning_rate": 2.6409385996261603e-05, "loss": 0.005, "step": 33680 }, { "grad_norm": 0.2550332844257355, "learning_rate": 2.638023390350148e-05, "loss": 0.0078, "step": 33690 }, { "grad_norm": 0.12824827432632446, "learning_rate": 2.635109214282948e-05, "loss": 0.0062, "step": 33700 }, { "grad_norm": 0.11904856562614441, "learning_rate": 2.6321960726993254e-05, "loss": 0.0064, "step": 33710 }, { "grad_norm": 0.301852822303772, "learning_rate": 2.6292839668735846e-05, "loss": 0.0064, "step": 33720 }, { "grad_norm": 0.25642088055610657, "learning_rate": 2.6263728980795778e-05, "loss": 0.0064, "step": 33730 }, { "grad_norm": 0.8940361738204956, "learning_rate": 2.6234628675907092e-05, "loss": 0.0057, "step": 33740 }, { "grad_norm": 0.17108193039894104, "learning_rate": 2.620553876679923e-05, "loss": 0.0062, "step": 33750 }, { "grad_norm": 0.11950923502445221, "learning_rate": 2.617645926619709e-05, "loss": 0.0067, "step": 33760 }, { "grad_norm": 0.4221552610397339, "learning_rate": 2.6147390186821064e-05, "loss": 0.0049, "step": 33770 }, { "grad_norm": 0.4391575753688812, "learning_rate": 2.6118331541386955e-05, "loss": 0.0062, "step": 33780 }, { "grad_norm": 0.1388387531042099, "learning_rate": 2.6089283342605953e-05, "loss": 0.0059, "step": 33790 }, { "grad_norm": 0.14629685878753662, "learning_rate": 2.6060245603184797e-05, "loss": 0.0069, "step": 33800 }, { "grad_norm": 0.5174705982208252, "learning_rate": 2.6031218335825536e-05, "loss": 0.0059, "step": 33810 }, { "grad_norm": 0.10365337878465652, "learning_rate": 2.600220155322568e-05, "loss": 0.0048, "step": 33820 }, { "grad_norm": 0.18213728070259094, "learning_rate": 2.5973195268078186e-05, "loss": 0.0063, "step": 33830 }, { "grad_norm": 0.1780465841293335, "learning_rate": 2.5944199493071352e-05, "loss": 0.0049, "step": 33840 }, { "grad_norm": 0.2329421490430832, "learning_rate": 2.591521424088895e-05, "loss": 0.0063, "step": 33850 }, { "grad_norm": 0.1523551046848297, "learning_rate": 2.5886239524210086e-05, "loss": 0.0055, "step": 33860 }, { "grad_norm": 0.5089626312255859, "learning_rate": 2.5857275355709317e-05, "loss": 0.0062, "step": 33870 }, { "grad_norm": 0.13242311775684357, "learning_rate": 2.5828321748056533e-05, "loss": 0.0055, "step": 33880 }, { "grad_norm": 0.16388484835624695, "learning_rate": 2.5799378713917012e-05, "loss": 0.0063, "step": 33890 }, { "grad_norm": 0.265689492225647, "learning_rate": 2.5770446265951455e-05, "loss": 0.0077, "step": 33900 }, { "grad_norm": 0.1060100570321083, "learning_rate": 2.5741524416815877e-05, "loss": 0.0054, "step": 33910 }, { "grad_norm": 0.126615509390831, "learning_rate": 2.571261317916166e-05, "loss": 0.0059, "step": 33920 }, { "grad_norm": 0.20453868806362152, "learning_rate": 2.5683712565635597e-05, "loss": 0.008, "step": 33930 }, { "grad_norm": 0.1501767337322235, "learning_rate": 2.565482258887979e-05, "loss": 0.0072, "step": 33940 }, { "grad_norm": 0.23420435190200806, "learning_rate": 2.5625943261531664e-05, "loss": 0.0062, "step": 33950 }, { "grad_norm": 0.1230296716094017, "learning_rate": 2.5597074596224067e-05, "loss": 0.0062, "step": 33960 }, { "grad_norm": 0.2572794258594513, "learning_rate": 2.5568216605585105e-05, "loss": 0.0079, "step": 33970 }, { "grad_norm": 0.487125426530838, "learning_rate": 2.5539369302238235e-05, "loss": 0.0078, "step": 33980 }, { "grad_norm": 0.24699532985687256, "learning_rate": 2.5510532698802276e-05, "loss": 0.0065, "step": 33990 }, { "grad_norm": 0.5142378211021423, "learning_rate": 2.5481706807891324e-05, "loss": 0.0068, "step": 34000 }, { "grad_norm": 0.13236355781555176, "learning_rate": 2.5452891642114783e-05, "loss": 0.0046, "step": 34010 }, { "grad_norm": 0.23529955744743347, "learning_rate": 2.542408721407742e-05, "loss": 0.0057, "step": 34020 }, { "grad_norm": 0.18956618010997772, "learning_rate": 2.539529353637926e-05, "loss": 0.0087, "step": 34030 }, { "grad_norm": 0.19079631567001343, "learning_rate": 2.5366510621615626e-05, "loss": 0.0062, "step": 34040 }, { "grad_norm": 0.10523567348718643, "learning_rate": 2.5337738482377132e-05, "loss": 0.0056, "step": 34050 }, { "grad_norm": 0.15172813832759857, "learning_rate": 2.5308977131249722e-05, "loss": 0.0071, "step": 34060 }, { "grad_norm": 0.23684127628803253, "learning_rate": 2.5280226580814554e-05, "loss": 0.0061, "step": 34070 }, { "grad_norm": 0.3217999041080475, "learning_rate": 2.525148684364813e-05, "loss": 0.0064, "step": 34080 }, { "grad_norm": 0.14769740402698517, "learning_rate": 2.522275793232215e-05, "loss": 0.0069, "step": 34090 }, { "grad_norm": 3.1555840969085693, "learning_rate": 2.519403985940365e-05, "loss": 0.0063, "step": 34100 }, { "grad_norm": 0.2332954853773117, "learning_rate": 2.5165332637454863e-05, "loss": 0.0078, "step": 34110 }, { "grad_norm": 0.13881081342697144, "learning_rate": 2.513663627903334e-05, "loss": 0.008, "step": 34120 }, { "grad_norm": 0.19286809861660004, "learning_rate": 2.5107950796691815e-05, "loss": 0.0071, "step": 34130 }, { "grad_norm": 0.14115333557128906, "learning_rate": 2.5079276202978276e-05, "loss": 0.0054, "step": 34140 }, { "grad_norm": 0.17587296664714813, "learning_rate": 2.505061251043601e-05, "loss": 0.0066, "step": 34150 }, { "grad_norm": 0.19561119377613068, "learning_rate": 2.502195973160346e-05, "loss": 0.006, "step": 34160 }, { "grad_norm": 0.12063077837228775, "learning_rate": 2.4993317879014316e-05, "loss": 0.0059, "step": 34170 }, { "grad_norm": 0.13008354604244232, "learning_rate": 2.4964686965197526e-05, "loss": 0.0065, "step": 34180 }, { "grad_norm": 0.8554925322532654, "learning_rate": 2.4936067002677217e-05, "loss": 0.0067, "step": 34190 }, { "grad_norm": 0.21614117920398712, "learning_rate": 2.4907458003972724e-05, "loss": 0.0059, "step": 34200 }, { "grad_norm": 0.3176872730255127, "learning_rate": 2.4878859981598572e-05, "loss": 0.0052, "step": 34210 }, { "grad_norm": 0.15751196444034576, "learning_rate": 2.4850272948064555e-05, "loss": 0.0056, "step": 34220 }, { "grad_norm": 0.2039312869310379, "learning_rate": 2.4821696915875596e-05, "loss": 0.006, "step": 34230 }, { "grad_norm": 0.20340703427791595, "learning_rate": 2.4793131897531798e-05, "loss": 0.0068, "step": 34240 }, { "grad_norm": 1.1748019456863403, "learning_rate": 2.47645779055285e-05, "loss": 0.0062, "step": 34250 }, { "grad_norm": 0.20338357985019684, "learning_rate": 2.4736034952356186e-05, "loss": 0.0062, "step": 34260 }, { "grad_norm": 0.20341533422470093, "learning_rate": 2.470750305050048e-05, "loss": 0.0065, "step": 34270 }, { "grad_norm": 0.26580435037612915, "learning_rate": 2.4678982212442253e-05, "loss": 0.008, "step": 34280 }, { "grad_norm": 0.38274139165878296, "learning_rate": 2.4650472450657463e-05, "loss": 0.0071, "step": 34290 }, { "grad_norm": 0.2559993267059326, "learning_rate": 2.4621973777617237e-05, "loss": 0.0054, "step": 34300 }, { "grad_norm": 0.2806370258331299, "learning_rate": 2.459348620578789e-05, "loss": 0.0065, "step": 34310 }, { "grad_norm": 0.20776067674160004, "learning_rate": 2.456500974763083e-05, "loss": 0.0072, "step": 34320 }, { "grad_norm": 0.36363160610198975, "learning_rate": 2.4536544415602653e-05, "loss": 0.0071, "step": 34330 }, { "grad_norm": 0.16578352451324463, "learning_rate": 2.4508090222155027e-05, "loss": 0.0057, "step": 34340 }, { "grad_norm": 0.13509055972099304, "learning_rate": 2.4479647179734822e-05, "loss": 0.0075, "step": 34350 }, { "grad_norm": 0.2111089825630188, "learning_rate": 2.4451215300783953e-05, "loss": 0.0073, "step": 34360 }, { "grad_norm": 0.22235903143882751, "learning_rate": 2.442279459773953e-05, "loss": 0.006, "step": 34370 }, { "grad_norm": 0.28362390398979187, "learning_rate": 2.4394385083033712e-05, "loss": 0.0062, "step": 34380 }, { "grad_norm": 0.22769224643707275, "learning_rate": 2.4365986769093792e-05, "loss": 0.0104, "step": 34390 }, { "grad_norm": 0.1265355944633484, "learning_rate": 2.4337599668342127e-05, "loss": 0.0051, "step": 34400 }, { "grad_norm": 0.33340540528297424, "learning_rate": 2.4309223793196245e-05, "loss": 0.0059, "step": 34410 }, { "grad_norm": 0.13801933825016022, "learning_rate": 2.4280859156068698e-05, "loss": 0.0061, "step": 34420 }, { "grad_norm": 0.15352767705917358, "learning_rate": 2.4252505769367123e-05, "loss": 0.0062, "step": 34430 }, { "grad_norm": 0.3096056282520294, "learning_rate": 2.422416364549429e-05, "loss": 0.0052, "step": 34440 }, { "grad_norm": 0.21131259202957153, "learning_rate": 2.4195832796847994e-05, "loss": 0.0063, "step": 34450 }, { "grad_norm": 0.20806622505187988, "learning_rate": 2.416751323582109e-05, "loss": 0.0055, "step": 34460 }, { "grad_norm": 0.15463927388191223, "learning_rate": 2.4139204974801544e-05, "loss": 0.0045, "step": 34470 }, { "grad_norm": 0.21422675251960754, "learning_rate": 2.4110908026172347e-05, "loss": 0.0067, "step": 34480 }, { "grad_norm": 0.2363048791885376, "learning_rate": 2.4082622402311518e-05, "loss": 0.0062, "step": 34490 }, { "grad_norm": 0.287845641374588, "learning_rate": 2.4054348115592183e-05, "loss": 0.0075, "step": 34500 }, { "grad_norm": 0.10283534228801727, "learning_rate": 2.4026085178382475e-05, "loss": 0.0059, "step": 34510 }, { "grad_norm": 0.2335999459028244, "learning_rate": 2.3997833603045523e-05, "loss": 0.0059, "step": 34520 }, { "grad_norm": 0.14227049052715302, "learning_rate": 2.396959340193958e-05, "loss": 0.0074, "step": 34530 }, { "grad_norm": 0.12491528689861298, "learning_rate": 2.3941364587417836e-05, "loss": 0.005, "step": 34540 }, { "grad_norm": 0.17995664477348328, "learning_rate": 2.3913147171828526e-05, "loss": 0.006, "step": 34550 }, { "grad_norm": 0.14066995680332184, "learning_rate": 2.388494116751494e-05, "loss": 0.0058, "step": 34560 }, { "grad_norm": 0.16874462366104126, "learning_rate": 2.3856746586815302e-05, "loss": 0.0075, "step": 34570 }, { "grad_norm": 0.15961194038391113, "learning_rate": 2.3828563442062924e-05, "loss": 0.0077, "step": 34580 }, { "grad_norm": 0.12893077731132507, "learning_rate": 2.3800391745586026e-05, "loss": 0.0064, "step": 34590 }, { "grad_norm": 0.23426587879657745, "learning_rate": 2.377223150970791e-05, "loss": 0.0054, "step": 34600 }, { "grad_norm": 0.1965402066707611, "learning_rate": 2.3744082746746793e-05, "loss": 0.0062, "step": 34610 }, { "grad_norm": 0.17264825105667114, "learning_rate": 2.3715945469015893e-05, "loss": 0.0069, "step": 34620 }, { "grad_norm": 0.13059303164482117, "learning_rate": 2.368781968882343e-05, "loss": 0.0061, "step": 34630 }, { "grad_norm": 0.17533209919929504, "learning_rate": 2.365970541847258e-05, "loss": 0.0069, "step": 34640 }, { "grad_norm": 0.6459860801696777, "learning_rate": 2.3631602670261445e-05, "loss": 0.0091, "step": 34650 }, { "grad_norm": 0.34529557824134827, "learning_rate": 2.3603511456483164e-05, "loss": 0.0057, "step": 34660 }, { "grad_norm": 0.18931038677692413, "learning_rate": 2.357543178942577e-05, "loss": 0.0063, "step": 34670 }, { "grad_norm": 0.14847764372825623, "learning_rate": 2.3547363681372242e-05, "loss": 0.0066, "step": 34680 }, { "grad_norm": 0.08371642231941223, "learning_rate": 2.3519307144600565e-05, "loss": 0.0052, "step": 34690 }, { "grad_norm": 0.23828674852848053, "learning_rate": 2.3491262191383607e-05, "loss": 0.006, "step": 34700 }, { "grad_norm": 0.18188616633415222, "learning_rate": 2.3463228833989154e-05, "loss": 0.0061, "step": 34710 }, { "grad_norm": 0.5549744963645935, "learning_rate": 2.343520708467999e-05, "loss": 0.0066, "step": 34720 }, { "grad_norm": 0.6190387606620789, "learning_rate": 2.3407196955713767e-05, "loss": 0.0068, "step": 34730 }, { "grad_norm": 0.1699715405702591, "learning_rate": 2.337919845934306e-05, "loss": 0.0071, "step": 34740 }, { "grad_norm": 0.8596966862678528, "learning_rate": 2.335121160781535e-05, "loss": 0.0067, "step": 34750 }, { "grad_norm": 0.4133213758468628, "learning_rate": 2.3323236413373063e-05, "loss": 0.0055, "step": 34760 }, { "grad_norm": 1.117254614830017, "learning_rate": 2.3295272888253494e-05, "loss": 0.0063, "step": 34770 }, { "grad_norm": 0.11289423704147339, "learning_rate": 2.3267321044688815e-05, "loss": 0.0047, "step": 34780 }, { "grad_norm": 0.17690306901931763, "learning_rate": 2.323938089490615e-05, "loss": 0.0048, "step": 34790 }, { "grad_norm": 0.09001410007476807, "learning_rate": 2.321145245112743e-05, "loss": 0.0055, "step": 34800 }, { "grad_norm": 0.18201717734336853, "learning_rate": 2.318353572556953e-05, "loss": 0.0059, "step": 34810 }, { "grad_norm": 0.8493418097496033, "learning_rate": 2.3155630730444183e-05, "loss": 0.0051, "step": 34820 }, { "grad_norm": 0.2326943576335907, "learning_rate": 2.3127737477957968e-05, "loss": 0.0068, "step": 34830 }, { "grad_norm": 0.14589597284793854, "learning_rate": 2.3099855980312324e-05, "loss": 0.0057, "step": 34840 }, { "grad_norm": 0.43321725726127625, "learning_rate": 2.3071986249703602e-05, "loss": 0.0064, "step": 34850 }, { "grad_norm": 0.09480030089616776, "learning_rate": 2.3044128298322953e-05, "loss": 0.0057, "step": 34860 }, { "grad_norm": 0.25182679295539856, "learning_rate": 2.301628213835637e-05, "loss": 0.0067, "step": 34870 }, { "grad_norm": 0.3782607316970825, "learning_rate": 2.2988447781984744e-05, "loss": 0.006, "step": 34880 }, { "grad_norm": 0.22266808152198792, "learning_rate": 2.2960625241383753e-05, "loss": 0.0056, "step": 34890 }, { "grad_norm": 0.4333478808403015, "learning_rate": 2.2932814528723922e-05, "loss": 0.0056, "step": 34900 }, { "grad_norm": 0.18290290236473083, "learning_rate": 2.2905015656170582e-05, "loss": 0.0059, "step": 34910 }, { "grad_norm": 0.24890266358852386, "learning_rate": 2.287722863588394e-05, "loss": 0.006, "step": 34920 }, { "grad_norm": 0.2653650939464569, "learning_rate": 2.2849453480018962e-05, "loss": 0.0069, "step": 34930 }, { "grad_norm": 0.12804684042930603, "learning_rate": 2.282169020072543e-05, "loss": 0.008, "step": 34940 }, { "grad_norm": 0.4567245543003082, "learning_rate": 2.2793938810147986e-05, "loss": 0.0056, "step": 34950 }, { "grad_norm": 1.7759138345718384, "learning_rate": 2.2766199320426007e-05, "loss": 0.0049, "step": 34960 }, { "grad_norm": 0.12130469083786011, "learning_rate": 2.2738471743693674e-05, "loss": 0.0053, "step": 34970 }, { "grad_norm": 0.25592222809791565, "learning_rate": 2.2710756092080006e-05, "loss": 0.0071, "step": 34980 }, { "grad_norm": 0.13732939958572388, "learning_rate": 2.2683052377708747e-05, "loss": 0.0056, "step": 34990 }, { "grad_norm": 0.18807245790958405, "learning_rate": 2.265536061269843e-05, "loss": 0.0057, "step": 35000 }, { "grad_norm": 0.22031469643115997, "learning_rate": 2.262768080916241e-05, "loss": 0.0105, "step": 35010 }, { "grad_norm": 0.3610115647315979, "learning_rate": 2.260001297920874e-05, "loss": 0.0067, "step": 35020 }, { "grad_norm": 0.4899102449417114, "learning_rate": 2.2572357134940297e-05, "loss": 0.0075, "step": 35030 }, { "grad_norm": 0.21940700709819794, "learning_rate": 2.2544713288454655e-05, "loss": 0.0061, "step": 35040 }, { "grad_norm": 0.14752313494682312, "learning_rate": 2.2517081451844213e-05, "loss": 0.0057, "step": 35050 }, { "grad_norm": 0.1681206375360489, "learning_rate": 2.248946163719605e-05, "loss": 0.005, "step": 35060 }, { "grad_norm": 0.2319275289773941, "learning_rate": 2.2461853856591996e-05, "loss": 0.0061, "step": 35070 }, { "grad_norm": 0.15879394114017487, "learning_rate": 2.2434258122108664e-05, "loss": 0.0079, "step": 35080 }, { "grad_norm": 0.21485617756843567, "learning_rate": 2.2406674445817356e-05, "loss": 0.0066, "step": 35090 }, { "grad_norm": 0.2508254051208496, "learning_rate": 2.2379102839784083e-05, "loss": 0.0055, "step": 35100 }, { "grad_norm": 0.1779729723930359, "learning_rate": 2.2351543316069645e-05, "loss": 0.0091, "step": 35110 }, { "grad_norm": 0.28217387199401855, "learning_rate": 2.2323995886729493e-05, "loss": 0.0066, "step": 35120 }, { "grad_norm": 0.18941770493984222, "learning_rate": 2.2296460563813786e-05, "loss": 0.0054, "step": 35130 }, { "grad_norm": 0.22841094434261322, "learning_rate": 2.2268937359367457e-05, "loss": 0.0056, "step": 35140 }, { "grad_norm": 0.4627249240875244, "learning_rate": 2.224142628543006e-05, "loss": 0.0058, "step": 35150 }, { "grad_norm": 0.22763164341449738, "learning_rate": 2.2213927354035863e-05, "loss": 0.0049, "step": 35160 }, { "grad_norm": 0.37665677070617676, "learning_rate": 2.218644057721387e-05, "loss": 0.0053, "step": 35170 }, { "grad_norm": 0.14022515714168549, "learning_rate": 2.2158965966987703e-05, "loss": 0.0077, "step": 35180 }, { "grad_norm": 0.17356552183628082, "learning_rate": 2.2131503535375687e-05, "loss": 0.0087, "step": 35190 }, { "grad_norm": 3.7775678634643555, "learning_rate": 2.2104053294390846e-05, "loss": 0.0077, "step": 35200 }, { "grad_norm": 0.24075718224048615, "learning_rate": 2.2076615256040834e-05, "loss": 0.0051, "step": 35210 }, { "grad_norm": 0.15221743285655975, "learning_rate": 2.2049189432327966e-05, "loss": 0.0052, "step": 35220 }, { "grad_norm": 0.1517089605331421, "learning_rate": 2.202177583524926e-05, "loss": 0.0059, "step": 35230 }, { "grad_norm": 0.1828600913286209, "learning_rate": 2.1994374476796343e-05, "loss": 0.0063, "step": 35240 }, { "grad_norm": 0.5394972562789917, "learning_rate": 2.1966985368955477e-05, "loss": 0.0057, "step": 35250 }, { "grad_norm": 0.092378169298172, "learning_rate": 2.193960852370762e-05, "loss": 0.0073, "step": 35260 }, { "grad_norm": 0.2314060628414154, "learning_rate": 2.1912243953028317e-05, "loss": 0.005, "step": 35270 }, { "grad_norm": 0.15656283497810364, "learning_rate": 2.1884891668887776e-05, "loss": 0.0058, "step": 35280 }, { "grad_norm": 0.4165196418762207, "learning_rate": 2.1857551683250787e-05, "loss": 0.0058, "step": 35290 }, { "grad_norm": 0.16834627091884613, "learning_rate": 2.1830224008076827e-05, "loss": 0.0064, "step": 35300 }, { "grad_norm": 0.18002986907958984, "learning_rate": 2.1802908655319924e-05, "loss": 0.0057, "step": 35310 }, { "grad_norm": 0.1767350137233734, "learning_rate": 2.1775605636928738e-05, "loss": 0.0071, "step": 35320 }, { "grad_norm": 0.4783501923084259, "learning_rate": 2.1748314964846555e-05, "loss": 0.007, "step": 35330 }, { "grad_norm": 0.15733805298805237, "learning_rate": 2.172103665101124e-05, "loss": 0.0057, "step": 35340 }, { "grad_norm": 0.1447518914937973, "learning_rate": 2.1693770707355226e-05, "loss": 0.007, "step": 35350 }, { "grad_norm": 0.10382869094610214, "learning_rate": 2.1666517145805605e-05, "loss": 0.0072, "step": 35360 }, { "grad_norm": 0.1494627147912979, "learning_rate": 2.1639275978283985e-05, "loss": 0.009, "step": 35370 }, { "grad_norm": 2.26249623298645, "learning_rate": 2.1612047216706567e-05, "loss": 0.0052, "step": 35380 }, { "grad_norm": 0.229799285531044, "learning_rate": 2.158483087298417e-05, "loss": 0.0078, "step": 35390 }, { "grad_norm": 0.240022674202919, "learning_rate": 2.155762695902213e-05, "loss": 0.0062, "step": 35400 }, { "grad_norm": 0.2654995918273926, "learning_rate": 2.1530435486720363e-05, "loss": 0.0062, "step": 35410 }, { "grad_norm": 0.15655027329921722, "learning_rate": 2.1503256467973327e-05, "loss": 0.0078, "step": 35420 }, { "grad_norm": 6.332845211029053, "learning_rate": 2.147608991467008e-05, "loss": 0.0075, "step": 35430 }, { "grad_norm": 0.1684940755367279, "learning_rate": 2.144893583869418e-05, "loss": 0.0089, "step": 35440 }, { "grad_norm": 0.28248342871665955, "learning_rate": 2.142179425192372e-05, "loss": 0.0059, "step": 35450 }, { "grad_norm": 0.15729756653308868, "learning_rate": 2.1394665166231393e-05, "loss": 0.007, "step": 35460 }, { "grad_norm": 0.1577041745185852, "learning_rate": 2.1367548593484364e-05, "loss": 0.0069, "step": 35470 }, { "grad_norm": 0.1539188176393509, "learning_rate": 2.134044454554433e-05, "loss": 0.0062, "step": 35480 }, { "grad_norm": 0.16292966902256012, "learning_rate": 2.1313353034267548e-05, "loss": 0.0042, "step": 35490 }, { "grad_norm": 0.11129474639892578, "learning_rate": 2.1286274071504735e-05, "loss": 0.0054, "step": 35500 }, { "grad_norm": 0.36508044600486755, "learning_rate": 2.125920766910119e-05, "loss": 0.0071, "step": 35510 }, { "grad_norm": 2.9318747520446777, "learning_rate": 2.1232153838896628e-05, "loss": 0.0066, "step": 35520 }, { "grad_norm": 0.27271538972854614, "learning_rate": 2.1205112592725366e-05, "loss": 0.0051, "step": 35530 }, { "grad_norm": 0.28427600860595703, "learning_rate": 2.1178083942416114e-05, "loss": 0.0061, "step": 35540 }, { "grad_norm": 0.16018015146255493, "learning_rate": 2.1151067899792164e-05, "loss": 0.0052, "step": 35550 }, { "grad_norm": 1.5866864919662476, "learning_rate": 2.1124064476671235e-05, "loss": 0.0058, "step": 35560 }, { "grad_norm": 3.375422954559326, "learning_rate": 2.1097073684865516e-05, "loss": 0.0056, "step": 35570 }, { "grad_norm": 0.13080130517482758, "learning_rate": 2.107009553618174e-05, "loss": 0.0054, "step": 35580 }, { "grad_norm": 0.1690950244665146, "learning_rate": 2.1043130042421044e-05, "loss": 0.0061, "step": 35590 }, { "grad_norm": 0.1943826824426651, "learning_rate": 2.1016177215379053e-05, "loss": 0.0073, "step": 35600 }, { "grad_norm": 0.171425923705101, "learning_rate": 2.0989237066845824e-05, "loss": 0.006, "step": 35610 }, { "grad_norm": 0.2531469166278839, "learning_rate": 2.096230960860594e-05, "loss": 0.0066, "step": 35620 }, { "grad_norm": 0.19014772772789001, "learning_rate": 2.0935394852438366e-05, "loss": 0.0084, "step": 35630 }, { "grad_norm": 0.18705864250659943, "learning_rate": 2.090849281011651e-05, "loss": 0.0078, "step": 35640 }, { "grad_norm": 0.16633771359920502, "learning_rate": 2.088160349340827e-05, "loss": 0.0047, "step": 35650 }, { "grad_norm": 0.8225829005241394, "learning_rate": 2.085472691407594e-05, "loss": 0.007, "step": 35660 }, { "grad_norm": 1.2561918497085571, "learning_rate": 2.082786308387622e-05, "loss": 0.0055, "step": 35670 }, { "grad_norm": 0.09755635261535645, "learning_rate": 2.0801012014560306e-05, "loss": 0.0062, "step": 35680 }, { "grad_norm": 0.5040331482887268, "learning_rate": 2.0774173717873746e-05, "loss": 0.0073, "step": 35690 }, { "grad_norm": 0.1534210443496704, "learning_rate": 2.074734820555651e-05, "loss": 0.0057, "step": 35700 }, { "grad_norm": 0.2259940207004547, "learning_rate": 2.0720535489343014e-05, "loss": 0.0059, "step": 35710 }, { "grad_norm": 0.1543004810810089, "learning_rate": 2.0693735580962044e-05, "loss": 0.0084, "step": 35720 }, { "grad_norm": 0.8160878419876099, "learning_rate": 2.0666948492136766e-05, "loss": 0.0066, "step": 35730 }, { "grad_norm": 0.2947297692298889, "learning_rate": 2.0640174234584797e-05, "loss": 0.0044, "step": 35740 }, { "grad_norm": 0.2645271122455597, "learning_rate": 2.061341282001808e-05, "loss": 0.0055, "step": 35750 }, { "grad_norm": 0.19945281744003296, "learning_rate": 2.0586664260142997e-05, "loss": 0.0056, "step": 35760 }, { "grad_norm": 1.823374629020691, "learning_rate": 2.0559928566660237e-05, "loss": 0.0079, "step": 35770 }, { "grad_norm": 0.12135347723960876, "learning_rate": 2.0533205751264945e-05, "loss": 0.0059, "step": 35780 }, { "grad_norm": 0.13918280601501465, "learning_rate": 2.050649582564656e-05, "loss": 0.0058, "step": 35790 }, { "grad_norm": 0.17311209440231323, "learning_rate": 2.04797988014889e-05, "loss": 0.0069, "step": 35800 }, { "grad_norm": 0.16787005960941315, "learning_rate": 2.0453114690470194e-05, "loss": 0.007, "step": 35810 }, { "grad_norm": 0.25197210907936096, "learning_rate": 2.0426443504262944e-05, "loss": 0.0057, "step": 35820 }, { "grad_norm": 0.11298413574695587, "learning_rate": 2.0399785254534027e-05, "loss": 0.0063, "step": 35830 }, { "grad_norm": 0.25732752680778503, "learning_rate": 2.0373139952944697e-05, "loss": 0.0056, "step": 35840 }, { "grad_norm": 0.1311044842004776, "learning_rate": 2.03465076111505e-05, "loss": 0.0064, "step": 35850 }, { "grad_norm": 0.13590306043624878, "learning_rate": 2.0319888240801305e-05, "loss": 0.0045, "step": 35860 }, { "grad_norm": 0.147650808095932, "learning_rate": 2.0293281853541375e-05, "loss": 0.0085, "step": 35870 }, { "grad_norm": 0.3189482092857361, "learning_rate": 2.0266688461009216e-05, "loss": 0.0064, "step": 35880 }, { "grad_norm": 0.12151923775672913, "learning_rate": 2.0240108074837666e-05, "loss": 0.0055, "step": 35890 }, { "grad_norm": 0.19967453181743622, "learning_rate": 2.021354070665393e-05, "loss": 0.0049, "step": 35900 }, { "grad_norm": 0.14432837069034576, "learning_rate": 2.0186986368079465e-05, "loss": 0.0067, "step": 35910 }, { "grad_norm": 0.16208310425281525, "learning_rate": 2.016044507073001e-05, "loss": 0.0053, "step": 35920 }, { "grad_norm": 0.21850311756134033, "learning_rate": 2.013391682621567e-05, "loss": 0.0051, "step": 35930 }, { "grad_norm": 0.14736701548099518, "learning_rate": 2.0107401646140787e-05, "loss": 0.0064, "step": 35940 }, { "grad_norm": 0.12233801931142807, "learning_rate": 2.0080899542104003e-05, "loss": 0.0051, "step": 35950 }, { "grad_norm": 0.1446283757686615, "learning_rate": 2.0054410525698215e-05, "loss": 0.0064, "step": 35960 }, { "grad_norm": 0.21161091327667236, "learning_rate": 2.0027934608510653e-05, "loss": 0.0066, "step": 35970 }, { "grad_norm": 0.15372416377067566, "learning_rate": 2.000147180212275e-05, "loss": 0.0083, "step": 35980 }, { "grad_norm": 0.1835821568965912, "learning_rate": 1.9975022118110275e-05, "loss": 0.0064, "step": 35990 }, { "grad_norm": 0.12086950242519379, "learning_rate": 1.994858556804318e-05, "loss": 0.0077, "step": 36000 }, { "grad_norm": 0.176798015832901, "learning_rate": 1.9922162163485748e-05, "loss": 0.0052, "step": 36010 }, { "grad_norm": 0.3386569619178772, "learning_rate": 1.989575191599643e-05, "loss": 0.0062, "step": 36020 }, { "grad_norm": 0.36662086844444275, "learning_rate": 1.9869354837128012e-05, "loss": 0.0082, "step": 36030 }, { "grad_norm": 1.422105312347412, "learning_rate": 1.984297093842745e-05, "loss": 0.0075, "step": 36040 }, { "grad_norm": 0.40828993916511536, "learning_rate": 1.981660023143594e-05, "loss": 0.0056, "step": 36050 }, { "grad_norm": 0.3742794990539551, "learning_rate": 1.9790242727688963e-05, "loss": 0.008, "step": 36060 }, { "grad_norm": 0.19766871631145477, "learning_rate": 1.9763898438716165e-05, "loss": 0.0053, "step": 36070 }, { "grad_norm": 0.4340510964393616, "learning_rate": 1.973756737604142e-05, "loss": 0.0073, "step": 36080 }, { "grad_norm": 0.12782388925552368, "learning_rate": 1.9711249551182858e-05, "loss": 0.0058, "step": 36090 }, { "grad_norm": 0.15129771828651428, "learning_rate": 1.968494497565278e-05, "loss": 0.0066, "step": 36100 }, { "grad_norm": 4.922865390777588, "learning_rate": 1.9658653660957698e-05, "loss": 0.0049, "step": 36110 }, { "grad_norm": 0.21194486320018768, "learning_rate": 1.9632375618598302e-05, "loss": 0.0056, "step": 36120 }, { "grad_norm": 0.1698664426803589, "learning_rate": 1.960611086006955e-05, "loss": 0.0049, "step": 36130 }, { "grad_norm": 0.14711657166481018, "learning_rate": 1.9579859396860513e-05, "loss": 0.0055, "step": 36140 }, { "grad_norm": 0.5330244898796082, "learning_rate": 1.9553621240454452e-05, "loss": 0.0058, "step": 36150 }, { "grad_norm": 1.5311013460159302, "learning_rate": 1.9527396402328864e-05, "loss": 0.0059, "step": 36160 }, { "grad_norm": 0.11414164304733276, "learning_rate": 1.9501184893955376e-05, "loss": 0.0064, "step": 36170 }, { "grad_norm": 0.29812091588974, "learning_rate": 1.9474986726799765e-05, "loss": 0.0065, "step": 36180 }, { "grad_norm": 0.16715799272060394, "learning_rate": 1.944880191232204e-05, "loss": 0.0058, "step": 36190 }, { "grad_norm": 0.20136186480522156, "learning_rate": 1.9422630461976304e-05, "loss": 0.0052, "step": 36200 }, { "grad_norm": 8.928753852844238, "learning_rate": 1.9396472387210835e-05, "loss": 0.0064, "step": 36210 }, { "grad_norm": 0.19302116334438324, "learning_rate": 1.9370327699468082e-05, "loss": 0.0052, "step": 36220 }, { "grad_norm": 0.15172573924064636, "learning_rate": 1.9344196410184594e-05, "loss": 0.0069, "step": 36230 }, { "grad_norm": 0.3660050332546234, "learning_rate": 1.9318078530791123e-05, "loss": 0.005, "step": 36240 }, { "grad_norm": 0.10053583979606628, "learning_rate": 1.9291974072712473e-05, "loss": 0.0051, "step": 36250 }, { "grad_norm": 0.15656454861164093, "learning_rate": 1.9265883047367656e-05, "loss": 0.0055, "step": 36260 }, { "grad_norm": 0.12205113470554352, "learning_rate": 1.9239805466169748e-05, "loss": 0.0053, "step": 36270 }, { "grad_norm": 0.28815391659736633, "learning_rate": 1.9213741340525992e-05, "loss": 0.0058, "step": 36280 }, { "grad_norm": 0.5088702440261841, "learning_rate": 1.9187690681837713e-05, "loss": 0.0054, "step": 36290 }, { "grad_norm": 0.1441364735364914, "learning_rate": 1.916165350150035e-05, "loss": 0.0051, "step": 36300 }, { "grad_norm": 0.2690218389034271, "learning_rate": 1.9135629810903434e-05, "loss": 0.0055, "step": 36310 }, { "grad_norm": 0.16282281279563904, "learning_rate": 1.9109619621430648e-05, "loss": 0.0061, "step": 36320 }, { "grad_norm": 0.20316565036773682, "learning_rate": 1.9083622944459712e-05, "loss": 0.0057, "step": 36330 }, { "grad_norm": 0.28121304512023926, "learning_rate": 1.9057639791362437e-05, "loss": 0.0064, "step": 36340 }, { "grad_norm": 0.12901484966278076, "learning_rate": 1.903167017350478e-05, "loss": 0.0049, "step": 36350 }, { "grad_norm": 0.4012267291545868, "learning_rate": 1.9005714102246707e-05, "loss": 0.0052, "step": 36360 }, { "grad_norm": 0.2941904067993164, "learning_rate": 1.897977158894227e-05, "loss": 0.0048, "step": 36370 }, { "grad_norm": 0.1460028886795044, "learning_rate": 1.8953842644939644e-05, "loss": 0.006, "step": 36380 }, { "grad_norm": 0.10777761042118073, "learning_rate": 1.8927927281581015e-05, "loss": 0.0067, "step": 36390 }, { "grad_norm": 0.37650108337402344, "learning_rate": 1.890202551020262e-05, "loss": 0.006, "step": 36400 }, { "grad_norm": 0.18149890005588531, "learning_rate": 1.8876137342134813e-05, "loss": 0.005, "step": 36410 }, { "grad_norm": 0.16541534662246704, "learning_rate": 1.885026278870194e-05, "loss": 0.0053, "step": 36420 }, { "grad_norm": 0.14289270341396332, "learning_rate": 1.8824401861222384e-05, "loss": 0.0059, "step": 36430 }, { "grad_norm": 0.4734090566635132, "learning_rate": 1.879855457100864e-05, "loss": 0.0056, "step": 36440 }, { "grad_norm": 0.27865031361579895, "learning_rate": 1.877272092936717e-05, "loss": 0.005, "step": 36450 }, { "grad_norm": 0.26095545291900635, "learning_rate": 1.8746900947598467e-05, "loss": 0.005, "step": 36460 }, { "grad_norm": 0.10315148532390594, "learning_rate": 1.87210946369971e-05, "loss": 0.0055, "step": 36470 }, { "grad_norm": 0.13533630967140198, "learning_rate": 1.8695302008851602e-05, "loss": 0.0053, "step": 36480 }, { "grad_norm": 0.12374185770750046, "learning_rate": 1.8669523074444568e-05, "loss": 0.0052, "step": 36490 }, { "grad_norm": 0.2218121439218521, "learning_rate": 1.8643757845052545e-05, "loss": 0.0057, "step": 36500 }, { "grad_norm": 0.10057786107063293, "learning_rate": 1.8618006331946164e-05, "loss": 0.0044, "step": 36510 }, { "grad_norm": 1.6252506971359253, "learning_rate": 1.859226854638999e-05, "loss": 0.0053, "step": 36520 }, { "grad_norm": 1.760777235031128, "learning_rate": 1.8566544499642587e-05, "loss": 0.0059, "step": 36530 }, { "grad_norm": 0.22267812490463257, "learning_rate": 1.854083420295656e-05, "loss": 0.0064, "step": 36540 }, { "grad_norm": 0.2660810053348541, "learning_rate": 1.8515137667578453e-05, "loss": 0.0051, "step": 36550 }, { "grad_norm": 0.9225879907608032, "learning_rate": 1.8489454904748784e-05, "loss": 0.0048, "step": 36560 }, { "grad_norm": 0.23431439697742462, "learning_rate": 1.8463785925702094e-05, "loss": 0.0043, "step": 36570 }, { "grad_norm": 0.1947307288646698, "learning_rate": 1.843813074166686e-05, "loss": 0.0064, "step": 36580 }, { "grad_norm": 0.16647179424762726, "learning_rate": 1.841248936386551e-05, "loss": 0.0075, "step": 36590 }, { "grad_norm": 0.19727462530136108, "learning_rate": 1.8386861803514483e-05, "loss": 0.0051, "step": 36600 }, { "grad_norm": 0.16973838210105896, "learning_rate": 1.836124807182414e-05, "loss": 0.0062, "step": 36610 }, { "grad_norm": 0.47753220796585083, "learning_rate": 1.8335648179998783e-05, "loss": 0.0063, "step": 36620 }, { "grad_norm": 0.16587227582931519, "learning_rate": 1.8310062139236667e-05, "loss": 0.0054, "step": 36630 }, { "grad_norm": 0.15804672241210938, "learning_rate": 1.828448996073002e-05, "loss": 0.0052, "step": 36640 }, { "grad_norm": 0.14325591921806335, "learning_rate": 1.825893165566498e-05, "loss": 0.0072, "step": 36650 }, { "grad_norm": 0.7622843980789185, "learning_rate": 1.823338723522159e-05, "loss": 0.0051, "step": 36660 }, { "grad_norm": 0.2070062756538391, "learning_rate": 1.8207856710573878e-05, "loss": 0.0063, "step": 36670 }, { "grad_norm": 0.13560838997364044, "learning_rate": 1.8182340092889756e-05, "loss": 0.0059, "step": 36680 }, { "grad_norm": 0.14477238059043884, "learning_rate": 1.8156837393331038e-05, "loss": 0.008, "step": 36690 }, { "grad_norm": 1.3663848638534546, "learning_rate": 1.81313486230535e-05, "loss": 0.0062, "step": 36700 }, { "grad_norm": 0.13645535707473755, "learning_rate": 1.8105873793206773e-05, "loss": 0.0048, "step": 36710 }, { "grad_norm": 0.1607402116060257, "learning_rate": 1.8080412914934436e-05, "loss": 0.0073, "step": 36720 }, { "grad_norm": 0.2082476019859314, "learning_rate": 1.805496599937392e-05, "loss": 0.005, "step": 36730 }, { "grad_norm": 0.36405855417251587, "learning_rate": 1.802953305765659e-05, "loss": 0.0045, "step": 36740 }, { "grad_norm": 0.19805742800235748, "learning_rate": 1.8004114100907653e-05, "loss": 0.0071, "step": 36750 }, { "grad_norm": 0.254256933927536, "learning_rate": 1.7978709140246263e-05, "loss": 0.006, "step": 36760 }, { "grad_norm": 0.1519225835800171, "learning_rate": 1.795331818678539e-05, "loss": 0.0043, "step": 36770 }, { "grad_norm": 0.1586618423461914, "learning_rate": 1.792794125163188e-05, "loss": 0.0046, "step": 36780 }, { "grad_norm": 0.6406729221343994, "learning_rate": 1.7902578345886507e-05, "loss": 0.0067, "step": 36790 }, { "grad_norm": 0.23510655760765076, "learning_rate": 1.7877229480643843e-05, "loss": 0.0068, "step": 36800 }, { "grad_norm": 0.0912669450044632, "learning_rate": 1.785189466699235e-05, "loss": 0.0055, "step": 36810 }, { "grad_norm": 0.36739668250083923, "learning_rate": 1.7826573916014317e-05, "loss": 0.0059, "step": 36820 }, { "grad_norm": 0.20436210930347443, "learning_rate": 1.7801267238785935e-05, "loss": 0.0072, "step": 36830 }, { "grad_norm": 0.2333240807056427, "learning_rate": 1.7775974646377187e-05, "loss": 0.0053, "step": 36840 }, { "grad_norm": 0.19974032044410706, "learning_rate": 1.775069614985189e-05, "loss": 0.0055, "step": 36850 }, { "grad_norm": 0.1580415517091751, "learning_rate": 1.7725431760267757e-05, "loss": 0.0059, "step": 36860 }, { "grad_norm": 0.1721716970205307, "learning_rate": 1.7700181488676277e-05, "loss": 0.0047, "step": 36870 }, { "grad_norm": 0.16550859808921814, "learning_rate": 1.767494534612275e-05, "loss": 0.0056, "step": 36880 }, { "grad_norm": 0.1991470903158188, "learning_rate": 1.7649723343646367e-05, "loss": 0.0055, "step": 36890 }, { "grad_norm": 0.21062445640563965, "learning_rate": 1.7624515492280063e-05, "loss": 0.0045, "step": 36900 }, { "grad_norm": 0.21747566759586334, "learning_rate": 1.7599321803050596e-05, "loss": 0.0085, "step": 36910 }, { "grad_norm": 0.13469244539737701, "learning_rate": 1.7574142286978574e-05, "loss": 0.0051, "step": 36920 }, { "grad_norm": 0.831383466720581, "learning_rate": 1.7548976955078354e-05, "loss": 0.0046, "step": 36930 }, { "grad_norm": 0.1732945591211319, "learning_rate": 1.7523825818358092e-05, "loss": 0.0059, "step": 36940 }, { "grad_norm": 0.1789679229259491, "learning_rate": 1.7498688887819785e-05, "loss": 0.0078, "step": 36950 }, { "grad_norm": 0.3609890043735504, "learning_rate": 1.7473566174459143e-05, "loss": 0.0063, "step": 36960 }, { "grad_norm": 0.14372076094150543, "learning_rate": 1.7448457689265735e-05, "loss": 0.005, "step": 36970 }, { "grad_norm": 0.13052639365196228, "learning_rate": 1.7423363443222823e-05, "loss": 0.0051, "step": 36980 }, { "grad_norm": 0.23554545640945435, "learning_rate": 1.7398283447307524e-05, "loss": 0.005, "step": 36990 }, { "grad_norm": 0.09368110448122025, "learning_rate": 1.737321771249066e-05, "loss": 0.0065, "step": 37000 }, { "grad_norm": 0.22103412449359894, "learning_rate": 1.7348166249736814e-05, "loss": 0.0057, "step": 37010 }, { "grad_norm": 0.26835498213768005, "learning_rate": 1.732312907000439e-05, "loss": 0.0054, "step": 37020 }, { "grad_norm": 0.18481209874153137, "learning_rate": 1.7298106184245478e-05, "loss": 0.0051, "step": 37030 }, { "grad_norm": 0.327660471200943, "learning_rate": 1.727309760340592e-05, "loss": 0.0073, "step": 37040 }, { "grad_norm": 0.13324189186096191, "learning_rate": 1.724810333842536e-05, "loss": 0.0059, "step": 37050 }, { "grad_norm": 0.1580764800310135, "learning_rate": 1.7223123400237112e-05, "loss": 0.0067, "step": 37060 }, { "grad_norm": 0.1220095083117485, "learning_rate": 1.7198157799768238e-05, "loss": 0.0059, "step": 37070 }, { "grad_norm": 0.13399319350719452, "learning_rate": 1.7173206547939576e-05, "loss": 0.0043, "step": 37080 }, { "grad_norm": 0.3585093319416046, "learning_rate": 1.7148269655665628e-05, "loss": 0.0063, "step": 37090 }, { "grad_norm": 4.7723164558410645, "learning_rate": 1.712334713385463e-05, "loss": 0.0057, "step": 37100 }, { "grad_norm": 0.6190882921218872, "learning_rate": 1.709843899340856e-05, "loss": 0.0055, "step": 37110 }, { "grad_norm": 0.143168643116951, "learning_rate": 1.7073545245223084e-05, "loss": 0.0067, "step": 37120 }, { "grad_norm": 0.21354687213897705, "learning_rate": 1.704866590018755e-05, "loss": 0.0051, "step": 37130 }, { "grad_norm": 0.11577665060758591, "learning_rate": 1.7023800969185054e-05, "loss": 0.0041, "step": 37140 }, { "grad_norm": 0.1648484766483307, "learning_rate": 1.6998950463092354e-05, "loss": 0.0054, "step": 37150 }, { "grad_norm": 0.6973658800125122, "learning_rate": 1.6974114392779907e-05, "loss": 0.0045, "step": 37160 }, { "grad_norm": 0.11278532445430756, "learning_rate": 1.694929276911183e-05, "loss": 0.0054, "step": 37170 }, { "grad_norm": 0.23674388229846954, "learning_rate": 1.6924485602945983e-05, "loss": 0.006, "step": 37180 }, { "grad_norm": 0.20323805510997772, "learning_rate": 1.6899692905133825e-05, "loss": 0.0069, "step": 37190 }, { "grad_norm": 0.24648931622505188, "learning_rate": 1.6874914686520565e-05, "loss": 0.0062, "step": 37200 }, { "grad_norm": 0.3843420445919037, "learning_rate": 1.6850150957945e-05, "loss": 0.0077, "step": 37210 }, { "grad_norm": 0.382053405046463, "learning_rate": 1.682540173023966e-05, "loss": 0.0065, "step": 37220 }, { "grad_norm": 0.20055758953094482, "learning_rate": 1.6800667014230665e-05, "loss": 0.0075, "step": 37230 }, { "grad_norm": 0.12941856682300568, "learning_rate": 1.677594682073785e-05, "loss": 0.006, "step": 37240 }, { "grad_norm": 0.12885962426662445, "learning_rate": 1.6751241160574653e-05, "loss": 0.0049, "step": 37250 }, { "grad_norm": 0.12363293021917343, "learning_rate": 1.6726550044548156e-05, "loss": 0.0051, "step": 37260 }, { "grad_norm": 0.15874169766902924, "learning_rate": 1.6701873483459125e-05, "loss": 0.0046, "step": 37270 }, { "grad_norm": 0.14366663992404938, "learning_rate": 1.6677211488101906e-05, "loss": 0.0053, "step": 37280 }, { "grad_norm": 0.2532627582550049, "learning_rate": 1.6652564069264475e-05, "loss": 0.004, "step": 37290 }, { "grad_norm": 0.18026360869407654, "learning_rate": 1.662793123772849e-05, "loss": 0.0056, "step": 37300 }, { "grad_norm": 0.10462182760238647, "learning_rate": 1.6603313004269168e-05, "loss": 0.0053, "step": 37310 }, { "grad_norm": 0.16260065138339996, "learning_rate": 1.657870937965536e-05, "loss": 0.0078, "step": 37320 }, { "grad_norm": 0.1528710275888443, "learning_rate": 1.655412037464952e-05, "loss": 0.0066, "step": 37330 }, { "grad_norm": 0.1358187049627304, "learning_rate": 1.652954600000773e-05, "loss": 0.0051, "step": 37340 }, { "grad_norm": 0.19231025874614716, "learning_rate": 1.650498626647965e-05, "loss": 0.0049, "step": 37350 }, { "grad_norm": 0.22855326533317566, "learning_rate": 1.6480441184808526e-05, "loss": 0.005, "step": 37360 }, { "grad_norm": 0.25470423698425293, "learning_rate": 1.6455910765731242e-05, "loss": 0.0048, "step": 37370 }, { "grad_norm": 0.162424698472023, "learning_rate": 1.6431395019978217e-05, "loss": 0.0053, "step": 37380 }, { "grad_norm": 0.41013282537460327, "learning_rate": 1.6406893958273467e-05, "loss": 0.0058, "step": 37390 }, { "grad_norm": 0.1532868593931198, "learning_rate": 1.6382407591334602e-05, "loss": 0.0056, "step": 37400 }, { "grad_norm": 0.17548033595085144, "learning_rate": 1.635793592987279e-05, "loss": 0.0071, "step": 37410 }, { "grad_norm": 0.1295548975467682, "learning_rate": 1.633347898459275e-05, "loss": 0.005, "step": 37420 }, { "grad_norm": 0.13442325592041016, "learning_rate": 1.6309036766192805e-05, "loss": 0.0053, "step": 37430 }, { "grad_norm": 0.16491727530956268, "learning_rate": 1.628460928536478e-05, "loss": 0.0081, "step": 37440 }, { "grad_norm": 0.23166252672672272, "learning_rate": 1.6260196552794128e-05, "loss": 0.0061, "step": 37450 }, { "grad_norm": 1.1204402446746826, "learning_rate": 1.6235798579159766e-05, "loss": 0.006, "step": 37460 }, { "grad_norm": 0.15987586975097656, "learning_rate": 1.6211415375134233e-05, "loss": 0.0068, "step": 37470 }, { "grad_norm": 0.26745039224624634, "learning_rate": 1.618704695138353e-05, "loss": 0.0049, "step": 37480 }, { "grad_norm": 0.6137509346008301, "learning_rate": 1.6162693318567273e-05, "loss": 0.0066, "step": 37490 }, { "grad_norm": 0.19049805402755737, "learning_rate": 1.613835448733856e-05, "loss": 0.0061, "step": 37500 }, { "grad_norm": 0.13386887311935425, "learning_rate": 1.6114030468344006e-05, "loss": 0.0058, "step": 37510 }, { "grad_norm": 0.2548896074295044, "learning_rate": 1.6089721272223755e-05, "loss": 0.0067, "step": 37520 }, { "grad_norm": 0.2358594685792923, "learning_rate": 1.606542690961151e-05, "loss": 0.0055, "step": 37530 }, { "grad_norm": 0.348614901304245, "learning_rate": 1.6041147391134438e-05, "loss": 0.0049, "step": 37540 }, { "grad_norm": 0.11187771707773209, "learning_rate": 1.6016882727413196e-05, "loss": 0.0052, "step": 37550 }, { "grad_norm": 0.18730932474136353, "learning_rate": 1.5992632929062014e-05, "loss": 0.0049, "step": 37560 }, { "grad_norm": 0.1468069702386856, "learning_rate": 1.5968398006688557e-05, "loss": 0.0064, "step": 37570 }, { "grad_norm": 0.12428442388772964, "learning_rate": 1.594417797089399e-05, "loss": 0.0053, "step": 37580 }, { "grad_norm": 0.21124149858951569, "learning_rate": 1.5919972832273015e-05, "loss": 0.0054, "step": 37590 }, { "grad_norm": 0.12185671180486679, "learning_rate": 1.5895782601413767e-05, "loss": 0.0038, "step": 37600 }, { "grad_norm": 0.18446838855743408, "learning_rate": 1.587160728889785e-05, "loss": 0.0052, "step": 37610 }, { "grad_norm": 0.3813819885253906, "learning_rate": 1.58474469053004e-05, "loss": 0.005, "step": 37620 }, { "grad_norm": 0.08896208554506302, "learning_rate": 1.5823301461189976e-05, "loss": 0.0043, "step": 37630 }, { "grad_norm": 0.281438946723938, "learning_rate": 1.5799170967128606e-05, "loss": 0.0054, "step": 37640 }, { "grad_norm": 0.28990620374679565, "learning_rate": 1.5775055433671814e-05, "loss": 0.0066, "step": 37650 }, { "grad_norm": 0.10765132308006287, "learning_rate": 1.5750954871368535e-05, "loss": 0.0054, "step": 37660 }, { "grad_norm": 0.1455782800912857, "learning_rate": 1.5726869290761158e-05, "loss": 0.0056, "step": 37670 }, { "grad_norm": 0.15914948284626007, "learning_rate": 1.5702798702385574e-05, "loss": 0.0078, "step": 37680 }, { "grad_norm": 0.271384596824646, "learning_rate": 1.5678743116771038e-05, "loss": 0.0064, "step": 37690 }, { "grad_norm": 0.3729017376899719, "learning_rate": 1.5654702544440313e-05, "loss": 0.0055, "step": 37700 }, { "grad_norm": 0.19838353991508484, "learning_rate": 1.5630676995909532e-05, "loss": 0.0057, "step": 37710 }, { "grad_norm": 0.16106098890304565, "learning_rate": 1.560666648168832e-05, "loss": 0.0051, "step": 37720 }, { "grad_norm": 0.16885638236999512, "learning_rate": 1.5582671012279663e-05, "loss": 0.0046, "step": 37730 }, { "grad_norm": 0.32449638843536377, "learning_rate": 1.5558690598179997e-05, "loss": 0.0045, "step": 37740 }, { "grad_norm": 0.7352672815322876, "learning_rate": 1.553472524987919e-05, "loss": 0.0048, "step": 37750 }, { "grad_norm": 0.23970402777194977, "learning_rate": 1.5510774977860485e-05, "loss": 0.0054, "step": 37760 }, { "grad_norm": 0.1773209273815155, "learning_rate": 1.5486839792600532e-05, "loss": 0.0055, "step": 37770 }, { "grad_norm": 0.1446879655122757, "learning_rate": 1.546291970456942e-05, "loss": 0.0061, "step": 37780 }, { "grad_norm": 0.20614562928676605, "learning_rate": 1.5439014724230606e-05, "loss": 0.0042, "step": 37790 }, { "grad_norm": 0.24796801805496216, "learning_rate": 1.5415124862040907e-05, "loss": 0.0047, "step": 37800 }, { "grad_norm": 0.24407152831554413, "learning_rate": 1.53912501284506e-05, "loss": 0.0061, "step": 37810 }, { "grad_norm": 0.38655510544776917, "learning_rate": 1.5367390533903292e-05, "loss": 0.005, "step": 37820 }, { "grad_norm": 0.24470500648021698, "learning_rate": 1.5343546088835964e-05, "loss": 0.0043, "step": 37830 }, { "grad_norm": 0.1941862404346466, "learning_rate": 1.531971680367901e-05, "loss": 0.0056, "step": 37840 }, { "grad_norm": 0.166323721408844, "learning_rate": 1.529590268885616e-05, "loss": 0.0048, "step": 37850 }, { "grad_norm": 0.16484901309013367, "learning_rate": 1.5272103754784517e-05, "loss": 0.0048, "step": 37860 }, { "grad_norm": 4.156395435333252, "learning_rate": 1.5248320011874523e-05, "loss": 0.005, "step": 37870 }, { "grad_norm": 0.12168317288160324, "learning_rate": 1.5224551470530035e-05, "loss": 0.0043, "step": 37880 }, { "grad_norm": 0.3345770239830017, "learning_rate": 1.5200798141148193e-05, "loss": 0.0067, "step": 37890 }, { "grad_norm": 0.18975165486335754, "learning_rate": 1.5177060034119505e-05, "loss": 0.0038, "step": 37900 }, { "grad_norm": 1.1231197118759155, "learning_rate": 1.5153337159827835e-05, "loss": 0.0046, "step": 37910 }, { "grad_norm": 0.4466826021671295, "learning_rate": 1.5129629528650397e-05, "loss": 0.0048, "step": 37920 }, { "grad_norm": 0.15559686720371246, "learning_rate": 1.5105937150957671e-05, "loss": 0.0052, "step": 37930 }, { "grad_norm": 0.11247314512729645, "learning_rate": 1.5082260037113548e-05, "loss": 0.0048, "step": 37940 }, { "grad_norm": 0.12634748220443726, "learning_rate": 1.5058598197475187e-05, "loss": 0.004, "step": 37950 }, { "grad_norm": 0.24332652986049652, "learning_rate": 1.5034951642393064e-05, "loss": 0.0061, "step": 37960 }, { "grad_norm": 1.0790754556655884, "learning_rate": 1.5011320382211013e-05, "loss": 0.005, "step": 37970 }, { "grad_norm": 0.23034806549549103, "learning_rate": 1.4987704427266136e-05, "loss": 0.0056, "step": 37980 }, { "grad_norm": 0.09563759714365005, "learning_rate": 1.4964103787888839e-05, "loss": 0.0046, "step": 37990 }, { "grad_norm": 0.24529044330120087, "learning_rate": 1.4940518474402882e-05, "loss": 0.0057, "step": 38000 }, { "grad_norm": 0.19562827050685883, "learning_rate": 1.4916948497125249e-05, "loss": 0.0057, "step": 38010 }, { "grad_norm": 0.3962743878364563, "learning_rate": 1.4893393866366267e-05, "loss": 0.0063, "step": 38020 }, { "grad_norm": 0.2077263444662094, "learning_rate": 1.4869854592429506e-05, "loss": 0.0051, "step": 38030 }, { "grad_norm": 0.24640601873397827, "learning_rate": 1.4846330685611875e-05, "loss": 0.0045, "step": 38040 }, { "grad_norm": 0.19920343160629272, "learning_rate": 1.482282215620352e-05, "loss": 0.0052, "step": 38050 }, { "grad_norm": 0.2712963819503784, "learning_rate": 1.4799329014487857e-05, "loss": 0.0048, "step": 38060 }, { "grad_norm": 0.22860875725746155, "learning_rate": 1.4775851270741603e-05, "loss": 0.0044, "step": 38070 }, { "grad_norm": 0.31241247057914734, "learning_rate": 1.4752388935234719e-05, "loss": 0.0053, "step": 38080 }, { "grad_norm": 1.135655164718628, "learning_rate": 1.4728942018230401e-05, "loss": 0.0068, "step": 38090 }, { "grad_norm": 1.7207478284835815, "learning_rate": 1.4705510529985167e-05, "loss": 0.0091, "step": 38100 }, { "grad_norm": 1.6448843479156494, "learning_rate": 1.4682094480748726e-05, "loss": 0.0053, "step": 38110 }, { "grad_norm": 4.239171981811523, "learning_rate": 1.4658693880764035e-05, "loss": 0.0067, "step": 38120 }, { "grad_norm": 0.16547854244709015, "learning_rate": 1.463530874026735e-05, "loss": 0.0071, "step": 38130 }, { "grad_norm": 0.09600922465324402, "learning_rate": 1.461193906948809e-05, "loss": 0.0045, "step": 38140 }, { "grad_norm": 0.519223153591156, "learning_rate": 1.4588584878648975e-05, "loss": 0.0042, "step": 38150 }, { "grad_norm": 0.14938203990459442, "learning_rate": 1.4565246177965896e-05, "loss": 0.0056, "step": 38160 }, { "grad_norm": 0.3116604685783386, "learning_rate": 1.4541922977648015e-05, "loss": 0.0061, "step": 38170 }, { "grad_norm": 0.22210104763507843, "learning_rate": 1.451861528789768e-05, "loss": 0.0054, "step": 38180 }, { "grad_norm": 0.19611547887325287, "learning_rate": 1.4495323118910454e-05, "loss": 0.005, "step": 38190 }, { "grad_norm": 12.07046127319336, "learning_rate": 1.4472046480875145e-05, "loss": 0.0066, "step": 38200 }, { "grad_norm": 0.2383204996585846, "learning_rate": 1.4448785383973745e-05, "loss": 0.0048, "step": 38210 }, { "grad_norm": 0.4310292899608612, "learning_rate": 1.4425539838381418e-05, "loss": 0.0046, "step": 38220 }, { "grad_norm": 1.1145246028900146, "learning_rate": 1.4402309854266594e-05, "loss": 0.0058, "step": 38230 }, { "grad_norm": 0.36388468742370605, "learning_rate": 1.4379095441790846e-05, "loss": 0.0049, "step": 38240 }, { "grad_norm": 0.3554711639881134, "learning_rate": 1.4355896611108921e-05, "loss": 0.0062, "step": 38250 }, { "grad_norm": 0.33920979499816895, "learning_rate": 1.433271337236881e-05, "loss": 0.0056, "step": 38260 }, { "grad_norm": 0.1838928759098053, "learning_rate": 1.4309545735711638e-05, "loss": 0.0055, "step": 38270 }, { "grad_norm": 0.20100268721580505, "learning_rate": 1.4286393711271696e-05, "loss": 0.0054, "step": 38280 }, { "grad_norm": 0.17744708061218262, "learning_rate": 1.4263257309176497e-05, "loss": 0.0054, "step": 38290 }, { "grad_norm": 0.12126003205776215, "learning_rate": 1.4240136539546678e-05, "loss": 0.0049, "step": 38300 }, { "grad_norm": 0.18075880408287048, "learning_rate": 1.4217031412496029e-05, "loss": 0.0053, "step": 38310 }, { "grad_norm": 0.11769892275333405, "learning_rate": 1.4193941938131554e-05, "loss": 0.004, "step": 38320 }, { "grad_norm": 0.3390062749385834, "learning_rate": 1.4170868126553356e-05, "loss": 0.0054, "step": 38330 }, { "grad_norm": 0.26942121982574463, "learning_rate": 1.4147809987854682e-05, "loss": 0.0082, "step": 38340 }, { "grad_norm": 0.13493454456329346, "learning_rate": 1.4124767532121985e-05, "loss": 0.0058, "step": 38350 }, { "grad_norm": 0.3334115445613861, "learning_rate": 1.41017407694348e-05, "loss": 0.0043, "step": 38360 }, { "grad_norm": 0.4270144999027252, "learning_rate": 1.4078729709865801e-05, "loss": 0.005, "step": 38370 }, { "grad_norm": 0.45628631114959717, "learning_rate": 1.4055734363480833e-05, "loss": 0.0052, "step": 38380 }, { "grad_norm": 0.2855791449546814, "learning_rate": 1.4032754740338815e-05, "loss": 0.0046, "step": 38390 }, { "grad_norm": 0.3521363437175751, "learning_rate": 1.4009790850491844e-05, "loss": 0.0049, "step": 38400 }, { "grad_norm": 0.384857714176178, "learning_rate": 1.3986842703985076e-05, "loss": 0.0069, "step": 38410 }, { "grad_norm": 0.3641919791698456, "learning_rate": 1.3963910310856842e-05, "loss": 0.0053, "step": 38420 }, { "grad_norm": 0.18435406684875488, "learning_rate": 1.394099368113853e-05, "loss": 0.0055, "step": 38430 }, { "grad_norm": 0.12490350008010864, "learning_rate": 1.3918092824854644e-05, "loss": 0.0036, "step": 38440 }, { "grad_norm": 0.15675707161426544, "learning_rate": 1.3895207752022821e-05, "loss": 0.0045, "step": 38450 }, { "grad_norm": 0.09712623059749603, "learning_rate": 1.3872338472653756e-05, "loss": 0.0047, "step": 38460 }, { "grad_norm": 0.16181093454360962, "learning_rate": 1.3849484996751233e-05, "loss": 0.0052, "step": 38470 }, { "grad_norm": 0.2885395586490631, "learning_rate": 1.382664733431217e-05, "loss": 0.005, "step": 38480 }, { "grad_norm": 0.4318568706512451, "learning_rate": 1.3803825495326522e-05, "loss": 0.0053, "step": 38490 }, { "grad_norm": 0.23205769062042236, "learning_rate": 1.3781019489777324e-05, "loss": 0.0066, "step": 38500 }, { "grad_norm": 0.1380842924118042, "learning_rate": 1.375822932764072e-05, "loss": 0.0067, "step": 38510 }, { "grad_norm": 0.15851260721683502, "learning_rate": 1.373545501888589e-05, "loss": 0.0051, "step": 38520 }, { "grad_norm": 4.241770267486572, "learning_rate": 1.3712696573475092e-05, "loss": 0.005, "step": 38530 }, { "grad_norm": 0.4191132187843323, "learning_rate": 1.368995400136363e-05, "loss": 0.0071, "step": 38540 }, { "grad_norm": 0.38829222321510315, "learning_rate": 1.3667227312499903e-05, "loss": 0.0073, "step": 38550 }, { "grad_norm": 0.13071587681770325, "learning_rate": 1.3644516516825323e-05, "loss": 0.0048, "step": 38560 }, { "grad_norm": 0.6081526875495911, "learning_rate": 1.3621821624274356e-05, "loss": 0.0059, "step": 38570 }, { "grad_norm": 0.14339013397693634, "learning_rate": 1.3599142644774537e-05, "loss": 0.0053, "step": 38580 }, { "grad_norm": 0.23397402465343475, "learning_rate": 1.357647958824641e-05, "loss": 0.0042, "step": 38590 }, { "grad_norm": 0.796136200428009, "learning_rate": 1.3553832464603555e-05, "loss": 0.0063, "step": 38600 }, { "grad_norm": 0.3157965838909149, "learning_rate": 1.3531201283752625e-05, "loss": 0.0052, "step": 38610 }, { "grad_norm": 0.14836567640304565, "learning_rate": 1.350858605559323e-05, "loss": 0.0065, "step": 38620 }, { "grad_norm": 0.13409291207790375, "learning_rate": 1.348598679001808e-05, "loss": 0.0053, "step": 38630 }, { "grad_norm": 0.3721282482147217, "learning_rate": 1.3463403496912819e-05, "loss": 0.0049, "step": 38640 }, { "grad_norm": 0.26590657234191895, "learning_rate": 1.3440836186156187e-05, "loss": 0.005, "step": 38650 }, { "grad_norm": 0.13475751876831055, "learning_rate": 1.3418284867619852e-05, "loss": 0.0042, "step": 38660 }, { "grad_norm": 0.3683088421821594, "learning_rate": 1.3395749551168567e-05, "loss": 0.0052, "step": 38670 }, { "grad_norm": 0.1829182356595993, "learning_rate": 1.3373230246660024e-05, "loss": 0.0066, "step": 38680 }, { "grad_norm": 0.10686242580413818, "learning_rate": 1.335072696394492e-05, "loss": 0.0061, "step": 38690 }, { "grad_norm": 0.11626376956701279, "learning_rate": 1.3328239712866986e-05, "loss": 0.0045, "step": 38700 }, { "grad_norm": 0.12643195688724518, "learning_rate": 1.3305768503262889e-05, "loss": 0.0044, "step": 38710 }, { "grad_norm": 0.4552939534187317, "learning_rate": 1.3283313344962305e-05, "loss": 0.0051, "step": 38720 }, { "grad_norm": 0.29704222083091736, "learning_rate": 1.3260874247787863e-05, "loss": 0.0052, "step": 38730 }, { "grad_norm": 0.23443108797073364, "learning_rate": 1.3238451221555226e-05, "loss": 0.0052, "step": 38740 }, { "grad_norm": 0.23148344457149506, "learning_rate": 1.3216044276072965e-05, "loss": 0.0055, "step": 38750 }, { "grad_norm": 0.19350141286849976, "learning_rate": 1.3193653421142622e-05, "loss": 0.0047, "step": 38760 }, { "grad_norm": 0.24927812814712524, "learning_rate": 1.3171278666558756e-05, "loss": 0.006, "step": 38770 }, { "grad_norm": 0.24143555760383606, "learning_rate": 1.3148920022108819e-05, "loss": 0.0045, "step": 38780 }, { "grad_norm": 0.973011314868927, "learning_rate": 1.312657749757324e-05, "loss": 0.0054, "step": 38790 }, { "grad_norm": 0.15692035853862762, "learning_rate": 1.3104251102725412e-05, "loss": 0.0053, "step": 38800 }, { "grad_norm": 0.1755124032497406, "learning_rate": 1.3081940847331659e-05, "loss": 0.0057, "step": 38810 }, { "grad_norm": 0.13255056738853455, "learning_rate": 1.3059646741151222e-05, "loss": 0.0057, "step": 38820 }, { "grad_norm": 0.7944989800453186, "learning_rate": 1.3037368793936334e-05, "loss": 0.0047, "step": 38830 }, { "grad_norm": 0.148971289396286, "learning_rate": 1.3015107015432104e-05, "loss": 0.006, "step": 38840 }, { "grad_norm": 0.15734733641147614, "learning_rate": 1.2992861415376584e-05, "loss": 0.0046, "step": 38850 }, { "grad_norm": 0.10133551061153412, "learning_rate": 1.2970632003500782e-05, "loss": 0.0066, "step": 38860 }, { "grad_norm": 0.5771617293357849, "learning_rate": 1.294841878952856e-05, "loss": 0.0054, "step": 38870 }, { "grad_norm": 1.6698887348175049, "learning_rate": 1.2926221783176779e-05, "loss": 0.0053, "step": 38880 }, { "grad_norm": 0.18741029500961304, "learning_rate": 1.290404099415512e-05, "loss": 0.0048, "step": 38890 }, { "grad_norm": 0.10219743102788925, "learning_rate": 1.2881876432166246e-05, "loss": 0.0054, "step": 38900 }, { "grad_norm": 0.22965176403522491, "learning_rate": 1.2859728106905678e-05, "loss": 0.0058, "step": 38910 }, { "grad_norm": 0.531288743019104, "learning_rate": 1.2837596028061832e-05, "loss": 0.0045, "step": 38920 }, { "grad_norm": 0.09334515780210495, "learning_rate": 1.2815480205316055e-05, "loss": 0.004, "step": 38930 }, { "grad_norm": 0.1292685866355896, "learning_rate": 1.279338064834255e-05, "loss": 0.0046, "step": 38940 }, { "grad_norm": 0.23406344652175903, "learning_rate": 1.27712973668084e-05, "loss": 0.0055, "step": 38950 }, { "grad_norm": 0.1169254407286644, "learning_rate": 1.2749230370373605e-05, "loss": 0.0041, "step": 38960 }, { "grad_norm": 0.31092023849487305, "learning_rate": 1.2727179668691009e-05, "loss": 0.0048, "step": 38970 }, { "grad_norm": 3.258925676345825, "learning_rate": 1.2705145271406327e-05, "loss": 0.0048, "step": 38980 }, { "grad_norm": 0.09011860191822052, "learning_rate": 1.2683127188158172e-05, "loss": 0.0064, "step": 38990 }, { "grad_norm": 0.14530633389949799, "learning_rate": 1.2661125428577997e-05, "loss": 0.0051, "step": 39000 }, { "grad_norm": 0.22834953665733337, "learning_rate": 1.2639140002290101e-05, "loss": 0.0056, "step": 39010 }, { "grad_norm": 0.09294521808624268, "learning_rate": 1.261717091891168e-05, "loss": 0.0051, "step": 39020 }, { "grad_norm": 1.8799153566360474, "learning_rate": 1.2595218188052755e-05, "loss": 0.0054, "step": 39030 }, { "grad_norm": 0.7290140390396118, "learning_rate": 1.2573281819316174e-05, "loss": 0.0064, "step": 39040 }, { "grad_norm": 0.22475263476371765, "learning_rate": 1.2551361822297674e-05, "loss": 0.0041, "step": 39050 }, { "grad_norm": 0.23321963846683502, "learning_rate": 1.2529458206585804e-05, "loss": 0.0068, "step": 39060 }, { "grad_norm": 0.14553648233413696, "learning_rate": 1.250757098176194e-05, "loss": 0.0051, "step": 39070 }, { "grad_norm": 0.3433811664581299, "learning_rate": 1.2485700157400294e-05, "loss": 0.0051, "step": 39080 }, { "grad_norm": 0.14136958122253418, "learning_rate": 1.246384574306792e-05, "loss": 0.0048, "step": 39090 }, { "grad_norm": 0.19643563032150269, "learning_rate": 1.2442007748324664e-05, "loss": 0.0058, "step": 39100 }, { "grad_norm": 0.239875927567482, "learning_rate": 1.2420186182723237e-05, "loss": 0.0049, "step": 39110 }, { "grad_norm": 0.193478524684906, "learning_rate": 1.23983810558091e-05, "loss": 0.0073, "step": 39120 }, { "grad_norm": 0.2036081999540329, "learning_rate": 1.2376592377120582e-05, "loss": 0.0054, "step": 39130 }, { "grad_norm": 2.577451467514038, "learning_rate": 1.2354820156188768e-05, "loss": 0.0063, "step": 39140 }, { "grad_norm": 0.31530648469924927, "learning_rate": 1.2333064402537597e-05, "loss": 0.0045, "step": 39150 }, { "grad_norm": 0.9233940243721008, "learning_rate": 1.2311325125683754e-05, "loss": 0.0051, "step": 39160 }, { "grad_norm": 0.2436484843492508, "learning_rate": 1.228960233513673e-05, "loss": 0.0036, "step": 39170 }, { "grad_norm": 0.10610280930995941, "learning_rate": 1.2267896040398836e-05, "loss": 0.0039, "step": 39180 }, { "grad_norm": 0.10209569334983826, "learning_rate": 1.2246206250965125e-05, "loss": 0.0055, "step": 39190 }, { "grad_norm": 0.3027985095977783, "learning_rate": 1.2224532976323444e-05, "loss": 0.0045, "step": 39200 }, { "grad_norm": 0.39932119846343994, "learning_rate": 1.220287622595444e-05, "loss": 0.0049, "step": 39210 }, { "grad_norm": 0.31280064582824707, "learning_rate": 1.2181236009331497e-05, "loss": 0.0051, "step": 39220 }, { "grad_norm": 0.23831108212471008, "learning_rate": 1.2159612335920784e-05, "loss": 0.0046, "step": 39230 }, { "grad_norm": 0.25206562876701355, "learning_rate": 1.2138005215181219e-05, "loss": 0.0067, "step": 39240 }, { "grad_norm": 0.15629924833774567, "learning_rate": 1.2116414656564512e-05, "loss": 0.004, "step": 39250 }, { "grad_norm": 0.1887701004743576, "learning_rate": 1.2094840669515095e-05, "loss": 0.0047, "step": 39260 }, { "grad_norm": 0.1919267326593399, "learning_rate": 1.2073283263470152e-05, "loss": 0.0049, "step": 39270 }, { "grad_norm": 0.15580184757709503, "learning_rate": 1.2051742447859653e-05, "loss": 0.0076, "step": 39280 }, { "grad_norm": 0.25243639945983887, "learning_rate": 1.203021823210626e-05, "loss": 0.0072, "step": 39290 }, { "grad_norm": 0.1749984622001648, "learning_rate": 1.2008710625625397e-05, "loss": 0.0045, "step": 39300 }, { "grad_norm": 0.15938745439052582, "learning_rate": 1.1987219637825236e-05, "loss": 0.0048, "step": 39310 }, { "grad_norm": 0.22096838057041168, "learning_rate": 1.1965745278106656e-05, "loss": 0.0054, "step": 39320 }, { "grad_norm": 0.13821715116500854, "learning_rate": 1.1944287555863259e-05, "loss": 0.0061, "step": 39330 }, { "grad_norm": 0.2525269091129303, "learning_rate": 1.1922846480481409e-05, "loss": 0.0048, "step": 39340 }, { "grad_norm": 0.30580827593803406, "learning_rate": 1.1901422061340133e-05, "loss": 0.0062, "step": 39350 }, { "grad_norm": 0.185045525431633, "learning_rate": 1.1880014307811227e-05, "loss": 0.0058, "step": 39360 }, { "grad_norm": 0.18961070477962494, "learning_rate": 1.1858623229259142e-05, "loss": 0.004, "step": 39370 }, { "grad_norm": 0.15233024954795837, "learning_rate": 1.1837248835041092e-05, "loss": 0.0043, "step": 39380 }, { "grad_norm": 0.12700259685516357, "learning_rate": 1.1815891134506935e-05, "loss": 0.0055, "step": 39390 }, { "grad_norm": 0.16711997985839844, "learning_rate": 1.1794550136999277e-05, "loss": 0.0046, "step": 39400 }, { "grad_norm": 0.7303370237350464, "learning_rate": 1.1773225851853387e-05, "loss": 0.0051, "step": 39410 }, { "grad_norm": 2.6138713359832764, "learning_rate": 1.1751918288397235e-05, "loss": 0.0059, "step": 39420 }, { "grad_norm": 0.25058814883232117, "learning_rate": 1.1730627455951449e-05, "loss": 0.0057, "step": 39430 }, { "grad_norm": 0.2322818636894226, "learning_rate": 1.17093533638294e-05, "loss": 0.0066, "step": 39440 }, { "grad_norm": 0.1436886489391327, "learning_rate": 1.168809602133708e-05, "loss": 0.0047, "step": 39450 }, { "grad_norm": 0.21872520446777344, "learning_rate": 1.1666855437773162e-05, "loss": 0.0042, "step": 39460 }, { "grad_norm": 0.11606911569833755, "learning_rate": 1.1645631622429026e-05, "loss": 0.0043, "step": 39470 }, { "grad_norm": 0.24330393970012665, "learning_rate": 1.1624424584588673e-05, "loss": 0.0043, "step": 39480 }, { "grad_norm": 0.2575834095478058, "learning_rate": 1.1603234333528768e-05, "loss": 0.0043, "step": 39490 }, { "grad_norm": 0.23474936187267303, "learning_rate": 1.1582060878518674e-05, "loss": 0.0049, "step": 39500 }, { "grad_norm": 0.21471315622329712, "learning_rate": 1.1560904228820375e-05, "loss": 0.0085, "step": 39510 }, { "grad_norm": 0.1922936737537384, "learning_rate": 1.1539764393688478e-05, "loss": 0.0055, "step": 39520 }, { "grad_norm": 0.12726567685604095, "learning_rate": 1.151864138237031e-05, "loss": 0.0052, "step": 39530 }, { "grad_norm": 0.16339829564094543, "learning_rate": 1.1497535204105769e-05, "loss": 0.0057, "step": 39540 }, { "grad_norm": 0.10817085951566696, "learning_rate": 1.14764458681274e-05, "loss": 0.004, "step": 39550 }, { "grad_norm": 0.12459404766559601, "learning_rate": 1.1455373383660434e-05, "loss": 0.0052, "step": 39560 }, { "grad_norm": 0.19203099608421326, "learning_rate": 1.1434317759922664e-05, "loss": 0.0043, "step": 39570 }, { "grad_norm": 2.625991106033325, "learning_rate": 1.1413279006124528e-05, "loss": 0.0072, "step": 39580 }, { "grad_norm": 0.2345021516084671, "learning_rate": 1.1392257131469119e-05, "loss": 0.0067, "step": 39590 }, { "grad_norm": 0.3266511857509613, "learning_rate": 1.1371252145152095e-05, "loss": 0.0049, "step": 39600 }, { "grad_norm": 0.1981060951948166, "learning_rate": 1.1350264056361776e-05, "loss": 0.0056, "step": 39610 }, { "grad_norm": 0.14928074181079865, "learning_rate": 1.1329292874279029e-05, "loss": 0.0046, "step": 39620 }, { "grad_norm": 0.12456248700618744, "learning_rate": 1.1308338608077401e-05, "loss": 0.0046, "step": 39630 }, { "grad_norm": 0.20438545942306519, "learning_rate": 1.1287401266922981e-05, "loss": 0.0055, "step": 39640 }, { "grad_norm": 0.08679072558879852, "learning_rate": 1.126648085997446e-05, "loss": 0.0044, "step": 39650 }, { "grad_norm": 0.712581217288971, "learning_rate": 1.124557739638316e-05, "loss": 0.0042, "step": 39660 }, { "grad_norm": 0.19501993060112, "learning_rate": 1.1224690885292955e-05, "loss": 0.004, "step": 39670 }, { "grad_norm": 0.4325558841228485, "learning_rate": 1.1203821335840303e-05, "loss": 0.0048, "step": 39680 }, { "grad_norm": 0.1272602677345276, "learning_rate": 1.1182968757154278e-05, "loss": 0.0047, "step": 39690 }, { "grad_norm": 1.3722269535064697, "learning_rate": 1.1162133158356492e-05, "loss": 0.0046, "step": 39700 }, { "grad_norm": 0.26291170716285706, "learning_rate": 1.1141314548561133e-05, "loss": 0.0065, "step": 39710 }, { "grad_norm": 1.225587248802185, "learning_rate": 1.1120512936874994e-05, "loss": 0.0038, "step": 39720 }, { "grad_norm": 0.22682303190231323, "learning_rate": 1.109972833239739e-05, "loss": 0.0056, "step": 39730 }, { "grad_norm": 0.21087646484375, "learning_rate": 1.1078960744220218e-05, "loss": 0.0062, "step": 39740 }, { "grad_norm": 0.15355171263217926, "learning_rate": 1.1058210181427914e-05, "loss": 0.0042, "step": 39750 }, { "grad_norm": 0.1890142560005188, "learning_rate": 1.10374766530975e-05, "loss": 0.0037, "step": 39760 }, { "grad_norm": 0.20967034995555878, "learning_rate": 1.1016760168298518e-05, "loss": 0.0048, "step": 39770 }, { "grad_norm": 0.2650945484638214, "learning_rate": 1.0996060736093044e-05, "loss": 0.0044, "step": 39780 }, { "grad_norm": 0.6226013898849487, "learning_rate": 1.0975378365535749e-05, "loss": 0.0072, "step": 39790 }, { "grad_norm": 0.10900340974330902, "learning_rate": 1.0954713065673777e-05, "loss": 0.0063, "step": 39800 }, { "grad_norm": 0.13616760075092316, "learning_rate": 1.0934064845546832e-05, "loss": 0.0044, "step": 39810 }, { "grad_norm": 0.19329868257045746, "learning_rate": 1.0913433714187166e-05, "loss": 0.0053, "step": 39820 }, { "grad_norm": 0.3518257141113281, "learning_rate": 1.089281968061951e-05, "loss": 0.0044, "step": 39830 }, { "grad_norm": 0.14941182732582092, "learning_rate": 1.0872222753861172e-05, "loss": 0.0053, "step": 39840 }, { "grad_norm": 0.5661031603813171, "learning_rate": 1.0851642942921925e-05, "loss": 0.0057, "step": 39850 }, { "grad_norm": 0.284862220287323, "learning_rate": 1.0831080256804093e-05, "loss": 0.004, "step": 39860 }, { "grad_norm": 0.6072126626968384, "learning_rate": 1.0810534704502478e-05, "loss": 0.0042, "step": 39870 }, { "grad_norm": 0.12107668071985245, "learning_rate": 1.0790006295004423e-05, "loss": 0.0039, "step": 39880 }, { "grad_norm": 0.26597005128860474, "learning_rate": 1.076949503728974e-05, "loss": 0.0053, "step": 39890 }, { "grad_norm": 0.09875788539648056, "learning_rate": 1.074900094033074e-05, "loss": 0.0056, "step": 39900 }, { "grad_norm": 0.1865628957748413, "learning_rate": 1.0728524013092256e-05, "loss": 0.0054, "step": 39910 }, { "grad_norm": 0.13926590979099274, "learning_rate": 1.0708064264531581e-05, "loss": 0.0062, "step": 39920 }, { "grad_norm": 0.2258056253194809, "learning_rate": 1.068762170359851e-05, "loss": 0.0068, "step": 39930 }, { "grad_norm": 0.2243901938199997, "learning_rate": 1.0667196339235297e-05, "loss": 0.0045, "step": 39940 }, { "grad_norm": 0.20346467196941376, "learning_rate": 1.0646788180376716e-05, "loss": 0.0043, "step": 39950 }, { "grad_norm": 2.095757246017456, "learning_rate": 1.0626397235949976e-05, "loss": 0.0061, "step": 39960 }, { "grad_norm": 3.3910012245178223, "learning_rate": 1.0606023514874751e-05, "loss": 0.0046, "step": 39970 }, { "grad_norm": 0.12065401673316956, "learning_rate": 1.0585667026063233e-05, "loss": 0.0054, "step": 39980 }, { "grad_norm": 1.0017447471618652, "learning_rate": 1.0565327778420025e-05, "loss": 0.0033, "step": 39990 }, { "grad_norm": 0.26032161712646484, "learning_rate": 1.0545005780842193e-05, "loss": 0.0035, "step": 40000 } ], "logging_steps": 10, "max_steps": 50000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }