{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.125, "eval_steps": 500, "global_step": 10000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.6393133401870728, "learning_rate": 2.25e-07, "loss": 1.3113, "step": 10 }, { "grad_norm": 1.767778992652893, "learning_rate": 4.75e-07, "loss": 1.3047, "step": 20 }, { "grad_norm": 1.566359519958496, "learning_rate": 7.25e-07, "loss": 1.3004, "step": 30 }, { "grad_norm": 1.3095495700836182, "learning_rate": 9.75e-07, "loss": 1.3043, "step": 40 }, { "grad_norm": 1.3919134140014648, "learning_rate": 1.2250000000000001e-06, "loss": 1.273, "step": 50 }, { "grad_norm": 0.7677862644195557, "learning_rate": 1.475e-06, "loss": 1.2406, "step": 60 }, { "grad_norm": 0.6092354655265808, "learning_rate": 1.7250000000000002e-06, "loss": 1.2496, "step": 70 }, { "grad_norm": 0.5461732149124146, "learning_rate": 1.975e-06, "loss": 1.2238, "step": 80 }, { "grad_norm": 0.42488712072372437, "learning_rate": 2.225e-06, "loss": 1.2281, "step": 90 }, { "grad_norm": 0.3467871844768524, "learning_rate": 2.4750000000000004e-06, "loss": 1.2195, "step": 100 }, { "grad_norm": 0.3987484574317932, "learning_rate": 2.725e-06, "loss": 1.2102, "step": 110 }, { "grad_norm": 0.5055514574050903, "learning_rate": 2.975e-06, "loss": 1.1973, "step": 120 }, { "grad_norm": 0.6315131783485413, "learning_rate": 3.225e-06, "loss": 1.1734, "step": 130 }, { "grad_norm": 0.643717885017395, "learning_rate": 3.4750000000000006e-06, "loss": 1.1187, "step": 140 }, { "grad_norm": 0.44285133481025696, "learning_rate": 3.725e-06, "loss": 1.0898, "step": 150 }, { "grad_norm": 0.42733797430992126, "learning_rate": 3.975e-06, "loss": 1.0934, "step": 160 }, { "grad_norm": 0.6901430487632751, "learning_rate": 4.225e-06, "loss": 1.0867, "step": 170 }, { "grad_norm": 0.5259752869606018, "learning_rate": 4.475e-06, "loss": 1.0836, "step": 180 }, { "grad_norm": 0.4607885479927063, "learning_rate": 4.7250000000000005e-06, "loss": 1.0793, "step": 190 }, { "grad_norm": 0.5349372625350952, "learning_rate": 4.975000000000001e-06, "loss": 1.0633, "step": 200 }, { "grad_norm": 0.5635078549385071, "learning_rate": 5.225e-06, "loss": 1.0684, "step": 210 }, { "grad_norm": 0.6118171215057373, "learning_rate": 5.475e-06, "loss": 1.0648, "step": 220 }, { "grad_norm": 0.5482243299484253, "learning_rate": 5.725e-06, "loss": 1.0539, "step": 230 }, { "grad_norm": 0.4757280945777893, "learning_rate": 5.975e-06, "loss": 1.0434, "step": 240 }, { "grad_norm": 0.5402773022651672, "learning_rate": 6.2250000000000005e-06, "loss": 1.0365, "step": 250 }, { "grad_norm": 0.7668147683143616, "learning_rate": 6.475000000000001e-06, "loss": 1.0406, "step": 260 }, { "grad_norm": 0.5882487893104553, "learning_rate": 6.725000000000001e-06, "loss": 1.043, "step": 270 }, { "grad_norm": 0.5864266157150269, "learning_rate": 6.975000000000001e-06, "loss": 1.0328, "step": 280 }, { "grad_norm": 0.4877640902996063, "learning_rate": 7.2249999999999994e-06, "loss": 1.0434, "step": 290 }, { "grad_norm": 0.5463629364967346, "learning_rate": 7.4750000000000004e-06, "loss": 1.0359, "step": 300 }, { "grad_norm": 0.6088712215423584, "learning_rate": 7.725e-06, "loss": 1.0375, "step": 310 }, { "grad_norm": 0.5386024117469788, "learning_rate": 7.975e-06, "loss": 1.0254, "step": 320 }, { "grad_norm": 0.616011917591095, "learning_rate": 8.225e-06, "loss": 1.0236, "step": 330 }, { "grad_norm": 0.6521902084350586, "learning_rate": 8.475000000000001e-06, "loss": 1.0111, "step": 340 }, { "grad_norm": 1.4101133346557617, "learning_rate": 8.725e-06, "loss": 1.0041, "step": 350 }, { "grad_norm": 1.1839807033538818, "learning_rate": 8.975e-06, "loss": 0.9824, "step": 360 }, { "grad_norm": 1.259799838066101, "learning_rate": 9.225e-06, "loss": 0.9549, "step": 370 }, { "grad_norm": 0.9416484832763672, "learning_rate": 9.475e-06, "loss": 0.9297, "step": 380 }, { "grad_norm": 1.2486101388931274, "learning_rate": 9.725000000000001e-06, "loss": 0.9111, "step": 390 }, { "grad_norm": 0.8262954354286194, "learning_rate": 9.975e-06, "loss": 0.9002, "step": 400 }, { "grad_norm": 1.1016004085540771, "learning_rate": 1.0225e-05, "loss": 0.8861, "step": 410 }, { "grad_norm": 1.1083495616912842, "learning_rate": 1.0475e-05, "loss": 0.8721, "step": 420 }, { "grad_norm": 1.6956247091293335, "learning_rate": 1.0725e-05, "loss": 0.8471, "step": 430 }, { "grad_norm": 1.498782753944397, "learning_rate": 1.0975e-05, "loss": 0.825, "step": 440 }, { "grad_norm": 1.4294670820236206, "learning_rate": 1.1225e-05, "loss": 0.7865, "step": 450 }, { "grad_norm": 1.6284973621368408, "learning_rate": 1.1475000000000001e-05, "loss": 0.773, "step": 460 }, { "grad_norm": 1.474381685256958, "learning_rate": 1.1725e-05, "loss": 0.7494, "step": 470 }, { "grad_norm": 1.5294127464294434, "learning_rate": 1.1975e-05, "loss": 0.733, "step": 480 }, { "grad_norm": 1.500434160232544, "learning_rate": 1.2225e-05, "loss": 0.7084, "step": 490 }, { "grad_norm": 1.5083955526351929, "learning_rate": 1.2475e-05, "loss": 0.6758, "step": 500 }, { "grad_norm": 1.600592017173767, "learning_rate": 1.2725000000000001e-05, "loss": 0.643, "step": 510 }, { "grad_norm": 2.0262179374694824, "learning_rate": 1.2975e-05, "loss": 0.6088, "step": 520 }, { "grad_norm": 2.1777470111846924, "learning_rate": 1.3225000000000001e-05, "loss": 0.5836, "step": 530 }, { "grad_norm": 2.2013309001922607, "learning_rate": 1.3475000000000002e-05, "loss": 0.5537, "step": 540 }, { "grad_norm": 1.9751819372177124, "learning_rate": 1.3725000000000002e-05, "loss": 0.5508, "step": 550 }, { "grad_norm": 2.0503947734832764, "learning_rate": 1.3975000000000003e-05, "loss": 0.5168, "step": 560 }, { "grad_norm": 1.8536869287490845, "learning_rate": 1.4225e-05, "loss": 0.507, "step": 570 }, { "grad_norm": 1.7867927551269531, "learning_rate": 1.4475e-05, "loss": 0.4876, "step": 580 }, { "grad_norm": 1.6938196420669556, "learning_rate": 1.4725e-05, "loss": 0.4615, "step": 590 }, { "grad_norm": 1.598246455192566, "learning_rate": 1.4975e-05, "loss": 0.4629, "step": 600 }, { "grad_norm": 1.6887547969818115, "learning_rate": 1.5225e-05, "loss": 0.434, "step": 610 }, { "grad_norm": 2.1538546085357666, "learning_rate": 1.5475e-05, "loss": 0.4415, "step": 620 }, { "grad_norm": 1.786858320236206, "learning_rate": 1.5725e-05, "loss": 0.3938, "step": 630 }, { "grad_norm": 1.5547300577163696, "learning_rate": 1.5975000000000002e-05, "loss": 0.4062, "step": 640 }, { "grad_norm": 1.5135575532913208, "learning_rate": 1.6225e-05, "loss": 0.3806, "step": 650 }, { "grad_norm": 1.5817131996154785, "learning_rate": 1.6475e-05, "loss": 0.3938, "step": 660 }, { "grad_norm": 1.5664219856262207, "learning_rate": 1.6725000000000003e-05, "loss": 0.3874, "step": 670 }, { "grad_norm": 1.3478021621704102, "learning_rate": 1.6975000000000003e-05, "loss": 0.3566, "step": 680 }, { "grad_norm": 1.7528225183486938, "learning_rate": 1.7225e-05, "loss": 0.3423, "step": 690 }, { "grad_norm": 1.5337364673614502, "learning_rate": 1.7475e-05, "loss": 0.3359, "step": 700 }, { "grad_norm": 2.4852774143218994, "learning_rate": 1.7725e-05, "loss": 0.3325, "step": 710 }, { "grad_norm": 1.831157922744751, "learning_rate": 1.7975e-05, "loss": 0.3099, "step": 720 }, { "grad_norm": 1.6502406597137451, "learning_rate": 1.8225e-05, "loss": 0.3021, "step": 730 }, { "grad_norm": 1.9696483612060547, "learning_rate": 1.8475000000000002e-05, "loss": 0.2734, "step": 740 }, { "grad_norm": 1.6383633613586426, "learning_rate": 1.8725e-05, "loss": 0.2622, "step": 750 }, { "grad_norm": 1.6707582473754883, "learning_rate": 1.8975e-05, "loss": 0.2438, "step": 760 }, { "grad_norm": 1.8254568576812744, "learning_rate": 1.9225e-05, "loss": 0.251, "step": 770 }, { "grad_norm": 2.3416049480438232, "learning_rate": 1.9475000000000002e-05, "loss": 0.226, "step": 780 }, { "grad_norm": 1.5340311527252197, "learning_rate": 1.9725000000000002e-05, "loss": 0.2138, "step": 790 }, { "grad_norm": 1.859876275062561, "learning_rate": 1.9975e-05, "loss": 0.2105, "step": 800 }, { "grad_norm": 2.154973030090332, "learning_rate": 2.0225000000000004e-05, "loss": 0.2118, "step": 810 }, { "grad_norm": 1.8348009586334229, "learning_rate": 2.0475e-05, "loss": 0.1903, "step": 820 }, { "grad_norm": 1.741173505783081, "learning_rate": 2.0725e-05, "loss": 0.1721, "step": 830 }, { "grad_norm": 2.5990681648254395, "learning_rate": 2.0975e-05, "loss": 0.1442, "step": 840 }, { "grad_norm": 1.6081572771072388, "learning_rate": 2.1225e-05, "loss": 0.1418, "step": 850 }, { "grad_norm": 1.6972485780715942, "learning_rate": 2.1475e-05, "loss": 0.1521, "step": 860 }, { "grad_norm": 1.9392932653427124, "learning_rate": 2.1725e-05, "loss": 0.1404, "step": 870 }, { "grad_norm": 1.553594708442688, "learning_rate": 2.1975000000000002e-05, "loss": 0.1287, "step": 880 }, { "grad_norm": 1.8011493682861328, "learning_rate": 2.2225e-05, "loss": 0.1453, "step": 890 }, { "grad_norm": 1.8249744176864624, "learning_rate": 2.2475e-05, "loss": 0.1411, "step": 900 }, { "grad_norm": 2.0196940898895264, "learning_rate": 2.2725000000000003e-05, "loss": 0.1329, "step": 910 }, { "grad_norm": 2.436136245727539, "learning_rate": 2.2975000000000003e-05, "loss": 0.1241, "step": 920 }, { "grad_norm": 1.6055057048797607, "learning_rate": 2.3225000000000002e-05, "loss": 0.1271, "step": 930 }, { "grad_norm": 1.7781250476837158, "learning_rate": 2.3475e-05, "loss": 0.1333, "step": 940 }, { "grad_norm": 1.3514556884765625, "learning_rate": 2.3725e-05, "loss": 0.1302, "step": 950 }, { "grad_norm": 1.6902070045471191, "learning_rate": 2.3975e-05, "loss": 0.1133, "step": 960 }, { "grad_norm": 1.4286205768585205, "learning_rate": 2.4225e-05, "loss": 0.1129, "step": 970 }, { "grad_norm": 1.6462979316711426, "learning_rate": 2.4475000000000002e-05, "loss": 0.1053, "step": 980 }, { "grad_norm": 1.651625633239746, "learning_rate": 2.4725e-05, "loss": 0.103, "step": 990 }, { "grad_norm": 1.5522713661193848, "learning_rate": 2.4975e-05, "loss": 0.1203, "step": 1000 }, { "grad_norm": 1.7168065309524536, "learning_rate": 2.5225e-05, "loss": 0.0971, "step": 1010 }, { "grad_norm": 1.537105679512024, "learning_rate": 2.5475e-05, "loss": 0.108, "step": 1020 }, { "grad_norm": 1.4447249174118042, "learning_rate": 2.5725e-05, "loss": 0.1107, "step": 1030 }, { "grad_norm": 1.6851987838745117, "learning_rate": 2.5974999999999998e-05, "loss": 0.0961, "step": 1040 }, { "grad_norm": 1.5686955451965332, "learning_rate": 2.6225e-05, "loss": 0.0958, "step": 1050 }, { "grad_norm": 1.5350412130355835, "learning_rate": 2.6475e-05, "loss": 0.1059, "step": 1060 }, { "grad_norm": 1.2784905433654785, "learning_rate": 2.6725e-05, "loss": 0.0998, "step": 1070 }, { "grad_norm": 1.1828101873397827, "learning_rate": 2.6975000000000002e-05, "loss": 0.1104, "step": 1080 }, { "grad_norm": 2.0655298233032227, "learning_rate": 2.7225e-05, "loss": 0.1091, "step": 1090 }, { "grad_norm": 1.5553728342056274, "learning_rate": 2.7475e-05, "loss": 0.0874, "step": 1100 }, { "grad_norm": 1.8540327548980713, "learning_rate": 2.7725e-05, "loss": 0.1103, "step": 1110 }, { "grad_norm": 1.1460381746292114, "learning_rate": 2.7975000000000002e-05, "loss": 0.0969, "step": 1120 }, { "grad_norm": 1.5079286098480225, "learning_rate": 2.8225e-05, "loss": 0.1096, "step": 1130 }, { "grad_norm": 1.2856305837631226, "learning_rate": 2.8475e-05, "loss": 0.0974, "step": 1140 }, { "grad_norm": 1.5520336627960205, "learning_rate": 2.8725e-05, "loss": 0.1015, "step": 1150 }, { "grad_norm": 1.449852705001831, "learning_rate": 2.8975000000000003e-05, "loss": 0.0964, "step": 1160 }, { "grad_norm": 1.3053845167160034, "learning_rate": 2.9225000000000002e-05, "loss": 0.0864, "step": 1170 }, { "grad_norm": 1.5947260856628418, "learning_rate": 2.9475e-05, "loss": 0.0917, "step": 1180 }, { "grad_norm": 1.4085862636566162, "learning_rate": 2.9725000000000004e-05, "loss": 0.0885, "step": 1190 }, { "grad_norm": 1.1937754154205322, "learning_rate": 2.9975000000000004e-05, "loss": 0.1008, "step": 1200 }, { "grad_norm": 1.3678373098373413, "learning_rate": 3.0225000000000003e-05, "loss": 0.0835, "step": 1210 }, { "grad_norm": 1.0101574659347534, "learning_rate": 3.0475000000000002e-05, "loss": 0.0907, "step": 1220 }, { "grad_norm": 1.0229358673095703, "learning_rate": 3.0725e-05, "loss": 0.0946, "step": 1230 }, { "grad_norm": 1.5056079626083374, "learning_rate": 3.0975e-05, "loss": 0.0855, "step": 1240 }, { "grad_norm": 1.1308737993240356, "learning_rate": 3.122500000000001e-05, "loss": 0.0853, "step": 1250 }, { "grad_norm": 1.1949784755706787, "learning_rate": 3.1475e-05, "loss": 0.0919, "step": 1260 }, { "grad_norm": 1.4017665386199951, "learning_rate": 3.1725e-05, "loss": 0.0902, "step": 1270 }, { "grad_norm": 1.5458481311798096, "learning_rate": 3.1975e-05, "loss": 0.0817, "step": 1280 }, { "grad_norm": 1.3599543571472168, "learning_rate": 3.2225e-05, "loss": 0.0852, "step": 1290 }, { "grad_norm": 1.450460433959961, "learning_rate": 3.2474999999999997e-05, "loss": 0.0855, "step": 1300 }, { "grad_norm": 1.5782848596572876, "learning_rate": 3.2725e-05, "loss": 0.0871, "step": 1310 }, { "grad_norm": 1.2380833625793457, "learning_rate": 3.2975e-05, "loss": 0.0942, "step": 1320 }, { "grad_norm": 1.3425542116165161, "learning_rate": 3.3225e-05, "loss": 0.0845, "step": 1330 }, { "grad_norm": 1.3883641958236694, "learning_rate": 3.3475e-05, "loss": 0.0867, "step": 1340 }, { "grad_norm": 1.3574200868606567, "learning_rate": 3.3725e-05, "loss": 0.0791, "step": 1350 }, { "grad_norm": 1.2991209030151367, "learning_rate": 3.3975e-05, "loss": 0.0845, "step": 1360 }, { "grad_norm": 1.212193489074707, "learning_rate": 3.4225e-05, "loss": 0.0802, "step": 1370 }, { "grad_norm": 1.2105095386505127, "learning_rate": 3.4475000000000005e-05, "loss": 0.0881, "step": 1380 }, { "grad_norm": 1.204946756362915, "learning_rate": 3.4725000000000004e-05, "loss": 0.0835, "step": 1390 }, { "grad_norm": 1.134838581085205, "learning_rate": 3.4975e-05, "loss": 0.075, "step": 1400 }, { "grad_norm": 1.597717523574829, "learning_rate": 3.5225e-05, "loss": 0.0941, "step": 1410 }, { "grad_norm": 1.120943546295166, "learning_rate": 3.5475e-05, "loss": 0.0957, "step": 1420 }, { "grad_norm": 1.286924123764038, "learning_rate": 3.5725e-05, "loss": 0.0782, "step": 1430 }, { "grad_norm": 1.2821955680847168, "learning_rate": 3.5975e-05, "loss": 0.0877, "step": 1440 }, { "grad_norm": 0.9786968231201172, "learning_rate": 3.6225000000000006e-05, "loss": 0.067, "step": 1450 }, { "grad_norm": 1.1541035175323486, "learning_rate": 3.6475000000000006e-05, "loss": 0.0731, "step": 1460 }, { "grad_norm": 1.0547113418579102, "learning_rate": 3.6725000000000005e-05, "loss": 0.0762, "step": 1470 }, { "grad_norm": 1.0338844060897827, "learning_rate": 3.6975000000000004e-05, "loss": 0.0862, "step": 1480 }, { "grad_norm": 0.9127304553985596, "learning_rate": 3.7225000000000004e-05, "loss": 0.0803, "step": 1490 }, { "grad_norm": 0.9178349375724792, "learning_rate": 3.7475e-05, "loss": 0.0832, "step": 1500 }, { "grad_norm": 1.0365991592407227, "learning_rate": 3.7725e-05, "loss": 0.0879, "step": 1510 }, { "grad_norm": 1.3525325059890747, "learning_rate": 3.7975e-05, "loss": 0.0891, "step": 1520 }, { "grad_norm": 1.3538330793380737, "learning_rate": 3.8225e-05, "loss": 0.0869, "step": 1530 }, { "grad_norm": 1.2002805471420288, "learning_rate": 3.8475e-05, "loss": 0.0745, "step": 1540 }, { "grad_norm": 1.0287679433822632, "learning_rate": 3.8725e-05, "loss": 0.08, "step": 1550 }, { "grad_norm": 1.310381531715393, "learning_rate": 3.8975e-05, "loss": 0.0789, "step": 1560 }, { "grad_norm": 0.927413821220398, "learning_rate": 3.9225e-05, "loss": 0.0847, "step": 1570 }, { "grad_norm": 0.945735514163971, "learning_rate": 3.9475000000000004e-05, "loss": 0.0823, "step": 1580 }, { "grad_norm": 1.4773881435394287, "learning_rate": 3.9725e-05, "loss": 0.0773, "step": 1590 }, { "grad_norm": 0.899528980255127, "learning_rate": 3.9975e-05, "loss": 0.0829, "step": 1600 }, { "grad_norm": 1.000025987625122, "learning_rate": 4.0225e-05, "loss": 0.0823, "step": 1610 }, { "grad_norm": 0.9304444789886475, "learning_rate": 4.0475e-05, "loss": 0.0722, "step": 1620 }, { "grad_norm": 1.2848438024520874, "learning_rate": 4.0725e-05, "loss": 0.0785, "step": 1630 }, { "grad_norm": 1.2650846242904663, "learning_rate": 4.0975e-05, "loss": 0.0748, "step": 1640 }, { "grad_norm": 1.0591596364974976, "learning_rate": 4.1225e-05, "loss": 0.0711, "step": 1650 }, { "grad_norm": 1.0950254201889038, "learning_rate": 4.1475000000000005e-05, "loss": 0.0797, "step": 1660 }, { "grad_norm": 1.232553243637085, "learning_rate": 4.1725000000000005e-05, "loss": 0.0793, "step": 1670 }, { "grad_norm": 1.0546668767929077, "learning_rate": 4.1975000000000004e-05, "loss": 0.0699, "step": 1680 }, { "grad_norm": 1.008373498916626, "learning_rate": 4.2225e-05, "loss": 0.0808, "step": 1690 }, { "grad_norm": 0.8450912833213806, "learning_rate": 4.2475e-05, "loss": 0.0731, "step": 1700 }, { "grad_norm": 1.1074550151824951, "learning_rate": 4.2725e-05, "loss": 0.0867, "step": 1710 }, { "grad_norm": 1.2273873090744019, "learning_rate": 4.2975e-05, "loss": 0.0782, "step": 1720 }, { "grad_norm": 1.0400255918502808, "learning_rate": 4.322500000000001e-05, "loss": 0.078, "step": 1730 }, { "grad_norm": 1.2490166425704956, "learning_rate": 4.3475000000000006e-05, "loss": 0.07, "step": 1740 }, { "grad_norm": 1.0737584829330444, "learning_rate": 4.3725000000000006e-05, "loss": 0.075, "step": 1750 }, { "grad_norm": 1.233413577079773, "learning_rate": 4.3975e-05, "loss": 0.0741, "step": 1760 }, { "grad_norm": 0.9239450097084045, "learning_rate": 4.4225e-05, "loss": 0.0663, "step": 1770 }, { "grad_norm": 1.1301162242889404, "learning_rate": 4.4475e-05, "loss": 0.0733, "step": 1780 }, { "grad_norm": 0.8251295685768127, "learning_rate": 4.4725e-05, "loss": 0.064, "step": 1790 }, { "grad_norm": 1.0031894445419312, "learning_rate": 4.4975e-05, "loss": 0.0669, "step": 1800 }, { "grad_norm": 0.8335981965065002, "learning_rate": 4.5225e-05, "loss": 0.064, "step": 1810 }, { "grad_norm": 0.8685722351074219, "learning_rate": 4.5475e-05, "loss": 0.0672, "step": 1820 }, { "grad_norm": 0.8541990518569946, "learning_rate": 4.5725e-05, "loss": 0.0602, "step": 1830 }, { "grad_norm": 0.8233271241188049, "learning_rate": 4.5975e-05, "loss": 0.0682, "step": 1840 }, { "grad_norm": 0.7718712687492371, "learning_rate": 4.6225e-05, "loss": 0.0651, "step": 1850 }, { "grad_norm": 1.4902597665786743, "learning_rate": 4.6475000000000005e-05, "loss": 0.073, "step": 1860 }, { "grad_norm": 0.8875091075897217, "learning_rate": 4.6725000000000004e-05, "loss": 0.0685, "step": 1870 }, { "grad_norm": 0.9251996874809265, "learning_rate": 4.6975000000000003e-05, "loss": 0.0613, "step": 1880 }, { "grad_norm": 1.170397400856018, "learning_rate": 4.7225e-05, "loss": 0.0711, "step": 1890 }, { "grad_norm": 0.8567249178886414, "learning_rate": 4.7475e-05, "loss": 0.0721, "step": 1900 }, { "grad_norm": 1.0828672647476196, "learning_rate": 4.7725e-05, "loss": 0.069, "step": 1910 }, { "grad_norm": 0.8799436688423157, "learning_rate": 4.7975e-05, "loss": 0.07, "step": 1920 }, { "grad_norm": 0.949116587638855, "learning_rate": 4.822500000000001e-05, "loss": 0.069, "step": 1930 }, { "grad_norm": 1.0246304273605347, "learning_rate": 4.8475000000000006e-05, "loss": 0.0693, "step": 1940 }, { "grad_norm": 0.8899738788604736, "learning_rate": 4.8725000000000005e-05, "loss": 0.0641, "step": 1950 }, { "grad_norm": 1.0466159582138062, "learning_rate": 4.8975000000000005e-05, "loss": 0.0752, "step": 1960 }, { "grad_norm": 0.899034321308136, "learning_rate": 4.9225000000000004e-05, "loss": 0.0803, "step": 1970 }, { "grad_norm": 0.8415388464927673, "learning_rate": 4.9475e-05, "loss": 0.0648, "step": 1980 }, { "grad_norm": 1.0137670040130615, "learning_rate": 4.9725e-05, "loss": 0.0651, "step": 1990 }, { "grad_norm": 0.8400244116783142, "learning_rate": 4.9975e-05, "loss": 0.0838, "step": 2000 }, { "grad_norm": 1.078927993774414, "learning_rate": 5.0225e-05, "loss": 0.0692, "step": 2010 }, { "grad_norm": 0.7297987937927246, "learning_rate": 5.047500000000001e-05, "loss": 0.0658, "step": 2020 }, { "grad_norm": 0.7978940606117249, "learning_rate": 5.0725e-05, "loss": 0.0704, "step": 2030 }, { "grad_norm": 0.7656873464584351, "learning_rate": 5.0975000000000006e-05, "loss": 0.0653, "step": 2040 }, { "grad_norm": 0.7548262476921082, "learning_rate": 5.1225e-05, "loss": 0.0696, "step": 2050 }, { "grad_norm": 1.0591721534729004, "learning_rate": 5.1475000000000004e-05, "loss": 0.0786, "step": 2060 }, { "grad_norm": 0.7878709435462952, "learning_rate": 5.1725000000000004e-05, "loss": 0.0698, "step": 2070 }, { "grad_norm": 0.897945761680603, "learning_rate": 5.197500000000001e-05, "loss": 0.0628, "step": 2080 }, { "grad_norm": 0.6967459917068481, "learning_rate": 5.2225e-05, "loss": 0.0612, "step": 2090 }, { "grad_norm": 0.8716256022453308, "learning_rate": 5.247500000000001e-05, "loss": 0.061, "step": 2100 }, { "grad_norm": 0.6177629828453064, "learning_rate": 5.2725e-05, "loss": 0.0569, "step": 2110 }, { "grad_norm": 0.9584032893180847, "learning_rate": 5.297500000000001e-05, "loss": 0.0667, "step": 2120 }, { "grad_norm": 0.9155662655830383, "learning_rate": 5.3225e-05, "loss": 0.0634, "step": 2130 }, { "grad_norm": 0.7499725222587585, "learning_rate": 5.3475e-05, "loss": 0.062, "step": 2140 }, { "grad_norm": 0.5990448594093323, "learning_rate": 5.3725000000000005e-05, "loss": 0.0594, "step": 2150 }, { "grad_norm": 0.8445324897766113, "learning_rate": 5.3975e-05, "loss": 0.0552, "step": 2160 }, { "grad_norm": 0.6003220677375793, "learning_rate": 5.4225000000000003e-05, "loss": 0.0584, "step": 2170 }, { "grad_norm": 1.0316931009292603, "learning_rate": 5.4474999999999996e-05, "loss": 0.0637, "step": 2180 }, { "grad_norm": 0.868081271648407, "learning_rate": 5.4725e-05, "loss": 0.0649, "step": 2190 }, { "grad_norm": 0.842610776424408, "learning_rate": 5.4975e-05, "loss": 0.0667, "step": 2200 }, { "grad_norm": 0.6579468250274658, "learning_rate": 5.522500000000001e-05, "loss": 0.0532, "step": 2210 }, { "grad_norm": 0.8250467777252197, "learning_rate": 5.5475e-05, "loss": 0.0577, "step": 2220 }, { "grad_norm": 0.882590115070343, "learning_rate": 5.5725000000000006e-05, "loss": 0.0618, "step": 2230 }, { "grad_norm": 0.6348373889923096, "learning_rate": 5.5975e-05, "loss": 0.0633, "step": 2240 }, { "grad_norm": 0.6598455309867859, "learning_rate": 5.6225000000000005e-05, "loss": 0.0693, "step": 2250 }, { "grad_norm": 0.8272908329963684, "learning_rate": 5.6475e-05, "loss": 0.0597, "step": 2260 }, { "grad_norm": 0.8021649718284607, "learning_rate": 5.6725e-05, "loss": 0.0551, "step": 2270 }, { "grad_norm": 0.7743063569068909, "learning_rate": 5.6975e-05, "loss": 0.0649, "step": 2280 }, { "grad_norm": 0.9049435257911682, "learning_rate": 5.722500000000001e-05, "loss": 0.0624, "step": 2290 }, { "grad_norm": 0.7708112597465515, "learning_rate": 5.7475e-05, "loss": 0.0647, "step": 2300 }, { "grad_norm": 0.9157259464263916, "learning_rate": 5.772500000000001e-05, "loss": 0.0642, "step": 2310 }, { "grad_norm": 0.7489880919456482, "learning_rate": 5.7975e-05, "loss": 0.0528, "step": 2320 }, { "grad_norm": 0.7524044513702393, "learning_rate": 5.8225000000000006e-05, "loss": 0.0594, "step": 2330 }, { "grad_norm": 0.9588838219642639, "learning_rate": 5.8475000000000005e-05, "loss": 0.0608, "step": 2340 }, { "grad_norm": 0.724116861820221, "learning_rate": 5.8725000000000004e-05, "loss": 0.0608, "step": 2350 }, { "grad_norm": 0.949410617351532, "learning_rate": 5.8975000000000004e-05, "loss": 0.0629, "step": 2360 }, { "grad_norm": 0.7337884306907654, "learning_rate": 5.922500000000001e-05, "loss": 0.0644, "step": 2370 }, { "grad_norm": 0.6380131244659424, "learning_rate": 5.9475e-05, "loss": 0.0605, "step": 2380 }, { "grad_norm": 0.6566464304924011, "learning_rate": 5.9724999999999995e-05, "loss": 0.0703, "step": 2390 }, { "grad_norm": 0.7548853158950806, "learning_rate": 5.9975e-05, "loss": 0.061, "step": 2400 }, { "grad_norm": 0.7135842442512512, "learning_rate": 6.0225e-05, "loss": 0.0594, "step": 2410 }, { "grad_norm": 0.6651610136032104, "learning_rate": 6.0475000000000006e-05, "loss": 0.0587, "step": 2420 }, { "grad_norm": 0.5913328528404236, "learning_rate": 6.0725e-05, "loss": 0.0555, "step": 2430 }, { "grad_norm": 0.6769335865974426, "learning_rate": 6.0975000000000005e-05, "loss": 0.0532, "step": 2440 }, { "grad_norm": 0.7440420985221863, "learning_rate": 6.1225e-05, "loss": 0.0555, "step": 2450 }, { "grad_norm": 0.5355431437492371, "learning_rate": 6.1475e-05, "loss": 0.0542, "step": 2460 }, { "grad_norm": 0.6097579598426819, "learning_rate": 6.1725e-05, "loss": 0.056, "step": 2470 }, { "grad_norm": 0.6164148449897766, "learning_rate": 6.1975e-05, "loss": 0.0638, "step": 2480 }, { "grad_norm": 0.8228153586387634, "learning_rate": 6.2225e-05, "loss": 0.0597, "step": 2490 }, { "grad_norm": 0.693198025226593, "learning_rate": 6.2475e-05, "loss": 0.0616, "step": 2500 }, { "grad_norm": 0.77430260181427, "learning_rate": 6.2725e-05, "loss": 0.048, "step": 2510 }, { "grad_norm": 0.7404604554176331, "learning_rate": 6.297500000000001e-05, "loss": 0.0541, "step": 2520 }, { "grad_norm": 0.7104705572128296, "learning_rate": 6.3225e-05, "loss": 0.0651, "step": 2530 }, { "grad_norm": 0.7111170887947083, "learning_rate": 6.347500000000001e-05, "loss": 0.0648, "step": 2540 }, { "grad_norm": 0.7771560549736023, "learning_rate": 6.3725e-05, "loss": 0.0578, "step": 2550 }, { "grad_norm": 0.7633046507835388, "learning_rate": 6.397500000000001e-05, "loss": 0.0579, "step": 2560 }, { "grad_norm": 0.8382501006126404, "learning_rate": 6.4225e-05, "loss": 0.0568, "step": 2570 }, { "grad_norm": 0.6458725929260254, "learning_rate": 6.447500000000001e-05, "loss": 0.0602, "step": 2580 }, { "grad_norm": 0.6697157621383667, "learning_rate": 6.4725e-05, "loss": 0.0563, "step": 2590 }, { "grad_norm": 0.6784992814064026, "learning_rate": 6.497500000000001e-05, "loss": 0.0558, "step": 2600 }, { "grad_norm": 0.6799176931381226, "learning_rate": 6.5225e-05, "loss": 0.0486, "step": 2610 }, { "grad_norm": 0.7377599477767944, "learning_rate": 6.5475e-05, "loss": 0.053, "step": 2620 }, { "grad_norm": 0.5490700602531433, "learning_rate": 6.5725e-05, "loss": 0.06, "step": 2630 }, { "grad_norm": 0.5960026979446411, "learning_rate": 6.5975e-05, "loss": 0.0747, "step": 2640 }, { "grad_norm": 0.6861433982849121, "learning_rate": 6.6225e-05, "loss": 0.0554, "step": 2650 }, { "grad_norm": 0.611458420753479, "learning_rate": 6.6475e-05, "loss": 0.0542, "step": 2660 }, { "grad_norm": 0.712469220161438, "learning_rate": 6.672500000000001e-05, "loss": 0.0484, "step": 2670 }, { "grad_norm": 0.6568604707717896, "learning_rate": 6.6975e-05, "loss": 0.0484, "step": 2680 }, { "grad_norm": 0.588529109954834, "learning_rate": 6.722500000000001e-05, "loss": 0.0515, "step": 2690 }, { "grad_norm": 0.9409171342849731, "learning_rate": 6.7475e-05, "loss": 0.0644, "step": 2700 }, { "grad_norm": 0.5707340240478516, "learning_rate": 6.7725e-05, "loss": 0.0612, "step": 2710 }, { "grad_norm": 0.7319713830947876, "learning_rate": 6.7975e-05, "loss": 0.0579, "step": 2720 }, { "grad_norm": 0.7075655460357666, "learning_rate": 6.8225e-05, "loss": 0.0561, "step": 2730 }, { "grad_norm": 0.6792818307876587, "learning_rate": 6.8475e-05, "loss": 0.0509, "step": 2740 }, { "grad_norm": 0.7108269333839417, "learning_rate": 6.8725e-05, "loss": 0.0531, "step": 2750 }, { "grad_norm": 0.5372207164764404, "learning_rate": 6.8975e-05, "loss": 0.0493, "step": 2760 }, { "grad_norm": 0.6156140565872192, "learning_rate": 6.9225e-05, "loss": 0.049, "step": 2770 }, { "grad_norm": 0.5341912508010864, "learning_rate": 6.9475e-05, "loss": 0.0438, "step": 2780 }, { "grad_norm": 0.5590257048606873, "learning_rate": 6.9725e-05, "loss": 0.0509, "step": 2790 }, { "grad_norm": 0.7167447209358215, "learning_rate": 6.997500000000001e-05, "loss": 0.0487, "step": 2800 }, { "grad_norm": 0.5477861166000366, "learning_rate": 7.022500000000001e-05, "loss": 0.0484, "step": 2810 }, { "grad_norm": 0.7099193930625916, "learning_rate": 7.0475e-05, "loss": 0.0484, "step": 2820 }, { "grad_norm": 0.815622866153717, "learning_rate": 7.072500000000001e-05, "loss": 0.0493, "step": 2830 }, { "grad_norm": 0.5379012227058411, "learning_rate": 7.0975e-05, "loss": 0.0535, "step": 2840 }, { "grad_norm": 0.4705035388469696, "learning_rate": 7.122500000000001e-05, "loss": 0.0527, "step": 2850 }, { "grad_norm": 0.5898609161376953, "learning_rate": 7.1475e-05, "loss": 0.0677, "step": 2860 }, { "grad_norm": 0.6316633224487305, "learning_rate": 7.172500000000001e-05, "loss": 0.0474, "step": 2870 }, { "grad_norm": 0.5751751661300659, "learning_rate": 7.1975e-05, "loss": 0.0519, "step": 2880 }, { "grad_norm": 0.6925854086875916, "learning_rate": 7.2225e-05, "loss": 0.0532, "step": 2890 }, { "grad_norm": 0.6490530371665955, "learning_rate": 7.2475e-05, "loss": 0.0431, "step": 2900 }, { "grad_norm": 0.5243796706199646, "learning_rate": 7.272499999999999e-05, "loss": 0.0594, "step": 2910 }, { "grad_norm": 0.4794312119483948, "learning_rate": 7.2975e-05, "loss": 0.0491, "step": 2920 }, { "grad_norm": 0.5939675569534302, "learning_rate": 7.3225e-05, "loss": 0.053, "step": 2930 }, { "grad_norm": 0.6829997301101685, "learning_rate": 7.347500000000001e-05, "loss": 0.0433, "step": 2940 }, { "grad_norm": 0.5756387710571289, "learning_rate": 7.3725e-05, "loss": 0.0453, "step": 2950 }, { "grad_norm": 0.6272783279418945, "learning_rate": 7.397500000000001e-05, "loss": 0.0495, "step": 2960 }, { "grad_norm": 0.6649883985519409, "learning_rate": 7.4225e-05, "loss": 0.0516, "step": 2970 }, { "grad_norm": 0.41673851013183594, "learning_rate": 7.447500000000001e-05, "loss": 0.0555, "step": 2980 }, { "grad_norm": 0.43925657868385315, "learning_rate": 7.4725e-05, "loss": 0.0422, "step": 2990 }, { "grad_norm": 0.5183777213096619, "learning_rate": 7.4975e-05, "loss": 0.0581, "step": 3000 }, { "grad_norm": 0.6291593909263611, "learning_rate": 7.5225e-05, "loss": 0.0502, "step": 3010 }, { "grad_norm": 0.6234071850776672, "learning_rate": 7.5475e-05, "loss": 0.0588, "step": 3020 }, { "grad_norm": 0.5591736435890198, "learning_rate": 7.5725e-05, "loss": 0.0491, "step": 3030 }, { "grad_norm": 0.570741593837738, "learning_rate": 7.5975e-05, "loss": 0.0546, "step": 3040 }, { "grad_norm": 0.6032003164291382, "learning_rate": 7.6225e-05, "loss": 0.0545, "step": 3050 }, { "grad_norm": 0.7455132007598877, "learning_rate": 7.6475e-05, "loss": 0.043, "step": 3060 }, { "grad_norm": 0.6484450697898865, "learning_rate": 7.672500000000001e-05, "loss": 0.0511, "step": 3070 }, { "grad_norm": 0.557806670665741, "learning_rate": 7.697500000000001e-05, "loss": 0.0549, "step": 3080 }, { "grad_norm": 0.7911346554756165, "learning_rate": 7.722500000000001e-05, "loss": 0.0447, "step": 3090 }, { "grad_norm": 0.5621020197868347, "learning_rate": 7.747500000000001e-05, "loss": 0.0545, "step": 3100 }, { "grad_norm": 0.7302330732345581, "learning_rate": 7.7725e-05, "loss": 0.0541, "step": 3110 }, { "grad_norm": 0.5450969934463501, "learning_rate": 7.797500000000001e-05, "loss": 0.0461, "step": 3120 }, { "grad_norm": 0.5978328585624695, "learning_rate": 7.8225e-05, "loss": 0.0542, "step": 3130 }, { "grad_norm": 0.6991857290267944, "learning_rate": 7.8475e-05, "loss": 0.0545, "step": 3140 }, { "grad_norm": 0.5713250041007996, "learning_rate": 7.8725e-05, "loss": 0.0429, "step": 3150 }, { "grad_norm": 0.5079202055931091, "learning_rate": 7.8975e-05, "loss": 0.0516, "step": 3160 }, { "grad_norm": 0.5190696716308594, "learning_rate": 7.9225e-05, "loss": 0.0448, "step": 3170 }, { "grad_norm": 0.6884950995445251, "learning_rate": 7.9475e-05, "loss": 0.0554, "step": 3180 }, { "grad_norm": 0.5221415162086487, "learning_rate": 7.9725e-05, "loss": 0.0573, "step": 3190 }, { "grad_norm": 0.4776185154914856, "learning_rate": 7.9975e-05, "loss": 0.0483, "step": 3200 }, { "grad_norm": 0.5168715119361877, "learning_rate": 8.022500000000001e-05, "loss": 0.0475, "step": 3210 }, { "grad_norm": 0.8133692741394043, "learning_rate": 8.0475e-05, "loss": 0.0501, "step": 3220 }, { "grad_norm": 0.5876872539520264, "learning_rate": 8.072500000000001e-05, "loss": 0.0498, "step": 3230 }, { "grad_norm": 0.5368916988372803, "learning_rate": 8.0975e-05, "loss": 0.0475, "step": 3240 }, { "grad_norm": 0.6654589176177979, "learning_rate": 8.122500000000001e-05, "loss": 0.0533, "step": 3250 }, { "grad_norm": 0.7075525522232056, "learning_rate": 8.1475e-05, "loss": 0.0506, "step": 3260 }, { "grad_norm": 0.7137773633003235, "learning_rate": 8.172500000000001e-05, "loss": 0.0473, "step": 3270 }, { "grad_norm": 0.6145274639129639, "learning_rate": 8.1975e-05, "loss": 0.053, "step": 3280 }, { "grad_norm": 0.5827512741088867, "learning_rate": 8.2225e-05, "loss": 0.0467, "step": 3290 }, { "grad_norm": 0.5741691589355469, "learning_rate": 8.2475e-05, "loss": 0.0477, "step": 3300 }, { "grad_norm": 0.6415374279022217, "learning_rate": 8.2725e-05, "loss": 0.0502, "step": 3310 }, { "grad_norm": 0.5508673787117004, "learning_rate": 8.2975e-05, "loss": 0.0436, "step": 3320 }, { "grad_norm": 0.5631225109100342, "learning_rate": 8.3225e-05, "loss": 0.051, "step": 3330 }, { "grad_norm": 0.6855165958404541, "learning_rate": 8.347500000000001e-05, "loss": 0.056, "step": 3340 }, { "grad_norm": 0.6133655309677124, "learning_rate": 8.3725e-05, "loss": 0.0498, "step": 3350 }, { "grad_norm": 0.5718126893043518, "learning_rate": 8.397500000000001e-05, "loss": 0.0472, "step": 3360 }, { "grad_norm": 0.6181097030639648, "learning_rate": 8.422500000000001e-05, "loss": 0.0448, "step": 3370 }, { "grad_norm": 0.6881830096244812, "learning_rate": 8.447500000000001e-05, "loss": 0.0492, "step": 3380 }, { "grad_norm": 0.7708672285079956, "learning_rate": 8.4725e-05, "loss": 0.0552, "step": 3390 }, { "grad_norm": 0.5261228680610657, "learning_rate": 8.4975e-05, "loss": 0.0661, "step": 3400 }, { "grad_norm": 0.532979428768158, "learning_rate": 8.5225e-05, "loss": 0.0533, "step": 3410 }, { "grad_norm": 0.4041130542755127, "learning_rate": 8.5475e-05, "loss": 0.0474, "step": 3420 }, { "grad_norm": 0.4805317521095276, "learning_rate": 8.5725e-05, "loss": 0.0542, "step": 3430 }, { "grad_norm": 0.4866563677787781, "learning_rate": 8.5975e-05, "loss": 0.0486, "step": 3440 }, { "grad_norm": 0.5513153076171875, "learning_rate": 8.6225e-05, "loss": 0.0534, "step": 3450 }, { "grad_norm": 0.5657448768615723, "learning_rate": 8.6475e-05, "loss": 0.0453, "step": 3460 }, { "grad_norm": 0.48526906967163086, "learning_rate": 8.672500000000001e-05, "loss": 0.0466, "step": 3470 }, { "grad_norm": 0.4409540891647339, "learning_rate": 8.6975e-05, "loss": 0.042, "step": 3480 }, { "grad_norm": 0.4810849130153656, "learning_rate": 8.7225e-05, "loss": 0.0509, "step": 3490 }, { "grad_norm": 0.5116516351699829, "learning_rate": 8.747500000000001e-05, "loss": 0.0412, "step": 3500 }, { "grad_norm": 0.39224880933761597, "learning_rate": 8.7725e-05, "loss": 0.049, "step": 3510 }, { "grad_norm": 0.44914278388023376, "learning_rate": 8.797500000000001e-05, "loss": 0.0394, "step": 3520 }, { "grad_norm": 0.6105808615684509, "learning_rate": 8.8225e-05, "loss": 0.0458, "step": 3530 }, { "grad_norm": 0.617580771446228, "learning_rate": 8.847500000000001e-05, "loss": 0.0645, "step": 3540 }, { "grad_norm": 0.5933853983879089, "learning_rate": 8.8725e-05, "loss": 0.0527, "step": 3550 }, { "grad_norm": 0.3423856794834137, "learning_rate": 8.897500000000001e-05, "loss": 0.0498, "step": 3560 }, { "grad_norm": 0.46960392594337463, "learning_rate": 8.9225e-05, "loss": 0.0574, "step": 3570 }, { "grad_norm": 0.579044759273529, "learning_rate": 8.9475e-05, "loss": 0.0453, "step": 3580 }, { "grad_norm": 0.6561214327812195, "learning_rate": 8.9725e-05, "loss": 0.0544, "step": 3590 }, { "grad_norm": 0.4810267388820648, "learning_rate": 8.9975e-05, "loss": 0.0533, "step": 3600 }, { "grad_norm": 0.5214345455169678, "learning_rate": 9.0225e-05, "loss": 0.0474, "step": 3610 }, { "grad_norm": 0.529765784740448, "learning_rate": 9.0475e-05, "loss": 0.0483, "step": 3620 }, { "grad_norm": 0.47406721115112305, "learning_rate": 9.072500000000001e-05, "loss": 0.05, "step": 3630 }, { "grad_norm": 0.5467729568481445, "learning_rate": 9.0975e-05, "loss": 0.0484, "step": 3640 }, { "grad_norm": 0.37928155064582825, "learning_rate": 9.122500000000001e-05, "loss": 0.0508, "step": 3650 }, { "grad_norm": 0.5016357898712158, "learning_rate": 9.1475e-05, "loss": 0.0396, "step": 3660 }, { "grad_norm": 0.5030633211135864, "learning_rate": 9.172500000000001e-05, "loss": 0.0413, "step": 3670 }, { "grad_norm": 0.5383405089378357, "learning_rate": 9.1975e-05, "loss": 0.0457, "step": 3680 }, { "grad_norm": 0.5056612491607666, "learning_rate": 9.2225e-05, "loss": 0.0464, "step": 3690 }, { "grad_norm": 0.5196707844734192, "learning_rate": 9.2475e-05, "loss": 0.0467, "step": 3700 }, { "grad_norm": 0.46504533290863037, "learning_rate": 9.2725e-05, "loss": 0.0499, "step": 3710 }, { "grad_norm": 0.5482833981513977, "learning_rate": 9.2975e-05, "loss": 0.0412, "step": 3720 }, { "grad_norm": 0.4397702217102051, "learning_rate": 9.3225e-05, "loss": 0.0462, "step": 3730 }, { "grad_norm": 0.5846696496009827, "learning_rate": 9.3475e-05, "loss": 0.047, "step": 3740 }, { "grad_norm": 0.4426489770412445, "learning_rate": 9.3725e-05, "loss": 0.0455, "step": 3750 }, { "grad_norm": 0.4533824920654297, "learning_rate": 9.397500000000001e-05, "loss": 0.0442, "step": 3760 }, { "grad_norm": 0.5120921730995178, "learning_rate": 9.422500000000001e-05, "loss": 0.0446, "step": 3770 }, { "grad_norm": 0.5631505250930786, "learning_rate": 9.4475e-05, "loss": 0.0445, "step": 3780 }, { "grad_norm": 0.41061311960220337, "learning_rate": 9.472500000000001e-05, "loss": 0.0476, "step": 3790 }, { "grad_norm": 0.5548518300056458, "learning_rate": 9.4975e-05, "loss": 0.0504, "step": 3800 }, { "grad_norm": 0.4881528317928314, "learning_rate": 9.522500000000001e-05, "loss": 0.0465, "step": 3810 }, { "grad_norm": 0.4558788239955902, "learning_rate": 9.5475e-05, "loss": 0.0457, "step": 3820 }, { "grad_norm": 0.5413118600845337, "learning_rate": 9.572500000000001e-05, "loss": 0.0432, "step": 3830 }, { "grad_norm": 0.46621590852737427, "learning_rate": 9.5975e-05, "loss": 0.04, "step": 3840 }, { "grad_norm": 0.46049967408180237, "learning_rate": 9.622500000000001e-05, "loss": 0.0424, "step": 3850 }, { "grad_norm": 0.5652264952659607, "learning_rate": 9.6475e-05, "loss": 0.0449, "step": 3860 }, { "grad_norm": 0.4800361692905426, "learning_rate": 9.6725e-05, "loss": 0.0527, "step": 3870 }, { "grad_norm": 0.43115052580833435, "learning_rate": 9.6975e-05, "loss": 0.0583, "step": 3880 }, { "grad_norm": 0.4887556731700897, "learning_rate": 9.7225e-05, "loss": 0.0444, "step": 3890 }, { "grad_norm": 0.6786840558052063, "learning_rate": 9.747500000000001e-05, "loss": 0.0459, "step": 3900 }, { "grad_norm": 0.5514159798622131, "learning_rate": 9.7725e-05, "loss": 0.044, "step": 3910 }, { "grad_norm": 0.44103673100471497, "learning_rate": 9.797500000000001e-05, "loss": 0.047, "step": 3920 }, { "grad_norm": 0.5125502347946167, "learning_rate": 9.8225e-05, "loss": 0.043, "step": 3930 }, { "grad_norm": 0.5543313026428223, "learning_rate": 9.847500000000001e-05, "loss": 0.0526, "step": 3940 }, { "grad_norm": 0.5096784234046936, "learning_rate": 9.8725e-05, "loss": 0.0499, "step": 3950 }, { "grad_norm": 0.49835607409477234, "learning_rate": 9.897500000000001e-05, "loss": 0.0516, "step": 3960 }, { "grad_norm": 0.3593589663505554, "learning_rate": 9.9225e-05, "loss": 0.0446, "step": 3970 }, { "grad_norm": 0.46587103605270386, "learning_rate": 9.9475e-05, "loss": 0.0418, "step": 3980 }, { "grad_norm": 0.4543410539627075, "learning_rate": 9.9725e-05, "loss": 0.0391, "step": 3990 }, { "grad_norm": 0.38223424553871155, "learning_rate": 9.9975e-05, "loss": 0.0473, "step": 4000 }, { "grad_norm": 0.43263593316078186, "learning_rate": 9.999999653982884e-05, "loss": 0.0451, "step": 4010 }, { "grad_norm": 0.3665847182273865, "learning_rate": 9.999998457874392e-05, "loss": 0.0504, "step": 4020 }, { "grad_norm": 0.35126611590385437, "learning_rate": 9.999996407402913e-05, "loss": 0.0505, "step": 4030 }, { "grad_norm": 0.6602134704589844, "learning_rate": 9.999993502568801e-05, "loss": 0.0477, "step": 4040 }, { "grad_norm": 0.45608222484588623, "learning_rate": 9.999989743372548e-05, "loss": 0.0422, "step": 4050 }, { "grad_norm": 0.45705121755599976, "learning_rate": 9.999985129814798e-05, "loss": 0.0446, "step": 4060 }, { "grad_norm": 0.41707882285118103, "learning_rate": 9.99997966189634e-05, "loss": 0.0526, "step": 4070 }, { "grad_norm": 0.6251623034477234, "learning_rate": 9.999973339618107e-05, "loss": 0.0413, "step": 4080 }, { "grad_norm": 0.47000041604042053, "learning_rate": 9.999966162981179e-05, "loss": 0.0547, "step": 4090 }, { "grad_norm": 0.4268397390842438, "learning_rate": 9.999958131986784e-05, "loss": 0.0435, "step": 4100 }, { "grad_norm": 0.4606395363807678, "learning_rate": 9.999949246636293e-05, "loss": 0.0441, "step": 4110 }, { "grad_norm": 0.6380859613418579, "learning_rate": 9.999939506931224e-05, "loss": 0.0498, "step": 4120 }, { "grad_norm": 0.506939709186554, "learning_rate": 9.999928912873243e-05, "loss": 0.0454, "step": 4130 }, { "grad_norm": 0.41200029850006104, "learning_rate": 9.999917464464159e-05, "loss": 0.0476, "step": 4140 }, { "grad_norm": 0.3817211389541626, "learning_rate": 9.999905161705929e-05, "loss": 0.0516, "step": 4150 }, { "grad_norm": 0.32852673530578613, "learning_rate": 9.999892004600653e-05, "loss": 0.04, "step": 4160 }, { "grad_norm": 0.45431870222091675, "learning_rate": 9.999877993150581e-05, "loss": 0.0456, "step": 4170 }, { "grad_norm": 0.4389994442462921, "learning_rate": 9.999863127358108e-05, "loss": 0.0471, "step": 4180 }, { "grad_norm": 0.3908551037311554, "learning_rate": 9.999847407225773e-05, "loss": 0.0432, "step": 4190 }, { "grad_norm": 0.371036559343338, "learning_rate": 9.999830832756262e-05, "loss": 0.0442, "step": 4200 }, { "grad_norm": 0.44369661808013916, "learning_rate": 9.999813403952407e-05, "loss": 0.036, "step": 4210 }, { "grad_norm": 0.5379936695098877, "learning_rate": 9.999795120817187e-05, "loss": 0.0539, "step": 4220 }, { "grad_norm": 0.43512627482414246, "learning_rate": 9.999775983353725e-05, "loss": 0.0478, "step": 4230 }, { "grad_norm": 0.40959644317626953, "learning_rate": 9.999755991565292e-05, "loss": 0.04, "step": 4240 }, { "grad_norm": 0.42856326699256897, "learning_rate": 9.999735145455303e-05, "loss": 0.0409, "step": 4250 }, { "grad_norm": 0.4295378029346466, "learning_rate": 9.99971344502732e-05, "loss": 0.0548, "step": 4260 }, { "grad_norm": 0.45623865723609924, "learning_rate": 9.999690890285053e-05, "loss": 0.0452, "step": 4270 }, { "grad_norm": 0.45868411660194397, "learning_rate": 9.999667481232356e-05, "loss": 0.0434, "step": 4280 }, { "grad_norm": 0.5010131001472473, "learning_rate": 9.999643217873225e-05, "loss": 0.0444, "step": 4290 }, { "grad_norm": 0.5755404829978943, "learning_rate": 9.999618100211809e-05, "loss": 0.047, "step": 4300 }, { "grad_norm": 0.3672829866409302, "learning_rate": 9.999592128252402e-05, "loss": 0.0353, "step": 4310 }, { "grad_norm": 0.4134146571159363, "learning_rate": 9.999565301999437e-05, "loss": 0.0347, "step": 4320 }, { "grad_norm": 0.4856330454349518, "learning_rate": 9.999537621457502e-05, "loss": 0.0359, "step": 4330 }, { "grad_norm": 0.4850810766220093, "learning_rate": 9.999509086631323e-05, "loss": 0.0527, "step": 4340 }, { "grad_norm": 0.38435518741607666, "learning_rate": 9.99947969752578e-05, "loss": 0.0419, "step": 4350 }, { "grad_norm": 0.5045213103294373, "learning_rate": 9.999449454145891e-05, "loss": 0.0448, "step": 4360 }, { "grad_norm": 0.31859809160232544, "learning_rate": 9.999418356496827e-05, "loss": 0.0392, "step": 4370 }, { "grad_norm": 0.5234799385070801, "learning_rate": 9.999386404583899e-05, "loss": 0.0465, "step": 4380 }, { "grad_norm": 0.36768561601638794, "learning_rate": 9.999353598412568e-05, "loss": 0.0381, "step": 4390 }, { "grad_norm": 0.5104751586914062, "learning_rate": 9.999319937988442e-05, "loss": 0.042, "step": 4400 }, { "grad_norm": 0.46488428115844727, "learning_rate": 9.999285423317268e-05, "loss": 0.0341, "step": 4410 }, { "grad_norm": 0.4123372733592987, "learning_rate": 9.999250054404947e-05, "loss": 0.044, "step": 4420 }, { "grad_norm": 0.4702058732509613, "learning_rate": 9.99921383125752e-05, "loss": 0.039, "step": 4430 }, { "grad_norm": 0.664667010307312, "learning_rate": 9.99917675388118e-05, "loss": 0.0444, "step": 4440 }, { "grad_norm": 0.48562461137771606, "learning_rate": 9.99913882228226e-05, "loss": 0.0414, "step": 4450 }, { "grad_norm": 0.43508341908454895, "learning_rate": 9.999100036467242e-05, "loss": 0.044, "step": 4460 }, { "grad_norm": 0.4537997245788574, "learning_rate": 9.999060396442753e-05, "loss": 0.0486, "step": 4470 }, { "grad_norm": 0.5342022776603699, "learning_rate": 9.999019902215566e-05, "loss": 0.0427, "step": 4480 }, { "grad_norm": 0.3794819414615631, "learning_rate": 9.998978553792602e-05, "loss": 0.0548, "step": 4490 }, { "grad_norm": 0.31241267919540405, "learning_rate": 9.998936351180926e-05, "loss": 0.0426, "step": 4500 }, { "grad_norm": 0.3834628164768219, "learning_rate": 9.998893294387747e-05, "loss": 0.0501, "step": 4510 }, { "grad_norm": 0.4144917130470276, "learning_rate": 9.998849383420426e-05, "loss": 0.0425, "step": 4520 }, { "grad_norm": 0.41814082860946655, "learning_rate": 9.998804618286465e-05, "loss": 0.045, "step": 4530 }, { "grad_norm": 0.4152892231941223, "learning_rate": 9.99875899899351e-05, "loss": 0.0548, "step": 4540 }, { "grad_norm": 0.5304038524627686, "learning_rate": 9.99871252554936e-05, "loss": 0.0442, "step": 4550 }, { "grad_norm": 0.4931137263774872, "learning_rate": 9.998665197961955e-05, "loss": 0.0423, "step": 4560 }, { "grad_norm": 0.36734575033187866, "learning_rate": 9.998617016239379e-05, "loss": 0.0412, "step": 4570 }, { "grad_norm": 0.4655676484107971, "learning_rate": 9.998567980389869e-05, "loss": 0.0363, "step": 4580 }, { "grad_norm": 0.3279894292354584, "learning_rate": 9.998518090421802e-05, "loss": 0.0445, "step": 4590 }, { "grad_norm": 0.4765828251838684, "learning_rate": 9.998467346343703e-05, "loss": 0.0373, "step": 4600 }, { "grad_norm": 0.41299644112586975, "learning_rate": 9.998415748164243e-05, "loss": 0.0429, "step": 4610 }, { "grad_norm": 0.3256024718284607, "learning_rate": 9.998363295892238e-05, "loss": 0.0414, "step": 4620 }, { "grad_norm": 0.5351561903953552, "learning_rate": 9.998309989536652e-05, "loss": 0.0441, "step": 4630 }, { "grad_norm": 0.311116486787796, "learning_rate": 9.998255829106593e-05, "loss": 0.0393, "step": 4640 }, { "grad_norm": 0.4408559501171112, "learning_rate": 9.998200814611316e-05, "loss": 0.0428, "step": 4650 }, { "grad_norm": 0.544904887676239, "learning_rate": 9.998144946060219e-05, "loss": 0.0544, "step": 4660 }, { "grad_norm": 0.382839173078537, "learning_rate": 9.998088223462852e-05, "loss": 0.0428, "step": 4670 }, { "grad_norm": 0.32513612508773804, "learning_rate": 9.998030646828905e-05, "loss": 0.0449, "step": 4680 }, { "grad_norm": 0.4306314289569855, "learning_rate": 9.997972216168217e-05, "loss": 0.0404, "step": 4690 }, { "grad_norm": 0.45632943511009216, "learning_rate": 9.997912931490771e-05, "loss": 0.0406, "step": 4700 }, { "grad_norm": 0.4381083548069, "learning_rate": 9.9978527928067e-05, "loss": 0.0448, "step": 4710 }, { "grad_norm": 0.3828885555267334, "learning_rate": 9.997791800126277e-05, "loss": 0.0463, "step": 4720 }, { "grad_norm": 0.4546440839767456, "learning_rate": 9.997729953459927e-05, "loss": 0.0413, "step": 4730 }, { "grad_norm": 0.36371952295303345, "learning_rate": 9.997667252818214e-05, "loss": 0.0423, "step": 4740 }, { "grad_norm": 0.4604377746582031, "learning_rate": 9.997603698211855e-05, "loss": 0.0415, "step": 4750 }, { "grad_norm": 0.4028468430042267, "learning_rate": 9.99753928965171e-05, "loss": 0.0448, "step": 4760 }, { "grad_norm": 0.5065643787384033, "learning_rate": 9.997474027148781e-05, "loss": 0.0427, "step": 4770 }, { "grad_norm": 0.3405841290950775, "learning_rate": 9.997407910714223e-05, "loss": 0.0451, "step": 4780 }, { "grad_norm": 0.44323331117630005, "learning_rate": 9.997340940359332e-05, "loss": 0.041, "step": 4790 }, { "grad_norm": 0.48448988795280457, "learning_rate": 9.997273116095552e-05, "loss": 0.0446, "step": 4800 }, { "grad_norm": 0.4945404529571533, "learning_rate": 9.997204437934473e-05, "loss": 0.0406, "step": 4810 }, { "grad_norm": 0.40779298543930054, "learning_rate": 9.997134905887829e-05, "loss": 0.0406, "step": 4820 }, { "grad_norm": 0.4229153096675873, "learning_rate": 9.997064519967501e-05, "loss": 0.0407, "step": 4830 }, { "grad_norm": 0.3593582212924957, "learning_rate": 9.996993280185517e-05, "loss": 0.043, "step": 4840 }, { "grad_norm": 0.3584595024585724, "learning_rate": 9.99692118655405e-05, "loss": 0.0414, "step": 4850 }, { "grad_norm": 0.2807089686393738, "learning_rate": 9.996848239085417e-05, "loss": 0.0386, "step": 4860 }, { "grad_norm": 0.35087722539901733, "learning_rate": 9.996774437792085e-05, "loss": 0.0423, "step": 4870 }, { "grad_norm": 0.3640047609806061, "learning_rate": 9.996699782686664e-05, "loss": 0.0361, "step": 4880 }, { "grad_norm": 0.4195345342159271, "learning_rate": 9.996624273781909e-05, "loss": 0.0409, "step": 4890 }, { "grad_norm": 0.34995055198669434, "learning_rate": 9.996547911090725e-05, "loss": 0.038, "step": 4900 }, { "grad_norm": 0.4135249853134155, "learning_rate": 9.996470694626157e-05, "loss": 0.0466, "step": 4910 }, { "grad_norm": 0.3482966125011444, "learning_rate": 9.996392624401403e-05, "loss": 0.044, "step": 4920 }, { "grad_norm": 0.4445941150188446, "learning_rate": 9.996313700429801e-05, "loss": 0.0377, "step": 4930 }, { "grad_norm": 0.4350826144218445, "learning_rate": 9.996233922724836e-05, "loss": 0.038, "step": 4940 }, { "grad_norm": 0.410921186208725, "learning_rate": 9.996153291300141e-05, "loss": 0.0542, "step": 4950 }, { "grad_norm": 0.3728349804878235, "learning_rate": 9.996071806169494e-05, "loss": 0.0361, "step": 4960 }, { "grad_norm": 0.35408514738082886, "learning_rate": 9.995989467346817e-05, "loss": 0.0399, "step": 4970 }, { "grad_norm": 0.36848753690719604, "learning_rate": 9.995906274846183e-05, "loss": 0.0402, "step": 4980 }, { "grad_norm": 0.3917164206504822, "learning_rate": 9.995822228681803e-05, "loss": 0.0436, "step": 4990 }, { "grad_norm": 0.4484070837497711, "learning_rate": 9.99573732886804e-05, "loss": 0.0413, "step": 5000 }, { "grad_norm": 0.3140717148780823, "learning_rate": 9.995651575419402e-05, "loss": 0.0337, "step": 5010 }, { "grad_norm": 0.37881579995155334, "learning_rate": 9.995564968350541e-05, "loss": 0.0544, "step": 5020 }, { "grad_norm": 0.49457451701164246, "learning_rate": 9.995477507676256e-05, "loss": 0.0434, "step": 5030 }, { "grad_norm": 0.3214081823825836, "learning_rate": 9.995389193411493e-05, "loss": 0.0377, "step": 5040 }, { "grad_norm": 0.47797009348869324, "learning_rate": 9.995300025571339e-05, "loss": 0.0363, "step": 5050 }, { "grad_norm": 0.3587394654750824, "learning_rate": 9.995210004171034e-05, "loss": 0.0393, "step": 5060 }, { "grad_norm": 0.29887068271636963, "learning_rate": 9.995119129225956e-05, "loss": 0.0413, "step": 5070 }, { "grad_norm": 0.3272266089916229, "learning_rate": 9.995027400751637e-05, "loss": 0.0345, "step": 5080 }, { "grad_norm": 0.33268240094184875, "learning_rate": 9.994934818763751e-05, "loss": 0.0431, "step": 5090 }, { "grad_norm": 0.4031470715999603, "learning_rate": 9.994841383278115e-05, "loss": 0.0424, "step": 5100 }, { "grad_norm": 0.3274897634983063, "learning_rate": 9.994747094310695e-05, "loss": 0.0382, "step": 5110 }, { "grad_norm": 0.388652503490448, "learning_rate": 9.994651951877604e-05, "loss": 0.0407, "step": 5120 }, { "grad_norm": 0.41498151421546936, "learning_rate": 9.994555955995099e-05, "loss": 0.0385, "step": 5130 }, { "grad_norm": 0.31065860390663147, "learning_rate": 9.994459106679581e-05, "loss": 0.0356, "step": 5140 }, { "grad_norm": 0.3383851647377014, "learning_rate": 9.994361403947603e-05, "loss": 0.0479, "step": 5150 }, { "grad_norm": 0.4299122393131256, "learning_rate": 9.994262847815854e-05, "loss": 0.0361, "step": 5160 }, { "grad_norm": 0.37227678298950195, "learning_rate": 9.99416343830118e-05, "loss": 0.0414, "step": 5170 }, { "grad_norm": 0.4289757013320923, "learning_rate": 9.994063175420565e-05, "loss": 0.0342, "step": 5180 }, { "grad_norm": 0.3861692249774933, "learning_rate": 9.99396205919114e-05, "loss": 0.0405, "step": 5190 }, { "grad_norm": 0.38395729660987854, "learning_rate": 9.993860089630185e-05, "loss": 0.039, "step": 5200 }, { "grad_norm": 0.3953026831150055, "learning_rate": 9.993757266755123e-05, "loss": 0.0387, "step": 5210 }, { "grad_norm": 0.4286213517189026, "learning_rate": 9.993653590583522e-05, "loss": 0.0397, "step": 5220 }, { "grad_norm": 0.38878172636032104, "learning_rate": 9.993549061133102e-05, "loss": 0.0351, "step": 5230 }, { "grad_norm": 0.39169883728027344, "learning_rate": 9.993443678421719e-05, "loss": 0.038, "step": 5240 }, { "grad_norm": 0.37418776750564575, "learning_rate": 9.993337442467384e-05, "loss": 0.0405, "step": 5250 }, { "grad_norm": 0.2671048045158386, "learning_rate": 9.993230353288248e-05, "loss": 0.0375, "step": 5260 }, { "grad_norm": 0.41562652587890625, "learning_rate": 9.993122410902608e-05, "loss": 0.0414, "step": 5270 }, { "grad_norm": 0.34002310037612915, "learning_rate": 9.993013615328912e-05, "loss": 0.0346, "step": 5280 }, { "grad_norm": 0.3519008457660675, "learning_rate": 9.992903966585747e-05, "loss": 0.0395, "step": 5290 }, { "grad_norm": 0.4274384081363678, "learning_rate": 9.992793464691852e-05, "loss": 0.0385, "step": 5300 }, { "grad_norm": 0.3572917878627777, "learning_rate": 9.992682109666105e-05, "loss": 0.0399, "step": 5310 }, { "grad_norm": 0.33959105610847473, "learning_rate": 9.992569901527538e-05, "loss": 0.0336, "step": 5320 }, { "grad_norm": 0.38588589429855347, "learning_rate": 9.99245684029532e-05, "loss": 0.0431, "step": 5330 }, { "grad_norm": 0.33187633752822876, "learning_rate": 9.992342925988774e-05, "loss": 0.0391, "step": 5340 }, { "grad_norm": 0.4083288311958313, "learning_rate": 9.992228158627361e-05, "loss": 0.0378, "step": 5350 }, { "grad_norm": 0.3457348942756653, "learning_rate": 9.992112538230693e-05, "loss": 0.0379, "step": 5360 }, { "grad_norm": 0.31987273693084717, "learning_rate": 9.991996064818527e-05, "loss": 0.0393, "step": 5370 }, { "grad_norm": 0.3721388578414917, "learning_rate": 9.991878738410768e-05, "loss": 0.0409, "step": 5380 }, { "grad_norm": 0.4058922231197357, "learning_rate": 9.991760559027457e-05, "loss": 0.0419, "step": 5390 }, { "grad_norm": 0.34973692893981934, "learning_rate": 9.991641526688793e-05, "loss": 0.0357, "step": 5400 }, { "grad_norm": 0.4551521837711334, "learning_rate": 9.991521641415113e-05, "loss": 0.0445, "step": 5410 }, { "grad_norm": 0.36950400471687317, "learning_rate": 9.991400903226904e-05, "loss": 0.042, "step": 5420 }, { "grad_norm": 0.4378596842288971, "learning_rate": 9.991279312144794e-05, "loss": 0.0401, "step": 5430 }, { "grad_norm": 0.32041800022125244, "learning_rate": 9.991156868189564e-05, "loss": 0.035, "step": 5440 }, { "grad_norm": 0.3157516121864319, "learning_rate": 9.991033571382131e-05, "loss": 0.0336, "step": 5450 }, { "grad_norm": 0.2467762529850006, "learning_rate": 9.990909421743569e-05, "loss": 0.0411, "step": 5460 }, { "grad_norm": 0.3263469338417053, "learning_rate": 9.990784419295085e-05, "loss": 0.0414, "step": 5470 }, { "grad_norm": 0.3765830397605896, "learning_rate": 9.990658564058044e-05, "loss": 0.0376, "step": 5480 }, { "grad_norm": 0.4481829106807709, "learning_rate": 9.990531856053948e-05, "loss": 0.0354, "step": 5490 }, { "grad_norm": 0.36389145255088806, "learning_rate": 9.99040429530445e-05, "loss": 0.0389, "step": 5500 }, { "grad_norm": 0.4072919189929962, "learning_rate": 9.990275881831346e-05, "loss": 0.0409, "step": 5510 }, { "grad_norm": 0.3611977696418762, "learning_rate": 9.990146615656577e-05, "loss": 0.0395, "step": 5520 }, { "grad_norm": 0.3288697600364685, "learning_rate": 9.990016496802233e-05, "loss": 0.0382, "step": 5530 }, { "grad_norm": 0.3143702745437622, "learning_rate": 9.989885525290548e-05, "loss": 0.0281, "step": 5540 }, { "grad_norm": 0.36469143629074097, "learning_rate": 9.989753701143897e-05, "loss": 0.0384, "step": 5550 }, { "grad_norm": 0.3968755602836609, "learning_rate": 9.989621024384812e-05, "loss": 0.0404, "step": 5560 }, { "grad_norm": 0.3524176776409149, "learning_rate": 9.989487495035959e-05, "loss": 0.0384, "step": 5570 }, { "grad_norm": 0.4100457727909088, "learning_rate": 9.989353113120156e-05, "loss": 0.0434, "step": 5580 }, { "grad_norm": 0.3930557370185852, "learning_rate": 9.989217878660366e-05, "loss": 0.0401, "step": 5590 }, { "grad_norm": 0.3679465353488922, "learning_rate": 9.989081791679695e-05, "loss": 0.0349, "step": 5600 }, { "grad_norm": 0.3893984854221344, "learning_rate": 9.988944852201397e-05, "loss": 0.0346, "step": 5610 }, { "grad_norm": 0.4222191572189331, "learning_rate": 9.988807060248873e-05, "loss": 0.0373, "step": 5620 }, { "grad_norm": 0.41161638498306274, "learning_rate": 9.988668415845665e-05, "loss": 0.0332, "step": 5630 }, { "grad_norm": 0.4656347632408142, "learning_rate": 9.988528919015466e-05, "loss": 0.0336, "step": 5640 }, { "grad_norm": 0.37889426946640015, "learning_rate": 9.988388569782112e-05, "loss": 0.0406, "step": 5650 }, { "grad_norm": 0.3573051989078522, "learning_rate": 9.988247368169583e-05, "loss": 0.0376, "step": 5660 }, { "grad_norm": 0.3604784309864044, "learning_rate": 9.988105314202007e-05, "loss": 0.0422, "step": 5670 }, { "grad_norm": 0.25682204961776733, "learning_rate": 9.987962407903659e-05, "loss": 0.0372, "step": 5680 }, { "grad_norm": 0.3097984194755554, "learning_rate": 9.987818649298957e-05, "loss": 0.0375, "step": 5690 }, { "grad_norm": 0.35520312190055847, "learning_rate": 9.987674038412465e-05, "loss": 0.0327, "step": 5700 }, { "grad_norm": 0.4087008535861969, "learning_rate": 9.987528575268891e-05, "loss": 0.0348, "step": 5710 }, { "grad_norm": 0.4199003577232361, "learning_rate": 9.987382259893095e-05, "loss": 0.037, "step": 5720 }, { "grad_norm": 0.24098661541938782, "learning_rate": 9.987235092310074e-05, "loss": 0.0354, "step": 5730 }, { "grad_norm": 0.3234051764011383, "learning_rate": 9.987087072544978e-05, "loss": 0.0371, "step": 5740 }, { "grad_norm": 0.3701421618461609, "learning_rate": 9.9869382006231e-05, "loss": 0.0384, "step": 5750 }, { "grad_norm": 0.37157562375068665, "learning_rate": 9.986788476569875e-05, "loss": 0.0311, "step": 5760 }, { "grad_norm": 0.3523711562156677, "learning_rate": 9.986637900410887e-05, "loss": 0.0356, "step": 5770 }, { "grad_norm": 0.30509835481643677, "learning_rate": 9.986486472171869e-05, "loss": 0.0309, "step": 5780 }, { "grad_norm": 0.34000104665756226, "learning_rate": 9.986334191878692e-05, "loss": 0.0354, "step": 5790 }, { "grad_norm": 0.36249038577079773, "learning_rate": 9.986181059557378e-05, "loss": 0.0343, "step": 5800 }, { "grad_norm": 0.3821953237056732, "learning_rate": 9.986027075234094e-05, "loss": 0.0323, "step": 5810 }, { "grad_norm": 0.38810038566589355, "learning_rate": 9.985872238935152e-05, "loss": 0.0367, "step": 5820 }, { "grad_norm": 0.25715023279190063, "learning_rate": 9.985716550687008e-05, "loss": 0.0318, "step": 5830 }, { "grad_norm": 0.3272154629230499, "learning_rate": 9.985560010516264e-05, "loss": 0.0347, "step": 5840 }, { "grad_norm": 0.48976990580558777, "learning_rate": 9.985402618449668e-05, "loss": 0.0373, "step": 5850 }, { "grad_norm": 0.3895314633846283, "learning_rate": 9.985244374514118e-05, "loss": 0.0451, "step": 5860 }, { "grad_norm": 0.410422146320343, "learning_rate": 9.985085278736651e-05, "loss": 0.0459, "step": 5870 }, { "grad_norm": 0.2891886830329895, "learning_rate": 9.984925331144452e-05, "loss": 0.0333, "step": 5880 }, { "grad_norm": 0.3112083673477173, "learning_rate": 9.984764531764851e-05, "loss": 0.039, "step": 5890 }, { "grad_norm": 0.3228941857814789, "learning_rate": 9.984602880625326e-05, "loss": 0.0383, "step": 5900 }, { "grad_norm": 0.27487504482269287, "learning_rate": 9.9844403777535e-05, "loss": 0.0356, "step": 5910 }, { "grad_norm": 0.38179752230644226, "learning_rate": 9.984277023177135e-05, "loss": 0.0457, "step": 5920 }, { "grad_norm": 0.33632954955101013, "learning_rate": 9.984112816924148e-05, "loss": 0.034, "step": 5930 }, { "grad_norm": 0.3125036954879761, "learning_rate": 9.983947759022596e-05, "loss": 0.0378, "step": 5940 }, { "grad_norm": 0.30616000294685364, "learning_rate": 9.983781849500682e-05, "loss": 0.0359, "step": 5950 }, { "grad_norm": 0.41578176617622375, "learning_rate": 9.98361508838676e-05, "loss": 0.0369, "step": 5960 }, { "grad_norm": 0.4629523456096649, "learning_rate": 9.98344747570932e-05, "loss": 0.036, "step": 5970 }, { "grad_norm": 0.3784680664539337, "learning_rate": 9.983279011497004e-05, "loss": 0.0332, "step": 5980 }, { "grad_norm": 0.30520278215408325, "learning_rate": 9.983109695778596e-05, "loss": 0.0357, "step": 5990 }, { "grad_norm": 0.40678441524505615, "learning_rate": 9.982939528583032e-05, "loss": 0.039, "step": 6000 }, { "grad_norm": 0.3706188499927521, "learning_rate": 9.982768509939385e-05, "loss": 0.0447, "step": 6010 }, { "grad_norm": 0.42652177810668945, "learning_rate": 9.982596639876879e-05, "loss": 0.0389, "step": 6020 }, { "grad_norm": 0.2830207049846649, "learning_rate": 9.982423918424881e-05, "loss": 0.0344, "step": 6030 }, { "grad_norm": 0.2619018852710724, "learning_rate": 9.982250345612908e-05, "loss": 0.0375, "step": 6040 }, { "grad_norm": 0.4262092113494873, "learning_rate": 9.982075921470611e-05, "loss": 0.0351, "step": 6050 }, { "grad_norm": 0.3694694936275482, "learning_rate": 9.981900646027802e-05, "loss": 0.0335, "step": 6060 }, { "grad_norm": 0.3520243167877197, "learning_rate": 9.981724519314425e-05, "loss": 0.0338, "step": 6070 }, { "grad_norm": 0.4404289722442627, "learning_rate": 9.981547541360581e-05, "loss": 0.0295, "step": 6080 }, { "grad_norm": 0.36102667450904846, "learning_rate": 9.981369712196508e-05, "loss": 0.0365, "step": 6090 }, { "grad_norm": 0.30017468333244324, "learning_rate": 9.981191031852592e-05, "loss": 0.031, "step": 6100 }, { "grad_norm": 0.29750964045524597, "learning_rate": 9.981011500359362e-05, "loss": 0.0357, "step": 6110 }, { "grad_norm": 0.34287959337234497, "learning_rate": 9.9808311177475e-05, "loss": 0.0357, "step": 6120 }, { "grad_norm": 0.3757396638393402, "learning_rate": 9.980649884047826e-05, "loss": 0.031, "step": 6130 }, { "grad_norm": 0.3396052122116089, "learning_rate": 9.980467799291307e-05, "loss": 0.0419, "step": 6140 }, { "grad_norm": 0.4638442397117615, "learning_rate": 9.980284863509058e-05, "loss": 0.0348, "step": 6150 }, { "grad_norm": 0.34926050901412964, "learning_rate": 9.980101076732334e-05, "loss": 0.0377, "step": 6160 }, { "grad_norm": 0.3208204209804535, "learning_rate": 9.979916438992544e-05, "loss": 0.0356, "step": 6170 }, { "grad_norm": 0.3122788369655609, "learning_rate": 9.979730950321237e-05, "loss": 0.0328, "step": 6180 }, { "grad_norm": 0.3575652539730072, "learning_rate": 9.979544610750104e-05, "loss": 0.038, "step": 6190 }, { "grad_norm": 0.37575212121009827, "learning_rate": 9.97935742031099e-05, "loss": 0.0398, "step": 6200 }, { "grad_norm": 0.5197086930274963, "learning_rate": 9.979169379035878e-05, "loss": 0.0484, "step": 6210 }, { "grad_norm": 0.35234495997428894, "learning_rate": 9.978980486956899e-05, "loss": 0.0365, "step": 6220 }, { "grad_norm": 0.3116982579231262, "learning_rate": 9.978790744106332e-05, "loss": 0.0347, "step": 6230 }, { "grad_norm": 0.3642233610153198, "learning_rate": 9.978600150516594e-05, "loss": 0.0289, "step": 6240 }, { "grad_norm": 0.36486566066741943, "learning_rate": 9.978408706220259e-05, "loss": 0.0401, "step": 6250 }, { "grad_norm": 0.3616689443588257, "learning_rate": 9.978216411250032e-05, "loss": 0.0454, "step": 6260 }, { "grad_norm": 0.28171005845069885, "learning_rate": 9.978023265638778e-05, "loss": 0.032, "step": 6270 }, { "grad_norm": 0.42742064595222473, "learning_rate": 9.977829269419495e-05, "loss": 0.034, "step": 6280 }, { "grad_norm": 0.44887232780456543, "learning_rate": 9.977634422625335e-05, "loss": 0.0393, "step": 6290 }, { "grad_norm": 0.3125928044319153, "learning_rate": 9.97743872528959e-05, "loss": 0.0372, "step": 6300 }, { "grad_norm": 0.2811821401119232, "learning_rate": 9.9772421774457e-05, "loss": 0.0326, "step": 6310 }, { "grad_norm": 0.37790635228157043, "learning_rate": 9.977044779127252e-05, "loss": 0.0308, "step": 6320 }, { "grad_norm": 0.3655528724193573, "learning_rate": 9.976846530367971e-05, "loss": 0.0333, "step": 6330 }, { "grad_norm": 0.28514018654823303, "learning_rate": 9.976647431201735e-05, "loss": 0.0382, "step": 6340 }, { "grad_norm": 0.36321479082107544, "learning_rate": 9.976447481662568e-05, "loss": 0.0358, "step": 6350 }, { "grad_norm": 0.30387699604034424, "learning_rate": 9.976246681784629e-05, "loss": 0.0351, "step": 6360 }, { "grad_norm": 0.31181055307388306, "learning_rate": 9.976045031602234e-05, "loss": 0.041, "step": 6370 }, { "grad_norm": 0.34030070900917053, "learning_rate": 9.975842531149837e-05, "loss": 0.0353, "step": 6380 }, { "grad_norm": 0.25730031728744507, "learning_rate": 9.975639180462043e-05, "loss": 0.0292, "step": 6390 }, { "grad_norm": 0.38701382279396057, "learning_rate": 9.975434979573596e-05, "loss": 0.0362, "step": 6400 }, { "grad_norm": 0.41777029633522034, "learning_rate": 9.97522992851939e-05, "loss": 0.0306, "step": 6410 }, { "grad_norm": 0.39182350039482117, "learning_rate": 9.975024027334461e-05, "loss": 0.0325, "step": 6420 }, { "grad_norm": 0.37735411524772644, "learning_rate": 9.974817276053993e-05, "loss": 0.0344, "step": 6430 }, { "grad_norm": 0.239715576171875, "learning_rate": 9.974609674713315e-05, "loss": 0.036, "step": 6440 }, { "grad_norm": 0.30965298414230347, "learning_rate": 9.9744012233479e-05, "loss": 0.0473, "step": 6450 }, { "grad_norm": 0.3785746693611145, "learning_rate": 9.974191921993366e-05, "loss": 0.0327, "step": 6460 }, { "grad_norm": 0.2892898619174957, "learning_rate": 9.973981770685474e-05, "loss": 0.0312, "step": 6470 }, { "grad_norm": 0.25268450379371643, "learning_rate": 9.97377076946014e-05, "loss": 0.0345, "step": 6480 }, { "grad_norm": 0.2839321792125702, "learning_rate": 9.973558918353412e-05, "loss": 0.0342, "step": 6490 }, { "grad_norm": 0.34005576372146606, "learning_rate": 9.973346217401494e-05, "loss": 0.0344, "step": 6500 }, { "grad_norm": 0.31852179765701294, "learning_rate": 9.973132666640726e-05, "loss": 0.0338, "step": 6510 }, { "grad_norm": 0.40457314252853394, "learning_rate": 9.972918266107602e-05, "loss": 0.0357, "step": 6520 }, { "grad_norm": 0.29993727803230286, "learning_rate": 9.972703015838756e-05, "loss": 0.0324, "step": 6530 }, { "grad_norm": 0.26681557297706604, "learning_rate": 9.97248691587097e-05, "loss": 0.0331, "step": 6540 }, { "grad_norm": 0.38051295280456543, "learning_rate": 9.972269966241166e-05, "loss": 0.0331, "step": 6550 }, { "grad_norm": 0.3518318831920624, "learning_rate": 9.972052166986417e-05, "loss": 0.0303, "step": 6560 }, { "grad_norm": 0.44882360100746155, "learning_rate": 9.971833518143938e-05, "loss": 0.0371, "step": 6570 }, { "grad_norm": 0.2745528519153595, "learning_rate": 9.971614019751093e-05, "loss": 0.0331, "step": 6580 }, { "grad_norm": 0.291486531496048, "learning_rate": 9.971393671845383e-05, "loss": 0.033, "step": 6590 }, { "grad_norm": 0.43581753969192505, "learning_rate": 9.971172474464464e-05, "loss": 0.0388, "step": 6600 }, { "grad_norm": 0.26633909344673157, "learning_rate": 9.97095042764613e-05, "loss": 0.0303, "step": 6610 }, { "grad_norm": 0.28575214743614197, "learning_rate": 9.970727531428324e-05, "loss": 0.0395, "step": 6620 }, { "grad_norm": 0.2557196319103241, "learning_rate": 9.970503785849132e-05, "loss": 0.0354, "step": 6630 }, { "grad_norm": 0.36666834354400635, "learning_rate": 9.970279190946788e-05, "loss": 0.0396, "step": 6640 }, { "grad_norm": 0.3738102912902832, "learning_rate": 9.970053746759667e-05, "loss": 0.0331, "step": 6650 }, { "grad_norm": 0.405752569437027, "learning_rate": 9.969827453326292e-05, "loss": 0.0337, "step": 6660 }, { "grad_norm": 0.3261372447013855, "learning_rate": 9.969600310685332e-05, "loss": 0.033, "step": 6670 }, { "grad_norm": 0.32274726033210754, "learning_rate": 9.969372318875596e-05, "loss": 0.0352, "step": 6680 }, { "grad_norm": 0.40957677364349365, "learning_rate": 9.969143477936043e-05, "loss": 0.0361, "step": 6690 }, { "grad_norm": 0.2990972697734833, "learning_rate": 9.968913787905775e-05, "loss": 0.0351, "step": 6700 }, { "grad_norm": 0.30667439103126526, "learning_rate": 9.968683248824045e-05, "loss": 0.0338, "step": 6710 }, { "grad_norm": 0.24451607465744019, "learning_rate": 9.968451860730238e-05, "loss": 0.0321, "step": 6720 }, { "grad_norm": 0.3071296215057373, "learning_rate": 9.968219623663896e-05, "loss": 0.0365, "step": 6730 }, { "grad_norm": 0.2676180601119995, "learning_rate": 9.967986537664702e-05, "loss": 0.0307, "step": 6740 }, { "grad_norm": 0.2845117449760437, "learning_rate": 9.967752602772483e-05, "loss": 0.0336, "step": 6750 }, { "grad_norm": 0.32477396726608276, "learning_rate": 9.967517819027212e-05, "loss": 0.0282, "step": 6760 }, { "grad_norm": 0.242864727973938, "learning_rate": 9.967282186469009e-05, "loss": 0.0362, "step": 6770 }, { "grad_norm": 0.267963707447052, "learning_rate": 9.967045705138135e-05, "loss": 0.0319, "step": 6780 }, { "grad_norm": 0.3003794252872467, "learning_rate": 9.966808375074998e-05, "loss": 0.0351, "step": 6790 }, { "grad_norm": 0.28945016860961914, "learning_rate": 9.966570196320154e-05, "loss": 0.0348, "step": 6800 }, { "grad_norm": 0.29786261916160583, "learning_rate": 9.966331168914299e-05, "loss": 0.031, "step": 6810 }, { "grad_norm": 0.3666224777698517, "learning_rate": 9.966091292898277e-05, "loss": 0.0429, "step": 6820 }, { "grad_norm": 0.25382450222969055, "learning_rate": 9.965850568313076e-05, "loss": 0.0337, "step": 6830 }, { "grad_norm": 0.2747704088687897, "learning_rate": 9.965608995199827e-05, "loss": 0.03, "step": 6840 }, { "grad_norm": 0.3063940107822418, "learning_rate": 9.965366573599812e-05, "loss": 0.0314, "step": 6850 }, { "grad_norm": 0.37189817428588867, "learning_rate": 9.965123303554453e-05, "loss": 0.0279, "step": 6860 }, { "grad_norm": 0.37349241971969604, "learning_rate": 9.964879185105317e-05, "loss": 0.037, "step": 6870 }, { "grad_norm": 0.26071450114250183, "learning_rate": 9.964634218294119e-05, "loss": 0.0318, "step": 6880 }, { "grad_norm": 0.23269088566303253, "learning_rate": 9.964388403162714e-05, "loss": 0.0241, "step": 6890 }, { "grad_norm": 0.2756706774234772, "learning_rate": 9.96414173975311e-05, "loss": 0.0291, "step": 6900 }, { "grad_norm": 0.42757686972618103, "learning_rate": 9.963894228107451e-05, "loss": 0.0337, "step": 6910 }, { "grad_norm": 0.37430834770202637, "learning_rate": 9.963645868268032e-05, "loss": 0.0345, "step": 6920 }, { "grad_norm": 0.36691534519195557, "learning_rate": 9.963396660277289e-05, "loss": 0.0381, "step": 6930 }, { "grad_norm": 0.3462618589401245, "learning_rate": 9.963146604177807e-05, "loss": 0.0391, "step": 6940 }, { "grad_norm": 0.33551061153411865, "learning_rate": 9.962895700012311e-05, "loss": 0.0318, "step": 6950 }, { "grad_norm": 0.36289486289024353, "learning_rate": 9.962643947823677e-05, "loss": 0.0343, "step": 6960 }, { "grad_norm": 0.29794469475746155, "learning_rate": 9.962391347654921e-05, "loss": 0.037, "step": 6970 }, { "grad_norm": 0.31927213072776794, "learning_rate": 9.962137899549204e-05, "loss": 0.0305, "step": 6980 }, { "grad_norm": 0.24957114458084106, "learning_rate": 9.961883603549835e-05, "loss": 0.0317, "step": 6990 }, { "grad_norm": 0.38821953535079956, "learning_rate": 9.961628459700267e-05, "loss": 0.0288, "step": 7000 }, { "grad_norm": 0.29203277826309204, "learning_rate": 9.961372468044095e-05, "loss": 0.0355, "step": 7010 }, { "grad_norm": 0.38052287697792053, "learning_rate": 9.961115628625062e-05, "loss": 0.038, "step": 7020 }, { "grad_norm": 0.3283199071884155, "learning_rate": 9.960857941487056e-05, "loss": 0.0346, "step": 7030 }, { "grad_norm": 0.2697531282901764, "learning_rate": 9.960599406674106e-05, "loss": 0.0345, "step": 7040 }, { "grad_norm": 0.2951621115207672, "learning_rate": 9.960340024230393e-05, "loss": 0.0348, "step": 7050 }, { "grad_norm": 0.31617775559425354, "learning_rate": 9.960079794200232e-05, "loss": 0.0328, "step": 7060 }, { "grad_norm": 0.2652266323566437, "learning_rate": 9.959818716628096e-05, "loss": 0.0331, "step": 7070 }, { "grad_norm": 0.40191444754600525, "learning_rate": 9.95955679155859e-05, "loss": 0.0328, "step": 7080 }, { "grad_norm": 0.3808077871799469, "learning_rate": 9.959294019036472e-05, "loss": 0.0344, "step": 7090 }, { "grad_norm": 0.32739973068237305, "learning_rate": 9.959030399106646e-05, "loss": 0.0331, "step": 7100 }, { "grad_norm": 0.34232816100120544, "learning_rate": 9.958765931814153e-05, "loss": 0.0331, "step": 7110 }, { "grad_norm": 0.6897105574607849, "learning_rate": 9.958500617204184e-05, "loss": 0.0358, "step": 7120 }, { "grad_norm": 0.3851407468318939, "learning_rate": 9.958234455322075e-05, "loss": 0.0376, "step": 7130 }, { "grad_norm": 0.4229067265987396, "learning_rate": 9.957967446213308e-05, "loss": 0.0483, "step": 7140 }, { "grad_norm": 0.3788696229457855, "learning_rate": 9.957699589923501e-05, "loss": 0.0403, "step": 7150 }, { "grad_norm": 0.3486098647117615, "learning_rate": 9.957430886498431e-05, "loss": 0.0369, "step": 7160 }, { "grad_norm": 0.43564796447753906, "learning_rate": 9.957161335984008e-05, "loss": 0.0355, "step": 7170 }, { "grad_norm": 0.3370746970176697, "learning_rate": 9.956890938426291e-05, "loss": 0.0296, "step": 7180 }, { "grad_norm": 0.3045687973499298, "learning_rate": 9.956619693871482e-05, "loss": 0.0323, "step": 7190 }, { "grad_norm": 0.31637901067733765, "learning_rate": 9.956347602365934e-05, "loss": 0.0354, "step": 7200 }, { "grad_norm": 0.27339494228363037, "learning_rate": 9.956074663956135e-05, "loss": 0.0356, "step": 7210 }, { "grad_norm": 0.395096093416214, "learning_rate": 9.955800878688726e-05, "loss": 0.031, "step": 7220 }, { "grad_norm": 0.330371618270874, "learning_rate": 9.955526246610489e-05, "loss": 0.0292, "step": 7230 }, { "grad_norm": 0.3775777816772461, "learning_rate": 9.955250767768349e-05, "loss": 0.0327, "step": 7240 }, { "grad_norm": 0.28562989830970764, "learning_rate": 9.95497444220938e-05, "loss": 0.0274, "step": 7250 }, { "grad_norm": 0.3032552897930145, "learning_rate": 9.954697269980797e-05, "loss": 0.0406, "step": 7260 }, { "grad_norm": 0.2830490469932556, "learning_rate": 9.954419251129962e-05, "loss": 0.0297, "step": 7270 }, { "grad_norm": 0.3309101462364197, "learning_rate": 9.95414038570438e-05, "loss": 0.0318, "step": 7280 }, { "grad_norm": 0.2936035096645355, "learning_rate": 9.953860673751703e-05, "loss": 0.0327, "step": 7290 }, { "grad_norm": 0.29450830817222595, "learning_rate": 9.953580115319725e-05, "loss": 0.0261, "step": 7300 }, { "grad_norm": 0.2364344447851181, "learning_rate": 9.953298710456387e-05, "loss": 0.0302, "step": 7310 }, { "grad_norm": 0.3390825390815735, "learning_rate": 9.953016459209771e-05, "loss": 0.0269, "step": 7320 }, { "grad_norm": 0.3496885597705841, "learning_rate": 9.952733361628108e-05, "loss": 0.0292, "step": 7330 }, { "grad_norm": 0.2891959249973297, "learning_rate": 9.952449417759772e-05, "loss": 0.03, "step": 7340 }, { "grad_norm": 0.21506375074386597, "learning_rate": 9.952164627653279e-05, "loss": 0.032, "step": 7350 }, { "grad_norm": 0.2982131540775299, "learning_rate": 9.951878991357292e-05, "loss": 0.0272, "step": 7360 }, { "grad_norm": 0.2923526465892792, "learning_rate": 9.951592508920622e-05, "loss": 0.0354, "step": 7370 }, { "grad_norm": 0.39301592111587524, "learning_rate": 9.951305180392219e-05, "loss": 0.0383, "step": 7380 }, { "grad_norm": 0.2739352583885193, "learning_rate": 9.951017005821178e-05, "loss": 0.0285, "step": 7390 }, { "grad_norm": 0.25622326135635376, "learning_rate": 9.95072798525674e-05, "loss": 0.0302, "step": 7400 }, { "grad_norm": 0.26457729935646057, "learning_rate": 9.950438118748293e-05, "loss": 0.0328, "step": 7410 }, { "grad_norm": 0.36722081899642944, "learning_rate": 9.950147406345366e-05, "loss": 0.0322, "step": 7420 }, { "grad_norm": 0.25892433524131775, "learning_rate": 9.949855848097635e-05, "loss": 0.0288, "step": 7430 }, { "grad_norm": 0.3595782220363617, "learning_rate": 9.949563444054916e-05, "loss": 0.0337, "step": 7440 }, { "grad_norm": 0.2523807883262634, "learning_rate": 9.949270194267178e-05, "loss": 0.0333, "step": 7450 }, { "grad_norm": 0.3690319061279297, "learning_rate": 9.948976098784526e-05, "loss": 0.0363, "step": 7460 }, { "grad_norm": 0.2910451591014862, "learning_rate": 9.948681157657213e-05, "loss": 0.038, "step": 7470 }, { "grad_norm": 0.37423786520957947, "learning_rate": 9.948385370935638e-05, "loss": 0.0315, "step": 7480 }, { "grad_norm": 0.5112977027893066, "learning_rate": 9.94808873867034e-05, "loss": 0.0362, "step": 7490 }, { "grad_norm": 0.32603180408477783, "learning_rate": 9.947791260912009e-05, "loss": 0.0303, "step": 7500 }, { "grad_norm": 0.2454315423965454, "learning_rate": 9.947492937711474e-05, "loss": 0.033, "step": 7510 }, { "grad_norm": 0.2839874029159546, "learning_rate": 9.947193769119707e-05, "loss": 0.0354, "step": 7520 }, { "grad_norm": 0.3385360836982727, "learning_rate": 9.946893755187834e-05, "loss": 0.0263, "step": 7530 }, { "grad_norm": 0.27848711609840393, "learning_rate": 9.946592895967115e-05, "loss": 0.0277, "step": 7540 }, { "grad_norm": 0.2697408199310303, "learning_rate": 9.94629119150896e-05, "loss": 0.0337, "step": 7550 }, { "grad_norm": 0.3225659728050232, "learning_rate": 9.94598864186492e-05, "loss": 0.0312, "step": 7560 }, { "grad_norm": 0.32454344630241394, "learning_rate": 9.945685247086696e-05, "loss": 0.0314, "step": 7570 }, { "grad_norm": 0.2538089156150818, "learning_rate": 9.945381007226129e-05, "loss": 0.0275, "step": 7580 }, { "grad_norm": 0.2913561165332794, "learning_rate": 9.945075922335203e-05, "loss": 0.0374, "step": 7590 }, { "grad_norm": 0.2999131977558136, "learning_rate": 9.944769992466049e-05, "loss": 0.0251, "step": 7600 }, { "grad_norm": 0.3628210723400116, "learning_rate": 9.944463217670945e-05, "loss": 0.0339, "step": 7610 }, { "grad_norm": 0.27503708004951477, "learning_rate": 9.944155598002307e-05, "loss": 0.0355, "step": 7620 }, { "grad_norm": 0.397616446018219, "learning_rate": 9.943847133512701e-05, "loss": 0.0324, "step": 7630 }, { "grad_norm": 0.33502161502838135, "learning_rate": 9.943537824254834e-05, "loss": 0.0393, "step": 7640 }, { "grad_norm": 0.2673455774784088, "learning_rate": 9.943227670281559e-05, "loss": 0.031, "step": 7650 }, { "grad_norm": 0.3250238001346588, "learning_rate": 9.942916671645873e-05, "loss": 0.0264, "step": 7660 }, { "grad_norm": 0.31812936067581177, "learning_rate": 9.942604828400916e-05, "loss": 0.0283, "step": 7670 }, { "grad_norm": 0.2739343047142029, "learning_rate": 9.942292140599975e-05, "loss": 0.0305, "step": 7680 }, { "grad_norm": 0.25117695331573486, "learning_rate": 9.94197860829648e-05, "loss": 0.0307, "step": 7690 }, { "grad_norm": 0.33927780389785767, "learning_rate": 9.941664231544004e-05, "loss": 0.0289, "step": 7700 }, { "grad_norm": 0.21605905890464783, "learning_rate": 9.941349010396264e-05, "loss": 0.0316, "step": 7710 }, { "grad_norm": 0.3564732074737549, "learning_rate": 9.941032944907125e-05, "loss": 0.0332, "step": 7720 }, { "grad_norm": 0.31643569469451904, "learning_rate": 9.940716035130596e-05, "loss": 0.029, "step": 7730 }, { "grad_norm": 0.33323702216148376, "learning_rate": 9.940398281120821e-05, "loss": 0.0266, "step": 7740 }, { "grad_norm": 0.26393526792526245, "learning_rate": 9.940079682932102e-05, "loss": 0.0299, "step": 7750 }, { "grad_norm": 0.3810127377510071, "learning_rate": 9.939760240618877e-05, "loss": 0.0287, "step": 7760 }, { "grad_norm": 0.2909797132015228, "learning_rate": 9.939439954235729e-05, "loss": 0.0355, "step": 7770 }, { "grad_norm": 0.3485502004623413, "learning_rate": 9.939118823837387e-05, "loss": 0.0328, "step": 7780 }, { "grad_norm": 0.36873164772987366, "learning_rate": 9.938796849478725e-05, "loss": 0.0328, "step": 7790 }, { "grad_norm": 0.2877669930458069, "learning_rate": 9.938474031214755e-05, "loss": 0.0393, "step": 7800 }, { "grad_norm": 0.2982155680656433, "learning_rate": 9.938150369100643e-05, "loss": 0.0317, "step": 7810 }, { "grad_norm": 0.277864933013916, "learning_rate": 9.93782586319169e-05, "loss": 0.0286, "step": 7820 }, { "grad_norm": 0.243008092045784, "learning_rate": 9.937500513543348e-05, "loss": 0.0281, "step": 7830 }, { "grad_norm": 0.28018298745155334, "learning_rate": 9.937174320211207e-05, "loss": 0.0453, "step": 7840 }, { "grad_norm": 0.2813608944416046, "learning_rate": 9.936847283251009e-05, "loss": 0.0287, "step": 7850 }, { "grad_norm": 0.33768826723098755, "learning_rate": 9.936519402718632e-05, "loss": 0.0299, "step": 7860 }, { "grad_norm": 0.29655835032463074, "learning_rate": 9.936190678670102e-05, "loss": 0.0349, "step": 7870 }, { "grad_norm": 0.3682067096233368, "learning_rate": 9.935861111161593e-05, "loss": 0.0323, "step": 7880 }, { "grad_norm": 0.2879161536693573, "learning_rate": 9.935530700249416e-05, "loss": 0.029, "step": 7890 }, { "grad_norm": 0.343533456325531, "learning_rate": 9.935199445990028e-05, "loss": 0.0319, "step": 7900 }, { "grad_norm": 0.312450110912323, "learning_rate": 9.934867348440033e-05, "loss": 0.0298, "step": 7910 }, { "grad_norm": 0.2762218713760376, "learning_rate": 9.934534407656176e-05, "loss": 0.0267, "step": 7920 }, { "grad_norm": 0.4004560112953186, "learning_rate": 9.93420062369535e-05, "loss": 0.0267, "step": 7930 }, { "grad_norm": 0.28360143303871155, "learning_rate": 9.933865996614589e-05, "loss": 0.0281, "step": 7940 }, { "grad_norm": 0.3423270881175995, "learning_rate": 9.933530526471068e-05, "loss": 0.0286, "step": 7950 }, { "grad_norm": 0.2437974214553833, "learning_rate": 9.933194213322114e-05, "loss": 0.0296, "step": 7960 }, { "grad_norm": 0.3338913023471832, "learning_rate": 9.932857057225192e-05, "loss": 0.0288, "step": 7970 }, { "grad_norm": 0.2605665624141693, "learning_rate": 9.932519058237912e-05, "loss": 0.0296, "step": 7980 }, { "grad_norm": 0.31775766611099243, "learning_rate": 9.932180216418032e-05, "loss": 0.0346, "step": 7990 }, { "grad_norm": 0.28003382682800293, "learning_rate": 9.931840531823446e-05, "loss": 0.0314, "step": 8000 }, { "grad_norm": 0.35633623600006104, "learning_rate": 9.9315000045122e-05, "loss": 0.0283, "step": 8010 }, { "grad_norm": 0.34563469886779785, "learning_rate": 9.931158634542481e-05, "loss": 0.0324, "step": 8020 }, { "grad_norm": 0.24726052582263947, "learning_rate": 9.930816421972617e-05, "loss": 0.0271, "step": 8030 }, { "grad_norm": 0.3022575080394745, "learning_rate": 9.930473366861086e-05, "loss": 0.0274, "step": 8040 }, { "grad_norm": 0.31050410866737366, "learning_rate": 9.930129469266505e-05, "loss": 0.0334, "step": 8050 }, { "grad_norm": 0.29892027378082275, "learning_rate": 9.929784729247638e-05, "loss": 0.0258, "step": 8060 }, { "grad_norm": 0.2688104510307312, "learning_rate": 9.929439146863389e-05, "loss": 0.0244, "step": 8070 }, { "grad_norm": 0.26449304819107056, "learning_rate": 9.92909272217281e-05, "loss": 0.0305, "step": 8080 }, { "grad_norm": 0.34047457575798035, "learning_rate": 9.928745455235097e-05, "loss": 0.0265, "step": 8090 }, { "grad_norm": 0.23641522228717804, "learning_rate": 9.928397346109588e-05, "loss": 0.0324, "step": 8100 }, { "grad_norm": 0.29577282071113586, "learning_rate": 9.928048394855762e-05, "loss": 0.0274, "step": 8110 }, { "grad_norm": 0.24676942825317383, "learning_rate": 9.92769860153325e-05, "loss": 0.0298, "step": 8120 }, { "grad_norm": 0.29635584354400635, "learning_rate": 9.927347966201819e-05, "loss": 0.0223, "step": 8130 }, { "grad_norm": 0.3972901999950409, "learning_rate": 9.926996488921383e-05, "loss": 0.0247, "step": 8140 }, { "grad_norm": 0.2868422269821167, "learning_rate": 9.926644169752001e-05, "loss": 0.0279, "step": 8150 }, { "grad_norm": 0.34539714455604553, "learning_rate": 9.926291008753875e-05, "loss": 0.0287, "step": 8160 }, { "grad_norm": 0.3036886751651764, "learning_rate": 9.92593700598735e-05, "loss": 0.0277, "step": 8170 }, { "grad_norm": 0.3932836353778839, "learning_rate": 9.925582161512915e-05, "loss": 0.03, "step": 8180 }, { "grad_norm": 0.3422134220600128, "learning_rate": 9.925226475391205e-05, "loss": 0.0264, "step": 8190 }, { "grad_norm": 0.2318727672100067, "learning_rate": 9.924869947682993e-05, "loss": 0.0249, "step": 8200 }, { "grad_norm": 0.2565622627735138, "learning_rate": 9.924512578449204e-05, "loss": 0.0323, "step": 8210 }, { "grad_norm": 0.21012654900550842, "learning_rate": 9.924154367750901e-05, "loss": 0.0279, "step": 8220 }, { "grad_norm": 0.28856876492500305, "learning_rate": 9.923795315649293e-05, "loss": 0.0302, "step": 8230 }, { "grad_norm": 0.3310989737510681, "learning_rate": 9.92343542220573e-05, "loss": 0.0299, "step": 8240 }, { "grad_norm": 0.35297778248786926, "learning_rate": 9.92307468748171e-05, "loss": 0.0319, "step": 8250 }, { "grad_norm": 0.41017183661460876, "learning_rate": 9.922713111538873e-05, "loss": 0.0348, "step": 8260 }, { "grad_norm": 0.29625335335731506, "learning_rate": 9.922350694439003e-05, "loss": 0.0274, "step": 8270 }, { "grad_norm": 0.2592681348323822, "learning_rate": 9.921987436244024e-05, "loss": 0.0286, "step": 8280 }, { "grad_norm": 0.33636289834976196, "learning_rate": 9.921623337016008e-05, "loss": 0.028, "step": 8290 }, { "grad_norm": 0.3335561156272888, "learning_rate": 9.921258396817172e-05, "loss": 0.0307, "step": 8300 }, { "grad_norm": 0.37211874127388, "learning_rate": 9.920892615709874e-05, "loss": 0.0251, "step": 8310 }, { "grad_norm": 0.2714262008666992, "learning_rate": 9.920525993756612e-05, "loss": 0.0304, "step": 8320 }, { "grad_norm": 0.26817065477371216, "learning_rate": 9.920158531020036e-05, "loss": 0.0307, "step": 8330 }, { "grad_norm": 0.2614506483078003, "learning_rate": 9.919790227562933e-05, "loss": 0.0279, "step": 8340 }, { "grad_norm": 0.2921057641506195, "learning_rate": 9.919421083448237e-05, "loss": 0.03, "step": 8350 }, { "grad_norm": 0.3357664942741394, "learning_rate": 9.919051098739022e-05, "loss": 0.0278, "step": 8360 }, { "grad_norm": 0.2271486222743988, "learning_rate": 9.918680273498514e-05, "loss": 0.0284, "step": 8370 }, { "grad_norm": 0.32124391198158264, "learning_rate": 9.918308607790072e-05, "loss": 0.0346, "step": 8380 }, { "grad_norm": 0.34780022501945496, "learning_rate": 9.917936101677205e-05, "loss": 0.0256, "step": 8390 }, { "grad_norm": 0.2884010374546051, "learning_rate": 9.917562755223564e-05, "loss": 0.0283, "step": 8400 }, { "grad_norm": 0.2519863545894623, "learning_rate": 9.917188568492944e-05, "loss": 0.0311, "step": 8410 }, { "grad_norm": 0.33243098855018616, "learning_rate": 9.916813541549283e-05, "loss": 0.0232, "step": 8420 }, { "grad_norm": 0.25748321413993835, "learning_rate": 9.916437674456663e-05, "loss": 0.0271, "step": 8430 }, { "grad_norm": 0.2719688415527344, "learning_rate": 9.916060967279308e-05, "loss": 0.0261, "step": 8440 }, { "grad_norm": 0.2374984472990036, "learning_rate": 9.91568342008159e-05, "loss": 0.0282, "step": 8450 }, { "grad_norm": 0.30480173230171204, "learning_rate": 9.915305032928019e-05, "loss": 0.0257, "step": 8460 }, { "grad_norm": 0.266640305519104, "learning_rate": 9.914925805883253e-05, "loss": 0.0258, "step": 8470 }, { "grad_norm": 0.222202867269516, "learning_rate": 9.914545739012088e-05, "loss": 0.0269, "step": 8480 }, { "grad_norm": 0.35895639657974243, "learning_rate": 9.91416483237947e-05, "loss": 0.0306, "step": 8490 }, { "grad_norm": 0.3159054219722748, "learning_rate": 9.913783086050485e-05, "loss": 0.0279, "step": 8500 }, { "grad_norm": 0.3683796525001526, "learning_rate": 9.913400500090364e-05, "loss": 0.0299, "step": 8510 }, { "grad_norm": 0.2813582122325897, "learning_rate": 9.913017074564479e-05, "loss": 0.0283, "step": 8520 }, { "grad_norm": 0.28818055987358093, "learning_rate": 9.912632809538348e-05, "loss": 0.0288, "step": 8530 }, { "grad_norm": 0.4443622827529907, "learning_rate": 9.912247705077629e-05, "loss": 0.0287, "step": 8540 }, { "grad_norm": 0.29019656777381897, "learning_rate": 9.911861761248127e-05, "loss": 0.0295, "step": 8550 }, { "grad_norm": 0.2530854642391205, "learning_rate": 9.91147497811579e-05, "loss": 0.0294, "step": 8560 }, { "grad_norm": 0.2650824189186096, "learning_rate": 9.911087355746709e-05, "loss": 0.025, "step": 8570 }, { "grad_norm": 0.2760615646839142, "learning_rate": 9.910698894207117e-05, "loss": 0.0264, "step": 8580 }, { "grad_norm": 0.21669961512088776, "learning_rate": 9.910309593563392e-05, "loss": 0.026, "step": 8590 }, { "grad_norm": 0.3154946267604828, "learning_rate": 9.909919453882057e-05, "loss": 0.0244, "step": 8600 }, { "grad_norm": 0.2173069268465042, "learning_rate": 9.90952847522977e-05, "loss": 0.0216, "step": 8610 }, { "grad_norm": 0.30173903703689575, "learning_rate": 9.909136657673346e-05, "loss": 0.0265, "step": 8620 }, { "grad_norm": 0.2571437358856201, "learning_rate": 9.908744001279731e-05, "loss": 0.0313, "step": 8630 }, { "grad_norm": 0.31306058168411255, "learning_rate": 9.90835050611602e-05, "loss": 0.0263, "step": 8640 }, { "grad_norm": 0.22509238123893738, "learning_rate": 9.90795617224945e-05, "loss": 0.0252, "step": 8650 }, { "grad_norm": 0.272987425327301, "learning_rate": 9.907560999747405e-05, "loss": 0.0268, "step": 8660 }, { "grad_norm": 0.438658207654953, "learning_rate": 9.907164988677408e-05, "loss": 0.0298, "step": 8670 }, { "grad_norm": 0.3824007511138916, "learning_rate": 9.906768139107124e-05, "loss": 0.0304, "step": 8680 }, { "grad_norm": 0.23737883567810059, "learning_rate": 9.906370451104367e-05, "loss": 0.0293, "step": 8690 }, { "grad_norm": 0.23873737454414368, "learning_rate": 9.905971924737088e-05, "loss": 0.0344, "step": 8700 }, { "grad_norm": 0.2209753692150116, "learning_rate": 9.905572560073387e-05, "loss": 0.0265, "step": 8710 }, { "grad_norm": 0.2742728888988495, "learning_rate": 9.905172357181501e-05, "loss": 0.0273, "step": 8720 }, { "grad_norm": 0.2416391372680664, "learning_rate": 9.904771316129817e-05, "loss": 0.027, "step": 8730 }, { "grad_norm": 0.345609575510025, "learning_rate": 9.904369436986862e-05, "loss": 0.0298, "step": 8740 }, { "grad_norm": 0.3573964536190033, "learning_rate": 9.903966719821303e-05, "loss": 0.0282, "step": 8750 }, { "grad_norm": 0.38187435269355774, "learning_rate": 9.903563164701956e-05, "loss": 0.0336, "step": 8760 }, { "grad_norm": 0.2990255653858185, "learning_rate": 9.903158771697778e-05, "loss": 0.0407, "step": 8770 }, { "grad_norm": 0.32644498348236084, "learning_rate": 9.902753540877867e-05, "loss": 0.0282, "step": 8780 }, { "grad_norm": 0.2971918284893036, "learning_rate": 9.902347472311466e-05, "loss": 0.0254, "step": 8790 }, { "grad_norm": 0.30210667848587036, "learning_rate": 9.901940566067962e-05, "loss": 0.0271, "step": 8800 }, { "grad_norm": 0.28629010915756226, "learning_rate": 9.901532822216883e-05, "loss": 0.0256, "step": 8810 }, { "grad_norm": 0.37315860390663147, "learning_rate": 9.901124240827904e-05, "loss": 0.0353, "step": 8820 }, { "grad_norm": 0.35837310552597046, "learning_rate": 9.900714821970835e-05, "loss": 0.0342, "step": 8830 }, { "grad_norm": 0.31172049045562744, "learning_rate": 9.900304565715641e-05, "loss": 0.0415, "step": 8840 }, { "grad_norm": 0.34397581219673157, "learning_rate": 9.899893472132419e-05, "loss": 0.0355, "step": 8850 }, { "grad_norm": 0.3559442460536957, "learning_rate": 9.899481541291415e-05, "loss": 0.0285, "step": 8860 }, { "grad_norm": 0.2742452323436737, "learning_rate": 9.899068773263016e-05, "loss": 0.0308, "step": 8870 }, { "grad_norm": 0.31892746686935425, "learning_rate": 9.898655168117754e-05, "loss": 0.0309, "step": 8880 }, { "grad_norm": 0.35528331995010376, "learning_rate": 9.898240725926302e-05, "loss": 0.0293, "step": 8890 }, { "grad_norm": 0.31735068559646606, "learning_rate": 9.897825446759478e-05, "loss": 0.035, "step": 8900 }, { "grad_norm": 0.3344995081424713, "learning_rate": 9.897409330688241e-05, "loss": 0.026, "step": 8910 }, { "grad_norm": 0.3340505063533783, "learning_rate": 9.896992377783692e-05, "loss": 0.0288, "step": 8920 }, { "grad_norm": 0.30629783868789673, "learning_rate": 9.89657458811708e-05, "loss": 0.0288, "step": 8930 }, { "grad_norm": 0.30604779720306396, "learning_rate": 9.896155961759792e-05, "loss": 0.0247, "step": 8940 }, { "grad_norm": 0.2330925613641739, "learning_rate": 9.895736498783361e-05, "loss": 0.0297, "step": 8950 }, { "grad_norm": 0.3215080201625824, "learning_rate": 9.895316199259462e-05, "loss": 0.0262, "step": 8960 }, { "grad_norm": 0.3021031320095062, "learning_rate": 9.894895063259909e-05, "loss": 0.0274, "step": 8970 }, { "grad_norm": 0.3261708617210388, "learning_rate": 9.894473090856667e-05, "loss": 0.0359, "step": 8980 }, { "grad_norm": 0.23329214751720428, "learning_rate": 9.894050282121839e-05, "loss": 0.0286, "step": 8990 }, { "grad_norm": 0.2671678364276886, "learning_rate": 9.893626637127668e-05, "loss": 0.023, "step": 9000 }, { "grad_norm": 0.20943698287010193, "learning_rate": 9.893202155946546e-05, "loss": 0.0218, "step": 9010 }, { "grad_norm": 0.25504496693611145, "learning_rate": 9.892776838651006e-05, "loss": 0.0264, "step": 9020 }, { "grad_norm": 0.2531440854072571, "learning_rate": 9.892350685313722e-05, "loss": 0.0249, "step": 9030 }, { "grad_norm": 0.4557083249092102, "learning_rate": 9.891923696007513e-05, "loss": 0.0245, "step": 9040 }, { "grad_norm": 0.33514922857284546, "learning_rate": 9.891495870805336e-05, "loss": 0.0276, "step": 9050 }, { "grad_norm": 0.29246994853019714, "learning_rate": 9.891067209780298e-05, "loss": 0.0313, "step": 9060 }, { "grad_norm": 0.31612539291381836, "learning_rate": 9.890637713005646e-05, "loss": 0.0312, "step": 9070 }, { "grad_norm": 0.24375128746032715, "learning_rate": 9.890207380554767e-05, "loss": 0.0248, "step": 9080 }, { "grad_norm": 0.2713509500026703, "learning_rate": 9.889776212501196e-05, "loss": 0.0262, "step": 9090 }, { "grad_norm": 0.2076655626296997, "learning_rate": 9.889344208918605e-05, "loss": 0.021, "step": 9100 }, { "grad_norm": 0.2124854177236557, "learning_rate": 9.888911369880812e-05, "loss": 0.0217, "step": 9110 }, { "grad_norm": 0.2542179524898529, "learning_rate": 9.888477695461777e-05, "loss": 0.0269, "step": 9120 }, { "grad_norm": 0.17360028624534607, "learning_rate": 9.888043185735607e-05, "loss": 0.0229, "step": 9130 }, { "grad_norm": 0.26906731724739075, "learning_rate": 9.887607840776542e-05, "loss": 0.0243, "step": 9140 }, { "grad_norm": 0.32335883378982544, "learning_rate": 9.887171660658975e-05, "loss": 0.0289, "step": 9150 }, { "grad_norm": 0.2963583171367645, "learning_rate": 9.886734645457435e-05, "loss": 0.0317, "step": 9160 }, { "grad_norm": 0.305130273103714, "learning_rate": 9.886296795246597e-05, "loss": 0.032, "step": 9170 }, { "grad_norm": 0.27309101819992065, "learning_rate": 9.885858110101276e-05, "loss": 0.0293, "step": 9180 }, { "grad_norm": 0.29315125942230225, "learning_rate": 9.885418590096434e-05, "loss": 0.0272, "step": 9190 }, { "grad_norm": 0.3027220666408539, "learning_rate": 9.88497823530717e-05, "loss": 0.0236, "step": 9200 }, { "grad_norm": 0.2568953037261963, "learning_rate": 9.884537045808732e-05, "loss": 0.0273, "step": 9210 }, { "grad_norm": 0.2810177803039551, "learning_rate": 9.884095021676502e-05, "loss": 0.0292, "step": 9220 }, { "grad_norm": 0.21563218533992767, "learning_rate": 9.883652162986017e-05, "loss": 0.0275, "step": 9230 }, { "grad_norm": 0.22166013717651367, "learning_rate": 9.883208469812943e-05, "loss": 0.0229, "step": 9240 }, { "grad_norm": 0.24490536749362946, "learning_rate": 9.882763942233098e-05, "loss": 0.0223, "step": 9250 }, { "grad_norm": 0.3001609742641449, "learning_rate": 9.882318580322441e-05, "loss": 0.0296, "step": 9260 }, { "grad_norm": 0.36867183446884155, "learning_rate": 9.881872384157067e-05, "loss": 0.0311, "step": 9270 }, { "grad_norm": 0.25284022092819214, "learning_rate": 9.881425353813225e-05, "loss": 0.0269, "step": 9280 }, { "grad_norm": 0.27816009521484375, "learning_rate": 9.880977489367296e-05, "loss": 0.0204, "step": 9290 }, { "grad_norm": 0.23137053847312927, "learning_rate": 9.88052879089581e-05, "loss": 0.0296, "step": 9300 }, { "grad_norm": 0.2683916389942169, "learning_rate": 9.880079258475434e-05, "loss": 0.0259, "step": 9310 }, { "grad_norm": 0.3088577091693878, "learning_rate": 9.879628892182985e-05, "loss": 0.0277, "step": 9320 }, { "grad_norm": 0.2414141595363617, "learning_rate": 9.879177692095416e-05, "loss": 0.0201, "step": 9330 }, { "grad_norm": 0.2326025813817978, "learning_rate": 9.878725658289825e-05, "loss": 0.0277, "step": 9340 }, { "grad_norm": 0.27317774295806885, "learning_rate": 9.878272790843454e-05, "loss": 0.0238, "step": 9350 }, { "grad_norm": 0.20756618678569794, "learning_rate": 9.877819089833682e-05, "loss": 0.0204, "step": 9360 }, { "grad_norm": 0.3260330855846405, "learning_rate": 9.877364555338038e-05, "loss": 0.0261, "step": 9370 }, { "grad_norm": 0.29582661390304565, "learning_rate": 9.876909187434186e-05, "loss": 0.022, "step": 9380 }, { "grad_norm": 0.2751789093017578, "learning_rate": 9.876452986199939e-05, "loss": 0.0309, "step": 9390 }, { "grad_norm": 0.2783774733543396, "learning_rate": 9.875995951713248e-05, "loss": 0.0348, "step": 9400 }, { "grad_norm": 0.36822712421417236, "learning_rate": 9.875538084052207e-05, "loss": 0.0245, "step": 9410 }, { "grad_norm": 0.26546820998191833, "learning_rate": 9.875079383295053e-05, "loss": 0.0284, "step": 9420 }, { "grad_norm": 0.2615055739879608, "learning_rate": 9.874619849520167e-05, "loss": 0.0225, "step": 9430 }, { "grad_norm": 0.2364216297864914, "learning_rate": 9.874159482806069e-05, "loss": 0.0221, "step": 9440 }, { "grad_norm": 0.33662593364715576, "learning_rate": 9.873698283231426e-05, "loss": 0.0232, "step": 9450 }, { "grad_norm": 0.28409042954444885, "learning_rate": 9.87323625087504e-05, "loss": 0.0234, "step": 9460 }, { "grad_norm": 0.31340229511260986, "learning_rate": 9.872773385815863e-05, "loss": 0.0244, "step": 9470 }, { "grad_norm": 0.2592726945877075, "learning_rate": 9.872309688132986e-05, "loss": 0.0254, "step": 9480 }, { "grad_norm": 0.250214546918869, "learning_rate": 9.871845157905639e-05, "loss": 0.0225, "step": 9490 }, { "grad_norm": 0.3336656391620636, "learning_rate": 9.871379795213201e-05, "loss": 0.0249, "step": 9500 }, { "grad_norm": 0.29983097314834595, "learning_rate": 9.87091360013519e-05, "loss": 0.0254, "step": 9510 }, { "grad_norm": 0.27386361360549927, "learning_rate": 9.870446572751262e-05, "loss": 0.0268, "step": 9520 }, { "grad_norm": 0.31065633893013, "learning_rate": 9.869978713141224e-05, "loss": 0.0311, "step": 9530 }, { "grad_norm": 0.2354942113161087, "learning_rate": 9.869510021385016e-05, "loss": 0.0303, "step": 9540 }, { "grad_norm": 0.26051679253578186, "learning_rate": 9.869040497562727e-05, "loss": 0.0234, "step": 9550 }, { "grad_norm": 0.2513461709022522, "learning_rate": 9.868570141754587e-05, "loss": 0.0249, "step": 9560 }, { "grad_norm": 0.24648208916187286, "learning_rate": 9.868098954040965e-05, "loss": 0.0242, "step": 9570 }, { "grad_norm": 0.2580796182155609, "learning_rate": 9.867626934502374e-05, "loss": 0.0304, "step": 9580 }, { "grad_norm": 0.2502591907978058, "learning_rate": 9.86715408321947e-05, "loss": 0.026, "step": 9590 }, { "grad_norm": 0.24327465891838074, "learning_rate": 9.86668040027305e-05, "loss": 0.0244, "step": 9600 }, { "grad_norm": 0.23552563786506653, "learning_rate": 9.866205885744053e-05, "loss": 0.0197, "step": 9610 }, { "grad_norm": 0.2684319317340851, "learning_rate": 9.865730539713563e-05, "loss": 0.0241, "step": 9620 }, { "grad_norm": 0.21562506258487701, "learning_rate": 9.8652543622628e-05, "loss": 0.0275, "step": 9630 }, { "grad_norm": 0.24486495554447174, "learning_rate": 9.864777353473132e-05, "loss": 0.024, "step": 9640 }, { "grad_norm": 0.2646384835243225, "learning_rate": 9.864299513426068e-05, "loss": 0.0224, "step": 9650 }, { "grad_norm": 0.19364383816719055, "learning_rate": 9.863820842203254e-05, "loss": 0.0213, "step": 9660 }, { "grad_norm": 0.21984915435314178, "learning_rate": 9.863341339886483e-05, "loss": 0.0215, "step": 9670 }, { "grad_norm": 0.28699374198913574, "learning_rate": 9.86286100655769e-05, "loss": 0.0261, "step": 9680 }, { "grad_norm": 0.27493321895599365, "learning_rate": 9.862379842298953e-05, "loss": 0.0257, "step": 9690 }, { "grad_norm": 0.21421775221824646, "learning_rate": 9.861897847192485e-05, "loss": 0.0231, "step": 9700 }, { "grad_norm": 0.228084534406662, "learning_rate": 9.86141502132065e-05, "loss": 0.0205, "step": 9710 }, { "grad_norm": 0.2654508650302887, "learning_rate": 9.860931364765946e-05, "loss": 0.0242, "step": 9720 }, { "grad_norm": 0.22129681706428528, "learning_rate": 9.860446877611021e-05, "loss": 0.0267, "step": 9730 }, { "grad_norm": 0.2933192253112793, "learning_rate": 9.859961559938655e-05, "loss": 0.0274, "step": 9740 }, { "grad_norm": 0.27277010679244995, "learning_rate": 9.85947541183178e-05, "loss": 0.0237, "step": 9750 }, { "grad_norm": 0.2468322217464447, "learning_rate": 9.858988433373463e-05, "loss": 0.0229, "step": 9760 }, { "grad_norm": 0.3467712104320526, "learning_rate": 9.858500624646918e-05, "loss": 0.0277, "step": 9770 }, { "grad_norm": 0.3158034384250641, "learning_rate": 9.858011985735497e-05, "loss": 0.0232, "step": 9780 }, { "grad_norm": 0.2458934336900711, "learning_rate": 9.857522516722693e-05, "loss": 0.0265, "step": 9790 }, { "grad_norm": 0.21453985571861267, "learning_rate": 9.857032217692145e-05, "loss": 0.0217, "step": 9800 }, { "grad_norm": 0.27818742394447327, "learning_rate": 9.856541088727631e-05, "loss": 0.0205, "step": 9810 }, { "grad_norm": 0.27790743112564087, "learning_rate": 9.856049129913072e-05, "loss": 0.0235, "step": 9820 }, { "grad_norm": 0.3291762173175812, "learning_rate": 9.85555634133253e-05, "loss": 0.0261, "step": 9830 }, { "grad_norm": 0.2756987512111664, "learning_rate": 9.855062723070208e-05, "loss": 0.0254, "step": 9840 }, { "grad_norm": 0.31815406680107117, "learning_rate": 9.854568275210454e-05, "loss": 0.0276, "step": 9850 }, { "grad_norm": 0.29460564255714417, "learning_rate": 9.854072997837754e-05, "loss": 0.0241, "step": 9860 }, { "grad_norm": 0.21283787488937378, "learning_rate": 9.853576891036737e-05, "loss": 0.0198, "step": 9870 }, { "grad_norm": 0.25143834948539734, "learning_rate": 9.853079954892177e-05, "loss": 0.0223, "step": 9880 }, { "grad_norm": 0.36423441767692566, "learning_rate": 9.852582189488983e-05, "loss": 0.026, "step": 9890 }, { "grad_norm": 0.17992782592773438, "learning_rate": 9.852083594912212e-05, "loss": 0.0232, "step": 9900 }, { "grad_norm": 0.2433064728975296, "learning_rate": 9.851584171247058e-05, "loss": 0.0285, "step": 9910 }, { "grad_norm": 0.21741041541099548, "learning_rate": 9.851083918578863e-05, "loss": 0.0211, "step": 9920 }, { "grad_norm": 0.19539839029312134, "learning_rate": 9.850582836993103e-05, "loss": 0.0217, "step": 9930 }, { "grad_norm": 0.15564468502998352, "learning_rate": 9.850080926575397e-05, "loss": 0.0181, "step": 9940 }, { "grad_norm": 0.29753050208091736, "learning_rate": 9.849578187411515e-05, "loss": 0.0322, "step": 9950 }, { "grad_norm": 0.27727940678596497, "learning_rate": 9.849074619587354e-05, "loss": 0.0225, "step": 9960 }, { "grad_norm": 0.3121736943721771, "learning_rate": 9.848570223188964e-05, "loss": 0.0241, "step": 9970 }, { "grad_norm": 0.2871214747428894, "learning_rate": 9.848064998302531e-05, "loss": 0.0273, "step": 9980 }, { "grad_norm": 0.28685495257377625, "learning_rate": 9.847558945014386e-05, "loss": 0.0245, "step": 9990 }, { "grad_norm": 0.23591387271881104, "learning_rate": 9.847052063410996e-05, "loss": 0.0245, "step": 10000 } ], "logging_steps": 10, "max_steps": 80000, "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": 24, "trial_name": null, "trial_params": null }