{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 15000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.4126290082931519, "learning_rate": 6.000000000000001e-07, "loss": 1.4568, "step": 10 }, { "grad_norm": 1.3735394477844238, "learning_rate": 1.2666666666666667e-06, "loss": 1.4539, "step": 20 }, { "grad_norm": 1.1533366441726685, "learning_rate": 1.9333333333333336e-06, "loss": 1.4465, "step": 30 }, { "grad_norm": 0.8541414737701416, "learning_rate": 2.6e-06, "loss": 1.4158, "step": 40 }, { "grad_norm": 0.6532411575317383, "learning_rate": 3.2666666666666666e-06, "loss": 1.366, "step": 50 }, { "grad_norm": 0.43932828307151794, "learning_rate": 3.9333333333333335e-06, "loss": 1.3258, "step": 60 }, { "grad_norm": 0.4106330871582031, "learning_rate": 4.6e-06, "loss": 1.2879, "step": 70 }, { "grad_norm": 0.46380355954170227, "learning_rate": 5.266666666666667e-06, "loss": 1.2307, "step": 80 }, { "grad_norm": 0.39837679266929626, "learning_rate": 5.933333333333334e-06, "loss": 1.1701, "step": 90 }, { "grad_norm": 0.5506791472434998, "learning_rate": 6.6e-06, "loss": 1.1316, "step": 100 }, { "grad_norm": 0.4063026010990143, "learning_rate": 7.266666666666668e-06, "loss": 1.0984, "step": 110 }, { "grad_norm": 0.5370113849639893, "learning_rate": 7.933333333333334e-06, "loss": 1.0844, "step": 120 }, { "grad_norm": 0.517879068851471, "learning_rate": 8.599999999999999e-06, "loss": 1.0775, "step": 130 }, { "grad_norm": 0.4200724959373474, "learning_rate": 9.266666666666667e-06, "loss": 1.0754, "step": 140 }, { "grad_norm": 0.6313542127609253, "learning_rate": 9.933333333333334e-06, "loss": 1.0645, "step": 150 }, { "grad_norm": 0.5384342074394226, "learning_rate": 1.06e-05, "loss": 1.0539, "step": 160 }, { "grad_norm": 0.47652921080589294, "learning_rate": 1.1266666666666667e-05, "loss": 1.0473, "step": 170 }, { "grad_norm": 0.6397068500518799, "learning_rate": 1.1933333333333333e-05, "loss": 1.0471, "step": 180 }, { "grad_norm": 0.33853602409362793, "learning_rate": 1.2600000000000001e-05, "loss": 1.0428, "step": 190 }, { "grad_norm": 0.5751429200172424, "learning_rate": 1.3266666666666666e-05, "loss": 1.0328, "step": 200 }, { "grad_norm": 0.5356873273849487, "learning_rate": 1.3933333333333334e-05, "loss": 1.0211, "step": 210 }, { "grad_norm": 0.5052016973495483, "learning_rate": 1.4599999999999999e-05, "loss": 1.0104, "step": 220 }, { "grad_norm": 0.4714415967464447, "learning_rate": 1.5266666666666667e-05, "loss": 0.9949, "step": 230 }, { "grad_norm": 0.5818572640419006, "learning_rate": 1.5933333333333332e-05, "loss": 0.9771, "step": 240 }, { "grad_norm": 0.6291383504867554, "learning_rate": 1.66e-05, "loss": 0.9692, "step": 250 }, { "grad_norm": 0.43864965438842773, "learning_rate": 1.726666666666667e-05, "loss": 0.9473, "step": 260 }, { "grad_norm": 0.6750267744064331, "learning_rate": 1.7933333333333337e-05, "loss": 0.9305, "step": 270 }, { "grad_norm": 1.1174927949905396, "learning_rate": 1.86e-05, "loss": 0.9073, "step": 280 }, { "grad_norm": 0.7326529622077942, "learning_rate": 1.926666666666667e-05, "loss": 0.8804, "step": 290 }, { "grad_norm": 0.7365794777870178, "learning_rate": 1.9933333333333334e-05, "loss": 0.8393, "step": 300 }, { "grad_norm": 0.6675243377685547, "learning_rate": 2.06e-05, "loss": 0.8023, "step": 310 }, { "grad_norm": 0.926457405090332, "learning_rate": 2.1266666666666667e-05, "loss": 0.7797, "step": 320 }, { "grad_norm": 0.7966919541358948, "learning_rate": 2.1933333333333332e-05, "loss": 0.7535, "step": 330 }, { "grad_norm": 0.7804962992668152, "learning_rate": 2.26e-05, "loss": 0.7258, "step": 340 }, { "grad_norm": 0.9337165951728821, "learning_rate": 2.326666666666667e-05, "loss": 0.6942, "step": 350 }, { "grad_norm": 0.7417464852333069, "learning_rate": 2.3933333333333337e-05, "loss": 0.6635, "step": 360 }, { "grad_norm": 1.0197713375091553, "learning_rate": 2.46e-05, "loss": 0.6353, "step": 370 }, { "grad_norm": 1.0989516973495483, "learning_rate": 2.5266666666666666e-05, "loss": 0.6127, "step": 380 }, { "grad_norm": 1.1899223327636719, "learning_rate": 2.5933333333333338e-05, "loss": 0.5922, "step": 390 }, { "grad_norm": 0.9237262010574341, "learning_rate": 2.6600000000000003e-05, "loss": 0.5758, "step": 400 }, { "grad_norm": 1.4765948057174683, "learning_rate": 2.7266666666666668e-05, "loss": 0.5479, "step": 410 }, { "grad_norm": 0.9667250514030457, "learning_rate": 2.7933333333333332e-05, "loss": 0.5333, "step": 420 }, { "grad_norm": 0.9509170651435852, "learning_rate": 2.86e-05, "loss": 0.5143, "step": 430 }, { "grad_norm": 1.2793920040130615, "learning_rate": 2.926666666666667e-05, "loss": 0.4968, "step": 440 }, { "grad_norm": 0.9938004016876221, "learning_rate": 2.9933333333333337e-05, "loss": 0.4869, "step": 450 }, { "grad_norm": 0.9902573227882385, "learning_rate": 3.06e-05, "loss": 0.4659, "step": 460 }, { "grad_norm": 0.9396094679832458, "learning_rate": 3.126666666666666e-05, "loss": 0.4555, "step": 470 }, { "grad_norm": 1.2180593013763428, "learning_rate": 3.1933333333333335e-05, "loss": 0.4367, "step": 480 }, { "grad_norm": 1.2830370664596558, "learning_rate": 3.26e-05, "loss": 0.4122, "step": 490 }, { "grad_norm": 1.0384615659713745, "learning_rate": 3.326666666666667e-05, "loss": 0.3994, "step": 500 }, { "grad_norm": 1.0901414155960083, "learning_rate": 3.3933333333333336e-05, "loss": 0.3788, "step": 510 }, { "grad_norm": 1.0764013528823853, "learning_rate": 3.46e-05, "loss": 0.36, "step": 520 }, { "grad_norm": 1.283591628074646, "learning_rate": 3.526666666666667e-05, "loss": 0.3453, "step": 530 }, { "grad_norm": 1.4109214544296265, "learning_rate": 3.593333333333334e-05, "loss": 0.3376, "step": 540 }, { "grad_norm": 1.2783732414245605, "learning_rate": 3.66e-05, "loss": 0.3195, "step": 550 }, { "grad_norm": 1.2349673509597778, "learning_rate": 3.726666666666667e-05, "loss": 0.3045, "step": 560 }, { "grad_norm": 1.6351103782653809, "learning_rate": 3.793333333333334e-05, "loss": 0.297, "step": 570 }, { "grad_norm": 1.413041591644287, "learning_rate": 3.86e-05, "loss": 0.2832, "step": 580 }, { "grad_norm": 1.7158143520355225, "learning_rate": 3.926666666666667e-05, "loss": 0.2675, "step": 590 }, { "grad_norm": 1.5770541429519653, "learning_rate": 3.993333333333333e-05, "loss": 0.2511, "step": 600 }, { "grad_norm": 1.2448288202285767, "learning_rate": 4.0600000000000004e-05, "loss": 0.2385, "step": 610 }, { "grad_norm": 1.0363948345184326, "learning_rate": 4.126666666666667e-05, "loss": 0.2324, "step": 620 }, { "grad_norm": 1.3222676515579224, "learning_rate": 4.1933333333333334e-05, "loss": 0.2232, "step": 630 }, { "grad_norm": 1.592990756034851, "learning_rate": 4.26e-05, "loss": 0.2109, "step": 640 }, { "grad_norm": 1.5724996328353882, "learning_rate": 4.3266666666666664e-05, "loss": 0.1945, "step": 650 }, { "grad_norm": 1.4436169862747192, "learning_rate": 4.3933333333333335e-05, "loss": 0.1779, "step": 660 }, { "grad_norm": 1.4387675523757935, "learning_rate": 4.46e-05, "loss": 0.1764, "step": 670 }, { "grad_norm": 1.2351089715957642, "learning_rate": 4.526666666666667e-05, "loss": 0.1615, "step": 680 }, { "grad_norm": 1.2490671873092651, "learning_rate": 4.5933333333333336e-05, "loss": 0.1529, "step": 690 }, { "grad_norm": 1.2988910675048828, "learning_rate": 4.660000000000001e-05, "loss": 0.1524, "step": 700 }, { "grad_norm": 1.1097872257232666, "learning_rate": 4.726666666666667e-05, "loss": 0.1448, "step": 710 }, { "grad_norm": 1.7465487718582153, "learning_rate": 4.793333333333334e-05, "loss": 0.1437, "step": 720 }, { "grad_norm": 1.3112348318099976, "learning_rate": 4.86e-05, "loss": 0.143, "step": 730 }, { "grad_norm": 1.1917496919631958, "learning_rate": 4.926666666666667e-05, "loss": 0.1295, "step": 740 }, { "grad_norm": 1.729387640953064, "learning_rate": 4.993333333333334e-05, "loss": 0.123, "step": 750 }, { "grad_norm": 1.7255092859268188, "learning_rate": 5.0600000000000003e-05, "loss": 0.1219, "step": 760 }, { "grad_norm": 1.6423131227493286, "learning_rate": 5.1266666666666675e-05, "loss": 0.1212, "step": 770 }, { "grad_norm": 1.2457637786865234, "learning_rate": 5.193333333333333e-05, "loss": 0.1153, "step": 780 }, { "grad_norm": 1.7827743291854858, "learning_rate": 5.2600000000000005e-05, "loss": 0.1132, "step": 790 }, { "grad_norm": 1.400822639465332, "learning_rate": 5.326666666666666e-05, "loss": 0.114, "step": 800 }, { "grad_norm": 1.4029743671417236, "learning_rate": 5.3933333333333334e-05, "loss": 0.1042, "step": 810 }, { "grad_norm": 1.3635342121124268, "learning_rate": 5.4600000000000006e-05, "loss": 0.1042, "step": 820 }, { "grad_norm": 2.1806066036224365, "learning_rate": 5.5266666666666664e-05, "loss": 0.0922, "step": 830 }, { "grad_norm": 1.3689796924591064, "learning_rate": 5.5933333333333335e-05, "loss": 0.0904, "step": 840 }, { "grad_norm": 1.4890871047973633, "learning_rate": 5.66e-05, "loss": 0.0928, "step": 850 }, { "grad_norm": 1.2499778270721436, "learning_rate": 5.726666666666667e-05, "loss": 0.0846, "step": 860 }, { "grad_norm": 1.0627532005310059, "learning_rate": 5.7933333333333337e-05, "loss": 0.081, "step": 870 }, { "grad_norm": 1.0695182085037231, "learning_rate": 5.86e-05, "loss": 0.0824, "step": 880 }, { "grad_norm": 1.68373441696167, "learning_rate": 5.926666666666667e-05, "loss": 0.0821, "step": 890 }, { "grad_norm": 1.0833736658096313, "learning_rate": 5.9933333333333345e-05, "loss": 0.0802, "step": 900 }, { "grad_norm": 1.1082783937454224, "learning_rate": 6.06e-05, "loss": 0.0766, "step": 910 }, { "grad_norm": 1.22420072555542, "learning_rate": 6.126666666666667e-05, "loss": 0.0772, "step": 920 }, { "grad_norm": 1.7994489669799805, "learning_rate": 6.193333333333333e-05, "loss": 0.0785, "step": 930 }, { "grad_norm": 0.9922900795936584, "learning_rate": 6.26e-05, "loss": 0.0775, "step": 940 }, { "grad_norm": 1.074314832687378, "learning_rate": 6.326666666666667e-05, "loss": 0.0761, "step": 950 }, { "grad_norm": 1.3164290189743042, "learning_rate": 6.393333333333333e-05, "loss": 0.0739, "step": 960 }, { "grad_norm": 1.3246431350708008, "learning_rate": 6.460000000000001e-05, "loss": 0.0764, "step": 970 }, { "grad_norm": 1.1399774551391602, "learning_rate": 6.526666666666666e-05, "loss": 0.0757, "step": 980 }, { "grad_norm": 0.9545037746429443, "learning_rate": 6.593333333333334e-05, "loss": 0.0769, "step": 990 }, { "grad_norm": 0.9759553670883179, "learning_rate": 6.66e-05, "loss": 0.0745, "step": 1000 }, { "grad_norm": 1.342103123664856, "learning_rate": 6.726666666666667e-05, "loss": 0.073, "step": 1010 }, { "grad_norm": 1.022963523864746, "learning_rate": 6.793333333333334e-05, "loss": 0.0711, "step": 1020 }, { "grad_norm": 1.0277061462402344, "learning_rate": 6.860000000000001e-05, "loss": 0.0714, "step": 1030 }, { "grad_norm": 1.4367462396621704, "learning_rate": 6.926666666666667e-05, "loss": 0.0708, "step": 1040 }, { "grad_norm": 1.2014946937561035, "learning_rate": 6.993333333333334e-05, "loss": 0.0705, "step": 1050 }, { "grad_norm": 1.111018419265747, "learning_rate": 7.06e-05, "loss": 0.0647, "step": 1060 }, { "grad_norm": 1.1501191854476929, "learning_rate": 7.126666666666667e-05, "loss": 0.0669, "step": 1070 }, { "grad_norm": 0.975878119468689, "learning_rate": 7.193333333333334e-05, "loss": 0.0648, "step": 1080 }, { "grad_norm": 1.1001148223876953, "learning_rate": 7.26e-05, "loss": 0.0662, "step": 1090 }, { "grad_norm": 0.9814893007278442, "learning_rate": 7.326666666666667e-05, "loss": 0.0679, "step": 1100 }, { "grad_norm": 1.2277650833129883, "learning_rate": 7.393333333333333e-05, "loss": 0.0688, "step": 1110 }, { "grad_norm": 0.9856839776039124, "learning_rate": 7.46e-05, "loss": 0.0648, "step": 1120 }, { "grad_norm": 1.1703674793243408, "learning_rate": 7.526666666666668e-05, "loss": 0.0687, "step": 1130 }, { "grad_norm": 1.1658049821853638, "learning_rate": 7.593333333333334e-05, "loss": 0.0646, "step": 1140 }, { "grad_norm": 1.1635832786560059, "learning_rate": 7.66e-05, "loss": 0.0657, "step": 1150 }, { "grad_norm": 1.1898596286773682, "learning_rate": 7.726666666666667e-05, "loss": 0.0679, "step": 1160 }, { "grad_norm": 1.04132878780365, "learning_rate": 7.793333333333333e-05, "loss": 0.0674, "step": 1170 }, { "grad_norm": 0.9637924432754517, "learning_rate": 7.860000000000001e-05, "loss": 0.0666, "step": 1180 }, { "grad_norm": 1.0488718748092651, "learning_rate": 7.926666666666666e-05, "loss": 0.0638, "step": 1190 }, { "grad_norm": 0.9946198463439941, "learning_rate": 7.993333333333334e-05, "loss": 0.0639, "step": 1200 }, { "grad_norm": 0.9626931548118591, "learning_rate": 8.060000000000001e-05, "loss": 0.0638, "step": 1210 }, { "grad_norm": 0.8904481530189514, "learning_rate": 8.126666666666667e-05, "loss": 0.0614, "step": 1220 }, { "grad_norm": 1.0349605083465576, "learning_rate": 8.193333333333334e-05, "loss": 0.0606, "step": 1230 }, { "grad_norm": 1.2032581567764282, "learning_rate": 8.26e-05, "loss": 0.06, "step": 1240 }, { "grad_norm": 1.0489283800125122, "learning_rate": 8.326666666666667e-05, "loss": 0.0629, "step": 1250 }, { "grad_norm": 0.8637824058532715, "learning_rate": 8.393333333333335e-05, "loss": 0.0606, "step": 1260 }, { "grad_norm": 0.9523459672927856, "learning_rate": 8.46e-05, "loss": 0.0608, "step": 1270 }, { "grad_norm": 0.910315990447998, "learning_rate": 8.526666666666667e-05, "loss": 0.0629, "step": 1280 }, { "grad_norm": 1.4048068523406982, "learning_rate": 8.593333333333333e-05, "loss": 0.0598, "step": 1290 }, { "grad_norm": 0.8386713862419128, "learning_rate": 8.66e-05, "loss": 0.0622, "step": 1300 }, { "grad_norm": 0.864817202091217, "learning_rate": 8.726666666666667e-05, "loss": 0.0598, "step": 1310 }, { "grad_norm": 0.9372347593307495, "learning_rate": 8.793333333333333e-05, "loss": 0.0594, "step": 1320 }, { "grad_norm": 0.9286304116249084, "learning_rate": 8.86e-05, "loss": 0.0621, "step": 1330 }, { "grad_norm": 0.8566858172416687, "learning_rate": 8.926666666666668e-05, "loss": 0.0587, "step": 1340 }, { "grad_norm": 1.0118904113769531, "learning_rate": 8.993333333333334e-05, "loss": 0.058, "step": 1350 }, { "grad_norm": 0.9429856538772583, "learning_rate": 9.06e-05, "loss": 0.0559, "step": 1360 }, { "grad_norm": 0.8768374919891357, "learning_rate": 9.126666666666667e-05, "loss": 0.0567, "step": 1370 }, { "grad_norm": 0.8839522004127502, "learning_rate": 9.193333333333334e-05, "loss": 0.0544, "step": 1380 }, { "grad_norm": 0.9084612131118774, "learning_rate": 9.260000000000001e-05, "loss": 0.0565, "step": 1390 }, { "grad_norm": 0.9119452238082886, "learning_rate": 9.326666666666667e-05, "loss": 0.0558, "step": 1400 }, { "grad_norm": 0.9000705480575562, "learning_rate": 9.393333333333334e-05, "loss": 0.0579, "step": 1410 }, { "grad_norm": 0.8316766023635864, "learning_rate": 9.46e-05, "loss": 0.0527, "step": 1420 }, { "grad_norm": 1.1867926120758057, "learning_rate": 9.526666666666667e-05, "loss": 0.0576, "step": 1430 }, { "grad_norm": 0.8146836161613464, "learning_rate": 9.593333333333334e-05, "loss": 0.054, "step": 1440 }, { "grad_norm": 1.0247355699539185, "learning_rate": 9.66e-05, "loss": 0.0543, "step": 1450 }, { "grad_norm": 0.990510880947113, "learning_rate": 9.726666666666667e-05, "loss": 0.0543, "step": 1460 }, { "grad_norm": 0.8735817074775696, "learning_rate": 9.793333333333333e-05, "loss": 0.0589, "step": 1470 }, { "grad_norm": 0.981984555721283, "learning_rate": 9.86e-05, "loss": 0.055, "step": 1480 }, { "grad_norm": 0.7862304449081421, "learning_rate": 9.926666666666668e-05, "loss": 0.051, "step": 1490 }, { "grad_norm": 1.1150578260421753, "learning_rate": 9.993333333333334e-05, "loss": 0.0527, "step": 1500 }, { "grad_norm": 0.7199082970619202, "learning_rate": 9.999997539434007e-05, "loss": 0.0551, "step": 1510 }, { "grad_norm": 0.9026912450790405, "learning_rate": 9.999989033776898e-05, "loss": 0.0533, "step": 1520 }, { "grad_norm": 0.6898888349533081, "learning_rate": 9.999974452661641e-05, "loss": 0.0532, "step": 1530 }, { "grad_norm": 0.8844292759895325, "learning_rate": 9.999953796105959e-05, "loss": 0.0521, "step": 1540 }, { "grad_norm": 0.9320904612541199, "learning_rate": 9.999927064134949e-05, "loss": 0.0556, "step": 1550 }, { "grad_norm": 0.9313002228736877, "learning_rate": 9.999894256781095e-05, "loss": 0.0541, "step": 1560 }, { "grad_norm": 0.8291693329811096, "learning_rate": 9.99985537408426e-05, "loss": 0.0536, "step": 1570 }, { "grad_norm": 0.5824195742607117, "learning_rate": 9.999810416091688e-05, "loss": 0.0501, "step": 1580 }, { "grad_norm": 0.7193527221679688, "learning_rate": 9.99975938285801e-05, "loss": 0.0488, "step": 1590 }, { "grad_norm": 0.6793032884597778, "learning_rate": 9.999702274445236e-05, "loss": 0.0504, "step": 1600 }, { "grad_norm": 0.8828063011169434, "learning_rate": 9.999639090922756e-05, "loss": 0.0491, "step": 1610 }, { "grad_norm": 0.9317867159843445, "learning_rate": 9.999569832367346e-05, "loss": 0.0483, "step": 1620 }, { "grad_norm": 0.7257526516914368, "learning_rate": 9.999494498863162e-05, "loss": 0.0501, "step": 1630 }, { "grad_norm": 1.0141751766204834, "learning_rate": 9.99941309050174e-05, "loss": 0.0502, "step": 1640 }, { "grad_norm": 0.7642090320587158, "learning_rate": 9.999325607381999e-05, "loss": 0.0539, "step": 1650 }, { "grad_norm": 0.6591696739196777, "learning_rate": 9.999232049610238e-05, "loss": 0.0492, "step": 1660 }, { "grad_norm": 0.8247687220573425, "learning_rate": 9.999132417300142e-05, "loss": 0.0459, "step": 1670 }, { "grad_norm": 0.912160336971283, "learning_rate": 9.99902671057277e-05, "loss": 0.0475, "step": 1680 }, { "grad_norm": 0.7522016763687134, "learning_rate": 9.998914929556569e-05, "loss": 0.0493, "step": 1690 }, { "grad_norm": 0.7177233099937439, "learning_rate": 9.998797074387361e-05, "loss": 0.0475, "step": 1700 }, { "grad_norm": 0.8339974880218506, "learning_rate": 9.99867314520835e-05, "loss": 0.0446, "step": 1710 }, { "grad_norm": 0.8249689340591431, "learning_rate": 9.998543142170126e-05, "loss": 0.0495, "step": 1720 }, { "grad_norm": 0.7558274865150452, "learning_rate": 9.99840706543065e-05, "loss": 0.0493, "step": 1730 }, { "grad_norm": 0.6998085379600525, "learning_rate": 9.998264915155274e-05, "loss": 0.0452, "step": 1740 }, { "grad_norm": 0.7514432072639465, "learning_rate": 9.998116691516718e-05, "loss": 0.0455, "step": 1750 }, { "grad_norm": 0.9076456427574158, "learning_rate": 9.997962394695091e-05, "loss": 0.0459, "step": 1760 }, { "grad_norm": 0.5168872475624084, "learning_rate": 9.997802024877875e-05, "loss": 0.0465, "step": 1770 }, { "grad_norm": 0.6847232580184937, "learning_rate": 9.99763558225994e-05, "loss": 0.046, "step": 1780 }, { "grad_norm": 0.6557778120040894, "learning_rate": 9.997463067043526e-05, "loss": 0.0485, "step": 1790 }, { "grad_norm": 0.7632912993431091, "learning_rate": 9.997284479438253e-05, "loss": 0.0466, "step": 1800 }, { "grad_norm": 0.8196837902069092, "learning_rate": 9.997099819661127e-05, "loss": 0.0445, "step": 1810 }, { "grad_norm": 0.7271484732627869, "learning_rate": 9.996909087936524e-05, "loss": 0.0438, "step": 1820 }, { "grad_norm": 0.7890754342079163, "learning_rate": 9.996712284496201e-05, "loss": 0.0461, "step": 1830 }, { "grad_norm": 0.6526637077331543, "learning_rate": 9.996509409579293e-05, "loss": 0.0444, "step": 1840 }, { "grad_norm": 0.8112354874610901, "learning_rate": 9.996300463432312e-05, "loss": 0.0452, "step": 1850 }, { "grad_norm": 0.6816654205322266, "learning_rate": 9.996085446309148e-05, "loss": 0.0457, "step": 1860 }, { "grad_norm": 0.7028149962425232, "learning_rate": 9.995864358471066e-05, "loss": 0.0449, "step": 1870 }, { "grad_norm": 0.7199757695198059, "learning_rate": 9.99563720018671e-05, "loss": 0.0441, "step": 1880 }, { "grad_norm": 0.6490606069564819, "learning_rate": 9.995403971732098e-05, "loss": 0.043, "step": 1890 }, { "grad_norm": 0.8152973651885986, "learning_rate": 9.995164673390625e-05, "loss": 0.0457, "step": 1900 }, { "grad_norm": 0.769864022731781, "learning_rate": 9.994919305453059e-05, "loss": 0.0425, "step": 1910 }, { "grad_norm": 0.6499236226081848, "learning_rate": 9.994667868217548e-05, "loss": 0.0435, "step": 1920 }, { "grad_norm": 0.7776778936386108, "learning_rate": 9.99441036198961e-05, "loss": 0.0429, "step": 1930 }, { "grad_norm": 0.7239458560943604, "learning_rate": 9.99414678708214e-05, "loss": 0.0435, "step": 1940 }, { "grad_norm": 0.601598858833313, "learning_rate": 9.993877143815407e-05, "loss": 0.042, "step": 1950 }, { "grad_norm": 0.7758349180221558, "learning_rate": 9.993601432517053e-05, "loss": 0.0476, "step": 1960 }, { "grad_norm": 0.6978219151496887, "learning_rate": 9.993319653522091e-05, "loss": 0.0418, "step": 1970 }, { "grad_norm": 0.7210837006568909, "learning_rate": 9.993031807172911e-05, "loss": 0.0428, "step": 1980 }, { "grad_norm": 0.7555508017539978, "learning_rate": 9.992737893819273e-05, "loss": 0.0468, "step": 1990 }, { "grad_norm": 0.6775459051132202, "learning_rate": 9.992437913818312e-05, "loss": 0.0447, "step": 2000 }, { "grad_norm": 0.7015236020088196, "learning_rate": 9.992131867534526e-05, "loss": 0.0426, "step": 2010 }, { "grad_norm": 0.7684049010276794, "learning_rate": 9.991819755339796e-05, "loss": 0.0425, "step": 2020 }, { "grad_norm": 0.6014548540115356, "learning_rate": 9.991501577613365e-05, "loss": 0.0428, "step": 2030 }, { "grad_norm": 0.6015975475311279, "learning_rate": 9.99117733474185e-05, "loss": 0.0412, "step": 2040 }, { "grad_norm": 0.6926084756851196, "learning_rate": 9.990847027119234e-05, "loss": 0.0409, "step": 2050 }, { "grad_norm": 0.5292817950248718, "learning_rate": 9.990510655146877e-05, "loss": 0.0404, "step": 2060 }, { "grad_norm": 0.842787504196167, "learning_rate": 9.990168219233496e-05, "loss": 0.0425, "step": 2070 }, { "grad_norm": 0.7378549575805664, "learning_rate": 9.989819719795188e-05, "loss": 0.0427, "step": 2080 }, { "grad_norm": 0.6256470680236816, "learning_rate": 9.989465157255412e-05, "loss": 0.0409, "step": 2090 }, { "grad_norm": 0.6263523697853088, "learning_rate": 9.989104532044994e-05, "loss": 0.0408, "step": 2100 }, { "grad_norm": 0.7206677794456482, "learning_rate": 9.988737844602128e-05, "loss": 0.0404, "step": 2110 }, { "grad_norm": 0.7646312713623047, "learning_rate": 9.988365095372372e-05, "loss": 0.0411, "step": 2120 }, { "grad_norm": 0.6586979031562805, "learning_rate": 9.987986284808654e-05, "loss": 0.0409, "step": 2130 }, { "grad_norm": 0.6952325701713562, "learning_rate": 9.987601413371264e-05, "loss": 0.0387, "step": 2140 }, { "grad_norm": 0.6108909249305725, "learning_rate": 9.987210481527855e-05, "loss": 0.0401, "step": 2150 }, { "grad_norm": 0.618565022945404, "learning_rate": 9.98681348975345e-05, "loss": 0.041, "step": 2160 }, { "grad_norm": 0.6037882566452026, "learning_rate": 9.986410438530427e-05, "loss": 0.0414, "step": 2170 }, { "grad_norm": 0.685737669467926, "learning_rate": 9.986001328348534e-05, "loss": 0.0379, "step": 2180 }, { "grad_norm": 0.6301174163818359, "learning_rate": 9.985586159704878e-05, "loss": 0.0415, "step": 2190 }, { "grad_norm": 0.7248358726501465, "learning_rate": 9.985164933103929e-05, "loss": 0.0407, "step": 2200 }, { "grad_norm": 0.6090236306190491, "learning_rate": 9.984737649057513e-05, "loss": 0.0404, "step": 2210 }, { "grad_norm": 0.6727856993675232, "learning_rate": 9.984304308084827e-05, "loss": 0.0407, "step": 2220 }, { "grad_norm": 0.5758669376373291, "learning_rate": 9.983864910712416e-05, "loss": 0.0392, "step": 2230 }, { "grad_norm": 0.6936368942260742, "learning_rate": 9.98341945747419e-05, "loss": 0.0396, "step": 2240 }, { "grad_norm": 0.7542185187339783, "learning_rate": 9.98296794891142e-05, "loss": 0.0392, "step": 2250 }, { "grad_norm": 0.7693107724189758, "learning_rate": 9.982510385572725e-05, "loss": 0.0371, "step": 2260 }, { "grad_norm": 0.6100176572799683, "learning_rate": 9.982046768014094e-05, "loss": 0.0383, "step": 2270 }, { "grad_norm": 0.8347607254981995, "learning_rate": 9.981577096798863e-05, "loss": 0.0398, "step": 2280 }, { "grad_norm": 0.7694611549377441, "learning_rate": 9.981101372497727e-05, "loss": 0.0393, "step": 2290 }, { "grad_norm": 0.7548644542694092, "learning_rate": 9.980619595688737e-05, "loss": 0.04, "step": 2300 }, { "grad_norm": 0.5724036693572998, "learning_rate": 9.980131766957295e-05, "loss": 0.0394, "step": 2310 }, { "grad_norm": 0.6039924621582031, "learning_rate": 9.979637886896163e-05, "loss": 0.038, "step": 2320 }, { "grad_norm": 0.5768024921417236, "learning_rate": 9.979137956105447e-05, "loss": 0.0392, "step": 2330 }, { "grad_norm": 0.6355780959129333, "learning_rate": 9.978631975192613e-05, "loss": 0.0384, "step": 2340 }, { "grad_norm": 0.7645574808120728, "learning_rate": 9.978119944772475e-05, "loss": 0.0409, "step": 2350 }, { "grad_norm": 0.6496811509132385, "learning_rate": 9.977601865467197e-05, "loss": 0.0393, "step": 2360 }, { "grad_norm": 0.5277372002601624, "learning_rate": 9.977077737906297e-05, "loss": 0.038, "step": 2370 }, { "grad_norm": 0.7123770117759705, "learning_rate": 9.976547562726636e-05, "loss": 0.0411, "step": 2380 }, { "grad_norm": 0.6734181642532349, "learning_rate": 9.976011340572429e-05, "loss": 0.0416, "step": 2390 }, { "grad_norm": 0.5722440481185913, "learning_rate": 9.975469072095237e-05, "loss": 0.039, "step": 2400 }, { "grad_norm": 0.7019569873809814, "learning_rate": 9.974920757953965e-05, "loss": 0.0403, "step": 2410 }, { "grad_norm": 0.6497634649276733, "learning_rate": 9.97436639881487e-05, "loss": 0.039, "step": 2420 }, { "grad_norm": 0.44458356499671936, "learning_rate": 9.973805995351545e-05, "loss": 0.0377, "step": 2430 }, { "grad_norm": 0.6675490140914917, "learning_rate": 9.973239548244939e-05, "loss": 0.0369, "step": 2440 }, { "grad_norm": 0.5237282514572144, "learning_rate": 9.972667058183333e-05, "loss": 0.0367, "step": 2450 }, { "grad_norm": 0.6908047795295715, "learning_rate": 9.972088525862362e-05, "loss": 0.0388, "step": 2460 }, { "grad_norm": 0.5526571273803711, "learning_rate": 9.971503951984995e-05, "loss": 0.0386, "step": 2470 }, { "grad_norm": 0.8267020583152771, "learning_rate": 9.970913337261543e-05, "loss": 0.0409, "step": 2480 }, { "grad_norm": 0.6126002073287964, "learning_rate": 9.97031668240966e-05, "loss": 0.0385, "step": 2490 }, { "grad_norm": 0.6993176341056824, "learning_rate": 9.969713988154339e-05, "loss": 0.0392, "step": 2500 }, { "grad_norm": 0.711223304271698, "learning_rate": 9.969105255227906e-05, "loss": 0.0394, "step": 2510 }, { "grad_norm": 0.5168538093566895, "learning_rate": 9.968490484370035e-05, "loss": 0.0372, "step": 2520 }, { "grad_norm": 0.6115143299102783, "learning_rate": 9.967869676327726e-05, "loss": 0.036, "step": 2530 }, { "grad_norm": 0.57094407081604, "learning_rate": 9.96724283185532e-05, "loss": 0.0352, "step": 2540 }, { "grad_norm": 0.5119908452033997, "learning_rate": 9.966609951714494e-05, "loss": 0.0362, "step": 2550 }, { "grad_norm": 0.7033934593200684, "learning_rate": 9.965971036674255e-05, "loss": 0.0373, "step": 2560 }, { "grad_norm": 0.6613783240318298, "learning_rate": 9.965326087510947e-05, "loss": 0.0391, "step": 2570 }, { "grad_norm": 0.6085628867149353, "learning_rate": 9.964675105008244e-05, "loss": 0.0376, "step": 2580 }, { "grad_norm": 0.5930835604667664, "learning_rate": 9.964018089957147e-05, "loss": 0.0344, "step": 2590 }, { "grad_norm": 0.5690402388572693, "learning_rate": 9.963355043155996e-05, "loss": 0.0368, "step": 2600 }, { "grad_norm": 0.6252157092094421, "learning_rate": 9.962685965410455e-05, "loss": 0.0383, "step": 2610 }, { "grad_norm": 0.690443754196167, "learning_rate": 9.962010857533514e-05, "loss": 0.0369, "step": 2620 }, { "grad_norm": 0.6155582070350647, "learning_rate": 9.961329720345493e-05, "loss": 0.0371, "step": 2630 }, { "grad_norm": 0.6473574042320251, "learning_rate": 9.96064255467404e-05, "loss": 0.0375, "step": 2640 }, { "grad_norm": 0.6162963509559631, "learning_rate": 9.959949361354126e-05, "loss": 0.0364, "step": 2650 }, { "grad_norm": 0.6339418888092041, "learning_rate": 9.959250141228045e-05, "loss": 0.0356, "step": 2660 }, { "grad_norm": 0.4604366719722748, "learning_rate": 9.958544895145414e-05, "loss": 0.0374, "step": 2670 }, { "grad_norm": 0.5806917548179626, "learning_rate": 9.957833623963177e-05, "loss": 0.0353, "step": 2680 }, { "grad_norm": 0.653438925743103, "learning_rate": 9.957116328545593e-05, "loss": 0.0351, "step": 2690 }, { "grad_norm": 0.6917750835418701, "learning_rate": 9.956393009764244e-05, "loss": 0.0353, "step": 2700 }, { "grad_norm": 0.626438558101654, "learning_rate": 9.955663668498032e-05, "loss": 0.0355, "step": 2710 }, { "grad_norm": 0.478832870721817, "learning_rate": 9.954928305633173e-05, "loss": 0.034, "step": 2720 }, { "grad_norm": 0.5281680226325989, "learning_rate": 9.954186922063204e-05, "loss": 0.0337, "step": 2730 }, { "grad_norm": 0.6203818917274475, "learning_rate": 9.953439518688974e-05, "loss": 0.0349, "step": 2740 }, { "grad_norm": 0.5640031695365906, "learning_rate": 9.952686096418652e-05, "loss": 0.0365, "step": 2750 }, { "grad_norm": 0.5402900576591492, "learning_rate": 9.951926656167714e-05, "loss": 0.0346, "step": 2760 }, { "grad_norm": 0.6839205026626587, "learning_rate": 9.951161198858953e-05, "loss": 0.0352, "step": 2770 }, { "grad_norm": 0.5779417753219604, "learning_rate": 9.950389725422471e-05, "loss": 0.0373, "step": 2780 }, { "grad_norm": 0.6267558336257935, "learning_rate": 9.949612236795682e-05, "loss": 0.0384, "step": 2790 }, { "grad_norm": 0.5585941076278687, "learning_rate": 9.948828733923305e-05, "loss": 0.0375, "step": 2800 }, { "grad_norm": 0.7183846831321716, "learning_rate": 9.948039217757374e-05, "loss": 0.0369, "step": 2810 }, { "grad_norm": 0.6528863906860352, "learning_rate": 9.947243689257225e-05, "loss": 0.0355, "step": 2820 }, { "grad_norm": 0.43302688002586365, "learning_rate": 9.946442149389497e-05, "loss": 0.0346, "step": 2830 }, { "grad_norm": 0.4747207462787628, "learning_rate": 9.945634599128139e-05, "loss": 0.0353, "step": 2840 }, { "grad_norm": 0.516400158405304, "learning_rate": 9.944821039454402e-05, "loss": 0.034, "step": 2850 }, { "grad_norm": 0.5410118699073792, "learning_rate": 9.944001471356835e-05, "loss": 0.034, "step": 2860 }, { "grad_norm": 0.5934143662452698, "learning_rate": 9.94317589583129e-05, "loss": 0.0376, "step": 2870 }, { "grad_norm": 0.5651362538337708, "learning_rate": 9.942344313880922e-05, "loss": 0.0345, "step": 2880 }, { "grad_norm": 0.7414537668228149, "learning_rate": 9.941506726516179e-05, "loss": 0.0355, "step": 2890 }, { "grad_norm": 0.4679614007472992, "learning_rate": 9.94066313475481e-05, "loss": 0.0349, "step": 2900 }, { "grad_norm": 0.5023335218429565, "learning_rate": 9.939813539621857e-05, "loss": 0.0352, "step": 2910 }, { "grad_norm": 0.5964629054069519, "learning_rate": 9.93895794214966e-05, "loss": 0.0334, "step": 2920 }, { "grad_norm": 0.5581484436988831, "learning_rate": 9.938096343377852e-05, "loss": 0.0352, "step": 2930 }, { "grad_norm": 0.5380269289016724, "learning_rate": 9.937228744353353e-05, "loss": 0.0349, "step": 2940 }, { "grad_norm": 0.4668833017349243, "learning_rate": 9.936355146130379e-05, "loss": 0.0328, "step": 2950 }, { "grad_norm": 0.43839725852012634, "learning_rate": 9.935475549770435e-05, "loss": 0.0335, "step": 2960 }, { "grad_norm": 0.5909776091575623, "learning_rate": 9.934589956342315e-05, "loss": 0.0356, "step": 2970 }, { "grad_norm": 0.46072277426719666, "learning_rate": 9.933698366922093e-05, "loss": 0.035, "step": 2980 }, { "grad_norm": 0.6207431554794312, "learning_rate": 9.93280078259314e-05, "loss": 0.0341, "step": 2990 }, { "grad_norm": 0.5444117188453674, "learning_rate": 9.931897204446104e-05, "loss": 0.033, "step": 3000 }, { "grad_norm": 0.5664342641830444, "learning_rate": 9.930987633578915e-05, "loss": 0.0344, "step": 3010 }, { "grad_norm": 0.7013835310935974, "learning_rate": 9.93007207109679e-05, "loss": 0.0342, "step": 3020 }, { "grad_norm": 0.41322097182273865, "learning_rate": 9.929150518112224e-05, "loss": 0.0338, "step": 3030 }, { "grad_norm": 0.4599977135658264, "learning_rate": 9.928222975744991e-05, "loss": 0.0349, "step": 3040 }, { "grad_norm": 0.4189639091491699, "learning_rate": 9.92728944512214e-05, "loss": 0.0334, "step": 3050 }, { "grad_norm": 0.6589005589485168, "learning_rate": 9.926349927378001e-05, "loss": 0.0348, "step": 3060 }, { "grad_norm": 0.4410434365272522, "learning_rate": 9.925404423654174e-05, "loss": 0.0349, "step": 3070 }, { "grad_norm": 0.5676413774490356, "learning_rate": 9.924452935099537e-05, "loss": 0.0323, "step": 3080 }, { "grad_norm": 0.5551483631134033, "learning_rate": 9.92349546287024e-05, "loss": 0.0316, "step": 3090 }, { "grad_norm": 0.6876076459884644, "learning_rate": 9.9225320081297e-05, "loss": 0.0339, "step": 3100 }, { "grad_norm": 0.5772173404693604, "learning_rate": 9.921562572048606e-05, "loss": 0.0347, "step": 3110 }, { "grad_norm": 0.5337933897972107, "learning_rate": 9.920587155804913e-05, "loss": 0.0329, "step": 3120 }, { "grad_norm": 0.5200109481811523, "learning_rate": 9.919605760583845e-05, "loss": 0.0314, "step": 3130 }, { "grad_norm": 0.4010808765888214, "learning_rate": 9.91861838757789e-05, "loss": 0.0312, "step": 3140 }, { "grad_norm": 0.602595329284668, "learning_rate": 9.917625037986798e-05, "loss": 0.0329, "step": 3150 }, { "grad_norm": 0.42525333166122437, "learning_rate": 9.916625713017583e-05, "loss": 0.0324, "step": 3160 }, { "grad_norm": 0.645604133605957, "learning_rate": 9.915620413884519e-05, "loss": 0.0328, "step": 3170 }, { "grad_norm": 0.5573173761367798, "learning_rate": 9.914609141809139e-05, "loss": 0.0347, "step": 3180 }, { "grad_norm": 0.5239332318305969, "learning_rate": 9.913591898020235e-05, "loss": 0.0326, "step": 3190 }, { "grad_norm": 0.6686249375343323, "learning_rate": 9.912568683753853e-05, "loss": 0.0354, "step": 3200 }, { "grad_norm": 0.5051627159118652, "learning_rate": 9.911539500253295e-05, "loss": 0.0341, "step": 3210 }, { "grad_norm": 0.49810031056404114, "learning_rate": 9.910504348769118e-05, "loss": 0.0315, "step": 3220 }, { "grad_norm": 0.6298574209213257, "learning_rate": 9.909463230559127e-05, "loss": 0.0319, "step": 3230 }, { "grad_norm": 0.4464428424835205, "learning_rate": 9.908416146888376e-05, "loss": 0.0307, "step": 3240 }, { "grad_norm": 0.5502170324325562, "learning_rate": 9.907363099029175e-05, "loss": 0.0336, "step": 3250 }, { "grad_norm": 0.5409323573112488, "learning_rate": 9.906304088261073e-05, "loss": 0.0333, "step": 3260 }, { "grad_norm": 0.5345064997673035, "learning_rate": 9.905239115870872e-05, "loss": 0.0306, "step": 3270 }, { "grad_norm": 0.710795521736145, "learning_rate": 9.90416818315261e-05, "loss": 0.0327, "step": 3280 }, { "grad_norm": 0.6401283144950867, "learning_rate": 9.903091291407573e-05, "loss": 0.0315, "step": 3290 }, { "grad_norm": 0.4734502136707306, "learning_rate": 9.902008441944286e-05, "loss": 0.0319, "step": 3300 }, { "grad_norm": 0.6390936970710754, "learning_rate": 9.900919636078512e-05, "loss": 0.0343, "step": 3310 }, { "grad_norm": 0.6288327574729919, "learning_rate": 9.899824875133255e-05, "loss": 0.0357, "step": 3320 }, { "grad_norm": 0.41811782121658325, "learning_rate": 9.898724160438749e-05, "loss": 0.0331, "step": 3330 }, { "grad_norm": 0.4712206721305847, "learning_rate": 9.89761749333247e-05, "loss": 0.0321, "step": 3340 }, { "grad_norm": 0.541106104850769, "learning_rate": 9.896504875159122e-05, "loss": 0.0317, "step": 3350 }, { "grad_norm": 0.6272668242454529, "learning_rate": 9.89538630727064e-05, "loss": 0.0337, "step": 3360 }, { "grad_norm": 0.4778730869293213, "learning_rate": 9.894261791026189e-05, "loss": 0.0329, "step": 3370 }, { "grad_norm": 0.5161771178245544, "learning_rate": 9.893131327792165e-05, "loss": 0.0358, "step": 3380 }, { "grad_norm": 0.5454795956611633, "learning_rate": 9.891994918942182e-05, "loss": 0.0337, "step": 3390 }, { "grad_norm": 0.5579949021339417, "learning_rate": 9.890852565857092e-05, "loss": 0.0325, "step": 3400 }, { "grad_norm": 0.43336305022239685, "learning_rate": 9.889704269924954e-05, "loss": 0.033, "step": 3410 }, { "grad_norm": 0.5180725455284119, "learning_rate": 9.888550032541059e-05, "loss": 0.0324, "step": 3420 }, { "grad_norm": 0.4674279987812042, "learning_rate": 9.887389855107916e-05, "loss": 0.0308, "step": 3430 }, { "grad_norm": 0.5220392346382141, "learning_rate": 9.886223739035248e-05, "loss": 0.032, "step": 3440 }, { "grad_norm": 0.5359625816345215, "learning_rate": 9.885051685739997e-05, "loss": 0.0312, "step": 3450 }, { "grad_norm": 0.5261549949645996, "learning_rate": 9.883873696646316e-05, "loss": 0.0327, "step": 3460 }, { "grad_norm": 0.43261486291885376, "learning_rate": 9.882689773185575e-05, "loss": 0.0313, "step": 3470 }, { "grad_norm": 0.5521197319030762, "learning_rate": 9.881499916796353e-05, "loss": 0.0314, "step": 3480 }, { "grad_norm": 0.6331434845924377, "learning_rate": 9.880304128924434e-05, "loss": 0.0313, "step": 3490 }, { "grad_norm": 0.49536824226379395, "learning_rate": 9.879102411022817e-05, "loss": 0.0292, "step": 3500 }, { "grad_norm": 0.5600954294204712, "learning_rate": 9.877894764551703e-05, "loss": 0.0308, "step": 3510 }, { "grad_norm": 0.46873748302459717, "learning_rate": 9.876681190978494e-05, "loss": 0.0305, "step": 3520 }, { "grad_norm": 0.5754407644271851, "learning_rate": 9.875461691777797e-05, "loss": 0.0296, "step": 3530 }, { "grad_norm": 0.552854597568512, "learning_rate": 9.874236268431417e-05, "loss": 0.0324, "step": 3540 }, { "grad_norm": 0.4065485894680023, "learning_rate": 9.873004922428361e-05, "loss": 0.0304, "step": 3550 }, { "grad_norm": 0.44858065247535706, "learning_rate": 9.871767655264829e-05, "loss": 0.0304, "step": 3560 }, { "grad_norm": 0.5338795781135559, "learning_rate": 9.87052446844422e-05, "loss": 0.0307, "step": 3570 }, { "grad_norm": 0.6342114210128784, "learning_rate": 9.869275363477122e-05, "loss": 0.0303, "step": 3580 }, { "grad_norm": 0.6753395199775696, "learning_rate": 9.868020341881312e-05, "loss": 0.0322, "step": 3590 }, { "grad_norm": 0.46530067920684814, "learning_rate": 9.866759405181765e-05, "loss": 0.032, "step": 3600 }, { "grad_norm": 0.5572906732559204, "learning_rate": 9.865492554910633e-05, "loss": 0.0299, "step": 3610 }, { "grad_norm": 0.5319712162017822, "learning_rate": 9.864219792607262e-05, "loss": 0.0335, "step": 3620 }, { "grad_norm": 0.5416569709777832, "learning_rate": 9.862941119818177e-05, "loss": 0.029, "step": 3630 }, { "grad_norm": 0.4529987871646881, "learning_rate": 9.861656538097086e-05, "loss": 0.0312, "step": 3640 }, { "grad_norm": 0.49536070227622986, "learning_rate": 9.860366049004877e-05, "loss": 0.0306, "step": 3650 }, { "grad_norm": 0.5354276895523071, "learning_rate": 9.859069654109615e-05, "loss": 0.0317, "step": 3660 }, { "grad_norm": 0.5388392806053162, "learning_rate": 9.857767354986545e-05, "loss": 0.029, "step": 3670 }, { "grad_norm": 0.4168851971626282, "learning_rate": 9.856459153218078e-05, "loss": 0.0331, "step": 3680 }, { "grad_norm": 0.5648592114448547, "learning_rate": 9.855145050393808e-05, "loss": 0.0326, "step": 3690 }, { "grad_norm": 0.51558917760849, "learning_rate": 9.85382504811049e-05, "loss": 0.0308, "step": 3700 }, { "grad_norm": 0.46784237027168274, "learning_rate": 9.852499147972054e-05, "loss": 0.0293, "step": 3710 }, { "grad_norm": 0.4863705337047577, "learning_rate": 9.851167351589592e-05, "loss": 0.0293, "step": 3720 }, { "grad_norm": 0.5770860910415649, "learning_rate": 9.849829660581363e-05, "loss": 0.0315, "step": 3730 }, { "grad_norm": 0.537215530872345, "learning_rate": 9.848486076572787e-05, "loss": 0.0294, "step": 3740 }, { "grad_norm": 0.5241726040840149, "learning_rate": 9.847136601196446e-05, "loss": 0.0294, "step": 3750 }, { "grad_norm": 0.5277009010314941, "learning_rate": 9.845781236092078e-05, "loss": 0.0297, "step": 3760 }, { "grad_norm": 0.6135159134864807, "learning_rate": 9.844419982906583e-05, "loss": 0.0293, "step": 3770 }, { "grad_norm": 0.5374872088432312, "learning_rate": 9.843052843294008e-05, "loss": 0.0314, "step": 3780 }, { "grad_norm": 0.561579704284668, "learning_rate": 9.841679818915559e-05, "loss": 0.0293, "step": 3790 }, { "grad_norm": 0.48467305302619934, "learning_rate": 9.840300911439591e-05, "loss": 0.0296, "step": 3800 }, { "grad_norm": 0.6467391848564148, "learning_rate": 9.838916122541603e-05, "loss": 0.029, "step": 3810 }, { "grad_norm": 0.5941223502159119, "learning_rate": 9.837525453904246e-05, "loss": 0.0308, "step": 3820 }, { "grad_norm": 0.6276981234550476, "learning_rate": 9.836128907217314e-05, "loss": 0.0309, "step": 3830 }, { "grad_norm": 0.576546847820282, "learning_rate": 9.834726484177743e-05, "loss": 0.0299, "step": 3840 }, { "grad_norm": 0.539531409740448, "learning_rate": 9.833318186489609e-05, "loss": 0.0308, "step": 3850 }, { "grad_norm": 0.5105360150337219, "learning_rate": 9.831904015864126e-05, "loss": 0.0302, "step": 3860 }, { "grad_norm": 0.4079248011112213, "learning_rate": 9.830483974019645e-05, "loss": 0.0294, "step": 3870 }, { "grad_norm": 0.5691694021224976, "learning_rate": 9.82905806268165e-05, "loss": 0.029, "step": 3880 }, { "grad_norm": 0.3930612802505493, "learning_rate": 9.82762628358276e-05, "loss": 0.0317, "step": 3890 }, { "grad_norm": 0.48195546865463257, "learning_rate": 9.826188638462718e-05, "loss": 0.0275, "step": 3900 }, { "grad_norm": 0.44009217619895935, "learning_rate": 9.824745129068402e-05, "loss": 0.0269, "step": 3910 }, { "grad_norm": 0.46501123905181885, "learning_rate": 9.82329575715381e-05, "loss": 0.0289, "step": 3920 }, { "grad_norm": 0.48853984475135803, "learning_rate": 9.821840524480066e-05, "loss": 0.029, "step": 3930 }, { "grad_norm": 0.4357215166091919, "learning_rate": 9.820379432815414e-05, "loss": 0.0283, "step": 3940 }, { "grad_norm": 0.5993157029151917, "learning_rate": 9.81891248393522e-05, "loss": 0.0297, "step": 3950 }, { "grad_norm": 0.48249173164367676, "learning_rate": 9.817439679621963e-05, "loss": 0.0295, "step": 3960 }, { "grad_norm": 0.5071114301681519, "learning_rate": 9.815961021665243e-05, "loss": 0.0297, "step": 3970 }, { "grad_norm": 0.5427010655403137, "learning_rate": 9.814476511861763e-05, "loss": 0.03, "step": 3980 }, { "grad_norm": 0.5940617322921753, "learning_rate": 9.812986152015348e-05, "loss": 0.0319, "step": 3990 }, { "grad_norm": 0.6007356643676758, "learning_rate": 9.811489943936922e-05, "loss": 0.0283, "step": 4000 }, { "grad_norm": 0.45132508873939514, "learning_rate": 9.809987889444522e-05, "loss": 0.0273, "step": 4010 }, { "grad_norm": 0.4298675060272217, "learning_rate": 9.808479990363282e-05, "loss": 0.0284, "step": 4020 }, { "grad_norm": 0.39303600788116455, "learning_rate": 9.806966248525445e-05, "loss": 0.028, "step": 4030 }, { "grad_norm": 0.5677952766418457, "learning_rate": 9.805446665770348e-05, "loss": 0.0284, "step": 4040 }, { "grad_norm": 0.41741976141929626, "learning_rate": 9.803921243944429e-05, "loss": 0.0288, "step": 4050 }, { "grad_norm": 0.4958172142505646, "learning_rate": 9.802389984901218e-05, "loss": 0.0279, "step": 4060 }, { "grad_norm": 0.5058831572532654, "learning_rate": 9.80085289050134e-05, "loss": 0.0277, "step": 4070 }, { "grad_norm": 0.4799667298793793, "learning_rate": 9.799309962612508e-05, "loss": 0.0293, "step": 4080 }, { "grad_norm": 0.4996417164802551, "learning_rate": 9.797761203109527e-05, "loss": 0.029, "step": 4090 }, { "grad_norm": 0.46496379375457764, "learning_rate": 9.796206613874283e-05, "loss": 0.0289, "step": 4100 }, { "grad_norm": 0.44552916288375854, "learning_rate": 9.794646196795754e-05, "loss": 0.0277, "step": 4110 }, { "grad_norm": 0.4660540819168091, "learning_rate": 9.793079953769987e-05, "loss": 0.0277, "step": 4120 }, { "grad_norm": 0.5758588910102844, "learning_rate": 9.79150788670012e-05, "loss": 0.0286, "step": 4130 }, { "grad_norm": 0.47274714708328247, "learning_rate": 9.78992999749636e-05, "loss": 0.028, "step": 4140 }, { "grad_norm": 0.5877339839935303, "learning_rate": 9.788346288075994e-05, "loss": 0.0281, "step": 4150 }, { "grad_norm": 0.6938865780830383, "learning_rate": 9.786756760363373e-05, "loss": 0.0278, "step": 4160 }, { "grad_norm": 0.4909265339374542, "learning_rate": 9.78516141628993e-05, "loss": 0.0275, "step": 4170 }, { "grad_norm": 0.4589878022670746, "learning_rate": 9.783560257794154e-05, "loss": 0.0284, "step": 4180 }, { "grad_norm": 0.6123479604721069, "learning_rate": 9.781953286821603e-05, "loss": 0.0257, "step": 4190 }, { "grad_norm": 0.46670061349868774, "learning_rate": 9.780340505324901e-05, "loss": 0.0272, "step": 4200 }, { "grad_norm": 0.6732054352760315, "learning_rate": 9.778721915263727e-05, "loss": 0.0281, "step": 4210 }, { "grad_norm": 0.5753781795501709, "learning_rate": 9.777097518604824e-05, "loss": 0.0274, "step": 4220 }, { "grad_norm": 0.6165180802345276, "learning_rate": 9.775467317321984e-05, "loss": 0.0303, "step": 4230 }, { "grad_norm": 0.5160939693450928, "learning_rate": 9.773831313396055e-05, "loss": 0.0273, "step": 4240 }, { "grad_norm": 0.5413023829460144, "learning_rate": 9.77218950881494e-05, "loss": 0.0286, "step": 4250 }, { "grad_norm": 0.5136318802833557, "learning_rate": 9.770541905573583e-05, "loss": 0.0271, "step": 4260 }, { "grad_norm": 0.662514865398407, "learning_rate": 9.768888505673976e-05, "loss": 0.0276, "step": 4270 }, { "grad_norm": 0.5162094235420227, "learning_rate": 9.767229311125162e-05, "loss": 0.0278, "step": 4280 }, { "grad_norm": 0.5230931639671326, "learning_rate": 9.765564323943211e-05, "loss": 0.0298, "step": 4290 }, { "grad_norm": 0.5208863615989685, "learning_rate": 9.763893546151244e-05, "loss": 0.0265, "step": 4300 }, { "grad_norm": 0.44033685326576233, "learning_rate": 9.762216979779412e-05, "loss": 0.0273, "step": 4310 }, { "grad_norm": 0.5224273204803467, "learning_rate": 9.760534626864902e-05, "loss": 0.0278, "step": 4320 }, { "grad_norm": 0.503278374671936, "learning_rate": 9.758846489451931e-05, "loss": 0.0264, "step": 4330 }, { "grad_norm": 0.43135321140289307, "learning_rate": 9.757152569591748e-05, "loss": 0.0278, "step": 4340 }, { "grad_norm": 0.41968685388565063, "learning_rate": 9.75545286934262e-05, "loss": 0.0247, "step": 4350 }, { "grad_norm": 0.5300660729408264, "learning_rate": 9.753747390769847e-05, "loss": 0.0265, "step": 4360 }, { "grad_norm": 0.5508757829666138, "learning_rate": 9.752036135945744e-05, "loss": 0.026, "step": 4370 }, { "grad_norm": 0.48966965079307556, "learning_rate": 9.750319106949649e-05, "loss": 0.024, "step": 4380 }, { "grad_norm": 0.4271833002567291, "learning_rate": 9.748596305867913e-05, "loss": 0.0281, "step": 4390 }, { "grad_norm": 0.4649449586868286, "learning_rate": 9.746867734793903e-05, "loss": 0.0249, "step": 4400 }, { "grad_norm": 0.6460062265396118, "learning_rate": 9.745133395827993e-05, "loss": 0.0256, "step": 4410 }, { "grad_norm": 0.5220698714256287, "learning_rate": 9.743393291077572e-05, "loss": 0.0273, "step": 4420 }, { "grad_norm": 0.4801582396030426, "learning_rate": 9.741647422657028e-05, "loss": 0.0275, "step": 4430 }, { "grad_norm": 0.4439697563648224, "learning_rate": 9.739895792687758e-05, "loss": 0.0271, "step": 4440 }, { "grad_norm": 0.5416855812072754, "learning_rate": 9.738138403298157e-05, "loss": 0.0288, "step": 4450 }, { "grad_norm": 0.47156187891960144, "learning_rate": 9.736375256623619e-05, "loss": 0.0267, "step": 4460 }, { "grad_norm": 0.48665690422058105, "learning_rate": 9.734606354806533e-05, "loss": 0.0281, "step": 4470 }, { "grad_norm": 0.4060969650745392, "learning_rate": 9.73283169999628e-05, "loss": 0.0272, "step": 4480 }, { "grad_norm": 0.4355490505695343, "learning_rate": 9.731051294349238e-05, "loss": 0.0272, "step": 4490 }, { "grad_norm": 0.5726194381713867, "learning_rate": 9.729265140028762e-05, "loss": 0.0281, "step": 4500 }, { "grad_norm": 0.49670320749282837, "learning_rate": 9.727473239205201e-05, "loss": 0.0285, "step": 4510 }, { "grad_norm": 0.5777426958084106, "learning_rate": 9.725675594055883e-05, "loss": 0.0273, "step": 4520 }, { "grad_norm": 0.36380887031555176, "learning_rate": 9.723872206765116e-05, "loss": 0.0253, "step": 4530 }, { "grad_norm": 0.5655593872070312, "learning_rate": 9.722063079524185e-05, "loss": 0.0257, "step": 4540 }, { "grad_norm": 0.5297718644142151, "learning_rate": 9.720248214531351e-05, "loss": 0.0248, "step": 4550 }, { "grad_norm": 0.6011912226676941, "learning_rate": 9.718427613991848e-05, "loss": 0.0247, "step": 4560 }, { "grad_norm": 0.46993371844291687, "learning_rate": 9.716601280117873e-05, "loss": 0.0243, "step": 4570 }, { "grad_norm": 0.6736533045768738, "learning_rate": 9.714769215128596e-05, "loss": 0.0257, "step": 4580 }, { "grad_norm": 0.4797600209712982, "learning_rate": 9.712931421250152e-05, "loss": 0.0248, "step": 4590 }, { "grad_norm": 0.5109550356864929, "learning_rate": 9.711087900715627e-05, "loss": 0.0251, "step": 4600 }, { "grad_norm": 0.5943213701248169, "learning_rate": 9.709238655765078e-05, "loss": 0.0257, "step": 4610 }, { "grad_norm": 0.4454824924468994, "learning_rate": 9.707383688645511e-05, "loss": 0.0249, "step": 4620 }, { "grad_norm": 0.5341625213623047, "learning_rate": 9.705523001610883e-05, "loss": 0.0266, "step": 4630 }, { "grad_norm": 0.7142083048820496, "learning_rate": 9.703656596922107e-05, "loss": 0.0279, "step": 4640 }, { "grad_norm": 0.5267425775527954, "learning_rate": 9.70178447684704e-05, "loss": 0.0254, "step": 4650 }, { "grad_norm": 0.4902753531932831, "learning_rate": 9.699906643660483e-05, "loss": 0.0236, "step": 4660 }, { "grad_norm": 0.456157922744751, "learning_rate": 9.698023099644185e-05, "loss": 0.0262, "step": 4670 }, { "grad_norm": 0.5741419196128845, "learning_rate": 9.696133847086823e-05, "loss": 0.0256, "step": 4680 }, { "grad_norm": 0.5706208348274231, "learning_rate": 9.694238888284022e-05, "loss": 0.0237, "step": 4690 }, { "grad_norm": 0.5595664381980896, "learning_rate": 9.692338225538333e-05, "loss": 0.0243, "step": 4700 }, { "grad_norm": 0.6362490653991699, "learning_rate": 9.690431861159241e-05, "loss": 0.0243, "step": 4710 }, { "grad_norm": 0.49317440390586853, "learning_rate": 9.688519797463161e-05, "loss": 0.0241, "step": 4720 }, { "grad_norm": 0.5495986938476562, "learning_rate": 9.686602036773426e-05, "loss": 0.0238, "step": 4730 }, { "grad_norm": 0.5672884583473206, "learning_rate": 9.684678581420302e-05, "loss": 0.0232, "step": 4740 }, { "grad_norm": 0.3791441321372986, "learning_rate": 9.682749433740962e-05, "loss": 0.0214, "step": 4750 }, { "grad_norm": 0.635084867477417, "learning_rate": 9.680814596079507e-05, "loss": 0.024, "step": 4760 }, { "grad_norm": 0.44687172770500183, "learning_rate": 9.678874070786945e-05, "loss": 0.0222, "step": 4770 }, { "grad_norm": 0.5880815982818604, "learning_rate": 9.676927860221199e-05, "loss": 0.0231, "step": 4780 }, { "grad_norm": 0.4695528745651245, "learning_rate": 9.674975966747097e-05, "loss": 0.0243, "step": 4790 }, { "grad_norm": 0.5622825026512146, "learning_rate": 9.673018392736374e-05, "loss": 0.0229, "step": 4800 }, { "grad_norm": 0.48140478134155273, "learning_rate": 9.671055140567667e-05, "loss": 0.0217, "step": 4810 }, { "grad_norm": 0.4679363965988159, "learning_rate": 9.669086212626511e-05, "loss": 0.023, "step": 4820 }, { "grad_norm": 0.6131110787391663, "learning_rate": 9.667111611305341e-05, "loss": 0.0225, "step": 4830 }, { "grad_norm": 0.4789270758628845, "learning_rate": 9.665131339003486e-05, "loss": 0.0232, "step": 4840 }, { "grad_norm": 0.6402637362480164, "learning_rate": 9.663145398127158e-05, "loss": 0.0219, "step": 4850 }, { "grad_norm": 0.47417306900024414, "learning_rate": 9.661153791089467e-05, "loss": 0.0226, "step": 4860 }, { "grad_norm": 0.5275768637657166, "learning_rate": 9.659156520310402e-05, "loss": 0.0232, "step": 4870 }, { "grad_norm": 0.5174130797386169, "learning_rate": 9.657153588216835e-05, "loss": 0.024, "step": 4880 }, { "grad_norm": 0.5008310675621033, "learning_rate": 9.655144997242516e-05, "loss": 0.0239, "step": 4890 }, { "grad_norm": 0.508402943611145, "learning_rate": 9.653130749828075e-05, "loss": 0.023, "step": 4900 }, { "grad_norm": 0.4857371747493744, "learning_rate": 9.65111084842101e-05, "loss": 0.0207, "step": 4910 }, { "grad_norm": 0.5261239409446716, "learning_rate": 9.649085295475695e-05, "loss": 0.0222, "step": 4920 }, { "grad_norm": 0.4184090197086334, "learning_rate": 9.647054093453365e-05, "loss": 0.023, "step": 4930 }, { "grad_norm": 0.6007316708564758, "learning_rate": 9.645017244822123e-05, "loss": 0.0231, "step": 4940 }, { "grad_norm": 0.44578850269317627, "learning_rate": 9.642974752056931e-05, "loss": 0.0221, "step": 4950 }, { "grad_norm": 0.5718969106674194, "learning_rate": 9.640926617639613e-05, "loss": 0.0228, "step": 4960 }, { "grad_norm": 0.47065165638923645, "learning_rate": 9.638872844058843e-05, "loss": 0.0226, "step": 4970 }, { "grad_norm": 0.6126935482025146, "learning_rate": 9.63681343381015e-05, "loss": 0.0225, "step": 4980 }, { "grad_norm": 0.424108624458313, "learning_rate": 9.634748389395914e-05, "loss": 0.0213, "step": 4990 }, { "grad_norm": 0.5950810313224792, "learning_rate": 9.632677713325353e-05, "loss": 0.0207, "step": 5000 }, { "grad_norm": 0.4594224691390991, "learning_rate": 9.63060140811454e-05, "loss": 0.021, "step": 5010 }, { "grad_norm": 0.46782585978507996, "learning_rate": 9.628519476286379e-05, "loss": 0.0214, "step": 5020 }, { "grad_norm": 0.4983988404273987, "learning_rate": 9.626431920370612e-05, "loss": 0.0201, "step": 5030 }, { "grad_norm": 0.6005197167396545, "learning_rate": 9.624338742903819e-05, "loss": 0.0213, "step": 5040 }, { "grad_norm": 0.6117371320724487, "learning_rate": 9.622239946429406e-05, "loss": 0.0231, "step": 5050 }, { "grad_norm": 0.5228967070579529, "learning_rate": 9.620135533497609e-05, "loss": 0.0207, "step": 5060 }, { "grad_norm": 0.6595703959465027, "learning_rate": 9.61802550666549e-05, "loss": 0.0194, "step": 5070 }, { "grad_norm": 0.4813227653503418, "learning_rate": 9.615909868496928e-05, "loss": 0.0197, "step": 5080 }, { "grad_norm": 0.41596779227256775, "learning_rate": 9.613788621562622e-05, "loss": 0.0224, "step": 5090 }, { "grad_norm": 0.45717138051986694, "learning_rate": 9.611661768440093e-05, "loss": 0.0195, "step": 5100 }, { "grad_norm": 0.5034703016281128, "learning_rate": 9.609529311713661e-05, "loss": 0.0202, "step": 5110 }, { "grad_norm": 0.40311121940612793, "learning_rate": 9.607391253974466e-05, "loss": 0.0198, "step": 5120 }, { "grad_norm": 0.5923986434936523, "learning_rate": 9.605247597820448e-05, "loss": 0.0199, "step": 5130 }, { "grad_norm": 0.4329419732093811, "learning_rate": 9.603098345856354e-05, "loss": 0.0204, "step": 5140 }, { "grad_norm": 0.4753156006336212, "learning_rate": 9.600943500693725e-05, "loss": 0.0213, "step": 5150 }, { "grad_norm": 0.42589521408081055, "learning_rate": 9.598783064950902e-05, "loss": 0.0194, "step": 5160 }, { "grad_norm": 0.5389040112495422, "learning_rate": 9.596617041253018e-05, "loss": 0.0202, "step": 5170 }, { "grad_norm": 0.491046279668808, "learning_rate": 9.594445432231996e-05, "loss": 0.0187, "step": 5180 }, { "grad_norm": 0.429523229598999, "learning_rate": 9.592268240526547e-05, "loss": 0.0177, "step": 5190 }, { "grad_norm": 0.498996376991272, "learning_rate": 9.590085468782162e-05, "loss": 0.0212, "step": 5200 }, { "grad_norm": 0.4649933874607086, "learning_rate": 9.587897119651116e-05, "loss": 0.0188, "step": 5210 }, { "grad_norm": 0.38946202397346497, "learning_rate": 9.585703195792459e-05, "loss": 0.0172, "step": 5220 }, { "grad_norm": 0.5122786164283752, "learning_rate": 9.583503699872016e-05, "loss": 0.0183, "step": 5230 }, { "grad_norm": 0.37925073504447937, "learning_rate": 9.581298634562381e-05, "loss": 0.0189, "step": 5240 }, { "grad_norm": 0.3267979025840759, "learning_rate": 9.579088002542917e-05, "loss": 0.0214, "step": 5250 }, { "grad_norm": 0.5075067281723022, "learning_rate": 9.57687180649975e-05, "loss": 0.0173, "step": 5260 }, { "grad_norm": 0.5464387536048889, "learning_rate": 9.574650049125768e-05, "loss": 0.0203, "step": 5270 }, { "grad_norm": 0.49496564269065857, "learning_rate": 9.572422733120614e-05, "loss": 0.0191, "step": 5280 }, { "grad_norm": 0.4231927990913391, "learning_rate": 9.570189861190689e-05, "loss": 0.0182, "step": 5290 }, { "grad_norm": 0.469819039106369, "learning_rate": 9.56795143604914e-05, "loss": 0.0181, "step": 5300 }, { "grad_norm": 0.5372419953346252, "learning_rate": 9.565707460415869e-05, "loss": 0.0186, "step": 5310 }, { "grad_norm": 0.46158352494239807, "learning_rate": 9.563457937017515e-05, "loss": 0.0197, "step": 5320 }, { "grad_norm": 0.43332675099372864, "learning_rate": 9.56120286858746e-05, "loss": 0.0176, "step": 5330 }, { "grad_norm": 0.605982780456543, "learning_rate": 9.558942257865829e-05, "loss": 0.0178, "step": 5340 }, { "grad_norm": 0.555808424949646, "learning_rate": 9.556676107599472e-05, "loss": 0.0182, "step": 5350 }, { "grad_norm": 0.44589316844940186, "learning_rate": 9.554404420541978e-05, "loss": 0.0185, "step": 5360 }, { "grad_norm": 0.44258564710617065, "learning_rate": 9.55212719945366e-05, "loss": 0.0187, "step": 5370 }, { "grad_norm": 0.5140361785888672, "learning_rate": 9.549844447101559e-05, "loss": 0.0182, "step": 5380 }, { "grad_norm": 0.6253278255462646, "learning_rate": 9.547556166259433e-05, "loss": 0.0166, "step": 5390 }, { "grad_norm": 0.46718496084213257, "learning_rate": 9.545262359707756e-05, "loss": 0.0188, "step": 5400 }, { "grad_norm": 0.383916974067688, "learning_rate": 9.542963030233724e-05, "loss": 0.0184, "step": 5410 }, { "grad_norm": 0.4682512581348419, "learning_rate": 9.540658180631237e-05, "loss": 0.0171, "step": 5420 }, { "grad_norm": 0.4831123948097229, "learning_rate": 9.538347813700904e-05, "loss": 0.0178, "step": 5430 }, { "grad_norm": 0.4267755150794983, "learning_rate": 9.536031932250036e-05, "loss": 0.0177, "step": 5440 }, { "grad_norm": 0.47811415791511536, "learning_rate": 9.533710539092653e-05, "loss": 0.017, "step": 5450 }, { "grad_norm": 0.5367674231529236, "learning_rate": 9.531383637049464e-05, "loss": 0.0173, "step": 5460 }, { "grad_norm": 0.45475828647613525, "learning_rate": 9.529051228947875e-05, "loss": 0.017, "step": 5470 }, { "grad_norm": 0.5281142592430115, "learning_rate": 9.52671331762198e-05, "loss": 0.0165, "step": 5480 }, { "grad_norm": 0.45434918999671936, "learning_rate": 9.524369905912565e-05, "loss": 0.0183, "step": 5490 }, { "grad_norm": 0.5526027083396912, "learning_rate": 9.522020996667092e-05, "loss": 0.0181, "step": 5500 }, { "grad_norm": 0.42288097739219666, "learning_rate": 9.519666592739709e-05, "loss": 0.0171, "step": 5510 }, { "grad_norm": 0.502322256565094, "learning_rate": 9.517306696991241e-05, "loss": 0.0163, "step": 5520 }, { "grad_norm": 0.4991615116596222, "learning_rate": 9.51494131228918e-05, "loss": 0.0183, "step": 5530 }, { "grad_norm": 0.46957167983055115, "learning_rate": 9.512570441507695e-05, "loss": 0.016, "step": 5540 }, { "grad_norm": 0.5327632427215576, "learning_rate": 9.510194087527615e-05, "loss": 0.0166, "step": 5550 }, { "grad_norm": 0.4170159697532654, "learning_rate": 9.507812253236435e-05, "loss": 0.017, "step": 5560 }, { "grad_norm": 0.4968635141849518, "learning_rate": 9.505424941528309e-05, "loss": 0.0164, "step": 5570 }, { "grad_norm": 0.4826819896697998, "learning_rate": 9.503032155304046e-05, "loss": 0.0162, "step": 5580 }, { "grad_norm": 0.3677280843257904, "learning_rate": 9.500633897471106e-05, "loss": 0.0164, "step": 5590 }, { "grad_norm": 0.4160604178905487, "learning_rate": 9.498230170943596e-05, "loss": 0.018, "step": 5600 }, { "grad_norm": 0.45811036229133606, "learning_rate": 9.495820978642275e-05, "loss": 0.0163, "step": 5610 }, { "grad_norm": 0.42683902382850647, "learning_rate": 9.493406323494535e-05, "loss": 0.0178, "step": 5620 }, { "grad_norm": 0.4382336735725403, "learning_rate": 9.490986208434413e-05, "loss": 0.0156, "step": 5630 }, { "grad_norm": 0.3592792749404907, "learning_rate": 9.488560636402577e-05, "loss": 0.0145, "step": 5640 }, { "grad_norm": 0.34286990761756897, "learning_rate": 9.486129610346321e-05, "loss": 0.0146, "step": 5650 }, { "grad_norm": 0.49023711681365967, "learning_rate": 9.483693133219576e-05, "loss": 0.0153, "step": 5660 }, { "grad_norm": 0.3683493435382843, "learning_rate": 9.481251207982888e-05, "loss": 0.0162, "step": 5670 }, { "grad_norm": 0.4639711380004883, "learning_rate": 9.47880383760343e-05, "loss": 0.0179, "step": 5680 }, { "grad_norm": 0.3855854868888855, "learning_rate": 9.476351025054983e-05, "loss": 0.0173, "step": 5690 }, { "grad_norm": 0.46952077746391296, "learning_rate": 9.473892773317952e-05, "loss": 0.0167, "step": 5700 }, { "grad_norm": 0.46543654799461365, "learning_rate": 9.471429085379338e-05, "loss": 0.0162, "step": 5710 }, { "grad_norm": 0.43368780612945557, "learning_rate": 9.468959964232757e-05, "loss": 0.015, "step": 5720 }, { "grad_norm": 0.6242442727088928, "learning_rate": 9.466485412878425e-05, "loss": 0.0163, "step": 5730 }, { "grad_norm": 0.46993181109428406, "learning_rate": 9.464005434323154e-05, "loss": 0.0147, "step": 5740 }, { "grad_norm": 0.4298034608364105, "learning_rate": 9.461520031580352e-05, "loss": 0.0149, "step": 5750 }, { "grad_norm": 0.679725170135498, "learning_rate": 9.459029207670019e-05, "loss": 0.0161, "step": 5760 }, { "grad_norm": 0.3906918466091156, "learning_rate": 9.456532965618737e-05, "loss": 0.0185, "step": 5770 }, { "grad_norm": 0.4709409177303314, "learning_rate": 9.454031308459681e-05, "loss": 0.017, "step": 5780 }, { "grad_norm": 0.4399077594280243, "learning_rate": 9.451524239232595e-05, "loss": 0.0165, "step": 5790 }, { "grad_norm": 0.4794357717037201, "learning_rate": 9.449011760983809e-05, "loss": 0.0163, "step": 5800 }, { "grad_norm": 0.37406089901924133, "learning_rate": 9.446493876766218e-05, "loss": 0.0192, "step": 5810 }, { "grad_norm": 0.4109271764755249, "learning_rate": 9.44397058963929e-05, "loss": 0.0185, "step": 5820 }, { "grad_norm": 0.42586711049079895, "learning_rate": 9.441441902669056e-05, "loss": 0.0161, "step": 5830 }, { "grad_norm": 0.35982322692871094, "learning_rate": 9.43890781892811e-05, "loss": 0.0152, "step": 5840 }, { "grad_norm": 0.302796870470047, "learning_rate": 9.436368341495603e-05, "loss": 0.0159, "step": 5850 }, { "grad_norm": 0.4041033685207367, "learning_rate": 9.43382347345724e-05, "loss": 0.0155, "step": 5860 }, { "grad_norm": 0.5660633444786072, "learning_rate": 9.431273217905273e-05, "loss": 0.0167, "step": 5870 }, { "grad_norm": 0.40607672929763794, "learning_rate": 9.428717577938504e-05, "loss": 0.0156, "step": 5880 }, { "grad_norm": 0.4061409831047058, "learning_rate": 9.426156556662276e-05, "loss": 0.014, "step": 5890 }, { "grad_norm": 0.4410693347454071, "learning_rate": 9.423590157188474e-05, "loss": 0.016, "step": 5900 }, { "grad_norm": 0.4097030758857727, "learning_rate": 9.421018382635513e-05, "loss": 0.0161, "step": 5910 }, { "grad_norm": 0.4834526479244232, "learning_rate": 9.418441236128343e-05, "loss": 0.0159, "step": 5920 }, { "grad_norm": 0.36734217405319214, "learning_rate": 9.41585872079844e-05, "loss": 0.0145, "step": 5930 }, { "grad_norm": 0.4003128707408905, "learning_rate": 9.413270839783802e-05, "loss": 0.0164, "step": 5940 }, { "grad_norm": 0.4849574863910675, "learning_rate": 9.41067759622895e-05, "loss": 0.0151, "step": 5950 }, { "grad_norm": 0.40424999594688416, "learning_rate": 9.408078993284917e-05, "loss": 0.0159, "step": 5960 }, { "grad_norm": 0.49761006236076355, "learning_rate": 9.405475034109254e-05, "loss": 0.0189, "step": 5970 }, { "grad_norm": 0.36773961782455444, "learning_rate": 9.402865721866015e-05, "loss": 0.0169, "step": 5980 }, { "grad_norm": 0.3295658826828003, "learning_rate": 9.400251059725762e-05, "loss": 0.0153, "step": 5990 }, { "grad_norm": 0.45937928557395935, "learning_rate": 9.397631050865554e-05, "loss": 0.0157, "step": 6000 }, { "grad_norm": 0.4183763861656189, "learning_rate": 9.395005698468949e-05, "loss": 0.0164, "step": 6010 }, { "grad_norm": 0.33879679441452026, "learning_rate": 9.392375005726e-05, "loss": 0.0137, "step": 6020 }, { "grad_norm": 0.37298041582107544, "learning_rate": 9.389738975833242e-05, "loss": 0.0147, "step": 6030 }, { "grad_norm": 0.4308323264122009, "learning_rate": 9.387097611993707e-05, "loss": 0.0144, "step": 6040 }, { "grad_norm": 0.437456339597702, "learning_rate": 9.384450917416894e-05, "loss": 0.0148, "step": 6050 }, { "grad_norm": 0.4685937166213989, "learning_rate": 9.381798895318792e-05, "loss": 0.0145, "step": 6060 }, { "grad_norm": 0.37430495023727417, "learning_rate": 9.379141548921853e-05, "loss": 0.0147, "step": 6070 }, { "grad_norm": 0.45154932141304016, "learning_rate": 9.376478881455009e-05, "loss": 0.0151, "step": 6080 }, { "grad_norm": 0.4732506573200226, "learning_rate": 9.373810896153647e-05, "loss": 0.0146, "step": 6090 }, { "grad_norm": 0.46154171228408813, "learning_rate": 9.371137596259623e-05, "loss": 0.0144, "step": 6100 }, { "grad_norm": 0.387769877910614, "learning_rate": 9.368458985021248e-05, "loss": 0.0144, "step": 6110 }, { "grad_norm": 0.3920918107032776, "learning_rate": 9.365775065693287e-05, "loss": 0.0164, "step": 6120 }, { "grad_norm": 0.43867599964141846, "learning_rate": 9.363085841536957e-05, "loss": 0.0135, "step": 6130 }, { "grad_norm": 0.44755738973617554, "learning_rate": 9.360391315819917e-05, "loss": 0.0149, "step": 6140 }, { "grad_norm": 0.48710331320762634, "learning_rate": 9.357691491816269e-05, "loss": 0.0141, "step": 6150 }, { "grad_norm": 0.41324517130851746, "learning_rate": 9.354986372806557e-05, "loss": 0.0149, "step": 6160 }, { "grad_norm": 0.3770032525062561, "learning_rate": 9.352275962077752e-05, "loss": 0.0138, "step": 6170 }, { "grad_norm": 0.44077855348587036, "learning_rate": 9.349560262923262e-05, "loss": 0.0152, "step": 6180 }, { "grad_norm": 0.40747708082199097, "learning_rate": 9.346839278642913e-05, "loss": 0.0133, "step": 6190 }, { "grad_norm": 0.4391169548034668, "learning_rate": 9.344113012542963e-05, "loss": 0.0145, "step": 6200 }, { "grad_norm": 0.4199869930744171, "learning_rate": 9.341381467936079e-05, "loss": 0.0133, "step": 6210 }, { "grad_norm": 0.4460112750530243, "learning_rate": 9.338644648141346e-05, "loss": 0.0152, "step": 6220 }, { "grad_norm": 0.30460286140441895, "learning_rate": 9.335902556484257e-05, "loss": 0.0131, "step": 6230 }, { "grad_norm": 0.37233051657676697, "learning_rate": 9.333155196296712e-05, "loss": 0.0141, "step": 6240 }, { "grad_norm": 0.3617931306362152, "learning_rate": 9.330402570917016e-05, "loss": 0.0146, "step": 6250 }, { "grad_norm": 0.3526821732521057, "learning_rate": 9.327644683689865e-05, "loss": 0.014, "step": 6260 }, { "grad_norm": 0.45122718811035156, "learning_rate": 9.324881537966354e-05, "loss": 0.0135, "step": 6270 }, { "grad_norm": 0.5205270648002625, "learning_rate": 9.322113137103964e-05, "loss": 0.0154, "step": 6280 }, { "grad_norm": 0.3603147566318512, "learning_rate": 9.319339484466564e-05, "loss": 0.0166, "step": 6290 }, { "grad_norm": 0.41952037811279297, "learning_rate": 9.316560583424403e-05, "loss": 0.0134, "step": 6300 }, { "grad_norm": 0.36976832151412964, "learning_rate": 9.313776437354108e-05, "loss": 0.0125, "step": 6310 }, { "grad_norm": 0.4265200197696686, "learning_rate": 9.31098704963868e-05, "loss": 0.0147, "step": 6320 }, { "grad_norm": 0.34672072529792786, "learning_rate": 9.308192423667485e-05, "loss": 0.0133, "step": 6330 }, { "grad_norm": 0.3036075532436371, "learning_rate": 9.305392562836262e-05, "loss": 0.0134, "step": 6340 }, { "grad_norm": 0.3667495846748352, "learning_rate": 9.3025874705471e-05, "loss": 0.0139, "step": 6350 }, { "grad_norm": 0.41114041209220886, "learning_rate": 9.299777150208457e-05, "loss": 0.0147, "step": 6360 }, { "grad_norm": 0.42483291029930115, "learning_rate": 9.296961605235133e-05, "loss": 0.0143, "step": 6370 }, { "grad_norm": 0.4150006175041199, "learning_rate": 9.29414083904828e-05, "loss": 0.0154, "step": 6380 }, { "grad_norm": 0.4819064736366272, "learning_rate": 9.291314855075397e-05, "loss": 0.0145, "step": 6390 }, { "grad_norm": 0.41249290108680725, "learning_rate": 9.288483656750322e-05, "loss": 0.0155, "step": 6400 }, { "grad_norm": 0.39605265855789185, "learning_rate": 9.285647247513225e-05, "loss": 0.0134, "step": 6410 }, { "grad_norm": 0.35166269540786743, "learning_rate": 9.282805630810614e-05, "loss": 0.0135, "step": 6420 }, { "grad_norm": 0.4163077771663666, "learning_rate": 9.279958810095317e-05, "loss": 0.0145, "step": 6430 }, { "grad_norm": 0.3122580349445343, "learning_rate": 9.277106788826494e-05, "loss": 0.0135, "step": 6440 }, { "grad_norm": 0.5408252477645874, "learning_rate": 9.274249570469618e-05, "loss": 0.0136, "step": 6450 }, { "grad_norm": 0.3665207624435425, "learning_rate": 9.271387158496476e-05, "loss": 0.0136, "step": 6460 }, { "grad_norm": 0.38548943400382996, "learning_rate": 9.268519556385173e-05, "loss": 0.0148, "step": 6470 }, { "grad_norm": 0.4338033199310303, "learning_rate": 9.265646767620113e-05, "loss": 0.0138, "step": 6480 }, { "grad_norm": 0.3697988986968994, "learning_rate": 9.262768795692006e-05, "loss": 0.0134, "step": 6490 }, { "grad_norm": 0.43436676263809204, "learning_rate": 9.259885644097861e-05, "loss": 0.0145, "step": 6500 }, { "grad_norm": 0.35602453351020813, "learning_rate": 9.256997316340976e-05, "loss": 0.0139, "step": 6510 }, { "grad_norm": 0.3094992935657501, "learning_rate": 9.254103815930943e-05, "loss": 0.0131, "step": 6520 }, { "grad_norm": 0.39769303798675537, "learning_rate": 9.251205146383636e-05, "loss": 0.0139, "step": 6530 }, { "grad_norm": 0.3360579013824463, "learning_rate": 9.248301311221216e-05, "loss": 0.0131, "step": 6540 }, { "grad_norm": 0.43902263045310974, "learning_rate": 9.245392313972115e-05, "loss": 0.0141, "step": 6550 }, { "grad_norm": 0.3751843273639679, "learning_rate": 9.242478158171037e-05, "loss": 0.0139, "step": 6560 }, { "grad_norm": 0.34897536039352417, "learning_rate": 9.239558847358957e-05, "loss": 0.0128, "step": 6570 }, { "grad_norm": 0.3229854106903076, "learning_rate": 9.236634385083114e-05, "loss": 0.013, "step": 6580 }, { "grad_norm": 0.33645179867744446, "learning_rate": 9.233704774897005e-05, "loss": 0.013, "step": 6590 }, { "grad_norm": 0.36346444487571716, "learning_rate": 9.230770020360382e-05, "loss": 0.013, "step": 6600 }, { "grad_norm": 0.38480672240257263, "learning_rate": 9.227830125039247e-05, "loss": 0.0129, "step": 6610 }, { "grad_norm": 0.42673030495643616, "learning_rate": 9.224885092505853e-05, "loss": 0.0133, "step": 6620 }, { "grad_norm": 0.4402414560317993, "learning_rate": 9.221934926338691e-05, "loss": 0.0139, "step": 6630 }, { "grad_norm": 0.39116808772087097, "learning_rate": 9.21897963012249e-05, "loss": 0.0133, "step": 6640 }, { "grad_norm": 0.28943920135498047, "learning_rate": 9.216019207448217e-05, "loss": 0.0147, "step": 6650 }, { "grad_norm": 0.3455876410007477, "learning_rate": 9.21305366191306e-05, "loss": 0.0147, "step": 6660 }, { "grad_norm": 0.44500064849853516, "learning_rate": 9.21008299712044e-05, "loss": 0.0137, "step": 6670 }, { "grad_norm": 0.5418192744255066, "learning_rate": 9.207107216679993e-05, "loss": 0.0141, "step": 6680 }, { "grad_norm": 0.44647619128227234, "learning_rate": 9.204126324207574e-05, "loss": 0.0134, "step": 6690 }, { "grad_norm": 0.33688196539878845, "learning_rate": 9.201140323325248e-05, "loss": 0.0137, "step": 6700 }, { "grad_norm": 0.4310382008552551, "learning_rate": 9.198149217661287e-05, "loss": 0.0142, "step": 6710 }, { "grad_norm": 0.39126962423324585, "learning_rate": 9.195153010850167e-05, "loss": 0.0151, "step": 6720 }, { "grad_norm": 0.46993929147720337, "learning_rate": 9.192151706532562e-05, "loss": 0.0133, "step": 6730 }, { "grad_norm": 0.4224424362182617, "learning_rate": 9.189145308355337e-05, "loss": 0.0146, "step": 6740 }, { "grad_norm": 0.3198283910751343, "learning_rate": 9.186133819971555e-05, "loss": 0.0138, "step": 6750 }, { "grad_norm": 0.45175471901893616, "learning_rate": 9.183117245040455e-05, "loss": 0.0133, "step": 6760 }, { "grad_norm": 0.3550010323524475, "learning_rate": 9.180095587227458e-05, "loss": 0.0133, "step": 6770 }, { "grad_norm": 0.36675015091896057, "learning_rate": 9.177068850204167e-05, "loss": 0.0144, "step": 6780 }, { "grad_norm": 0.322945773601532, "learning_rate": 9.174037037648351e-05, "loss": 0.0134, "step": 6790 }, { "grad_norm": 0.4099406898021698, "learning_rate": 9.171000153243947e-05, "loss": 0.0139, "step": 6800 }, { "grad_norm": 0.32047194242477417, "learning_rate": 9.167958200681057e-05, "loss": 0.012, "step": 6810 }, { "grad_norm": 0.3311690390110016, "learning_rate": 9.164911183655943e-05, "loss": 0.0132, "step": 6820 }, { "grad_norm": 0.34048816561698914, "learning_rate": 9.161859105871013e-05, "loss": 0.0135, "step": 6830 }, { "grad_norm": 0.3637918531894684, "learning_rate": 9.158801971034832e-05, "loss": 0.0138, "step": 6840 }, { "grad_norm": 0.3191274404525757, "learning_rate": 9.155739782862107e-05, "loss": 0.0132, "step": 6850 }, { "grad_norm": 0.30677056312561035, "learning_rate": 9.152672545073687e-05, "loss": 0.0121, "step": 6860 }, { "grad_norm": 0.327228844165802, "learning_rate": 9.149600261396551e-05, "loss": 0.0139, "step": 6870 }, { "grad_norm": 0.4942983090877533, "learning_rate": 9.146522935563817e-05, "loss": 0.0145, "step": 6880 }, { "grad_norm": 0.43315163254737854, "learning_rate": 9.143440571314723e-05, "loss": 0.0119, "step": 6890 }, { "grad_norm": 0.27714836597442627, "learning_rate": 9.140353172394636e-05, "loss": 0.0119, "step": 6900 }, { "grad_norm": 0.2607634663581848, "learning_rate": 9.137260742555032e-05, "loss": 0.0123, "step": 6910 }, { "grad_norm": 0.40268388390541077, "learning_rate": 9.13416328555351e-05, "loss": 0.0133, "step": 6920 }, { "grad_norm": 0.35649073123931885, "learning_rate": 9.131060805153769e-05, "loss": 0.0118, "step": 6930 }, { "grad_norm": 0.389036625623703, "learning_rate": 9.127953305125618e-05, "loss": 0.0136, "step": 6940 }, { "grad_norm": 0.40989917516708374, "learning_rate": 9.124840789244958e-05, "loss": 0.0122, "step": 6950 }, { "grad_norm": 0.40351611375808716, "learning_rate": 9.121723261293794e-05, "loss": 0.0135, "step": 6960 }, { "grad_norm": 0.3241773843765259, "learning_rate": 9.118600725060214e-05, "loss": 0.0117, "step": 6970 }, { "grad_norm": 0.3243855834007263, "learning_rate": 9.115473184338392e-05, "loss": 0.0128, "step": 6980 }, { "grad_norm": 0.4551032781600952, "learning_rate": 9.11234064292859e-05, "loss": 0.0135, "step": 6990 }, { "grad_norm": 0.3283938765525818, "learning_rate": 9.109203104637137e-05, "loss": 0.0118, "step": 7000 }, { "grad_norm": 0.5484637022018433, "learning_rate": 9.10606057327644e-05, "loss": 0.0136, "step": 7010 }, { "grad_norm": 0.39262712001800537, "learning_rate": 9.10291305266497e-05, "loss": 0.0123, "step": 7020 }, { "grad_norm": 0.39485543966293335, "learning_rate": 9.099760546627261e-05, "loss": 0.0133, "step": 7030 }, { "grad_norm": 0.4103463590145111, "learning_rate": 9.096603058993907e-05, "loss": 0.011, "step": 7040 }, { "grad_norm": 0.445277601480484, "learning_rate": 9.093440593601552e-05, "loss": 0.0117, "step": 7050 }, { "grad_norm": 0.48510605096817017, "learning_rate": 9.090273154292889e-05, "loss": 0.013, "step": 7060 }, { "grad_norm": 0.3229779899120331, "learning_rate": 9.087100744916656e-05, "loss": 0.0121, "step": 7070 }, { "grad_norm": 0.34501615166664124, "learning_rate": 9.08392336932763e-05, "loss": 0.012, "step": 7080 }, { "grad_norm": 0.2764420211315155, "learning_rate": 9.080741031386619e-05, "loss": 0.0133, "step": 7090 }, { "grad_norm": 0.3332461416721344, "learning_rate": 9.077553734960468e-05, "loss": 0.0128, "step": 7100 }, { "grad_norm": 0.3700748383998871, "learning_rate": 9.074361483922041e-05, "loss": 0.0129, "step": 7110 }, { "grad_norm": 0.39643654227256775, "learning_rate": 9.071164282150224e-05, "loss": 0.0125, "step": 7120 }, { "grad_norm": 0.2998582720756531, "learning_rate": 9.067962133529919e-05, "loss": 0.0122, "step": 7130 }, { "grad_norm": 0.29886800050735474, "learning_rate": 9.064755041952036e-05, "loss": 0.0118, "step": 7140 }, { "grad_norm": 0.3528709411621094, "learning_rate": 9.061543011313497e-05, "loss": 0.0118, "step": 7150 }, { "grad_norm": 0.3907919228076935, "learning_rate": 9.05832604551722e-05, "loss": 0.0135, "step": 7160 }, { "grad_norm": 0.276755690574646, "learning_rate": 9.055104148472124e-05, "loss": 0.0119, "step": 7170 }, { "grad_norm": 0.37680584192276, "learning_rate": 9.051877324093114e-05, "loss": 0.0114, "step": 7180 }, { "grad_norm": 0.36407947540283203, "learning_rate": 9.04864557630109e-05, "loss": 0.0116, "step": 7190 }, { "grad_norm": 0.4268858730792999, "learning_rate": 9.045408909022928e-05, "loss": 0.0125, "step": 7200 }, { "grad_norm": 0.34331417083740234, "learning_rate": 9.042167326191484e-05, "loss": 0.0108, "step": 7210 }, { "grad_norm": 0.42327192425727844, "learning_rate": 9.038920831745587e-05, "loss": 0.0121, "step": 7220 }, { "grad_norm": 0.2950005531311035, "learning_rate": 9.035669429630036e-05, "loss": 0.0122, "step": 7230 }, { "grad_norm": 0.31393739581108093, "learning_rate": 9.032413123795588e-05, "loss": 0.0113, "step": 7240 }, { "grad_norm": 0.290615051984787, "learning_rate": 9.029151918198962e-05, "loss": 0.0124, "step": 7250 }, { "grad_norm": 0.4486510753631592, "learning_rate": 9.025885816802833e-05, "loss": 0.0117, "step": 7260 }, { "grad_norm": 0.3639298379421234, "learning_rate": 9.02261482357582e-05, "loss": 0.0114, "step": 7270 }, { "grad_norm": 0.3103725016117096, "learning_rate": 9.019338942492485e-05, "loss": 0.0136, "step": 7280 }, { "grad_norm": 0.39008089900016785, "learning_rate": 9.01605817753334e-05, "loss": 0.0116, "step": 7290 }, { "grad_norm": 0.37852030992507935, "learning_rate": 9.012772532684818e-05, "loss": 0.0124, "step": 7300 }, { "grad_norm": 0.29308199882507324, "learning_rate": 9.00948201193929e-05, "loss": 0.0118, "step": 7310 }, { "grad_norm": 0.42500293254852295, "learning_rate": 9.006186619295047e-05, "loss": 0.012, "step": 7320 }, { "grad_norm": 0.34131696820259094, "learning_rate": 9.002886358756305e-05, "loss": 0.0115, "step": 7330 }, { "grad_norm": 0.36960646510124207, "learning_rate": 8.999581234333188e-05, "loss": 0.0125, "step": 7340 }, { "grad_norm": 0.34096142649650574, "learning_rate": 8.996271250041734e-05, "loss": 0.0116, "step": 7350 }, { "grad_norm": 0.33445802330970764, "learning_rate": 8.99295640990389e-05, "loss": 0.011, "step": 7360 }, { "grad_norm": 0.28105318546295166, "learning_rate": 8.989636717947494e-05, "loss": 0.0115, "step": 7370 }, { "grad_norm": 0.3778381943702698, "learning_rate": 8.986312178206291e-05, "loss": 0.0123, "step": 7380 }, { "grad_norm": 0.3436017334461212, "learning_rate": 8.982982794719904e-05, "loss": 0.0114, "step": 7390 }, { "grad_norm": 0.33777639269828796, "learning_rate": 8.979648571533852e-05, "loss": 0.0124, "step": 7400 }, { "grad_norm": 0.39780232310295105, "learning_rate": 8.976309512699528e-05, "loss": 0.0105, "step": 7410 }, { "grad_norm": 0.336264044046402, "learning_rate": 8.972965622274205e-05, "loss": 0.0113, "step": 7420 }, { "grad_norm": 0.4956374168395996, "learning_rate": 8.969616904321026e-05, "loss": 0.011, "step": 7430 }, { "grad_norm": 0.4496384561061859, "learning_rate": 8.966263362908998e-05, "loss": 0.0118, "step": 7440 }, { "grad_norm": 0.3216903805732727, "learning_rate": 8.962905002112989e-05, "loss": 0.012, "step": 7450 }, { "grad_norm": 0.3218890130519867, "learning_rate": 8.959541826013724e-05, "loss": 0.0124, "step": 7460 }, { "grad_norm": 0.35675048828125, "learning_rate": 8.956173838697781e-05, "loss": 0.0128, "step": 7470 }, { "grad_norm": 0.32960838079452515, "learning_rate": 8.952801044257581e-05, "loss": 0.0111, "step": 7480 }, { "grad_norm": 0.35388556122779846, "learning_rate": 8.949423446791387e-05, "loss": 0.0119, "step": 7490 }, { "grad_norm": 0.32837608456611633, "learning_rate": 8.9460410504033e-05, "loss": 0.0115, "step": 7500 }, { "grad_norm": 0.31394892930984497, "learning_rate": 8.942653859203248e-05, "loss": 0.0136, "step": 7510 }, { "grad_norm": 0.28792908787727356, "learning_rate": 8.93926187730699e-05, "loss": 0.0107, "step": 7520 }, { "grad_norm": 0.2549435794353485, "learning_rate": 8.935865108836102e-05, "loss": 0.0124, "step": 7530 }, { "grad_norm": 0.3519458770751953, "learning_rate": 8.932463557917981e-05, "loss": 0.0142, "step": 7540 }, { "grad_norm": 0.34155184030532837, "learning_rate": 8.929057228685829e-05, "loss": 0.0135, "step": 7550 }, { "grad_norm": 0.37917962670326233, "learning_rate": 8.925646125278658e-05, "loss": 0.0124, "step": 7560 }, { "grad_norm": 0.3055376410484314, "learning_rate": 8.92223025184128e-05, "loss": 0.0111, "step": 7570 }, { "grad_norm": 0.2570594549179077, "learning_rate": 8.918809612524303e-05, "loss": 0.0111, "step": 7580 }, { "grad_norm": 0.37878426909446716, "learning_rate": 8.91538421148413e-05, "loss": 0.0112, "step": 7590 }, { "grad_norm": 0.36314377188682556, "learning_rate": 8.91195405288294e-05, "loss": 0.0131, "step": 7600 }, { "grad_norm": 0.3628675639629364, "learning_rate": 8.908519140888705e-05, "loss": 0.0108, "step": 7610 }, { "grad_norm": 0.35584089159965515, "learning_rate": 8.905079479675164e-05, "loss": 0.0121, "step": 7620 }, { "grad_norm": 0.40685468912124634, "learning_rate": 8.901635073421831e-05, "loss": 0.0119, "step": 7630 }, { "grad_norm": 0.344491183757782, "learning_rate": 8.898185926313983e-05, "loss": 0.0127, "step": 7640 }, { "grad_norm": 0.34751275181770325, "learning_rate": 8.894732042542659e-05, "loss": 0.013, "step": 7650 }, { "grad_norm": 0.2718348801136017, "learning_rate": 8.891273426304654e-05, "loss": 0.0098, "step": 7660 }, { "grad_norm": 0.35238078236579895, "learning_rate": 8.887810081802515e-05, "loss": 0.0113, "step": 7670 }, { "grad_norm": 0.3028378486633301, "learning_rate": 8.884342013244529e-05, "loss": 0.0119, "step": 7680 }, { "grad_norm": 0.3731449544429779, "learning_rate": 8.880869224844725e-05, "loss": 0.0123, "step": 7690 }, { "grad_norm": 0.3167056739330292, "learning_rate": 8.877391720822875e-05, "loss": 0.0132, "step": 7700 }, { "grad_norm": 0.3493010699748993, "learning_rate": 8.873909505404468e-05, "loss": 0.012, "step": 7710 }, { "grad_norm": 0.34445878863334656, "learning_rate": 8.870422582820726e-05, "loss": 0.0126, "step": 7720 }, { "grad_norm": 0.33221665024757385, "learning_rate": 8.866930957308588e-05, "loss": 0.0106, "step": 7730 }, { "grad_norm": 0.334201842546463, "learning_rate": 8.86343463311071e-05, "loss": 0.0114, "step": 7740 }, { "grad_norm": 0.3063510060310364, "learning_rate": 8.859933614475452e-05, "loss": 0.0131, "step": 7750 }, { "grad_norm": 0.5343878865242004, "learning_rate": 8.85642790565689e-05, "loss": 0.0123, "step": 7760 }, { "grad_norm": 0.32829737663269043, "learning_rate": 8.852917510914782e-05, "loss": 0.0112, "step": 7770 }, { "grad_norm": 0.31713294982910156, "learning_rate": 8.84940243451459e-05, "loss": 0.0129, "step": 7780 }, { "grad_norm": 0.44927090406417847, "learning_rate": 8.845882680727469e-05, "loss": 0.0127, "step": 7790 }, { "grad_norm": 0.29228851199150085, "learning_rate": 8.842358253830246e-05, "loss": 0.0115, "step": 7800 }, { "grad_norm": 0.3498315215110779, "learning_rate": 8.838829158105434e-05, "loss": 0.0121, "step": 7810 }, { "grad_norm": 0.3420019745826721, "learning_rate": 8.835295397841217e-05, "loss": 0.0115, "step": 7820 }, { "grad_norm": 0.30869531631469727, "learning_rate": 8.831756977331447e-05, "loss": 0.0117, "step": 7830 }, { "grad_norm": 0.37823766469955444, "learning_rate": 8.828213900875638e-05, "loss": 0.0108, "step": 7840 }, { "grad_norm": 0.3189360499382019, "learning_rate": 8.824666172778964e-05, "loss": 0.0104, "step": 7850 }, { "grad_norm": 0.476632297039032, "learning_rate": 8.821113797352245e-05, "loss": 0.0118, "step": 7860 }, { "grad_norm": 0.2909100353717804, "learning_rate": 8.817556778911956e-05, "loss": 0.0112, "step": 7870 }, { "grad_norm": 0.3038470149040222, "learning_rate": 8.81399512178021e-05, "loss": 0.012, "step": 7880 }, { "grad_norm": 0.3964444696903229, "learning_rate": 8.810428830284752e-05, "loss": 0.0116, "step": 7890 }, { "grad_norm": 0.6055150032043457, "learning_rate": 8.806857908758967e-05, "loss": 0.0129, "step": 7900 }, { "grad_norm": 0.3260117173194885, "learning_rate": 8.80328236154186e-05, "loss": 0.0117, "step": 7910 }, { "grad_norm": 0.19797098636627197, "learning_rate": 8.799702192978056e-05, "loss": 0.0117, "step": 7920 }, { "grad_norm": 0.3774504065513611, "learning_rate": 8.7961174074178e-05, "loss": 0.0119, "step": 7930 }, { "grad_norm": 0.40719759464263916, "learning_rate": 8.792528009216942e-05, "loss": 0.0117, "step": 7940 }, { "grad_norm": 0.41024884581565857, "learning_rate": 8.788934002736943e-05, "loss": 0.012, "step": 7950 }, { "grad_norm": 0.34125572443008423, "learning_rate": 8.785335392344857e-05, "loss": 0.0108, "step": 7960 }, { "grad_norm": 0.20444460213184357, "learning_rate": 8.781732182413336e-05, "loss": 0.0109, "step": 7970 }, { "grad_norm": 0.38777855038642883, "learning_rate": 8.778124377320618e-05, "loss": 0.0112, "step": 7980 }, { "grad_norm": 0.3357694745063782, "learning_rate": 8.774511981450529e-05, "loss": 0.0116, "step": 7990 }, { "grad_norm": 0.3549976348876953, "learning_rate": 8.770894999192467e-05, "loss": 0.0109, "step": 8000 }, { "grad_norm": 0.3552756607532501, "learning_rate": 8.767273434941412e-05, "loss": 0.0111, "step": 8010 }, { "grad_norm": 0.3434692621231079, "learning_rate": 8.7636472930979e-05, "loss": 0.0108, "step": 8020 }, { "grad_norm": 0.3151755630970001, "learning_rate": 8.76001657806804e-05, "loss": 0.0112, "step": 8030 }, { "grad_norm": 0.40102118253707886, "learning_rate": 8.756381294263491e-05, "loss": 0.0108, "step": 8040 }, { "grad_norm": 0.31904926896095276, "learning_rate": 8.752741446101464e-05, "loss": 0.0108, "step": 8050 }, { "grad_norm": 0.3375629484653473, "learning_rate": 8.749097038004722e-05, "loss": 0.0105, "step": 8060 }, { "grad_norm": 0.36504364013671875, "learning_rate": 8.745448074401561e-05, "loss": 0.0115, "step": 8070 }, { "grad_norm": 0.275889128446579, "learning_rate": 8.741794559725818e-05, "loss": 0.0106, "step": 8080 }, { "grad_norm": 0.3825940191745758, "learning_rate": 8.738136498416856e-05, "loss": 0.0105, "step": 8090 }, { "grad_norm": 0.4263371229171753, "learning_rate": 8.734473894919564e-05, "loss": 0.0125, "step": 8100 }, { "grad_norm": 0.3766123950481415, "learning_rate": 8.730806753684353e-05, "loss": 0.0112, "step": 8110 }, { "grad_norm": 0.27596545219421387, "learning_rate": 8.727135079167144e-05, "loss": 0.0098, "step": 8120 }, { "grad_norm": 0.26507529616355896, "learning_rate": 8.723458875829368e-05, "loss": 0.0112, "step": 8130 }, { "grad_norm": 0.32793647050857544, "learning_rate": 8.719778148137958e-05, "loss": 0.0117, "step": 8140 }, { "grad_norm": 0.5006524920463562, "learning_rate": 8.716092900565346e-05, "loss": 0.0141, "step": 8150 }, { "grad_norm": 0.29135891795158386, "learning_rate": 8.712403137589456e-05, "loss": 0.0117, "step": 8160 }, { "grad_norm": 0.3051094114780426, "learning_rate": 8.708708863693697e-05, "loss": 0.0125, "step": 8170 }, { "grad_norm": 0.33508554100990295, "learning_rate": 8.70501008336696e-05, "loss": 0.0107, "step": 8180 }, { "grad_norm": 0.33325785398483276, "learning_rate": 8.701306801103611e-05, "loss": 0.0101, "step": 8190 }, { "grad_norm": 0.3076310455799103, "learning_rate": 8.697599021403489e-05, "loss": 0.0119, "step": 8200 }, { "grad_norm": 0.2822284996509552, "learning_rate": 8.693886748771896e-05, "loss": 0.0102, "step": 8210 }, { "grad_norm": 0.29608970880508423, "learning_rate": 8.690169987719593e-05, "loss": 0.011, "step": 8220 }, { "grad_norm": 0.36461561918258667, "learning_rate": 8.686448742762791e-05, "loss": 0.0107, "step": 8230 }, { "grad_norm": 0.3555935323238373, "learning_rate": 8.682723018423161e-05, "loss": 0.0106, "step": 8240 }, { "grad_norm": 0.2973068356513977, "learning_rate": 8.678992819227804e-05, "loss": 0.011, "step": 8250 }, { "grad_norm": 0.3376704156398773, "learning_rate": 8.675258149709265e-05, "loss": 0.0104, "step": 8260 }, { "grad_norm": 0.340263694524765, "learning_rate": 8.67151901440552e-05, "loss": 0.0101, "step": 8270 }, { "grad_norm": 0.293972373008728, "learning_rate": 8.667775417859971e-05, "loss": 0.0097, "step": 8280 }, { "grad_norm": 0.27219337224960327, "learning_rate": 8.664027364621441e-05, "loss": 0.0105, "step": 8290 }, { "grad_norm": 0.27255040407180786, "learning_rate": 8.660274859244168e-05, "loss": 0.0123, "step": 8300 }, { "grad_norm": 0.3804474472999573, "learning_rate": 8.656517906287799e-05, "loss": 0.0103, "step": 8310 }, { "grad_norm": 0.4029916226863861, "learning_rate": 8.652756510317387e-05, "loss": 0.0116, "step": 8320 }, { "grad_norm": 0.3002796173095703, "learning_rate": 8.648990675903382e-05, "loss": 0.0106, "step": 8330 }, { "grad_norm": 0.2673972249031067, "learning_rate": 8.64522040762163e-05, "loss": 0.0105, "step": 8340 }, { "grad_norm": 0.31608739495277405, "learning_rate": 8.64144571005336e-05, "loss": 0.0127, "step": 8350 }, { "grad_norm": 0.2921021580696106, "learning_rate": 8.637666587785184e-05, "loss": 0.011, "step": 8360 }, { "grad_norm": 0.2442844808101654, "learning_rate": 8.633883045409096e-05, "loss": 0.0101, "step": 8370 }, { "grad_norm": 0.2928016483783722, "learning_rate": 8.630095087522457e-05, "loss": 0.0109, "step": 8380 }, { "grad_norm": 0.3342704474925995, "learning_rate": 8.62630271872799e-05, "loss": 0.0104, "step": 8390 }, { "grad_norm": 0.30834606289863586, "learning_rate": 8.62250594363378e-05, "loss": 0.0114, "step": 8400 }, { "grad_norm": 0.2779328227043152, "learning_rate": 8.61870476685327e-05, "loss": 0.0122, "step": 8410 }, { "grad_norm": 0.37507641315460205, "learning_rate": 8.614899193005248e-05, "loss": 0.0105, "step": 8420 }, { "grad_norm": 0.3060291111469269, "learning_rate": 8.611089226713842e-05, "loss": 0.0097, "step": 8430 }, { "grad_norm": 0.3532843291759491, "learning_rate": 8.607274872608521e-05, "loss": 0.0101, "step": 8440 }, { "grad_norm": 0.3082657754421234, "learning_rate": 8.603456135324088e-05, "loss": 0.0108, "step": 8450 }, { "grad_norm": 0.27097657322883606, "learning_rate": 8.599633019500665e-05, "loss": 0.0097, "step": 8460 }, { "grad_norm": 0.4173792600631714, "learning_rate": 8.595805529783702e-05, "loss": 0.0097, "step": 8470 }, { "grad_norm": 0.48021912574768066, "learning_rate": 8.591973670823958e-05, "loss": 0.0107, "step": 8480 }, { "grad_norm": 0.2856282889842987, "learning_rate": 8.588137447277503e-05, "loss": 0.0104, "step": 8490 }, { "grad_norm": 0.3201524317264557, "learning_rate": 8.584296863805709e-05, "loss": 0.0119, "step": 8500 }, { "grad_norm": 0.3024746775627136, "learning_rate": 8.580451925075248e-05, "loss": 0.0113, "step": 8510 }, { "grad_norm": 0.37119582295417786, "learning_rate": 8.576602635758085e-05, "loss": 0.0107, "step": 8520 }, { "grad_norm": 0.3784583806991577, "learning_rate": 8.572749000531469e-05, "loss": 0.0108, "step": 8530 }, { "grad_norm": 0.2842702567577362, "learning_rate": 8.568891024077926e-05, "loss": 0.0105, "step": 8540 }, { "grad_norm": 0.382940411567688, "learning_rate": 8.565028711085265e-05, "loss": 0.0105, "step": 8550 }, { "grad_norm": 0.30965447425842285, "learning_rate": 8.561162066246561e-05, "loss": 0.0105, "step": 8560 }, { "grad_norm": 0.32315099239349365, "learning_rate": 8.557291094260152e-05, "loss": 0.0108, "step": 8570 }, { "grad_norm": 0.3546385169029236, "learning_rate": 8.553415799829629e-05, "loss": 0.0107, "step": 8580 }, { "grad_norm": 0.4725189208984375, "learning_rate": 8.549536187663847e-05, "loss": 0.011, "step": 8590 }, { "grad_norm": 0.37511712312698364, "learning_rate": 8.545652262476897e-05, "loss": 0.0104, "step": 8600 }, { "grad_norm": 0.38934120535850525, "learning_rate": 8.541764028988115e-05, "loss": 0.0115, "step": 8610 }, { "grad_norm": 0.3524245023727417, "learning_rate": 8.537871491922071e-05, "loss": 0.0099, "step": 8620 }, { "grad_norm": 0.30628371238708496, "learning_rate": 8.533974656008566e-05, "loss": 0.0103, "step": 8630 }, { "grad_norm": 0.31296461820602417, "learning_rate": 8.530073525982621e-05, "loss": 0.0095, "step": 8640 }, { "grad_norm": 0.35006681084632874, "learning_rate": 8.526168106584476e-05, "loss": 0.0106, "step": 8650 }, { "grad_norm": 0.3150104880332947, "learning_rate": 8.522258402559586e-05, "loss": 0.011, "step": 8660 }, { "grad_norm": 0.34715455770492554, "learning_rate": 8.518344418658613e-05, "loss": 0.0102, "step": 8670 }, { "grad_norm": 0.2967567443847656, "learning_rate": 8.51442615963741e-05, "loss": 0.0103, "step": 8680 }, { "grad_norm": 0.32458803057670593, "learning_rate": 8.510503630257034e-05, "loss": 0.0102, "step": 8690 }, { "grad_norm": 0.3020278215408325, "learning_rate": 8.50657683528373e-05, "loss": 0.0121, "step": 8700 }, { "grad_norm": 0.243942990899086, "learning_rate": 8.502645779488922e-05, "loss": 0.0095, "step": 8710 }, { "grad_norm": 0.2725197374820709, "learning_rate": 8.498710467649213e-05, "loss": 0.0093, "step": 8720 }, { "grad_norm": 0.341840922832489, "learning_rate": 8.494770904546381e-05, "loss": 0.0098, "step": 8730 }, { "grad_norm": 0.34156742691993713, "learning_rate": 8.490827094967363e-05, "loss": 0.0099, "step": 8740 }, { "grad_norm": 0.2555910050868988, "learning_rate": 8.486879043704263e-05, "loss": 0.0098, "step": 8750 }, { "grad_norm": 0.2553873658180237, "learning_rate": 8.482926755554332e-05, "loss": 0.0093, "step": 8760 }, { "grad_norm": 0.3187851011753082, "learning_rate": 8.478970235319975e-05, "loss": 0.011, "step": 8770 }, { "grad_norm": 0.4050900936126709, "learning_rate": 8.475009487808737e-05, "loss": 0.0105, "step": 8780 }, { "grad_norm": 0.36939358711242676, "learning_rate": 8.471044517833298e-05, "loss": 0.0115, "step": 8790 }, { "grad_norm": 0.3504938781261444, "learning_rate": 8.467075330211474e-05, "loss": 0.0117, "step": 8800 }, { "grad_norm": 0.3362886905670166, "learning_rate": 8.463101929766198e-05, "loss": 0.0094, "step": 8810 }, { "grad_norm": 0.3649793267250061, "learning_rate": 8.459124321325528e-05, "loss": 0.0111, "step": 8820 }, { "grad_norm": 0.24022740125656128, "learning_rate": 8.455142509722634e-05, "loss": 0.0109, "step": 8830 }, { "grad_norm": 0.33446770906448364, "learning_rate": 8.451156499795792e-05, "loss": 0.0094, "step": 8840 }, { "grad_norm": 0.3062475621700287, "learning_rate": 8.447166296388381e-05, "loss": 0.0111, "step": 8850 }, { "grad_norm": 0.36087968945503235, "learning_rate": 8.443171904348872e-05, "loss": 0.0089, "step": 8860 }, { "grad_norm": 0.282789021730423, "learning_rate": 8.43917332853083e-05, "loss": 0.0097, "step": 8870 }, { "grad_norm": 0.2516290545463562, "learning_rate": 8.435170573792901e-05, "loss": 0.0089, "step": 8880 }, { "grad_norm": 0.29854705929756165, "learning_rate": 8.431163644998808e-05, "loss": 0.009, "step": 8890 }, { "grad_norm": 0.3754189908504486, "learning_rate": 8.42715254701735e-05, "loss": 0.0096, "step": 8900 }, { "grad_norm": 0.33703139424324036, "learning_rate": 8.42313728472239e-05, "loss": 0.011, "step": 8910 }, { "grad_norm": 0.3538967967033386, "learning_rate": 8.419117862992844e-05, "loss": 0.0096, "step": 8920 }, { "grad_norm": 0.3214809000492096, "learning_rate": 8.415094286712694e-05, "loss": 0.0098, "step": 8930 }, { "grad_norm": 0.38746607303619385, "learning_rate": 8.411066560770965e-05, "loss": 0.0099, "step": 8940 }, { "grad_norm": 0.23421555757522583, "learning_rate": 8.407034690061721e-05, "loss": 0.0093, "step": 8950 }, { "grad_norm": 0.25281527638435364, "learning_rate": 8.402998679484067e-05, "loss": 0.01, "step": 8960 }, { "grad_norm": 0.3989478647708893, "learning_rate": 8.398958533942135e-05, "loss": 0.0099, "step": 8970 }, { "grad_norm": 0.24676960706710815, "learning_rate": 8.394914258345083e-05, "loss": 0.0091, "step": 8980 }, { "grad_norm": 0.28471508622169495, "learning_rate": 8.390865857607089e-05, "loss": 0.0096, "step": 8990 }, { "grad_norm": 0.3082621991634369, "learning_rate": 8.386813336647339e-05, "loss": 0.0095, "step": 9000 }, { "grad_norm": 0.2667902410030365, "learning_rate": 8.38275670039003e-05, "loss": 0.0098, "step": 9010 }, { "grad_norm": 0.2579623758792877, "learning_rate": 8.378695953764356e-05, "loss": 0.0098, "step": 9020 }, { "grad_norm": 0.42782002687454224, "learning_rate": 8.374631101704509e-05, "loss": 0.0099, "step": 9030 }, { "grad_norm": 0.2860717475414276, "learning_rate": 8.370562149149665e-05, "loss": 0.0102, "step": 9040 }, { "grad_norm": 0.25799623131752014, "learning_rate": 8.366489101043989e-05, "loss": 0.0099, "step": 9050 }, { "grad_norm": 0.31252026557922363, "learning_rate": 8.362411962336614e-05, "loss": 0.0096, "step": 9060 }, { "grad_norm": 0.23370763659477234, "learning_rate": 8.358330737981651e-05, "loss": 0.01, "step": 9070 }, { "grad_norm": 0.3088170886039734, "learning_rate": 8.35424543293817e-05, "loss": 0.009, "step": 9080 }, { "grad_norm": 0.29783859848976135, "learning_rate": 8.350156052170206e-05, "loss": 0.0115, "step": 9090 }, { "grad_norm": 0.3804084360599518, "learning_rate": 8.346062600646738e-05, "loss": 0.0096, "step": 9100 }, { "grad_norm": 0.30282413959503174, "learning_rate": 8.341965083341697e-05, "loss": 0.0103, "step": 9110 }, { "grad_norm": 0.38241881132125854, "learning_rate": 8.337863505233953e-05, "loss": 0.0102, "step": 9120 }, { "grad_norm": 0.3249412178993225, "learning_rate": 8.33375787130731e-05, "loss": 0.009, "step": 9130 }, { "grad_norm": 0.27458664774894714, "learning_rate": 8.3296481865505e-05, "loss": 0.0099, "step": 9140 }, { "grad_norm": 0.4108836352825165, "learning_rate": 8.325534455957181e-05, "loss": 0.0101, "step": 9150 }, { "grad_norm": 0.4223904311656952, "learning_rate": 8.321416684525917e-05, "loss": 0.0105, "step": 9160 }, { "grad_norm": 0.32968321442604065, "learning_rate": 8.317294877260193e-05, "loss": 0.0108, "step": 9170 }, { "grad_norm": 0.26523229479789734, "learning_rate": 8.313169039168394e-05, "loss": 0.0092, "step": 9180 }, { "grad_norm": 0.2565534710884094, "learning_rate": 8.3090391752638e-05, "loss": 0.0099, "step": 9190 }, { "grad_norm": 0.308163583278656, "learning_rate": 8.304905290564586e-05, "loss": 0.01, "step": 9200 }, { "grad_norm": 0.3425454795360565, "learning_rate": 8.300767390093813e-05, "loss": 0.0113, "step": 9210 }, { "grad_norm": 0.19908463954925537, "learning_rate": 8.296625478879417e-05, "loss": 0.0103, "step": 9220 }, { "grad_norm": 0.316671758890152, "learning_rate": 8.292479561954214e-05, "loss": 0.0102, "step": 9230 }, { "grad_norm": 0.3078480660915375, "learning_rate": 8.288329644355883e-05, "loss": 0.0096, "step": 9240 }, { "grad_norm": 0.29191339015960693, "learning_rate": 8.284175731126965e-05, "loss": 0.0096, "step": 9250 }, { "grad_norm": 0.35205793380737305, "learning_rate": 8.280017827314854e-05, "loss": 0.0109, "step": 9260 }, { "grad_norm": 0.36453571915626526, "learning_rate": 8.2758559379718e-05, "loss": 0.0104, "step": 9270 }, { "grad_norm": 0.2580679655075073, "learning_rate": 8.271690068154887e-05, "loss": 0.0098, "step": 9280 }, { "grad_norm": 0.2611638605594635, "learning_rate": 8.26752022292604e-05, "loss": 0.0101, "step": 9290 }, { "grad_norm": 0.29041630029678345, "learning_rate": 8.263346407352018e-05, "loss": 0.0097, "step": 9300 }, { "grad_norm": 0.2506728172302246, "learning_rate": 8.259168626504395e-05, "loss": 0.0095, "step": 9310 }, { "grad_norm": 0.2942103445529938, "learning_rate": 8.254986885459572e-05, "loss": 0.0093, "step": 9320 }, { "grad_norm": 0.30901575088500977, "learning_rate": 8.250801189298757e-05, "loss": 0.0092, "step": 9330 }, { "grad_norm": 0.381786584854126, "learning_rate": 8.246611543107967e-05, "loss": 0.0087, "step": 9340 }, { "grad_norm": 0.3429117500782013, "learning_rate": 8.242417951978018e-05, "loss": 0.0098, "step": 9350 }, { "grad_norm": 0.30365362763404846, "learning_rate": 8.238220421004517e-05, "loss": 0.0093, "step": 9360 }, { "grad_norm": 0.3773025572299957, "learning_rate": 8.234018955287859e-05, "loss": 0.0098, "step": 9370 }, { "grad_norm": 0.2867434620857239, "learning_rate": 8.229813559933224e-05, "loss": 0.0096, "step": 9380 }, { "grad_norm": 0.36871495842933655, "learning_rate": 8.22560424005056e-05, "loss": 0.0097, "step": 9390 }, { "grad_norm": 0.29946839809417725, "learning_rate": 8.22139100075459e-05, "loss": 0.0101, "step": 9400 }, { "grad_norm": 0.33686700463294983, "learning_rate": 8.217173847164799e-05, "loss": 0.0104, "step": 9410 }, { "grad_norm": 0.3452436923980713, "learning_rate": 8.212952784405422e-05, "loss": 0.0095, "step": 9420 }, { "grad_norm": 0.3390505313873291, "learning_rate": 8.208727817605451e-05, "loss": 0.0103, "step": 9430 }, { "grad_norm": 0.2770306468009949, "learning_rate": 8.204498951898618e-05, "loss": 0.0089, "step": 9440 }, { "grad_norm": 0.26577332615852356, "learning_rate": 8.200266192423395e-05, "loss": 0.0091, "step": 9450 }, { "grad_norm": 0.31393006443977356, "learning_rate": 8.196029544322981e-05, "loss": 0.0098, "step": 9460 }, { "grad_norm": 0.2694907784461975, "learning_rate": 8.191789012745305e-05, "loss": 0.0088, "step": 9470 }, { "grad_norm": 0.28339341282844543, "learning_rate": 8.187544602843014e-05, "loss": 0.0091, "step": 9480 }, { "grad_norm": 0.28915339708328247, "learning_rate": 8.183296319773466e-05, "loss": 0.0104, "step": 9490 }, { "grad_norm": 0.2694634199142456, "learning_rate": 8.179044168698721e-05, "loss": 0.0096, "step": 9500 }, { "grad_norm": 0.31086036562919617, "learning_rate": 8.174788154785548e-05, "loss": 0.0088, "step": 9510 }, { "grad_norm": 0.31560900807380676, "learning_rate": 8.170528283205403e-05, "loss": 0.0103, "step": 9520 }, { "grad_norm": 0.2754019498825073, "learning_rate": 8.166264559134435e-05, "loss": 0.0113, "step": 9530 }, { "grad_norm": 0.2802848815917969, "learning_rate": 8.161996987753467e-05, "loss": 0.0104, "step": 9540 }, { "grad_norm": 0.2925932705402374, "learning_rate": 8.157725574248e-05, "loss": 0.0103, "step": 9550 }, { "grad_norm": 0.3291204869747162, "learning_rate": 8.153450323808209e-05, "loss": 0.0115, "step": 9560 }, { "grad_norm": 0.3283328711986542, "learning_rate": 8.149171241628923e-05, "loss": 0.0106, "step": 9570 }, { "grad_norm": 0.3074251413345337, "learning_rate": 8.144888332909631e-05, "loss": 0.0098, "step": 9580 }, { "grad_norm": 0.27979621291160583, "learning_rate": 8.14060160285447e-05, "loss": 0.0089, "step": 9590 }, { "grad_norm": 0.26148489117622375, "learning_rate": 8.136311056672223e-05, "loss": 0.0116, "step": 9600 }, { "grad_norm": 0.3204076886177063, "learning_rate": 8.132016699576308e-05, "loss": 0.0097, "step": 9610 }, { "grad_norm": 0.3098124861717224, "learning_rate": 8.127718536784771e-05, "loss": 0.0094, "step": 9620 }, { "grad_norm": 0.3159780204296112, "learning_rate": 8.12341657352029e-05, "loss": 0.0109, "step": 9630 }, { "grad_norm": 0.3584921956062317, "learning_rate": 8.119110815010151e-05, "loss": 0.0099, "step": 9640 }, { "grad_norm": 0.2659761905670166, "learning_rate": 8.114801266486261e-05, "loss": 0.0096, "step": 9650 }, { "grad_norm": 0.3183245360851288, "learning_rate": 8.110487933185122e-05, "loss": 0.0104, "step": 9660 }, { "grad_norm": 0.3073471486568451, "learning_rate": 8.106170820347847e-05, "loss": 0.0094, "step": 9670 }, { "grad_norm": 0.2520548105239868, "learning_rate": 8.101849933220132e-05, "loss": 0.0085, "step": 9680 }, { "grad_norm": 0.270577609539032, "learning_rate": 8.097525277052264e-05, "loss": 0.0109, "step": 9690 }, { "grad_norm": 0.3279707133769989, "learning_rate": 8.093196857099106e-05, "loss": 0.0105, "step": 9700 }, { "grad_norm": 0.41190341114997864, "learning_rate": 8.088864678620096e-05, "loss": 0.0102, "step": 9710 }, { "grad_norm": 0.32842499017715454, "learning_rate": 8.084528746879243e-05, "loss": 0.01, "step": 9720 }, { "grad_norm": 0.3014456331729889, "learning_rate": 8.080189067145108e-05, "loss": 0.0083, "step": 9730 }, { "grad_norm": 0.259885311126709, "learning_rate": 8.075845644690813e-05, "loss": 0.0082, "step": 9740 }, { "grad_norm": 0.2711023986339569, "learning_rate": 8.07149848479403e-05, "loss": 0.0079, "step": 9750 }, { "grad_norm": 0.29589828848838806, "learning_rate": 8.067147592736962e-05, "loss": 0.0095, "step": 9760 }, { "grad_norm": 0.24040688574314117, "learning_rate": 8.062792973806357e-05, "loss": 0.0087, "step": 9770 }, { "grad_norm": 0.3271734118461609, "learning_rate": 8.058434633293485e-05, "loss": 0.0097, "step": 9780 }, { "grad_norm": 0.2870268225669861, "learning_rate": 8.054072576494143e-05, "loss": 0.0092, "step": 9790 }, { "grad_norm": 0.23761416971683502, "learning_rate": 8.04970680870864e-05, "loss": 0.0084, "step": 9800 }, { "grad_norm": 0.31880688667297363, "learning_rate": 8.045337335241793e-05, "loss": 0.0089, "step": 9810 }, { "grad_norm": 0.2461557537317276, "learning_rate": 8.040964161402932e-05, "loss": 0.0088, "step": 9820 }, { "grad_norm": 0.34059616923332214, "learning_rate": 8.036587292505869e-05, "loss": 0.0104, "step": 9830 }, { "grad_norm": 0.2754742503166199, "learning_rate": 8.032206733868912e-05, "loss": 0.0094, "step": 9840 }, { "grad_norm": 0.33969369530677795, "learning_rate": 8.027822490814858e-05, "loss": 0.0089, "step": 9850 }, { "grad_norm": 0.28036972880363464, "learning_rate": 8.023434568670972e-05, "loss": 0.0087, "step": 9860 }, { "grad_norm": 0.3701011538505554, "learning_rate": 8.019042972768992e-05, "loss": 0.0092, "step": 9870 }, { "grad_norm": 0.27006396651268005, "learning_rate": 8.014647708445124e-05, "loss": 0.011, "step": 9880 }, { "grad_norm": 0.31688278913497925, "learning_rate": 8.010248781040027e-05, "loss": 0.0088, "step": 9890 }, { "grad_norm": 0.2984282374382019, "learning_rate": 8.005846195898814e-05, "loss": 0.0098, "step": 9900 }, { "grad_norm": 0.34147998690605164, "learning_rate": 8.00143995837104e-05, "loss": 0.0095, "step": 9910 }, { "grad_norm": 0.30249854922294617, "learning_rate": 7.997030073810698e-05, "loss": 0.0096, "step": 9920 }, { "grad_norm": 0.3344995081424713, "learning_rate": 7.992616547576217e-05, "loss": 0.0088, "step": 9930 }, { "grad_norm": 0.2232092171907425, "learning_rate": 7.988199385030445e-05, "loss": 0.0093, "step": 9940 }, { "grad_norm": 0.28129565715789795, "learning_rate": 7.98377859154065e-05, "loss": 0.0088, "step": 9950 }, { "grad_norm": 0.3453581631183624, "learning_rate": 7.979354172478516e-05, "loss": 0.0106, "step": 9960 }, { "grad_norm": 0.3263741731643677, "learning_rate": 7.974926133220127e-05, "loss": 0.0105, "step": 9970 }, { "grad_norm": 0.32977229356765747, "learning_rate": 7.970494479145967e-05, "loss": 0.011, "step": 9980 }, { "grad_norm": 0.3182139992713928, "learning_rate": 7.966059215640919e-05, "loss": 0.009, "step": 9990 }, { "grad_norm": 0.25233185291290283, "learning_rate": 7.961620348094241e-05, "loss": 0.0088, "step": 10000 }, { "grad_norm": 0.28312796354293823, "learning_rate": 7.957177881899577e-05, "loss": 0.0082, "step": 10010 }, { "grad_norm": 0.2667032778263092, "learning_rate": 7.952731822454944e-05, "loss": 0.0092, "step": 10020 }, { "grad_norm": 0.2328304648399353, "learning_rate": 7.948282175162722e-05, "loss": 0.008, "step": 10030 }, { "grad_norm": 0.26578933000564575, "learning_rate": 7.943828945429652e-05, "loss": 0.0086, "step": 10040 }, { "grad_norm": 0.3218497335910797, "learning_rate": 7.939372138666827e-05, "loss": 0.0087, "step": 10050 }, { "grad_norm": 0.2721211016178131, "learning_rate": 7.934911760289692e-05, "loss": 0.0093, "step": 10060 }, { "grad_norm": 0.2776382565498352, "learning_rate": 7.930447815718022e-05, "loss": 0.0089, "step": 10070 }, { "grad_norm": 0.3087117075920105, "learning_rate": 7.925980310375934e-05, "loss": 0.0094, "step": 10080 }, { "grad_norm": 0.3460996150970459, "learning_rate": 7.921509249691865e-05, "loss": 0.0097, "step": 10090 }, { "grad_norm": 0.345557302236557, "learning_rate": 7.917034639098579e-05, "loss": 0.0091, "step": 10100 }, { "grad_norm": 0.32438960671424866, "learning_rate": 7.912556484033146e-05, "loss": 0.0098, "step": 10110 }, { "grad_norm": 0.2700251340866089, "learning_rate": 7.908074789936952e-05, "loss": 0.0088, "step": 10120 }, { "grad_norm": 0.259725958108902, "learning_rate": 7.903589562255673e-05, "loss": 0.0086, "step": 10130 }, { "grad_norm": 0.2990911304950714, "learning_rate": 7.899100806439285e-05, "loss": 0.0087, "step": 10140 }, { "grad_norm": 0.32164642214775085, "learning_rate": 7.894608527942049e-05, "loss": 0.0092, "step": 10150 }, { "grad_norm": 0.2375648319721222, "learning_rate": 7.89011273222251e-05, "loss": 0.0081, "step": 10160 }, { "grad_norm": 0.2520039677619934, "learning_rate": 7.885613424743482e-05, "loss": 0.0082, "step": 10170 }, { "grad_norm": 0.28493061661720276, "learning_rate": 7.881110610972044e-05, "loss": 0.0091, "step": 10180 }, { "grad_norm": 0.2752293050289154, "learning_rate": 7.876604296379544e-05, "loss": 0.0101, "step": 10190 }, { "grad_norm": 0.2828473448753357, "learning_rate": 7.872094486441579e-05, "loss": 0.0087, "step": 10200 }, { "grad_norm": 0.2889810800552368, "learning_rate": 7.867581186637991e-05, "loss": 0.0082, "step": 10210 }, { "grad_norm": 0.2776321768760681, "learning_rate": 7.863064402452866e-05, "loss": 0.0088, "step": 10220 }, { "grad_norm": 0.2540366053581238, "learning_rate": 7.858544139374524e-05, "loss": 0.0091, "step": 10230 }, { "grad_norm": 0.31785818934440613, "learning_rate": 7.854020402895508e-05, "loss": 0.0109, "step": 10240 }, { "grad_norm": 0.24002966284751892, "learning_rate": 7.849493198512587e-05, "loss": 0.0083, "step": 10250 }, { "grad_norm": 0.249736025929451, "learning_rate": 7.84496253172674e-05, "loss": 0.0076, "step": 10260 }, { "grad_norm": 0.3291330337524414, "learning_rate": 7.840428408043155e-05, "loss": 0.0095, "step": 10270 }, { "grad_norm": 0.22596527636051178, "learning_rate": 7.835890832971218e-05, "loss": 0.0086, "step": 10280 }, { "grad_norm": 0.3245750963687897, "learning_rate": 7.831349812024512e-05, "loss": 0.0095, "step": 10290 }, { "grad_norm": 0.23606570065021515, "learning_rate": 7.826805350720807e-05, "loss": 0.0091, "step": 10300 }, { "grad_norm": 0.2839299142360687, "learning_rate": 7.822257454582049e-05, "loss": 0.0085, "step": 10310 }, { "grad_norm": 0.22629347443580627, "learning_rate": 7.817706129134363e-05, "loss": 0.0087, "step": 10320 }, { "grad_norm": 0.375497967004776, "learning_rate": 7.813151379908036e-05, "loss": 0.0079, "step": 10330 }, { "grad_norm": 0.2748655378818512, "learning_rate": 7.808593212437523e-05, "loss": 0.0082, "step": 10340 }, { "grad_norm": 0.3295239806175232, "learning_rate": 7.804031632261421e-05, "loss": 0.0096, "step": 10350 }, { "grad_norm": 0.33085861802101135, "learning_rate": 7.799466644922484e-05, "loss": 0.011, "step": 10360 }, { "grad_norm": 0.2350122630596161, "learning_rate": 7.794898255967602e-05, "loss": 0.0086, "step": 10370 }, { "grad_norm": 0.27413979172706604, "learning_rate": 7.790326470947795e-05, "loss": 0.0084, "step": 10380 }, { "grad_norm": 0.2807300090789795, "learning_rate": 7.785751295418217e-05, "loss": 0.0091, "step": 10390 }, { "grad_norm": 0.24402906000614166, "learning_rate": 7.781172734938136e-05, "loss": 0.0078, "step": 10400 }, { "grad_norm": 0.1861390620470047, "learning_rate": 7.776590795070933e-05, "loss": 0.0086, "step": 10410 }, { "grad_norm": 0.2955159842967987, "learning_rate": 7.772005481384099e-05, "loss": 0.009, "step": 10420 }, { "grad_norm": 0.3031066954135895, "learning_rate": 7.767416799449223e-05, "loss": 0.0088, "step": 10430 }, { "grad_norm": 0.30942216515541077, "learning_rate": 7.762824754841984e-05, "loss": 0.0084, "step": 10440 }, { "grad_norm": 0.27772602438926697, "learning_rate": 7.758229353142152e-05, "loss": 0.0088, "step": 10450 }, { "grad_norm": 0.18874435126781464, "learning_rate": 7.753630599933572e-05, "loss": 0.009, "step": 10460 }, { "grad_norm": 0.20754432678222656, "learning_rate": 7.749028500804161e-05, "loss": 0.0093, "step": 10470 }, { "grad_norm": 0.3341434895992279, "learning_rate": 7.744423061345906e-05, "loss": 0.0092, "step": 10480 }, { "grad_norm": 0.24594378471374512, "learning_rate": 7.73981428715485e-05, "loss": 0.0087, "step": 10490 }, { "grad_norm": 0.2970474362373352, "learning_rate": 7.735202183831085e-05, "loss": 0.0087, "step": 10500 }, { "grad_norm": 0.2976945638656616, "learning_rate": 7.730586756978757e-05, "loss": 0.0088, "step": 10510 }, { "grad_norm": 0.2520144581794739, "learning_rate": 7.72596801220604e-05, "loss": 0.009, "step": 10520 }, { "grad_norm": 0.21115921437740326, "learning_rate": 7.721345955125147e-05, "loss": 0.0076, "step": 10530 }, { "grad_norm": 0.2453879415988922, "learning_rate": 7.71672059135231e-05, "loss": 0.0082, "step": 10540 }, { "grad_norm": 0.333455890417099, "learning_rate": 7.712091926507787e-05, "loss": 0.0082, "step": 10550 }, { "grad_norm": 0.20065860450267792, "learning_rate": 7.70745996621584e-05, "loss": 0.0082, "step": 10560 }, { "grad_norm": 0.3492734134197235, "learning_rate": 7.702824716104735e-05, "loss": 0.0091, "step": 10570 }, { "grad_norm": 0.2791425585746765, "learning_rate": 7.698186181806743e-05, "loss": 0.0087, "step": 10580 }, { "grad_norm": 0.2732599079608917, "learning_rate": 7.693544368958116e-05, "loss": 0.0083, "step": 10590 }, { "grad_norm": 0.30734577775001526, "learning_rate": 7.688899283199096e-05, "loss": 0.0087, "step": 10600 }, { "grad_norm": 0.29557982087135315, "learning_rate": 7.684250930173901e-05, "loss": 0.0087, "step": 10610 }, { "grad_norm": 0.27425116300582886, "learning_rate": 7.679599315530718e-05, "loss": 0.0099, "step": 10620 }, { "grad_norm": 0.2503387928009033, "learning_rate": 7.674944444921695e-05, "loss": 0.0089, "step": 10630 }, { "grad_norm": 0.22014181315898895, "learning_rate": 7.670286324002944e-05, "loss": 0.0083, "step": 10640 }, { "grad_norm": 0.2801256477832794, "learning_rate": 7.665624958434514e-05, "loss": 0.0085, "step": 10650 }, { "grad_norm": 0.2603195905685425, "learning_rate": 7.66096035388041e-05, "loss": 0.0101, "step": 10660 }, { "grad_norm": 0.32898423075675964, "learning_rate": 7.656292516008563e-05, "loss": 0.0091, "step": 10670 }, { "grad_norm": 0.2944939136505127, "learning_rate": 7.651621450490837e-05, "loss": 0.0093, "step": 10680 }, { "grad_norm": 0.3123464584350586, "learning_rate": 7.646947163003017e-05, "loss": 0.0086, "step": 10690 }, { "grad_norm": 0.29304465651512146, "learning_rate": 7.642269659224803e-05, "loss": 0.0087, "step": 10700 }, { "grad_norm": 0.3500259220600128, "learning_rate": 7.637588944839803e-05, "loss": 0.009, "step": 10710 }, { "grad_norm": 0.26453834772109985, "learning_rate": 7.632905025535528e-05, "loss": 0.0089, "step": 10720 }, { "grad_norm": 0.25538867712020874, "learning_rate": 7.628217907003378e-05, "loss": 0.0082, "step": 10730 }, { "grad_norm": 0.21325016021728516, "learning_rate": 7.623527594938649e-05, "loss": 0.0084, "step": 10740 }, { "grad_norm": 0.2767277657985687, "learning_rate": 7.618834095040509e-05, "loss": 0.008, "step": 10750 }, { "grad_norm": 0.2665347158908844, "learning_rate": 7.614137413012e-05, "loss": 0.0091, "step": 10760 }, { "grad_norm": 0.2695922255516052, "learning_rate": 7.609437554560042e-05, "loss": 0.0078, "step": 10770 }, { "grad_norm": 0.23686271905899048, "learning_rate": 7.604734525395398e-05, "loss": 0.0083, "step": 10780 }, { "grad_norm": 0.23653416335582733, "learning_rate": 7.600028331232699e-05, "loss": 0.0085, "step": 10790 }, { "grad_norm": 0.20893239974975586, "learning_rate": 7.595318977790408e-05, "loss": 0.0081, "step": 10800 }, { "grad_norm": 0.348706990480423, "learning_rate": 7.590606470790836e-05, "loss": 0.0086, "step": 10810 }, { "grad_norm": 0.27015960216522217, "learning_rate": 7.585890815960125e-05, "loss": 0.0082, "step": 10820 }, { "grad_norm": 0.22292320430278778, "learning_rate": 7.581172019028238e-05, "loss": 0.0077, "step": 10830 }, { "grad_norm": 0.3161787688732147, "learning_rate": 7.576450085728959e-05, "loss": 0.0079, "step": 10840 }, { "grad_norm": 0.4077858626842499, "learning_rate": 7.571725021799885e-05, "loss": 0.0084, "step": 10850 }, { "grad_norm": 0.3302555978298187, "learning_rate": 7.566996832982409e-05, "loss": 0.0094, "step": 10860 }, { "grad_norm": 0.2723492980003357, "learning_rate": 7.56226552502173e-05, "loss": 0.0082, "step": 10870 }, { "grad_norm": 0.2489624172449112, "learning_rate": 7.557531103666833e-05, "loss": 0.0096, "step": 10880 }, { "grad_norm": 0.2531261146068573, "learning_rate": 7.552793574670485e-05, "loss": 0.0087, "step": 10890 }, { "grad_norm": 0.3141566812992096, "learning_rate": 7.54805294378923e-05, "loss": 0.0078, "step": 10900 }, { "grad_norm": 0.2886313498020172, "learning_rate": 7.543309216783384e-05, "loss": 0.0084, "step": 10910 }, { "grad_norm": 0.33143842220306396, "learning_rate": 7.53856239941702e-05, "loss": 0.0084, "step": 10920 }, { "grad_norm": 0.3636597692966461, "learning_rate": 7.533812497457972e-05, "loss": 0.009, "step": 10930 }, { "grad_norm": 0.2903405725955963, "learning_rate": 7.529059516677814e-05, "loss": 0.0088, "step": 10940 }, { "grad_norm": 0.2362307757139206, "learning_rate": 7.524303462851872e-05, "loss": 0.0078, "step": 10950 }, { "grad_norm": 0.2994009256362915, "learning_rate": 7.519544341759192e-05, "loss": 0.0084, "step": 10960 }, { "grad_norm": 0.22063317894935608, "learning_rate": 7.514782159182562e-05, "loss": 0.0096, "step": 10970 }, { "grad_norm": 0.22048303484916687, "learning_rate": 7.510016920908481e-05, "loss": 0.0078, "step": 10980 }, { "grad_norm": 0.3061755895614624, "learning_rate": 7.505248632727158e-05, "loss": 0.0086, "step": 10990 }, { "grad_norm": 0.2510005235671997, "learning_rate": 7.50047730043252e-05, "loss": 0.0085, "step": 11000 }, { "grad_norm": 0.2653082609176636, "learning_rate": 7.495702929822183e-05, "loss": 0.0078, "step": 11010 }, { "grad_norm": 0.24982401728630066, "learning_rate": 7.490925526697455e-05, "loss": 0.008, "step": 11020 }, { "grad_norm": 0.33277788758277893, "learning_rate": 7.486145096863334e-05, "loss": 0.0075, "step": 11030 }, { "grad_norm": 0.2504388689994812, "learning_rate": 7.481361646128491e-05, "loss": 0.0085, "step": 11040 }, { "grad_norm": 0.264525443315506, "learning_rate": 7.476575180305271e-05, "loss": 0.0081, "step": 11050 }, { "grad_norm": 0.25063708424568176, "learning_rate": 7.471785705209682e-05, "loss": 0.0076, "step": 11060 }, { "grad_norm": 0.3602023124694824, "learning_rate": 7.466993226661387e-05, "loss": 0.0089, "step": 11070 }, { "grad_norm": 0.3457334339618683, "learning_rate": 7.462197750483699e-05, "loss": 0.0084, "step": 11080 }, { "grad_norm": 0.3209291398525238, "learning_rate": 7.457399282503574e-05, "loss": 0.0084, "step": 11090 }, { "grad_norm": 0.2748512923717499, "learning_rate": 7.452597828551604e-05, "loss": 0.0079, "step": 11100 }, { "grad_norm": 0.29751718044281006, "learning_rate": 7.447793394462006e-05, "loss": 0.0089, "step": 11110 }, { "grad_norm": 0.25582262873649597, "learning_rate": 7.442985986072624e-05, "loss": 0.008, "step": 11120 }, { "grad_norm": 0.2571868300437927, "learning_rate": 7.438175609224908e-05, "loss": 0.0077, "step": 11130 }, { "grad_norm": 0.24162840843200684, "learning_rate": 7.433362269763924e-05, "loss": 0.0083, "step": 11140 }, { "grad_norm": 0.2867334187030792, "learning_rate": 7.428545973538329e-05, "loss": 0.0084, "step": 11150 }, { "grad_norm": 0.2729656994342804, "learning_rate": 7.42372672640038e-05, "loss": 0.0082, "step": 11160 }, { "grad_norm": 0.26211899518966675, "learning_rate": 7.418904534205917e-05, "loss": 0.0086, "step": 11170 }, { "grad_norm": 0.2512105405330658, "learning_rate": 7.414079402814356e-05, "loss": 0.0079, "step": 11180 }, { "grad_norm": 0.28634539246559143, "learning_rate": 7.409251338088687e-05, "loss": 0.0079, "step": 11190 }, { "grad_norm": 0.28483590483665466, "learning_rate": 7.404420345895467e-05, "loss": 0.009, "step": 11200 }, { "grad_norm": 0.27629390358924866, "learning_rate": 7.399586432104804e-05, "loss": 0.0093, "step": 11210 }, { "grad_norm": 0.32245320081710815, "learning_rate": 7.394749602590358e-05, "loss": 0.0085, "step": 11220 }, { "grad_norm": 0.37746337056159973, "learning_rate": 7.389909863229336e-05, "loss": 0.0104, "step": 11230 }, { "grad_norm": 0.2625359892845154, "learning_rate": 7.385067219902477e-05, "loss": 0.0074, "step": 11240 }, { "grad_norm": 0.3064190149307251, "learning_rate": 7.380221678494049e-05, "loss": 0.0078, "step": 11250 }, { "grad_norm": 0.3272659480571747, "learning_rate": 7.37537324489184e-05, "loss": 0.0092, "step": 11260 }, { "grad_norm": 0.23744618892669678, "learning_rate": 7.370521924987155e-05, "loss": 0.0078, "step": 11270 }, { "grad_norm": 0.2976019084453583, "learning_rate": 7.365667724674806e-05, "loss": 0.0084, "step": 11280 }, { "grad_norm": 0.2628666162490845, "learning_rate": 7.360810649853105e-05, "loss": 0.0089, "step": 11290 }, { "grad_norm": 0.249125674366951, "learning_rate": 7.355950706423853e-05, "loss": 0.0076, "step": 11300 }, { "grad_norm": 0.22084388136863708, "learning_rate": 7.351087900292342e-05, "loss": 0.0095, "step": 11310 }, { "grad_norm": 0.2580353021621704, "learning_rate": 7.346222237367339e-05, "loss": 0.0082, "step": 11320 }, { "grad_norm": 0.2719237506389618, "learning_rate": 7.341353723561081e-05, "loss": 0.0085, "step": 11330 }, { "grad_norm": 0.3120487630367279, "learning_rate": 7.336482364789277e-05, "loss": 0.0088, "step": 11340 }, { "grad_norm": 0.2787841558456421, "learning_rate": 7.331608166971082e-05, "loss": 0.0097, "step": 11350 }, { "grad_norm": 0.3463439643383026, "learning_rate": 7.326731136029108e-05, "loss": 0.0092, "step": 11360 }, { "grad_norm": 0.38997554779052734, "learning_rate": 7.321851277889408e-05, "loss": 0.0082, "step": 11370 }, { "grad_norm": 0.239556223154068, "learning_rate": 7.316968598481469e-05, "loss": 0.0087, "step": 11380 }, { "grad_norm": 0.3022553622722626, "learning_rate": 7.312083103738206e-05, "loss": 0.009, "step": 11390 }, { "grad_norm": 0.2007443904876709, "learning_rate": 7.307194799595958e-05, "loss": 0.0075, "step": 11400 }, { "grad_norm": 0.2427217811346054, "learning_rate": 7.302303691994473e-05, "loss": 0.0074, "step": 11410 }, { "grad_norm": 0.2285958081483841, "learning_rate": 7.29740978687691e-05, "loss": 0.0066, "step": 11420 }, { "grad_norm": 0.30119019746780396, "learning_rate": 7.292513090189825e-05, "loss": 0.0071, "step": 11430 }, { "grad_norm": 0.26628202199935913, "learning_rate": 7.287613607883163e-05, "loss": 0.0076, "step": 11440 }, { "grad_norm": 0.21646924316883087, "learning_rate": 7.282711345910263e-05, "loss": 0.0077, "step": 11450 }, { "grad_norm": 0.2584187090396881, "learning_rate": 7.27780631022783e-05, "loss": 0.0081, "step": 11460 }, { "grad_norm": 0.25080376863479614, "learning_rate": 7.272898506795948e-05, "loss": 0.0093, "step": 11470 }, { "grad_norm": 0.24241487681865692, "learning_rate": 7.267987941578058e-05, "loss": 0.008, "step": 11480 }, { "grad_norm": 0.2837870419025421, "learning_rate": 7.263074620540962e-05, "loss": 0.0086, "step": 11490 }, { "grad_norm": 0.23800799250602722, "learning_rate": 7.25815854965481e-05, "loss": 0.0077, "step": 11500 }, { "grad_norm": 0.20950715243816376, "learning_rate": 7.253239734893089e-05, "loss": 0.0079, "step": 11510 }, { "grad_norm": 0.27476486563682556, "learning_rate": 7.248318182232623e-05, "loss": 0.0081, "step": 11520 }, { "grad_norm": 0.22000481188297272, "learning_rate": 7.243393897653565e-05, "loss": 0.0078, "step": 11530 }, { "grad_norm": 0.22252796590328217, "learning_rate": 7.23846688713938e-05, "loss": 0.0078, "step": 11540 }, { "grad_norm": 0.24442344903945923, "learning_rate": 7.233537156676854e-05, "loss": 0.0082, "step": 11550 }, { "grad_norm": 0.28191429376602173, "learning_rate": 7.228604712256076e-05, "loss": 0.0078, "step": 11560 }, { "grad_norm": 0.24343600869178772, "learning_rate": 7.223669559870426e-05, "loss": 0.0075, "step": 11570 }, { "grad_norm": 0.2522268295288086, "learning_rate": 7.218731705516585e-05, "loss": 0.0094, "step": 11580 }, { "grad_norm": 0.25070708990097046, "learning_rate": 7.21379115519451e-05, "loss": 0.0078, "step": 11590 }, { "grad_norm": 0.21620993316173553, "learning_rate": 7.20884791490743e-05, "loss": 0.0079, "step": 11600 }, { "grad_norm": 0.2193872481584549, "learning_rate": 7.203901990661857e-05, "loss": 0.0077, "step": 11610 }, { "grad_norm": 0.2501574158668518, "learning_rate": 7.198953388467549e-05, "loss": 0.0076, "step": 11620 }, { "grad_norm": 0.3814297914505005, "learning_rate": 7.194002114337526e-05, "loss": 0.0085, "step": 11630 }, { "grad_norm": 0.3631609380245209, "learning_rate": 7.189048174288054e-05, "loss": 0.009, "step": 11640 }, { "grad_norm": 0.35405850410461426, "learning_rate": 7.184091574338636e-05, "loss": 0.0085, "step": 11650 }, { "grad_norm": 0.26702335476875305, "learning_rate": 7.179132320512009e-05, "loss": 0.0083, "step": 11660 }, { "grad_norm": 0.24685519933700562, "learning_rate": 7.174170418834134e-05, "loss": 0.0081, "step": 11670 }, { "grad_norm": 0.24434106051921844, "learning_rate": 7.169205875334189e-05, "loss": 0.0074, "step": 11680 }, { "grad_norm": 0.3144436776638031, "learning_rate": 7.164238696044562e-05, "loss": 0.0081, "step": 11690 }, { "grad_norm": 0.19701911509037018, "learning_rate": 7.159268887000846e-05, "loss": 0.0078, "step": 11700 }, { "grad_norm": 0.25164148211479187, "learning_rate": 7.154296454241827e-05, "loss": 0.0084, "step": 11710 }, { "grad_norm": 0.28079986572265625, "learning_rate": 7.149321403809479e-05, "loss": 0.0088, "step": 11720 }, { "grad_norm": 0.2987631857395172, "learning_rate": 7.144343741748956e-05, "loss": 0.0088, "step": 11730 }, { "grad_norm": 0.25953763723373413, "learning_rate": 7.139363474108589e-05, "loss": 0.0086, "step": 11740 }, { "grad_norm": 0.2453896552324295, "learning_rate": 7.134380606939874e-05, "loss": 0.0074, "step": 11750 }, { "grad_norm": 0.36023178696632385, "learning_rate": 7.12939514629746e-05, "loss": 0.0084, "step": 11760 }, { "grad_norm": 0.24952717125415802, "learning_rate": 7.124407098239155e-05, "loss": 0.0088, "step": 11770 }, { "grad_norm": 0.2500772774219513, "learning_rate": 7.119416468825908e-05, "loss": 0.0084, "step": 11780 }, { "grad_norm": 0.29303258657455444, "learning_rate": 7.114423264121804e-05, "loss": 0.0086, "step": 11790 }, { "grad_norm": 0.25186672806739807, "learning_rate": 7.109427490194056e-05, "loss": 0.0076, "step": 11800 }, { "grad_norm": 0.2728545367717743, "learning_rate": 7.104429153113001e-05, "loss": 0.0076, "step": 11810 }, { "grad_norm": 0.23277583718299866, "learning_rate": 7.099428258952092e-05, "loss": 0.0073, "step": 11820 }, { "grad_norm": 0.3232274055480957, "learning_rate": 7.094424813787883e-05, "loss": 0.0085, "step": 11830 }, { "grad_norm": 0.24028784036636353, "learning_rate": 7.089418823700034e-05, "loss": 0.0074, "step": 11840 }, { "grad_norm": 0.17265081405639648, "learning_rate": 7.084410294771298e-05, "loss": 0.0071, "step": 11850 }, { "grad_norm": 0.2584075331687927, "learning_rate": 7.079399233087504e-05, "loss": 0.0076, "step": 11860 }, { "grad_norm": 0.2501102387905121, "learning_rate": 7.074385644737568e-05, "loss": 0.0078, "step": 11870 }, { "grad_norm": 0.25903502106666565, "learning_rate": 7.069369535813473e-05, "loss": 0.0086, "step": 11880 }, { "grad_norm": 0.1961066722869873, "learning_rate": 7.06435091241026e-05, "loss": 0.0072, "step": 11890 }, { "grad_norm": 0.3219556212425232, "learning_rate": 7.059329780626034e-05, "loss": 0.0069, "step": 11900 }, { "grad_norm": 0.29326993227005005, "learning_rate": 7.054306146561944e-05, "loss": 0.0078, "step": 11910 }, { "grad_norm": 0.3069089353084564, "learning_rate": 7.049280016322178e-05, "loss": 0.0078, "step": 11920 }, { "grad_norm": 0.2479066103696823, "learning_rate": 7.044251396013957e-05, "loss": 0.0077, "step": 11930 }, { "grad_norm": 0.2422267347574234, "learning_rate": 7.039220291747528e-05, "loss": 0.0071, "step": 11940 }, { "grad_norm": 0.21664093434810638, "learning_rate": 7.034186709636159e-05, "loss": 0.007, "step": 11950 }, { "grad_norm": 0.300942599773407, "learning_rate": 7.029150655796129e-05, "loss": 0.0063, "step": 11960 }, { "grad_norm": 0.2523380219936371, "learning_rate": 7.024112136346712e-05, "loss": 0.007, "step": 11970 }, { "grad_norm": 0.25909993052482605, "learning_rate": 7.01907115741019e-05, "loss": 0.0078, "step": 11980 }, { "grad_norm": 0.3041617274284363, "learning_rate": 7.014027725111825e-05, "loss": 0.0085, "step": 11990 }, { "grad_norm": 0.2782539129257202, "learning_rate": 7.008981845579864e-05, "loss": 0.0075, "step": 12000 }, { "grad_norm": 0.3118305802345276, "learning_rate": 7.003933524945528e-05, "loss": 0.0081, "step": 12010 }, { "grad_norm": 0.26557084918022156, "learning_rate": 6.998882769342998e-05, "loss": 0.0072, "step": 12020 }, { "grad_norm": 0.23537853360176086, "learning_rate": 6.993829584909423e-05, "loss": 0.0078, "step": 12030 }, { "grad_norm": 0.22218097746372223, "learning_rate": 6.988773977784895e-05, "loss": 0.0073, "step": 12040 }, { "grad_norm": 0.22848103940486908, "learning_rate": 6.983715954112454e-05, "loss": 0.0077, "step": 12050 }, { "grad_norm": 0.21472135186195374, "learning_rate": 6.978655520038079e-05, "loss": 0.0077, "step": 12060 }, { "grad_norm": 0.26741930842399597, "learning_rate": 6.97359268171067e-05, "loss": 0.007, "step": 12070 }, { "grad_norm": 0.20105968415737152, "learning_rate": 6.968527445282056e-05, "loss": 0.0083, "step": 12080 }, { "grad_norm": 0.3267427682876587, "learning_rate": 6.963459816906977e-05, "loss": 0.0076, "step": 12090 }, { "grad_norm": 0.19545020163059235, "learning_rate": 6.958389802743077e-05, "loss": 0.0083, "step": 12100 }, { "grad_norm": 0.2549017369747162, "learning_rate": 6.953317408950903e-05, "loss": 0.0083, "step": 12110 }, { "grad_norm": 0.23277691006660461, "learning_rate": 6.94824264169389e-05, "loss": 0.0071, "step": 12120 }, { "grad_norm": 0.2652171850204468, "learning_rate": 6.94316550713836e-05, "loss": 0.0072, "step": 12130 }, { "grad_norm": 0.23182570934295654, "learning_rate": 6.938086011453513e-05, "loss": 0.0071, "step": 12140 }, { "grad_norm": 0.23776176571846008, "learning_rate": 6.93300416081141e-05, "loss": 0.0083, "step": 12150 }, { "grad_norm": 0.2618558406829834, "learning_rate": 6.927919961386984e-05, "loss": 0.0085, "step": 12160 }, { "grad_norm": 0.28387129306793213, "learning_rate": 6.922833419358013e-05, "loss": 0.0076, "step": 12170 }, { "grad_norm": 0.257773220539093, "learning_rate": 6.917744540905124e-05, "loss": 0.0083, "step": 12180 }, { "grad_norm": 0.23209062218666077, "learning_rate": 6.912653332211787e-05, "loss": 0.0067, "step": 12190 }, { "grad_norm": 0.24082745611667633, "learning_rate": 6.9075597994643e-05, "loss": 0.0079, "step": 12200 }, { "grad_norm": 0.2725173532962799, "learning_rate": 6.902463948851787e-05, "loss": 0.0077, "step": 12210 }, { "grad_norm": 0.2495144158601761, "learning_rate": 6.897365786566184e-05, "loss": 0.0083, "step": 12220 }, { "grad_norm": 0.270396888256073, "learning_rate": 6.892265318802242e-05, "loss": 0.0083, "step": 12230 }, { "grad_norm": 0.3428340554237366, "learning_rate": 6.887162551757507e-05, "loss": 0.0075, "step": 12240 }, { "grad_norm": 0.28904852271080017, "learning_rate": 6.882057491632326e-05, "loss": 0.0075, "step": 12250 }, { "grad_norm": 0.23016634583473206, "learning_rate": 6.876950144629824e-05, "loss": 0.0079, "step": 12260 }, { "grad_norm": 0.24582698941230774, "learning_rate": 6.87184051695591e-05, "loss": 0.0071, "step": 12270 }, { "grad_norm": 0.2511565685272217, "learning_rate": 6.866728614819268e-05, "loss": 0.0081, "step": 12280 }, { "grad_norm": 0.21527400612831116, "learning_rate": 6.861614444431337e-05, "loss": 0.0079, "step": 12290 }, { "grad_norm": 0.21463048458099365, "learning_rate": 6.856498012006318e-05, "loss": 0.0067, "step": 12300 }, { "grad_norm": 0.24686744809150696, "learning_rate": 6.851379323761157e-05, "loss": 0.0076, "step": 12310 }, { "grad_norm": 0.2256661057472229, "learning_rate": 6.846258385915545e-05, "loss": 0.0071, "step": 12320 }, { "grad_norm": 0.2230372130870819, "learning_rate": 6.841135204691902e-05, "loss": 0.0074, "step": 12330 }, { "grad_norm": 0.2623993754386902, "learning_rate": 6.836009786315377e-05, "loss": 0.0066, "step": 12340 }, { "grad_norm": 0.28395700454711914, "learning_rate": 6.830882137013839e-05, "loss": 0.0067, "step": 12350 }, { "grad_norm": 0.3418019115924835, "learning_rate": 6.825752263017863e-05, "loss": 0.0081, "step": 12360 }, { "grad_norm": 0.2549242675304413, "learning_rate": 6.820620170560731e-05, "loss": 0.0073, "step": 12370 }, { "grad_norm": 0.2908019423484802, "learning_rate": 6.815485865878418e-05, "loss": 0.0075, "step": 12380 }, { "grad_norm": 0.1923939436674118, "learning_rate": 6.810349355209588e-05, "loss": 0.0082, "step": 12390 }, { "grad_norm": 0.28272947669029236, "learning_rate": 6.805210644795588e-05, "loss": 0.0077, "step": 12400 }, { "grad_norm": 0.2060726433992386, "learning_rate": 6.800069740880433e-05, "loss": 0.0083, "step": 12410 }, { "grad_norm": 0.22574064135551453, "learning_rate": 6.794926649710807e-05, "loss": 0.0072, "step": 12420 }, { "grad_norm": 0.2950822114944458, "learning_rate": 6.789781377536052e-05, "loss": 0.008, "step": 12430 }, { "grad_norm": 0.25280317664146423, "learning_rate": 6.784633930608158e-05, "loss": 0.0078, "step": 12440 }, { "grad_norm": 0.26652100682258606, "learning_rate": 6.779484315181759e-05, "loss": 0.0072, "step": 12450 }, { "grad_norm": 0.3308437764644623, "learning_rate": 6.774332537514122e-05, "loss": 0.0077, "step": 12460 }, { "grad_norm": 0.21044404804706573, "learning_rate": 6.769178603865143e-05, "loss": 0.0084, "step": 12470 }, { "grad_norm": 0.24346868693828583, "learning_rate": 6.764022520497337e-05, "loss": 0.0078, "step": 12480 }, { "grad_norm": 0.25575554370880127, "learning_rate": 6.758864293675833e-05, "loss": 0.007, "step": 12490 }, { "grad_norm": 0.26012754440307617, "learning_rate": 6.753703929668362e-05, "loss": 0.008, "step": 12500 }, { "grad_norm": 0.27795130014419556, "learning_rate": 6.748541434745253e-05, "loss": 0.0067, "step": 12510 }, { "grad_norm": 0.22350628674030304, "learning_rate": 6.743376815179424e-05, "loss": 0.0077, "step": 12520 }, { "grad_norm": 0.24714377522468567, "learning_rate": 6.738210077246376e-05, "loss": 0.0071, "step": 12530 }, { "grad_norm": 0.2536865472793579, "learning_rate": 6.733041227224181e-05, "loss": 0.0082, "step": 12540 }, { "grad_norm": 0.28324517607688904, "learning_rate": 6.72787027139348e-05, "loss": 0.0071, "step": 12550 }, { "grad_norm": 0.25217628479003906, "learning_rate": 6.72269721603747e-05, "loss": 0.0073, "step": 12560 }, { "grad_norm": 0.2609931528568268, "learning_rate": 6.717522067441904e-05, "loss": 0.0076, "step": 12570 }, { "grad_norm": 0.22231438755989075, "learning_rate": 6.712344831895074e-05, "loss": 0.0077, "step": 12580 }, { "grad_norm": 0.2796629071235657, "learning_rate": 6.707165515687811e-05, "loss": 0.007, "step": 12590 }, { "grad_norm": 0.20269301533699036, "learning_rate": 6.70198412511347e-05, "loss": 0.0067, "step": 12600 }, { "grad_norm": 0.2400275319814682, "learning_rate": 6.69680066646793e-05, "loss": 0.0077, "step": 12610 }, { "grad_norm": 0.2624037563800812, "learning_rate": 6.691615146049583e-05, "loss": 0.0069, "step": 12620 }, { "grad_norm": 0.2606382966041565, "learning_rate": 6.686427570159324e-05, "loss": 0.0073, "step": 12630 }, { "grad_norm": 0.23165848851203918, "learning_rate": 6.681237945100548e-05, "loss": 0.0075, "step": 12640 }, { "grad_norm": 0.24344083666801453, "learning_rate": 6.676046277179139e-05, "loss": 0.007, "step": 12650 }, { "grad_norm": 0.21247272193431854, "learning_rate": 6.670852572703462e-05, "loss": 0.0068, "step": 12660 }, { "grad_norm": 0.2795293629169464, "learning_rate": 6.665656837984359e-05, "loss": 0.0066, "step": 12670 }, { "grad_norm": 0.21867285668849945, "learning_rate": 6.660459079335135e-05, "loss": 0.0073, "step": 12680 }, { "grad_norm": 0.23383796215057373, "learning_rate": 6.655259303071558e-05, "loss": 0.0063, "step": 12690 }, { "grad_norm": 0.1923694759607315, "learning_rate": 6.650057515511848e-05, "loss": 0.0064, "step": 12700 }, { "grad_norm": 0.23878122866153717, "learning_rate": 6.644853722976667e-05, "loss": 0.0068, "step": 12710 }, { "grad_norm": 0.19844236969947815, "learning_rate": 6.639647931789113e-05, "loss": 0.0068, "step": 12720 }, { "grad_norm": 0.3030112385749817, "learning_rate": 6.634440148274713e-05, "loss": 0.0072, "step": 12730 }, { "grad_norm": 0.29436954855918884, "learning_rate": 6.629230378761415e-05, "loss": 0.0085, "step": 12740 }, { "grad_norm": 0.251598984003067, "learning_rate": 6.624018629579581e-05, "loss": 0.0078, "step": 12750 }, { "grad_norm": 0.22182928025722504, "learning_rate": 6.618804907061976e-05, "loss": 0.0069, "step": 12760 }, { "grad_norm": 0.2940364181995392, "learning_rate": 6.613589217543766e-05, "loss": 0.0076, "step": 12770 }, { "grad_norm": 0.20308679342269897, "learning_rate": 6.608371567362504e-05, "loss": 0.0077, "step": 12780 }, { "grad_norm": 0.3306617736816406, "learning_rate": 6.60315196285813e-05, "loss": 0.0069, "step": 12790 }, { "grad_norm": 0.27817803621292114, "learning_rate": 6.597930410372954e-05, "loss": 0.008, "step": 12800 }, { "grad_norm": 0.22359612584114075, "learning_rate": 6.592706916251653e-05, "loss": 0.0067, "step": 12810 }, { "grad_norm": 0.2547834813594818, "learning_rate": 6.587481486841267e-05, "loss": 0.0069, "step": 12820 }, { "grad_norm": 0.23497673869132996, "learning_rate": 6.582254128491184e-05, "loss": 0.006, "step": 12830 }, { "grad_norm": 0.2852087616920471, "learning_rate": 6.577024847553139e-05, "loss": 0.0073, "step": 12840 }, { "grad_norm": 0.3237002491950989, "learning_rate": 6.571793650381201e-05, "loss": 0.0077, "step": 12850 }, { "grad_norm": 0.2525921165943146, "learning_rate": 6.566560543331766e-05, "loss": 0.0078, "step": 12860 }, { "grad_norm": 0.29381003975868225, "learning_rate": 6.561325532763554e-05, "loss": 0.0074, "step": 12870 }, { "grad_norm": 0.2689189016819, "learning_rate": 6.556088625037597e-05, "loss": 0.0072, "step": 12880 }, { "grad_norm": 0.26632052659988403, "learning_rate": 6.55084982651723e-05, "loss": 0.0067, "step": 12890 }, { "grad_norm": 0.19162796437740326, "learning_rate": 6.54560914356809e-05, "loss": 0.0077, "step": 12900 }, { "grad_norm": 0.2577548027038574, "learning_rate": 6.540366582558097e-05, "loss": 0.0066, "step": 12910 }, { "grad_norm": 0.2573550045490265, "learning_rate": 6.53512214985746e-05, "loss": 0.0064, "step": 12920 }, { "grad_norm": 0.27676379680633545, "learning_rate": 6.529875851838658e-05, "loss": 0.0064, "step": 12930 }, { "grad_norm": 0.23883642256259918, "learning_rate": 6.52462769487644e-05, "loss": 0.007, "step": 12940 }, { "grad_norm": 0.2487030029296875, "learning_rate": 6.519377685347807e-05, "loss": 0.0074, "step": 12950 }, { "grad_norm": 0.2392069399356842, "learning_rate": 6.514125829632021e-05, "loss": 0.007, "step": 12960 }, { "grad_norm": 0.2278149276971817, "learning_rate": 6.508872134110577e-05, "loss": 0.0077, "step": 12970 }, { "grad_norm": 0.26694855093955994, "learning_rate": 6.503616605167213e-05, "loss": 0.0072, "step": 12980 }, { "grad_norm": 0.2206481695175171, "learning_rate": 6.498359249187893e-05, "loss": 0.0079, "step": 12990 }, { "grad_norm": 0.2661944627761841, "learning_rate": 6.493100072560799e-05, "loss": 0.0075, "step": 13000 }, { "grad_norm": 0.2793487012386322, "learning_rate": 6.487839081676327e-05, "loss": 0.0069, "step": 13010 }, { "grad_norm": 0.28858742117881775, "learning_rate": 6.482576282927076e-05, "loss": 0.0076, "step": 13020 }, { "grad_norm": 0.3059164583683014, "learning_rate": 6.477311682707845e-05, "loss": 0.0079, "step": 13030 }, { "grad_norm": 0.2655072510242462, "learning_rate": 6.472045287415616e-05, "loss": 0.0067, "step": 13040 }, { "grad_norm": 0.20142611861228943, "learning_rate": 6.466777103449558e-05, "loss": 0.006, "step": 13050 }, { "grad_norm": 0.24640636146068573, "learning_rate": 6.461507137211012e-05, "loss": 0.0069, "step": 13060 }, { "grad_norm": 0.26599934697151184, "learning_rate": 6.456235395103483e-05, "loss": 0.0062, "step": 13070 }, { "grad_norm": 0.2124205380678177, "learning_rate": 6.450961883532634e-05, "loss": 0.0066, "step": 13080 }, { "grad_norm": 0.2221207469701767, "learning_rate": 6.445686608906283e-05, "loss": 0.0064, "step": 13090 }, { "grad_norm": 0.22872713208198547, "learning_rate": 6.44040957763438e-05, "loss": 0.0068, "step": 13100 }, { "grad_norm": 0.24420705437660217, "learning_rate": 6.435130796129018e-05, "loss": 0.0062, "step": 13110 }, { "grad_norm": 0.21613144874572754, "learning_rate": 6.429850270804416e-05, "loss": 0.0072, "step": 13120 }, { "grad_norm": 0.22340647876262665, "learning_rate": 6.424568008076909e-05, "loss": 0.0084, "step": 13130 }, { "grad_norm": 0.26508018374443054, "learning_rate": 6.419284014364943e-05, "loss": 0.007, "step": 13140 }, { "grad_norm": 0.2540311813354492, "learning_rate": 6.41399829608907e-05, "loss": 0.0063, "step": 13150 }, { "grad_norm": 0.24151162803173065, "learning_rate": 6.408710859671938e-05, "loss": 0.0085, "step": 13160 }, { "grad_norm": 0.23258985579013824, "learning_rate": 6.403421711538278e-05, "loss": 0.008, "step": 13170 }, { "grad_norm": 0.21937763690948486, "learning_rate": 6.398130858114903e-05, "loss": 0.0061, "step": 13180 }, { "grad_norm": 0.2620888948440552, "learning_rate": 6.392838305830701e-05, "loss": 0.0067, "step": 13190 }, { "grad_norm": 0.25849398970603943, "learning_rate": 6.387544061116621e-05, "loss": 0.0083, "step": 13200 }, { "grad_norm": 0.19696420431137085, "learning_rate": 6.382248130405671e-05, "loss": 0.0067, "step": 13210 }, { "grad_norm": 0.20594185590744019, "learning_rate": 6.376950520132905e-05, "loss": 0.0073, "step": 13220 }, { "grad_norm": 0.2561691701412201, "learning_rate": 6.371651236735418e-05, "loss": 0.0062, "step": 13230 }, { "grad_norm": 0.2856631875038147, "learning_rate": 6.366350286652341e-05, "loss": 0.0074, "step": 13240 }, { "grad_norm": 0.22115884721279144, "learning_rate": 6.361047676324827e-05, "loss": 0.0072, "step": 13250 }, { "grad_norm": 0.21611319482326508, "learning_rate": 6.355743412196047e-05, "loss": 0.007, "step": 13260 }, { "grad_norm": 0.25783073902130127, "learning_rate": 6.350437500711184e-05, "loss": 0.008, "step": 13270 }, { "grad_norm": 0.2509796619415283, "learning_rate": 6.345129948317418e-05, "loss": 0.0066, "step": 13280 }, { "grad_norm": 0.20480823516845703, "learning_rate": 6.33982076146393e-05, "loss": 0.0078, "step": 13290 }, { "grad_norm": 0.22611509263515472, "learning_rate": 6.334509946601879e-05, "loss": 0.0077, "step": 13300 }, { "grad_norm": 0.2350149154663086, "learning_rate": 6.329197510184406e-05, "loss": 0.0073, "step": 13310 }, { "grad_norm": 0.1989920735359192, "learning_rate": 6.323883458666624e-05, "loss": 0.0075, "step": 13320 }, { "grad_norm": 0.24860800802707672, "learning_rate": 6.318567798505605e-05, "loss": 0.0069, "step": 13330 }, { "grad_norm": 0.14864976704120636, "learning_rate": 6.313250536160378e-05, "loss": 0.0067, "step": 13340 }, { "grad_norm": 0.2616286873817444, "learning_rate": 6.307931678091917e-05, "loss": 0.0065, "step": 13350 }, { "grad_norm": 0.27229225635528564, "learning_rate": 6.302611230763138e-05, "loss": 0.0071, "step": 13360 }, { "grad_norm": 0.25116783380508423, "learning_rate": 6.297289200638887e-05, "loss": 0.0083, "step": 13370 }, { "grad_norm": 0.24302475154399872, "learning_rate": 6.29196559418593e-05, "loss": 0.0068, "step": 13380 }, { "grad_norm": 0.18964216113090515, "learning_rate": 6.286640417872951e-05, "loss": 0.0072, "step": 13390 }, { "grad_norm": 0.21581992506980896, "learning_rate": 6.281313678170542e-05, "loss": 0.0083, "step": 13400 }, { "grad_norm": 0.270751416683197, "learning_rate": 6.275985381551194e-05, "loss": 0.0066, "step": 13410 }, { "grad_norm": 0.24866518378257751, "learning_rate": 6.270655534489292e-05, "loss": 0.0065, "step": 13420 }, { "grad_norm": 0.2527863383293152, "learning_rate": 6.265324143461098e-05, "loss": 0.0071, "step": 13430 }, { "grad_norm": 0.20228345692157745, "learning_rate": 6.259991214944758e-05, "loss": 0.0065, "step": 13440 }, { "grad_norm": 0.23330335319042206, "learning_rate": 6.254656755420283e-05, "loss": 0.0064, "step": 13450 }, { "grad_norm": 0.23438553512096405, "learning_rate": 6.249320771369545e-05, "loss": 0.0067, "step": 13460 }, { "grad_norm": 0.20001573860645294, "learning_rate": 6.243983269276263e-05, "loss": 0.0068, "step": 13470 }, { "grad_norm": 0.23964233696460724, "learning_rate": 6.238644255626012e-05, "loss": 0.0061, "step": 13480 }, { "grad_norm": 0.21075086295604706, "learning_rate": 6.233303736906193e-05, "loss": 0.0061, "step": 13490 }, { "grad_norm": 0.2517348527908325, "learning_rate": 6.227961719606039e-05, "loss": 0.0075, "step": 13500 }, { "grad_norm": 0.19695648550987244, "learning_rate": 6.222618210216608e-05, "loss": 0.0066, "step": 13510 }, { "grad_norm": 0.2171541452407837, "learning_rate": 6.217273215230767e-05, "loss": 0.008, "step": 13520 }, { "grad_norm": 0.23161739110946655, "learning_rate": 6.211926741143188e-05, "loss": 0.0065, "step": 13530 }, { "grad_norm": 0.2714325487613678, "learning_rate": 6.20657879445034e-05, "loss": 0.0075, "step": 13540 }, { "grad_norm": 0.30451586842536926, "learning_rate": 6.201229381650485e-05, "loss": 0.0064, "step": 13550 }, { "grad_norm": 0.2556527256965637, "learning_rate": 6.195878509243661e-05, "loss": 0.0061, "step": 13560 }, { "grad_norm": 0.253057599067688, "learning_rate": 6.190526183731685e-05, "loss": 0.007, "step": 13570 }, { "grad_norm": 0.19131217896938324, "learning_rate": 6.185172411618138e-05, "loss": 0.0061, "step": 13580 }, { "grad_norm": 0.30017247796058655, "learning_rate": 6.179817199408355e-05, "loss": 0.0071, "step": 13590 }, { "grad_norm": 0.21754956245422363, "learning_rate": 6.174460553609426e-05, "loss": 0.0065, "step": 13600 }, { "grad_norm": 0.2323118895292282, "learning_rate": 6.16910248073018e-05, "loss": 0.0072, "step": 13610 }, { "grad_norm": 0.273316890001297, "learning_rate": 6.16374298728118e-05, "loss": 0.0061, "step": 13620 }, { "grad_norm": 0.21758437156677246, "learning_rate": 6.158382079774716e-05, "loss": 0.0072, "step": 13630 }, { "grad_norm": 0.21978215873241425, "learning_rate": 6.153019764724799e-05, "loss": 0.0063, "step": 13640 }, { "grad_norm": 0.27762216329574585, "learning_rate": 6.147656048647144e-05, "loss": 0.0071, "step": 13650 }, { "grad_norm": 0.26004552841186523, "learning_rate": 6.142290938059173e-05, "loss": 0.0067, "step": 13660 }, { "grad_norm": 0.26268357038497925, "learning_rate": 6.136924439480001e-05, "loss": 0.0071, "step": 13670 }, { "grad_norm": 0.2489512413740158, "learning_rate": 6.13155655943043e-05, "loss": 0.0061, "step": 13680 }, { "grad_norm": 0.19315683841705322, "learning_rate": 6.126187304432941e-05, "loss": 0.0056, "step": 13690 }, { "grad_norm": 0.19838185608386993, "learning_rate": 6.120816681011684e-05, "loss": 0.0058, "step": 13700 }, { "grad_norm": 0.22212015092372894, "learning_rate": 6.115444695692474e-05, "loss": 0.0057, "step": 13710 }, { "grad_norm": 0.27834391593933105, "learning_rate": 6.110071355002779e-05, "loss": 0.007, "step": 13720 }, { "grad_norm": 0.21555571258068085, "learning_rate": 6.104696665471714e-05, "loss": 0.0073, "step": 13730 }, { "grad_norm": 0.25469234585762024, "learning_rate": 6.099320633630036e-05, "loss": 0.0062, "step": 13740 }, { "grad_norm": 0.24184435606002808, "learning_rate": 6.093943266010128e-05, "loss": 0.0062, "step": 13750 }, { "grad_norm": 0.20043589174747467, "learning_rate": 6.088564569146e-05, "loss": 0.0066, "step": 13760 }, { "grad_norm": 0.19360867142677307, "learning_rate": 6.083184549573274e-05, "loss": 0.006, "step": 13770 }, { "grad_norm": 0.257904052734375, "learning_rate": 6.077803213829184e-05, "loss": 0.0068, "step": 13780 }, { "grad_norm": 0.228912353515625, "learning_rate": 6.0724205684525594e-05, "loss": 0.0061, "step": 13790 }, { "grad_norm": 0.19492770731449127, "learning_rate": 6.0670366199838215e-05, "loss": 0.0068, "step": 13800 }, { "grad_norm": 0.2720438539981842, "learning_rate": 6.061651374964974e-05, "loss": 0.0068, "step": 13810 }, { "grad_norm": 0.26898762583732605, "learning_rate": 6.0562648399396004e-05, "loss": 0.0067, "step": 13820 }, { "grad_norm": 0.24783796072006226, "learning_rate": 6.050877021452845e-05, "loss": 0.0064, "step": 13830 }, { "grad_norm": 0.22201155126094818, "learning_rate": 6.04548792605142e-05, "loss": 0.0061, "step": 13840 }, { "grad_norm": 0.1546681523323059, "learning_rate": 6.04009756028358e-05, "loss": 0.0068, "step": 13850 }, { "grad_norm": 0.25796008110046387, "learning_rate": 6.0347059306991294e-05, "loss": 0.0083, "step": 13860 }, { "grad_norm": 0.2468590885400772, "learning_rate": 6.029313043849407e-05, "loss": 0.0066, "step": 13870 }, { "grad_norm": 0.29722049832344055, "learning_rate": 6.023918906287279e-05, "loss": 0.0069, "step": 13880 }, { "grad_norm": 0.23768748342990875, "learning_rate": 6.01852352456713e-05, "loss": 0.0067, "step": 13890 }, { "grad_norm": 0.23087532818317413, "learning_rate": 6.013126905244858e-05, "loss": 0.0074, "step": 13900 }, { "grad_norm": 0.2194361388683319, "learning_rate": 6.007729054877864e-05, "loss": 0.0067, "step": 13910 }, { "grad_norm": 0.25105589628219604, "learning_rate": 6.0023299800250466e-05, "loss": 0.0062, "step": 13920 }, { "grad_norm": 0.2088538259267807, "learning_rate": 5.9969296872467894e-05, "loss": 0.0058, "step": 13930 }, { "grad_norm": 0.19388312101364136, "learning_rate": 5.9915281831049585e-05, "loss": 0.0074, "step": 13940 }, { "grad_norm": 0.25578510761260986, "learning_rate": 5.98612547416289e-05, "loss": 0.0073, "step": 13950 }, { "grad_norm": 0.2675463557243347, "learning_rate": 5.9807215669853855e-05, "loss": 0.0084, "step": 13960 }, { "grad_norm": 0.23419392108917236, "learning_rate": 5.9753164681387e-05, "loss": 0.0062, "step": 13970 }, { "grad_norm": 0.21404705941677094, "learning_rate": 5.969910184190539e-05, "loss": 0.0062, "step": 13980 }, { "grad_norm": 0.20408913493156433, "learning_rate": 5.9645027217100477e-05, "loss": 0.0056, "step": 13990 }, { "grad_norm": 0.28686222434043884, "learning_rate": 5.959094087267804e-05, "loss": 0.0076, "step": 14000 }, { "grad_norm": 0.20522059500217438, "learning_rate": 5.9536842874358065e-05, "loss": 0.0072, "step": 14010 }, { "grad_norm": 0.2946614921092987, "learning_rate": 5.9482733287874734e-05, "loss": 0.0068, "step": 14020 }, { "grad_norm": 0.3324415981769562, "learning_rate": 5.942861217897631e-05, "loss": 0.0074, "step": 14030 }, { "grad_norm": 0.27357664704322815, "learning_rate": 5.9374479613425e-05, "loss": 0.0071, "step": 14040 }, { "grad_norm": 0.21822060644626617, "learning_rate": 5.932033565699704e-05, "loss": 0.0066, "step": 14050 }, { "grad_norm": 0.22324790060520172, "learning_rate": 5.926618037548237e-05, "loss": 0.0072, "step": 14060 }, { "grad_norm": 0.22239309549331665, "learning_rate": 5.9212013834684824e-05, "loss": 0.0076, "step": 14070 }, { "grad_norm": 0.20068594813346863, "learning_rate": 5.9157836100421824e-05, "loss": 0.007, "step": 14080 }, { "grad_norm": 0.18320992588996887, "learning_rate": 5.910364723852444e-05, "loss": 0.0069, "step": 14090 }, { "grad_norm": 0.2011718899011612, "learning_rate": 5.904944731483723e-05, "loss": 0.0057, "step": 14100 }, { "grad_norm": 0.2642921805381775, "learning_rate": 5.899523639521825e-05, "loss": 0.006, "step": 14110 }, { "grad_norm": 0.2612369954586029, "learning_rate": 5.894101454553883e-05, "loss": 0.0063, "step": 14120 }, { "grad_norm": 0.2539650499820709, "learning_rate": 5.888678183168368e-05, "loss": 0.0078, "step": 14130 }, { "grad_norm": 0.24627746641635895, "learning_rate": 5.883253831955061e-05, "loss": 0.0066, "step": 14140 }, { "grad_norm": 0.2440510392189026, "learning_rate": 5.8778284075050625e-05, "loss": 0.0058, "step": 14150 }, { "grad_norm": 0.20609524846076965, "learning_rate": 5.872401916410777e-05, "loss": 0.0064, "step": 14160 }, { "grad_norm": 0.265025794506073, "learning_rate": 5.8669743652659014e-05, "loss": 0.0062, "step": 14170 }, { "grad_norm": 0.1870139241218567, "learning_rate": 5.86154576066542e-05, "loss": 0.0062, "step": 14180 }, { "grad_norm": 0.25761324167251587, "learning_rate": 5.856116109205602e-05, "loss": 0.0062, "step": 14190 }, { "grad_norm": 0.20956628024578094, "learning_rate": 5.850685417483983e-05, "loss": 0.0063, "step": 14200 }, { "grad_norm": 0.20504550635814667, "learning_rate": 5.8452536920993684e-05, "loss": 0.0056, "step": 14210 }, { "grad_norm": 0.20799843966960907, "learning_rate": 5.839820939651817e-05, "loss": 0.0054, "step": 14220 }, { "grad_norm": 0.1858857274055481, "learning_rate": 5.834387166742632e-05, "loss": 0.0056, "step": 14230 }, { "grad_norm": 0.19511808454990387, "learning_rate": 5.828952379974364e-05, "loss": 0.0066, "step": 14240 }, { "grad_norm": 0.2534532845020294, "learning_rate": 5.8235165859507864e-05, "loss": 0.0064, "step": 14250 }, { "grad_norm": 0.2107435017824173, "learning_rate": 5.818079791276907e-05, "loss": 0.0057, "step": 14260 }, { "grad_norm": 0.23028282821178436, "learning_rate": 5.8126420025589415e-05, "loss": 0.006, "step": 14270 }, { "grad_norm": 0.1975833773612976, "learning_rate": 5.807203226404313e-05, "loss": 0.0057, "step": 14280 }, { "grad_norm": 0.2542743682861328, "learning_rate": 5.801763469421652e-05, "loss": 0.0078, "step": 14290 }, { "grad_norm": 0.23853014409542084, "learning_rate": 5.796322738220774e-05, "loss": 0.007, "step": 14300 }, { "grad_norm": 0.2695060670375824, "learning_rate": 5.79088103941268e-05, "loss": 0.0063, "step": 14310 }, { "grad_norm": 0.23394541442394257, "learning_rate": 5.785438379609549e-05, "loss": 0.0063, "step": 14320 }, { "grad_norm": 0.19033122062683105, "learning_rate": 5.779994765424724e-05, "loss": 0.0056, "step": 14330 }, { "grad_norm": 0.19316981732845306, "learning_rate": 5.7745502034727114e-05, "loss": 0.0061, "step": 14340 }, { "grad_norm": 0.24229130148887634, "learning_rate": 5.769104700369164e-05, "loss": 0.0063, "step": 14350 }, { "grad_norm": 0.23179645836353302, "learning_rate": 5.7636582627308854e-05, "loss": 0.0059, "step": 14360 }, { "grad_norm": 0.24487490952014923, "learning_rate": 5.7582108971758095e-05, "loss": 0.0061, "step": 14370 }, { "grad_norm": 0.2611483633518219, "learning_rate": 5.752762610322999e-05, "loss": 0.0075, "step": 14380 }, { "grad_norm": 0.247283935546875, "learning_rate": 5.747313408792636e-05, "loss": 0.0063, "step": 14390 }, { "grad_norm": 0.2301962524652481, "learning_rate": 5.741863299206014e-05, "loss": 0.0058, "step": 14400 }, { "grad_norm": 0.25670161843299866, "learning_rate": 5.73641228818553e-05, "loss": 0.0072, "step": 14410 }, { "grad_norm": 0.20070618391036987, "learning_rate": 5.730960382354676e-05, "loss": 0.0061, "step": 14420 }, { "grad_norm": 0.2672232687473297, "learning_rate": 5.725507588338035e-05, "loss": 0.0066, "step": 14430 }, { "grad_norm": 0.3072107136249542, "learning_rate": 5.7200539127612604e-05, "loss": 0.0076, "step": 14440 }, { "grad_norm": 0.2517894506454468, "learning_rate": 5.714599362251087e-05, "loss": 0.0061, "step": 14450 }, { "grad_norm": 0.13219939172267914, "learning_rate": 5.709143943435307e-05, "loss": 0.007, "step": 14460 }, { "grad_norm": 0.19510792195796967, "learning_rate": 5.7036876629427646e-05, "loss": 0.006, "step": 14470 }, { "grad_norm": 0.22808505594730377, "learning_rate": 5.698230527403361e-05, "loss": 0.0073, "step": 14480 }, { "grad_norm": 0.22400452196598053, "learning_rate": 5.6927725434480274e-05, "loss": 0.006, "step": 14490 }, { "grad_norm": 0.23427478969097137, "learning_rate": 5.687313717708728e-05, "loss": 0.0061, "step": 14500 }, { "grad_norm": 0.17823611199855804, "learning_rate": 5.681854056818453e-05, "loss": 0.0062, "step": 14510 }, { "grad_norm": 0.21296104788780212, "learning_rate": 5.676393567411205e-05, "loss": 0.0062, "step": 14520 }, { "grad_norm": 0.2652239501476288, "learning_rate": 5.670932256121991e-05, "loss": 0.007, "step": 14530 }, { "grad_norm": 0.2570689022541046, "learning_rate": 5.6654701295868216e-05, "loss": 0.0068, "step": 14540 }, { "grad_norm": 0.22959604859352112, "learning_rate": 5.660007194442697e-05, "loss": 0.0057, "step": 14550 }, { "grad_norm": 0.2390802949666977, "learning_rate": 5.6545434573275946e-05, "loss": 0.0063, "step": 14560 }, { "grad_norm": 0.23341608047485352, "learning_rate": 5.649078924880472e-05, "loss": 0.0066, "step": 14570 }, { "grad_norm": 0.213889017701149, "learning_rate": 5.643613603741252e-05, "loss": 0.0072, "step": 14580 }, { "grad_norm": 0.29240983724594116, "learning_rate": 5.6381475005508154e-05, "loss": 0.0067, "step": 14590 }, { "grad_norm": 0.2720557451248169, "learning_rate": 5.632680621950992e-05, "loss": 0.0058, "step": 14600 }, { "grad_norm": 0.21651709079742432, "learning_rate": 5.627212974584555e-05, "loss": 0.0067, "step": 14610 }, { "grad_norm": 0.2748400568962097, "learning_rate": 5.62174456509521e-05, "loss": 0.0063, "step": 14620 }, { "grad_norm": 0.22031934559345245, "learning_rate": 5.616275400127594e-05, "loss": 0.0065, "step": 14630 }, { "grad_norm": 0.2475021928548813, "learning_rate": 5.610805486327254e-05, "loss": 0.0068, "step": 14640 }, { "grad_norm": 0.2812642753124237, "learning_rate": 5.6053348303406536e-05, "loss": 0.0068, "step": 14650 }, { "grad_norm": 0.2446913719177246, "learning_rate": 5.599863438815156e-05, "loss": 0.0065, "step": 14660 }, { "grad_norm": 0.24713528156280518, "learning_rate": 5.594391318399017e-05, "loss": 0.0064, "step": 14670 }, { "grad_norm": 0.223225399851799, "learning_rate": 5.588918475741377e-05, "loss": 0.0068, "step": 14680 }, { "grad_norm": 0.21757501363754272, "learning_rate": 5.5834449174922586e-05, "loss": 0.0056, "step": 14690 }, { "grad_norm": 0.2678114175796509, "learning_rate": 5.57797065030255e-05, "loss": 0.0059, "step": 14700 }, { "grad_norm": 0.26760223507881165, "learning_rate": 5.5724956808240016e-05, "loss": 0.0061, "step": 14710 }, { "grad_norm": 0.1902293711900711, "learning_rate": 5.567020015709219e-05, "loss": 0.0064, "step": 14720 }, { "grad_norm": 0.2306230366230011, "learning_rate": 5.561543661611649e-05, "loss": 0.007, "step": 14730 }, { "grad_norm": 0.2748568058013916, "learning_rate": 5.556066625185583e-05, "loss": 0.007, "step": 14740 }, { "grad_norm": 0.17141860723495483, "learning_rate": 5.550588913086131e-05, "loss": 0.0072, "step": 14750 }, { "grad_norm": 0.22994983196258545, "learning_rate": 5.545110531969234e-05, "loss": 0.0064, "step": 14760 }, { "grad_norm": 0.24939128756523132, "learning_rate": 5.539631488491641e-05, "loss": 0.0059, "step": 14770 }, { "grad_norm": 0.22121156752109528, "learning_rate": 5.534151789310904e-05, "loss": 0.006, "step": 14780 }, { "grad_norm": 0.3021535277366638, "learning_rate": 5.528671441085376e-05, "loss": 0.0068, "step": 14790 }, { "grad_norm": 0.21206553280353546, "learning_rate": 5.523190450474197e-05, "loss": 0.0056, "step": 14800 }, { "grad_norm": 0.20038598775863647, "learning_rate": 5.5177088241372865e-05, "loss": 0.0055, "step": 14810 }, { "grad_norm": 0.24848155677318573, "learning_rate": 5.512226568735338e-05, "loss": 0.0074, "step": 14820 }, { "grad_norm": 0.20866146683692932, "learning_rate": 5.506743690929809e-05, "loss": 0.0055, "step": 14830 }, { "grad_norm": 0.207877978682518, "learning_rate": 5.501260197382912e-05, "loss": 0.0054, "step": 14840 }, { "grad_norm": 0.20717130601406097, "learning_rate": 5.49577609475761e-05, "loss": 0.0067, "step": 14850 }, { "grad_norm": 0.24062608182430267, "learning_rate": 5.490291389717603e-05, "loss": 0.0059, "step": 14860 }, { "grad_norm": 0.21328893303871155, "learning_rate": 5.484806088927329e-05, "loss": 0.0053, "step": 14870 }, { "grad_norm": 0.2662334442138672, "learning_rate": 5.479320199051942e-05, "loss": 0.0062, "step": 14880 }, { "grad_norm": 0.1990499496459961, "learning_rate": 5.473833726757314e-05, "loss": 0.0064, "step": 14890 }, { "grad_norm": 0.19696538150310516, "learning_rate": 5.4683466787100304e-05, "loss": 0.0058, "step": 14900 }, { "grad_norm": 0.1558971107006073, "learning_rate": 5.462859061577369e-05, "loss": 0.0064, "step": 14910 }, { "grad_norm": 0.19675497710704803, "learning_rate": 5.4573708820273026e-05, "loss": 0.0059, "step": 14920 }, { "grad_norm": 0.20575101673603058, "learning_rate": 5.451882146728488e-05, "loss": 0.0066, "step": 14930 }, { "grad_norm": 0.1856343150138855, "learning_rate": 5.446392862350255e-05, "loss": 0.0058, "step": 14940 }, { "grad_norm": 0.22339941561222076, "learning_rate": 5.440903035562603e-05, "loss": 0.0061, "step": 14950 }, { "grad_norm": 0.19470913708209991, "learning_rate": 5.435412673036188e-05, "loss": 0.0062, "step": 14960 }, { "grad_norm": 0.21660403907299042, "learning_rate": 5.429921781442318e-05, "loss": 0.0064, "step": 14970 }, { "grad_norm": 0.1847912073135376, "learning_rate": 5.424430367452945e-05, "loss": 0.006, "step": 14980 }, { "grad_norm": 0.19388660788536072, "learning_rate": 5.418938437740655e-05, "loss": 0.006, "step": 14990 }, { "grad_norm": 0.17676885426044464, "learning_rate": 5.413445998978658e-05, "loss": 0.0051, "step": 15000 } ], "logging_steps": 10, "max_steps": 30000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 7500, "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": 128, "trial_name": null, "trial_params": null }