{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 80000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.5073959231376648, "learning_rate": 2.25e-07, "loss": 1.1644, "step": 10 }, { "grad_norm": 0.7803415060043335, "learning_rate": 4.75e-07, "loss": 1.1669, "step": 20 }, { "grad_norm": 0.5123212933540344, "learning_rate": 7.25e-07, "loss": 1.1606, "step": 30 }, { "grad_norm": 0.7255907654762268, "learning_rate": 9.75e-07, "loss": 1.1548, "step": 40 }, { "grad_norm": 0.5409747958183289, "learning_rate": 1.2250000000000001e-06, "loss": 1.1439, "step": 50 }, { "grad_norm": 0.7347477674484253, "learning_rate": 1.475e-06, "loss": 1.1311, "step": 60 }, { "grad_norm": 0.9941715598106384, "learning_rate": 1.7250000000000002e-06, "loss": 1.1048, "step": 70 }, { "grad_norm": 1.030922770500183, "learning_rate": 1.975e-06, "loss": 1.0868, "step": 80 }, { "grad_norm": 1.0312676429748535, "learning_rate": 2.225e-06, "loss": 1.1055, "step": 90 }, { "grad_norm": 0.7102149128913879, "learning_rate": 2.4750000000000004e-06, "loss": 1.1082, "step": 100 }, { "grad_norm": 0.8581516742706299, "learning_rate": 2.725e-06, "loss": 1.0982, "step": 110 }, { "grad_norm": 0.7224396467208862, "learning_rate": 2.975e-06, "loss": 1.1086, "step": 120 }, { "grad_norm": 1.6907237768173218, "learning_rate": 3.225e-06, "loss": 1.0943, "step": 130 }, { "grad_norm": 0.949960470199585, "learning_rate": 3.4750000000000006e-06, "loss": 1.1082, "step": 140 }, { "grad_norm": 0.8837343454360962, "learning_rate": 3.725e-06, "loss": 1.0896, "step": 150 }, { "grad_norm": 1.0260722637176514, "learning_rate": 3.975e-06, "loss": 1.1136, "step": 160 }, { "grad_norm": 0.9111544489860535, "learning_rate": 4.225e-06, "loss": 1.0758, "step": 170 }, { "grad_norm": 1.1038429737091064, "learning_rate": 4.475e-06, "loss": 1.067, "step": 180 }, { "grad_norm": 0.941226601600647, "learning_rate": 4.7250000000000005e-06, "loss": 1.0838, "step": 190 }, { "grad_norm": 0.8749931454658508, "learning_rate": 4.975000000000001e-06, "loss": 1.069, "step": 200 }, { "grad_norm": 0.8463923931121826, "learning_rate": 5.225e-06, "loss": 1.0536, "step": 210 }, { "grad_norm": 1.1122890710830688, "learning_rate": 5.475e-06, "loss": 1.0708, "step": 220 }, { "grad_norm": 1.1898643970489502, "learning_rate": 5.725e-06, "loss": 1.0771, "step": 230 }, { "grad_norm": 0.9497122168540955, "learning_rate": 5.975e-06, "loss": 1.0747, "step": 240 }, { "grad_norm": 0.9883389472961426, "learning_rate": 6.2250000000000005e-06, "loss": 1.0806, "step": 250 }, { "grad_norm": 1.2806713581085205, "learning_rate": 6.475000000000001e-06, "loss": 1.092, "step": 260 }, { "grad_norm": 1.154653787612915, "learning_rate": 6.725000000000001e-06, "loss": 1.0527, "step": 270 }, { "grad_norm": 0.9450279474258423, "learning_rate": 6.975000000000001e-06, "loss": 1.0615, "step": 280 }, { "grad_norm": 0.9994856119155884, "learning_rate": 7.2249999999999994e-06, "loss": 1.0572, "step": 290 }, { "grad_norm": 1.0168260335922241, "learning_rate": 7.4750000000000004e-06, "loss": 1.0374, "step": 300 }, { "grad_norm": 1.120961308479309, "learning_rate": 7.725e-06, "loss": 1.0541, "step": 310 }, { "grad_norm": 1.6963564157485962, "learning_rate": 7.975e-06, "loss": 1.0524, "step": 320 }, { "grad_norm": 1.204584002494812, "learning_rate": 8.225e-06, "loss": 1.0393, "step": 330 }, { "grad_norm": 1.9084389209747314, "learning_rate": 8.475000000000001e-06, "loss": 1.0239, "step": 340 }, { "grad_norm": 1.2236602306365967, "learning_rate": 8.725e-06, "loss": 1.0186, "step": 350 }, { "grad_norm": 1.9821852445602417, "learning_rate": 8.975e-06, "loss": 0.976, "step": 360 }, { "grad_norm": 1.6828573942184448, "learning_rate": 9.225e-06, "loss": 0.9655, "step": 370 }, { "grad_norm": 2.0583863258361816, "learning_rate": 9.475e-06, "loss": 0.9307, "step": 380 }, { "grad_norm": 1.91310453414917, "learning_rate": 9.725000000000001e-06, "loss": 0.9079, "step": 390 }, { "grad_norm": 2.340939998626709, "learning_rate": 9.975e-06, "loss": 0.8755, "step": 400 }, { "grad_norm": 2.187577962875366, "learning_rate": 1.0225e-05, "loss": 0.8267, "step": 410 }, { "grad_norm": 2.2650794982910156, "learning_rate": 1.0475e-05, "loss": 0.8105, "step": 420 }, { "grad_norm": 2.461749315261841, "learning_rate": 1.0725e-05, "loss": 0.7679, "step": 430 }, { "grad_norm": 2.3694913387298584, "learning_rate": 1.0975e-05, "loss": 0.7381, "step": 440 }, { "grad_norm": 2.450118064880371, "learning_rate": 1.1225e-05, "loss": 0.6994, "step": 450 }, { "grad_norm": 3.0052988529205322, "learning_rate": 1.1475000000000001e-05, "loss": 0.6666, "step": 460 }, { "grad_norm": 3.196892499923706, "learning_rate": 1.1725e-05, "loss": 0.6598, "step": 470 }, { "grad_norm": 2.691117525100708, "learning_rate": 1.1975e-05, "loss": 0.6414, "step": 480 }, { "grad_norm": 2.3986222743988037, "learning_rate": 1.2225e-05, "loss": 0.6289, "step": 490 }, { "grad_norm": 3.2106850147247314, "learning_rate": 1.2475e-05, "loss": 0.5924, "step": 500 }, { "grad_norm": 2.7443888187408447, "learning_rate": 1.2725000000000001e-05, "loss": 0.5846, "step": 510 }, { "grad_norm": 2.702709674835205, "learning_rate": 1.2975e-05, "loss": 0.5609, "step": 520 }, { "grad_norm": 2.7888636589050293, "learning_rate": 1.3225000000000001e-05, "loss": 0.5578, "step": 530 }, { "grad_norm": 2.2206408977508545, "learning_rate": 1.3475000000000002e-05, "loss": 0.5005, "step": 540 }, { "grad_norm": 3.0677924156188965, "learning_rate": 1.3725000000000002e-05, "loss": 0.5133, "step": 550 }, { "grad_norm": 3.49368953704834, "learning_rate": 1.3975000000000003e-05, "loss": 0.498, "step": 560 }, { "grad_norm": 3.335202217102051, "learning_rate": 1.4225e-05, "loss": 0.4418, "step": 570 }, { "grad_norm": 2.4725992679595947, "learning_rate": 1.4475e-05, "loss": 0.4169, "step": 580 }, { "grad_norm": 2.685983657836914, "learning_rate": 1.4725e-05, "loss": 0.3992, "step": 590 }, { "grad_norm": 2.2743890285491943, "learning_rate": 1.4975e-05, "loss": 0.3921, "step": 600 }, { "grad_norm": 2.9696595668792725, "learning_rate": 1.5225e-05, "loss": 0.3512, "step": 610 }, { "grad_norm": 3.002302646636963, "learning_rate": 1.5475e-05, "loss": 0.3432, "step": 620 }, { "grad_norm": 2.3584959506988525, "learning_rate": 1.5725e-05, "loss": 0.3347, "step": 630 }, { "grad_norm": 3.8261184692382812, "learning_rate": 1.5975000000000002e-05, "loss": 0.2967, "step": 640 }, { "grad_norm": 3.3849244117736816, "learning_rate": 1.6225e-05, "loss": 0.2835, "step": 650 }, { "grad_norm": 3.039376974105835, "learning_rate": 1.6475e-05, "loss": 0.2857, "step": 660 }, { "grad_norm": 2.9371492862701416, "learning_rate": 1.6725000000000003e-05, "loss": 0.2454, "step": 670 }, { "grad_norm": 2.852123737335205, "learning_rate": 1.6975000000000003e-05, "loss": 0.2586, "step": 680 }, { "grad_norm": 2.8384900093078613, "learning_rate": 1.7225e-05, "loss": 0.2373, "step": 690 }, { "grad_norm": 3.4945240020751953, "learning_rate": 1.7475e-05, "loss": 0.255, "step": 700 }, { "grad_norm": 3.1713826656341553, "learning_rate": 1.7725e-05, "loss": 0.2082, "step": 710 }, { "grad_norm": 3.5380451679229736, "learning_rate": 1.7975e-05, "loss": 0.2109, "step": 720 }, { "grad_norm": 2.5445046424865723, "learning_rate": 1.8225e-05, "loss": 0.1959, "step": 730 }, { "grad_norm": 2.1349852085113525, "learning_rate": 1.8475000000000002e-05, "loss": 0.1606, "step": 740 }, { "grad_norm": 2.6936252117156982, "learning_rate": 1.8725e-05, "loss": 0.1915, "step": 750 }, { "grad_norm": 1.9734646081924438, "learning_rate": 1.8975e-05, "loss": 0.1791, "step": 760 }, { "grad_norm": 2.524522542953491, "learning_rate": 1.9225e-05, "loss": 0.1992, "step": 770 }, { "grad_norm": 2.409587860107422, "learning_rate": 1.9475000000000002e-05, "loss": 0.2303, "step": 780 }, { "grad_norm": 1.848709225654602, "learning_rate": 1.9725000000000002e-05, "loss": 0.1719, "step": 790 }, { "grad_norm": 2.4158613681793213, "learning_rate": 1.9975e-05, "loss": 0.1635, "step": 800 }, { "grad_norm": 2.468670606613159, "learning_rate": 2.0225000000000004e-05, "loss": 0.1633, "step": 810 }, { "grad_norm": 2.067035675048828, "learning_rate": 2.0475e-05, "loss": 0.1828, "step": 820 }, { "grad_norm": 3.2508981227874756, "learning_rate": 2.0725e-05, "loss": 0.2121, "step": 830 }, { "grad_norm": 1.859985589981079, "learning_rate": 2.0975e-05, "loss": 0.178, "step": 840 }, { "grad_norm": 2.3349030017852783, "learning_rate": 2.1225e-05, "loss": 0.1844, "step": 850 }, { "grad_norm": 2.579805850982666, "learning_rate": 2.1475e-05, "loss": 0.1595, "step": 860 }, { "grad_norm": 2.408214807510376, "learning_rate": 2.1725e-05, "loss": 0.1952, "step": 870 }, { "grad_norm": 2.4160408973693848, "learning_rate": 2.1975000000000002e-05, "loss": 0.1635, "step": 880 }, { "grad_norm": 2.977281332015991, "learning_rate": 2.2225e-05, "loss": 0.1778, "step": 890 }, { "grad_norm": 2.2975680828094482, "learning_rate": 2.2475e-05, "loss": 0.1769, "step": 900 }, { "grad_norm": 1.944863200187683, "learning_rate": 2.2725000000000003e-05, "loss": 0.167, "step": 910 }, { "grad_norm": 1.5339910984039307, "learning_rate": 2.2975000000000003e-05, "loss": 0.1428, "step": 920 }, { "grad_norm": 2.0159671306610107, "learning_rate": 2.3225000000000002e-05, "loss": 0.1564, "step": 930 }, { "grad_norm": 2.1091220378875732, "learning_rate": 2.3475e-05, "loss": 0.1602, "step": 940 }, { "grad_norm": 1.5893237590789795, "learning_rate": 2.3725e-05, "loss": 0.1418, "step": 950 }, { "grad_norm": 2.135547161102295, "learning_rate": 2.3975e-05, "loss": 0.1508, "step": 960 }, { "grad_norm": 1.8646960258483887, "learning_rate": 2.4225e-05, "loss": 0.18, "step": 970 }, { "grad_norm": 1.9566240310668945, "learning_rate": 2.4475000000000002e-05, "loss": 0.1412, "step": 980 }, { "grad_norm": 2.004134178161621, "learning_rate": 2.4725e-05, "loss": 0.1302, "step": 990 }, { "grad_norm": 1.8272979259490967, "learning_rate": 2.4975e-05, "loss": 0.1436, "step": 1000 }, { "grad_norm": 1.6809829473495483, "learning_rate": 2.5225e-05, "loss": 0.1351, "step": 1010 }, { "grad_norm": 1.743357539176941, "learning_rate": 2.5475e-05, "loss": 0.172, "step": 1020 }, { "grad_norm": 2.02821683883667, "learning_rate": 2.5725e-05, "loss": 0.1277, "step": 1030 }, { "grad_norm": 1.5678287744522095, "learning_rate": 2.5974999999999998e-05, "loss": 0.144, "step": 1040 }, { "grad_norm": 2.3435487747192383, "learning_rate": 2.6225e-05, "loss": 0.1528, "step": 1050 }, { "grad_norm": 1.496344804763794, "learning_rate": 2.6475e-05, "loss": 0.1678, "step": 1060 }, { "grad_norm": 1.979750156402588, "learning_rate": 2.6725e-05, "loss": 0.1726, "step": 1070 }, { "grad_norm": 2.092949390411377, "learning_rate": 2.6975000000000002e-05, "loss": 0.1824, "step": 1080 }, { "grad_norm": 1.5520745515823364, "learning_rate": 2.7225e-05, "loss": 0.1514, "step": 1090 }, { "grad_norm": 2.168461561203003, "learning_rate": 2.7475e-05, "loss": 0.1206, "step": 1100 }, { "grad_norm": 2.1470136642456055, "learning_rate": 2.7725e-05, "loss": 0.1766, "step": 1110 }, { "grad_norm": 1.1509313583374023, "learning_rate": 2.7975000000000002e-05, "loss": 0.1277, "step": 1120 }, { "grad_norm": 1.4077762365341187, "learning_rate": 2.8225e-05, "loss": 0.1483, "step": 1130 }, { "grad_norm": 2.173135757446289, "learning_rate": 2.8475e-05, "loss": 0.1828, "step": 1140 }, { "grad_norm": 1.5813802480697632, "learning_rate": 2.8725e-05, "loss": 0.174, "step": 1150 }, { "grad_norm": 1.6651506423950195, "learning_rate": 2.8975000000000003e-05, "loss": 0.1495, "step": 1160 }, { "grad_norm": 1.8753128051757812, "learning_rate": 2.9225000000000002e-05, "loss": 0.1499, "step": 1170 }, { "grad_norm": 1.8960762023925781, "learning_rate": 2.9475e-05, "loss": 0.1324, "step": 1180 }, { "grad_norm": 1.6046335697174072, "learning_rate": 2.9725000000000004e-05, "loss": 0.1496, "step": 1190 }, { "grad_norm": 2.733522891998291, "learning_rate": 2.9975000000000004e-05, "loss": 0.1534, "step": 1200 }, { "grad_norm": 3.0834760665893555, "learning_rate": 3.0225000000000003e-05, "loss": 0.1624, "step": 1210 }, { "grad_norm": 1.1160377264022827, "learning_rate": 3.0475000000000002e-05, "loss": 0.1466, "step": 1220 }, { "grad_norm": 1.9310168027877808, "learning_rate": 3.0725e-05, "loss": 0.1425, "step": 1230 }, { "grad_norm": 1.213556170463562, "learning_rate": 3.0975e-05, "loss": 0.1525, "step": 1240 }, { "grad_norm": 1.9900226593017578, "learning_rate": 3.122500000000001e-05, "loss": 0.1323, "step": 1250 }, { "grad_norm": 2.1951401233673096, "learning_rate": 3.1475e-05, "loss": 0.141, "step": 1260 }, { "grad_norm": 1.9855964183807373, "learning_rate": 3.1725e-05, "loss": 0.1325, "step": 1270 }, { "grad_norm": 1.4207353591918945, "learning_rate": 3.1975e-05, "loss": 0.1531, "step": 1280 }, { "grad_norm": 1.2286968231201172, "learning_rate": 3.2225e-05, "loss": 0.1445, "step": 1290 }, { "grad_norm": 1.464707374572754, "learning_rate": 3.2474999999999997e-05, "loss": 0.1417, "step": 1300 }, { "grad_norm": 1.2808053493499756, "learning_rate": 3.2725e-05, "loss": 0.1332, "step": 1310 }, { "grad_norm": 1.0549663305282593, "learning_rate": 3.2975e-05, "loss": 0.1421, "step": 1320 }, { "grad_norm": 1.4115490913391113, "learning_rate": 3.3225e-05, "loss": 0.1151, "step": 1330 }, { "grad_norm": 1.3394752740859985, "learning_rate": 3.3475e-05, "loss": 0.1102, "step": 1340 }, { "grad_norm": 1.3478564023971558, "learning_rate": 3.3725e-05, "loss": 0.1481, "step": 1350 }, { "grad_norm": 2.127822160720825, "learning_rate": 3.3975e-05, "loss": 0.1406, "step": 1360 }, { "grad_norm": 1.4861711263656616, "learning_rate": 3.4225e-05, "loss": 0.1161, "step": 1370 }, { "grad_norm": 1.5479358434677124, "learning_rate": 3.4475000000000005e-05, "loss": 0.1168, "step": 1380 }, { "grad_norm": 1.6797703504562378, "learning_rate": 3.4725000000000004e-05, "loss": 0.1181, "step": 1390 }, { "grad_norm": 1.3760179281234741, "learning_rate": 3.4975e-05, "loss": 0.1212, "step": 1400 }, { "grad_norm": 1.2973918914794922, "learning_rate": 3.5225e-05, "loss": 0.1315, "step": 1410 }, { "grad_norm": 1.2697198390960693, "learning_rate": 3.5475e-05, "loss": 0.1299, "step": 1420 }, { "grad_norm": 1.3843201398849487, "learning_rate": 3.5725e-05, "loss": 0.1162, "step": 1430 }, { "grad_norm": 1.3361221551895142, "learning_rate": 3.5975e-05, "loss": 0.1314, "step": 1440 }, { "grad_norm": 1.040103554725647, "learning_rate": 3.6225000000000006e-05, "loss": 0.1452, "step": 1450 }, { "grad_norm": 1.0761785507202148, "learning_rate": 3.6475000000000006e-05, "loss": 0.1404, "step": 1460 }, { "grad_norm": 1.0334229469299316, "learning_rate": 3.6725000000000005e-05, "loss": 0.1384, "step": 1470 }, { "grad_norm": 1.421263575553894, "learning_rate": 3.6975000000000004e-05, "loss": 0.1181, "step": 1480 }, { "grad_norm": 1.5024892091751099, "learning_rate": 3.7225000000000004e-05, "loss": 0.1339, "step": 1490 }, { "grad_norm": 1.1826684474945068, "learning_rate": 3.7475e-05, "loss": 0.1237, "step": 1500 }, { "grad_norm": 0.9820448756217957, "learning_rate": 3.7725e-05, "loss": 0.111, "step": 1510 }, { "grad_norm": 1.6007694005966187, "learning_rate": 3.7975e-05, "loss": 0.1283, "step": 1520 }, { "grad_norm": 0.9924052953720093, "learning_rate": 3.8225e-05, "loss": 0.1297, "step": 1530 }, { "grad_norm": 0.972622811794281, "learning_rate": 3.8475e-05, "loss": 0.1223, "step": 1540 }, { "grad_norm": 1.1575090885162354, "learning_rate": 3.8725e-05, "loss": 0.1165, "step": 1550 }, { "grad_norm": 0.9879919290542603, "learning_rate": 3.8975e-05, "loss": 0.1427, "step": 1560 }, { "grad_norm": 0.9486808180809021, "learning_rate": 3.9225e-05, "loss": 0.1082, "step": 1570 }, { "grad_norm": 1.5942504405975342, "learning_rate": 3.9475000000000004e-05, "loss": 0.1227, "step": 1580 }, { "grad_norm": 1.2734758853912354, "learning_rate": 3.9725e-05, "loss": 0.1343, "step": 1590 }, { "grad_norm": 0.9524399638175964, "learning_rate": 3.9975e-05, "loss": 0.1245, "step": 1600 }, { "grad_norm": 0.8637304902076721, "learning_rate": 4.0225e-05, "loss": 0.1032, "step": 1610 }, { "grad_norm": 1.5943416357040405, "learning_rate": 4.0475e-05, "loss": 0.1233, "step": 1620 }, { "grad_norm": 0.9994795322418213, "learning_rate": 4.0725e-05, "loss": 0.1549, "step": 1630 }, { "grad_norm": 1.044713020324707, "learning_rate": 4.0975e-05, "loss": 0.123, "step": 1640 }, { "grad_norm": 1.0975927114486694, "learning_rate": 4.1225e-05, "loss": 0.1335, "step": 1650 }, { "grad_norm": 1.0759559869766235, "learning_rate": 4.1475000000000005e-05, "loss": 0.1287, "step": 1660 }, { "grad_norm": 1.2826054096221924, "learning_rate": 4.1725000000000005e-05, "loss": 0.1074, "step": 1670 }, { "grad_norm": 1.3388053178787231, "learning_rate": 4.1975000000000004e-05, "loss": 0.1158, "step": 1680 }, { "grad_norm": 0.9900388717651367, "learning_rate": 4.2225e-05, "loss": 0.1237, "step": 1690 }, { "grad_norm": 1.2066534757614136, "learning_rate": 4.2475e-05, "loss": 0.1239, "step": 1700 }, { "grad_norm": 1.7479703426361084, "learning_rate": 4.2725e-05, "loss": 0.1489, "step": 1710 }, { "grad_norm": 0.9357154965400696, "learning_rate": 4.2975e-05, "loss": 0.1293, "step": 1720 }, { "grad_norm": 1.323936104774475, "learning_rate": 4.322500000000001e-05, "loss": 0.1199, "step": 1730 }, { "grad_norm": 1.0725867748260498, "learning_rate": 4.3475000000000006e-05, "loss": 0.1131, "step": 1740 }, { "grad_norm": 1.11674165725708, "learning_rate": 4.3725000000000006e-05, "loss": 0.1167, "step": 1750 }, { "grad_norm": 1.2877349853515625, "learning_rate": 4.3975e-05, "loss": 0.1206, "step": 1760 }, { "grad_norm": 1.1202073097229004, "learning_rate": 4.4225e-05, "loss": 0.1036, "step": 1770 }, { "grad_norm": 0.9302598834037781, "learning_rate": 4.4475e-05, "loss": 0.131, "step": 1780 }, { "grad_norm": 1.5441420078277588, "learning_rate": 4.4725e-05, "loss": 0.1142, "step": 1790 }, { "grad_norm": 1.1070986986160278, "learning_rate": 4.4975e-05, "loss": 0.1235, "step": 1800 }, { "grad_norm": 1.0408257246017456, "learning_rate": 4.5225e-05, "loss": 0.1044, "step": 1810 }, { "grad_norm": 1.2668044567108154, "learning_rate": 4.5475e-05, "loss": 0.1353, "step": 1820 }, { "grad_norm": 1.5305765867233276, "learning_rate": 4.5725e-05, "loss": 0.1197, "step": 1830 }, { "grad_norm": 1.2396742105484009, "learning_rate": 4.5975e-05, "loss": 0.1273, "step": 1840 }, { "grad_norm": 0.9998034238815308, "learning_rate": 4.6225e-05, "loss": 0.1499, "step": 1850 }, { "grad_norm": 1.20533287525177, "learning_rate": 4.6475000000000005e-05, "loss": 0.143, "step": 1860 }, { "grad_norm": 1.2437342405319214, "learning_rate": 4.6725000000000004e-05, "loss": 0.119, "step": 1870 }, { "grad_norm": 1.237125039100647, "learning_rate": 4.6975000000000003e-05, "loss": 0.0935, "step": 1880 }, { "grad_norm": 1.082076072692871, "learning_rate": 4.7225e-05, "loss": 0.1225, "step": 1890 }, { "grad_norm": 1.2639411687850952, "learning_rate": 4.7475e-05, "loss": 0.1098, "step": 1900 }, { "grad_norm": 1.1132287979125977, "learning_rate": 4.7725e-05, "loss": 0.1341, "step": 1910 }, { "grad_norm": 1.0274171829223633, "learning_rate": 4.7975e-05, "loss": 0.1186, "step": 1920 }, { "grad_norm": 0.919283390045166, "learning_rate": 4.822500000000001e-05, "loss": 0.1069, "step": 1930 }, { "grad_norm": 0.9867066144943237, "learning_rate": 4.8475000000000006e-05, "loss": 0.1077, "step": 1940 }, { "grad_norm": 1.0011707544326782, "learning_rate": 4.8725000000000005e-05, "loss": 0.104, "step": 1950 }, { "grad_norm": 1.5317845344543457, "learning_rate": 4.8975000000000005e-05, "loss": 0.1404, "step": 1960 }, { "grad_norm": 0.8776949644088745, "learning_rate": 4.9225000000000004e-05, "loss": 0.0966, "step": 1970 }, { "grad_norm": 1.4857360124588013, "learning_rate": 4.9475e-05, "loss": 0.1249, "step": 1980 }, { "grad_norm": 1.0491628646850586, "learning_rate": 4.9725e-05, "loss": 0.0979, "step": 1990 }, { "grad_norm": 0.9116098880767822, "learning_rate": 4.9975e-05, "loss": 0.1178, "step": 2000 }, { "grad_norm": 1.2793480157852173, "learning_rate": 5.0225e-05, "loss": 0.1451, "step": 2010 }, { "grad_norm": 1.1006876230239868, "learning_rate": 5.047500000000001e-05, "loss": 0.1188, "step": 2020 }, { "grad_norm": 1.1112377643585205, "learning_rate": 5.0725e-05, "loss": 0.093, "step": 2030 }, { "grad_norm": 0.6314584612846375, "learning_rate": 5.0975000000000006e-05, "loss": 0.1194, "step": 2040 }, { "grad_norm": 1.026397705078125, "learning_rate": 5.1225e-05, "loss": 0.1184, "step": 2050 }, { "grad_norm": 1.1197253465652466, "learning_rate": 5.1475000000000004e-05, "loss": 0.1051, "step": 2060 }, { "grad_norm": 0.7780553102493286, "learning_rate": 5.1725000000000004e-05, "loss": 0.0974, "step": 2070 }, { "grad_norm": 0.9816649556159973, "learning_rate": 5.197500000000001e-05, "loss": 0.1049, "step": 2080 }, { "grad_norm": 0.891352653503418, "learning_rate": 5.2225e-05, "loss": 0.1031, "step": 2090 }, { "grad_norm": 1.2035142183303833, "learning_rate": 5.247500000000001e-05, "loss": 0.1163, "step": 2100 }, { "grad_norm": 0.7586325407028198, "learning_rate": 5.2725e-05, "loss": 0.1324, "step": 2110 }, { "grad_norm": 0.7698188424110413, "learning_rate": 5.297500000000001e-05, "loss": 0.1008, "step": 2120 }, { "grad_norm": 1.1944496631622314, "learning_rate": 5.3225e-05, "loss": 0.1065, "step": 2130 }, { "grad_norm": 0.7752849459648132, "learning_rate": 5.3475e-05, "loss": 0.0871, "step": 2140 }, { "grad_norm": 0.8189555406570435, "learning_rate": 5.3725000000000005e-05, "loss": 0.1317, "step": 2150 }, { "grad_norm": 0.78727787733078, "learning_rate": 5.3975e-05, "loss": 0.0999, "step": 2160 }, { "grad_norm": 0.7721348404884338, "learning_rate": 5.4225000000000003e-05, "loss": 0.0989, "step": 2170 }, { "grad_norm": 0.6773834228515625, "learning_rate": 5.4474999999999996e-05, "loss": 0.1251, "step": 2180 }, { "grad_norm": 0.8012632131576538, "learning_rate": 5.4725e-05, "loss": 0.1029, "step": 2190 }, { "grad_norm": 0.9874103665351868, "learning_rate": 5.4975e-05, "loss": 0.107, "step": 2200 }, { "grad_norm": 0.8125993013381958, "learning_rate": 5.522500000000001e-05, "loss": 0.1107, "step": 2210 }, { "grad_norm": 1.1150137186050415, "learning_rate": 5.5475e-05, "loss": 0.0944, "step": 2220 }, { "grad_norm": 0.7674887180328369, "learning_rate": 5.5725000000000006e-05, "loss": 0.1002, "step": 2230 }, { "grad_norm": 0.6333231329917908, "learning_rate": 5.5975e-05, "loss": 0.098, "step": 2240 }, { "grad_norm": 0.8490403890609741, "learning_rate": 5.6225000000000005e-05, "loss": 0.1097, "step": 2250 }, { "grad_norm": 0.8553760647773743, "learning_rate": 5.6475e-05, "loss": 0.0922, "step": 2260 }, { "grad_norm": 0.7606274485588074, "learning_rate": 5.6725e-05, "loss": 0.0943, "step": 2270 }, { "grad_norm": 0.7568450570106506, "learning_rate": 5.6975e-05, "loss": 0.1034, "step": 2280 }, { "grad_norm": 1.2010023593902588, "learning_rate": 5.722500000000001e-05, "loss": 0.1298, "step": 2290 }, { "grad_norm": 0.7152898907661438, "learning_rate": 5.7475e-05, "loss": 0.0908, "step": 2300 }, { "grad_norm": 0.8346425294876099, "learning_rate": 5.772500000000001e-05, "loss": 0.0978, "step": 2310 }, { "grad_norm": 0.689307689666748, "learning_rate": 5.7975e-05, "loss": 0.0824, "step": 2320 }, { "grad_norm": 0.735334575176239, "learning_rate": 5.8225000000000006e-05, "loss": 0.1143, "step": 2330 }, { "grad_norm": 0.6812131404876709, "learning_rate": 5.8475000000000005e-05, "loss": 0.1033, "step": 2340 }, { "grad_norm": 0.8845462203025818, "learning_rate": 5.8725000000000004e-05, "loss": 0.1034, "step": 2350 }, { "grad_norm": 0.887540340423584, "learning_rate": 5.8975000000000004e-05, "loss": 0.0977, "step": 2360 }, { "grad_norm": 0.7078615427017212, "learning_rate": 5.922500000000001e-05, "loss": 0.1072, "step": 2370 }, { "grad_norm": 0.7703161835670471, "learning_rate": 5.9475e-05, "loss": 0.1131, "step": 2380 }, { "grad_norm": 0.8678466081619263, "learning_rate": 5.9724999999999995e-05, "loss": 0.1186, "step": 2390 }, { "grad_norm": 0.6814717650413513, "learning_rate": 5.9975e-05, "loss": 0.1028, "step": 2400 }, { "grad_norm": 0.8692256808280945, "learning_rate": 6.0225e-05, "loss": 0.1042, "step": 2410 }, { "grad_norm": 0.8696652054786682, "learning_rate": 6.0475000000000006e-05, "loss": 0.1369, "step": 2420 }, { "grad_norm": 0.9023002982139587, "learning_rate": 6.0725e-05, "loss": 0.1085, "step": 2430 }, { "grad_norm": 0.9979051947593689, "learning_rate": 6.0975000000000005e-05, "loss": 0.1206, "step": 2440 }, { "grad_norm": 0.8776962757110596, "learning_rate": 6.1225e-05, "loss": 0.0973, "step": 2450 }, { "grad_norm": 0.6429100632667542, "learning_rate": 6.1475e-05, "loss": 0.0933, "step": 2460 }, { "grad_norm": 0.7949334383010864, "learning_rate": 6.1725e-05, "loss": 0.0997, "step": 2470 }, { "grad_norm": 0.8405593037605286, "learning_rate": 6.1975e-05, "loss": 0.121, "step": 2480 }, { "grad_norm": 0.9628981947898865, "learning_rate": 6.2225e-05, "loss": 0.1003, "step": 2490 }, { "grad_norm": 0.74796462059021, "learning_rate": 6.2475e-05, "loss": 0.0905, "step": 2500 }, { "grad_norm": 0.8433115482330322, "learning_rate": 6.2725e-05, "loss": 0.0876, "step": 2510 }, { "grad_norm": 0.7153704166412354, "learning_rate": 6.297500000000001e-05, "loss": 0.0877, "step": 2520 }, { "grad_norm": 0.7668380737304688, "learning_rate": 6.3225e-05, "loss": 0.0831, "step": 2530 }, { "grad_norm": 0.6176163554191589, "learning_rate": 6.347500000000001e-05, "loss": 0.0949, "step": 2540 }, { "grad_norm": 0.5773432850837708, "learning_rate": 6.3725e-05, "loss": 0.1045, "step": 2550 }, { "grad_norm": 0.949303150177002, "learning_rate": 6.397500000000001e-05, "loss": 0.105, "step": 2560 }, { "grad_norm": 0.8517661690711975, "learning_rate": 6.4225e-05, "loss": 0.1039, "step": 2570 }, { "grad_norm": 0.6492380499839783, "learning_rate": 6.447500000000001e-05, "loss": 0.1041, "step": 2580 }, { "grad_norm": 0.7586159110069275, "learning_rate": 6.4725e-05, "loss": 0.1006, "step": 2590 }, { "grad_norm": 0.8489075303077698, "learning_rate": 6.497500000000001e-05, "loss": 0.105, "step": 2600 }, { "grad_norm": 0.8067915439605713, "learning_rate": 6.5225e-05, "loss": 0.1067, "step": 2610 }, { "grad_norm": 0.8483464121818542, "learning_rate": 6.5475e-05, "loss": 0.0898, "step": 2620 }, { "grad_norm": 0.7459204792976379, "learning_rate": 6.5725e-05, "loss": 0.0959, "step": 2630 }, { "grad_norm": 0.7203967571258545, "learning_rate": 6.5975e-05, "loss": 0.1162, "step": 2640 }, { "grad_norm": 0.5645008087158203, "learning_rate": 6.6225e-05, "loss": 0.1074, "step": 2650 }, { "grad_norm": 0.7819217443466187, "learning_rate": 6.6475e-05, "loss": 0.1123, "step": 2660 }, { "grad_norm": 0.6358935236930847, "learning_rate": 6.672500000000001e-05, "loss": 0.112, "step": 2670 }, { "grad_norm": 0.8776800632476807, "learning_rate": 6.6975e-05, "loss": 0.1234, "step": 2680 }, { "grad_norm": 0.6010076403617859, "learning_rate": 6.722500000000001e-05, "loss": 0.1008, "step": 2690 }, { "grad_norm": 0.7875289916992188, "learning_rate": 6.7475e-05, "loss": 0.0892, "step": 2700 }, { "grad_norm": 0.7562458515167236, "learning_rate": 6.7725e-05, "loss": 0.0987, "step": 2710 }, { "grad_norm": 0.5519564747810364, "learning_rate": 6.7975e-05, "loss": 0.1207, "step": 2720 }, { "grad_norm": 0.7309026718139648, "learning_rate": 6.8225e-05, "loss": 0.1019, "step": 2730 }, { "grad_norm": 0.7063409686088562, "learning_rate": 6.8475e-05, "loss": 0.0958, "step": 2740 }, { "grad_norm": 0.9421608448028564, "learning_rate": 6.8725e-05, "loss": 0.1082, "step": 2750 }, { "grad_norm": 0.7463240623474121, "learning_rate": 6.8975e-05, "loss": 0.0895, "step": 2760 }, { "grad_norm": 0.684904158115387, "learning_rate": 6.9225e-05, "loss": 0.0902, "step": 2770 }, { "grad_norm": 0.7690800428390503, "learning_rate": 6.9475e-05, "loss": 0.1072, "step": 2780 }, { "grad_norm": 0.6356088519096375, "learning_rate": 6.9725e-05, "loss": 0.0984, "step": 2790 }, { "grad_norm": 0.8271968364715576, "learning_rate": 6.997500000000001e-05, "loss": 0.0903, "step": 2800 }, { "grad_norm": 0.86630779504776, "learning_rate": 7.022500000000001e-05, "loss": 0.1181, "step": 2810 }, { "grad_norm": 0.8602688908576965, "learning_rate": 7.0475e-05, "loss": 0.1075, "step": 2820 }, { "grad_norm": 0.740068793296814, "learning_rate": 7.072500000000001e-05, "loss": 0.0989, "step": 2830 }, { "grad_norm": 0.622437596321106, "learning_rate": 7.0975e-05, "loss": 0.1118, "step": 2840 }, { "grad_norm": 0.7722273468971252, "learning_rate": 7.122500000000001e-05, "loss": 0.1006, "step": 2850 }, { "grad_norm": 0.7438315749168396, "learning_rate": 7.1475e-05, "loss": 0.0921, "step": 2860 }, { "grad_norm": 0.6028250455856323, "learning_rate": 7.172500000000001e-05, "loss": 0.0912, "step": 2870 }, { "grad_norm": 0.583326518535614, "learning_rate": 7.1975e-05, "loss": 0.0878, "step": 2880 }, { "grad_norm": 0.7564980983734131, "learning_rate": 7.2225e-05, "loss": 0.0952, "step": 2890 }, { "grad_norm": 0.5247694253921509, "learning_rate": 7.2475e-05, "loss": 0.1048, "step": 2900 }, { "grad_norm": 0.9172112941741943, "learning_rate": 7.272499999999999e-05, "loss": 0.0928, "step": 2910 }, { "grad_norm": 0.5955166220664978, "learning_rate": 7.2975e-05, "loss": 0.0959, "step": 2920 }, { "grad_norm": 0.6690652966499329, "learning_rate": 7.3225e-05, "loss": 0.0778, "step": 2930 }, { "grad_norm": 0.6502119898796082, "learning_rate": 7.347500000000001e-05, "loss": 0.0777, "step": 2940 }, { "grad_norm": 0.7560237050056458, "learning_rate": 7.3725e-05, "loss": 0.0846, "step": 2950 }, { "grad_norm": 0.6795781254768372, "learning_rate": 7.397500000000001e-05, "loss": 0.0958, "step": 2960 }, { "grad_norm": 0.6178928017616272, "learning_rate": 7.4225e-05, "loss": 0.0852, "step": 2970 }, { "grad_norm": 0.800875723361969, "learning_rate": 7.447500000000001e-05, "loss": 0.098, "step": 2980 }, { "grad_norm": 0.7155036926269531, "learning_rate": 7.4725e-05, "loss": 0.0985, "step": 2990 }, { "grad_norm": 0.6415840983390808, "learning_rate": 7.4975e-05, "loss": 0.1017, "step": 3000 }, { "grad_norm": 0.7156231999397278, "learning_rate": 7.5225e-05, "loss": 0.0755, "step": 3010 }, { "grad_norm": 0.6197031140327454, "learning_rate": 7.5475e-05, "loss": 0.0881, "step": 3020 }, { "grad_norm": 0.7514368891716003, "learning_rate": 7.5725e-05, "loss": 0.0875, "step": 3030 }, { "grad_norm": 0.6251776218414307, "learning_rate": 7.5975e-05, "loss": 0.0955, "step": 3040 }, { "grad_norm": 0.7700625061988831, "learning_rate": 7.6225e-05, "loss": 0.0979, "step": 3050 }, { "grad_norm": 0.8082242608070374, "learning_rate": 7.6475e-05, "loss": 0.0942, "step": 3060 }, { "grad_norm": 0.6902333498001099, "learning_rate": 7.672500000000001e-05, "loss": 0.0716, "step": 3070 }, { "grad_norm": 0.7287672162055969, "learning_rate": 7.697500000000001e-05, "loss": 0.0922, "step": 3080 }, { "grad_norm": 0.7879593372344971, "learning_rate": 7.722500000000001e-05, "loss": 0.1104, "step": 3090 }, { "grad_norm": 0.7823359370231628, "learning_rate": 7.747500000000001e-05, "loss": 0.0942, "step": 3100 }, { "grad_norm": 0.7347156405448914, "learning_rate": 7.7725e-05, "loss": 0.0875, "step": 3110 }, { "grad_norm": 0.6600995063781738, "learning_rate": 7.797500000000001e-05, "loss": 0.0834, "step": 3120 }, { "grad_norm": 0.6094813346862793, "learning_rate": 7.8225e-05, "loss": 0.0866, "step": 3130 }, { "grad_norm": 0.6312695741653442, "learning_rate": 7.8475e-05, "loss": 0.0896, "step": 3140 }, { "grad_norm": 0.6996814012527466, "learning_rate": 7.8725e-05, "loss": 0.0873, "step": 3150 }, { "grad_norm": 0.6207980513572693, "learning_rate": 7.8975e-05, "loss": 0.0853, "step": 3160 }, { "grad_norm": 0.682526707649231, "learning_rate": 7.9225e-05, "loss": 0.0806, "step": 3170 }, { "grad_norm": 0.8607361316680908, "learning_rate": 7.9475e-05, "loss": 0.1037, "step": 3180 }, { "grad_norm": 0.8278142809867859, "learning_rate": 7.9725e-05, "loss": 0.0943, "step": 3190 }, { "grad_norm": 0.7249951958656311, "learning_rate": 7.9975e-05, "loss": 0.0955, "step": 3200 }, { "grad_norm": 0.885858952999115, "learning_rate": 8.022500000000001e-05, "loss": 0.0879, "step": 3210 }, { "grad_norm": 0.9108887910842896, "learning_rate": 8.0475e-05, "loss": 0.1163, "step": 3220 }, { "grad_norm": 0.7320034503936768, "learning_rate": 8.072500000000001e-05, "loss": 0.0973, "step": 3230 }, { "grad_norm": 0.5469526648521423, "learning_rate": 8.0975e-05, "loss": 0.0863, "step": 3240 }, { "grad_norm": 0.5540540218353271, "learning_rate": 8.122500000000001e-05, "loss": 0.0953, "step": 3250 }, { "grad_norm": 1.115841031074524, "learning_rate": 8.1475e-05, "loss": 0.1047, "step": 3260 }, { "grad_norm": 0.8033411502838135, "learning_rate": 8.172500000000001e-05, "loss": 0.0963, "step": 3270 }, { "grad_norm": 0.545085072517395, "learning_rate": 8.1975e-05, "loss": 0.0857, "step": 3280 }, { "grad_norm": 0.514118492603302, "learning_rate": 8.2225e-05, "loss": 0.0865, "step": 3290 }, { "grad_norm": 0.5399650931358337, "learning_rate": 8.2475e-05, "loss": 0.0817, "step": 3300 }, { "grad_norm": 0.7783600091934204, "learning_rate": 8.2725e-05, "loss": 0.0954, "step": 3310 }, { "grad_norm": 0.7915538549423218, "learning_rate": 8.2975e-05, "loss": 0.1122, "step": 3320 }, { "grad_norm": 0.7557186484336853, "learning_rate": 8.3225e-05, "loss": 0.0988, "step": 3330 }, { "grad_norm": 0.7403407096862793, "learning_rate": 8.347500000000001e-05, "loss": 0.0943, "step": 3340 }, { "grad_norm": 0.6044697761535645, "learning_rate": 8.3725e-05, "loss": 0.0851, "step": 3350 }, { "grad_norm": 0.7047536969184875, "learning_rate": 8.397500000000001e-05, "loss": 0.0944, "step": 3360 }, { "grad_norm": 0.5471181273460388, "learning_rate": 8.422500000000001e-05, "loss": 0.0934, "step": 3370 }, { "grad_norm": 0.7111130952835083, "learning_rate": 8.447500000000001e-05, "loss": 0.1436, "step": 3380 }, { "grad_norm": 0.6257320046424866, "learning_rate": 8.4725e-05, "loss": 0.0908, "step": 3390 }, { "grad_norm": 0.6705989241600037, "learning_rate": 8.4975e-05, "loss": 0.0929, "step": 3400 }, { "grad_norm": 0.6962879300117493, "learning_rate": 8.5225e-05, "loss": 0.0916, "step": 3410 }, { "grad_norm": 0.7273101806640625, "learning_rate": 8.5475e-05, "loss": 0.0838, "step": 3420 }, { "grad_norm": 0.5893482565879822, "learning_rate": 8.5725e-05, "loss": 0.0859, "step": 3430 }, { "grad_norm": 0.6493954062461853, "learning_rate": 8.5975e-05, "loss": 0.111, "step": 3440 }, { "grad_norm": 0.7825080156326294, "learning_rate": 8.6225e-05, "loss": 0.0921, "step": 3450 }, { "grad_norm": 0.49176713824272156, "learning_rate": 8.6475e-05, "loss": 0.0768, "step": 3460 }, { "grad_norm": 0.6128948330879211, "learning_rate": 8.672500000000001e-05, "loss": 0.1002, "step": 3470 }, { "grad_norm": 0.7143135666847229, "learning_rate": 8.6975e-05, "loss": 0.0942, "step": 3480 }, { "grad_norm": 0.5378119349479675, "learning_rate": 8.7225e-05, "loss": 0.0912, "step": 3490 }, { "grad_norm": 0.5702638030052185, "learning_rate": 8.747500000000001e-05, "loss": 0.076, "step": 3500 }, { "grad_norm": 0.5983698964118958, "learning_rate": 8.7725e-05, "loss": 0.0802, "step": 3510 }, { "grad_norm": 0.8389286398887634, "learning_rate": 8.797500000000001e-05, "loss": 0.0872, "step": 3520 }, { "grad_norm": 0.6865819692611694, "learning_rate": 8.8225e-05, "loss": 0.0816, "step": 3530 }, { "grad_norm": 0.46662089228630066, "learning_rate": 8.847500000000001e-05, "loss": 0.0806, "step": 3540 }, { "grad_norm": 0.6724150776863098, "learning_rate": 8.8725e-05, "loss": 0.0915, "step": 3550 }, { "grad_norm": 0.619030773639679, "learning_rate": 8.897500000000001e-05, "loss": 0.0713, "step": 3560 }, { "grad_norm": 0.6042924523353577, "learning_rate": 8.9225e-05, "loss": 0.0954, "step": 3570 }, { "grad_norm": 0.5618426203727722, "learning_rate": 8.9475e-05, "loss": 0.0965, "step": 3580 }, { "grad_norm": 0.7726119160652161, "learning_rate": 8.9725e-05, "loss": 0.1005, "step": 3590 }, { "grad_norm": 0.7333736419677734, "learning_rate": 8.9975e-05, "loss": 0.0887, "step": 3600 }, { "grad_norm": 0.6378642320632935, "learning_rate": 9.0225e-05, "loss": 0.0888, "step": 3610 }, { "grad_norm": 0.8210988640785217, "learning_rate": 9.0475e-05, "loss": 0.0895, "step": 3620 }, { "grad_norm": 0.6252534985542297, "learning_rate": 9.072500000000001e-05, "loss": 0.0724, "step": 3630 }, { "grad_norm": 0.4078366458415985, "learning_rate": 9.0975e-05, "loss": 0.073, "step": 3640 }, { "grad_norm": 0.5861696004867554, "learning_rate": 9.122500000000001e-05, "loss": 0.0781, "step": 3650 }, { "grad_norm": 0.7361677885055542, "learning_rate": 9.1475e-05, "loss": 0.0866, "step": 3660 }, { "grad_norm": 0.656457781791687, "learning_rate": 9.172500000000001e-05, "loss": 0.088, "step": 3670 }, { "grad_norm": 0.6332527995109558, "learning_rate": 9.1975e-05, "loss": 0.09, "step": 3680 }, { "grad_norm": 0.6745708584785461, "learning_rate": 9.2225e-05, "loss": 0.0861, "step": 3690 }, { "grad_norm": 0.8137019276618958, "learning_rate": 9.2475e-05, "loss": 0.0853, "step": 3700 }, { "grad_norm": 0.6163873672485352, "learning_rate": 9.2725e-05, "loss": 0.0823, "step": 3710 }, { "grad_norm": 0.5831032395362854, "learning_rate": 9.2975e-05, "loss": 0.0905, "step": 3720 }, { "grad_norm": 0.6978399157524109, "learning_rate": 9.3225e-05, "loss": 0.0734, "step": 3730 }, { "grad_norm": 0.6969825029373169, "learning_rate": 9.3475e-05, "loss": 0.0785, "step": 3740 }, { "grad_norm": 0.6341997385025024, "learning_rate": 9.3725e-05, "loss": 0.0901, "step": 3750 }, { "grad_norm": 0.6204107403755188, "learning_rate": 9.397500000000001e-05, "loss": 0.0849, "step": 3760 }, { "grad_norm": 0.7125059366226196, "learning_rate": 9.422500000000001e-05, "loss": 0.1046, "step": 3770 }, { "grad_norm": 0.5445437431335449, "learning_rate": 9.4475e-05, "loss": 0.0968, "step": 3780 }, { "grad_norm": 0.8947967886924744, "learning_rate": 9.472500000000001e-05, "loss": 0.0917, "step": 3790 }, { "grad_norm": 0.566863477230072, "learning_rate": 9.4975e-05, "loss": 0.1168, "step": 3800 }, { "grad_norm": 0.5553996562957764, "learning_rate": 9.522500000000001e-05, "loss": 0.0785, "step": 3810 }, { "grad_norm": 0.681637704372406, "learning_rate": 9.5475e-05, "loss": 0.096, "step": 3820 }, { "grad_norm": 0.5515270233154297, "learning_rate": 9.572500000000001e-05, "loss": 0.0839, "step": 3830 }, { "grad_norm": 0.8616570830345154, "learning_rate": 9.5975e-05, "loss": 0.0701, "step": 3840 }, { "grad_norm": 0.6740477085113525, "learning_rate": 9.622500000000001e-05, "loss": 0.0925, "step": 3850 }, { "grad_norm": 0.6459686160087585, "learning_rate": 9.6475e-05, "loss": 0.0783, "step": 3860 }, { "grad_norm": 0.6132193207740784, "learning_rate": 9.6725e-05, "loss": 0.0774, "step": 3870 }, { "grad_norm": 0.5325903296470642, "learning_rate": 9.6975e-05, "loss": 0.0668, "step": 3880 }, { "grad_norm": 0.7001843452453613, "learning_rate": 9.7225e-05, "loss": 0.0954, "step": 3890 }, { "grad_norm": 0.674278974533081, "learning_rate": 9.747500000000001e-05, "loss": 0.0879, "step": 3900 }, { "grad_norm": 0.5039471387863159, "learning_rate": 9.7725e-05, "loss": 0.0853, "step": 3910 }, { "grad_norm": 0.779964804649353, "learning_rate": 9.797500000000001e-05, "loss": 0.0912, "step": 3920 }, { "grad_norm": 0.5644387602806091, "learning_rate": 9.8225e-05, "loss": 0.0835, "step": 3930 }, { "grad_norm": 0.6944144368171692, "learning_rate": 9.847500000000001e-05, "loss": 0.0961, "step": 3940 }, { "grad_norm": 0.5628661513328552, "learning_rate": 9.8725e-05, "loss": 0.0921, "step": 3950 }, { "grad_norm": 0.5846566557884216, "learning_rate": 9.897500000000001e-05, "loss": 0.1007, "step": 3960 }, { "grad_norm": 0.6191354393959045, "learning_rate": 9.9225e-05, "loss": 0.0784, "step": 3970 }, { "grad_norm": 0.5801456570625305, "learning_rate": 9.9475e-05, "loss": 0.0913, "step": 3980 }, { "grad_norm": 0.8229723572731018, "learning_rate": 9.9725e-05, "loss": 0.0914, "step": 3990 }, { "grad_norm": 0.8687177896499634, "learning_rate": 9.9975e-05, "loss": 0.0928, "step": 4000 }, { "grad_norm": 0.6486630439758301, "learning_rate": 9.999999653982884e-05, "loss": 0.0878, "step": 4010 }, { "grad_norm": 0.6722474098205566, "learning_rate": 9.999998457874392e-05, "loss": 0.0902, "step": 4020 }, { "grad_norm": 0.5630568265914917, "learning_rate": 9.999996407402913e-05, "loss": 0.0904, "step": 4030 }, { "grad_norm": 0.51105135679245, "learning_rate": 9.999993502568801e-05, "loss": 0.0752, "step": 4040 }, { "grad_norm": 0.6104747653007507, "learning_rate": 9.999989743372548e-05, "loss": 0.0868, "step": 4050 }, { "grad_norm": 0.7069773077964783, "learning_rate": 9.999985129814798e-05, "loss": 0.0832, "step": 4060 }, { "grad_norm": 0.6861071586608887, "learning_rate": 9.99997966189634e-05, "loss": 0.0785, "step": 4070 }, { "grad_norm": 0.6679325699806213, "learning_rate": 9.999973339618107e-05, "loss": 0.0946, "step": 4080 }, { "grad_norm": 0.6111060976982117, "learning_rate": 9.999966162981179e-05, "loss": 0.0898, "step": 4090 }, { "grad_norm": 0.5768477916717529, "learning_rate": 9.999958131986784e-05, "loss": 0.077, "step": 4100 }, { "grad_norm": 0.6191268563270569, "learning_rate": 9.999949246636293e-05, "loss": 0.0833, "step": 4110 }, { "grad_norm": 0.9080634713172913, "learning_rate": 9.999939506931224e-05, "loss": 0.0979, "step": 4120 }, { "grad_norm": 0.5218828320503235, "learning_rate": 9.999928912873243e-05, "loss": 0.0888, "step": 4130 }, { "grad_norm": 0.6252426505088806, "learning_rate": 9.999917464464159e-05, "loss": 0.0931, "step": 4140 }, { "grad_norm": 0.596807062625885, "learning_rate": 9.999905161705929e-05, "loss": 0.0993, "step": 4150 }, { "grad_norm": 0.5591884851455688, "learning_rate": 9.999892004600653e-05, "loss": 0.0815, "step": 4160 }, { "grad_norm": 0.8507593870162964, "learning_rate": 9.999877993150581e-05, "loss": 0.0847, "step": 4170 }, { "grad_norm": 0.5746834874153137, "learning_rate": 9.999863127358108e-05, "loss": 0.0715, "step": 4180 }, { "grad_norm": 0.5294283032417297, "learning_rate": 9.999847407225773e-05, "loss": 0.0913, "step": 4190 }, { "grad_norm": 0.6025252938270569, "learning_rate": 9.999830832756262e-05, "loss": 0.0847, "step": 4200 }, { "grad_norm": 0.5119889378547668, "learning_rate": 9.999813403952407e-05, "loss": 0.0886, "step": 4210 }, { "grad_norm": 0.5102826952934265, "learning_rate": 9.999795120817187e-05, "loss": 0.1004, "step": 4220 }, { "grad_norm": 0.7677639722824097, "learning_rate": 9.999775983353725e-05, "loss": 0.1127, "step": 4230 }, { "grad_norm": 0.8045722246170044, "learning_rate": 9.999755991565292e-05, "loss": 0.0963, "step": 4240 }, { "grad_norm": 0.5217555165290833, "learning_rate": 9.999735145455303e-05, "loss": 0.0978, "step": 4250 }, { "grad_norm": 0.6077550649642944, "learning_rate": 9.99971344502732e-05, "loss": 0.0983, "step": 4260 }, { "grad_norm": 0.6040435433387756, "learning_rate": 9.999690890285053e-05, "loss": 0.0958, "step": 4270 }, { "grad_norm": 0.5658352971076965, "learning_rate": 9.999667481232356e-05, "loss": 0.085, "step": 4280 }, { "grad_norm": 0.7794869542121887, "learning_rate": 9.999643217873225e-05, "loss": 0.0916, "step": 4290 }, { "grad_norm": 0.7427794933319092, "learning_rate": 9.999618100211809e-05, "loss": 0.0793, "step": 4300 }, { "grad_norm": 0.5041927099227905, "learning_rate": 9.999592128252402e-05, "loss": 0.0873, "step": 4310 }, { "grad_norm": 0.7539704442024231, "learning_rate": 9.999565301999437e-05, "loss": 0.1128, "step": 4320 }, { "grad_norm": 0.5182201862335205, "learning_rate": 9.999537621457502e-05, "loss": 0.0738, "step": 4330 }, { "grad_norm": 0.5854272842407227, "learning_rate": 9.999509086631323e-05, "loss": 0.0827, "step": 4340 }, { "grad_norm": 0.47699281573295593, "learning_rate": 9.99947969752578e-05, "loss": 0.0655, "step": 4350 }, { "grad_norm": 0.5443515777587891, "learning_rate": 9.999449454145891e-05, "loss": 0.0766, "step": 4360 }, { "grad_norm": 0.6993736624717712, "learning_rate": 9.999418356496827e-05, "loss": 0.0783, "step": 4370 }, { "grad_norm": 0.6094298362731934, "learning_rate": 9.999386404583899e-05, "loss": 0.091, "step": 4380 }, { "grad_norm": 0.6208835244178772, "learning_rate": 9.999353598412568e-05, "loss": 0.0746, "step": 4390 }, { "grad_norm": 0.6582993865013123, "learning_rate": 9.999319937988442e-05, "loss": 0.0928, "step": 4400 }, { "grad_norm": 0.7512438297271729, "learning_rate": 9.999285423317268e-05, "loss": 0.0838, "step": 4410 }, { "grad_norm": 0.5576356649398804, "learning_rate": 9.999250054404947e-05, "loss": 0.1043, "step": 4420 }, { "grad_norm": 0.4824795424938202, "learning_rate": 9.99921383125752e-05, "loss": 0.0873, "step": 4430 }, { "grad_norm": 0.5182933807373047, "learning_rate": 9.99917675388118e-05, "loss": 0.0942, "step": 4440 }, { "grad_norm": 0.8796399831771851, "learning_rate": 9.99913882228226e-05, "loss": 0.1072, "step": 4450 }, { "grad_norm": 0.49580875039100647, "learning_rate": 9.999100036467242e-05, "loss": 0.0939, "step": 4460 }, { "grad_norm": 0.5728574991226196, "learning_rate": 9.999060396442753e-05, "loss": 0.0884, "step": 4470 }, { "grad_norm": 0.4653300344944, "learning_rate": 9.999019902215566e-05, "loss": 0.0937, "step": 4480 }, { "grad_norm": 0.5733027458190918, "learning_rate": 9.998978553792602e-05, "loss": 0.0958, "step": 4490 }, { "grad_norm": 0.6129554510116577, "learning_rate": 9.998936351180926e-05, "loss": 0.0849, "step": 4500 }, { "grad_norm": 0.461276113986969, "learning_rate": 9.998893294387747e-05, "loss": 0.1141, "step": 4510 }, { "grad_norm": 0.6880885362625122, "learning_rate": 9.998849383420426e-05, "loss": 0.0809, "step": 4520 }, { "grad_norm": 0.6465590000152588, "learning_rate": 9.998804618286465e-05, "loss": 0.0761, "step": 4530 }, { "grad_norm": 0.6465198397636414, "learning_rate": 9.99875899899351e-05, "loss": 0.1, "step": 4540 }, { "grad_norm": 0.5432243347167969, "learning_rate": 9.99871252554936e-05, "loss": 0.0932, "step": 4550 }, { "grad_norm": 0.7017683982849121, "learning_rate": 9.998665197961955e-05, "loss": 0.0912, "step": 4560 }, { "grad_norm": 0.627629280090332, "learning_rate": 9.998617016239379e-05, "loss": 0.0933, "step": 4570 }, { "grad_norm": 0.9303929209709167, "learning_rate": 9.998567980389869e-05, "loss": 0.0848, "step": 4580 }, { "grad_norm": 0.6766685843467712, "learning_rate": 9.998518090421802e-05, "loss": 0.0925, "step": 4590 }, { "grad_norm": 0.6262253522872925, "learning_rate": 9.998467346343703e-05, "loss": 0.0847, "step": 4600 }, { "grad_norm": 0.562491774559021, "learning_rate": 9.998415748164243e-05, "loss": 0.0839, "step": 4610 }, { "grad_norm": 0.6020750403404236, "learning_rate": 9.998363295892238e-05, "loss": 0.0717, "step": 4620 }, { "grad_norm": 0.6641767024993896, "learning_rate": 9.998309989536652e-05, "loss": 0.0894, "step": 4630 }, { "grad_norm": 0.491462767124176, "learning_rate": 9.998255829106593e-05, "loss": 0.0847, "step": 4640 }, { "grad_norm": 0.6818451881408691, "learning_rate": 9.998200814611316e-05, "loss": 0.0733, "step": 4650 }, { "grad_norm": 0.7361763715744019, "learning_rate": 9.998144946060219e-05, "loss": 0.0836, "step": 4660 }, { "grad_norm": 0.8766361474990845, "learning_rate": 9.998088223462852e-05, "loss": 0.0816, "step": 4670 }, { "grad_norm": 0.5376171469688416, "learning_rate": 9.998030646828905e-05, "loss": 0.078, "step": 4680 }, { "grad_norm": 0.6451379060745239, "learning_rate": 9.997972216168217e-05, "loss": 0.0943, "step": 4690 }, { "grad_norm": 0.6088747382164001, "learning_rate": 9.997912931490771e-05, "loss": 0.0789, "step": 4700 }, { "grad_norm": 0.531538724899292, "learning_rate": 9.9978527928067e-05, "loss": 0.0771, "step": 4710 }, { "grad_norm": 0.434615820646286, "learning_rate": 9.997791800126277e-05, "loss": 0.0744, "step": 4720 }, { "grad_norm": 0.3629869818687439, "learning_rate": 9.997729953459927e-05, "loss": 0.0866, "step": 4730 }, { "grad_norm": 0.46210455894470215, "learning_rate": 9.997667252818214e-05, "loss": 0.0838, "step": 4740 }, { "grad_norm": 0.657033383846283, "learning_rate": 9.997603698211855e-05, "loss": 0.0736, "step": 4750 }, { "grad_norm": 0.6309747695922852, "learning_rate": 9.99753928965171e-05, "loss": 0.0893, "step": 4760 }, { "grad_norm": 0.47228509187698364, "learning_rate": 9.997474027148781e-05, "loss": 0.0872, "step": 4770 }, { "grad_norm": 0.8225560188293457, "learning_rate": 9.997407910714223e-05, "loss": 0.1022, "step": 4780 }, { "grad_norm": 0.5831339359283447, "learning_rate": 9.997340940359332e-05, "loss": 0.0824, "step": 4790 }, { "grad_norm": 0.5864866375923157, "learning_rate": 9.997273116095552e-05, "loss": 0.0814, "step": 4800 }, { "grad_norm": 0.4298469126224518, "learning_rate": 9.997204437934473e-05, "loss": 0.0823, "step": 4810 }, { "grad_norm": 0.5572091341018677, "learning_rate": 9.997134905887829e-05, "loss": 0.0946, "step": 4820 }, { "grad_norm": 0.48601824045181274, "learning_rate": 9.997064519967501e-05, "loss": 0.0907, "step": 4830 }, { "grad_norm": 0.669722855091095, "learning_rate": 9.996993280185517e-05, "loss": 0.0931, "step": 4840 }, { "grad_norm": 0.5567987561225891, "learning_rate": 9.99692118655405e-05, "loss": 0.0859, "step": 4850 }, { "grad_norm": 0.5270654559135437, "learning_rate": 9.996848239085417e-05, "loss": 0.0808, "step": 4860 }, { "grad_norm": 0.5638974905014038, "learning_rate": 9.996774437792085e-05, "loss": 0.0907, "step": 4870 }, { "grad_norm": 0.49142542481422424, "learning_rate": 9.996699782686664e-05, "loss": 0.0841, "step": 4880 }, { "grad_norm": 0.8461046814918518, "learning_rate": 9.996624273781909e-05, "loss": 0.0977, "step": 4890 }, { "grad_norm": 0.6340373158454895, "learning_rate": 9.996547911090725e-05, "loss": 0.072, "step": 4900 }, { "grad_norm": 0.43583932518959045, "learning_rate": 9.996470694626157e-05, "loss": 0.0848, "step": 4910 }, { "grad_norm": 0.541293203830719, "learning_rate": 9.996392624401403e-05, "loss": 0.092, "step": 4920 }, { "grad_norm": 0.5116671919822693, "learning_rate": 9.996313700429801e-05, "loss": 0.078, "step": 4930 }, { "grad_norm": 0.5400022864341736, "learning_rate": 9.996233922724836e-05, "loss": 0.0728, "step": 4940 }, { "grad_norm": 0.5716456174850464, "learning_rate": 9.996153291300141e-05, "loss": 0.0762, "step": 4950 }, { "grad_norm": 0.42475733160972595, "learning_rate": 9.996071806169494e-05, "loss": 0.0794, "step": 4960 }, { "grad_norm": 0.5726726651191711, "learning_rate": 9.995989467346817e-05, "loss": 0.088, "step": 4970 }, { "grad_norm": 0.38884237408638, "learning_rate": 9.995906274846183e-05, "loss": 0.0799, "step": 4980 }, { "grad_norm": 0.5371545553207397, "learning_rate": 9.995822228681803e-05, "loss": 0.0776, "step": 4990 }, { "grad_norm": 0.5012737512588501, "learning_rate": 9.99573732886804e-05, "loss": 0.0824, "step": 5000 }, { "grad_norm": 0.5204139947891235, "learning_rate": 9.995651575419402e-05, "loss": 0.0863, "step": 5010 }, { "grad_norm": 0.8084962368011475, "learning_rate": 9.995564968350541e-05, "loss": 0.0874, "step": 5020 }, { "grad_norm": 0.47391277551651, "learning_rate": 9.995477507676256e-05, "loss": 0.0787, "step": 5030 }, { "grad_norm": 0.5946065187454224, "learning_rate": 9.995389193411493e-05, "loss": 0.08, "step": 5040 }, { "grad_norm": 0.5285128355026245, "learning_rate": 9.995300025571339e-05, "loss": 0.0866, "step": 5050 }, { "grad_norm": 0.4374072253704071, "learning_rate": 9.995210004171034e-05, "loss": 0.0694, "step": 5060 }, { "grad_norm": 0.5473783612251282, "learning_rate": 9.995119129225956e-05, "loss": 0.1032, "step": 5070 }, { "grad_norm": 0.5987237691879272, "learning_rate": 9.995027400751637e-05, "loss": 0.0945, "step": 5080 }, { "grad_norm": 0.716157853603363, "learning_rate": 9.994934818763751e-05, "loss": 0.0781, "step": 5090 }, { "grad_norm": 0.6150620579719543, "learning_rate": 9.994841383278115e-05, "loss": 0.0729, "step": 5100 }, { "grad_norm": 0.6588375568389893, "learning_rate": 9.994747094310695e-05, "loss": 0.0825, "step": 5110 }, { "grad_norm": 0.5789560675621033, "learning_rate": 9.994651951877604e-05, "loss": 0.0973, "step": 5120 }, { "grad_norm": 0.5507475733757019, "learning_rate": 9.994555955995099e-05, "loss": 0.0925, "step": 5130 }, { "grad_norm": 0.5058655142784119, "learning_rate": 9.994459106679581e-05, "loss": 0.0816, "step": 5140 }, { "grad_norm": 0.45602911710739136, "learning_rate": 9.994361403947603e-05, "loss": 0.0758, "step": 5150 }, { "grad_norm": 0.5895761847496033, "learning_rate": 9.994262847815854e-05, "loss": 0.1057, "step": 5160 }, { "grad_norm": 0.48596590757369995, "learning_rate": 9.99416343830118e-05, "loss": 0.0846, "step": 5170 }, { "grad_norm": 0.6095079183578491, "learning_rate": 9.994063175420565e-05, "loss": 0.0727, "step": 5180 }, { "grad_norm": 0.5912063121795654, "learning_rate": 9.99396205919114e-05, "loss": 0.1032, "step": 5190 }, { "grad_norm": 0.7843122482299805, "learning_rate": 9.993860089630185e-05, "loss": 0.0697, "step": 5200 }, { "grad_norm": 0.5337680578231812, "learning_rate": 9.993757266755123e-05, "loss": 0.0875, "step": 5210 }, { "grad_norm": 0.4889732003211975, "learning_rate": 9.993653590583522e-05, "loss": 0.0784, "step": 5220 }, { "grad_norm": 0.4468344748020172, "learning_rate": 9.993549061133102e-05, "loss": 0.0726, "step": 5230 }, { "grad_norm": 0.5112206935882568, "learning_rate": 9.993443678421719e-05, "loss": 0.0741, "step": 5240 }, { "grad_norm": 0.49959343671798706, "learning_rate": 9.993337442467384e-05, "loss": 0.0917, "step": 5250 }, { "grad_norm": 0.6393494009971619, "learning_rate": 9.993230353288248e-05, "loss": 0.0788, "step": 5260 }, { "grad_norm": 0.5618520379066467, "learning_rate": 9.993122410902608e-05, "loss": 0.08, "step": 5270 }, { "grad_norm": 0.584959089756012, "learning_rate": 9.993013615328912e-05, "loss": 0.095, "step": 5280 }, { "grad_norm": 0.5268732905387878, "learning_rate": 9.992903966585747e-05, "loss": 0.0725, "step": 5290 }, { "grad_norm": 0.4745140075683594, "learning_rate": 9.992793464691852e-05, "loss": 0.0725, "step": 5300 }, { "grad_norm": 0.5246791243553162, "learning_rate": 9.992682109666105e-05, "loss": 0.0689, "step": 5310 }, { "grad_norm": 0.641170859336853, "learning_rate": 9.992569901527538e-05, "loss": 0.0735, "step": 5320 }, { "grad_norm": 0.5502227544784546, "learning_rate": 9.99245684029532e-05, "loss": 0.088, "step": 5330 }, { "grad_norm": 0.5642456412315369, "learning_rate": 9.992342925988774e-05, "loss": 0.0761, "step": 5340 }, { "grad_norm": 0.7248629927635193, "learning_rate": 9.992228158627361e-05, "loss": 0.0741, "step": 5350 }, { "grad_norm": 0.4263325333595276, "learning_rate": 9.992112538230693e-05, "loss": 0.0711, "step": 5360 }, { "grad_norm": 0.4350776672363281, "learning_rate": 9.991996064818527e-05, "loss": 0.0679, "step": 5370 }, { "grad_norm": 0.5506912469863892, "learning_rate": 9.991878738410768e-05, "loss": 0.0733, "step": 5380 }, { "grad_norm": 0.7301594018936157, "learning_rate": 9.991760559027457e-05, "loss": 0.0763, "step": 5390 }, { "grad_norm": 0.45027852058410645, "learning_rate": 9.991641526688793e-05, "loss": 0.0857, "step": 5400 }, { "grad_norm": 0.4788757860660553, "learning_rate": 9.991521641415113e-05, "loss": 0.0708, "step": 5410 }, { "grad_norm": 0.41393736004829407, "learning_rate": 9.991400903226904e-05, "loss": 0.0678, "step": 5420 }, { "grad_norm": 0.6515522599220276, "learning_rate": 9.991279312144794e-05, "loss": 0.0776, "step": 5430 }, { "grad_norm": 0.4541871249675751, "learning_rate": 9.991156868189564e-05, "loss": 0.0889, "step": 5440 }, { "grad_norm": 0.5893957018852234, "learning_rate": 9.991033571382131e-05, "loss": 0.0554, "step": 5450 }, { "grad_norm": 0.6253201961517334, "learning_rate": 9.990909421743569e-05, "loss": 0.0675, "step": 5460 }, { "grad_norm": 0.44191673398017883, "learning_rate": 9.990784419295085e-05, "loss": 0.0648, "step": 5470 }, { "grad_norm": 0.7284402847290039, "learning_rate": 9.990658564058044e-05, "loss": 0.0775, "step": 5480 }, { "grad_norm": 0.586794912815094, "learning_rate": 9.990531856053948e-05, "loss": 0.0644, "step": 5490 }, { "grad_norm": 0.6796877384185791, "learning_rate": 9.99040429530445e-05, "loss": 0.0727, "step": 5500 }, { "grad_norm": 0.5081077814102173, "learning_rate": 9.990275881831346e-05, "loss": 0.0767, "step": 5510 }, { "grad_norm": 0.44701218605041504, "learning_rate": 9.990146615656577e-05, "loss": 0.07, "step": 5520 }, { "grad_norm": 0.5687649250030518, "learning_rate": 9.990016496802233e-05, "loss": 0.0709, "step": 5530 }, { "grad_norm": 0.442732572555542, "learning_rate": 9.989885525290548e-05, "loss": 0.0602, "step": 5540 }, { "grad_norm": 0.48847028613090515, "learning_rate": 9.989753701143897e-05, "loss": 0.0638, "step": 5550 }, { "grad_norm": 0.5844025015830994, "learning_rate": 9.989621024384812e-05, "loss": 0.0745, "step": 5560 }, { "grad_norm": 0.49832648038864136, "learning_rate": 9.989487495035959e-05, "loss": 0.0676, "step": 5570 }, { "grad_norm": 0.6486583352088928, "learning_rate": 9.989353113120156e-05, "loss": 0.0915, "step": 5580 }, { "grad_norm": 0.5814209580421448, "learning_rate": 9.989217878660366e-05, "loss": 0.067, "step": 5590 }, { "grad_norm": 0.5083304643630981, "learning_rate": 9.989081791679695e-05, "loss": 0.0734, "step": 5600 }, { "grad_norm": 0.5369959473609924, "learning_rate": 9.988944852201397e-05, "loss": 0.0787, "step": 5610 }, { "grad_norm": 0.5771442651748657, "learning_rate": 9.988807060248873e-05, "loss": 0.0683, "step": 5620 }, { "grad_norm": 0.4718175232410431, "learning_rate": 9.988668415845665e-05, "loss": 0.0716, "step": 5630 }, { "grad_norm": 0.3627550005912781, "learning_rate": 9.988528919015466e-05, "loss": 0.0801, "step": 5640 }, { "grad_norm": 0.4617774784564972, "learning_rate": 9.988388569782112e-05, "loss": 0.084, "step": 5650 }, { "grad_norm": 0.5343621969223022, "learning_rate": 9.988247368169583e-05, "loss": 0.0752, "step": 5660 }, { "grad_norm": 0.5654984712600708, "learning_rate": 9.988105314202007e-05, "loss": 0.0719, "step": 5670 }, { "grad_norm": 0.4744129478931427, "learning_rate": 9.987962407903659e-05, "loss": 0.0829, "step": 5680 }, { "grad_norm": 0.4832727611064911, "learning_rate": 9.987818649298957e-05, "loss": 0.0734, "step": 5690 }, { "grad_norm": 0.4578443169593811, "learning_rate": 9.987674038412465e-05, "loss": 0.072, "step": 5700 }, { "grad_norm": 0.5139937400817871, "learning_rate": 9.987528575268891e-05, "loss": 0.076, "step": 5710 }, { "grad_norm": 0.47352954745292664, "learning_rate": 9.987382259893095e-05, "loss": 0.0654, "step": 5720 }, { "grad_norm": 0.5637082457542419, "learning_rate": 9.987235092310074e-05, "loss": 0.0696, "step": 5730 }, { "grad_norm": 0.5025337934494019, "learning_rate": 9.987087072544978e-05, "loss": 0.0832, "step": 5740 }, { "grad_norm": 0.6389997601509094, "learning_rate": 9.9869382006231e-05, "loss": 0.0738, "step": 5750 }, { "grad_norm": 0.6681616306304932, "learning_rate": 9.986788476569875e-05, "loss": 0.0845, "step": 5760 }, { "grad_norm": 0.4920739233493805, "learning_rate": 9.986637900410887e-05, "loss": 0.0778, "step": 5770 }, { "grad_norm": 0.5525789856910706, "learning_rate": 9.986486472171869e-05, "loss": 0.0913, "step": 5780 }, { "grad_norm": 0.49310189485549927, "learning_rate": 9.986334191878692e-05, "loss": 0.072, "step": 5790 }, { "grad_norm": 0.5177735090255737, "learning_rate": 9.986181059557378e-05, "loss": 0.0885, "step": 5800 }, { "grad_norm": 0.47781679034233093, "learning_rate": 9.986027075234094e-05, "loss": 0.0793, "step": 5810 }, { "grad_norm": 0.6172356605529785, "learning_rate": 9.985872238935152e-05, "loss": 0.0737, "step": 5820 }, { "grad_norm": 0.509569525718689, "learning_rate": 9.985716550687008e-05, "loss": 0.082, "step": 5830 }, { "grad_norm": 0.4610617160797119, "learning_rate": 9.985560010516264e-05, "loss": 0.0706, "step": 5840 }, { "grad_norm": 0.5688307285308838, "learning_rate": 9.985402618449668e-05, "loss": 0.0703, "step": 5850 }, { "grad_norm": 0.463290810585022, "learning_rate": 9.985244374514118e-05, "loss": 0.0858, "step": 5860 }, { "grad_norm": 0.6022486686706543, "learning_rate": 9.985085278736651e-05, "loss": 0.0762, "step": 5870 }, { "grad_norm": 0.6292734146118164, "learning_rate": 9.984925331144452e-05, "loss": 0.0845, "step": 5880 }, { "grad_norm": 0.5954769253730774, "learning_rate": 9.984764531764851e-05, "loss": 0.0766, "step": 5890 }, { "grad_norm": 0.546191930770874, "learning_rate": 9.984602880625326e-05, "loss": 0.0902, "step": 5900 }, { "grad_norm": 0.4542558193206787, "learning_rate": 9.9844403777535e-05, "loss": 0.0723, "step": 5910 }, { "grad_norm": 0.5400745868682861, "learning_rate": 9.984277023177135e-05, "loss": 0.0748, "step": 5920 }, { "grad_norm": 0.7082663774490356, "learning_rate": 9.984112816924148e-05, "loss": 0.0671, "step": 5930 }, { "grad_norm": 0.5205890536308289, "learning_rate": 9.983947759022596e-05, "loss": 0.0787, "step": 5940 }, { "grad_norm": 0.7037428021430969, "learning_rate": 9.983781849500682e-05, "loss": 0.0781, "step": 5950 }, { "grad_norm": 0.4560556411743164, "learning_rate": 9.98361508838676e-05, "loss": 0.0692, "step": 5960 }, { "grad_norm": 0.6069479584693909, "learning_rate": 9.98344747570932e-05, "loss": 0.0777, "step": 5970 }, { "grad_norm": 0.6198562979698181, "learning_rate": 9.983279011497004e-05, "loss": 0.081, "step": 5980 }, { "grad_norm": 0.5022843480110168, "learning_rate": 9.983109695778596e-05, "loss": 0.0851, "step": 5990 }, { "grad_norm": 0.5292416214942932, "learning_rate": 9.982939528583032e-05, "loss": 0.0981, "step": 6000 }, { "grad_norm": 0.4370587170124054, "learning_rate": 9.982768509939385e-05, "loss": 0.0768, "step": 6010 }, { "grad_norm": 0.732629120349884, "learning_rate": 9.982596639876879e-05, "loss": 0.0707, "step": 6020 }, { "grad_norm": 0.5164657831192017, "learning_rate": 9.982423918424881e-05, "loss": 0.0814, "step": 6030 }, { "grad_norm": 0.5695842504501343, "learning_rate": 9.982250345612908e-05, "loss": 0.07, "step": 6040 }, { "grad_norm": 0.5473247170448303, "learning_rate": 9.982075921470611e-05, "loss": 0.0805, "step": 6050 }, { "grad_norm": 0.48645853996276855, "learning_rate": 9.981900646027802e-05, "loss": 0.0665, "step": 6060 }, { "grad_norm": 0.4421829581260681, "learning_rate": 9.981724519314425e-05, "loss": 0.0694, "step": 6070 }, { "grad_norm": 0.5293723940849304, "learning_rate": 9.981547541360581e-05, "loss": 0.0752, "step": 6080 }, { "grad_norm": 0.44883397221565247, "learning_rate": 9.981369712196508e-05, "loss": 0.0791, "step": 6090 }, { "grad_norm": 0.6710265278816223, "learning_rate": 9.981191031852592e-05, "loss": 0.0891, "step": 6100 }, { "grad_norm": 0.5190449357032776, "learning_rate": 9.981011500359362e-05, "loss": 0.0659, "step": 6110 }, { "grad_norm": 0.47535139322280884, "learning_rate": 9.9808311177475e-05, "loss": 0.0712, "step": 6120 }, { "grad_norm": 0.46057409048080444, "learning_rate": 9.980649884047826e-05, "loss": 0.0628, "step": 6130 }, { "grad_norm": 0.5029507279396057, "learning_rate": 9.980467799291307e-05, "loss": 0.0706, "step": 6140 }, { "grad_norm": 0.44730356335639954, "learning_rate": 9.980284863509058e-05, "loss": 0.0845, "step": 6150 }, { "grad_norm": 0.6177793145179749, "learning_rate": 9.980101076732334e-05, "loss": 0.0938, "step": 6160 }, { "grad_norm": 0.5602224469184875, "learning_rate": 9.979916438992544e-05, "loss": 0.0912, "step": 6170 }, { "grad_norm": 0.5715076327323914, "learning_rate": 9.979730950321237e-05, "loss": 0.0699, "step": 6180 }, { "grad_norm": 0.5572122931480408, "learning_rate": 9.979544610750104e-05, "loss": 0.0896, "step": 6190 }, { "grad_norm": 0.5672398805618286, "learning_rate": 9.97935742031099e-05, "loss": 0.0771, "step": 6200 }, { "grad_norm": 0.6285523176193237, "learning_rate": 9.979169379035878e-05, "loss": 0.0811, "step": 6210 }, { "grad_norm": 0.536139726638794, "learning_rate": 9.978980486956899e-05, "loss": 0.064, "step": 6220 }, { "grad_norm": 0.3833799362182617, "learning_rate": 9.978790744106332e-05, "loss": 0.0723, "step": 6230 }, { "grad_norm": 0.560405433177948, "learning_rate": 9.978600150516594e-05, "loss": 0.0695, "step": 6240 }, { "grad_norm": 0.6019940376281738, "learning_rate": 9.978408706220259e-05, "loss": 0.0751, "step": 6250 }, { "grad_norm": 0.30245646834373474, "learning_rate": 9.978216411250032e-05, "loss": 0.0574, "step": 6260 }, { "grad_norm": 0.4767233729362488, "learning_rate": 9.978023265638778e-05, "loss": 0.0607, "step": 6270 }, { "grad_norm": 0.7099765539169312, "learning_rate": 9.977829269419495e-05, "loss": 0.0693, "step": 6280 }, { "grad_norm": 0.5014215707778931, "learning_rate": 9.977634422625335e-05, "loss": 0.0676, "step": 6290 }, { "grad_norm": 0.5026882886886597, "learning_rate": 9.97743872528959e-05, "loss": 0.0729, "step": 6300 }, { "grad_norm": 0.44441261887550354, "learning_rate": 9.9772421774457e-05, "loss": 0.0632, "step": 6310 }, { "grad_norm": 0.62713623046875, "learning_rate": 9.977044779127252e-05, "loss": 0.0657, "step": 6320 }, { "grad_norm": 0.5576682686805725, "learning_rate": 9.976846530367971e-05, "loss": 0.0707, "step": 6330 }, { "grad_norm": 0.5727444887161255, "learning_rate": 9.976647431201735e-05, "loss": 0.0887, "step": 6340 }, { "grad_norm": 0.6435531377792358, "learning_rate": 9.976447481662568e-05, "loss": 0.0729, "step": 6350 }, { "grad_norm": 0.5089086294174194, "learning_rate": 9.976246681784629e-05, "loss": 0.0679, "step": 6360 }, { "grad_norm": 0.5029277801513672, "learning_rate": 9.976045031602234e-05, "loss": 0.0561, "step": 6370 }, { "grad_norm": 0.42792779207229614, "learning_rate": 9.975842531149837e-05, "loss": 0.0611, "step": 6380 }, { "grad_norm": 0.6331509947776794, "learning_rate": 9.975639180462043e-05, "loss": 0.0735, "step": 6390 }, { "grad_norm": 0.5917182564735413, "learning_rate": 9.975434979573596e-05, "loss": 0.0773, "step": 6400 }, { "grad_norm": 0.44716015458106995, "learning_rate": 9.97522992851939e-05, "loss": 0.0565, "step": 6410 }, { "grad_norm": 0.5291863083839417, "learning_rate": 9.975024027334461e-05, "loss": 0.0693, "step": 6420 }, { "grad_norm": 0.43258893489837646, "learning_rate": 9.974817276053993e-05, "loss": 0.0698, "step": 6430 }, { "grad_norm": 0.46597152948379517, "learning_rate": 9.974609674713315e-05, "loss": 0.0653, "step": 6440 }, { "grad_norm": 0.5337729454040527, "learning_rate": 9.9744012233479e-05, "loss": 0.0596, "step": 6450 }, { "grad_norm": 0.43993350863456726, "learning_rate": 9.974191921993366e-05, "loss": 0.0632, "step": 6460 }, { "grad_norm": 0.43363407254219055, "learning_rate": 9.973981770685474e-05, "loss": 0.0639, "step": 6470 }, { "grad_norm": 0.5119966864585876, "learning_rate": 9.97377076946014e-05, "loss": 0.0644, "step": 6480 }, { "grad_norm": 0.6266461610794067, "learning_rate": 9.973558918353412e-05, "loss": 0.073, "step": 6490 }, { "grad_norm": 0.5414121747016907, "learning_rate": 9.973346217401494e-05, "loss": 0.0622, "step": 6500 }, { "grad_norm": 0.44482892751693726, "learning_rate": 9.973132666640726e-05, "loss": 0.0617, "step": 6510 }, { "grad_norm": 0.5286384224891663, "learning_rate": 9.972918266107602e-05, "loss": 0.0689, "step": 6520 }, { "grad_norm": 0.5695348978042603, "learning_rate": 9.972703015838756e-05, "loss": 0.0669, "step": 6530 }, { "grad_norm": 0.4239356815814972, "learning_rate": 9.97248691587097e-05, "loss": 0.0767, "step": 6540 }, { "grad_norm": 0.6522865295410156, "learning_rate": 9.972269966241166e-05, "loss": 0.0686, "step": 6550 }, { "grad_norm": 0.6588292717933655, "learning_rate": 9.972052166986417e-05, "loss": 0.0707, "step": 6560 }, { "grad_norm": 0.5377855896949768, "learning_rate": 9.971833518143938e-05, "loss": 0.0753, "step": 6570 }, { "grad_norm": 0.44505342841148376, "learning_rate": 9.971614019751093e-05, "loss": 0.0588, "step": 6580 }, { "grad_norm": 0.54701167345047, "learning_rate": 9.971393671845383e-05, "loss": 0.075, "step": 6590 }, { "grad_norm": 0.527596116065979, "learning_rate": 9.971172474464464e-05, "loss": 0.0752, "step": 6600 }, { "grad_norm": 0.453477144241333, "learning_rate": 9.97095042764613e-05, "loss": 0.0594, "step": 6610 }, { "grad_norm": 0.4244938790798187, "learning_rate": 9.970727531428324e-05, "loss": 0.0563, "step": 6620 }, { "grad_norm": 0.5544337630271912, "learning_rate": 9.970503785849132e-05, "loss": 0.0647, "step": 6630 }, { "grad_norm": 0.49020254611968994, "learning_rate": 9.970279190946788e-05, "loss": 0.0595, "step": 6640 }, { "grad_norm": 0.6523012518882751, "learning_rate": 9.970053746759667e-05, "loss": 0.0795, "step": 6650 }, { "grad_norm": 0.5250698924064636, "learning_rate": 9.969827453326292e-05, "loss": 0.0684, "step": 6660 }, { "grad_norm": 0.5076116323471069, "learning_rate": 9.969600310685332e-05, "loss": 0.0656, "step": 6670 }, { "grad_norm": 0.6399413347244263, "learning_rate": 9.969372318875596e-05, "loss": 0.0796, "step": 6680 }, { "grad_norm": 0.48697221279144287, "learning_rate": 9.969143477936043e-05, "loss": 0.0712, "step": 6690 }, { "grad_norm": 0.5766555070877075, "learning_rate": 9.968913787905775e-05, "loss": 0.0733, "step": 6700 }, { "grad_norm": 0.703494668006897, "learning_rate": 9.968683248824045e-05, "loss": 0.0702, "step": 6710 }, { "grad_norm": 0.3872264325618744, "learning_rate": 9.968451860730238e-05, "loss": 0.0811, "step": 6720 }, { "grad_norm": 0.4199783504009247, "learning_rate": 9.968219623663896e-05, "loss": 0.0676, "step": 6730 }, { "grad_norm": 0.45422613620758057, "learning_rate": 9.967986537664702e-05, "loss": 0.0919, "step": 6740 }, { "grad_norm": 0.599388599395752, "learning_rate": 9.967752602772483e-05, "loss": 0.0629, "step": 6750 }, { "grad_norm": 0.40662965178489685, "learning_rate": 9.967517819027212e-05, "loss": 0.0696, "step": 6760 }, { "grad_norm": 0.5443575978279114, "learning_rate": 9.967282186469009e-05, "loss": 0.0637, "step": 6770 }, { "grad_norm": 0.675692081451416, "learning_rate": 9.967045705138135e-05, "loss": 0.0793, "step": 6780 }, { "grad_norm": 0.6569861769676208, "learning_rate": 9.966808375074998e-05, "loss": 0.0719, "step": 6790 }, { "grad_norm": 0.5583736896514893, "learning_rate": 9.966570196320154e-05, "loss": 0.0867, "step": 6800 }, { "grad_norm": 0.5363962054252625, "learning_rate": 9.966331168914299e-05, "loss": 0.0584, "step": 6810 }, { "grad_norm": 0.7393838763237, "learning_rate": 9.966091292898277e-05, "loss": 0.0814, "step": 6820 }, { "grad_norm": 0.6039823293685913, "learning_rate": 9.965850568313076e-05, "loss": 0.0651, "step": 6830 }, { "grad_norm": 0.7039910554885864, "learning_rate": 9.965608995199827e-05, "loss": 0.0788, "step": 6840 }, { "grad_norm": 0.4252396523952484, "learning_rate": 9.965366573599812e-05, "loss": 0.0648, "step": 6850 }, { "grad_norm": 0.6517411470413208, "learning_rate": 9.965123303554453e-05, "loss": 0.0657, "step": 6860 }, { "grad_norm": 0.5993431806564331, "learning_rate": 9.964879185105317e-05, "loss": 0.0932, "step": 6870 }, { "grad_norm": 0.4742625057697296, "learning_rate": 9.964634218294119e-05, "loss": 0.0815, "step": 6880 }, { "grad_norm": 0.5321409702301025, "learning_rate": 9.964388403162714e-05, "loss": 0.0697, "step": 6890 }, { "grad_norm": 0.7384690046310425, "learning_rate": 9.96414173975311e-05, "loss": 0.0815, "step": 6900 }, { "grad_norm": 0.3786472976207733, "learning_rate": 9.963894228107451e-05, "loss": 0.0705, "step": 6910 }, { "grad_norm": 0.41563448309898376, "learning_rate": 9.963645868268032e-05, "loss": 0.0728, "step": 6920 }, { "grad_norm": 0.48719727993011475, "learning_rate": 9.963396660277289e-05, "loss": 0.0733, "step": 6930 }, { "grad_norm": 0.49982085824012756, "learning_rate": 9.963146604177807e-05, "loss": 0.0565, "step": 6940 }, { "grad_norm": 0.43706533312797546, "learning_rate": 9.962895700012311e-05, "loss": 0.0622, "step": 6950 }, { "grad_norm": 0.37476569414138794, "learning_rate": 9.962643947823677e-05, "loss": 0.0586, "step": 6960 }, { "grad_norm": 0.4607299864292145, "learning_rate": 9.962391347654921e-05, "loss": 0.0735, "step": 6970 }, { "grad_norm": 0.5145196318626404, "learning_rate": 9.962137899549204e-05, "loss": 0.0813, "step": 6980 }, { "grad_norm": 0.4408886134624481, "learning_rate": 9.961883603549835e-05, "loss": 0.0574, "step": 6990 }, { "grad_norm": 0.5569041967391968, "learning_rate": 9.961628459700267e-05, "loss": 0.0747, "step": 7000 }, { "grad_norm": 0.5446930527687073, "learning_rate": 9.961372468044095e-05, "loss": 0.0801, "step": 7010 }, { "grad_norm": 0.5529423952102661, "learning_rate": 9.961115628625062e-05, "loss": 0.0621, "step": 7020 }, { "grad_norm": 0.468223512172699, "learning_rate": 9.960857941487056e-05, "loss": 0.088, "step": 7030 }, { "grad_norm": 0.5051165819168091, "learning_rate": 9.960599406674106e-05, "loss": 0.0598, "step": 7040 }, { "grad_norm": 0.6131184101104736, "learning_rate": 9.960340024230393e-05, "loss": 0.0553, "step": 7050 }, { "grad_norm": 0.44106611609458923, "learning_rate": 9.960079794200232e-05, "loss": 0.0604, "step": 7060 }, { "grad_norm": 0.3928908705711365, "learning_rate": 9.959818716628096e-05, "loss": 0.063, "step": 7070 }, { "grad_norm": 0.3982239365577698, "learning_rate": 9.95955679155859e-05, "loss": 0.0912, "step": 7080 }, { "grad_norm": 0.8146026730537415, "learning_rate": 9.959294019036472e-05, "loss": 0.0776, "step": 7090 }, { "grad_norm": 0.6021227836608887, "learning_rate": 9.959030399106646e-05, "loss": 0.0633, "step": 7100 }, { "grad_norm": 0.44157716631889343, "learning_rate": 9.958765931814153e-05, "loss": 0.0779, "step": 7110 }, { "grad_norm": 0.4046730399131775, "learning_rate": 9.958500617204184e-05, "loss": 0.0575, "step": 7120 }, { "grad_norm": 0.52311772108078, "learning_rate": 9.958234455322075e-05, "loss": 0.0671, "step": 7130 }, { "grad_norm": 0.6020161509513855, "learning_rate": 9.957967446213308e-05, "loss": 0.0786, "step": 7140 }, { "grad_norm": 0.5623763799667358, "learning_rate": 9.957699589923501e-05, "loss": 0.0817, "step": 7150 }, { "grad_norm": 0.6758502125740051, "learning_rate": 9.957430886498431e-05, "loss": 0.0723, "step": 7160 }, { "grad_norm": 0.4276138246059418, "learning_rate": 9.957161335984008e-05, "loss": 0.0712, "step": 7170 }, { "grad_norm": 0.5295581817626953, "learning_rate": 9.956890938426291e-05, "loss": 0.0707, "step": 7180 }, { "grad_norm": 0.4468136727809906, "learning_rate": 9.956619693871482e-05, "loss": 0.0711, "step": 7190 }, { "grad_norm": 0.4902241826057434, "learning_rate": 9.956347602365934e-05, "loss": 0.0648, "step": 7200 }, { "grad_norm": 0.39685526490211487, "learning_rate": 9.956074663956135e-05, "loss": 0.0609, "step": 7210 }, { "grad_norm": 0.6127735376358032, "learning_rate": 9.955800878688726e-05, "loss": 0.0655, "step": 7220 }, { "grad_norm": 0.41581276059150696, "learning_rate": 9.955526246610489e-05, "loss": 0.0613, "step": 7230 }, { "grad_norm": 0.6451886296272278, "learning_rate": 9.955250767768349e-05, "loss": 0.067, "step": 7240 }, { "grad_norm": 0.4269467890262604, "learning_rate": 9.95497444220938e-05, "loss": 0.0724, "step": 7250 }, { "grad_norm": 0.439544677734375, "learning_rate": 9.954697269980797e-05, "loss": 0.057, "step": 7260 }, { "grad_norm": 0.3689979314804077, "learning_rate": 9.954419251129962e-05, "loss": 0.0559, "step": 7270 }, { "grad_norm": 0.5276520848274231, "learning_rate": 9.95414038570438e-05, "loss": 0.06, "step": 7280 }, { "grad_norm": 0.5722265839576721, "learning_rate": 9.953860673751703e-05, "loss": 0.0662, "step": 7290 }, { "grad_norm": 0.3625737428665161, "learning_rate": 9.953580115319725e-05, "loss": 0.062, "step": 7300 }, { "grad_norm": 0.4630908966064453, "learning_rate": 9.953298710456387e-05, "loss": 0.0618, "step": 7310 }, { "grad_norm": 0.5401329398155212, "learning_rate": 9.953016459209771e-05, "loss": 0.054, "step": 7320 }, { "grad_norm": 0.638853132724762, "learning_rate": 9.952733361628108e-05, "loss": 0.0627, "step": 7330 }, { "grad_norm": 0.5337995886802673, "learning_rate": 9.952449417759772e-05, "loss": 0.075, "step": 7340 }, { "grad_norm": 0.43395620584487915, "learning_rate": 9.952164627653279e-05, "loss": 0.0628, "step": 7350 }, { "grad_norm": 0.5646661520004272, "learning_rate": 9.951878991357292e-05, "loss": 0.0615, "step": 7360 }, { "grad_norm": 0.3716937303543091, "learning_rate": 9.951592508920622e-05, "loss": 0.0705, "step": 7370 }, { "grad_norm": 0.4022165536880493, "learning_rate": 9.951305180392219e-05, "loss": 0.0546, "step": 7380 }, { "grad_norm": 0.4523867666721344, "learning_rate": 9.951017005821178e-05, "loss": 0.0608, "step": 7390 }, { "grad_norm": 0.3953218162059784, "learning_rate": 9.95072798525674e-05, "loss": 0.0598, "step": 7400 }, { "grad_norm": 0.524406373500824, "learning_rate": 9.950438118748293e-05, "loss": 0.0642, "step": 7410 }, { "grad_norm": 0.6465882062911987, "learning_rate": 9.950147406345366e-05, "loss": 0.0747, "step": 7420 }, { "grad_norm": 0.7912021279335022, "learning_rate": 9.949855848097635e-05, "loss": 0.0668, "step": 7430 }, { "grad_norm": 0.42495548725128174, "learning_rate": 9.949563444054916e-05, "loss": 0.0636, "step": 7440 }, { "grad_norm": 0.5707964897155762, "learning_rate": 9.949270194267178e-05, "loss": 0.0583, "step": 7450 }, { "grad_norm": 0.4841402471065521, "learning_rate": 9.948976098784526e-05, "loss": 0.0618, "step": 7460 }, { "grad_norm": 0.4177761673927307, "learning_rate": 9.948681157657213e-05, "loss": 0.066, "step": 7470 }, { "grad_norm": 0.5400902628898621, "learning_rate": 9.948385370935638e-05, "loss": 0.0713, "step": 7480 }, { "grad_norm": 0.5477775931358337, "learning_rate": 9.94808873867034e-05, "loss": 0.061, "step": 7490 }, { "grad_norm": 0.4944971799850464, "learning_rate": 9.947791260912009e-05, "loss": 0.0571, "step": 7500 }, { "grad_norm": 0.5852222442626953, "learning_rate": 9.947492937711474e-05, "loss": 0.0587, "step": 7510 }, { "grad_norm": 0.48948585987091064, "learning_rate": 9.947193769119707e-05, "loss": 0.0521, "step": 7520 }, { "grad_norm": 0.4256834089756012, "learning_rate": 9.946893755187834e-05, "loss": 0.0536, "step": 7530 }, { "grad_norm": 0.3768213987350464, "learning_rate": 9.946592895967115e-05, "loss": 0.0633, "step": 7540 }, { "grad_norm": 0.47808757424354553, "learning_rate": 9.94629119150896e-05, "loss": 0.0628, "step": 7550 }, { "grad_norm": 0.48061680793762207, "learning_rate": 9.94598864186492e-05, "loss": 0.0589, "step": 7560 }, { "grad_norm": 0.44125881791114807, "learning_rate": 9.945685247086696e-05, "loss": 0.0622, "step": 7570 }, { "grad_norm": 0.5013756155967712, "learning_rate": 9.945381007226129e-05, "loss": 0.075, "step": 7580 }, { "grad_norm": 0.42641201615333557, "learning_rate": 9.945075922335203e-05, "loss": 0.0693, "step": 7590 }, { "grad_norm": 0.42772185802459717, "learning_rate": 9.944769992466049e-05, "loss": 0.0837, "step": 7600 }, { "grad_norm": 0.4323573410511017, "learning_rate": 9.944463217670945e-05, "loss": 0.0709, "step": 7610 }, { "grad_norm": 0.3500499427318573, "learning_rate": 9.944155598002307e-05, "loss": 0.0641, "step": 7620 }, { "grad_norm": 0.36832502484321594, "learning_rate": 9.943847133512701e-05, "loss": 0.066, "step": 7630 }, { "grad_norm": 0.5067250728607178, "learning_rate": 9.943537824254834e-05, "loss": 0.0611, "step": 7640 }, { "grad_norm": 0.461847722530365, "learning_rate": 9.943227670281559e-05, "loss": 0.0571, "step": 7650 }, { "grad_norm": 0.439995676279068, "learning_rate": 9.942916671645873e-05, "loss": 0.0607, "step": 7660 }, { "grad_norm": 0.5223570466041565, "learning_rate": 9.942604828400916e-05, "loss": 0.067, "step": 7670 }, { "grad_norm": 0.6113464832305908, "learning_rate": 9.942292140599975e-05, "loss": 0.0745, "step": 7680 }, { "grad_norm": 0.45780158042907715, "learning_rate": 9.94197860829648e-05, "loss": 0.0633, "step": 7690 }, { "grad_norm": 0.44215044379234314, "learning_rate": 9.941664231544004e-05, "loss": 0.0533, "step": 7700 }, { "grad_norm": 0.43736228346824646, "learning_rate": 9.941349010396264e-05, "loss": 0.0753, "step": 7710 }, { "grad_norm": 0.42800211906433105, "learning_rate": 9.941032944907125e-05, "loss": 0.0694, "step": 7720 }, { "grad_norm": 0.3939782679080963, "learning_rate": 9.940716035130596e-05, "loss": 0.0644, "step": 7730 }, { "grad_norm": 0.5622648596763611, "learning_rate": 9.940398281120821e-05, "loss": 0.0672, "step": 7740 }, { "grad_norm": 0.47955086827278137, "learning_rate": 9.940079682932102e-05, "loss": 0.0595, "step": 7750 }, { "grad_norm": 0.4686276614665985, "learning_rate": 9.939760240618877e-05, "loss": 0.0672, "step": 7760 }, { "grad_norm": 0.4772428870201111, "learning_rate": 9.939439954235729e-05, "loss": 0.067, "step": 7770 }, { "grad_norm": 0.45899879932403564, "learning_rate": 9.939118823837387e-05, "loss": 0.07, "step": 7780 }, { "grad_norm": 0.5096253156661987, "learning_rate": 9.938796849478725e-05, "loss": 0.0625, "step": 7790 }, { "grad_norm": 0.5427016615867615, "learning_rate": 9.938474031214755e-05, "loss": 0.0565, "step": 7800 }, { "grad_norm": 0.32928094267845154, "learning_rate": 9.938150369100643e-05, "loss": 0.0636, "step": 7810 }, { "grad_norm": 0.3956354260444641, "learning_rate": 9.93782586319169e-05, "loss": 0.0506, "step": 7820 }, { "grad_norm": 0.4345318377017975, "learning_rate": 9.937500513543348e-05, "loss": 0.0554, "step": 7830 }, { "grad_norm": 0.46664732694625854, "learning_rate": 9.937174320211207e-05, "loss": 0.0507, "step": 7840 }, { "grad_norm": 0.39379948377609253, "learning_rate": 9.936847283251009e-05, "loss": 0.0554, "step": 7850 }, { "grad_norm": 0.47593116760253906, "learning_rate": 9.936519402718632e-05, "loss": 0.0704, "step": 7860 }, { "grad_norm": 0.5252768397331238, "learning_rate": 9.936190678670102e-05, "loss": 0.0723, "step": 7870 }, { "grad_norm": 0.5388329029083252, "learning_rate": 9.935861111161593e-05, "loss": 0.07, "step": 7880 }, { "grad_norm": 0.4319884777069092, "learning_rate": 9.935530700249416e-05, "loss": 0.0565, "step": 7890 }, { "grad_norm": 0.47093483805656433, "learning_rate": 9.935199445990028e-05, "loss": 0.0709, "step": 7900 }, { "grad_norm": 0.4184587299823761, "learning_rate": 9.934867348440033e-05, "loss": 0.0687, "step": 7910 }, { "grad_norm": 0.4491914212703705, "learning_rate": 9.934534407656176e-05, "loss": 0.0814, "step": 7920 }, { "grad_norm": 0.45227494835853577, "learning_rate": 9.93420062369535e-05, "loss": 0.0606, "step": 7930 }, { "grad_norm": 0.47251078486442566, "learning_rate": 9.933865996614589e-05, "loss": 0.0697, "step": 7940 }, { "grad_norm": 0.44612911343574524, "learning_rate": 9.933530526471068e-05, "loss": 0.0611, "step": 7950 }, { "grad_norm": 0.3449079990386963, "learning_rate": 9.933194213322114e-05, "loss": 0.0594, "step": 7960 }, { "grad_norm": 0.531134843826294, "learning_rate": 9.932857057225192e-05, "loss": 0.0545, "step": 7970 }, { "grad_norm": 0.5288393497467041, "learning_rate": 9.932519058237912e-05, "loss": 0.0676, "step": 7980 }, { "grad_norm": 0.5131773948669434, "learning_rate": 9.932180216418032e-05, "loss": 0.0589, "step": 7990 }, { "grad_norm": 0.5218172669410706, "learning_rate": 9.931840531823446e-05, "loss": 0.0562, "step": 8000 }, { "grad_norm": 0.3923073709011078, "learning_rate": 9.9315000045122e-05, "loss": 0.0554, "step": 8010 }, { "grad_norm": 0.5288413763046265, "learning_rate": 9.931158634542481e-05, "loss": 0.0653, "step": 8020 }, { "grad_norm": 0.48843270540237427, "learning_rate": 9.930816421972617e-05, "loss": 0.0603, "step": 8030 }, { "grad_norm": 0.47065499424934387, "learning_rate": 9.930473366861086e-05, "loss": 0.0613, "step": 8040 }, { "grad_norm": 0.5197038650512695, "learning_rate": 9.930129469266505e-05, "loss": 0.0604, "step": 8050 }, { "grad_norm": 0.453200101852417, "learning_rate": 9.929784729247638e-05, "loss": 0.0505, "step": 8060 }, { "grad_norm": 0.43006375432014465, "learning_rate": 9.929439146863389e-05, "loss": 0.0556, "step": 8070 }, { "grad_norm": 0.4926486015319824, "learning_rate": 9.92909272217281e-05, "loss": 0.0694, "step": 8080 }, { "grad_norm": 0.42049747705459595, "learning_rate": 9.928745455235097e-05, "loss": 0.0655, "step": 8090 }, { "grad_norm": 0.42021095752716064, "learning_rate": 9.928397346109588e-05, "loss": 0.0666, "step": 8100 }, { "grad_norm": 0.4836280345916748, "learning_rate": 9.928048394855762e-05, "loss": 0.0706, "step": 8110 }, { "grad_norm": 0.36174941062927246, "learning_rate": 9.92769860153325e-05, "loss": 0.0599, "step": 8120 }, { "grad_norm": 0.4855441749095917, "learning_rate": 9.927347966201819e-05, "loss": 0.0774, "step": 8130 }, { "grad_norm": 0.46640151739120483, "learning_rate": 9.926996488921383e-05, "loss": 0.0673, "step": 8140 }, { "grad_norm": 0.5571258664131165, "learning_rate": 9.926644169752001e-05, "loss": 0.0627, "step": 8150 }, { "grad_norm": 0.38499385118484497, "learning_rate": 9.926291008753875e-05, "loss": 0.0719, "step": 8160 }, { "grad_norm": 0.5366449952125549, "learning_rate": 9.92593700598735e-05, "loss": 0.0716, "step": 8170 }, { "grad_norm": 0.37004023790359497, "learning_rate": 9.925582161512915e-05, "loss": 0.0534, "step": 8180 }, { "grad_norm": 0.4097578823566437, "learning_rate": 9.925226475391205e-05, "loss": 0.0591, "step": 8190 }, { "grad_norm": 0.5100091099739075, "learning_rate": 9.924869947682993e-05, "loss": 0.065, "step": 8200 }, { "grad_norm": 0.45731642842292786, "learning_rate": 9.924512578449204e-05, "loss": 0.0484, "step": 8210 }, { "grad_norm": 0.4205667972564697, "learning_rate": 9.924154367750901e-05, "loss": 0.0573, "step": 8220 }, { "grad_norm": 0.4803036153316498, "learning_rate": 9.923795315649293e-05, "loss": 0.0636, "step": 8230 }, { "grad_norm": 0.2887707054615021, "learning_rate": 9.92343542220573e-05, "loss": 0.0588, "step": 8240 }, { "grad_norm": 0.3687312602996826, "learning_rate": 9.92307468748171e-05, "loss": 0.0553, "step": 8250 }, { "grad_norm": 0.47687849402427673, "learning_rate": 9.922713111538873e-05, "loss": 0.0528, "step": 8260 }, { "grad_norm": 0.4092696011066437, "learning_rate": 9.922350694439003e-05, "loss": 0.0552, "step": 8270 }, { "grad_norm": 0.4059515595436096, "learning_rate": 9.921987436244024e-05, "loss": 0.0608, "step": 8280 }, { "grad_norm": 0.5099157094955444, "learning_rate": 9.921623337016008e-05, "loss": 0.0498, "step": 8290 }, { "grad_norm": 0.5113776326179504, "learning_rate": 9.921258396817172e-05, "loss": 0.0646, "step": 8300 }, { "grad_norm": 0.5002313256263733, "learning_rate": 9.920892615709874e-05, "loss": 0.0599, "step": 8310 }, { "grad_norm": 0.5032743215560913, "learning_rate": 9.920525993756612e-05, "loss": 0.051, "step": 8320 }, { "grad_norm": 0.5051451325416565, "learning_rate": 9.920158531020036e-05, "loss": 0.058, "step": 8330 }, { "grad_norm": 0.47269952297210693, "learning_rate": 9.919790227562933e-05, "loss": 0.0541, "step": 8340 }, { "grad_norm": 0.43650004267692566, "learning_rate": 9.919421083448237e-05, "loss": 0.0643, "step": 8350 }, { "grad_norm": 0.4929603636264801, "learning_rate": 9.919051098739022e-05, "loss": 0.0534, "step": 8360 }, { "grad_norm": 0.48525750637054443, "learning_rate": 9.918680273498514e-05, "loss": 0.0569, "step": 8370 }, { "grad_norm": 0.5295612812042236, "learning_rate": 9.918308607790072e-05, "loss": 0.0681, "step": 8380 }, { "grad_norm": 0.5033278465270996, "learning_rate": 9.917936101677205e-05, "loss": 0.0611, "step": 8390 }, { "grad_norm": 0.47828206419944763, "learning_rate": 9.917562755223564e-05, "loss": 0.0622, "step": 8400 }, { "grad_norm": 0.5523513555526733, "learning_rate": 9.917188568492944e-05, "loss": 0.059, "step": 8410 }, { "grad_norm": 0.4805269241333008, "learning_rate": 9.916813541549283e-05, "loss": 0.0598, "step": 8420 }, { "grad_norm": 0.5504578948020935, "learning_rate": 9.916437674456663e-05, "loss": 0.0607, "step": 8430 }, { "grad_norm": 0.4023531377315521, "learning_rate": 9.916060967279308e-05, "loss": 0.0621, "step": 8440 }, { "grad_norm": 0.37021932005882263, "learning_rate": 9.91568342008159e-05, "loss": 0.0563, "step": 8450 }, { "grad_norm": 0.4524683952331543, "learning_rate": 9.915305032928019e-05, "loss": 0.0598, "step": 8460 }, { "grad_norm": 0.48902085423469543, "learning_rate": 9.914925805883253e-05, "loss": 0.0517, "step": 8470 }, { "grad_norm": 0.42045408487319946, "learning_rate": 9.914545739012088e-05, "loss": 0.0575, "step": 8480 }, { "grad_norm": 0.4086402952671051, "learning_rate": 9.91416483237947e-05, "loss": 0.0662, "step": 8490 }, { "grad_norm": 0.5043709874153137, "learning_rate": 9.913783086050485e-05, "loss": 0.0663, "step": 8500 }, { "grad_norm": 0.5037598013877869, "learning_rate": 9.913400500090364e-05, "loss": 0.0553, "step": 8510 }, { "grad_norm": 0.5742251873016357, "learning_rate": 9.913017074564479e-05, "loss": 0.0485, "step": 8520 }, { "grad_norm": 0.35522377490997314, "learning_rate": 9.912632809538348e-05, "loss": 0.0528, "step": 8530 }, { "grad_norm": 0.4088614881038666, "learning_rate": 9.912247705077629e-05, "loss": 0.0666, "step": 8540 }, { "grad_norm": 0.49467653036117554, "learning_rate": 9.911861761248127e-05, "loss": 0.0596, "step": 8550 }, { "grad_norm": 0.3475607931613922, "learning_rate": 9.91147497811579e-05, "loss": 0.0588, "step": 8560 }, { "grad_norm": 0.4604014754295349, "learning_rate": 9.911087355746709e-05, "loss": 0.0898, "step": 8570 }, { "grad_norm": 0.5929560661315918, "learning_rate": 9.910698894207117e-05, "loss": 0.07, "step": 8580 }, { "grad_norm": 0.41813012957572937, "learning_rate": 9.910309593563392e-05, "loss": 0.06, "step": 8590 }, { "grad_norm": 0.3136797249317169, "learning_rate": 9.909919453882057e-05, "loss": 0.0482, "step": 8600 }, { "grad_norm": 0.5554121136665344, "learning_rate": 9.90952847522977e-05, "loss": 0.0532, "step": 8610 }, { "grad_norm": 0.4952928423881531, "learning_rate": 9.909136657673346e-05, "loss": 0.0778, "step": 8620 }, { "grad_norm": 0.3980344533920288, "learning_rate": 9.908744001279731e-05, "loss": 0.0531, "step": 8630 }, { "grad_norm": 0.42486685514450073, "learning_rate": 9.90835050611602e-05, "loss": 0.0645, "step": 8640 }, { "grad_norm": 0.42865562438964844, "learning_rate": 9.90795617224945e-05, "loss": 0.0647, "step": 8650 }, { "grad_norm": 0.5066950917243958, "learning_rate": 9.907560999747405e-05, "loss": 0.0612, "step": 8660 }, { "grad_norm": 0.3675713539123535, "learning_rate": 9.907164988677408e-05, "loss": 0.0568, "step": 8670 }, { "grad_norm": 0.5558947920799255, "learning_rate": 9.906768139107124e-05, "loss": 0.0544, "step": 8680 }, { "grad_norm": 0.4715644121170044, "learning_rate": 9.906370451104367e-05, "loss": 0.0579, "step": 8690 }, { "grad_norm": 0.2937060296535492, "learning_rate": 9.905971924737088e-05, "loss": 0.0606, "step": 8700 }, { "grad_norm": 0.33401817083358765, "learning_rate": 9.905572560073387e-05, "loss": 0.0611, "step": 8710 }, { "grad_norm": 0.6436197757720947, "learning_rate": 9.905172357181501e-05, "loss": 0.0753, "step": 8720 }, { "grad_norm": 0.45227745175361633, "learning_rate": 9.904771316129817e-05, "loss": 0.0575, "step": 8730 }, { "grad_norm": 0.5521301627159119, "learning_rate": 9.904369436986862e-05, "loss": 0.0548, "step": 8740 }, { "grad_norm": 0.5437580943107605, "learning_rate": 9.903966719821303e-05, "loss": 0.0654, "step": 8750 }, { "grad_norm": 0.31113019585609436, "learning_rate": 9.903563164701956e-05, "loss": 0.0511, "step": 8760 }, { "grad_norm": 0.4817245602607727, "learning_rate": 9.903158771697778e-05, "loss": 0.0649, "step": 8770 }, { "grad_norm": 0.4836362898349762, "learning_rate": 9.902753540877867e-05, "loss": 0.0683, "step": 8780 }, { "grad_norm": 0.43733498454093933, "learning_rate": 9.902347472311466e-05, "loss": 0.058, "step": 8790 }, { "grad_norm": 0.4433652460575104, "learning_rate": 9.901940566067962e-05, "loss": 0.0439, "step": 8800 }, { "grad_norm": 0.43895578384399414, "learning_rate": 9.901532822216883e-05, "loss": 0.051, "step": 8810 }, { "grad_norm": 0.4183765649795532, "learning_rate": 9.901124240827904e-05, "loss": 0.0543, "step": 8820 }, { "grad_norm": 0.4335339069366455, "learning_rate": 9.900714821970835e-05, "loss": 0.0599, "step": 8830 }, { "grad_norm": 0.45780184864997864, "learning_rate": 9.900304565715641e-05, "loss": 0.0524, "step": 8840 }, { "grad_norm": 0.430409699678421, "learning_rate": 9.899893472132419e-05, "loss": 0.0543, "step": 8850 }, { "grad_norm": 0.5251228213310242, "learning_rate": 9.899481541291415e-05, "loss": 0.0548, "step": 8860 }, { "grad_norm": 0.6075785756111145, "learning_rate": 9.899068773263016e-05, "loss": 0.0625, "step": 8870 }, { "grad_norm": 0.4726378321647644, "learning_rate": 9.898655168117754e-05, "loss": 0.0582, "step": 8880 }, { "grad_norm": 0.42911389470100403, "learning_rate": 9.898240725926302e-05, "loss": 0.0622, "step": 8890 }, { "grad_norm": 0.5441592931747437, "learning_rate": 9.897825446759478e-05, "loss": 0.0637, "step": 8900 }, { "grad_norm": 0.34584230184555054, "learning_rate": 9.897409330688241e-05, "loss": 0.0533, "step": 8910 }, { "grad_norm": 0.5441373586654663, "learning_rate": 9.896992377783692e-05, "loss": 0.0723, "step": 8920 }, { "grad_norm": 0.4632478654384613, "learning_rate": 9.89657458811708e-05, "loss": 0.0615, "step": 8930 }, { "grad_norm": 0.3643001616001129, "learning_rate": 9.896155961759792e-05, "loss": 0.065, "step": 8940 }, { "grad_norm": 0.350515753030777, "learning_rate": 9.895736498783361e-05, "loss": 0.0546, "step": 8950 }, { "grad_norm": 0.37960436940193176, "learning_rate": 9.895316199259462e-05, "loss": 0.0485, "step": 8960 }, { "grad_norm": 0.3468948304653168, "learning_rate": 9.894895063259909e-05, "loss": 0.0494, "step": 8970 }, { "grad_norm": 0.4744904637336731, "learning_rate": 9.894473090856667e-05, "loss": 0.0643, "step": 8980 }, { "grad_norm": 0.5059979557991028, "learning_rate": 9.894050282121839e-05, "loss": 0.0796, "step": 8990 }, { "grad_norm": 0.47976475954055786, "learning_rate": 9.893626637127668e-05, "loss": 0.0606, "step": 9000 }, { "grad_norm": 1.122956395149231, "learning_rate": 9.893202155946546e-05, "loss": 0.0535, "step": 9010 }, { "grad_norm": 0.5506441593170166, "learning_rate": 9.892776838651006e-05, "loss": 0.0663, "step": 9020 }, { "grad_norm": 0.5711438655853271, "learning_rate": 9.892350685313722e-05, "loss": 0.0747, "step": 9030 }, { "grad_norm": 0.5419245958328247, "learning_rate": 9.891923696007513e-05, "loss": 0.0732, "step": 9040 }, { "grad_norm": 0.7961545586585999, "learning_rate": 9.891495870805336e-05, "loss": 0.0621, "step": 9050 }, { "grad_norm": 0.4729606807231903, "learning_rate": 9.891067209780298e-05, "loss": 0.0752, "step": 9060 }, { "grad_norm": 0.3959084749221802, "learning_rate": 9.890637713005646e-05, "loss": 0.0565, "step": 9070 }, { "grad_norm": 0.47326406836509705, "learning_rate": 9.890207380554767e-05, "loss": 0.0623, "step": 9080 }, { "grad_norm": 0.5323147773742676, "learning_rate": 9.889776212501196e-05, "loss": 0.0655, "step": 9090 }, { "grad_norm": 0.5140562057495117, "learning_rate": 9.889344208918605e-05, "loss": 0.07, "step": 9100 }, { "grad_norm": 0.495865136384964, "learning_rate": 9.888911369880812e-05, "loss": 0.0618, "step": 9110 }, { "grad_norm": 0.40728136897087097, "learning_rate": 9.888477695461777e-05, "loss": 0.0597, "step": 9120 }, { "grad_norm": 0.6221647262573242, "learning_rate": 9.888043185735607e-05, "loss": 0.0723, "step": 9130 }, { "grad_norm": 0.5027464628219604, "learning_rate": 9.887607840776542e-05, "loss": 0.0566, "step": 9140 }, { "grad_norm": 0.5109729766845703, "learning_rate": 9.887171660658975e-05, "loss": 0.0763, "step": 9150 }, { "grad_norm": 0.544312596321106, "learning_rate": 9.886734645457435e-05, "loss": 0.0589, "step": 9160 }, { "grad_norm": 0.39493608474731445, "learning_rate": 9.886296795246597e-05, "loss": 0.0545, "step": 9170 }, { "grad_norm": 0.6962580680847168, "learning_rate": 9.885858110101276e-05, "loss": 0.0581, "step": 9180 }, { "grad_norm": 0.5010082125663757, "learning_rate": 9.885418590096434e-05, "loss": 0.0513, "step": 9190 }, { "grad_norm": 0.4526444673538208, "learning_rate": 9.88497823530717e-05, "loss": 0.0615, "step": 9200 }, { "grad_norm": 0.2848573923110962, "learning_rate": 9.884537045808732e-05, "loss": 0.0656, "step": 9210 }, { "grad_norm": 0.42657285928726196, "learning_rate": 9.884095021676502e-05, "loss": 0.0624, "step": 9220 }, { "grad_norm": 0.3145913779735565, "learning_rate": 9.883652162986017e-05, "loss": 0.0543, "step": 9230 }, { "grad_norm": 0.32874080538749695, "learning_rate": 9.883208469812943e-05, "loss": 0.063, "step": 9240 }, { "grad_norm": 0.4018154442310333, "learning_rate": 9.882763942233098e-05, "loss": 0.0494, "step": 9250 }, { "grad_norm": 0.5344663858413696, "learning_rate": 9.882318580322441e-05, "loss": 0.0566, "step": 9260 }, { "grad_norm": 0.6002445816993713, "learning_rate": 9.881872384157067e-05, "loss": 0.0561, "step": 9270 }, { "grad_norm": 0.38654759526252747, "learning_rate": 9.881425353813225e-05, "loss": 0.0605, "step": 9280 }, { "grad_norm": 0.5475775599479675, "learning_rate": 9.880977489367296e-05, "loss": 0.0736, "step": 9290 }, { "grad_norm": 0.3896369934082031, "learning_rate": 9.88052879089581e-05, "loss": 0.0657, "step": 9300 }, { "grad_norm": 0.4910360276699066, "learning_rate": 9.880079258475434e-05, "loss": 0.0553, "step": 9310 }, { "grad_norm": 0.43140339851379395, "learning_rate": 9.879628892182985e-05, "loss": 0.07, "step": 9320 }, { "grad_norm": 0.33755800127983093, "learning_rate": 9.879177692095416e-05, "loss": 0.0552, "step": 9330 }, { "grad_norm": 0.4203072190284729, "learning_rate": 9.878725658289825e-05, "loss": 0.0638, "step": 9340 }, { "grad_norm": 0.543766975402832, "learning_rate": 9.878272790843454e-05, "loss": 0.0563, "step": 9350 }, { "grad_norm": 0.6821854710578918, "learning_rate": 9.877819089833682e-05, "loss": 0.058, "step": 9360 }, { "grad_norm": 0.42083609104156494, "learning_rate": 9.877364555338038e-05, "loss": 0.0699, "step": 9370 }, { "grad_norm": 0.3585001528263092, "learning_rate": 9.876909187434186e-05, "loss": 0.0619, "step": 9380 }, { "grad_norm": 0.6836966276168823, "learning_rate": 9.876452986199939e-05, "loss": 0.0599, "step": 9390 }, { "grad_norm": 0.41200268268585205, "learning_rate": 9.875995951713248e-05, "loss": 0.0705, "step": 9400 }, { "grad_norm": 0.4871574938297272, "learning_rate": 9.875538084052207e-05, "loss": 0.0655, "step": 9410 }, { "grad_norm": 0.6059675812721252, "learning_rate": 9.875079383295053e-05, "loss": 0.0565, "step": 9420 }, { "grad_norm": 0.36636388301849365, "learning_rate": 9.874619849520167e-05, "loss": 0.0558, "step": 9430 }, { "grad_norm": 0.48130717873573303, "learning_rate": 9.874159482806069e-05, "loss": 0.0593, "step": 9440 }, { "grad_norm": 0.36688074469566345, "learning_rate": 9.873698283231426e-05, "loss": 0.0492, "step": 9450 }, { "grad_norm": 0.5678653120994568, "learning_rate": 9.87323625087504e-05, "loss": 0.0491, "step": 9460 }, { "grad_norm": 0.4267147183418274, "learning_rate": 9.872773385815863e-05, "loss": 0.0762, "step": 9470 }, { "grad_norm": 0.4516391158103943, "learning_rate": 9.872309688132986e-05, "loss": 0.0537, "step": 9480 }, { "grad_norm": 0.5072023272514343, "learning_rate": 9.871845157905639e-05, "loss": 0.0633, "step": 9490 }, { "grad_norm": 0.4486435651779175, "learning_rate": 9.871379795213201e-05, "loss": 0.0557, "step": 9500 }, { "grad_norm": 0.3954801857471466, "learning_rate": 9.87091360013519e-05, "loss": 0.0543, "step": 9510 }, { "grad_norm": 0.43340665102005005, "learning_rate": 9.870446572751262e-05, "loss": 0.0586, "step": 9520 }, { "grad_norm": 0.379962682723999, "learning_rate": 9.869978713141224e-05, "loss": 0.0575, "step": 9530 }, { "grad_norm": 0.6431499123573303, "learning_rate": 9.869510021385016e-05, "loss": 0.0765, "step": 9540 }, { "grad_norm": 0.4511128067970276, "learning_rate": 9.869040497562727e-05, "loss": 0.0577, "step": 9550 }, { "grad_norm": 0.3485644459724426, "learning_rate": 9.868570141754587e-05, "loss": 0.0627, "step": 9560 }, { "grad_norm": 0.3629051744937897, "learning_rate": 9.868098954040965e-05, "loss": 0.0752, "step": 9570 }, { "grad_norm": 0.3694360554218292, "learning_rate": 9.867626934502374e-05, "loss": 0.0641, "step": 9580 }, { "grad_norm": 0.40874165296554565, "learning_rate": 9.86715408321947e-05, "loss": 0.0547, "step": 9590 }, { "grad_norm": 0.46394842863082886, "learning_rate": 9.86668040027305e-05, "loss": 0.0532, "step": 9600 }, { "grad_norm": 0.45124754309654236, "learning_rate": 9.866205885744053e-05, "loss": 0.0527, "step": 9610 }, { "grad_norm": 0.5352866649627686, "learning_rate": 9.865730539713563e-05, "loss": 0.0549, "step": 9620 }, { "grad_norm": 0.3600561320781708, "learning_rate": 9.8652543622628e-05, "loss": 0.0553, "step": 9630 }, { "grad_norm": 0.49924734234809875, "learning_rate": 9.864777353473132e-05, "loss": 0.056, "step": 9640 }, { "grad_norm": 0.5463290810585022, "learning_rate": 9.864299513426068e-05, "loss": 0.06, "step": 9650 }, { "grad_norm": 0.31315892934799194, "learning_rate": 9.863820842203254e-05, "loss": 0.0519, "step": 9660 }, { "grad_norm": 0.37181010842323303, "learning_rate": 9.863341339886483e-05, "loss": 0.0619, "step": 9670 }, { "grad_norm": 0.5278286337852478, "learning_rate": 9.86286100655769e-05, "loss": 0.0559, "step": 9680 }, { "grad_norm": 0.5751702785491943, "learning_rate": 9.862379842298953e-05, "loss": 0.0591, "step": 9690 }, { "grad_norm": 0.44032564759254456, "learning_rate": 9.861897847192485e-05, "loss": 0.0569, "step": 9700 }, { "grad_norm": 0.4759109616279602, "learning_rate": 9.86141502132065e-05, "loss": 0.0551, "step": 9710 }, { "grad_norm": 0.5131281614303589, "learning_rate": 9.860931364765946e-05, "loss": 0.0575, "step": 9720 }, { "grad_norm": 0.4861328601837158, "learning_rate": 9.860446877611021e-05, "loss": 0.0632, "step": 9730 }, { "grad_norm": 0.5383393168449402, "learning_rate": 9.859961559938655e-05, "loss": 0.0552, "step": 9740 }, { "grad_norm": 0.4293195307254791, "learning_rate": 9.85947541183178e-05, "loss": 0.0521, "step": 9750 }, { "grad_norm": 0.4951362609863281, "learning_rate": 9.858988433373463e-05, "loss": 0.0593, "step": 9760 }, { "grad_norm": 0.3804342746734619, "learning_rate": 9.858500624646918e-05, "loss": 0.0615, "step": 9770 }, { "grad_norm": 0.3618844449520111, "learning_rate": 9.858011985735497e-05, "loss": 0.0586, "step": 9780 }, { "grad_norm": 0.46017152070999146, "learning_rate": 9.857522516722693e-05, "loss": 0.0547, "step": 9790 }, { "grad_norm": 0.43261608481407166, "learning_rate": 9.857032217692145e-05, "loss": 0.0495, "step": 9800 }, { "grad_norm": 0.45438966155052185, "learning_rate": 9.856541088727631e-05, "loss": 0.0579, "step": 9810 }, { "grad_norm": 0.45502182841300964, "learning_rate": 9.856049129913072e-05, "loss": 0.0656, "step": 9820 }, { "grad_norm": 0.42832043766975403, "learning_rate": 9.85555634133253e-05, "loss": 0.0604, "step": 9830 }, { "grad_norm": 0.4670901298522949, "learning_rate": 9.855062723070208e-05, "loss": 0.0744, "step": 9840 }, { "grad_norm": 0.3795892596244812, "learning_rate": 9.854568275210454e-05, "loss": 0.0648, "step": 9850 }, { "grad_norm": 0.3189364969730377, "learning_rate": 9.854072997837754e-05, "loss": 0.0484, "step": 9860 }, { "grad_norm": 0.30879148840904236, "learning_rate": 9.853576891036737e-05, "loss": 0.0549, "step": 9870 }, { "grad_norm": 0.3750562369823456, "learning_rate": 9.853079954892177e-05, "loss": 0.0638, "step": 9880 }, { "grad_norm": 0.35535728931427, "learning_rate": 9.852582189488983e-05, "loss": 0.0487, "step": 9890 }, { "grad_norm": 0.3642033636569977, "learning_rate": 9.852083594912212e-05, "loss": 0.0578, "step": 9900 }, { "grad_norm": 0.32696449756622314, "learning_rate": 9.851584171247058e-05, "loss": 0.0518, "step": 9910 }, { "grad_norm": 0.5918006300926208, "learning_rate": 9.851083918578863e-05, "loss": 0.0707, "step": 9920 }, { "grad_norm": 0.44025105237960815, "learning_rate": 9.850582836993103e-05, "loss": 0.049, "step": 9930 }, { "grad_norm": 0.6183627843856812, "learning_rate": 9.850080926575397e-05, "loss": 0.0549, "step": 9940 }, { "grad_norm": 0.45941099524497986, "learning_rate": 9.849578187411515e-05, "loss": 0.05, "step": 9950 }, { "grad_norm": 0.37986934185028076, "learning_rate": 9.849074619587354e-05, "loss": 0.0516, "step": 9960 }, { "grad_norm": 0.34755825996398926, "learning_rate": 9.848570223188964e-05, "loss": 0.0621, "step": 9970 }, { "grad_norm": 0.42294907569885254, "learning_rate": 9.848064998302531e-05, "loss": 0.057, "step": 9980 }, { "grad_norm": 0.5003387331962585, "learning_rate": 9.847558945014386e-05, "loss": 0.06, "step": 9990 }, { "grad_norm": 0.5029525756835938, "learning_rate": 9.847052063410996e-05, "loss": 0.0643, "step": 10000 }, { "grad_norm": 0.5062949657440186, "learning_rate": 9.846544353578977e-05, "loss": 0.0584, "step": 10010 }, { "grad_norm": 0.4371069073677063, "learning_rate": 9.846035815605081e-05, "loss": 0.058, "step": 10020 }, { "grad_norm": 0.4119616746902466, "learning_rate": 9.845526449576204e-05, "loss": 0.0454, "step": 10030 }, { "grad_norm": 0.3703286945819855, "learning_rate": 9.845016255579383e-05, "loss": 0.0593, "step": 10040 }, { "grad_norm": 0.5655228495597839, "learning_rate": 9.844505233701794e-05, "loss": 0.0557, "step": 10050 }, { "grad_norm": 0.46740955114364624, "learning_rate": 9.843993384030757e-05, "loss": 0.0567, "step": 10060 }, { "grad_norm": 0.2956545948982239, "learning_rate": 9.843480706653737e-05, "loss": 0.0505, "step": 10070 }, { "grad_norm": 0.3872145712375641, "learning_rate": 9.84296720165833e-05, "loss": 0.0546, "step": 10080 }, { "grad_norm": 0.503693163394928, "learning_rate": 9.842452869132286e-05, "loss": 0.0559, "step": 10090 }, { "grad_norm": 0.424137681722641, "learning_rate": 9.841937709163489e-05, "loss": 0.0585, "step": 10100 }, { "grad_norm": 0.34537747502326965, "learning_rate": 9.841421721839962e-05, "loss": 0.0507, "step": 10110 }, { "grad_norm": 0.39741072058677673, "learning_rate": 9.840904907249879e-05, "loss": 0.0539, "step": 10120 }, { "grad_norm": 0.5649335980415344, "learning_rate": 9.840387265481545e-05, "loss": 0.0639, "step": 10130 }, { "grad_norm": 0.49150073528289795, "learning_rate": 9.839868796623411e-05, "loss": 0.0529, "step": 10140 }, { "grad_norm": 0.3739422857761383, "learning_rate": 9.839349500764072e-05, "loss": 0.0639, "step": 10150 }, { "grad_norm": 0.5887473225593567, "learning_rate": 9.83882937799226e-05, "loss": 0.0584, "step": 10160 }, { "grad_norm": 0.321927934885025, "learning_rate": 9.838308428396849e-05, "loss": 0.0539, "step": 10170 }, { "grad_norm": 0.41461971402168274, "learning_rate": 9.837786652066854e-05, "loss": 0.0704, "step": 10180 }, { "grad_norm": 0.4189949929714203, "learning_rate": 9.837264049091437e-05, "loss": 0.0712, "step": 10190 }, { "grad_norm": 0.4983333647251129, "learning_rate": 9.836740619559893e-05, "loss": 0.0564, "step": 10200 }, { "grad_norm": 0.49896422028541565, "learning_rate": 9.836216363561659e-05, "loss": 0.0579, "step": 10210 }, { "grad_norm": 0.34528398513793945, "learning_rate": 9.835691281186322e-05, "loss": 0.0465, "step": 10220 }, { "grad_norm": 0.40845245122909546, "learning_rate": 9.8351653725236e-05, "loss": 0.057, "step": 10230 }, { "grad_norm": 0.45859938859939575, "learning_rate": 9.83463863766336e-05, "loss": 0.0589, "step": 10240 }, { "grad_norm": 0.49283894896507263, "learning_rate": 9.834111076695602e-05, "loss": 0.0605, "step": 10250 }, { "grad_norm": 0.3845052421092987, "learning_rate": 9.833582689710477e-05, "loss": 0.051, "step": 10260 }, { "grad_norm": 0.36395010352134705, "learning_rate": 9.833053476798268e-05, "loss": 0.059, "step": 10270 }, { "grad_norm": 0.5350288152694702, "learning_rate": 9.832523438049404e-05, "loss": 0.0567, "step": 10280 }, { "grad_norm": 0.5128869414329529, "learning_rate": 9.831992573554454e-05, "loss": 0.0652, "step": 10290 }, { "grad_norm": 0.5054547786712646, "learning_rate": 9.831460883404128e-05, "loss": 0.0547, "step": 10300 }, { "grad_norm": 0.43813398480415344, "learning_rate": 9.830928367689278e-05, "loss": 0.0529, "step": 10310 }, { "grad_norm": 0.35673975944519043, "learning_rate": 9.830395026500896e-05, "loss": 0.062, "step": 10320 }, { "grad_norm": 0.413614422082901, "learning_rate": 9.829860859930115e-05, "loss": 0.0567, "step": 10330 }, { "grad_norm": 0.41404086351394653, "learning_rate": 9.829325868068212e-05, "loss": 0.0578, "step": 10340 }, { "grad_norm": 0.4736163318157196, "learning_rate": 9.8287900510066e-05, "loss": 0.0754, "step": 10350 }, { "grad_norm": 0.41841715574264526, "learning_rate": 9.828253408836834e-05, "loss": 0.051, "step": 10360 }, { "grad_norm": 0.3907080888748169, "learning_rate": 9.827715941650615e-05, "loss": 0.0524, "step": 10370 }, { "grad_norm": 0.5006973743438721, "learning_rate": 9.82717764953978e-05, "loss": 0.0616, "step": 10380 }, { "grad_norm": 0.43180859088897705, "learning_rate": 9.826638532596308e-05, "loss": 0.0556, "step": 10390 }, { "grad_norm": 0.4847463369369507, "learning_rate": 9.82609859091232e-05, "loss": 0.0602, "step": 10400 }, { "grad_norm": 0.5669083595275879, "learning_rate": 9.825557824580076e-05, "loss": 0.0654, "step": 10410 }, { "grad_norm": 0.3305533230304718, "learning_rate": 9.82501623369198e-05, "loss": 0.0586, "step": 10420 }, { "grad_norm": 0.38766470551490784, "learning_rate": 9.824473818340574e-05, "loss": 0.0483, "step": 10430 }, { "grad_norm": 0.6765963435173035, "learning_rate": 9.823930578618541e-05, "loss": 0.0522, "step": 10440 }, { "grad_norm": 0.4206819534301758, "learning_rate": 9.823386514618709e-05, "loss": 0.0601, "step": 10450 }, { "grad_norm": 0.41518890857696533, "learning_rate": 9.82284162643404e-05, "loss": 0.0529, "step": 10460 }, { "grad_norm": 0.5392363667488098, "learning_rate": 9.822295914157642e-05, "loss": 0.0626, "step": 10470 }, { "grad_norm": 0.3258773386478424, "learning_rate": 9.821749377882763e-05, "loss": 0.0492, "step": 10480 }, { "grad_norm": 0.39749833941459656, "learning_rate": 9.821202017702791e-05, "loss": 0.0494, "step": 10490 }, { "grad_norm": 0.7785617113113403, "learning_rate": 9.820653833711253e-05, "loss": 0.0513, "step": 10500 }, { "grad_norm": 0.6920726895332336, "learning_rate": 9.820104826001822e-05, "loss": 0.0645, "step": 10510 }, { "grad_norm": 0.379166841506958, "learning_rate": 9.819554994668305e-05, "loss": 0.0441, "step": 10520 }, { "grad_norm": 0.4621697962284088, "learning_rate": 9.819004339804654e-05, "loss": 0.0541, "step": 10530 }, { "grad_norm": 0.39015641808509827, "learning_rate": 9.818452861504961e-05, "loss": 0.0467, "step": 10540 }, { "grad_norm": 0.41982513666152954, "learning_rate": 9.81790055986346e-05, "loss": 0.0482, "step": 10550 }, { "grad_norm": 0.49144747853279114, "learning_rate": 9.817347434974523e-05, "loss": 0.0538, "step": 10560 }, { "grad_norm": 0.3972775936126709, "learning_rate": 9.816793486932664e-05, "loss": 0.0479, "step": 10570 }, { "grad_norm": 0.3529851734638214, "learning_rate": 9.816238715832538e-05, "loss": 0.0566, "step": 10580 }, { "grad_norm": 0.5645320415496826, "learning_rate": 9.815683121768939e-05, "loss": 0.0613, "step": 10590 }, { "grad_norm": 0.36421942710876465, "learning_rate": 9.815126704836804e-05, "loss": 0.0704, "step": 10600 }, { "grad_norm": 0.7019299864768982, "learning_rate": 9.81456946513121e-05, "loss": 0.0821, "step": 10610 }, { "grad_norm": 0.6391369104385376, "learning_rate": 9.814011402747373e-05, "loss": 0.0715, "step": 10620 }, { "grad_norm": 0.4190993905067444, "learning_rate": 9.813452517780651e-05, "loss": 0.0493, "step": 10630 }, { "grad_norm": 0.48832589387893677, "learning_rate": 9.81289281032654e-05, "loss": 0.0724, "step": 10640 }, { "grad_norm": 0.3881099224090576, "learning_rate": 9.812332280480683e-05, "loss": 0.0648, "step": 10650 }, { "grad_norm": 0.4545387923717499, "learning_rate": 9.811770928338854e-05, "loss": 0.0623, "step": 10660 }, { "grad_norm": 0.51077800989151, "learning_rate": 9.811208753996979e-05, "loss": 0.056, "step": 10670 }, { "grad_norm": 0.3232453167438507, "learning_rate": 9.810645757551113e-05, "loss": 0.0501, "step": 10680 }, { "grad_norm": 0.4621141850948334, "learning_rate": 9.810081939097459e-05, "loss": 0.0548, "step": 10690 }, { "grad_norm": 0.507266104221344, "learning_rate": 9.809517298732356e-05, "loss": 0.0812, "step": 10700 }, { "grad_norm": 0.34652239084243774, "learning_rate": 9.80895183655229e-05, "loss": 0.0623, "step": 10710 }, { "grad_norm": 0.4222243130207062, "learning_rate": 9.808385552653877e-05, "loss": 0.0588, "step": 10720 }, { "grad_norm": 0.41094115376472473, "learning_rate": 9.807818447133886e-05, "loss": 0.0533, "step": 10730 }, { "grad_norm": 0.4246993064880371, "learning_rate": 9.807250520089215e-05, "loss": 0.0524, "step": 10740 }, { "grad_norm": 0.45378726720809937, "learning_rate": 9.806681771616908e-05, "loss": 0.0592, "step": 10750 }, { "grad_norm": 0.460653156042099, "learning_rate": 9.80611220181415e-05, "loss": 0.0697, "step": 10760 }, { "grad_norm": 0.6050897240638733, "learning_rate": 9.805541810778264e-05, "loss": 0.0655, "step": 10770 }, { "grad_norm": 0.4356127381324768, "learning_rate": 9.804970598606716e-05, "loss": 0.0624, "step": 10780 }, { "grad_norm": 0.39845404028892517, "learning_rate": 9.804398565397106e-05, "loss": 0.0551, "step": 10790 }, { "grad_norm": 0.6390231847763062, "learning_rate": 9.803825711247183e-05, "loss": 0.0707, "step": 10800 }, { "grad_norm": 0.648217499256134, "learning_rate": 9.803252036254831e-05, "loss": 0.0593, "step": 10810 }, { "grad_norm": 0.4851435124874115, "learning_rate": 9.802677540518076e-05, "loss": 0.0636, "step": 10820 }, { "grad_norm": 0.6034203171730042, "learning_rate": 9.802102224135081e-05, "loss": 0.0673, "step": 10830 }, { "grad_norm": 0.3073295056819916, "learning_rate": 9.801526087204155e-05, "loss": 0.0635, "step": 10840 }, { "grad_norm": 0.5849307775497437, "learning_rate": 9.800949129823743e-05, "loss": 0.0618, "step": 10850 }, { "grad_norm": 0.41362473368644714, "learning_rate": 9.80037135209243e-05, "loss": 0.0555, "step": 10860 }, { "grad_norm": 0.43270963430404663, "learning_rate": 9.799792754108946e-05, "loss": 0.0596, "step": 10870 }, { "grad_norm": 0.3962707221508026, "learning_rate": 9.799213335972152e-05, "loss": 0.0553, "step": 10880 }, { "grad_norm": 0.3622930645942688, "learning_rate": 9.798633097781058e-05, "loss": 0.0552, "step": 10890 }, { "grad_norm": 0.4341129660606384, "learning_rate": 9.79805203963481e-05, "loss": 0.0575, "step": 10900 }, { "grad_norm": 0.5534108877182007, "learning_rate": 9.797470161632697e-05, "loss": 0.0536, "step": 10910 }, { "grad_norm": 0.37151068449020386, "learning_rate": 9.796887463874145e-05, "loss": 0.0539, "step": 10920 }, { "grad_norm": 0.4427274465560913, "learning_rate": 9.796303946458718e-05, "loss": 0.0673, "step": 10930 }, { "grad_norm": 0.4677928388118744, "learning_rate": 9.795719609486127e-05, "loss": 0.0669, "step": 10940 }, { "grad_norm": 0.47822198271751404, "learning_rate": 9.795134453056219e-05, "loss": 0.0587, "step": 10950 }, { "grad_norm": 0.34578588604927063, "learning_rate": 9.794548477268979e-05, "loss": 0.0555, "step": 10960 }, { "grad_norm": 0.41567280888557434, "learning_rate": 9.793961682224537e-05, "loss": 0.0567, "step": 10970 }, { "grad_norm": 0.33626997470855713, "learning_rate": 9.793374068023156e-05, "loss": 0.0564, "step": 10980 }, { "grad_norm": 0.4508557617664337, "learning_rate": 9.792785634765247e-05, "loss": 0.0515, "step": 10990 }, { "grad_norm": 0.48642081022262573, "learning_rate": 9.792196382551357e-05, "loss": 0.0592, "step": 11000 }, { "grad_norm": 0.6468857526779175, "learning_rate": 9.791606311482171e-05, "loss": 0.0567, "step": 11010 }, { "grad_norm": 0.4135489761829376, "learning_rate": 9.791015421658518e-05, "loss": 0.0616, "step": 11020 }, { "grad_norm": 0.6094536781311035, "learning_rate": 9.790423713181362e-05, "loss": 0.0554, "step": 11030 }, { "grad_norm": 0.3795141279697418, "learning_rate": 9.789831186151814e-05, "loss": 0.0556, "step": 11040 }, { "grad_norm": 0.41814640164375305, "learning_rate": 9.789237840671118e-05, "loss": 0.0505, "step": 11050 }, { "grad_norm": 0.5335978865623474, "learning_rate": 9.78864367684066e-05, "loss": 0.0595, "step": 11060 }, { "grad_norm": 0.3555239737033844, "learning_rate": 9.788048694761968e-05, "loss": 0.0493, "step": 11070 }, { "grad_norm": 0.45342928171157837, "learning_rate": 9.787452894536709e-05, "loss": 0.0575, "step": 11080 }, { "grad_norm": 0.36121881008148193, "learning_rate": 9.786856276266685e-05, "loss": 0.0503, "step": 11090 }, { "grad_norm": 0.4260951578617096, "learning_rate": 9.786258840053845e-05, "loss": 0.0568, "step": 11100 }, { "grad_norm": 0.3685038387775421, "learning_rate": 9.785660586000273e-05, "loss": 0.0552, "step": 11110 }, { "grad_norm": 0.42220503091812134, "learning_rate": 9.785061514208196e-05, "loss": 0.0527, "step": 11120 }, { "grad_norm": 0.434721440076828, "learning_rate": 9.784461624779977e-05, "loss": 0.0626, "step": 11130 }, { "grad_norm": 0.5165106654167175, "learning_rate": 9.783860917818123e-05, "loss": 0.0521, "step": 11140 }, { "grad_norm": 0.5050967335700989, "learning_rate": 9.783259393425277e-05, "loss": 0.0412, "step": 11150 }, { "grad_norm": 0.5958642959594727, "learning_rate": 9.782657051704221e-05, "loss": 0.0529, "step": 11160 }, { "grad_norm": 0.46789634227752686, "learning_rate": 9.782053892757883e-05, "loss": 0.0485, "step": 11170 }, { "grad_norm": 0.43901094794273376, "learning_rate": 9.781449916689324e-05, "loss": 0.0504, "step": 11180 }, { "grad_norm": 0.332599937915802, "learning_rate": 9.780845123601746e-05, "loss": 0.056, "step": 11190 }, { "grad_norm": 0.6372300982475281, "learning_rate": 9.780239513598492e-05, "loss": 0.0612, "step": 11200 }, { "grad_norm": 0.5010126829147339, "learning_rate": 9.779633086783047e-05, "loss": 0.0572, "step": 11210 }, { "grad_norm": 0.7651643753051758, "learning_rate": 9.779025843259031e-05, "loss": 0.0527, "step": 11220 }, { "grad_norm": 0.4795134961605072, "learning_rate": 9.778417783130204e-05, "loss": 0.0648, "step": 11230 }, { "grad_norm": 0.42991554737091064, "learning_rate": 9.777808906500468e-05, "loss": 0.0523, "step": 11240 }, { "grad_norm": 0.48731908202171326, "learning_rate": 9.777199213473862e-05, "loss": 0.0497, "step": 11250 }, { "grad_norm": 0.4929375648498535, "learning_rate": 9.77658870415457e-05, "loss": 0.0457, "step": 11260 }, { "grad_norm": 0.5855218172073364, "learning_rate": 9.775977378646906e-05, "loss": 0.0498, "step": 11270 }, { "grad_norm": 0.47217321395874023, "learning_rate": 9.775365237055331e-05, "loss": 0.0584, "step": 11280 }, { "grad_norm": 0.45357993245124817, "learning_rate": 9.774752279484445e-05, "loss": 0.0613, "step": 11290 }, { "grad_norm": 0.4787278473377228, "learning_rate": 9.774138506038984e-05, "loss": 0.0451, "step": 11300 }, { "grad_norm": 0.4793242812156677, "learning_rate": 9.773523916823826e-05, "loss": 0.0561, "step": 11310 }, { "grad_norm": 0.5883529782295227, "learning_rate": 9.772908511943986e-05, "loss": 0.0632, "step": 11320 }, { "grad_norm": 0.3872091770172119, "learning_rate": 9.77229229150462e-05, "loss": 0.0544, "step": 11330 }, { "grad_norm": 0.3500083088874817, "learning_rate": 9.771675255611024e-05, "loss": 0.0451, "step": 11340 }, { "grad_norm": 0.45945826172828674, "learning_rate": 9.771057404368632e-05, "loss": 0.0747, "step": 11350 }, { "grad_norm": 0.5197332501411438, "learning_rate": 9.770438737883018e-05, "loss": 0.0584, "step": 11360 }, { "grad_norm": 0.34130969643592834, "learning_rate": 9.769819256259898e-05, "loss": 0.0412, "step": 11370 }, { "grad_norm": 0.4344630837440491, "learning_rate": 9.769198959605119e-05, "loss": 0.0472, "step": 11380 }, { "grad_norm": 0.38674336671829224, "learning_rate": 9.768577848024678e-05, "loss": 0.0645, "step": 11390 }, { "grad_norm": 0.634168267250061, "learning_rate": 9.767955921624702e-05, "loss": 0.069, "step": 11400 }, { "grad_norm": 0.40684154629707336, "learning_rate": 9.767333180511465e-05, "loss": 0.0463, "step": 11410 }, { "grad_norm": 0.35300981998443604, "learning_rate": 9.766709624791373e-05, "loss": 0.064, "step": 11420 }, { "grad_norm": 0.2765166461467743, "learning_rate": 9.766085254570975e-05, "loss": 0.053, "step": 11430 }, { "grad_norm": 0.4438316226005554, "learning_rate": 9.76546006995696e-05, "loss": 0.0484, "step": 11440 }, { "grad_norm": 0.46002891659736633, "learning_rate": 9.764834071056155e-05, "loss": 0.0458, "step": 11450 }, { "grad_norm": 0.5446194410324097, "learning_rate": 9.764207257975526e-05, "loss": 0.0595, "step": 11460 }, { "grad_norm": 0.442137748003006, "learning_rate": 9.763579630822179e-05, "loss": 0.0466, "step": 11470 }, { "grad_norm": 0.5057682991027832, "learning_rate": 9.762951189703356e-05, "loss": 0.0514, "step": 11480 }, { "grad_norm": 0.4334900677204132, "learning_rate": 9.762321934726442e-05, "loss": 0.0577, "step": 11490 }, { "grad_norm": 0.4889335632324219, "learning_rate": 9.761691865998959e-05, "loss": 0.053, "step": 11500 }, { "grad_norm": 0.40103253722190857, "learning_rate": 9.76106098362857e-05, "loss": 0.0429, "step": 11510 }, { "grad_norm": 0.31766319274902344, "learning_rate": 9.760429287723072e-05, "loss": 0.0518, "step": 11520 }, { "grad_norm": 0.41593238711357117, "learning_rate": 9.759796778390406e-05, "loss": 0.0451, "step": 11530 }, { "grad_norm": 0.46325448155403137, "learning_rate": 9.759163455738653e-05, "loss": 0.0537, "step": 11540 }, { "grad_norm": 0.5031986236572266, "learning_rate": 9.75852931987603e-05, "loss": 0.0607, "step": 11550 }, { "grad_norm": 0.5930668711662292, "learning_rate": 9.757894370910891e-05, "loss": 0.0584, "step": 11560 }, { "grad_norm": 0.5218968987464905, "learning_rate": 9.757258608951733e-05, "loss": 0.0516, "step": 11570 }, { "grad_norm": 0.6189503073692322, "learning_rate": 9.75662203410719e-05, "loss": 0.0622, "step": 11580 }, { "grad_norm": 0.44555097818374634, "learning_rate": 9.755984646486034e-05, "loss": 0.0607, "step": 11590 }, { "grad_norm": 0.4600214958190918, "learning_rate": 9.75534644619718e-05, "loss": 0.0494, "step": 11600 }, { "grad_norm": 0.4284780025482178, "learning_rate": 9.754707433349676e-05, "loss": 0.0737, "step": 11610 }, { "grad_norm": 0.47870737314224243, "learning_rate": 9.754067608052715e-05, "loss": 0.0648, "step": 11620 }, { "grad_norm": 0.5755010843276978, "learning_rate": 9.753426970415622e-05, "loss": 0.0744, "step": 11630 }, { "grad_norm": 0.4970603585243225, "learning_rate": 9.752785520547868e-05, "loss": 0.0547, "step": 11640 }, { "grad_norm": 0.4155946373939514, "learning_rate": 9.752143258559056e-05, "loss": 0.0572, "step": 11650 }, { "grad_norm": 0.35152533650398254, "learning_rate": 9.751500184558933e-05, "loss": 0.0506, "step": 11660 }, { "grad_norm": 0.34461709856987, "learning_rate": 9.750856298657383e-05, "loss": 0.0436, "step": 11670 }, { "grad_norm": 0.34051713347435, "learning_rate": 9.750211600964428e-05, "loss": 0.052, "step": 11680 }, { "grad_norm": 0.41016659140586853, "learning_rate": 9.749566091590226e-05, "loss": 0.0516, "step": 11690 }, { "grad_norm": 0.4237092137336731, "learning_rate": 9.748919770645083e-05, "loss": 0.0487, "step": 11700 }, { "grad_norm": 0.4282430410385132, "learning_rate": 9.748272638239432e-05, "loss": 0.0468, "step": 11710 }, { "grad_norm": 0.5277729034423828, "learning_rate": 9.747624694483855e-05, "loss": 0.0453, "step": 11720 }, { "grad_norm": 0.3529506027698517, "learning_rate": 9.746975939489065e-05, "loss": 0.042, "step": 11730 }, { "grad_norm": 0.6057960391044617, "learning_rate": 9.746326373365918e-05, "loss": 0.0581, "step": 11740 }, { "grad_norm": 0.41606810688972473, "learning_rate": 9.745675996225403e-05, "loss": 0.0617, "step": 11750 }, { "grad_norm": 0.44787290692329407, "learning_rate": 9.745024808178657e-05, "loss": 0.0576, "step": 11760 }, { "grad_norm": 0.4436747431755066, "learning_rate": 9.744372809336947e-05, "loss": 0.0573, "step": 11770 }, { "grad_norm": 0.5531265735626221, "learning_rate": 9.743719999811682e-05, "loss": 0.0453, "step": 11780 }, { "grad_norm": 0.5176612138748169, "learning_rate": 9.743066379714412e-05, "loss": 0.0475, "step": 11790 }, { "grad_norm": 0.310330331325531, "learning_rate": 9.74241194915682e-05, "loss": 0.0553, "step": 11800 }, { "grad_norm": 0.35178473591804504, "learning_rate": 9.741756708250731e-05, "loss": 0.0656, "step": 11810 }, { "grad_norm": 0.36779704689979553, "learning_rate": 9.741100657108109e-05, "loss": 0.0522, "step": 11820 }, { "grad_norm": 0.47580209374427795, "learning_rate": 9.740443795841054e-05, "loss": 0.0643, "step": 11830 }, { "grad_norm": 0.3517252206802368, "learning_rate": 9.739786124561805e-05, "loss": 0.0476, "step": 11840 }, { "grad_norm": 0.38903531432151794, "learning_rate": 9.73912764338274e-05, "loss": 0.0543, "step": 11850 }, { "grad_norm": 0.48615631461143494, "learning_rate": 9.738468352416377e-05, "loss": 0.0641, "step": 11860 }, { "grad_norm": 0.48408570885658264, "learning_rate": 9.737808251775369e-05, "loss": 0.0478, "step": 11870 }, { "grad_norm": 0.39901474118232727, "learning_rate": 9.737147341572512e-05, "loss": 0.0623, "step": 11880 }, { "grad_norm": 0.3756915032863617, "learning_rate": 9.736485621920735e-05, "loss": 0.0455, "step": 11890 }, { "grad_norm": 0.4413887858390808, "learning_rate": 9.735823092933108e-05, "loss": 0.0517, "step": 11900 }, { "grad_norm": 0.42510566115379333, "learning_rate": 9.735159754722838e-05, "loss": 0.0502, "step": 11910 }, { "grad_norm": 0.370809942483902, "learning_rate": 9.734495607403275e-05, "loss": 0.0499, "step": 11920 }, { "grad_norm": 0.39076292514801025, "learning_rate": 9.733830651087901e-05, "loss": 0.0555, "step": 11930 }, { "grad_norm": 0.5764327645301819, "learning_rate": 9.733164885890338e-05, "loss": 0.0482, "step": 11940 }, { "grad_norm": 0.4518195688724518, "learning_rate": 9.732498311924349e-05, "loss": 0.0425, "step": 11950 }, { "grad_norm": 0.35466307401657104, "learning_rate": 9.731830929303833e-05, "loss": 0.0471, "step": 11960 }, { "grad_norm": 0.3543587923049927, "learning_rate": 9.731162738142827e-05, "loss": 0.0475, "step": 11970 }, { "grad_norm": 0.3775603771209717, "learning_rate": 9.730493738555506e-05, "loss": 0.0522, "step": 11980 }, { "grad_norm": 0.3980860114097595, "learning_rate": 9.729823930656186e-05, "loss": 0.0512, "step": 11990 }, { "grad_norm": 0.5234410762786865, "learning_rate": 9.729153314559316e-05, "loss": 0.0628, "step": 12000 }, { "grad_norm": 0.5548250675201416, "learning_rate": 9.728481890379486e-05, "loss": 0.0504, "step": 12010 }, { "grad_norm": 0.3075947165489197, "learning_rate": 9.727809658231428e-05, "loss": 0.0464, "step": 12020 }, { "grad_norm": 0.4553605020046234, "learning_rate": 9.727136618230003e-05, "loss": 0.0458, "step": 12030 }, { "grad_norm": 0.7177439332008362, "learning_rate": 9.726462770490219e-05, "loss": 0.0488, "step": 12040 }, { "grad_norm": 0.4482293128967285, "learning_rate": 9.725788115127214e-05, "loss": 0.0506, "step": 12050 }, { "grad_norm": 0.4333193898200989, "learning_rate": 9.725112652256274e-05, "loss": 0.0549, "step": 12060 }, { "grad_norm": 0.3872765898704529, "learning_rate": 9.724436381992812e-05, "loss": 0.0708, "step": 12070 }, { "grad_norm": 0.5031154155731201, "learning_rate": 9.723759304452387e-05, "loss": 0.0488, "step": 12080 }, { "grad_norm": 0.3430350720882416, "learning_rate": 9.72308141975069e-05, "loss": 0.0468, "step": 12090 }, { "grad_norm": 0.4087543785572052, "learning_rate": 9.722402728003557e-05, "loss": 0.0539, "step": 12100 }, { "grad_norm": 0.6738215684890747, "learning_rate": 9.721723229326953e-05, "loss": 0.0547, "step": 12110 }, { "grad_norm": 0.492762953042984, "learning_rate": 9.721042923836992e-05, "loss": 0.0448, "step": 12120 }, { "grad_norm": 0.42791736125946045, "learning_rate": 9.720361811649914e-05, "loss": 0.0518, "step": 12130 }, { "grad_norm": 0.44797462224960327, "learning_rate": 9.719679892882106e-05, "loss": 0.0477, "step": 12140 }, { "grad_norm": 0.41706526279449463, "learning_rate": 9.718997167650085e-05, "loss": 0.0413, "step": 12150 }, { "grad_norm": 0.39196670055389404, "learning_rate": 9.718313636070515e-05, "loss": 0.0469, "step": 12160 }, { "grad_norm": 0.41630902886390686, "learning_rate": 9.717629298260192e-05, "loss": 0.0527, "step": 12170 }, { "grad_norm": 0.42392465472221375, "learning_rate": 9.716944154336047e-05, "loss": 0.055, "step": 12180 }, { "grad_norm": 0.3740271329879761, "learning_rate": 9.716258204415157e-05, "loss": 0.0464, "step": 12190 }, { "grad_norm": 0.41008588671684265, "learning_rate": 9.715571448614728e-05, "loss": 0.0597, "step": 12200 }, { "grad_norm": 0.40151265263557434, "learning_rate": 9.71488388705211e-05, "loss": 0.0678, "step": 12210 }, { "grad_norm": 0.45577237010002136, "learning_rate": 9.714195519844788e-05, "loss": 0.0482, "step": 12220 }, { "grad_norm": 0.4886590838432312, "learning_rate": 9.713506347110386e-05, "loss": 0.0514, "step": 12230 }, { "grad_norm": 0.4707130193710327, "learning_rate": 9.712816368966662e-05, "loss": 0.0498, "step": 12240 }, { "grad_norm": 0.3872677683830261, "learning_rate": 9.712125585531517e-05, "loss": 0.064, "step": 12250 }, { "grad_norm": 0.5612844228744507, "learning_rate": 9.711433996922988e-05, "loss": 0.0733, "step": 12260 }, { "grad_norm": 0.40948185324668884, "learning_rate": 9.710741603259245e-05, "loss": 0.0533, "step": 12270 }, { "grad_norm": 0.42486199736595154, "learning_rate": 9.710048404658603e-05, "loss": 0.0567, "step": 12280 }, { "grad_norm": 0.3417401611804962, "learning_rate": 9.709354401239508e-05, "loss": 0.058, "step": 12290 }, { "grad_norm": 0.3666415810585022, "learning_rate": 9.708659593120546e-05, "loss": 0.0444, "step": 12300 }, { "grad_norm": 0.4455275535583496, "learning_rate": 9.707963980420443e-05, "loss": 0.0475, "step": 12310 }, { "grad_norm": 0.41144296526908875, "learning_rate": 9.707267563258058e-05, "loss": 0.0549, "step": 12320 }, { "grad_norm": 0.3629525303840637, "learning_rate": 9.70657034175239e-05, "loss": 0.0409, "step": 12330 }, { "grad_norm": 0.48039138317108154, "learning_rate": 9.705872316022577e-05, "loss": 0.0563, "step": 12340 }, { "grad_norm": 0.41358011960983276, "learning_rate": 9.705173486187891e-05, "loss": 0.0519, "step": 12350 }, { "grad_norm": 0.35323861241340637, "learning_rate": 9.704473852367741e-05, "loss": 0.0603, "step": 12360 }, { "grad_norm": 0.3187739849090576, "learning_rate": 9.70377341468168e-05, "loss": 0.048, "step": 12370 }, { "grad_norm": 0.40090039372444153, "learning_rate": 9.703072173249389e-05, "loss": 0.0511, "step": 12380 }, { "grad_norm": 0.47929179668426514, "learning_rate": 9.702370128190693e-05, "loss": 0.0623, "step": 12390 }, { "grad_norm": 0.4549730122089386, "learning_rate": 9.701667279625552e-05, "loss": 0.0596, "step": 12400 }, { "grad_norm": 0.4848211705684662, "learning_rate": 9.700963627674065e-05, "loss": 0.0566, "step": 12410 }, { "grad_norm": 0.3672427833080292, "learning_rate": 9.700259172456466e-05, "loss": 0.0487, "step": 12420 }, { "grad_norm": 0.5273815989494324, "learning_rate": 9.699553914093124e-05, "loss": 0.0695, "step": 12430 }, { "grad_norm": 0.5076151490211487, "learning_rate": 9.698847852704553e-05, "loss": 0.0668, "step": 12440 }, { "grad_norm": 0.48393186926841736, "learning_rate": 9.6981409884114e-05, "loss": 0.0501, "step": 12450 }, { "grad_norm": 0.37472841143608093, "learning_rate": 9.697433321334443e-05, "loss": 0.0568, "step": 12460 }, { "grad_norm": 0.2745518386363983, "learning_rate": 9.696724851594607e-05, "loss": 0.0472, "step": 12470 }, { "grad_norm": 0.294425904750824, "learning_rate": 9.696015579312952e-05, "loss": 0.0425, "step": 12480 }, { "grad_norm": 0.38828736543655396, "learning_rate": 9.695305504610668e-05, "loss": 0.0478, "step": 12490 }, { "grad_norm": 0.35990607738494873, "learning_rate": 9.694594627609092e-05, "loss": 0.051, "step": 12500 }, { "grad_norm": 0.4120553135871887, "learning_rate": 9.693882948429691e-05, "loss": 0.066, "step": 12510 }, { "grad_norm": 0.48542240262031555, "learning_rate": 9.693170467194071e-05, "loss": 0.0512, "step": 12520 }, { "grad_norm": 0.6171020865440369, "learning_rate": 9.692457184023977e-05, "loss": 0.0514, "step": 12530 }, { "grad_norm": 0.4153648912906647, "learning_rate": 9.691743099041291e-05, "loss": 0.0586, "step": 12540 }, { "grad_norm": 0.6141236424446106, "learning_rate": 9.691028212368027e-05, "loss": 0.0564, "step": 12550 }, { "grad_norm": 0.47928425669670105, "learning_rate": 9.690312524126342e-05, "loss": 0.0522, "step": 12560 }, { "grad_norm": 0.44296547770500183, "learning_rate": 9.689596034438527e-05, "loss": 0.0523, "step": 12570 }, { "grad_norm": 0.268115758895874, "learning_rate": 9.688878743427012e-05, "loss": 0.0469, "step": 12580 }, { "grad_norm": 0.3851206302642822, "learning_rate": 9.688160651214359e-05, "loss": 0.0498, "step": 12590 }, { "grad_norm": 0.4025841951370239, "learning_rate": 9.687441757923273e-05, "loss": 0.0542, "step": 12600 }, { "grad_norm": 0.41910675168037415, "learning_rate": 9.68672206367659e-05, "loss": 0.0446, "step": 12610 }, { "grad_norm": 0.4515206515789032, "learning_rate": 9.686001568597291e-05, "loss": 0.0459, "step": 12620 }, { "grad_norm": 0.30085667967796326, "learning_rate": 9.685280272808486e-05, "loss": 0.0398, "step": 12630 }, { "grad_norm": 0.4601069688796997, "learning_rate": 9.684558176433424e-05, "loss": 0.0548, "step": 12640 }, { "grad_norm": 0.54300856590271, "learning_rate": 9.683835279595495e-05, "loss": 0.048, "step": 12650 }, { "grad_norm": 0.5263824462890625, "learning_rate": 9.683111582418216e-05, "loss": 0.0562, "step": 12660 }, { "grad_norm": 0.45691755414009094, "learning_rate": 9.682387085025254e-05, "loss": 0.0494, "step": 12670 }, { "grad_norm": 0.3696897029876709, "learning_rate": 9.681661787540401e-05, "loss": 0.0611, "step": 12680 }, { "grad_norm": 0.3684198260307312, "learning_rate": 9.680935690087593e-05, "loss": 0.0511, "step": 12690 }, { "grad_norm": 0.4471553862094879, "learning_rate": 9.680208792790901e-05, "loss": 0.0559, "step": 12700 }, { "grad_norm": 0.7865198254585266, "learning_rate": 9.679481095774529e-05, "loss": 0.0566, "step": 12710 }, { "grad_norm": 0.38431721925735474, "learning_rate": 9.678752599162822e-05, "loss": 0.0512, "step": 12720 }, { "grad_norm": 0.45565634965896606, "learning_rate": 9.678023303080259e-05, "loss": 0.0476, "step": 12730 }, { "grad_norm": 0.39118024706840515, "learning_rate": 9.677293207651459e-05, "loss": 0.0552, "step": 12740 }, { "grad_norm": 0.402032732963562, "learning_rate": 9.676562313001173e-05, "loss": 0.0495, "step": 12750 }, { "grad_norm": 0.6205902695655823, "learning_rate": 9.675830619254293e-05, "loss": 0.0594, "step": 12760 }, { "grad_norm": 0.4884233772754669, "learning_rate": 9.675098126535843e-05, "loss": 0.0577, "step": 12770 }, { "grad_norm": 0.4817465841770172, "learning_rate": 9.674364834970988e-05, "loss": 0.0553, "step": 12780 }, { "grad_norm": 0.5414949059486389, "learning_rate": 9.673630744685028e-05, "loss": 0.0548, "step": 12790 }, { "grad_norm": 0.3678781986236572, "learning_rate": 9.672895855803397e-05, "loss": 0.0498, "step": 12800 }, { "grad_norm": 0.5519404411315918, "learning_rate": 9.672160168451667e-05, "loss": 0.0542, "step": 12810 }, { "grad_norm": 0.3657751977443695, "learning_rate": 9.671423682755549e-05, "loss": 0.0518, "step": 12820 }, { "grad_norm": 0.5591874718666077, "learning_rate": 9.670686398840888e-05, "loss": 0.0565, "step": 12830 }, { "grad_norm": 0.3438622057437897, "learning_rate": 9.669948316833664e-05, "loss": 0.0479, "step": 12840 }, { "grad_norm": 0.36001941561698914, "learning_rate": 9.669209436859997e-05, "loss": 0.048, "step": 12850 }, { "grad_norm": 0.4363784193992615, "learning_rate": 9.66846975904614e-05, "loss": 0.0572, "step": 12860 }, { "grad_norm": 0.406424343585968, "learning_rate": 9.667729283518483e-05, "loss": 0.0493, "step": 12870 }, { "grad_norm": 0.44831207394599915, "learning_rate": 9.666988010403557e-05, "loss": 0.0422, "step": 12880 }, { "grad_norm": 0.34375861287117004, "learning_rate": 9.66624593982802e-05, "loss": 0.0538, "step": 12890 }, { "grad_norm": 0.4239714741706848, "learning_rate": 9.665503071918675e-05, "loss": 0.0572, "step": 12900 }, { "grad_norm": 0.45530804991722107, "learning_rate": 9.664759406802456e-05, "loss": 0.0511, "step": 12910 }, { "grad_norm": 0.4781215488910675, "learning_rate": 9.664014944606437e-05, "loss": 0.0466, "step": 12920 }, { "grad_norm": 0.3125785291194916, "learning_rate": 9.663269685457822e-05, "loss": 0.0422, "step": 12930 }, { "grad_norm": 0.6326125264167786, "learning_rate": 9.662523629483962e-05, "loss": 0.0489, "step": 12940 }, { "grad_norm": 0.4334132969379425, "learning_rate": 9.661776776812333e-05, "loss": 0.0425, "step": 12950 }, { "grad_norm": 0.4959748089313507, "learning_rate": 9.661029127570553e-05, "loss": 0.0443, "step": 12960 }, { "grad_norm": 0.4485611617565155, "learning_rate": 9.660280681886373e-05, "loss": 0.0463, "step": 12970 }, { "grad_norm": 0.3815354108810425, "learning_rate": 9.659531439887685e-05, "loss": 0.0444, "step": 12980 }, { "grad_norm": 0.44503018260002136, "learning_rate": 9.658781401702511e-05, "loss": 0.0466, "step": 12990 }, { "grad_norm": 0.3816302716732025, "learning_rate": 9.658030567459015e-05, "loss": 0.0488, "step": 13000 }, { "grad_norm": 0.4347861409187317, "learning_rate": 9.65727893728549e-05, "loss": 0.07, "step": 13010 }, { "grad_norm": 0.5163665413856506, "learning_rate": 9.656526511310375e-05, "loss": 0.0648, "step": 13020 }, { "grad_norm": 0.45969268679618835, "learning_rate": 9.655773289662233e-05, "loss": 0.0682, "step": 13030 }, { "grad_norm": 0.500993549823761, "learning_rate": 9.655019272469772e-05, "loss": 0.0487, "step": 13040 }, { "grad_norm": 0.5975098609924316, "learning_rate": 9.654264459861832e-05, "loss": 0.0456, "step": 13050 }, { "grad_norm": 0.5115871429443359, "learning_rate": 9.653508851967391e-05, "loss": 0.0472, "step": 13060 }, { "grad_norm": 0.38931789994239807, "learning_rate": 9.65275244891556e-05, "loss": 0.0609, "step": 13070 }, { "grad_norm": 0.4399626553058624, "learning_rate": 9.651995250835591e-05, "loss": 0.0473, "step": 13080 }, { "grad_norm": 0.4849909543991089, "learning_rate": 9.651237257856862e-05, "loss": 0.0521, "step": 13090 }, { "grad_norm": 0.4051220118999481, "learning_rate": 9.6504784701089e-05, "loss": 0.0482, "step": 13100 }, { "grad_norm": 0.2836496829986572, "learning_rate": 9.649718887721357e-05, "loss": 0.046, "step": 13110 }, { "grad_norm": 0.4197397828102112, "learning_rate": 9.648958510824028e-05, "loss": 0.0462, "step": 13120 }, { "grad_norm": 0.42242658138275146, "learning_rate": 9.648197339546837e-05, "loss": 0.0525, "step": 13130 }, { "grad_norm": 0.2932666838169098, "learning_rate": 9.647435374019851e-05, "loss": 0.0497, "step": 13140 }, { "grad_norm": 0.37803053855895996, "learning_rate": 9.646672614373266e-05, "loss": 0.0502, "step": 13150 }, { "grad_norm": 0.4012050926685333, "learning_rate": 9.645909060737418e-05, "loss": 0.0391, "step": 13160 }, { "grad_norm": 0.4073188304901123, "learning_rate": 9.645144713242778e-05, "loss": 0.0454, "step": 13170 }, { "grad_norm": 0.46537333726882935, "learning_rate": 9.64437957201995e-05, "loss": 0.0465, "step": 13180 }, { "grad_norm": 0.4425288140773773, "learning_rate": 9.643613637199678e-05, "loss": 0.0412, "step": 13190 }, { "grad_norm": 0.3590709865093231, "learning_rate": 9.642846908912839e-05, "loss": 0.0525, "step": 13200 }, { "grad_norm": 0.4061567187309265, "learning_rate": 9.642079387290444e-05, "loss": 0.0364, "step": 13210 }, { "grad_norm": 0.3705960810184479, "learning_rate": 9.641311072463644e-05, "loss": 0.0429, "step": 13220 }, { "grad_norm": 0.581544816493988, "learning_rate": 9.640541964563722e-05, "loss": 0.0491, "step": 13230 }, { "grad_norm": 0.5175280570983887, "learning_rate": 9.639772063722096e-05, "loss": 0.0584, "step": 13240 }, { "grad_norm": 0.36531198024749756, "learning_rate": 9.639001370070324e-05, "loss": 0.0432, "step": 13250 }, { "grad_norm": 0.4849933981895447, "learning_rate": 9.638229883740095e-05, "loss": 0.0491, "step": 13260 }, { "grad_norm": 0.3879377543926239, "learning_rate": 9.637457604863233e-05, "loss": 0.0457, "step": 13270 }, { "grad_norm": 0.4325227737426758, "learning_rate": 9.636684533571703e-05, "loss": 0.0463, "step": 13280 }, { "grad_norm": 0.3474501073360443, "learning_rate": 9.635910669997599e-05, "loss": 0.0436, "step": 13290 }, { "grad_norm": 0.3876914381980896, "learning_rate": 9.635136014273154e-05, "loss": 0.0455, "step": 13300 }, { "grad_norm": 0.4264110028743744, "learning_rate": 9.634360566530735e-05, "loss": 0.051, "step": 13310 }, { "grad_norm": 0.6488595008850098, "learning_rate": 9.633584326902845e-05, "loss": 0.06, "step": 13320 }, { "grad_norm": 0.3730056881904602, "learning_rate": 9.632807295522124e-05, "loss": 0.0482, "step": 13330 }, { "grad_norm": 0.5143369436264038, "learning_rate": 9.632029472521342e-05, "loss": 0.0468, "step": 13340 }, { "grad_norm": 0.25752004981040955, "learning_rate": 9.631250858033409e-05, "loss": 0.0473, "step": 13350 }, { "grad_norm": 0.3423303961753845, "learning_rate": 9.630471452191371e-05, "loss": 0.0482, "step": 13360 }, { "grad_norm": 0.43088462948799133, "learning_rate": 9.629691255128405e-05, "loss": 0.0529, "step": 13370 }, { "grad_norm": 0.5564923286437988, "learning_rate": 9.628910266977825e-05, "loss": 0.0465, "step": 13380 }, { "grad_norm": 0.40494972467422485, "learning_rate": 9.628128487873083e-05, "loss": 0.0565, "step": 13390 }, { "grad_norm": 0.35353249311447144, "learning_rate": 9.627345917947761e-05, "loss": 0.0559, "step": 13400 }, { "grad_norm": 0.35125473141670227, "learning_rate": 9.626562557335579e-05, "loss": 0.049, "step": 13410 }, { "grad_norm": 0.3305366635322571, "learning_rate": 9.625778406170393e-05, "loss": 0.0525, "step": 13420 }, { "grad_norm": 0.41589823365211487, "learning_rate": 9.624993464586193e-05, "loss": 0.0532, "step": 13430 }, { "grad_norm": 0.40280526876449585, "learning_rate": 9.624207732717105e-05, "loss": 0.0493, "step": 13440 }, { "grad_norm": 0.553024411201477, "learning_rate": 9.623421210697386e-05, "loss": 0.0589, "step": 13450 }, { "grad_norm": 0.2807592749595642, "learning_rate": 9.622633898661434e-05, "loss": 0.0612, "step": 13460 }, { "grad_norm": 0.3414004147052765, "learning_rate": 9.621845796743778e-05, "loss": 0.067, "step": 13470 }, { "grad_norm": 0.29216983914375305, "learning_rate": 9.621056905079082e-05, "loss": 0.0529, "step": 13480 }, { "grad_norm": 0.41028478741645813, "learning_rate": 9.620267223802149e-05, "loss": 0.0583, "step": 13490 }, { "grad_norm": 0.3673085868358612, "learning_rate": 9.619476753047911e-05, "loss": 0.0439, "step": 13500 }, { "grad_norm": 0.40739431977272034, "learning_rate": 9.618685492951438e-05, "loss": 0.0514, "step": 13510 }, { "grad_norm": 0.35084661841392517, "learning_rate": 9.617893443647938e-05, "loss": 0.0583, "step": 13520 }, { "grad_norm": 0.3999849259853363, "learning_rate": 9.617100605272746e-05, "loss": 0.0528, "step": 13530 }, { "grad_norm": 0.35168805718421936, "learning_rate": 9.616306977961338e-05, "loss": 0.0481, "step": 13540 }, { "grad_norm": 0.3374437093734741, "learning_rate": 9.615512561849326e-05, "loss": 0.0394, "step": 13550 }, { "grad_norm": 0.4102657437324524, "learning_rate": 9.61471735707245e-05, "loss": 0.0544, "step": 13560 }, { "grad_norm": 0.3532998859882355, "learning_rate": 9.613921363766592e-05, "loss": 0.0418, "step": 13570 }, { "grad_norm": 0.41592082381248474, "learning_rate": 9.613124582067763e-05, "loss": 0.0486, "step": 13580 }, { "grad_norm": 0.4322201609611511, "learning_rate": 9.612327012112112e-05, "loss": 0.0549, "step": 13590 }, { "grad_norm": 0.4439077079296112, "learning_rate": 9.611528654035921e-05, "loss": 0.0705, "step": 13600 }, { "grad_norm": 0.46936899423599243, "learning_rate": 9.610729507975611e-05, "loss": 0.0467, "step": 13610 }, { "grad_norm": 0.3713732063770294, "learning_rate": 9.609929574067731e-05, "loss": 0.0667, "step": 13620 }, { "grad_norm": 0.31264403462409973, "learning_rate": 9.609128852448967e-05, "loss": 0.047, "step": 13630 }, { "grad_norm": 0.40778160095214844, "learning_rate": 9.608327343256143e-05, "loss": 0.0481, "step": 13640 }, { "grad_norm": 0.42713606357574463, "learning_rate": 9.607525046626216e-05, "loss": 0.0557, "step": 13650 }, { "grad_norm": 0.5862718820571899, "learning_rate": 9.606721962696272e-05, "loss": 0.0568, "step": 13660 }, { "grad_norm": 0.48016878962516785, "learning_rate": 9.60591809160354e-05, "loss": 0.0461, "step": 13670 }, { "grad_norm": 0.3830547630786896, "learning_rate": 9.605113433485378e-05, "loss": 0.0471, "step": 13680 }, { "grad_norm": 0.5757759213447571, "learning_rate": 9.604307988479279e-05, "loss": 0.0497, "step": 13690 }, { "grad_norm": 0.4723908603191376, "learning_rate": 9.603501756722876e-05, "loss": 0.0512, "step": 13700 }, { "grad_norm": 0.3885815143585205, "learning_rate": 9.602694738353927e-05, "loss": 0.0501, "step": 13710 }, { "grad_norm": 0.37895599007606506, "learning_rate": 9.601886933510331e-05, "loss": 0.0442, "step": 13720 }, { "grad_norm": 0.3326095938682556, "learning_rate": 9.60107834233012e-05, "loss": 0.0379, "step": 13730 }, { "grad_norm": 0.377339243888855, "learning_rate": 9.60026896495146e-05, "loss": 0.0414, "step": 13740 }, { "grad_norm": 0.2243943065404892, "learning_rate": 9.599458801512652e-05, "loss": 0.039, "step": 13750 }, { "grad_norm": 0.3575519919395447, "learning_rate": 9.598647852152129e-05, "loss": 0.0672, "step": 13760 }, { "grad_norm": 0.3920873701572418, "learning_rate": 9.597836117008462e-05, "loss": 0.0476, "step": 13770 }, { "grad_norm": 0.28096500039100647, "learning_rate": 9.597023596220356e-05, "loss": 0.0445, "step": 13780 }, { "grad_norm": 0.4927907884120941, "learning_rate": 9.596210289926643e-05, "loss": 0.0491, "step": 13790 }, { "grad_norm": 0.3872702717781067, "learning_rate": 9.5953961982663e-05, "loss": 0.044, "step": 13800 }, { "grad_norm": 0.6126062273979187, "learning_rate": 9.594581321378431e-05, "loss": 0.054, "step": 13810 }, { "grad_norm": 0.45314663648605347, "learning_rate": 9.593765659402276e-05, "loss": 0.0527, "step": 13820 }, { "grad_norm": 0.40250343084335327, "learning_rate": 9.59294921247721e-05, "loss": 0.0507, "step": 13830 }, { "grad_norm": 0.3826405704021454, "learning_rate": 9.59213198074274e-05, "loss": 0.0478, "step": 13840 }, { "grad_norm": 0.48315611481666565, "learning_rate": 9.59131396433851e-05, "loss": 0.0411, "step": 13850 }, { "grad_norm": 0.4225694537162781, "learning_rate": 9.590495163404297e-05, "loss": 0.0506, "step": 13860 }, { "grad_norm": 0.32770848274230957, "learning_rate": 9.589675578080009e-05, "loss": 0.0454, "step": 13870 }, { "grad_norm": 0.35880258679389954, "learning_rate": 9.588855208505694e-05, "loss": 0.0377, "step": 13880 }, { "grad_norm": 0.4800691604614258, "learning_rate": 9.588034054821529e-05, "loss": 0.0534, "step": 13890 }, { "grad_norm": 0.33709996938705444, "learning_rate": 9.587212117167826e-05, "loss": 0.0352, "step": 13900 }, { "grad_norm": 0.395236611366272, "learning_rate": 9.586389395685033e-05, "loss": 0.0461, "step": 13910 }, { "grad_norm": 0.48640817403793335, "learning_rate": 9.585565890513733e-05, "loss": 0.0519, "step": 13920 }, { "grad_norm": 0.3849732279777527, "learning_rate": 9.584741601794636e-05, "loss": 0.0486, "step": 13930 }, { "grad_norm": 0.5261255502700806, "learning_rate": 9.58391652966859e-05, "loss": 0.0476, "step": 13940 }, { "grad_norm": 0.4128359854221344, "learning_rate": 9.583090674276583e-05, "loss": 0.0466, "step": 13950 }, { "grad_norm": 0.3698306977748871, "learning_rate": 9.582264035759726e-05, "loss": 0.0452, "step": 13960 }, { "grad_norm": 0.30111536383628845, "learning_rate": 9.58143661425927e-05, "loss": 0.0595, "step": 13970 }, { "grad_norm": 0.45516085624694824, "learning_rate": 9.580608409916601e-05, "loss": 0.0582, "step": 13980 }, { "grad_norm": 0.36068016290664673, "learning_rate": 9.579779422873233e-05, "loss": 0.0532, "step": 13990 }, { "grad_norm": 0.40655380487442017, "learning_rate": 9.578949653270819e-05, "loss": 0.0511, "step": 14000 }, { "grad_norm": 0.38917431235313416, "learning_rate": 9.578119101251144e-05, "loss": 0.0437, "step": 14010 }, { "grad_norm": 0.3813334107398987, "learning_rate": 9.577287766956127e-05, "loss": 0.0585, "step": 14020 }, { "grad_norm": 0.4013017416000366, "learning_rate": 9.57645565052782e-05, "loss": 0.0506, "step": 14030 }, { "grad_norm": 0.4230312705039978, "learning_rate": 9.575622752108407e-05, "loss": 0.0418, "step": 14040 }, { "grad_norm": 0.3719237744808197, "learning_rate": 9.57478907184021e-05, "loss": 0.0407, "step": 14050 }, { "grad_norm": 0.275958389043808, "learning_rate": 9.573954609865681e-05, "loss": 0.0502, "step": 14060 }, { "grad_norm": 0.36407965421676636, "learning_rate": 9.573119366327408e-05, "loss": 0.0464, "step": 14070 }, { "grad_norm": 0.35692912340164185, "learning_rate": 9.57228334136811e-05, "loss": 0.047, "step": 14080 }, { "grad_norm": 0.44371405243873596, "learning_rate": 9.571446535130641e-05, "loss": 0.0466, "step": 14090 }, { "grad_norm": 0.3650364875793457, "learning_rate": 9.570608947757988e-05, "loss": 0.0514, "step": 14100 }, { "grad_norm": 0.41035810112953186, "learning_rate": 9.569770579393274e-05, "loss": 0.0459, "step": 14110 }, { "grad_norm": 0.34269508719444275, "learning_rate": 9.56893143017975e-05, "loss": 0.0466, "step": 14120 }, { "grad_norm": 0.3412228524684906, "learning_rate": 9.568091500260806e-05, "loss": 0.0513, "step": 14130 }, { "grad_norm": 0.30729520320892334, "learning_rate": 9.567250789779961e-05, "loss": 0.051, "step": 14140 }, { "grad_norm": 0.3763951063156128, "learning_rate": 9.566409298880872e-05, "loss": 0.0405, "step": 14150 }, { "grad_norm": 0.4269770085811615, "learning_rate": 9.565567027707326e-05, "loss": 0.0631, "step": 14160 }, { "grad_norm": 0.35958999395370483, "learning_rate": 9.56472397640324e-05, "loss": 0.0478, "step": 14170 }, { "grad_norm": 0.31860706210136414, "learning_rate": 9.563880145112675e-05, "loss": 0.0511, "step": 14180 }, { "grad_norm": 0.35230764746665955, "learning_rate": 9.563035533979814e-05, "loss": 0.05, "step": 14190 }, { "grad_norm": 0.3362249732017517, "learning_rate": 9.562190143148981e-05, "loss": 0.0466, "step": 14200 }, { "grad_norm": 0.3719567060470581, "learning_rate": 9.561343972764627e-05, "loss": 0.0547, "step": 14210 }, { "grad_norm": 0.461235910654068, "learning_rate": 9.560497022971343e-05, "loss": 0.0609, "step": 14220 }, { "grad_norm": 0.28196635842323303, "learning_rate": 9.559649293913847e-05, "loss": 0.0501, "step": 14230 }, { "grad_norm": 0.41258928179740906, "learning_rate": 9.558800785736993e-05, "loss": 0.0503, "step": 14240 }, { "grad_norm": 0.4199092388153076, "learning_rate": 9.557951498585767e-05, "loss": 0.0567, "step": 14250 }, { "grad_norm": 0.42940470576286316, "learning_rate": 9.557101432605293e-05, "loss": 0.0476, "step": 14260 }, { "grad_norm": 0.5870382785797119, "learning_rate": 9.556250587940818e-05, "loss": 0.0623, "step": 14270 }, { "grad_norm": 0.4043969511985779, "learning_rate": 9.555398964737734e-05, "loss": 0.0476, "step": 14280 }, { "grad_norm": 0.439555823802948, "learning_rate": 9.554546563141555e-05, "loss": 0.0422, "step": 14290 }, { "grad_norm": 0.3517785370349884, "learning_rate": 9.553693383297937e-05, "loss": 0.048, "step": 14300 }, { "grad_norm": 0.41349801421165466, "learning_rate": 9.552839425352663e-05, "loss": 0.0416, "step": 14310 }, { "grad_norm": 0.37404245138168335, "learning_rate": 9.551984689451652e-05, "loss": 0.0557, "step": 14320 }, { "grad_norm": 0.4774573743343353, "learning_rate": 9.551129175740953e-05, "loss": 0.0428, "step": 14330 }, { "grad_norm": 0.28095948696136475, "learning_rate": 9.550272884366754e-05, "loss": 0.0477, "step": 14340 }, { "grad_norm": 0.441459059715271, "learning_rate": 9.549415815475369e-05, "loss": 0.0461, "step": 14350 }, { "grad_norm": 0.3844626545906067, "learning_rate": 9.548557969213247e-05, "loss": 0.035, "step": 14360 }, { "grad_norm": 0.30973076820373535, "learning_rate": 9.547699345726972e-05, "loss": 0.0422, "step": 14370 }, { "grad_norm": 0.3486696779727936, "learning_rate": 9.546839945163257e-05, "loss": 0.0442, "step": 14380 }, { "grad_norm": 0.4203062355518341, "learning_rate": 9.545979767668953e-05, "loss": 0.0603, "step": 14390 }, { "grad_norm": 0.3897950053215027, "learning_rate": 9.54511881339104e-05, "loss": 0.0518, "step": 14400 }, { "grad_norm": 0.37809720635414124, "learning_rate": 9.54425708247663e-05, "loss": 0.0445, "step": 14410 }, { "grad_norm": 0.43537071347236633, "learning_rate": 9.543394575072972e-05, "loss": 0.0446, "step": 14420 }, { "grad_norm": 0.3312765061855316, "learning_rate": 9.542531291327441e-05, "loss": 0.0503, "step": 14430 }, { "grad_norm": 0.5354935526847839, "learning_rate": 9.541667231387552e-05, "loss": 0.044, "step": 14440 }, { "grad_norm": 0.4314918518066406, "learning_rate": 9.540802395400949e-05, "loss": 0.0496, "step": 14450 }, { "grad_norm": 0.5341212153434753, "learning_rate": 9.539936783515406e-05, "loss": 0.0544, "step": 14460 }, { "grad_norm": 0.43150296807289124, "learning_rate": 9.539070395878835e-05, "loss": 0.0589, "step": 14470 }, { "grad_norm": 0.4805160164833069, "learning_rate": 9.538203232639277e-05, "loss": 0.0487, "step": 14480 }, { "grad_norm": 0.43065714836120605, "learning_rate": 9.537335293944907e-05, "loss": 0.047, "step": 14490 }, { "grad_norm": 0.41533347964286804, "learning_rate": 9.536466579944032e-05, "loss": 0.07, "step": 14500 }, { "grad_norm": 0.35014376044273376, "learning_rate": 9.535597090785091e-05, "loss": 0.0509, "step": 14510 }, { "grad_norm": 0.40082502365112305, "learning_rate": 9.534726826616656e-05, "loss": 0.039, "step": 14520 }, { "grad_norm": 0.33705925941467285, "learning_rate": 9.53385578758743e-05, "loss": 0.071, "step": 14530 }, { "grad_norm": 0.4502198398113251, "learning_rate": 9.532983973846252e-05, "loss": 0.0487, "step": 14540 }, { "grad_norm": 0.47115448117256165, "learning_rate": 9.53211138554209e-05, "loss": 0.0484, "step": 14550 }, { "grad_norm": 0.35732030868530273, "learning_rate": 9.531238022824047e-05, "loss": 0.0506, "step": 14560 }, { "grad_norm": 0.3419415354728699, "learning_rate": 9.530363885841355e-05, "loss": 0.0412, "step": 14570 }, { "grad_norm": 0.36515143513679504, "learning_rate": 9.52948897474338e-05, "loss": 0.0474, "step": 14580 }, { "grad_norm": 0.42453697323799133, "learning_rate": 9.528613289679622e-05, "loss": 0.0585, "step": 14590 }, { "grad_norm": 0.38910484313964844, "learning_rate": 9.52773683079971e-05, "loss": 0.0571, "step": 14600 }, { "grad_norm": 0.312469482421875, "learning_rate": 9.526859598253407e-05, "loss": 0.0417, "step": 14610 }, { "grad_norm": 0.41147977113723755, "learning_rate": 9.525981592190609e-05, "loss": 0.0465, "step": 14620 }, { "grad_norm": 0.46917054057121277, "learning_rate": 9.525102812761342e-05, "loss": 0.053, "step": 14630 }, { "grad_norm": 0.3898388743400574, "learning_rate": 9.524223260115768e-05, "loss": 0.0404, "step": 14640 }, { "grad_norm": 0.3458184599876404, "learning_rate": 9.523342934404175e-05, "loss": 0.0466, "step": 14650 }, { "grad_norm": 0.38495591282844543, "learning_rate": 9.522461835776989e-05, "loss": 0.0492, "step": 14660 }, { "grad_norm": 0.41919761896133423, "learning_rate": 9.521579964384764e-05, "loss": 0.062, "step": 14670 }, { "grad_norm": 0.5257139801979065, "learning_rate": 9.52069732037819e-05, "loss": 0.0448, "step": 14680 }, { "grad_norm": 0.32991334795951843, "learning_rate": 9.519813903908083e-05, "loss": 0.0452, "step": 14690 }, { "grad_norm": 0.3379595875740051, "learning_rate": 9.5189297151254e-05, "loss": 0.049, "step": 14700 }, { "grad_norm": 0.6116548180580139, "learning_rate": 9.518044754181218e-05, "loss": 0.0611, "step": 14710 }, { "grad_norm": 0.30249840021133423, "learning_rate": 9.51715902122676e-05, "loss": 0.0494, "step": 14720 }, { "grad_norm": 0.3663947284221649, "learning_rate": 9.516272516413368e-05, "loss": 0.0495, "step": 14730 }, { "grad_norm": 0.2631785571575165, "learning_rate": 9.515385239892525e-05, "loss": 0.0529, "step": 14740 }, { "grad_norm": 0.4310569167137146, "learning_rate": 9.514497191815839e-05, "loss": 0.0534, "step": 14750 }, { "grad_norm": 0.4983830153942108, "learning_rate": 9.513608372335055e-05, "loss": 0.0565, "step": 14760 }, { "grad_norm": 0.43089038133621216, "learning_rate": 9.512718781602045e-05, "loss": 0.0497, "step": 14770 }, { "grad_norm": 0.45937880873680115, "learning_rate": 9.511828419768823e-05, "loss": 0.0501, "step": 14780 }, { "grad_norm": 0.31677067279815674, "learning_rate": 9.510937286987521e-05, "loss": 0.0401, "step": 14790 }, { "grad_norm": 0.41055983304977417, "learning_rate": 9.510045383410408e-05, "loss": 0.0507, "step": 14800 }, { "grad_norm": 0.34889960289001465, "learning_rate": 9.509152709189892e-05, "loss": 0.0431, "step": 14810 }, { "grad_norm": 0.4375223219394684, "learning_rate": 9.508259264478504e-05, "loss": 0.0431, "step": 14820 }, { "grad_norm": 0.4926651418209076, "learning_rate": 9.507365049428909e-05, "loss": 0.0636, "step": 14830 }, { "grad_norm": 0.4076637327671051, "learning_rate": 9.506470064193902e-05, "loss": 0.0631, "step": 14840 }, { "grad_norm": 0.5131556987762451, "learning_rate": 9.505574308926414e-05, "loss": 0.0682, "step": 14850 }, { "grad_norm": 0.4786784052848816, "learning_rate": 9.504677783779505e-05, "loss": 0.0447, "step": 14860 }, { "grad_norm": 0.4247991740703583, "learning_rate": 9.503780488906365e-05, "loss": 0.0479, "step": 14870 }, { "grad_norm": 0.4948933720588684, "learning_rate": 9.502882424460319e-05, "loss": 0.0541, "step": 14880 }, { "grad_norm": 0.9383213520050049, "learning_rate": 9.501983590594821e-05, "loss": 0.0563, "step": 14890 }, { "grad_norm": 0.5537415146827698, "learning_rate": 9.501083987463455e-05, "loss": 0.0507, "step": 14900 }, { "grad_norm": 0.3960493505001068, "learning_rate": 9.500183615219942e-05, "loss": 0.0486, "step": 14910 }, { "grad_norm": 0.4254415035247803, "learning_rate": 9.49928247401813e-05, "loss": 0.0455, "step": 14920 }, { "grad_norm": 0.5638425350189209, "learning_rate": 9.498380564011997e-05, "loss": 0.0545, "step": 14930 }, { "grad_norm": 0.44203120470046997, "learning_rate": 9.497477885355656e-05, "loss": 0.0594, "step": 14940 }, { "grad_norm": 0.5479838252067566, "learning_rate": 9.496574438203353e-05, "loss": 0.0601, "step": 14950 }, { "grad_norm": 0.34689924120903015, "learning_rate": 9.495670222709459e-05, "loss": 0.0535, "step": 14960 }, { "grad_norm": 0.5044149160385132, "learning_rate": 9.494765239028483e-05, "loss": 0.0574, "step": 14970 }, { "grad_norm": 0.4121340811252594, "learning_rate": 9.493859487315057e-05, "loss": 0.0542, "step": 14980 }, { "grad_norm": 0.41954219341278076, "learning_rate": 9.492952967723953e-05, "loss": 0.0499, "step": 14990 }, { "grad_norm": 0.3499198853969574, "learning_rate": 9.492045680410068e-05, "loss": 0.0542, "step": 15000 }, { "grad_norm": 0.4652712941169739, "learning_rate": 9.491137625528436e-05, "loss": 0.0418, "step": 15010 }, { "grad_norm": 0.41413071751594543, "learning_rate": 9.490228803234215e-05, "loss": 0.0409, "step": 15020 }, { "grad_norm": 0.960139811038971, "learning_rate": 9.489319213682701e-05, "loss": 0.0451, "step": 15030 }, { "grad_norm": 0.35239869356155396, "learning_rate": 9.488408857029316e-05, "loss": 0.0589, "step": 15040 }, { "grad_norm": 0.42842698097229004, "learning_rate": 9.487497733429616e-05, "loss": 0.0619, "step": 15050 }, { "grad_norm": 0.5162450671195984, "learning_rate": 9.486585843039286e-05, "loss": 0.0648, "step": 15060 }, { "grad_norm": 0.25464075803756714, "learning_rate": 9.485673186014143e-05, "loss": 0.0491, "step": 15070 }, { "grad_norm": 0.3871200382709503, "learning_rate": 9.484759762510137e-05, "loss": 0.0495, "step": 15080 }, { "grad_norm": 0.6277713179588318, "learning_rate": 9.483845572683346e-05, "loss": 0.0621, "step": 15090 }, { "grad_norm": 0.41140130162239075, "learning_rate": 9.48293061668998e-05, "loss": 0.047, "step": 15100 }, { "grad_norm": 0.44160765409469604, "learning_rate": 9.48201489468638e-05, "loss": 0.0541, "step": 15110 }, { "grad_norm": 0.3089779019355774, "learning_rate": 9.481098406829016e-05, "loss": 0.0477, "step": 15120 }, { "grad_norm": 0.47231221199035645, "learning_rate": 9.480181153274495e-05, "loss": 0.0455, "step": 15130 }, { "grad_norm": 0.339372843503952, "learning_rate": 9.479263134179548e-05, "loss": 0.0413, "step": 15140 }, { "grad_norm": 0.6150843501091003, "learning_rate": 9.478344349701039e-05, "loss": 0.0573, "step": 15150 }, { "grad_norm": 0.32713583111763, "learning_rate": 9.477424799995964e-05, "loss": 0.0525, "step": 15160 }, { "grad_norm": 0.45450273156166077, "learning_rate": 9.476504485221448e-05, "loss": 0.0499, "step": 15170 }, { "grad_norm": 0.35841459035873413, "learning_rate": 9.475583405534748e-05, "loss": 0.0473, "step": 15180 }, { "grad_norm": 0.4086077809333801, "learning_rate": 9.474661561093251e-05, "loss": 0.0427, "step": 15190 }, { "grad_norm": 0.4135025441646576, "learning_rate": 9.473738952054478e-05, "loss": 0.0523, "step": 15200 }, { "grad_norm": 0.4020475447177887, "learning_rate": 9.472815578576073e-05, "loss": 0.043, "step": 15210 }, { "grad_norm": 0.5021268725395203, "learning_rate": 9.471891440815817e-05, "loss": 0.0545, "step": 15220 }, { "grad_norm": 0.41970399022102356, "learning_rate": 9.470966538931621e-05, "loss": 0.0397, "step": 15230 }, { "grad_norm": 0.32239797711372375, "learning_rate": 9.470040873081525e-05, "loss": 0.0452, "step": 15240 }, { "grad_norm": 0.5036735534667969, "learning_rate": 9.469114443423698e-05, "loss": 0.0617, "step": 15250 }, { "grad_norm": 0.4329770505428314, "learning_rate": 9.468187250116445e-05, "loss": 0.0554, "step": 15260 }, { "grad_norm": 0.34834107756614685, "learning_rate": 9.467259293318197e-05, "loss": 0.0433, "step": 15270 }, { "grad_norm": 0.5051876902580261, "learning_rate": 9.466330573187514e-05, "loss": 0.0466, "step": 15280 }, { "grad_norm": 0.4659854471683502, "learning_rate": 9.46540108988309e-05, "loss": 0.0626, "step": 15290 }, { "grad_norm": 0.5298144817352295, "learning_rate": 9.46447084356375e-05, "loss": 0.0509, "step": 15300 }, { "grad_norm": 0.4605858623981476, "learning_rate": 9.463539834388447e-05, "loss": 0.0558, "step": 15310 }, { "grad_norm": 0.404693603515625, "learning_rate": 9.462608062516263e-05, "loss": 0.0405, "step": 15320 }, { "grad_norm": 0.4100753962993622, "learning_rate": 9.461675528106413e-05, "loss": 0.0448, "step": 15330 }, { "grad_norm": 0.3448340594768524, "learning_rate": 9.460742231318244e-05, "loss": 0.0473, "step": 15340 }, { "grad_norm": 0.3479556739330292, "learning_rate": 9.459808172311229e-05, "loss": 0.0522, "step": 15350 }, { "grad_norm": 0.46060290932655334, "learning_rate": 9.458873351244972e-05, "loss": 0.0431, "step": 15360 }, { "grad_norm": 0.4061931371688843, "learning_rate": 9.457937768279211e-05, "loss": 0.0488, "step": 15370 }, { "grad_norm": 0.4031650424003601, "learning_rate": 9.45700142357381e-05, "loss": 0.0467, "step": 15380 }, { "grad_norm": 0.48161330819129944, "learning_rate": 9.456064317288765e-05, "loss": 0.0427, "step": 15390 }, { "grad_norm": 0.4094424247741699, "learning_rate": 9.455126449584201e-05, "loss": 0.045, "step": 15400 }, { "grad_norm": 0.33341264724731445, "learning_rate": 9.454187820620375e-05, "loss": 0.0358, "step": 15410 }, { "grad_norm": 0.3696576952934265, "learning_rate": 9.453248430557673e-05, "loss": 0.0514, "step": 15420 }, { "grad_norm": 0.4652194380760193, "learning_rate": 9.452308279556611e-05, "loss": 0.0678, "step": 15430 }, { "grad_norm": 0.3781914710998535, "learning_rate": 9.451367367777835e-05, "loss": 0.0473, "step": 15440 }, { "grad_norm": 0.5419948697090149, "learning_rate": 9.450425695382122e-05, "loss": 0.0597, "step": 15450 }, { "grad_norm": 0.3342730402946472, "learning_rate": 9.449483262530375e-05, "loss": 0.046, "step": 15460 }, { "grad_norm": 0.3465645909309387, "learning_rate": 9.448540069383633e-05, "loss": 0.0391, "step": 15470 }, { "grad_norm": 0.3617420196533203, "learning_rate": 9.447596116103061e-05, "loss": 0.0501, "step": 15480 }, { "grad_norm": 0.4771029055118561, "learning_rate": 9.446651402849955e-05, "loss": 0.0522, "step": 15490 }, { "grad_norm": 0.43842658400535583, "learning_rate": 9.44570592978574e-05, "loss": 0.0436, "step": 15500 }, { "grad_norm": 0.3321513831615448, "learning_rate": 9.444759697071972e-05, "loss": 0.0492, "step": 15510 }, { "grad_norm": 0.340935081243515, "learning_rate": 9.443812704870336e-05, "loss": 0.0383, "step": 15520 }, { "grad_norm": 0.41091233491897583, "learning_rate": 9.442864953342649e-05, "loss": 0.0394, "step": 15530 }, { "grad_norm": 0.43434372544288635, "learning_rate": 9.441916442650852e-05, "loss": 0.0469, "step": 15540 }, { "grad_norm": 0.345905601978302, "learning_rate": 9.440967172957023e-05, "loss": 0.043, "step": 15550 }, { "grad_norm": 0.44068658351898193, "learning_rate": 9.440017144423364e-05, "loss": 0.0527, "step": 15560 }, { "grad_norm": 0.42031121253967285, "learning_rate": 9.439066357212209e-05, "loss": 0.052, "step": 15570 }, { "grad_norm": 0.348583847284317, "learning_rate": 9.438114811486022e-05, "loss": 0.0582, "step": 15580 }, { "grad_norm": 0.36845120787620544, "learning_rate": 9.4371625074074e-05, "loss": 0.0437, "step": 15590 }, { "grad_norm": 0.4044170677661896, "learning_rate": 9.436209445139059e-05, "loss": 0.0368, "step": 15600 }, { "grad_norm": 0.32234156131744385, "learning_rate": 9.435255624843855e-05, "loss": 0.0484, "step": 15610 }, { "grad_norm": 0.23194867372512817, "learning_rate": 9.43430104668477e-05, "loss": 0.0401, "step": 15620 }, { "grad_norm": 0.49449288845062256, "learning_rate": 9.433345710824914e-05, "loss": 0.0491, "step": 15630 }, { "grad_norm": 0.4833472967147827, "learning_rate": 9.432389617427529e-05, "loss": 0.0497, "step": 15640 }, { "grad_norm": 0.289554238319397, "learning_rate": 9.431432766655984e-05, "loss": 0.0486, "step": 15650 }, { "grad_norm": 0.5020430088043213, "learning_rate": 9.430475158673778e-05, "loss": 0.0496, "step": 15660 }, { "grad_norm": 0.4437330961227417, "learning_rate": 9.429516793644542e-05, "loss": 0.0421, "step": 15670 }, { "grad_norm": 0.39976540207862854, "learning_rate": 9.428557671732034e-05, "loss": 0.0395, "step": 15680 }, { "grad_norm": 0.3531879782676697, "learning_rate": 9.42759779310014e-05, "loss": 0.0433, "step": 15690 }, { "grad_norm": 0.38674014806747437, "learning_rate": 9.426637157912879e-05, "loss": 0.0479, "step": 15700 }, { "grad_norm": 0.38119420409202576, "learning_rate": 9.425675766334397e-05, "loss": 0.0523, "step": 15710 }, { "grad_norm": 0.43723323941230774, "learning_rate": 9.424713618528968e-05, "loss": 0.0445, "step": 15720 }, { "grad_norm": 0.4838997423648834, "learning_rate": 9.423750714661e-05, "loss": 0.0601, "step": 15730 }, { "grad_norm": 0.37525978684425354, "learning_rate": 9.422787054895022e-05, "loss": 0.0451, "step": 15740 }, { "grad_norm": 0.40821653604507446, "learning_rate": 9.4218226393957e-05, "loss": 0.0521, "step": 15750 }, { "grad_norm": 0.2976524233818054, "learning_rate": 9.420857468327828e-05, "loss": 0.0472, "step": 15760 }, { "grad_norm": 0.5709309577941895, "learning_rate": 9.419891541856323e-05, "loss": 0.0554, "step": 15770 }, { "grad_norm": 0.4562196433544159, "learning_rate": 9.41892486014624e-05, "loss": 0.0497, "step": 15780 }, { "grad_norm": 0.46806320548057556, "learning_rate": 9.417957423362756e-05, "loss": 0.0542, "step": 15790 }, { "grad_norm": 0.39357680082321167, "learning_rate": 9.416989231671178e-05, "loss": 0.0455, "step": 15800 }, { "grad_norm": 0.2845017910003662, "learning_rate": 9.416020285236946e-05, "loss": 0.0458, "step": 15810 }, { "grad_norm": 0.400729775428772, "learning_rate": 9.415050584225626e-05, "loss": 0.0535, "step": 15820 }, { "grad_norm": 0.4101080298423767, "learning_rate": 9.414080128802914e-05, "loss": 0.0425, "step": 15830 }, { "grad_norm": 0.33988264203071594, "learning_rate": 9.413108919134632e-05, "loss": 0.0427, "step": 15840 }, { "grad_norm": 0.3702412247657776, "learning_rate": 9.412136955386734e-05, "loss": 0.0477, "step": 15850 }, { "grad_norm": 0.28774383664131165, "learning_rate": 9.411164237725303e-05, "loss": 0.0426, "step": 15860 }, { "grad_norm": 0.4400995671749115, "learning_rate": 9.41019076631655e-05, "loss": 0.0488, "step": 15870 }, { "grad_norm": 0.38163620233535767, "learning_rate": 9.409216541326815e-05, "loss": 0.0502, "step": 15880 }, { "grad_norm": 0.39709609746932983, "learning_rate": 9.408241562922564e-05, "loss": 0.051, "step": 15890 }, { "grad_norm": 0.39012956619262695, "learning_rate": 9.407265831270395e-05, "loss": 0.0377, "step": 15900 }, { "grad_norm": 0.38051682710647583, "learning_rate": 9.406289346537035e-05, "loss": 0.0496, "step": 15910 }, { "grad_norm": 0.5675883293151855, "learning_rate": 9.405312108889339e-05, "loss": 0.0516, "step": 15920 }, { "grad_norm": 0.4480844736099243, "learning_rate": 9.404334118494288e-05, "loss": 0.051, "step": 15930 }, { "grad_norm": 0.2967241406440735, "learning_rate": 9.403355375518995e-05, "loss": 0.0475, "step": 15940 }, { "grad_norm": 0.44353607296943665, "learning_rate": 9.4023758801307e-05, "loss": 0.04, "step": 15950 }, { "grad_norm": 0.36436551809310913, "learning_rate": 9.401395632496774e-05, "loss": 0.043, "step": 15960 }, { "grad_norm": 0.4211932122707367, "learning_rate": 9.400414632784711e-05, "loss": 0.0515, "step": 15970 }, { "grad_norm": 0.38510292768478394, "learning_rate": 9.39943288116214e-05, "loss": 0.0451, "step": 15980 }, { "grad_norm": 0.3534134030342102, "learning_rate": 9.398450377796815e-05, "loss": 0.0557, "step": 15990 }, { "grad_norm": 0.3480151891708374, "learning_rate": 9.397467122856616e-05, "loss": 0.0458, "step": 16000 }, { "grad_norm": 0.4515029788017273, "learning_rate": 9.396483116509558e-05, "loss": 0.0484, "step": 16010 }, { "grad_norm": 0.4168718457221985, "learning_rate": 9.39549835892378e-05, "loss": 0.0441, "step": 16020 }, { "grad_norm": 0.6776199340820312, "learning_rate": 9.39451285026755e-05, "loss": 0.0532, "step": 16030 }, { "grad_norm": 0.3284358084201813, "learning_rate": 9.393526590709262e-05, "loss": 0.0437, "step": 16040 }, { "grad_norm": 0.44154778122901917, "learning_rate": 9.392539580417444e-05, "loss": 0.0549, "step": 16050 }, { "grad_norm": 0.3439365327358246, "learning_rate": 9.391551819560747e-05, "loss": 0.0493, "step": 16060 }, { "grad_norm": 0.3662562370300293, "learning_rate": 9.390563308307955e-05, "loss": 0.0446, "step": 16070 }, { "grad_norm": 0.4558155834674835, "learning_rate": 9.389574046827974e-05, "loss": 0.0481, "step": 16080 }, { "grad_norm": 0.5193330645561218, "learning_rate": 9.388584035289845e-05, "loss": 0.0531, "step": 16090 }, { "grad_norm": 0.4920704960823059, "learning_rate": 9.387593273862732e-05, "loss": 0.0482, "step": 16100 }, { "grad_norm": 0.34582945704460144, "learning_rate": 9.386601762715929e-05, "loss": 0.042, "step": 16110 }, { "grad_norm": 0.418066143989563, "learning_rate": 9.38560950201886e-05, "loss": 0.0558, "step": 16120 }, { "grad_norm": 0.5332428812980652, "learning_rate": 9.384616491941071e-05, "loss": 0.0465, "step": 16130 }, { "grad_norm": 0.4025322198867798, "learning_rate": 9.383622732652245e-05, "loss": 0.0456, "step": 16140 }, { "grad_norm": 0.7799068093299866, "learning_rate": 9.382628224322187e-05, "loss": 0.0577, "step": 16150 }, { "grad_norm": 0.36506378650665283, "learning_rate": 9.381632967120829e-05, "loss": 0.0501, "step": 16160 }, { "grad_norm": 0.45038944482803345, "learning_rate": 9.380636961218235e-05, "loss": 0.0556, "step": 16170 }, { "grad_norm": 0.3291938006877899, "learning_rate": 9.379640206784597e-05, "loss": 0.0435, "step": 16180 }, { "grad_norm": 0.41967153549194336, "learning_rate": 9.378642703990229e-05, "loss": 0.0504, "step": 16190 }, { "grad_norm": 0.4110299348831177, "learning_rate": 9.37764445300558e-05, "loss": 0.0434, "step": 16200 }, { "grad_norm": 0.5220197439193726, "learning_rate": 9.376645454001222e-05, "loss": 0.0546, "step": 16210 }, { "grad_norm": 0.42515185475349426, "learning_rate": 9.375645707147858e-05, "loss": 0.0446, "step": 16220 }, { "grad_norm": 0.3396822214126587, "learning_rate": 9.374645212616316e-05, "loss": 0.0609, "step": 16230 }, { "grad_norm": 0.5092000365257263, "learning_rate": 9.373643970577555e-05, "loss": 0.0571, "step": 16240 }, { "grad_norm": 0.6701273322105408, "learning_rate": 9.372641981202659e-05, "loss": 0.0464, "step": 16250 }, { "grad_norm": 0.5061848759651184, "learning_rate": 9.37163924466284e-05, "loss": 0.0501, "step": 16260 }, { "grad_norm": 0.31880897283554077, "learning_rate": 9.370635761129438e-05, "loss": 0.045, "step": 16270 }, { "grad_norm": 0.3371732831001282, "learning_rate": 9.36963153077392e-05, "loss": 0.0564, "step": 16280 }, { "grad_norm": 0.4113680124282837, "learning_rate": 9.368626553767888e-05, "loss": 0.0523, "step": 16290 }, { "grad_norm": 0.4088797867298126, "learning_rate": 9.367620830283057e-05, "loss": 0.0406, "step": 16300 }, { "grad_norm": 0.4081205129623413, "learning_rate": 9.366614360491281e-05, "loss": 0.0526, "step": 16310 }, { "grad_norm": 0.41397958993911743, "learning_rate": 9.365607144564539e-05, "loss": 0.048, "step": 16320 }, { "grad_norm": 0.4578011631965637, "learning_rate": 9.364599182674934e-05, "loss": 0.0546, "step": 16330 }, { "grad_norm": 0.3494669497013092, "learning_rate": 9.3635904749947e-05, "loss": 0.0435, "step": 16340 }, { "grad_norm": 0.3170452117919922, "learning_rate": 9.362581021696202e-05, "loss": 0.043, "step": 16350 }, { "grad_norm": 0.44718217849731445, "learning_rate": 9.361570822951921e-05, "loss": 0.0545, "step": 16360 }, { "grad_norm": 0.3870565593242645, "learning_rate": 9.360559878934476e-05, "loss": 0.0563, "step": 16370 }, { "grad_norm": 0.5148109197616577, "learning_rate": 9.359548189816611e-05, "loss": 0.05, "step": 16380 }, { "grad_norm": 0.5685315132141113, "learning_rate": 9.358535755771193e-05, "loss": 0.0594, "step": 16390 }, { "grad_norm": 0.3671525716781616, "learning_rate": 9.357522576971221e-05, "loss": 0.0542, "step": 16400 }, { "grad_norm": 0.3875155746936798, "learning_rate": 9.356508653589819e-05, "loss": 0.0521, "step": 16410 }, { "grad_norm": 0.44781553745269775, "learning_rate": 9.355493985800237e-05, "loss": 0.0457, "step": 16420 }, { "grad_norm": 0.4224380850791931, "learning_rate": 9.354478573775857e-05, "loss": 0.0543, "step": 16430 }, { "grad_norm": 0.3051585555076599, "learning_rate": 9.353462417690186e-05, "loss": 0.0456, "step": 16440 }, { "grad_norm": 0.4762795567512512, "learning_rate": 9.352445517716853e-05, "loss": 0.0456, "step": 16450 }, { "grad_norm": 0.42510783672332764, "learning_rate": 9.351427874029621e-05, "loss": 0.0456, "step": 16460 }, { "grad_norm": 0.35180091857910156, "learning_rate": 9.350409486802379e-05, "loss": 0.0464, "step": 16470 }, { "grad_norm": 0.4340033233165741, "learning_rate": 9.349390356209138e-05, "loss": 0.0558, "step": 16480 }, { "grad_norm": 0.3922746777534485, "learning_rate": 9.348370482424042e-05, "loss": 0.051, "step": 16490 }, { "grad_norm": 0.42816075682640076, "learning_rate": 9.347349865621357e-05, "loss": 0.0431, "step": 16500 }, { "grad_norm": 0.5247583985328674, "learning_rate": 9.346328505975481e-05, "loss": 0.0482, "step": 16510 }, { "grad_norm": 0.4724240303039551, "learning_rate": 9.345306403660936e-05, "loss": 0.0507, "step": 16520 }, { "grad_norm": 0.5446886420249939, "learning_rate": 9.344283558852371e-05, "loss": 0.053, "step": 16530 }, { "grad_norm": 0.4147014617919922, "learning_rate": 9.343259971724563e-05, "loss": 0.0483, "step": 16540 }, { "grad_norm": 0.4019239842891693, "learning_rate": 9.342235642452413e-05, "loss": 0.0464, "step": 16550 }, { "grad_norm": 0.6095350384712219, "learning_rate": 9.341210571210954e-05, "loss": 0.0581, "step": 16560 }, { "grad_norm": 0.4188694953918457, "learning_rate": 9.340184758175338e-05, "loss": 0.0513, "step": 16570 }, { "grad_norm": 0.3675832748413086, "learning_rate": 9.339158203520854e-05, "loss": 0.0506, "step": 16580 }, { "grad_norm": 0.40626928210258484, "learning_rate": 9.338130907422908e-05, "loss": 0.0373, "step": 16590 }, { "grad_norm": 0.5153055787086487, "learning_rate": 9.337102870057037e-05, "loss": 0.0345, "step": 16600 }, { "grad_norm": 0.4657212793827057, "learning_rate": 9.336074091598907e-05, "loss": 0.0371, "step": 16610 }, { "grad_norm": 0.3050200045108795, "learning_rate": 9.335044572224306e-05, "loss": 0.0405, "step": 16620 }, { "grad_norm": 0.38840487599372864, "learning_rate": 9.334014312109151e-05, "loss": 0.0382, "step": 16630 }, { "grad_norm": 0.49427342414855957, "learning_rate": 9.332983311429486e-05, "loss": 0.0532, "step": 16640 }, { "grad_norm": 0.34831857681274414, "learning_rate": 9.33195157036148e-05, "loss": 0.0423, "step": 16650 }, { "grad_norm": 0.3544004559516907, "learning_rate": 9.330919089081432e-05, "loss": 0.0392, "step": 16660 }, { "grad_norm": 0.30270716547966003, "learning_rate": 9.32988586776576e-05, "loss": 0.0559, "step": 16670 }, { "grad_norm": 0.43688732385635376, "learning_rate": 9.328851906591016e-05, "loss": 0.0388, "step": 16680 }, { "grad_norm": 0.3951958119869232, "learning_rate": 9.327817205733875e-05, "loss": 0.045, "step": 16690 }, { "grad_norm": 0.4002712070941925, "learning_rate": 9.326781765371142e-05, "loss": 0.0435, "step": 16700 }, { "grad_norm": 0.33034974336624146, "learning_rate": 9.325745585679741e-05, "loss": 0.0439, "step": 16710 }, { "grad_norm": 0.5749499797821045, "learning_rate": 9.32470866683673e-05, "loss": 0.0591, "step": 16720 }, { "grad_norm": 0.5818949937820435, "learning_rate": 9.323671009019288e-05, "loss": 0.0497, "step": 16730 }, { "grad_norm": 0.29080531001091003, "learning_rate": 9.322632612404725e-05, "loss": 0.0437, "step": 16740 }, { "grad_norm": 0.36766520142555237, "learning_rate": 9.321593477170471e-05, "loss": 0.0549, "step": 16750 }, { "grad_norm": 0.3587471544742584, "learning_rate": 9.320553603494088e-05, "loss": 0.0402, "step": 16760 }, { "grad_norm": 0.3865559995174408, "learning_rate": 9.319512991553261e-05, "loss": 0.0391, "step": 16770 }, { "grad_norm": 0.424536794424057, "learning_rate": 9.318471641525803e-05, "loss": 0.0476, "step": 16780 }, { "grad_norm": 0.35113123059272766, "learning_rate": 9.317429553589652e-05, "loss": 0.0388, "step": 16790 }, { "grad_norm": 0.5464462637901306, "learning_rate": 9.316386727922873e-05, "loss": 0.0483, "step": 16800 }, { "grad_norm": 0.3583579361438751, "learning_rate": 9.315343164703656e-05, "loss": 0.0422, "step": 16810 }, { "grad_norm": 0.6014124155044556, "learning_rate": 9.314298864110316e-05, "loss": 0.0432, "step": 16820 }, { "grad_norm": 0.3332412838935852, "learning_rate": 9.313253826321295e-05, "loss": 0.0449, "step": 16830 }, { "grad_norm": 0.3309617340564728, "learning_rate": 9.312208051515165e-05, "loss": 0.0415, "step": 16840 }, { "grad_norm": 0.5172247886657715, "learning_rate": 9.311161539870618e-05, "loss": 0.0484, "step": 16850 }, { "grad_norm": 0.34086447954177856, "learning_rate": 9.310114291566474e-05, "loss": 0.0522, "step": 16860 }, { "grad_norm": 0.3515569567680359, "learning_rate": 9.309066306781679e-05, "loss": 0.0397, "step": 16870 }, { "grad_norm": 0.2873685657978058, "learning_rate": 9.308017585695306e-05, "loss": 0.0378, "step": 16880 }, { "grad_norm": 0.4072122573852539, "learning_rate": 9.306968128486552e-05, "loss": 0.0449, "step": 16890 }, { "grad_norm": 0.4156700670719147, "learning_rate": 9.30591793533474e-05, "loss": 0.0456, "step": 16900 }, { "grad_norm": 0.4378160834312439, "learning_rate": 9.304867006419321e-05, "loss": 0.042, "step": 16910 }, { "grad_norm": 0.5101292133331299, "learning_rate": 9.303815341919868e-05, "loss": 0.0469, "step": 16920 }, { "grad_norm": 0.35458725690841675, "learning_rate": 9.302762942016084e-05, "loss": 0.0447, "step": 16930 }, { "grad_norm": 0.3353431224822998, "learning_rate": 9.301709806887792e-05, "loss": 0.0428, "step": 16940 }, { "grad_norm": 0.4215874671936035, "learning_rate": 9.300655936714948e-05, "loss": 0.0587, "step": 16950 }, { "grad_norm": 0.40563279390335083, "learning_rate": 9.299601331677627e-05, "loss": 0.0489, "step": 16960 }, { "grad_norm": 0.46726930141448975, "learning_rate": 9.298545991956033e-05, "loss": 0.0552, "step": 16970 }, { "grad_norm": 0.36153584718704224, "learning_rate": 9.297489917730493e-05, "loss": 0.0429, "step": 16980 }, { "grad_norm": 0.4149998128414154, "learning_rate": 9.296433109181464e-05, "loss": 0.0544, "step": 16990 }, { "grad_norm": 0.5364313125610352, "learning_rate": 9.295375566489523e-05, "loss": 0.0568, "step": 17000 }, { "grad_norm": 0.32383057475090027, "learning_rate": 9.294317289835379e-05, "loss": 0.056, "step": 17010 }, { "grad_norm": 0.4465278685092926, "learning_rate": 9.293258279399859e-05, "loss": 0.0603, "step": 17020 }, { "grad_norm": 0.2677239775657654, "learning_rate": 9.292198535363919e-05, "loss": 0.042, "step": 17030 }, { "grad_norm": 0.36786550283432007, "learning_rate": 9.291138057908641e-05, "loss": 0.0433, "step": 17040 }, { "grad_norm": 0.36141231656074524, "learning_rate": 9.290076847215234e-05, "loss": 0.0395, "step": 17050 }, { "grad_norm": 0.5603335499763489, "learning_rate": 9.289014903465025e-05, "loss": 0.0583, "step": 17060 }, { "grad_norm": 0.3586393892765045, "learning_rate": 9.287952226839475e-05, "loss": 0.0385, "step": 17070 }, { "grad_norm": 0.4231727421283722, "learning_rate": 9.286888817520164e-05, "loss": 0.0427, "step": 17080 }, { "grad_norm": 0.2722272574901581, "learning_rate": 9.285824675688803e-05, "loss": 0.0426, "step": 17090 }, { "grad_norm": 0.4129244089126587, "learning_rate": 9.28475980152722e-05, "loss": 0.0424, "step": 17100 }, { "grad_norm": 0.5249737501144409, "learning_rate": 9.283694195217379e-05, "loss": 0.0438, "step": 17110 }, { "grad_norm": 0.438594788312912, "learning_rate": 9.282627856941356e-05, "loss": 0.0412, "step": 17120 }, { "grad_norm": 0.3802977204322815, "learning_rate": 9.281560786881363e-05, "loss": 0.042, "step": 17130 }, { "grad_norm": 0.386523962020874, "learning_rate": 9.280492985219733e-05, "loss": 0.0377, "step": 17140 }, { "grad_norm": 0.5018362998962402, "learning_rate": 9.279424452138924e-05, "loss": 0.0528, "step": 17150 }, { "grad_norm": 0.37506622076034546, "learning_rate": 9.278355187821517e-05, "loss": 0.0466, "step": 17160 }, { "grad_norm": 0.2826026678085327, "learning_rate": 9.277285192450224e-05, "loss": 0.0527, "step": 17170 }, { "grad_norm": 0.43043333292007446, "learning_rate": 9.276214466207875e-05, "loss": 0.0418, "step": 17180 }, { "grad_norm": 0.4530573785305023, "learning_rate": 9.275143009277427e-05, "loss": 0.0422, "step": 17190 }, { "grad_norm": 0.5631917119026184, "learning_rate": 9.274070821841964e-05, "loss": 0.0483, "step": 17200 }, { "grad_norm": 0.3463212847709656, "learning_rate": 9.272997904084696e-05, "loss": 0.0383, "step": 17210 }, { "grad_norm": 0.3466903567314148, "learning_rate": 9.271924256188951e-05, "loss": 0.0427, "step": 17220 }, { "grad_norm": 0.4858320951461792, "learning_rate": 9.270849878338189e-05, "loss": 0.0411, "step": 17230 }, { "grad_norm": 0.5150187015533447, "learning_rate": 9.269774770715991e-05, "loss": 0.0485, "step": 17240 }, { "grad_norm": 0.4255411922931671, "learning_rate": 9.268698933506061e-05, "loss": 0.054, "step": 17250 }, { "grad_norm": 0.5422551035881042, "learning_rate": 9.267622366892235e-05, "loss": 0.0417, "step": 17260 }, { "grad_norm": 0.370106965303421, "learning_rate": 9.266545071058465e-05, "loss": 0.0475, "step": 17270 }, { "grad_norm": 0.35295870900154114, "learning_rate": 9.265467046188833e-05, "loss": 0.0351, "step": 17280 }, { "grad_norm": 0.30540788173675537, "learning_rate": 9.264388292467543e-05, "loss": 0.0507, "step": 17290 }, { "grad_norm": 0.3414306342601776, "learning_rate": 9.263308810078926e-05, "loss": 0.0434, "step": 17300 }, { "grad_norm": 0.23686596751213074, "learning_rate": 9.262228599207434e-05, "loss": 0.0473, "step": 17310 }, { "grad_norm": 0.3807498514652252, "learning_rate": 9.261147660037647e-05, "loss": 0.0341, "step": 17320 }, { "grad_norm": 0.30639907717704773, "learning_rate": 9.26006599275427e-05, "loss": 0.0388, "step": 17330 }, { "grad_norm": 0.4319237768650055, "learning_rate": 9.258983597542124e-05, "loss": 0.0461, "step": 17340 }, { "grad_norm": 0.32253748178482056, "learning_rate": 9.257900474586167e-05, "loss": 0.0399, "step": 17350 }, { "grad_norm": 0.44698405265808105, "learning_rate": 9.256816624071471e-05, "loss": 0.0526, "step": 17360 }, { "grad_norm": 0.5550304055213928, "learning_rate": 9.25573204618324e-05, "loss": 0.0488, "step": 17370 }, { "grad_norm": 0.41105788946151733, "learning_rate": 9.254646741106796e-05, "loss": 0.0418, "step": 17380 }, { "grad_norm": 0.2816503047943115, "learning_rate": 9.253560709027589e-05, "loss": 0.0382, "step": 17390 }, { "grad_norm": 0.5917705297470093, "learning_rate": 9.252473950131192e-05, "loss": 0.0527, "step": 17400 }, { "grad_norm": 0.463128000497818, "learning_rate": 9.251386464603302e-05, "loss": 0.0368, "step": 17410 }, { "grad_norm": 0.3394263982772827, "learning_rate": 9.250298252629741e-05, "loss": 0.0475, "step": 17420 }, { "grad_norm": 0.4658971428871155, "learning_rate": 9.249209314396454e-05, "loss": 0.039, "step": 17430 }, { "grad_norm": 0.3817553222179413, "learning_rate": 9.248119650089513e-05, "loss": 0.0484, "step": 17440 }, { "grad_norm": 0.5541427135467529, "learning_rate": 9.247029259895108e-05, "loss": 0.0502, "step": 17450 }, { "grad_norm": 0.4152420163154602, "learning_rate": 9.24593814399956e-05, "loss": 0.0454, "step": 17460 }, { "grad_norm": 0.4089938700199127, "learning_rate": 9.244846302589309e-05, "loss": 0.0353, "step": 17470 }, { "grad_norm": 0.4692973792552948, "learning_rate": 9.243753735850923e-05, "loss": 0.0469, "step": 17480 }, { "grad_norm": 0.5425073504447937, "learning_rate": 9.24266044397109e-05, "loss": 0.037, "step": 17490 }, { "grad_norm": 0.3606676459312439, "learning_rate": 9.241566427136624e-05, "loss": 0.0333, "step": 17500 }, { "grad_norm": 0.46345287561416626, "learning_rate": 9.240471685534463e-05, "loss": 0.042, "step": 17510 }, { "grad_norm": 0.465634822845459, "learning_rate": 9.239376219351667e-05, "loss": 0.0431, "step": 17520 }, { "grad_norm": 0.418846994638443, "learning_rate": 9.238280028775425e-05, "loss": 0.0397, "step": 17530 }, { "grad_norm": 0.3841538429260254, "learning_rate": 9.237183113993041e-05, "loss": 0.0519, "step": 17540 }, { "grad_norm": 0.43342074751853943, "learning_rate": 9.236085475191952e-05, "loss": 0.0397, "step": 17550 }, { "grad_norm": 0.48073849081993103, "learning_rate": 9.234987112559709e-05, "loss": 0.0603, "step": 17560 }, { "grad_norm": 0.48265963792800903, "learning_rate": 9.233888026283999e-05, "loss": 0.0533, "step": 17570 }, { "grad_norm": 0.35319143533706665, "learning_rate": 9.232788216552619e-05, "loss": 0.0467, "step": 17580 }, { "grad_norm": 0.42752817273139954, "learning_rate": 9.231687683553502e-05, "loss": 0.0478, "step": 17590 }, { "grad_norm": 0.35506850481033325, "learning_rate": 9.230586427474698e-05, "loss": 0.059, "step": 17600 }, { "grad_norm": 0.3374832272529602, "learning_rate": 9.229484448504379e-05, "loss": 0.0469, "step": 17610 }, { "grad_norm": 0.5712307691574097, "learning_rate": 9.228381746830843e-05, "loss": 0.0483, "step": 17620 }, { "grad_norm": 0.3497712314128876, "learning_rate": 9.227278322642514e-05, "loss": 0.0359, "step": 17630 }, { "grad_norm": 0.4290991723537445, "learning_rate": 9.226174176127937e-05, "loss": 0.0453, "step": 17640 }, { "grad_norm": 0.30924659967422485, "learning_rate": 9.22506930747578e-05, "loss": 0.04, "step": 17650 }, { "grad_norm": 0.45242226123809814, "learning_rate": 9.223963716874831e-05, "loss": 0.0518, "step": 17660 }, { "grad_norm": 0.33190295100212097, "learning_rate": 9.222857404514012e-05, "loss": 0.0417, "step": 17670 }, { "grad_norm": 0.38160473108291626, "learning_rate": 9.221750370582355e-05, "loss": 0.0422, "step": 17680 }, { "grad_norm": 0.3844849467277527, "learning_rate": 9.220642615269028e-05, "loss": 0.0409, "step": 17690 }, { "grad_norm": 0.2982814610004425, "learning_rate": 9.219534138763311e-05, "loss": 0.0435, "step": 17700 }, { "grad_norm": 0.4142303168773651, "learning_rate": 9.218424941254613e-05, "loss": 0.0445, "step": 17710 }, { "grad_norm": 0.5142948627471924, "learning_rate": 9.217315022932468e-05, "loss": 0.0523, "step": 17720 }, { "grad_norm": 0.3257816433906555, "learning_rate": 9.216204383986528e-05, "loss": 0.0454, "step": 17730 }, { "grad_norm": 0.36639130115509033, "learning_rate": 9.215093024606574e-05, "loss": 0.0629, "step": 17740 }, { "grad_norm": 0.38774698972702026, "learning_rate": 9.213980944982506e-05, "loss": 0.0716, "step": 17750 }, { "grad_norm": 0.5866062045097351, "learning_rate": 9.212868145304346e-05, "loss": 0.0487, "step": 17760 }, { "grad_norm": 0.4801490604877472, "learning_rate": 9.211754625762241e-05, "loss": 0.0417, "step": 17770 }, { "grad_norm": 0.409999281167984, "learning_rate": 9.210640386546463e-05, "loss": 0.0591, "step": 17780 }, { "grad_norm": 0.37742429971694946, "learning_rate": 9.209525427847405e-05, "loss": 0.0384, "step": 17790 }, { "grad_norm": 0.7934357523918152, "learning_rate": 9.208409749855583e-05, "loss": 0.0478, "step": 17800 }, { "grad_norm": 0.503709077835083, "learning_rate": 9.207293352761633e-05, "loss": 0.0565, "step": 17810 }, { "grad_norm": 0.47765985131263733, "learning_rate": 9.206176236756319e-05, "loss": 0.0542, "step": 17820 }, { "grad_norm": 0.4221409857273102, "learning_rate": 9.205058402030525e-05, "loss": 0.0497, "step": 17830 }, { "grad_norm": 0.27990156412124634, "learning_rate": 9.203939848775259e-05, "loss": 0.0543, "step": 17840 }, { "grad_norm": 0.3694227337837219, "learning_rate": 9.202820577181652e-05, "loss": 0.0435, "step": 17850 }, { "grad_norm": 0.4312697947025299, "learning_rate": 9.201700587440953e-05, "loss": 0.0446, "step": 17860 }, { "grad_norm": 0.38566818833351135, "learning_rate": 9.200579879744544e-05, "loss": 0.0451, "step": 17870 }, { "grad_norm": 0.34547463059425354, "learning_rate": 9.199458454283918e-05, "loss": 0.0454, "step": 17880 }, { "grad_norm": 0.48654744029045105, "learning_rate": 9.198336311250697e-05, "loss": 0.0484, "step": 17890 }, { "grad_norm": 0.4620773494243622, "learning_rate": 9.197213450836626e-05, "loss": 0.0432, "step": 17900 }, { "grad_norm": 0.49586501717567444, "learning_rate": 9.19608987323357e-05, "loss": 0.0556, "step": 17910 }, { "grad_norm": 0.5069105625152588, "learning_rate": 9.194965578633517e-05, "loss": 0.0481, "step": 17920 }, { "grad_norm": 0.309783935546875, "learning_rate": 9.193840567228582e-05, "loss": 0.0472, "step": 17930 }, { "grad_norm": 0.5137965083122253, "learning_rate": 9.192714839210994e-05, "loss": 0.0498, "step": 17940 }, { "grad_norm": 0.6071773171424866, "learning_rate": 9.19158839477311e-05, "loss": 0.0585, "step": 17950 }, { "grad_norm": 0.4420047700405121, "learning_rate": 9.190461234107411e-05, "loss": 0.0598, "step": 17960 }, { "grad_norm": 0.2912691533565521, "learning_rate": 9.189333357406496e-05, "loss": 0.039, "step": 17970 }, { "grad_norm": 0.4677204191684723, "learning_rate": 9.188204764863089e-05, "loss": 0.0567, "step": 17980 }, { "grad_norm": 0.32337766885757446, "learning_rate": 9.187075456670033e-05, "loss": 0.0412, "step": 17990 }, { "grad_norm": 0.413387656211853, "learning_rate": 9.1859454330203e-05, "loss": 0.0451, "step": 18000 }, { "grad_norm": 0.4007972776889801, "learning_rate": 9.18481469410698e-05, "loss": 0.0601, "step": 18010 }, { "grad_norm": 0.38561534881591797, "learning_rate": 9.183683240123281e-05, "loss": 0.0418, "step": 18020 }, { "grad_norm": 0.3848193883895874, "learning_rate": 9.182551071262541e-05, "loss": 0.049, "step": 18030 }, { "grad_norm": 0.39562907814979553, "learning_rate": 9.181418187718218e-05, "loss": 0.0488, "step": 18040 }, { "grad_norm": 0.37272945046424866, "learning_rate": 9.180284589683888e-05, "loss": 0.0507, "step": 18050 }, { "grad_norm": 0.47173672914505005, "learning_rate": 9.17915027735325e-05, "loss": 0.0436, "step": 18060 }, { "grad_norm": 0.3203021287918091, "learning_rate": 9.178015250920133e-05, "loss": 0.0447, "step": 18070 }, { "grad_norm": 0.2826497256755829, "learning_rate": 9.176879510578477e-05, "loss": 0.0415, "step": 18080 }, { "grad_norm": 0.3159111738204956, "learning_rate": 9.17574305652235e-05, "loss": 0.0394, "step": 18090 }, { "grad_norm": 0.3283838927745819, "learning_rate": 9.174605888945942e-05, "loss": 0.0492, "step": 18100 }, { "grad_norm": 0.4019433856010437, "learning_rate": 9.173468008043564e-05, "loss": 0.0354, "step": 18110 }, { "grad_norm": 0.3285483717918396, "learning_rate": 9.172329414009648e-05, "loss": 0.043, "step": 18120 }, { "grad_norm": 0.38524743914604187, "learning_rate": 9.171190107038747e-05, "loss": 0.0413, "step": 18130 }, { "grad_norm": 0.37110522389411926, "learning_rate": 9.170050087325541e-05, "loss": 0.0356, "step": 18140 }, { "grad_norm": 0.35192376375198364, "learning_rate": 9.168909355064824e-05, "loss": 0.0381, "step": 18150 }, { "grad_norm": 0.26871782541275024, "learning_rate": 9.167767910451519e-05, "loss": 0.043, "step": 18160 }, { "grad_norm": 0.5211079120635986, "learning_rate": 9.166625753680669e-05, "loss": 0.0453, "step": 18170 }, { "grad_norm": 0.49040767550468445, "learning_rate": 9.165482884947431e-05, "loss": 0.042, "step": 18180 }, { "grad_norm": 0.3356230854988098, "learning_rate": 9.164339304447098e-05, "loss": 0.0455, "step": 18190 }, { "grad_norm": 0.3549482524394989, "learning_rate": 9.163195012375072e-05, "loss": 0.037, "step": 18200 }, { "grad_norm": 0.3502342700958252, "learning_rate": 9.16205000892688e-05, "loss": 0.0449, "step": 18210 }, { "grad_norm": 0.31965723633766174, "learning_rate": 9.160904294298175e-05, "loss": 0.0347, "step": 18220 }, { "grad_norm": 0.44199076294898987, "learning_rate": 9.159757868684727e-05, "loss": 0.046, "step": 18230 }, { "grad_norm": 0.6916593313217163, "learning_rate": 9.15861073228243e-05, "loss": 0.0504, "step": 18240 }, { "grad_norm": 0.4489550292491913, "learning_rate": 9.157462885287296e-05, "loss": 0.0522, "step": 18250 }, { "grad_norm": 0.4008234441280365, "learning_rate": 9.156314327895461e-05, "loss": 0.0383, "step": 18260 }, { "grad_norm": 0.4458562135696411, "learning_rate": 9.155165060303185e-05, "loss": 0.0465, "step": 18270 }, { "grad_norm": 0.43301865458488464, "learning_rate": 9.154015082706841e-05, "loss": 0.0381, "step": 18280 }, { "grad_norm": 0.321919709444046, "learning_rate": 9.152864395302936e-05, "loss": 0.0406, "step": 18290 }, { "grad_norm": 0.3730536997318268, "learning_rate": 9.151712998288085e-05, "loss": 0.039, "step": 18300 }, { "grad_norm": 0.3612805902957916, "learning_rate": 9.150560891859031e-05, "loss": 0.0454, "step": 18310 }, { "grad_norm": 0.4281851649284363, "learning_rate": 9.14940807621264e-05, "loss": 0.0398, "step": 18320 }, { "grad_norm": 0.2928585112094879, "learning_rate": 9.148254551545894e-05, "loss": 0.047, "step": 18330 }, { "grad_norm": 0.3819984495639801, "learning_rate": 9.147100318055901e-05, "loss": 0.0507, "step": 18340 }, { "grad_norm": 0.29400673508644104, "learning_rate": 9.145945375939888e-05, "loss": 0.0428, "step": 18350 }, { "grad_norm": 0.3896501660346985, "learning_rate": 9.144789725395203e-05, "loss": 0.0457, "step": 18360 }, { "grad_norm": 0.5316767692565918, "learning_rate": 9.14363336661931e-05, "loss": 0.0382, "step": 18370 }, { "grad_norm": 0.5342324376106262, "learning_rate": 9.142476299809806e-05, "loss": 0.0421, "step": 18380 }, { "grad_norm": 0.2932133078575134, "learning_rate": 9.1413185251644e-05, "loss": 0.042, "step": 18390 }, { "grad_norm": 0.39078113436698914, "learning_rate": 9.140160042880923e-05, "loss": 0.0474, "step": 18400 }, { "grad_norm": 0.3092840909957886, "learning_rate": 9.139000853157327e-05, "loss": 0.0477, "step": 18410 }, { "grad_norm": 0.33695095777511597, "learning_rate": 9.137840956191688e-05, "loss": 0.0482, "step": 18420 }, { "grad_norm": 0.3620397746562958, "learning_rate": 9.136680352182199e-05, "loss": 0.0392, "step": 18430 }, { "grad_norm": 0.3553537428379059, "learning_rate": 9.135519041327177e-05, "loss": 0.0364, "step": 18440 }, { "grad_norm": 0.3491050601005554, "learning_rate": 9.134357023825058e-05, "loss": 0.0472, "step": 18450 }, { "grad_norm": 0.365303099155426, "learning_rate": 9.133194299874398e-05, "loss": 0.0412, "step": 18460 }, { "grad_norm": 0.5238720774650574, "learning_rate": 9.132030869673876e-05, "loss": 0.051, "step": 18470 }, { "grad_norm": 0.38298216462135315, "learning_rate": 9.130866733422288e-05, "loss": 0.0439, "step": 18480 }, { "grad_norm": 0.3939999043941498, "learning_rate": 9.129701891318556e-05, "loss": 0.0438, "step": 18490 }, { "grad_norm": 0.32359373569488525, "learning_rate": 9.128536343561718e-05, "loss": 0.0512, "step": 18500 }, { "grad_norm": 0.48186424374580383, "learning_rate": 9.127370090350934e-05, "loss": 0.0456, "step": 18510 }, { "grad_norm": 0.34243595600128174, "learning_rate": 9.126203131885487e-05, "loss": 0.0431, "step": 18520 }, { "grad_norm": 0.37168312072753906, "learning_rate": 9.125035468364775e-05, "loss": 0.0451, "step": 18530 }, { "grad_norm": 0.30324941873550415, "learning_rate": 9.123867099988322e-05, "loss": 0.0355, "step": 18540 }, { "grad_norm": 0.30525100231170654, "learning_rate": 9.122698026955769e-05, "loss": 0.039, "step": 18550 }, { "grad_norm": 0.3254189193248749, "learning_rate": 9.12152824946688e-05, "loss": 0.0581, "step": 18560 }, { "grad_norm": 0.4388691782951355, "learning_rate": 9.120357767721538e-05, "loss": 0.0453, "step": 18570 }, { "grad_norm": 0.5080212950706482, "learning_rate": 9.119186581919745e-05, "loss": 0.0429, "step": 18580 }, { "grad_norm": 0.45977476239204407, "learning_rate": 9.118014692261624e-05, "loss": 0.0424, "step": 18590 }, { "grad_norm": 0.41906824707984924, "learning_rate": 9.116842098947422e-05, "loss": 0.0599, "step": 18600 }, { "grad_norm": 0.4744398891925812, "learning_rate": 9.115668802177499e-05, "loss": 0.039, "step": 18610 }, { "grad_norm": 0.31763389706611633, "learning_rate": 9.114494802152342e-05, "loss": 0.0577, "step": 18620 }, { "grad_norm": 0.5818264484405518, "learning_rate": 9.113320099072555e-05, "loss": 0.0697, "step": 18630 }, { "grad_norm": 0.45183995366096497, "learning_rate": 9.112144693138864e-05, "loss": 0.0479, "step": 18640 }, { "grad_norm": 0.33835354447364807, "learning_rate": 9.110968584552111e-05, "loss": 0.0498, "step": 18650 }, { "grad_norm": 0.5182276964187622, "learning_rate": 9.109791773513264e-05, "loss": 0.0538, "step": 18660 }, { "grad_norm": 0.5133038759231567, "learning_rate": 9.108614260223403e-05, "loss": 0.048, "step": 18670 }, { "grad_norm": 0.36949431896209717, "learning_rate": 9.107436044883738e-05, "loss": 0.0451, "step": 18680 }, { "grad_norm": 0.4013812839984894, "learning_rate": 9.10625712769559e-05, "loss": 0.04, "step": 18690 }, { "grad_norm": 0.30634841322898865, "learning_rate": 9.105077508860406e-05, "loss": 0.0362, "step": 18700 }, { "grad_norm": 0.31909167766571045, "learning_rate": 9.103897188579751e-05, "loss": 0.0334, "step": 18710 }, { "grad_norm": 0.23747442662715912, "learning_rate": 9.102716167055308e-05, "loss": 0.0363, "step": 18720 }, { "grad_norm": 0.35197913646698, "learning_rate": 9.10153444448888e-05, "loss": 0.0649, "step": 18730 }, { "grad_norm": 0.30421626567840576, "learning_rate": 9.100352021082393e-05, "loss": 0.033, "step": 18740 }, { "grad_norm": 0.4439198672771454, "learning_rate": 9.099168897037891e-05, "loss": 0.0352, "step": 18750 }, { "grad_norm": 0.43803292512893677, "learning_rate": 9.097985072557538e-05, "loss": 0.0455, "step": 18760 }, { "grad_norm": 0.382721483707428, "learning_rate": 9.096800547843615e-05, "loss": 0.0595, "step": 18770 }, { "grad_norm": 0.6737100481987, "learning_rate": 9.095615323098526e-05, "loss": 0.05, "step": 18780 }, { "grad_norm": 0.30420276522636414, "learning_rate": 9.094429398524795e-05, "loss": 0.0388, "step": 18790 }, { "grad_norm": 0.2679481506347656, "learning_rate": 9.093242774325061e-05, "loss": 0.036, "step": 18800 }, { "grad_norm": 0.3715372681617737, "learning_rate": 9.092055450702088e-05, "loss": 0.0381, "step": 18810 }, { "grad_norm": 0.291696697473526, "learning_rate": 9.090867427858756e-05, "loss": 0.0369, "step": 18820 }, { "grad_norm": 0.35231414437294006, "learning_rate": 9.089678705998066e-05, "loss": 0.0393, "step": 18830 }, { "grad_norm": 0.41830340027809143, "learning_rate": 9.088489285323139e-05, "loss": 0.0408, "step": 18840 }, { "grad_norm": 0.5147258043289185, "learning_rate": 9.087299166037212e-05, "loss": 0.0444, "step": 18850 }, { "grad_norm": 0.4573446810245514, "learning_rate": 9.086108348343647e-05, "loss": 0.0487, "step": 18860 }, { "grad_norm": 0.4023536443710327, "learning_rate": 9.08491683244592e-05, "loss": 0.0451, "step": 18870 }, { "grad_norm": 0.3110596239566803, "learning_rate": 9.08372461854763e-05, "loss": 0.0348, "step": 18880 }, { "grad_norm": 0.448354572057724, "learning_rate": 9.082531706852492e-05, "loss": 0.0364, "step": 18890 }, { "grad_norm": 0.4312792718410492, "learning_rate": 9.081338097564342e-05, "loss": 0.0595, "step": 18900 }, { "grad_norm": 0.3109110891819, "learning_rate": 9.080143790887137e-05, "loss": 0.0437, "step": 18910 }, { "grad_norm": 0.3404797911643982, "learning_rate": 9.07894878702495e-05, "loss": 0.0383, "step": 18920 }, { "grad_norm": 0.42709001898765564, "learning_rate": 9.077753086181974e-05, "loss": 0.0456, "step": 18930 }, { "grad_norm": 0.3660978674888611, "learning_rate": 9.076556688562524e-05, "loss": 0.0384, "step": 18940 }, { "grad_norm": 0.45759421586990356, "learning_rate": 9.075359594371029e-05, "loss": 0.0374, "step": 18950 }, { "grad_norm": 0.492327481508255, "learning_rate": 9.07416180381204e-05, "loss": 0.0429, "step": 18960 }, { "grad_norm": 0.33899950981140137, "learning_rate": 9.072963317090228e-05, "loss": 0.0387, "step": 18970 }, { "grad_norm": 0.3034154176712036, "learning_rate": 9.071764134410382e-05, "loss": 0.044, "step": 18980 }, { "grad_norm": 0.3417646884918213, "learning_rate": 9.070564255977407e-05, "loss": 0.0398, "step": 18990 }, { "grad_norm": 0.3029663562774658, "learning_rate": 9.06936368199633e-05, "loss": 0.0468, "step": 19000 }, { "grad_norm": 0.3115640878677368, "learning_rate": 9.0681624126723e-05, "loss": 0.0449, "step": 19010 }, { "grad_norm": 0.4512065052986145, "learning_rate": 9.066960448210576e-05, "loss": 0.0439, "step": 19020 }, { "grad_norm": 0.346085786819458, "learning_rate": 9.065757788816543e-05, "loss": 0.0358, "step": 19030 }, { "grad_norm": 0.32831594347953796, "learning_rate": 9.064554434695705e-05, "loss": 0.0447, "step": 19040 }, { "grad_norm": 0.42507025599479675, "learning_rate": 9.063350386053677e-05, "loss": 0.0434, "step": 19050 }, { "grad_norm": 0.43074464797973633, "learning_rate": 9.062145643096202e-05, "loss": 0.0481, "step": 19060 }, { "grad_norm": 0.35712409019470215, "learning_rate": 9.060940206029136e-05, "loss": 0.0472, "step": 19070 }, { "grad_norm": 0.4280473589897156, "learning_rate": 9.059734075058457e-05, "loss": 0.0402, "step": 19080 }, { "grad_norm": 0.49733009934425354, "learning_rate": 9.058527250390257e-05, "loss": 0.0469, "step": 19090 }, { "grad_norm": 0.3968447744846344, "learning_rate": 9.057319732230752e-05, "loss": 0.0484, "step": 19100 }, { "grad_norm": 0.2721068263053894, "learning_rate": 9.056111520786273e-05, "loss": 0.0383, "step": 19110 }, { "grad_norm": 0.4637555778026581, "learning_rate": 9.054902616263268e-05, "loss": 0.0416, "step": 19120 }, { "grad_norm": 0.2778015434741974, "learning_rate": 9.05369301886831e-05, "loss": 0.0535, "step": 19130 }, { "grad_norm": 0.3983023464679718, "learning_rate": 9.052482728808083e-05, "loss": 0.0391, "step": 19140 }, { "grad_norm": 0.35446321964263916, "learning_rate": 9.051271746289391e-05, "loss": 0.0288, "step": 19150 }, { "grad_norm": 0.43416059017181396, "learning_rate": 9.050060071519162e-05, "loss": 0.0445, "step": 19160 }, { "grad_norm": 0.2643897831439972, "learning_rate": 9.048847704704437e-05, "loss": 0.0342, "step": 19170 }, { "grad_norm": 0.3623649477958679, "learning_rate": 9.047634646052376e-05, "loss": 0.041, "step": 19180 }, { "grad_norm": 0.44789931178092957, "learning_rate": 9.046420895770256e-05, "loss": 0.0546, "step": 19190 }, { "grad_norm": 0.5132743716239929, "learning_rate": 9.045206454065473e-05, "loss": 0.0438, "step": 19200 }, { "grad_norm": 0.48844969272613525, "learning_rate": 9.043991321145546e-05, "loss": 0.0422, "step": 19210 }, { "grad_norm": 0.539111852645874, "learning_rate": 9.042775497218105e-05, "loss": 0.0453, "step": 19220 }, { "grad_norm": 0.6202012300491333, "learning_rate": 9.041558982490901e-05, "loss": 0.0513, "step": 19230 }, { "grad_norm": 0.3635985553264618, "learning_rate": 9.040341777171805e-05, "loss": 0.0502, "step": 19240 }, { "grad_norm": 0.3376377820968628, "learning_rate": 9.039123881468802e-05, "loss": 0.0401, "step": 19250 }, { "grad_norm": 0.349570631980896, "learning_rate": 9.037905295589998e-05, "loss": 0.0411, "step": 19260 }, { "grad_norm": 0.32835379242897034, "learning_rate": 9.036686019743617e-05, "loss": 0.036, "step": 19270 }, { "grad_norm": 0.4472915530204773, "learning_rate": 9.035466054137997e-05, "loss": 0.0461, "step": 19280 }, { "grad_norm": 0.5473878979682922, "learning_rate": 9.0342453989816e-05, "loss": 0.0544, "step": 19290 }, { "grad_norm": 0.35233888030052185, "learning_rate": 9.033024054483e-05, "loss": 0.0418, "step": 19300 }, { "grad_norm": 0.37867608666419983, "learning_rate": 9.031802020850894e-05, "loss": 0.0413, "step": 19310 }, { "grad_norm": 0.3319167196750641, "learning_rate": 9.030579298294092e-05, "loss": 0.0507, "step": 19320 }, { "grad_norm": 0.3680018186569214, "learning_rate": 9.029355887021524e-05, "loss": 0.0556, "step": 19330 }, { "grad_norm": 0.45647817850112915, "learning_rate": 9.028131787242238e-05, "loss": 0.067, "step": 19340 }, { "grad_norm": 0.42750340700149536, "learning_rate": 9.026906999165399e-05, "loss": 0.0523, "step": 19350 }, { "grad_norm": 0.2325245887041092, "learning_rate": 9.025681523000291e-05, "loss": 0.0364, "step": 19360 }, { "grad_norm": 0.3565565049648285, "learning_rate": 9.024455358956315e-05, "loss": 0.0395, "step": 19370 }, { "grad_norm": 0.42194169759750366, "learning_rate": 9.023228507242984e-05, "loss": 0.052, "step": 19380 }, { "grad_norm": 0.4616215229034424, "learning_rate": 9.022000968069937e-05, "loss": 0.0482, "step": 19390 }, { "grad_norm": 0.266264945268631, "learning_rate": 9.020772741646928e-05, "loss": 0.0428, "step": 19400 }, { "grad_norm": 0.31947365403175354, "learning_rate": 9.019543828183826e-05, "loss": 0.0384, "step": 19410 }, { "grad_norm": 0.3115067780017853, "learning_rate": 9.018314227890616e-05, "loss": 0.0344, "step": 19420 }, { "grad_norm": 0.4392133355140686, "learning_rate": 9.017083940977408e-05, "loss": 0.0606, "step": 19430 }, { "grad_norm": 0.2827661633491516, "learning_rate": 9.015852967654422e-05, "loss": 0.04, "step": 19440 }, { "grad_norm": 0.5408406853675842, "learning_rate": 9.014621308131996e-05, "loss": 0.0426, "step": 19450 }, { "grad_norm": 0.39852604269981384, "learning_rate": 9.01338896262059e-05, "loss": 0.0482, "step": 19460 }, { "grad_norm": 0.37820345163345337, "learning_rate": 9.012155931330777e-05, "loss": 0.0529, "step": 19470 }, { "grad_norm": 0.3933204114437103, "learning_rate": 9.010922214473246e-05, "loss": 0.0369, "step": 19480 }, { "grad_norm": 0.44558975100517273, "learning_rate": 9.009687812258808e-05, "loss": 0.0556, "step": 19490 }, { "grad_norm": 0.3257337510585785, "learning_rate": 9.00845272489839e-05, "loss": 0.0413, "step": 19500 }, { "grad_norm": 0.49663659930229187, "learning_rate": 9.007216952603031e-05, "loss": 0.0587, "step": 19510 }, { "grad_norm": 0.4508957266807556, "learning_rate": 9.005980495583894e-05, "loss": 0.0443, "step": 19520 }, { "grad_norm": 0.47022339701652527, "learning_rate": 9.004743354052252e-05, "loss": 0.042, "step": 19530 }, { "grad_norm": 0.4297194480895996, "learning_rate": 9.003505528219503e-05, "loss": 0.0425, "step": 19540 }, { "grad_norm": 0.31774312257766724, "learning_rate": 9.002267018297154e-05, "loss": 0.0481, "step": 19550 }, { "grad_norm": 0.3709782063961029, "learning_rate": 9.001027824496834e-05, "loss": 0.057, "step": 19560 }, { "grad_norm": 0.34220224618911743, "learning_rate": 8.999787947030287e-05, "loss": 0.0453, "step": 19570 }, { "grad_norm": 0.3640087842941284, "learning_rate": 8.998547386109376e-05, "loss": 0.0459, "step": 19580 }, { "grad_norm": 0.3256826400756836, "learning_rate": 8.997306141946073e-05, "loss": 0.0388, "step": 19590 }, { "grad_norm": 0.36512312293052673, "learning_rate": 8.996064214752481e-05, "loss": 0.0474, "step": 19600 }, { "grad_norm": 0.39477166533470154, "learning_rate": 8.994821604740806e-05, "loss": 0.0434, "step": 19610 }, { "grad_norm": 0.3769446313381195, "learning_rate": 8.993578312123377e-05, "loss": 0.0394, "step": 19620 }, { "grad_norm": 0.31065472960472107, "learning_rate": 8.992334337112639e-05, "loss": 0.0461, "step": 19630 }, { "grad_norm": 0.3636997938156128, "learning_rate": 8.991089679921154e-05, "loss": 0.0369, "step": 19640 }, { "grad_norm": 0.35839715600013733, "learning_rate": 8.989844340761599e-05, "loss": 0.0464, "step": 19650 }, { "grad_norm": 0.4701409339904785, "learning_rate": 8.988598319846768e-05, "loss": 0.0388, "step": 19660 }, { "grad_norm": 0.3046249449253082, "learning_rate": 8.987351617389574e-05, "loss": 0.0325, "step": 19670 }, { "grad_norm": 0.2794276177883148, "learning_rate": 8.98610423360304e-05, "loss": 0.0359, "step": 19680 }, { "grad_norm": 0.34957173466682434, "learning_rate": 8.984856168700317e-05, "loss": 0.042, "step": 19690 }, { "grad_norm": 0.31548017263412476, "learning_rate": 8.983607422894658e-05, "loss": 0.0377, "step": 19700 }, { "grad_norm": 0.31685176491737366, "learning_rate": 8.982357996399442e-05, "loss": 0.045, "step": 19710 }, { "grad_norm": 0.3655830919742584, "learning_rate": 8.981107889428164e-05, "loss": 0.0456, "step": 19720 }, { "grad_norm": 0.36335518956184387, "learning_rate": 8.979857102194428e-05, "loss": 0.0398, "step": 19730 }, { "grad_norm": 0.31021228432655334, "learning_rate": 8.978605634911968e-05, "loss": 0.0391, "step": 19740 }, { "grad_norm": 0.24813474714756012, "learning_rate": 8.977353487794616e-05, "loss": 0.0353, "step": 19750 }, { "grad_norm": 0.41594064235687256, "learning_rate": 8.976100661056334e-05, "loss": 0.0523, "step": 19760 }, { "grad_norm": 0.2559102773666382, "learning_rate": 8.974847154911197e-05, "loss": 0.036, "step": 19770 }, { "grad_norm": 0.3531963527202606, "learning_rate": 8.973592969573393e-05, "loss": 0.0466, "step": 19780 }, { "grad_norm": 0.4210762083530426, "learning_rate": 8.972338105257228e-05, "loss": 0.0485, "step": 19790 }, { "grad_norm": 0.3306858539581299, "learning_rate": 8.971082562177125e-05, "loss": 0.04, "step": 19800 }, { "grad_norm": 0.3127327263355255, "learning_rate": 8.96982634054762e-05, "loss": 0.0438, "step": 19810 }, { "grad_norm": 0.38720932602882385, "learning_rate": 8.96856944058337e-05, "loss": 0.0427, "step": 19820 }, { "grad_norm": 0.3254213333129883, "learning_rate": 8.967311862499144e-05, "loss": 0.0314, "step": 19830 }, { "grad_norm": 0.21706761419773102, "learning_rate": 8.966053606509825e-05, "loss": 0.0364, "step": 19840 }, { "grad_norm": 0.4226604998111725, "learning_rate": 8.964794672830417e-05, "loss": 0.0464, "step": 19850 }, { "grad_norm": 0.3817647695541382, "learning_rate": 8.963535061676038e-05, "loss": 0.0535, "step": 19860 }, { "grad_norm": 0.4624723196029663, "learning_rate": 8.962274773261918e-05, "loss": 0.0413, "step": 19870 }, { "grad_norm": 0.5627330541610718, "learning_rate": 8.961013807803409e-05, "loss": 0.0558, "step": 19880 }, { "grad_norm": 0.35219714045524597, "learning_rate": 8.959752165515973e-05, "loss": 0.0487, "step": 19890 }, { "grad_norm": 0.4106142818927765, "learning_rate": 8.958489846615193e-05, "loss": 0.0596, "step": 19900 }, { "grad_norm": 0.35747969150543213, "learning_rate": 8.957226851316762e-05, "loss": 0.0447, "step": 19910 }, { "grad_norm": 0.43494120240211487, "learning_rate": 8.955963179836493e-05, "loss": 0.042, "step": 19920 }, { "grad_norm": 0.7158629894256592, "learning_rate": 8.954698832390312e-05, "loss": 0.0398, "step": 19930 }, { "grad_norm": 0.6183689832687378, "learning_rate": 8.953433809194263e-05, "loss": 0.042, "step": 19940 }, { "grad_norm": 0.45778295397758484, "learning_rate": 8.9521681104645e-05, "loss": 0.0409, "step": 19950 }, { "grad_norm": 0.4849797487258911, "learning_rate": 8.9509017364173e-05, "loss": 0.0379, "step": 19960 }, { "grad_norm": 0.4377289116382599, "learning_rate": 8.949634687269052e-05, "loss": 0.0435, "step": 19970 }, { "grad_norm": 0.37428516149520874, "learning_rate": 8.948366963236259e-05, "loss": 0.0458, "step": 19980 }, { "grad_norm": 0.3744920492172241, "learning_rate": 8.947098564535538e-05, "loss": 0.0457, "step": 19990 }, { "grad_norm": 0.43413805961608887, "learning_rate": 8.945829491383627e-05, "loss": 0.0441, "step": 20000 }, { "grad_norm": 0.387794554233551, "learning_rate": 8.944559743997374e-05, "loss": 0.0376, "step": 20010 }, { "grad_norm": 0.3492976427078247, "learning_rate": 8.943289322593746e-05, "loss": 0.0433, "step": 20020 }, { "grad_norm": 0.37445440888404846, "learning_rate": 8.942018227389821e-05, "loss": 0.0379, "step": 20030 }, { "grad_norm": 0.4317210912704468, "learning_rate": 8.940746458602795e-05, "loss": 0.0622, "step": 20040 }, { "grad_norm": 0.42304643988609314, "learning_rate": 8.939474016449979e-05, "loss": 0.0551, "step": 20050 }, { "grad_norm": 0.41639772057533264, "learning_rate": 8.938200901148799e-05, "loss": 0.0367, "step": 20060 }, { "grad_norm": 0.5143146514892578, "learning_rate": 8.936927112916795e-05, "loss": 0.0511, "step": 20070 }, { "grad_norm": 0.37981247901916504, "learning_rate": 8.935652651971622e-05, "loss": 0.0371, "step": 20080 }, { "grad_norm": 0.5283191800117493, "learning_rate": 8.934377518531052e-05, "loss": 0.0519, "step": 20090 }, { "grad_norm": 0.36992496252059937, "learning_rate": 8.933101712812967e-05, "loss": 0.0387, "step": 20100 }, { "grad_norm": 0.3001583516597748, "learning_rate": 8.931825235035374e-05, "loss": 0.0435, "step": 20110 }, { "grad_norm": 0.45357242226600647, "learning_rate": 8.930548085416382e-05, "loss": 0.0437, "step": 20120 }, { "grad_norm": 0.2815125584602356, "learning_rate": 8.92927026417422e-05, "loss": 0.0452, "step": 20130 }, { "grad_norm": 0.36474665999412537, "learning_rate": 8.92799177152724e-05, "loss": 0.0362, "step": 20140 }, { "grad_norm": 0.3312701880931854, "learning_rate": 8.926712607693895e-05, "loss": 0.0387, "step": 20150 }, { "grad_norm": 0.3992764949798584, "learning_rate": 8.925432772892762e-05, "loss": 0.0412, "step": 20160 }, { "grad_norm": 0.528748095035553, "learning_rate": 8.924152267342529e-05, "loss": 0.0447, "step": 20170 }, { "grad_norm": 0.34439003467559814, "learning_rate": 8.922871091261998e-05, "loss": 0.0413, "step": 20180 }, { "grad_norm": 0.385511189699173, "learning_rate": 8.92158924487009e-05, "loss": 0.0351, "step": 20190 }, { "grad_norm": 0.5201751589775085, "learning_rate": 8.920306728385834e-05, "loss": 0.043, "step": 20200 }, { "grad_norm": 0.4094677269458771, "learning_rate": 8.919023542028379e-05, "loss": 0.0405, "step": 20210 }, { "grad_norm": 0.32750385999679565, "learning_rate": 8.917739686016988e-05, "loss": 0.0412, "step": 20220 }, { "grad_norm": 0.4405117332935333, "learning_rate": 8.916455160571033e-05, "loss": 0.0559, "step": 20230 }, { "grad_norm": 0.33073821663856506, "learning_rate": 8.915169965910008e-05, "loss": 0.0405, "step": 20240 }, { "grad_norm": 0.30817845463752747, "learning_rate": 8.913884102253514e-05, "loss": 0.0388, "step": 20250 }, { "grad_norm": 0.40459540486335754, "learning_rate": 8.912597569821273e-05, "loss": 0.0316, "step": 20260 }, { "grad_norm": 0.2952406704425812, "learning_rate": 8.911310368833118e-05, "loss": 0.0338, "step": 20270 }, { "grad_norm": 0.6386471390724182, "learning_rate": 8.910022499508994e-05, "loss": 0.0311, "step": 20280 }, { "grad_norm": 0.31243082880973816, "learning_rate": 8.908733962068965e-05, "loss": 0.0498, "step": 20290 }, { "grad_norm": 0.3139176070690155, "learning_rate": 8.907444756733207e-05, "loss": 0.0389, "step": 20300 }, { "grad_norm": 0.31814417243003845, "learning_rate": 8.906154883722006e-05, "loss": 0.0402, "step": 20310 }, { "grad_norm": 0.34970182180404663, "learning_rate": 8.904864343255773e-05, "loss": 0.0427, "step": 20320 }, { "grad_norm": 0.35952049493789673, "learning_rate": 8.90357313555502e-05, "loss": 0.0333, "step": 20330 }, { "grad_norm": 0.3439064025878906, "learning_rate": 8.90228126084038e-05, "loss": 0.0458, "step": 20340 }, { "grad_norm": 0.3804868459701538, "learning_rate": 8.900988719332601e-05, "loss": 0.0459, "step": 20350 }, { "grad_norm": 0.4386235773563385, "learning_rate": 8.899695511252542e-05, "loss": 0.041, "step": 20360 }, { "grad_norm": 0.35995376110076904, "learning_rate": 8.898401636821176e-05, "loss": 0.0532, "step": 20370 }, { "grad_norm": 0.6951944828033447, "learning_rate": 8.897107096259593e-05, "loss": 0.0583, "step": 20380 }, { "grad_norm": 0.4014672040939331, "learning_rate": 8.895811889788994e-05, "loss": 0.0445, "step": 20390 }, { "grad_norm": 0.3392268717288971, "learning_rate": 8.894516017630692e-05, "loss": 0.0388, "step": 20400 }, { "grad_norm": 0.47892194986343384, "learning_rate": 8.893219480006118e-05, "loss": 0.0526, "step": 20410 }, { "grad_norm": 0.4146702289581299, "learning_rate": 8.891922277136817e-05, "loss": 0.0401, "step": 20420 }, { "grad_norm": 0.27754124999046326, "learning_rate": 8.890624409244441e-05, "loss": 0.0394, "step": 20430 }, { "grad_norm": 0.42535725235939026, "learning_rate": 8.889325876550763e-05, "loss": 0.0465, "step": 20440 }, { "grad_norm": 0.32492178678512573, "learning_rate": 8.888026679277666e-05, "loss": 0.0459, "step": 20450 }, { "grad_norm": 0.31606119871139526, "learning_rate": 8.886726817647147e-05, "loss": 0.0711, "step": 20460 }, { "grad_norm": 0.2516234219074249, "learning_rate": 8.885426291881319e-05, "loss": 0.051, "step": 20470 }, { "grad_norm": 0.5146047472953796, "learning_rate": 8.884125102202401e-05, "loss": 0.0358, "step": 20480 }, { "grad_norm": 0.40816232562065125, "learning_rate": 8.882823248832736e-05, "loss": 0.036, "step": 20490 }, { "grad_norm": 0.4206562042236328, "learning_rate": 8.881520731994772e-05, "loss": 0.0447, "step": 20500 }, { "grad_norm": 0.2636395990848541, "learning_rate": 8.880217551911077e-05, "loss": 0.0448, "step": 20510 }, { "grad_norm": 0.4109051823616028, "learning_rate": 8.878913708804323e-05, "loss": 0.0388, "step": 20520 }, { "grad_norm": 0.30645468831062317, "learning_rate": 8.877609202897308e-05, "loss": 0.0415, "step": 20530 }, { "grad_norm": 0.40703198313713074, "learning_rate": 8.876304034412933e-05, "loss": 0.035, "step": 20540 }, { "grad_norm": 0.3150538504123688, "learning_rate": 8.874998203574214e-05, "loss": 0.0495, "step": 20550 }, { "grad_norm": 0.425525039434433, "learning_rate": 8.873691710604284e-05, "loss": 0.0413, "step": 20560 }, { "grad_norm": 0.44036346673965454, "learning_rate": 8.872384555726387e-05, "loss": 0.0355, "step": 20570 }, { "grad_norm": 0.48263925313949585, "learning_rate": 8.871076739163878e-05, "loss": 0.0396, "step": 20580 }, { "grad_norm": 0.3532925844192505, "learning_rate": 8.86976826114023e-05, "loss": 0.0399, "step": 20590 }, { "grad_norm": 0.34109631180763245, "learning_rate": 8.868459121879023e-05, "loss": 0.0506, "step": 20600 }, { "grad_norm": 0.4493064284324646, "learning_rate": 8.867149321603956e-05, "loss": 0.0468, "step": 20610 }, { "grad_norm": 0.2567494511604309, "learning_rate": 8.865838860538835e-05, "loss": 0.0409, "step": 20620 }, { "grad_norm": 0.399948388338089, "learning_rate": 8.864527738907585e-05, "loss": 0.0426, "step": 20630 }, { "grad_norm": 0.354536235332489, "learning_rate": 8.863215956934239e-05, "loss": 0.0386, "step": 20640 }, { "grad_norm": 0.3307906985282898, "learning_rate": 8.861903514842947e-05, "loss": 0.0318, "step": 20650 }, { "grad_norm": 0.323828786611557, "learning_rate": 8.860590412857966e-05, "loss": 0.0356, "step": 20660 }, { "grad_norm": 0.5197577476501465, "learning_rate": 8.85927665120367e-05, "loss": 0.0386, "step": 20670 }, { "grad_norm": 0.4473459720611572, "learning_rate": 8.857962230104546e-05, "loss": 0.041, "step": 20680 }, { "grad_norm": 0.4178787171840668, "learning_rate": 8.856647149785193e-05, "loss": 0.0375, "step": 20690 }, { "grad_norm": 0.236542746424675, "learning_rate": 8.855331410470322e-05, "loss": 0.0438, "step": 20700 }, { "grad_norm": 0.41362348198890686, "learning_rate": 8.854015012384756e-05, "loss": 0.0361, "step": 20710 }, { "grad_norm": 0.38061708211898804, "learning_rate": 8.852697955753433e-05, "loss": 0.0493, "step": 20720 }, { "grad_norm": 0.4350025951862335, "learning_rate": 8.851380240801399e-05, "loss": 0.0501, "step": 20730 }, { "grad_norm": 0.4095286726951599, "learning_rate": 8.850061867753818e-05, "loss": 0.036, "step": 20740 }, { "grad_norm": 0.41567787528038025, "learning_rate": 8.848742836835963e-05, "loss": 0.0485, "step": 20750 }, { "grad_norm": 0.37137553095817566, "learning_rate": 8.847423148273221e-05, "loss": 0.0443, "step": 20760 }, { "grad_norm": 0.5091126561164856, "learning_rate": 8.846102802291092e-05, "loss": 0.0507, "step": 20770 }, { "grad_norm": 0.34991008043289185, "learning_rate": 8.844781799115183e-05, "loss": 0.0435, "step": 20780 }, { "grad_norm": 0.35027632117271423, "learning_rate": 8.84346013897122e-05, "loss": 0.0387, "step": 20790 }, { "grad_norm": 0.331842303276062, "learning_rate": 8.842137822085038e-05, "loss": 0.0538, "step": 20800 }, { "grad_norm": 0.3434692323207855, "learning_rate": 8.840814848682585e-05, "loss": 0.0396, "step": 20810 }, { "grad_norm": 0.3335796892642975, "learning_rate": 8.83949121898992e-05, "loss": 0.0456, "step": 20820 }, { "grad_norm": 0.3673268258571625, "learning_rate": 8.838166933233217e-05, "loss": 0.0407, "step": 20830 }, { "grad_norm": 0.295770525932312, "learning_rate": 8.83684199163876e-05, "loss": 0.0446, "step": 20840 }, { "grad_norm": 0.3486168086528778, "learning_rate": 8.835516394432943e-05, "loss": 0.0417, "step": 20850 }, { "grad_norm": 0.3780852258205414, "learning_rate": 8.834190141842276e-05, "loss": 0.0415, "step": 20860 }, { "grad_norm": 0.4269368052482605, "learning_rate": 8.83286323409338e-05, "loss": 0.0408, "step": 20870 }, { "grad_norm": 0.37096723914146423, "learning_rate": 8.831535671412986e-05, "loss": 0.0416, "step": 20880 }, { "grad_norm": 0.36144042015075684, "learning_rate": 8.830207454027938e-05, "loss": 0.0431, "step": 20890 }, { "grad_norm": 0.308226078748703, "learning_rate": 8.828878582165192e-05, "loss": 0.0404, "step": 20900 }, { "grad_norm": 0.4884698987007141, "learning_rate": 8.827549056051818e-05, "loss": 0.0444, "step": 20910 }, { "grad_norm": 0.37099918723106384, "learning_rate": 8.826218875914993e-05, "loss": 0.0381, "step": 20920 }, { "grad_norm": 0.2802234888076782, "learning_rate": 8.82488804198201e-05, "loss": 0.0373, "step": 20930 }, { "grad_norm": 0.2959311008453369, "learning_rate": 8.82355655448027e-05, "loss": 0.042, "step": 20940 }, { "grad_norm": 0.41041237115859985, "learning_rate": 8.822224413637293e-05, "loss": 0.0421, "step": 20950 }, { "grad_norm": 0.34879204630851746, "learning_rate": 8.820891619680697e-05, "loss": 0.0413, "step": 20960 }, { "grad_norm": 0.306095153093338, "learning_rate": 8.819558172838227e-05, "loss": 0.0375, "step": 20970 }, { "grad_norm": 0.5195566415786743, "learning_rate": 8.818224073337731e-05, "loss": 0.0371, "step": 20980 }, { "grad_norm": 0.4506552219390869, "learning_rate": 8.816889321407169e-05, "loss": 0.0388, "step": 20990 }, { "grad_norm": 0.4051350951194763, "learning_rate": 8.815553917274615e-05, "loss": 0.0433, "step": 21000 }, { "grad_norm": 0.4733479619026184, "learning_rate": 8.81421786116825e-05, "loss": 0.0422, "step": 21010 }, { "grad_norm": 0.3486434817314148, "learning_rate": 8.812881153316373e-05, "loss": 0.0332, "step": 21020 }, { "grad_norm": 0.30931252241134644, "learning_rate": 8.81154379394739e-05, "loss": 0.0464, "step": 21030 }, { "grad_norm": 0.4020349979400635, "learning_rate": 8.810205783289818e-05, "loss": 0.0407, "step": 21040 }, { "grad_norm": 0.3397813141345978, "learning_rate": 8.808867121572286e-05, "loss": 0.0469, "step": 21050 }, { "grad_norm": 0.42453107237815857, "learning_rate": 8.807527809023537e-05, "loss": 0.051, "step": 21060 }, { "grad_norm": 0.42833420634269714, "learning_rate": 8.80618784587242e-05, "loss": 0.0371, "step": 21070 }, { "grad_norm": 0.36068302392959595, "learning_rate": 8.804847232347902e-05, "loss": 0.038, "step": 21080 }, { "grad_norm": 0.2949095070362091, "learning_rate": 8.803505968679054e-05, "loss": 0.0439, "step": 21090 }, { "grad_norm": 0.435809463262558, "learning_rate": 8.802164055095061e-05, "loss": 0.0483, "step": 21100 }, { "grad_norm": 0.34112271666526794, "learning_rate": 8.80082149182522e-05, "loss": 0.0399, "step": 21110 }, { "grad_norm": 0.4234452545642853, "learning_rate": 8.79947827909894e-05, "loss": 0.0438, "step": 21120 }, { "grad_norm": 0.4117744565010071, "learning_rate": 8.798134417145738e-05, "loss": 0.045, "step": 21130 }, { "grad_norm": 0.425105482339859, "learning_rate": 8.796789906195243e-05, "loss": 0.0468, "step": 21140 }, { "grad_norm": 0.5083947777748108, "learning_rate": 8.795444746477195e-05, "loss": 0.0483, "step": 21150 }, { "grad_norm": 0.2896019220352173, "learning_rate": 8.794098938221446e-05, "loss": 0.0396, "step": 21160 }, { "grad_norm": 0.3009101450443268, "learning_rate": 8.792752481657957e-05, "loss": 0.0456, "step": 21170 }, { "grad_norm": 0.4236001968383789, "learning_rate": 8.791405377016802e-05, "loss": 0.0388, "step": 21180 }, { "grad_norm": 0.40842607617378235, "learning_rate": 8.790057624528163e-05, "loss": 0.0484, "step": 21190 }, { "grad_norm": 0.9146794080734253, "learning_rate": 8.788709224422333e-05, "loss": 0.0431, "step": 21200 }, { "grad_norm": 0.41976797580718994, "learning_rate": 8.787360176929717e-05, "loss": 0.0425, "step": 21210 }, { "grad_norm": 0.3830180764198303, "learning_rate": 8.786010482280834e-05, "loss": 0.0461, "step": 21220 }, { "grad_norm": 0.3310227692127228, "learning_rate": 8.784660140706306e-05, "loss": 0.0458, "step": 21230 }, { "grad_norm": 0.36421477794647217, "learning_rate": 8.783309152436872e-05, "loss": 0.0405, "step": 21240 }, { "grad_norm": 0.3417850434780121, "learning_rate": 8.781957517703375e-05, "loss": 0.0372, "step": 21250 }, { "grad_norm": 0.41681015491485596, "learning_rate": 8.780605236736776e-05, "loss": 0.0524, "step": 21260 }, { "grad_norm": 0.2228592187166214, "learning_rate": 8.779252309768142e-05, "loss": 0.0425, "step": 21270 }, { "grad_norm": 0.46776485443115234, "learning_rate": 8.777898737028652e-05, "loss": 0.0468, "step": 21280 }, { "grad_norm": 0.38369569182395935, "learning_rate": 8.776544518749591e-05, "loss": 0.0411, "step": 21290 }, { "grad_norm": 0.31993335485458374, "learning_rate": 8.775189655162364e-05, "loss": 0.0395, "step": 21300 }, { "grad_norm": 0.36869776248931885, "learning_rate": 8.773834146498474e-05, "loss": 0.0365, "step": 21310 }, { "grad_norm": 0.4338088929653168, "learning_rate": 8.772477992989545e-05, "loss": 0.038, "step": 21320 }, { "grad_norm": 0.6916678547859192, "learning_rate": 8.771121194867304e-05, "loss": 0.051, "step": 21330 }, { "grad_norm": 0.37099021673202515, "learning_rate": 8.769763752363589e-05, "loss": 0.0404, "step": 21340 }, { "grad_norm": 0.3031005263328552, "learning_rate": 8.768405665710352e-05, "loss": 0.0461, "step": 21350 }, { "grad_norm": 0.480080246925354, "learning_rate": 8.767046935139655e-05, "loss": 0.051, "step": 21360 }, { "grad_norm": 0.3326179087162018, "learning_rate": 8.765687560883666e-05, "loss": 0.0397, "step": 21370 }, { "grad_norm": 0.4037152826786041, "learning_rate": 8.764327543174664e-05, "loss": 0.0369, "step": 21380 }, { "grad_norm": 0.3543984889984131, "learning_rate": 8.762966882245038e-05, "loss": 0.0374, "step": 21390 }, { "grad_norm": 0.39467549324035645, "learning_rate": 8.761605578327291e-05, "loss": 0.0457, "step": 21400 }, { "grad_norm": 0.5660311579704285, "learning_rate": 8.76024363165403e-05, "loss": 0.0563, "step": 21410 }, { "grad_norm": 0.3657277524471283, "learning_rate": 8.758881042457976e-05, "loss": 0.0484, "step": 21420 }, { "grad_norm": 0.33198949694633484, "learning_rate": 8.757517810971957e-05, "loss": 0.0438, "step": 21430 }, { "grad_norm": 0.29138830304145813, "learning_rate": 8.756153937428913e-05, "loss": 0.04, "step": 21440 }, { "grad_norm": 0.5014550089836121, "learning_rate": 8.754789422061889e-05, "loss": 0.056, "step": 21450 }, { "grad_norm": 0.3196139335632324, "learning_rate": 8.753424265104052e-05, "loss": 0.0422, "step": 21460 }, { "grad_norm": 0.4802767336368561, "learning_rate": 8.752058466788659e-05, "loss": 0.0544, "step": 21470 }, { "grad_norm": 0.4103253185749054, "learning_rate": 8.750692027349097e-05, "loss": 0.0398, "step": 21480 }, { "grad_norm": 0.5150931477546692, "learning_rate": 8.749324947018847e-05, "loss": 0.0381, "step": 21490 }, { "grad_norm": 0.4969213604927063, "learning_rate": 8.747957226031507e-05, "loss": 0.047, "step": 21500 }, { "grad_norm": 0.37318184971809387, "learning_rate": 8.746588864620787e-05, "loss": 0.0437, "step": 21510 }, { "grad_norm": 0.32031211256980896, "learning_rate": 8.745219863020498e-05, "loss": 0.0347, "step": 21520 }, { "grad_norm": 0.36749058961868286, "learning_rate": 8.743850221464564e-05, "loss": 0.0378, "step": 21530 }, { "grad_norm": 0.3676793575286865, "learning_rate": 8.742479940187026e-05, "loss": 0.0438, "step": 21540 }, { "grad_norm": 0.33825019001960754, "learning_rate": 8.74110901942202e-05, "loss": 0.0426, "step": 21550 }, { "grad_norm": 0.27610138058662415, "learning_rate": 8.739737459403803e-05, "loss": 0.0428, "step": 21560 }, { "grad_norm": 0.42978769540786743, "learning_rate": 8.738365260366737e-05, "loss": 0.04, "step": 21570 }, { "grad_norm": 0.38095465302467346, "learning_rate": 8.736992422545292e-05, "loss": 0.0388, "step": 21580 }, { "grad_norm": 0.4261016547679901, "learning_rate": 8.73561894617405e-05, "loss": 0.0525, "step": 21590 }, { "grad_norm": 0.2840716242790222, "learning_rate": 8.734244831487697e-05, "loss": 0.0351, "step": 21600 }, { "grad_norm": 0.42852187156677246, "learning_rate": 8.732870078721035e-05, "loss": 0.0462, "step": 21610 }, { "grad_norm": 0.48396414518356323, "learning_rate": 8.731494688108972e-05, "loss": 0.0391, "step": 21620 }, { "grad_norm": 0.443623423576355, "learning_rate": 8.730118659886523e-05, "loss": 0.0493, "step": 21630 }, { "grad_norm": 0.3801542818546295, "learning_rate": 8.728741994288814e-05, "loss": 0.0369, "step": 21640 }, { "grad_norm": 0.5763310194015503, "learning_rate": 8.727364691551079e-05, "loss": 0.0344, "step": 21650 }, { "grad_norm": 0.27141907811164856, "learning_rate": 8.725986751908661e-05, "loss": 0.0404, "step": 21660 }, { "grad_norm": 0.38547980785369873, "learning_rate": 8.724608175597016e-05, "loss": 0.0375, "step": 21670 }, { "grad_norm": 0.381242573261261, "learning_rate": 8.723228962851699e-05, "loss": 0.0402, "step": 21680 }, { "grad_norm": 0.3088046610355377, "learning_rate": 8.721849113908385e-05, "loss": 0.0446, "step": 21690 }, { "grad_norm": 0.3399161696434021, "learning_rate": 8.720468629002848e-05, "loss": 0.0391, "step": 21700 }, { "grad_norm": 0.4690243601799011, "learning_rate": 8.719087508370978e-05, "loss": 0.0341, "step": 21710 }, { "grad_norm": 0.4215295612812042, "learning_rate": 8.717705752248772e-05, "loss": 0.0487, "step": 21720 }, { "grad_norm": 0.5325270891189575, "learning_rate": 8.71632336087233e-05, "loss": 0.0337, "step": 21730 }, { "grad_norm": 0.3691488802433014, "learning_rate": 8.71494033447787e-05, "loss": 0.0422, "step": 21740 }, { "grad_norm": 0.36854299902915955, "learning_rate": 8.713556673301708e-05, "loss": 0.0353, "step": 21750 }, { "grad_norm": 0.42735815048217773, "learning_rate": 8.712172377580278e-05, "loss": 0.0545, "step": 21760 }, { "grad_norm": 0.4140470325946808, "learning_rate": 8.710787447550114e-05, "loss": 0.0481, "step": 21770 }, { "grad_norm": 0.46764686703681946, "learning_rate": 8.70940188344787e-05, "loss": 0.056, "step": 21780 }, { "grad_norm": 0.5003660321235657, "learning_rate": 8.708015685510293e-05, "loss": 0.0402, "step": 21790 }, { "grad_norm": 0.4477996826171875, "learning_rate": 8.706628853974252e-05, "loss": 0.0428, "step": 21800 }, { "grad_norm": 0.2909353971481323, "learning_rate": 8.705241389076715e-05, "loss": 0.0259, "step": 21810 }, { "grad_norm": 0.3756578266620636, "learning_rate": 8.703853291054764e-05, "loss": 0.0354, "step": 21820 }, { "grad_norm": 0.5933502912521362, "learning_rate": 8.702464560145587e-05, "loss": 0.0473, "step": 21830 }, { "grad_norm": 0.3806300759315491, "learning_rate": 8.701075196586476e-05, "loss": 0.0364, "step": 21840 }, { "grad_norm": 0.37282443046569824, "learning_rate": 8.699685200614842e-05, "loss": 0.0345, "step": 21850 }, { "grad_norm": 0.4787067174911499, "learning_rate": 8.698294572468193e-05, "loss": 0.0488, "step": 21860 }, { "grad_norm": 0.31947124004364014, "learning_rate": 8.696903312384148e-05, "loss": 0.0428, "step": 21870 }, { "grad_norm": 0.3805564045906067, "learning_rate": 8.695511420600439e-05, "loss": 0.0334, "step": 21880 }, { "grad_norm": 0.2963232100009918, "learning_rate": 8.694118897354901e-05, "loss": 0.0387, "step": 21890 }, { "grad_norm": 0.414899080991745, "learning_rate": 8.692725742885478e-05, "loss": 0.0373, "step": 21900 }, { "grad_norm": 0.4719719886779785, "learning_rate": 8.691331957430221e-05, "loss": 0.0479, "step": 21910 }, { "grad_norm": 0.4174164831638336, "learning_rate": 8.68993754122729e-05, "loss": 0.0411, "step": 21920 }, { "grad_norm": 0.4918383061885834, "learning_rate": 8.688542494514955e-05, "loss": 0.0347, "step": 21930 }, { "grad_norm": 0.45207908749580383, "learning_rate": 8.68714681753159e-05, "loss": 0.0417, "step": 21940 }, { "grad_norm": 0.32498565316200256, "learning_rate": 8.685750510515676e-05, "loss": 0.0304, "step": 21950 }, { "grad_norm": 0.44260773062705994, "learning_rate": 8.684353573705805e-05, "loss": 0.0375, "step": 21960 }, { "grad_norm": 0.3957160413265228, "learning_rate": 8.682956007340677e-05, "loss": 0.0356, "step": 21970 }, { "grad_norm": 0.42306050658226013, "learning_rate": 8.681557811659095e-05, "loss": 0.0353, "step": 21980 }, { "grad_norm": 0.41711342334747314, "learning_rate": 8.680158986899974e-05, "loss": 0.0349, "step": 21990 }, { "grad_norm": 0.4232124984264374, "learning_rate": 8.678759533302335e-05, "loss": 0.0406, "step": 22000 }, { "grad_norm": 0.39855650067329407, "learning_rate": 8.677359451105308e-05, "loss": 0.0307, "step": 22010 }, { "grad_norm": 0.313954621553421, "learning_rate": 8.675958740548123e-05, "loss": 0.0497, "step": 22020 }, { "grad_norm": 0.31016290187835693, "learning_rate": 8.674557401870129e-05, "loss": 0.0413, "step": 22030 }, { "grad_norm": 0.2937169373035431, "learning_rate": 8.673155435310775e-05, "loss": 0.0418, "step": 22040 }, { "grad_norm": 0.3003093898296356, "learning_rate": 8.671752841109617e-05, "loss": 0.0364, "step": 22050 }, { "grad_norm": 0.39263081550598145, "learning_rate": 8.670349619506321e-05, "loss": 0.0357, "step": 22060 }, { "grad_norm": 0.48682937026023865, "learning_rate": 8.66894577074066e-05, "loss": 0.0419, "step": 22070 }, { "grad_norm": 0.4499397873878479, "learning_rate": 8.667541295052513e-05, "loss": 0.0372, "step": 22080 }, { "grad_norm": 0.3887929320335388, "learning_rate": 8.666136192681865e-05, "loss": 0.0485, "step": 22090 }, { "grad_norm": 0.3225102722644806, "learning_rate": 8.664730463868811e-05, "loss": 0.0378, "step": 22100 }, { "grad_norm": 0.37264689803123474, "learning_rate": 8.663324108853552e-05, "loss": 0.0428, "step": 22110 }, { "grad_norm": 0.3142189681529999, "learning_rate": 8.661917127876395e-05, "loss": 0.0505, "step": 22120 }, { "grad_norm": 0.37197184562683105, "learning_rate": 8.660509521177754e-05, "loss": 0.0462, "step": 22130 }, { "grad_norm": 0.33539700508117676, "learning_rate": 8.65910128899815e-05, "loss": 0.0364, "step": 22140 }, { "grad_norm": 0.3083111047744751, "learning_rate": 8.657692431578214e-05, "loss": 0.039, "step": 22150 }, { "grad_norm": 0.38744908571243286, "learning_rate": 8.656282949158679e-05, "loss": 0.0354, "step": 22160 }, { "grad_norm": 0.28293362259864807, "learning_rate": 8.654872841980388e-05, "loss": 0.0329, "step": 22170 }, { "grad_norm": 0.45414435863494873, "learning_rate": 8.653462110284289e-05, "loss": 0.0333, "step": 22180 }, { "grad_norm": 0.33212709426879883, "learning_rate": 8.652050754311437e-05, "loss": 0.0393, "step": 22190 }, { "grad_norm": 0.424123078584671, "learning_rate": 8.650638774302995e-05, "loss": 0.0399, "step": 22200 }, { "grad_norm": 0.3025352656841278, "learning_rate": 8.649226170500233e-05, "loss": 0.0422, "step": 22210 }, { "grad_norm": 0.3062354624271393, "learning_rate": 8.647812943144524e-05, "loss": 0.0406, "step": 22220 }, { "grad_norm": 0.2891155779361725, "learning_rate": 8.646399092477351e-05, "loss": 0.0306, "step": 22230 }, { "grad_norm": 0.3669663667678833, "learning_rate": 8.644984618740301e-05, "loss": 0.05, "step": 22240 }, { "grad_norm": 0.3991347849369049, "learning_rate": 8.643569522175073e-05, "loss": 0.0338, "step": 22250 }, { "grad_norm": 0.3346879184246063, "learning_rate": 8.642153803023463e-05, "loss": 0.0442, "step": 22260 }, { "grad_norm": 0.3311840891838074, "learning_rate": 8.640737461527383e-05, "loss": 0.0359, "step": 22270 }, { "grad_norm": 0.3827577233314514, "learning_rate": 8.639320497928845e-05, "loss": 0.0376, "step": 22280 }, { "grad_norm": 0.34823042154312134, "learning_rate": 8.637902912469969e-05, "loss": 0.0358, "step": 22290 }, { "grad_norm": 0.3133566975593567, "learning_rate": 8.636484705392982e-05, "loss": 0.0373, "step": 22300 }, { "grad_norm": 0.3115680515766144, "learning_rate": 8.635065876940216e-05, "loss": 0.0356, "step": 22310 }, { "grad_norm": 0.39751386642456055, "learning_rate": 8.633646427354112e-05, "loss": 0.0417, "step": 22320 }, { "grad_norm": 0.45514845848083496, "learning_rate": 8.632226356877213e-05, "loss": 0.0328, "step": 22330 }, { "grad_norm": 0.33311542868614197, "learning_rate": 8.630805665752173e-05, "loss": 0.0466, "step": 22340 }, { "grad_norm": 0.3525981903076172, "learning_rate": 8.629384354221748e-05, "loss": 0.0403, "step": 22350 }, { "grad_norm": 0.3535078167915344, "learning_rate": 8.627962422528797e-05, "loss": 0.0369, "step": 22360 }, { "grad_norm": 0.3151554763317108, "learning_rate": 8.626539870916296e-05, "loss": 0.0539, "step": 22370 }, { "grad_norm": 0.3335074186325073, "learning_rate": 8.625116699627317e-05, "loss": 0.0416, "step": 22380 }, { "grad_norm": 0.528399646282196, "learning_rate": 8.623692908905041e-05, "loss": 0.0363, "step": 22390 }, { "grad_norm": 0.39988407492637634, "learning_rate": 8.622268498992755e-05, "loss": 0.0444, "step": 22400 }, { "grad_norm": 0.3582186698913574, "learning_rate": 8.620843470133851e-05, "loss": 0.053, "step": 22410 }, { "grad_norm": 0.35210421681404114, "learning_rate": 8.619417822571829e-05, "loss": 0.0392, "step": 22420 }, { "grad_norm": 0.5580580830574036, "learning_rate": 8.617991556550292e-05, "loss": 0.0368, "step": 22430 }, { "grad_norm": 0.42661648988723755, "learning_rate": 8.616564672312952e-05, "loss": 0.036, "step": 22440 }, { "grad_norm": 0.3531917631626129, "learning_rate": 8.61513717010362e-05, "loss": 0.0412, "step": 22450 }, { "grad_norm": 0.42926958203315735, "learning_rate": 8.613709050166221e-05, "loss": 0.0423, "step": 22460 }, { "grad_norm": 0.30733969807624817, "learning_rate": 8.61228031274478e-05, "loss": 0.0369, "step": 22470 }, { "grad_norm": 0.3439446985721588, "learning_rate": 8.610850958083431e-05, "loss": 0.0365, "step": 22480 }, { "grad_norm": 0.42236077785491943, "learning_rate": 8.609420986426409e-05, "loss": 0.0364, "step": 22490 }, { "grad_norm": 0.42711761593818665, "learning_rate": 8.60799039801806e-05, "loss": 0.035, "step": 22500 }, { "grad_norm": 0.38648414611816406, "learning_rate": 8.606559193102828e-05, "loss": 0.032, "step": 22510 }, { "grad_norm": 0.38117650151252747, "learning_rate": 8.605127371925273e-05, "loss": 0.0417, "step": 22520 }, { "grad_norm": 0.37771323323249817, "learning_rate": 8.603694934730047e-05, "loss": 0.0443, "step": 22530 }, { "grad_norm": 0.618434488773346, "learning_rate": 8.602261881761919e-05, "loss": 0.0324, "step": 22540 }, { "grad_norm": 0.36504125595092773, "learning_rate": 8.600828213265759e-05, "loss": 0.0445, "step": 22550 }, { "grad_norm": 0.41492095589637756, "learning_rate": 8.599393929486539e-05, "loss": 0.0522, "step": 22560 }, { "grad_norm": 0.4469957649707794, "learning_rate": 8.59795903066934e-05, "loss": 0.0375, "step": 22570 }, { "grad_norm": 0.38498803973197937, "learning_rate": 8.596523517059347e-05, "loss": 0.0293, "step": 22580 }, { "grad_norm": 0.3485424518585205, "learning_rate": 8.59508738890185e-05, "loss": 0.0335, "step": 22590 }, { "grad_norm": 0.38221463561058044, "learning_rate": 8.593650646442246e-05, "loss": 0.0349, "step": 22600 }, { "grad_norm": 0.3656059503555298, "learning_rate": 8.59221328992603e-05, "loss": 0.0315, "step": 22610 }, { "grad_norm": 0.41283363103866577, "learning_rate": 8.590775319598813e-05, "loss": 0.0383, "step": 22620 }, { "grad_norm": 0.4620700776576996, "learning_rate": 8.589336735706301e-05, "loss": 0.0384, "step": 22630 }, { "grad_norm": 0.3957197666168213, "learning_rate": 8.587897538494307e-05, "loss": 0.0457, "step": 22640 }, { "grad_norm": 0.36472973227500916, "learning_rate": 8.586457728208756e-05, "loss": 0.0402, "step": 22650 }, { "grad_norm": 0.37954777479171753, "learning_rate": 8.585017305095667e-05, "loss": 0.0412, "step": 22660 }, { "grad_norm": 0.39871373772621155, "learning_rate": 8.583576269401173e-05, "loss": 0.039, "step": 22670 }, { "grad_norm": 0.38557013869285583, "learning_rate": 8.582134621371504e-05, "loss": 0.0387, "step": 22680 }, { "grad_norm": 0.35018035769462585, "learning_rate": 8.580692361253e-05, "loss": 0.0317, "step": 22690 }, { "grad_norm": 0.3837898373603821, "learning_rate": 8.579249489292104e-05, "loss": 0.0395, "step": 22700 }, { "grad_norm": 0.36323636770248413, "learning_rate": 8.577806005735363e-05, "loss": 0.041, "step": 22710 }, { "grad_norm": 0.5464754104614258, "learning_rate": 8.576361910829429e-05, "loss": 0.0315, "step": 22720 }, { "grad_norm": 0.3956581652164459, "learning_rate": 8.574917204821057e-05, "loss": 0.0372, "step": 22730 }, { "grad_norm": 0.3279998302459717, "learning_rate": 8.57347188795711e-05, "loss": 0.0321, "step": 22740 }, { "grad_norm": 0.561747133731842, "learning_rate": 8.572025960484551e-05, "loss": 0.0457, "step": 22750 }, { "grad_norm": 0.6542598009109497, "learning_rate": 8.57057942265045e-05, "loss": 0.0562, "step": 22760 }, { "grad_norm": 0.350781112909317, "learning_rate": 8.569132274701984e-05, "loss": 0.0396, "step": 22770 }, { "grad_norm": 0.44922250509262085, "learning_rate": 8.567684516886427e-05, "loss": 0.0398, "step": 22780 }, { "grad_norm": 0.3992965817451477, "learning_rate": 8.56623614945116e-05, "loss": 0.0373, "step": 22790 }, { "grad_norm": 0.339305579662323, "learning_rate": 8.564787172643675e-05, "loss": 0.0351, "step": 22800 }, { "grad_norm": 0.30042266845703125, "learning_rate": 8.563337586711559e-05, "loss": 0.0414, "step": 22810 }, { "grad_norm": 0.3215928375720978, "learning_rate": 8.561887391902506e-05, "loss": 0.0455, "step": 22820 }, { "grad_norm": 0.4454922676086426, "learning_rate": 8.560436588464316e-05, "loss": 0.0296, "step": 22830 }, { "grad_norm": 0.3871561884880066, "learning_rate": 8.55898517664489e-05, "loss": 0.0403, "step": 22840 }, { "grad_norm": 0.42701658606529236, "learning_rate": 8.557533156692236e-05, "loss": 0.0495, "step": 22850 }, { "grad_norm": 0.3720012605190277, "learning_rate": 8.556080528854467e-05, "loss": 0.0304, "step": 22860 }, { "grad_norm": 0.30344319343566895, "learning_rate": 8.554627293379791e-05, "loss": 0.04, "step": 22870 }, { "grad_norm": 0.3498762547969818, "learning_rate": 8.553173450516531e-05, "loss": 0.0487, "step": 22880 }, { "grad_norm": 0.29829058051109314, "learning_rate": 8.551719000513108e-05, "loss": 0.0361, "step": 22890 }, { "grad_norm": 0.40355029702186584, "learning_rate": 8.550263943618049e-05, "loss": 0.053, "step": 22900 }, { "grad_norm": 0.4343678057193756, "learning_rate": 8.54880828007998e-05, "loss": 0.0406, "step": 22910 }, { "grad_norm": 0.4013167917728424, "learning_rate": 8.547352010147637e-05, "loss": 0.0452, "step": 22920 }, { "grad_norm": 0.41550007462501526, "learning_rate": 8.545895134069855e-05, "loss": 0.0366, "step": 22930 }, { "grad_norm": 0.38255831599235535, "learning_rate": 8.544437652095576e-05, "loss": 0.0307, "step": 22940 }, { "grad_norm": 0.35456421971321106, "learning_rate": 8.542979564473843e-05, "loss": 0.0483, "step": 22950 }, { "grad_norm": 0.3631385564804077, "learning_rate": 8.541520871453802e-05, "loss": 0.0286, "step": 22960 }, { "grad_norm": 0.4218404293060303, "learning_rate": 8.540061573284705e-05, "loss": 0.0307, "step": 22970 }, { "grad_norm": 0.4623669385910034, "learning_rate": 8.538601670215906e-05, "loss": 0.0371, "step": 22980 }, { "grad_norm": 0.40289410948753357, "learning_rate": 8.537141162496864e-05, "loss": 0.0438, "step": 22990 }, { "grad_norm": 0.4400940239429474, "learning_rate": 8.535680050377137e-05, "loss": 0.032, "step": 23000 }, { "grad_norm": 0.356539785861969, "learning_rate": 8.534218334106391e-05, "loss": 0.0368, "step": 23010 }, { "grad_norm": 0.34105247259140015, "learning_rate": 8.532756013934393e-05, "loss": 0.0348, "step": 23020 }, { "grad_norm": 0.4015099108219147, "learning_rate": 8.531293090111012e-05, "loss": 0.043, "step": 23030 }, { "grad_norm": 0.4142247140407562, "learning_rate": 8.529829562886225e-05, "loss": 0.0376, "step": 23040 }, { "grad_norm": 0.4011203348636627, "learning_rate": 8.528365432510105e-05, "loss": 0.0504, "step": 23050 }, { "grad_norm": 0.34107279777526855, "learning_rate": 8.526900699232833e-05, "loss": 0.0385, "step": 23060 }, { "grad_norm": 0.4815182089805603, "learning_rate": 8.525435363304695e-05, "loss": 0.0376, "step": 23070 }, { "grad_norm": 0.33544594049453735, "learning_rate": 8.523969424976072e-05, "loss": 0.0365, "step": 23080 }, { "grad_norm": 0.37008583545684814, "learning_rate": 8.522502884497457e-05, "loss": 0.0377, "step": 23090 }, { "grad_norm": 0.42611435055732727, "learning_rate": 8.521035742119437e-05, "loss": 0.0444, "step": 23100 }, { "grad_norm": 0.3506646156311035, "learning_rate": 8.519567998092712e-05, "loss": 0.05, "step": 23110 }, { "grad_norm": 0.29969364404678345, "learning_rate": 8.518099652668075e-05, "loss": 0.0396, "step": 23120 }, { "grad_norm": 0.27346110343933105, "learning_rate": 8.516630706096429e-05, "loss": 0.0355, "step": 23130 }, { "grad_norm": 0.35979223251342773, "learning_rate": 8.515161158628773e-05, "loss": 0.0295, "step": 23140 }, { "grad_norm": 0.5405411720275879, "learning_rate": 8.513691010516216e-05, "loss": 0.0526, "step": 23150 }, { "grad_norm": 0.3220446705818176, "learning_rate": 8.512220262009966e-05, "loss": 0.0416, "step": 23160 }, { "grad_norm": 0.43073615431785583, "learning_rate": 8.510748913361332e-05, "loss": 0.0443, "step": 23170 }, { "grad_norm": 0.42975398898124695, "learning_rate": 8.509276964821726e-05, "loss": 0.0353, "step": 23180 }, { "grad_norm": 0.41236361861228943, "learning_rate": 8.507804416642669e-05, "loss": 0.0438, "step": 23190 }, { "grad_norm": 0.495741069316864, "learning_rate": 8.506331269075774e-05, "loss": 0.0395, "step": 23200 }, { "grad_norm": 0.23717144131660461, "learning_rate": 8.504857522372765e-05, "loss": 0.0339, "step": 23210 }, { "grad_norm": 0.3372161388397217, "learning_rate": 8.503383176785461e-05, "loss": 0.0326, "step": 23220 }, { "grad_norm": 0.36592885851860046, "learning_rate": 8.501908232565792e-05, "loss": 0.0357, "step": 23230 }, { "grad_norm": 0.3256005644798279, "learning_rate": 8.50043268996578e-05, "loss": 0.0326, "step": 23240 }, { "grad_norm": 0.2870590090751648, "learning_rate": 8.498956549237562e-05, "loss": 0.0354, "step": 23250 }, { "grad_norm": 0.28205758333206177, "learning_rate": 8.497479810633366e-05, "loss": 0.0318, "step": 23260 }, { "grad_norm": 0.3792552649974823, "learning_rate": 8.496002474405525e-05, "loss": 0.0335, "step": 23270 }, { "grad_norm": 0.439423531293869, "learning_rate": 8.494524540806478e-05, "loss": 0.043, "step": 23280 }, { "grad_norm": 0.2763785719871521, "learning_rate": 8.493046010088761e-05, "loss": 0.0485, "step": 23290 }, { "grad_norm": 0.36452656984329224, "learning_rate": 8.491566882505018e-05, "loss": 0.0386, "step": 23300 }, { "grad_norm": 0.39563503861427307, "learning_rate": 8.490087158307988e-05, "loss": 0.0413, "step": 23310 }, { "grad_norm": 0.6306521892547607, "learning_rate": 8.488606837750518e-05, "loss": 0.0403, "step": 23320 }, { "grad_norm": 0.4139973521232605, "learning_rate": 8.487125921085552e-05, "loss": 0.0432, "step": 23330 }, { "grad_norm": 0.2379145473241806, "learning_rate": 8.485644408566141e-05, "loss": 0.0438, "step": 23340 }, { "grad_norm": 0.32447126507759094, "learning_rate": 8.484162300445431e-05, "loss": 0.0431, "step": 23350 }, { "grad_norm": 0.4747713804244995, "learning_rate": 8.482679596976676e-05, "loss": 0.0421, "step": 23360 }, { "grad_norm": 0.3097721338272095, "learning_rate": 8.48119629841323e-05, "loss": 0.0461, "step": 23370 }, { "grad_norm": 0.34914296865463257, "learning_rate": 8.479712405008547e-05, "loss": 0.0517, "step": 23380 }, { "grad_norm": 0.3911955952644348, "learning_rate": 8.478227917016184e-05, "loss": 0.0484, "step": 23390 }, { "grad_norm": 0.28815460205078125, "learning_rate": 8.476742834689801e-05, "loss": 0.0476, "step": 23400 }, { "grad_norm": 0.3456789553165436, "learning_rate": 8.475257158283157e-05, "loss": 0.045, "step": 23410 }, { "grad_norm": 0.3569360673427582, "learning_rate": 8.473770888050112e-05, "loss": 0.0368, "step": 23420 }, { "grad_norm": 0.27215415239334106, "learning_rate": 8.47228402424463e-05, "loss": 0.0404, "step": 23430 }, { "grad_norm": 0.3015184998512268, "learning_rate": 8.470796567120775e-05, "loss": 0.0384, "step": 23440 }, { "grad_norm": 0.39913079142570496, "learning_rate": 8.469308516932714e-05, "loss": 0.0469, "step": 23450 }, { "grad_norm": 0.46880197525024414, "learning_rate": 8.467819873934714e-05, "loss": 0.04, "step": 23460 }, { "grad_norm": 0.45341256260871887, "learning_rate": 8.466330638381143e-05, "loss": 0.0405, "step": 23470 }, { "grad_norm": 0.4190179109573364, "learning_rate": 8.464840810526469e-05, "loss": 0.04, "step": 23480 }, { "grad_norm": 0.3892393410205841, "learning_rate": 8.463350390625264e-05, "loss": 0.0408, "step": 23490 }, { "grad_norm": 0.3353746831417084, "learning_rate": 8.4618593789322e-05, "loss": 0.0563, "step": 23500 }, { "grad_norm": 0.3500174582004547, "learning_rate": 8.46036777570205e-05, "loss": 0.0411, "step": 23510 }, { "grad_norm": 0.4396076500415802, "learning_rate": 8.458875581189688e-05, "loss": 0.0605, "step": 23520 }, { "grad_norm": 0.3689144551753998, "learning_rate": 8.457382795650092e-05, "loss": 0.0412, "step": 23530 }, { "grad_norm": 0.46727290749549866, "learning_rate": 8.455889419338335e-05, "loss": 0.0474, "step": 23540 }, { "grad_norm": 0.4136229157447815, "learning_rate": 8.454395452509593e-05, "loss": 0.0394, "step": 23550 }, { "grad_norm": 0.3535958230495453, "learning_rate": 8.452900895419146e-05, "loss": 0.0383, "step": 23560 }, { "grad_norm": 0.31150108575820923, "learning_rate": 8.451405748322376e-05, "loss": 0.0412, "step": 23570 }, { "grad_norm": 0.4583132565021515, "learning_rate": 8.449910011474759e-05, "loss": 0.0412, "step": 23580 }, { "grad_norm": 0.44240278005599976, "learning_rate": 8.448413685131876e-05, "loss": 0.0455, "step": 23590 }, { "grad_norm": 0.3491165041923523, "learning_rate": 8.446916769549407e-05, "loss": 0.0366, "step": 23600 }, { "grad_norm": 0.263944149017334, "learning_rate": 8.445419264983136e-05, "loss": 0.043, "step": 23610 }, { "grad_norm": 0.30194738507270813, "learning_rate": 8.443921171688947e-05, "loss": 0.0425, "step": 23620 }, { "grad_norm": 0.388961523771286, "learning_rate": 8.442422489922819e-05, "loss": 0.0347, "step": 23630 }, { "grad_norm": 0.42216286063194275, "learning_rate": 8.440923219940838e-05, "loss": 0.0565, "step": 23640 }, { "grad_norm": 0.6105095744132996, "learning_rate": 8.439423361999189e-05, "loss": 0.0562, "step": 23650 }, { "grad_norm": 0.38634398579597473, "learning_rate": 8.437922916354155e-05, "loss": 0.0453, "step": 23660 }, { "grad_norm": 0.324392706155777, "learning_rate": 8.436421883262123e-05, "loss": 0.0407, "step": 23670 }, { "grad_norm": 0.30379459261894226, "learning_rate": 8.434920262979577e-05, "loss": 0.0556, "step": 23680 }, { "grad_norm": 0.43797826766967773, "learning_rate": 8.433418055763104e-05, "loss": 0.0478, "step": 23690 }, { "grad_norm": 0.35024866461753845, "learning_rate": 8.431915261869389e-05, "loss": 0.045, "step": 23700 }, { "grad_norm": 0.36255937814712524, "learning_rate": 8.43041188155522e-05, "loss": 0.0418, "step": 23710 }, { "grad_norm": 0.43557387590408325, "learning_rate": 8.428907915077481e-05, "loss": 0.0424, "step": 23720 }, { "grad_norm": 0.4772786796092987, "learning_rate": 8.42740336269316e-05, "loss": 0.0511, "step": 23730 }, { "grad_norm": 0.5120342373847961, "learning_rate": 8.425898224659345e-05, "loss": 0.034, "step": 23740 }, { "grad_norm": 0.35042130947113037, "learning_rate": 8.42439250123322e-05, "loss": 0.0424, "step": 23750 }, { "grad_norm": 0.563557505607605, "learning_rate": 8.422886192672076e-05, "loss": 0.0447, "step": 23760 }, { "grad_norm": 0.31430578231811523, "learning_rate": 8.421379299233297e-05, "loss": 0.0378, "step": 23770 }, { "grad_norm": 0.3722459375858307, "learning_rate": 8.419871821174371e-05, "loss": 0.0383, "step": 23780 }, { "grad_norm": 0.3776662051677704, "learning_rate": 8.418363758752884e-05, "loss": 0.0371, "step": 23790 }, { "grad_norm": 0.31939584016799927, "learning_rate": 8.416855112226523e-05, "loss": 0.0421, "step": 23800 }, { "grad_norm": 0.4017135202884674, "learning_rate": 8.415345881853075e-05, "loss": 0.0366, "step": 23810 }, { "grad_norm": 0.4640204906463623, "learning_rate": 8.413836067890426e-05, "loss": 0.0365, "step": 23820 }, { "grad_norm": 0.3655495047569275, "learning_rate": 8.41232567059656e-05, "loss": 0.038, "step": 23830 }, { "grad_norm": 0.28555649518966675, "learning_rate": 8.410814690229565e-05, "loss": 0.0433, "step": 23840 }, { "grad_norm": 0.3641728162765503, "learning_rate": 8.409303127047626e-05, "loss": 0.0399, "step": 23850 }, { "grad_norm": 0.47697630524635315, "learning_rate": 8.407790981309028e-05, "loss": 0.0394, "step": 23860 }, { "grad_norm": 0.31464821100234985, "learning_rate": 8.406278253272153e-05, "loss": 0.0435, "step": 23870 }, { "grad_norm": 0.32221370935440063, "learning_rate": 8.404764943195487e-05, "loss": 0.0396, "step": 23880 }, { "grad_norm": 0.4256116449832916, "learning_rate": 8.403251051337613e-05, "loss": 0.0408, "step": 23890 }, { "grad_norm": 0.32697802782058716, "learning_rate": 8.401736577957214e-05, "loss": 0.0341, "step": 23900 }, { "grad_norm": 0.3488301634788513, "learning_rate": 8.40022152331307e-05, "loss": 0.0375, "step": 23910 }, { "grad_norm": 0.56453937292099, "learning_rate": 8.398705887664064e-05, "loss": 0.0464, "step": 23920 }, { "grad_norm": 0.2651853561401367, "learning_rate": 8.397189671269177e-05, "loss": 0.0335, "step": 23930 }, { "grad_norm": 0.3397473692893982, "learning_rate": 8.395672874387488e-05, "loss": 0.0276, "step": 23940 }, { "grad_norm": 0.3236231803894043, "learning_rate": 8.394155497278177e-05, "loss": 0.0309, "step": 23950 }, { "grad_norm": 0.5639981031417847, "learning_rate": 8.392637540200523e-05, "loss": 0.0566, "step": 23960 }, { "grad_norm": 0.31851956248283386, "learning_rate": 8.391119003413902e-05, "loss": 0.0445, "step": 23970 }, { "grad_norm": 0.3534126579761505, "learning_rate": 8.38959988717779e-05, "loss": 0.0329, "step": 23980 }, { "grad_norm": 0.30622121691703796, "learning_rate": 8.388080191751764e-05, "loss": 0.0419, "step": 23990 }, { "grad_norm": 0.36993902921676636, "learning_rate": 8.386559917395496e-05, "loss": 0.0341, "step": 24000 }, { "grad_norm": 0.274532675743103, "learning_rate": 8.385039064368761e-05, "loss": 0.0377, "step": 24010 }, { "grad_norm": 0.26382631063461304, "learning_rate": 8.383517632931431e-05, "loss": 0.0339, "step": 24020 }, { "grad_norm": 0.38836798071861267, "learning_rate": 8.381995623343477e-05, "loss": 0.0411, "step": 24030 }, { "grad_norm": 0.48026782274246216, "learning_rate": 8.380473035864968e-05, "loss": 0.0358, "step": 24040 }, { "grad_norm": 0.3013838827610016, "learning_rate": 8.378949870756076e-05, "loss": 0.0398, "step": 24050 }, { "grad_norm": 0.33148249983787537, "learning_rate": 8.377426128277063e-05, "loss": 0.0418, "step": 24060 }, { "grad_norm": 0.3197381794452667, "learning_rate": 8.375901808688298e-05, "loss": 0.0295, "step": 24070 }, { "grad_norm": 0.40156203508377075, "learning_rate": 8.374376912250246e-05, "loss": 0.0493, "step": 24080 }, { "grad_norm": 0.28702646493911743, "learning_rate": 8.372851439223468e-05, "loss": 0.0344, "step": 24090 }, { "grad_norm": 0.38651660084724426, "learning_rate": 8.371325389868627e-05, "loss": 0.0455, "step": 24100 }, { "grad_norm": 0.6556715369224548, "learning_rate": 8.369798764446482e-05, "loss": 0.0378, "step": 24110 }, { "grad_norm": 0.4283306300640106, "learning_rate": 8.368271563217893e-05, "loss": 0.0309, "step": 24120 }, { "grad_norm": 0.29921290278434753, "learning_rate": 8.366743786443817e-05, "loss": 0.0427, "step": 24130 }, { "grad_norm": 0.3301067650318146, "learning_rate": 8.365215434385309e-05, "loss": 0.0419, "step": 24140 }, { "grad_norm": 0.31924113631248474, "learning_rate": 8.36368650730352e-05, "loss": 0.0336, "step": 24150 }, { "grad_norm": 0.3219988942146301, "learning_rate": 8.362157005459705e-05, "loss": 0.0427, "step": 24160 }, { "grad_norm": 0.47154465317726135, "learning_rate": 8.360626929115213e-05, "loss": 0.0374, "step": 24170 }, { "grad_norm": 0.3367515206336975, "learning_rate": 8.359096278531492e-05, "loss": 0.0394, "step": 24180 }, { "grad_norm": 0.3942396938800812, "learning_rate": 8.357565053970088e-05, "loss": 0.0364, "step": 24190 }, { "grad_norm": 0.28665316104888916, "learning_rate": 8.356033255692647e-05, "loss": 0.0319, "step": 24200 }, { "grad_norm": 0.36011558771133423, "learning_rate": 8.354500883960911e-05, "loss": 0.0335, "step": 24210 }, { "grad_norm": 0.35091620683670044, "learning_rate": 8.352967939036717e-05, "loss": 0.0392, "step": 24220 }, { "grad_norm": 0.4321282207965851, "learning_rate": 8.35143442118201e-05, "loss": 0.0462, "step": 24230 }, { "grad_norm": 0.37432971596717834, "learning_rate": 8.349900330658819e-05, "loss": 0.04, "step": 24240 }, { "grad_norm": 0.3596404790878296, "learning_rate": 8.348365667729284e-05, "loss": 0.0453, "step": 24250 }, { "grad_norm": 0.36690250039100647, "learning_rate": 8.346830432655633e-05, "loss": 0.032, "step": 24260 }, { "grad_norm": 0.39830124378204346, "learning_rate": 8.345294625700195e-05, "loss": 0.0374, "step": 24270 }, { "grad_norm": 0.4679827392101288, "learning_rate": 8.343758247125402e-05, "loss": 0.0341, "step": 24280 }, { "grad_norm": 0.5275177359580994, "learning_rate": 8.342221297193776e-05, "loss": 0.0479, "step": 24290 }, { "grad_norm": 0.24078798294067383, "learning_rate": 8.34068377616794e-05, "loss": 0.0446, "step": 24300 }, { "grad_norm": 0.4122019410133362, "learning_rate": 8.339145684310615e-05, "loss": 0.039, "step": 24310 }, { "grad_norm": 0.3424845337867737, "learning_rate": 8.337607021884618e-05, "loss": 0.0341, "step": 24320 }, { "grad_norm": 0.28680840134620667, "learning_rate": 8.336067789152867e-05, "loss": 0.0274, "step": 24330 }, { "grad_norm": 0.30638548731803894, "learning_rate": 8.334527986378369e-05, "loss": 0.039, "step": 24340 }, { "grad_norm": 0.40284016728401184, "learning_rate": 8.332987613824239e-05, "loss": 0.0478, "step": 24350 }, { "grad_norm": 0.4369588792324066, "learning_rate": 8.331446671753685e-05, "loss": 0.0349, "step": 24360 }, { "grad_norm": 0.31089386343955994, "learning_rate": 8.329905160430007e-05, "loss": 0.0371, "step": 24370 }, { "grad_norm": 0.5273174047470093, "learning_rate": 8.328363080116611e-05, "loss": 0.0349, "step": 24380 }, { "grad_norm": 0.31856438517570496, "learning_rate": 8.326820431076997e-05, "loss": 0.0368, "step": 24390 }, { "grad_norm": 0.4168931245803833, "learning_rate": 8.325277213574759e-05, "loss": 0.0414, "step": 24400 }, { "grad_norm": 0.5163846015930176, "learning_rate": 8.32373342787359e-05, "loss": 0.035, "step": 24410 }, { "grad_norm": 0.40572282671928406, "learning_rate": 8.322189074237285e-05, "loss": 0.0359, "step": 24420 }, { "grad_norm": 0.4560743570327759, "learning_rate": 8.32064415292973e-05, "loss": 0.0462, "step": 24430 }, { "grad_norm": 0.3135559558868408, "learning_rate": 8.319098664214907e-05, "loss": 0.0309, "step": 24440 }, { "grad_norm": 0.36122798919677734, "learning_rate": 8.3175526083569e-05, "loss": 0.045, "step": 24450 }, { "grad_norm": 0.2741647958755493, "learning_rate": 8.316005985619889e-05, "loss": 0.0351, "step": 24460 }, { "grad_norm": 0.35958048701286316, "learning_rate": 8.314458796268147e-05, "loss": 0.0422, "step": 24470 }, { "grad_norm": 0.30307623744010925, "learning_rate": 8.312911040566047e-05, "loss": 0.0415, "step": 24480 }, { "grad_norm": 0.3811935484409332, "learning_rate": 8.31136271877806e-05, "loss": 0.0346, "step": 24490 }, { "grad_norm": 0.41277623176574707, "learning_rate": 8.309813831168748e-05, "loss": 0.0453, "step": 24500 }, { "grad_norm": 0.4568603038787842, "learning_rate": 8.308264378002777e-05, "loss": 0.0352, "step": 24510 }, { "grad_norm": 0.37363091111183167, "learning_rate": 8.306714359544906e-05, "loss": 0.049, "step": 24520 }, { "grad_norm": 0.389091432094574, "learning_rate": 8.30516377605999e-05, "loss": 0.0349, "step": 24530 }, { "grad_norm": 0.36732184886932373, "learning_rate": 8.30361262781298e-05, "loss": 0.0322, "step": 24540 }, { "grad_norm": 0.3067933917045593, "learning_rate": 8.302060915068924e-05, "loss": 0.0359, "step": 24550 }, { "grad_norm": 0.2393970489501953, "learning_rate": 8.300508638092972e-05, "loss": 0.0292, "step": 24560 }, { "grad_norm": 0.32076236605644226, "learning_rate": 8.298955797150361e-05, "loss": 0.0337, "step": 24570 }, { "grad_norm": 0.5177778005599976, "learning_rate": 8.297402392506433e-05, "loss": 0.0352, "step": 24580 }, { "grad_norm": 0.4178493916988373, "learning_rate": 8.295848424426617e-05, "loss": 0.0284, "step": 24590 }, { "grad_norm": 0.33221039175987244, "learning_rate": 8.29429389317645e-05, "loss": 0.0318, "step": 24600 }, { "grad_norm": 0.25546935200691223, "learning_rate": 8.292738799021556e-05, "loss": 0.0414, "step": 24610 }, { "grad_norm": 0.26992595195770264, "learning_rate": 8.291183142227656e-05, "loss": 0.0305, "step": 24620 }, { "grad_norm": 0.30381715297698975, "learning_rate": 8.289626923060572e-05, "loss": 0.0466, "step": 24630 }, { "grad_norm": 0.3839510977268219, "learning_rate": 8.288070141786218e-05, "loss": 0.0417, "step": 24640 }, { "grad_norm": 0.301843523979187, "learning_rate": 8.286512798670605e-05, "loss": 0.0368, "step": 24650 }, { "grad_norm": 0.4570848047733307, "learning_rate": 8.284954893979842e-05, "loss": 0.0455, "step": 24660 }, { "grad_norm": 0.32159578800201416, "learning_rate": 8.283396427980131e-05, "loss": 0.0418, "step": 24670 }, { "grad_norm": 0.3833651840686798, "learning_rate": 8.281837400937771e-05, "loss": 0.0362, "step": 24680 }, { "grad_norm": 0.36977308988571167, "learning_rate": 8.28027781311916e-05, "loss": 0.0355, "step": 24690 }, { "grad_norm": 0.4166921079158783, "learning_rate": 8.278717664790785e-05, "loss": 0.0381, "step": 24700 }, { "grad_norm": 0.42737701535224915, "learning_rate": 8.277156956219234e-05, "loss": 0.0452, "step": 24710 }, { "grad_norm": 0.46604350209236145, "learning_rate": 8.275595687671189e-05, "loss": 0.0369, "step": 24720 }, { "grad_norm": 0.2596409320831299, "learning_rate": 8.27403385941343e-05, "loss": 0.036, "step": 24730 }, { "grad_norm": 0.5402652621269226, "learning_rate": 8.272471471712828e-05, "loss": 0.0351, "step": 24740 }, { "grad_norm": 0.3058589994907379, "learning_rate": 8.270908524836355e-05, "loss": 0.0403, "step": 24750 }, { "grad_norm": 0.3583117425441742, "learning_rate": 8.269345019051074e-05, "loss": 0.0465, "step": 24760 }, { "grad_norm": 0.39004501700401306, "learning_rate": 8.267780954624147e-05, "loss": 0.0377, "step": 24770 }, { "grad_norm": 0.4254257380962372, "learning_rate": 8.266216331822827e-05, "loss": 0.0438, "step": 24780 }, { "grad_norm": 0.3907501697540283, "learning_rate": 8.264651150914469e-05, "loss": 0.0368, "step": 24790 }, { "grad_norm": 0.4190019369125366, "learning_rate": 8.263085412166517e-05, "loss": 0.0356, "step": 24800 }, { "grad_norm": 0.3351989686489105, "learning_rate": 8.261519115846514e-05, "loss": 0.038, "step": 24810 }, { "grad_norm": 0.4328312873840332, "learning_rate": 8.259952262222096e-05, "loss": 0.0355, "step": 24820 }, { "grad_norm": 0.27332547307014465, "learning_rate": 8.258384851560997e-05, "loss": 0.0435, "step": 24830 }, { "grad_norm": 0.3922070562839508, "learning_rate": 8.256816884131044e-05, "loss": 0.0367, "step": 24840 }, { "grad_norm": 0.43185505270957947, "learning_rate": 8.255248360200159e-05, "loss": 0.0375, "step": 24850 }, { "grad_norm": 0.5363869071006775, "learning_rate": 8.253679280036359e-05, "loss": 0.0319, "step": 24860 }, { "grad_norm": 0.36312010884284973, "learning_rate": 8.252109643907762e-05, "loss": 0.048, "step": 24870 }, { "grad_norm": 0.32812443375587463, "learning_rate": 8.250539452082569e-05, "loss": 0.0301, "step": 24880 }, { "grad_norm": 0.3943787217140198, "learning_rate": 8.248968704829087e-05, "loss": 0.0412, "step": 24890 }, { "grad_norm": 0.2823750078678131, "learning_rate": 8.247397402415714e-05, "loss": 0.0509, "step": 24900 }, { "grad_norm": 0.54642653465271, "learning_rate": 8.24582554511094e-05, "loss": 0.0358, "step": 24910 }, { "grad_norm": 0.3138369023799896, "learning_rate": 8.244253133183355e-05, "loss": 0.0307, "step": 24920 }, { "grad_norm": 0.3771836459636688, "learning_rate": 8.24268016690164e-05, "loss": 0.0339, "step": 24930 }, { "grad_norm": 0.4068857729434967, "learning_rate": 8.241106646534571e-05, "loss": 0.0518, "step": 24940 }, { "grad_norm": 0.5794603824615479, "learning_rate": 8.23953257235102e-05, "loss": 0.0495, "step": 24950 }, { "grad_norm": 0.31331193447113037, "learning_rate": 8.237957944619956e-05, "loss": 0.0371, "step": 24960 }, { "grad_norm": 0.39770662784576416, "learning_rate": 8.236382763610437e-05, "loss": 0.0395, "step": 24970 }, { "grad_norm": 0.4207095503807068, "learning_rate": 8.234807029591619e-05, "loss": 0.0363, "step": 24980 }, { "grad_norm": 0.41365846991539, "learning_rate": 8.233230742832752e-05, "loss": 0.0482, "step": 24990 }, { "grad_norm": 0.41928473114967346, "learning_rate": 8.231653903603178e-05, "loss": 0.0379, "step": 25000 }, { "grad_norm": 0.33148834109306335, "learning_rate": 8.23007651217234e-05, "loss": 0.0386, "step": 25010 }, { "grad_norm": 0.37171563506126404, "learning_rate": 8.228498568809769e-05, "loss": 0.0393, "step": 25020 }, { "grad_norm": 0.4247572720050812, "learning_rate": 8.22692007378509e-05, "loss": 0.0407, "step": 25030 }, { "grad_norm": 0.36958062648773193, "learning_rate": 8.225341027368028e-05, "loss": 0.0368, "step": 25040 }, { "grad_norm": 0.36790990829467773, "learning_rate": 8.223761429828399e-05, "loss": 0.0379, "step": 25050 }, { "grad_norm": 0.4005896747112274, "learning_rate": 8.22218128143611e-05, "loss": 0.0399, "step": 25060 }, { "grad_norm": 0.375597208738327, "learning_rate": 8.220600582461166e-05, "loss": 0.0356, "step": 25070 }, { "grad_norm": 0.23129917681217194, "learning_rate": 8.219019333173668e-05, "loss": 0.0378, "step": 25080 }, { "grad_norm": 0.5707178115844727, "learning_rate": 8.217437533843805e-05, "loss": 0.0487, "step": 25090 }, { "grad_norm": 0.47403523325920105, "learning_rate": 8.215855184741867e-05, "loss": 0.0454, "step": 25100 }, { "grad_norm": 0.48936229944229126, "learning_rate": 8.21427228613823e-05, "loss": 0.0367, "step": 25110 }, { "grad_norm": 0.4165499806404114, "learning_rate": 8.21268883830337e-05, "loss": 0.0405, "step": 25120 }, { "grad_norm": 0.326204776763916, "learning_rate": 8.211104841507855e-05, "loss": 0.0459, "step": 25130 }, { "grad_norm": 0.3730461895465851, "learning_rate": 8.209520296022346e-05, "loss": 0.0369, "step": 25140 }, { "grad_norm": 0.5954432487487793, "learning_rate": 8.207935202117599e-05, "loss": 0.0399, "step": 25150 }, { "grad_norm": 0.2895020544528961, "learning_rate": 8.206349560064463e-05, "loss": 0.0375, "step": 25160 }, { "grad_norm": 0.43661418557167053, "learning_rate": 8.204763370133881e-05, "loss": 0.0347, "step": 25170 }, { "grad_norm": 0.33309757709503174, "learning_rate": 8.203176632596892e-05, "loss": 0.0394, "step": 25180 }, { "grad_norm": 0.44551658630371094, "learning_rate": 8.20158934772462e-05, "loss": 0.0525, "step": 25190 }, { "grad_norm": 0.35299888253211975, "learning_rate": 8.200001515788294e-05, "loss": 0.0346, "step": 25200 }, { "grad_norm": 0.4110110402107239, "learning_rate": 8.198413137059228e-05, "loss": 0.0421, "step": 25210 }, { "grad_norm": 0.24410638213157654, "learning_rate": 8.196824211808835e-05, "loss": 0.0337, "step": 25220 }, { "grad_norm": 0.44860604405403137, "learning_rate": 8.195234740308617e-05, "loss": 0.0498, "step": 25230 }, { "grad_norm": 0.4491843283176422, "learning_rate": 8.193644722830171e-05, "loss": 0.0343, "step": 25240 }, { "grad_norm": 0.32048606872558594, "learning_rate": 8.19205415964519e-05, "loss": 0.0317, "step": 25250 }, { "grad_norm": 0.30388766527175903, "learning_rate": 8.190463051025456e-05, "loss": 0.0328, "step": 25260 }, { "grad_norm": 0.30014029145240784, "learning_rate": 8.188871397242843e-05, "loss": 0.0361, "step": 25270 }, { "grad_norm": 0.3719833493232727, "learning_rate": 8.187279198569326e-05, "loss": 0.0364, "step": 25280 }, { "grad_norm": 0.33194348216056824, "learning_rate": 8.185686455276966e-05, "loss": 0.0314, "step": 25290 }, { "grad_norm": 0.287749707698822, "learning_rate": 8.184093167637921e-05, "loss": 0.0429, "step": 25300 }, { "grad_norm": 0.35625913739204407, "learning_rate": 8.182499335924437e-05, "loss": 0.0374, "step": 25310 }, { "grad_norm": 0.44134533405303955, "learning_rate": 8.18090496040886e-05, "loss": 0.0321, "step": 25320 }, { "grad_norm": 0.4514225721359253, "learning_rate": 8.179310041363621e-05, "loss": 0.0517, "step": 25330 }, { "grad_norm": 0.39103466272354126, "learning_rate": 8.17771457906125e-05, "loss": 0.0379, "step": 25340 }, { "grad_norm": 0.32766619324684143, "learning_rate": 8.176118573774371e-05, "loss": 0.0336, "step": 25350 }, { "grad_norm": 0.3986833691596985, "learning_rate": 8.174522025775692e-05, "loss": 0.0336, "step": 25360 }, { "grad_norm": 0.3779258131980896, "learning_rate": 8.172924935338022e-05, "loss": 0.0376, "step": 25370 }, { "grad_norm": 0.3363264799118042, "learning_rate": 8.171327302734262e-05, "loss": 0.0358, "step": 25380 }, { "grad_norm": 0.3750261962413788, "learning_rate": 8.169729128237401e-05, "loss": 0.0389, "step": 25390 }, { "grad_norm": 0.36107221245765686, "learning_rate": 8.168130412120525e-05, "loss": 0.0332, "step": 25400 }, { "grad_norm": 0.45020249485969543, "learning_rate": 8.16653115465681e-05, "loss": 0.0345, "step": 25410 }, { "grad_norm": 0.4982074797153473, "learning_rate": 8.164931356119526e-05, "loss": 0.0354, "step": 25420 }, { "grad_norm": 0.44438862800598145, "learning_rate": 8.163331016782032e-05, "loss": 0.0498, "step": 25430 }, { "grad_norm": 0.38684889674186707, "learning_rate": 8.161730136917785e-05, "loss": 0.0353, "step": 25440 }, { "grad_norm": 0.31371280550956726, "learning_rate": 8.160128716800333e-05, "loss": 0.0445, "step": 25450 }, { "grad_norm": 0.3660390079021454, "learning_rate": 8.158526756703313e-05, "loss": 0.0297, "step": 25460 }, { "grad_norm": 0.3935338854789734, "learning_rate": 8.156924256900455e-05, "loss": 0.044, "step": 25470 }, { "grad_norm": 0.4511154294013977, "learning_rate": 8.155321217665584e-05, "loss": 0.0371, "step": 25480 }, { "grad_norm": 0.2529715597629547, "learning_rate": 8.153717639272614e-05, "loss": 0.0349, "step": 25490 }, { "grad_norm": 0.4790585935115814, "learning_rate": 8.152113521995555e-05, "loss": 0.043, "step": 25500 }, { "grad_norm": 0.4349554777145386, "learning_rate": 8.150508866108505e-05, "loss": 0.0357, "step": 25510 }, { "grad_norm": 0.3739681839942932, "learning_rate": 8.148903671885657e-05, "loss": 0.0333, "step": 25520 }, { "grad_norm": 0.5277025103569031, "learning_rate": 8.147297939601292e-05, "loss": 0.0413, "step": 25530 }, { "grad_norm": 0.4492665231227875, "learning_rate": 8.145691669529792e-05, "loss": 0.0355, "step": 25540 }, { "grad_norm": 0.43140730261802673, "learning_rate": 8.144084861945618e-05, "loss": 0.0312, "step": 25550 }, { "grad_norm": 0.3350101709365845, "learning_rate": 8.142477517123333e-05, "loss": 0.0375, "step": 25560 }, { "grad_norm": 0.46503710746765137, "learning_rate": 8.140869635337586e-05, "loss": 0.0363, "step": 25570 }, { "grad_norm": 0.3441832363605499, "learning_rate": 8.139261216863123e-05, "loss": 0.0344, "step": 25580 }, { "grad_norm": 0.24411389231681824, "learning_rate": 8.137652261974776e-05, "loss": 0.038, "step": 25590 }, { "grad_norm": 0.3171435296535492, "learning_rate": 8.136042770947472e-05, "loss": 0.0323, "step": 25600 }, { "grad_norm": 0.288541316986084, "learning_rate": 8.134432744056228e-05, "loss": 0.0343, "step": 25610 }, { "grad_norm": 0.5807042717933655, "learning_rate": 8.132822181576158e-05, "loss": 0.0422, "step": 25620 }, { "grad_norm": 0.3020765483379364, "learning_rate": 8.131211083782459e-05, "loss": 0.0355, "step": 25630 }, { "grad_norm": 0.3948400318622589, "learning_rate": 8.129599450950424e-05, "loss": 0.0337, "step": 25640 }, { "grad_norm": 0.39932364225387573, "learning_rate": 8.127987283355438e-05, "loss": 0.0405, "step": 25650 }, { "grad_norm": 0.3889129161834717, "learning_rate": 8.126374581272976e-05, "loss": 0.0293, "step": 25660 }, { "grad_norm": 0.32412806153297424, "learning_rate": 8.124761344978605e-05, "loss": 0.047, "step": 25670 }, { "grad_norm": 0.37887465953826904, "learning_rate": 8.12314757474798e-05, "loss": 0.0288, "step": 25680 }, { "grad_norm": 0.3234838545322418, "learning_rate": 8.121533270856856e-05, "loss": 0.0321, "step": 25690 }, { "grad_norm": 0.4064747393131256, "learning_rate": 8.119918433581069e-05, "loss": 0.0324, "step": 25700 }, { "grad_norm": 0.3785143494606018, "learning_rate": 8.118303063196551e-05, "loss": 0.0302, "step": 25710 }, { "grad_norm": 0.4258919358253479, "learning_rate": 8.116687159979326e-05, "loss": 0.034, "step": 25720 }, { "grad_norm": 0.37841421365737915, "learning_rate": 8.115070724205508e-05, "loss": 0.0376, "step": 25730 }, { "grad_norm": 0.41836869716644287, "learning_rate": 8.113453756151296e-05, "loss": 0.0313, "step": 25740 }, { "grad_norm": 0.47415784001350403, "learning_rate": 8.111836256092995e-05, "loss": 0.0288, "step": 25750 }, { "grad_norm": 0.3687664866447449, "learning_rate": 8.110218224306985e-05, "loss": 0.0349, "step": 25760 }, { "grad_norm": 0.3343181610107422, "learning_rate": 8.108599661069745e-05, "loss": 0.0302, "step": 25770 }, { "grad_norm": 0.43976402282714844, "learning_rate": 8.106980566657845e-05, "loss": 0.0306, "step": 25780 }, { "grad_norm": 0.3592569828033447, "learning_rate": 8.10536094134794e-05, "loss": 0.0272, "step": 25790 }, { "grad_norm": 0.3597477078437805, "learning_rate": 8.103740785416783e-05, "loss": 0.0301, "step": 25800 }, { "grad_norm": 0.34747177362442017, "learning_rate": 8.102120099141212e-05, "loss": 0.0308, "step": 25810 }, { "grad_norm": 0.2590218484401703, "learning_rate": 8.100498882798163e-05, "loss": 0.0267, "step": 25820 }, { "grad_norm": 0.33262813091278076, "learning_rate": 8.09887713666465e-05, "loss": 0.0305, "step": 25830 }, { "grad_norm": 0.2938781678676605, "learning_rate": 8.09725486101779e-05, "loss": 0.0438, "step": 25840 }, { "grad_norm": 0.27309107780456543, "learning_rate": 8.095632056134784e-05, "loss": 0.0363, "step": 25850 }, { "grad_norm": 0.48759764432907104, "learning_rate": 8.094008722292925e-05, "loss": 0.0314, "step": 25860 }, { "grad_norm": 0.3860134780406952, "learning_rate": 8.092384859769598e-05, "loss": 0.0379, "step": 25870 }, { "grad_norm": 0.4532302916049957, "learning_rate": 8.090760468842275e-05, "loss": 0.0525, "step": 25880 }, { "grad_norm": 0.26998135447502136, "learning_rate": 8.089135549788521e-05, "loss": 0.0324, "step": 25890 }, { "grad_norm": 0.49438080191612244, "learning_rate": 8.087510102885987e-05, "loss": 0.0323, "step": 25900 }, { "grad_norm": 0.23743480443954468, "learning_rate": 8.085884128412422e-05, "loss": 0.0308, "step": 25910 }, { "grad_norm": 0.4531894028186798, "learning_rate": 8.084257626645659e-05, "loss": 0.0348, "step": 25920 }, { "grad_norm": 0.3902311325073242, "learning_rate": 8.08263059786362e-05, "loss": 0.034, "step": 25930 }, { "grad_norm": 0.39534321427345276, "learning_rate": 8.081003042344325e-05, "loss": 0.0252, "step": 25940 }, { "grad_norm": 0.3409890830516815, "learning_rate": 8.079374960365872e-05, "loss": 0.0362, "step": 25950 }, { "grad_norm": 0.41484174132347107, "learning_rate": 8.077746352206463e-05, "loss": 0.0415, "step": 25960 }, { "grad_norm": 0.4331217110157013, "learning_rate": 8.076117218144377e-05, "loss": 0.0316, "step": 25970 }, { "grad_norm": 0.38115257024765015, "learning_rate": 8.074487558457991e-05, "loss": 0.0319, "step": 25980 }, { "grad_norm": 0.2873573303222656, "learning_rate": 8.072857373425768e-05, "loss": 0.0348, "step": 25990 }, { "grad_norm": 0.44626981019973755, "learning_rate": 8.071226663326264e-05, "loss": 0.0333, "step": 26000 }, { "grad_norm": 0.4406629502773285, "learning_rate": 8.069595428438121e-05, "loss": 0.0346, "step": 26010 }, { "grad_norm": 0.32575222849845886, "learning_rate": 8.067963669040072e-05, "loss": 0.0283, "step": 26020 }, { "grad_norm": 0.45244747400283813, "learning_rate": 8.066331385410942e-05, "loss": 0.0387, "step": 26030 }, { "grad_norm": 0.3053961992263794, "learning_rate": 8.064698577829641e-05, "loss": 0.0373, "step": 26040 }, { "grad_norm": 0.3329750597476959, "learning_rate": 8.063065246575175e-05, "loss": 0.0399, "step": 26050 }, { "grad_norm": 0.5798983573913574, "learning_rate": 8.061431391926631e-05, "loss": 0.0324, "step": 26060 }, { "grad_norm": 0.24122315645217896, "learning_rate": 8.059797014163195e-05, "loss": 0.0397, "step": 26070 }, { "grad_norm": 0.4569953680038452, "learning_rate": 8.058162113564133e-05, "loss": 0.0468, "step": 26080 }, { "grad_norm": 0.2765665352344513, "learning_rate": 8.056526690408806e-05, "loss": 0.0392, "step": 26090 }, { "grad_norm": 0.28487181663513184, "learning_rate": 8.054890744976666e-05, "loss": 0.0403, "step": 26100 }, { "grad_norm": 0.2756977081298828, "learning_rate": 8.053254277547248e-05, "loss": 0.0332, "step": 26110 }, { "grad_norm": 0.4899536669254303, "learning_rate": 8.051617288400182e-05, "loss": 0.0365, "step": 26120 }, { "grad_norm": 0.3708132207393646, "learning_rate": 8.049979777815182e-05, "loss": 0.0378, "step": 26130 }, { "grad_norm": 0.24313804507255554, "learning_rate": 8.048341746072054e-05, "loss": 0.04, "step": 26140 }, { "grad_norm": 0.2864110469818115, "learning_rate": 8.046703193450696e-05, "loss": 0.0482, "step": 26150 }, { "grad_norm": 0.32194676995277405, "learning_rate": 8.04506412023109e-05, "loss": 0.0305, "step": 26160 }, { "grad_norm": 0.27694541215896606, "learning_rate": 8.043424526693306e-05, "loss": 0.0339, "step": 26170 }, { "grad_norm": 0.3432939350605011, "learning_rate": 8.04178441311751e-05, "loss": 0.0357, "step": 26180 }, { "grad_norm": 0.44760996103286743, "learning_rate": 8.04014377978395e-05, "loss": 0.0422, "step": 26190 }, { "grad_norm": 0.3424156904220581, "learning_rate": 8.038502626972967e-05, "loss": 0.0377, "step": 26200 }, { "grad_norm": 0.312992125749588, "learning_rate": 8.036860954964989e-05, "loss": 0.0419, "step": 26210 }, { "grad_norm": 0.3550204038619995, "learning_rate": 8.035218764040531e-05, "loss": 0.0402, "step": 26220 }, { "grad_norm": 0.3166965842247009, "learning_rate": 8.033576054480199e-05, "loss": 0.0346, "step": 26230 }, { "grad_norm": 0.3681778609752655, "learning_rate": 8.031932826564688e-05, "loss": 0.0314, "step": 26240 }, { "grad_norm": 0.3851892948150635, "learning_rate": 8.030289080574782e-05, "loss": 0.0394, "step": 26250 }, { "grad_norm": 0.3246735632419586, "learning_rate": 8.028644816791349e-05, "loss": 0.0373, "step": 26260 }, { "grad_norm": 0.2874853014945984, "learning_rate": 8.027000035495351e-05, "loss": 0.0315, "step": 26270 }, { "grad_norm": 0.32086846232414246, "learning_rate": 8.025354736967836e-05, "loss": 0.0509, "step": 26280 }, { "grad_norm": 0.30210602283477783, "learning_rate": 8.023708921489941e-05, "loss": 0.0442, "step": 26290 }, { "grad_norm": 0.4319862127304077, "learning_rate": 8.022062589342887e-05, "loss": 0.0451, "step": 26300 }, { "grad_norm": 0.41925421357154846, "learning_rate": 8.020415740807993e-05, "loss": 0.0556, "step": 26310 }, { "grad_norm": 0.42554354667663574, "learning_rate": 8.018768376166656e-05, "loss": 0.0294, "step": 26320 }, { "grad_norm": 0.3663954734802246, "learning_rate": 8.017120495700368e-05, "loss": 0.0369, "step": 26330 }, { "grad_norm": 0.37426698207855225, "learning_rate": 8.015472099690704e-05, "loss": 0.0354, "step": 26340 }, { "grad_norm": 0.41321808099746704, "learning_rate": 8.013823188419332e-05, "loss": 0.0315, "step": 26350 }, { "grad_norm": 0.4142768085002899, "learning_rate": 8.012173762168006e-05, "loss": 0.0385, "step": 26360 }, { "grad_norm": 0.23811385035514832, "learning_rate": 8.010523821218567e-05, "loss": 0.0324, "step": 26370 }, { "grad_norm": 0.37286725640296936, "learning_rate": 8.008873365852945e-05, "loss": 0.031, "step": 26380 }, { "grad_norm": 0.43289968371391296, "learning_rate": 8.007222396353157e-05, "loss": 0.0359, "step": 26390 }, { "grad_norm": 0.42114901542663574, "learning_rate": 8.00557091300131e-05, "loss": 0.0382, "step": 26400 }, { "grad_norm": 0.2754996120929718, "learning_rate": 8.003918916079597e-05, "loss": 0.0353, "step": 26410 }, { "grad_norm": 0.3544955849647522, "learning_rate": 8.002266405870298e-05, "loss": 0.0349, "step": 26420 }, { "grad_norm": 0.39434853196144104, "learning_rate": 8.000613382655782e-05, "loss": 0.0406, "step": 26430 }, { "grad_norm": 0.43428584933280945, "learning_rate": 7.998959846718505e-05, "loss": 0.0393, "step": 26440 }, { "grad_norm": 0.3714704215526581, "learning_rate": 7.997305798341012e-05, "loss": 0.0398, "step": 26450 }, { "grad_norm": 0.37669217586517334, "learning_rate": 7.995651237805937e-05, "loss": 0.0424, "step": 26460 }, { "grad_norm": 0.342828631401062, "learning_rate": 7.993996165395996e-05, "loss": 0.0317, "step": 26470 }, { "grad_norm": 0.4345220923423767, "learning_rate": 7.992340581393996e-05, "loss": 0.0327, "step": 26480 }, { "grad_norm": 0.32164937257766724, "learning_rate": 7.990684486082831e-05, "loss": 0.0346, "step": 26490 }, { "grad_norm": 0.404118150472641, "learning_rate": 7.989027879745482e-05, "loss": 0.0404, "step": 26500 }, { "grad_norm": 0.35268598794937134, "learning_rate": 7.98737076266502e-05, "loss": 0.0488, "step": 26510 }, { "grad_norm": 0.29377108812332153, "learning_rate": 7.985713135124598e-05, "loss": 0.0393, "step": 26520 }, { "grad_norm": 0.3578587472438812, "learning_rate": 7.98405499740746e-05, "loss": 0.0428, "step": 26530 }, { "grad_norm": 0.4542444944381714, "learning_rate": 7.98239634979694e-05, "loss": 0.0372, "step": 26540 }, { "grad_norm": 0.5136612057685852, "learning_rate": 7.98073719257645e-05, "loss": 0.0366, "step": 26550 }, { "grad_norm": 0.37096935510635376, "learning_rate": 7.979077526029499e-05, "loss": 0.0374, "step": 26560 }, { "grad_norm": 0.30696749687194824, "learning_rate": 7.977417350439675e-05, "loss": 0.0294, "step": 26570 }, { "grad_norm": 0.33801478147506714, "learning_rate": 7.97575666609066e-05, "loss": 0.0335, "step": 26580 }, { "grad_norm": 0.36334341764450073, "learning_rate": 7.974095473266216e-05, "loss": 0.037, "step": 26590 }, { "grad_norm": 0.36929479241371155, "learning_rate": 7.972433772250198e-05, "loss": 0.0317, "step": 26600 }, { "grad_norm": 0.3251532316207886, "learning_rate": 7.970771563326544e-05, "loss": 0.0296, "step": 26610 }, { "grad_norm": 0.3098278045654297, "learning_rate": 7.96910884677928e-05, "loss": 0.0354, "step": 26620 }, { "grad_norm": 0.37688764929771423, "learning_rate": 7.967445622892523e-05, "loss": 0.0348, "step": 26630 }, { "grad_norm": 0.46456077694892883, "learning_rate": 7.965781891950465e-05, "loss": 0.036, "step": 26640 }, { "grad_norm": 0.3023759722709656, "learning_rate": 7.964117654237397e-05, "loss": 0.026, "step": 26650 }, { "grad_norm": 0.37512853741645813, "learning_rate": 7.962452910037692e-05, "loss": 0.0399, "step": 26660 }, { "grad_norm": 0.39518019556999207, "learning_rate": 7.96078765963581e-05, "loss": 0.0343, "step": 26670 }, { "grad_norm": 0.3340742290019989, "learning_rate": 7.95912190331629e-05, "loss": 0.0356, "step": 26680 }, { "grad_norm": 0.3496148884296417, "learning_rate": 7.957455641363772e-05, "loss": 0.0356, "step": 26690 }, { "grad_norm": 0.5742669105529785, "learning_rate": 7.955788874062968e-05, "loss": 0.0473, "step": 26700 }, { "grad_norm": 0.4894200265407562, "learning_rate": 7.95412160169869e-05, "loss": 0.0411, "step": 26710 }, { "grad_norm": 0.3889892101287842, "learning_rate": 7.952453824555824e-05, "loss": 0.0328, "step": 26720 }, { "grad_norm": 0.32553163170814514, "learning_rate": 7.95078554291935e-05, "loss": 0.0378, "step": 26730 }, { "grad_norm": 0.41073694825172424, "learning_rate": 7.94911675707433e-05, "loss": 0.0305, "step": 26740 }, { "grad_norm": 0.3915424942970276, "learning_rate": 7.947447467305915e-05, "loss": 0.0424, "step": 26750 }, { "grad_norm": 0.2711355686187744, "learning_rate": 7.94577767389934e-05, "loss": 0.035, "step": 26760 }, { "grad_norm": 0.3518739342689514, "learning_rate": 7.944107377139928e-05, "loss": 0.0311, "step": 26770 }, { "grad_norm": 0.34472307562828064, "learning_rate": 7.942436577313088e-05, "loss": 0.0338, "step": 26780 }, { "grad_norm": 0.3660496473312378, "learning_rate": 7.940765274704312e-05, "loss": 0.0363, "step": 26790 }, { "grad_norm": 0.24865996837615967, "learning_rate": 7.939093469599181e-05, "loss": 0.0314, "step": 26800 }, { "grad_norm": 0.33950090408325195, "learning_rate": 7.93742116228336e-05, "loss": 0.0344, "step": 26810 }, { "grad_norm": 0.37387099862098694, "learning_rate": 7.935748353042602e-05, "loss": 0.027, "step": 26820 }, { "grad_norm": 0.3343324363231659, "learning_rate": 7.934075042162744e-05, "loss": 0.0317, "step": 26830 }, { "grad_norm": 0.320907860994339, "learning_rate": 7.932401229929705e-05, "loss": 0.0392, "step": 26840 }, { "grad_norm": 0.37097933888435364, "learning_rate": 7.9307269166295e-05, "loss": 0.0366, "step": 26850 }, { "grad_norm": 0.24075092375278473, "learning_rate": 7.92905210254822e-05, "loss": 0.0319, "step": 26860 }, { "grad_norm": 0.4558104872703552, "learning_rate": 7.927376787972045e-05, "loss": 0.0296, "step": 26870 }, { "grad_norm": 0.40787336230278015, "learning_rate": 7.92570097318724e-05, "loss": 0.0403, "step": 26880 }, { "grad_norm": 0.25000298023223877, "learning_rate": 7.924024658480158e-05, "loss": 0.0311, "step": 26890 }, { "grad_norm": 0.4641658067703247, "learning_rate": 7.922347844137233e-05, "loss": 0.0296, "step": 26900 }, { "grad_norm": 0.295892596244812, "learning_rate": 7.92067053044499e-05, "loss": 0.0358, "step": 26910 }, { "grad_norm": 0.39900559186935425, "learning_rate": 7.918992717690031e-05, "loss": 0.0376, "step": 26920 }, { "grad_norm": 0.40183281898498535, "learning_rate": 7.917314406159053e-05, "loss": 0.0355, "step": 26930 }, { "grad_norm": 0.4928407073020935, "learning_rate": 7.915635596138832e-05, "loss": 0.0473, "step": 26940 }, { "grad_norm": 0.4176950752735138, "learning_rate": 7.913956287916228e-05, "loss": 0.043, "step": 26950 }, { "grad_norm": 0.2909531593322754, "learning_rate": 7.912276481778193e-05, "loss": 0.031, "step": 26960 }, { "grad_norm": 0.31027451157569885, "learning_rate": 7.910596178011759e-05, "loss": 0.0401, "step": 26970 }, { "grad_norm": 0.273788720369339, "learning_rate": 7.908915376904043e-05, "loss": 0.0281, "step": 26980 }, { "grad_norm": 0.470488965511322, "learning_rate": 7.907234078742247e-05, "loss": 0.0374, "step": 26990 }, { "grad_norm": 0.334475576877594, "learning_rate": 7.90555228381366e-05, "loss": 0.0281, "step": 27000 }, { "grad_norm": 0.2731702923774719, "learning_rate": 7.903869992405656e-05, "loss": 0.0297, "step": 27010 }, { "grad_norm": 0.30989760160446167, "learning_rate": 7.902187204805691e-05, "loss": 0.0343, "step": 27020 }, { "grad_norm": 0.4759176969528198, "learning_rate": 7.900503921301308e-05, "loss": 0.032, "step": 27030 }, { "grad_norm": 0.3164946436882019, "learning_rate": 7.898820142180133e-05, "loss": 0.0405, "step": 27040 }, { "grad_norm": 0.3444637358188629, "learning_rate": 7.897135867729879e-05, "loss": 0.0352, "step": 27050 }, { "grad_norm": 0.3636505603790283, "learning_rate": 7.89545109823834e-05, "loss": 0.0302, "step": 27060 }, { "grad_norm": 0.2797556519508362, "learning_rate": 7.8937658339934e-05, "loss": 0.0314, "step": 27070 }, { "grad_norm": 0.3740761876106262, "learning_rate": 7.892080075283026e-05, "loss": 0.0345, "step": 27080 }, { "grad_norm": 0.3841857314109802, "learning_rate": 7.890393822395263e-05, "loss": 0.034, "step": 27090 }, { "grad_norm": 0.30566778779029846, "learning_rate": 7.88870707561825e-05, "loss": 0.0329, "step": 27100 }, { "grad_norm": 0.3941406309604645, "learning_rate": 7.887019835240203e-05, "loss": 0.0352, "step": 27110 }, { "grad_norm": 0.4022711217403412, "learning_rate": 7.885332101549427e-05, "loss": 0.0395, "step": 27120 }, { "grad_norm": 0.4008924067020416, "learning_rate": 7.883643874834308e-05, "loss": 0.0304, "step": 27130 }, { "grad_norm": 0.3614582121372223, "learning_rate": 7.881955155383321e-05, "loss": 0.0438, "step": 27140 }, { "grad_norm": 0.39218419790267944, "learning_rate": 7.880265943485017e-05, "loss": 0.0339, "step": 27150 }, { "grad_norm": 0.391654372215271, "learning_rate": 7.878576239428038e-05, "loss": 0.0339, "step": 27160 }, { "grad_norm": 0.3658292591571808, "learning_rate": 7.87688604350111e-05, "loss": 0.0375, "step": 27170 }, { "grad_norm": 0.40951645374298096, "learning_rate": 7.875195355993042e-05, "loss": 0.0403, "step": 27180 }, { "grad_norm": 0.3709970712661743, "learning_rate": 7.873504177192724e-05, "loss": 0.0405, "step": 27190 }, { "grad_norm": 0.44801610708236694, "learning_rate": 7.87181250738913e-05, "loss": 0.0521, "step": 27200 }, { "grad_norm": 0.38546934723854065, "learning_rate": 7.870120346871324e-05, "loss": 0.0343, "step": 27210 }, { "grad_norm": 0.3684806823730469, "learning_rate": 7.86842769592845e-05, "loss": 0.0327, "step": 27220 }, { "grad_norm": 0.346295028924942, "learning_rate": 7.866734554849732e-05, "loss": 0.0303, "step": 27230 }, { "grad_norm": 0.26429614424705505, "learning_rate": 7.865040923924486e-05, "loss": 0.0319, "step": 27240 }, { "grad_norm": 0.35391944646835327, "learning_rate": 7.863346803442104e-05, "loss": 0.0459, "step": 27250 }, { "grad_norm": 0.4162650406360626, "learning_rate": 7.861652193692067e-05, "loss": 0.0328, "step": 27260 }, { "grad_norm": 0.403413325548172, "learning_rate": 7.859957094963937e-05, "loss": 0.0308, "step": 27270 }, { "grad_norm": 0.4106557369232178, "learning_rate": 7.858261507547357e-05, "loss": 0.0364, "step": 27280 }, { "grad_norm": 0.3603224754333496, "learning_rate": 7.856565431732061e-05, "loss": 0.0339, "step": 27290 }, { "grad_norm": 0.3972238600254059, "learning_rate": 7.854868867807859e-05, "loss": 0.0291, "step": 27300 }, { "grad_norm": 0.3915458917617798, "learning_rate": 7.85317181606465e-05, "loss": 0.0378, "step": 27310 }, { "grad_norm": 0.3697916865348816, "learning_rate": 7.85147427679241e-05, "loss": 0.0357, "step": 27320 }, { "grad_norm": 0.27382543683052063, "learning_rate": 7.849776250281205e-05, "loss": 0.0279, "step": 27330 }, { "grad_norm": 0.23472075164318085, "learning_rate": 7.84807773682118e-05, "loss": 0.032, "step": 27340 }, { "grad_norm": 0.2745887339115143, "learning_rate": 7.846378736702565e-05, "loss": 0.0345, "step": 27350 }, { "grad_norm": 0.22632138431072235, "learning_rate": 7.844679250215671e-05, "loss": 0.041, "step": 27360 }, { "grad_norm": 0.24411825835704803, "learning_rate": 7.842979277650898e-05, "loss": 0.0286, "step": 27370 }, { "grad_norm": 0.3467039465904236, "learning_rate": 7.84127881929872e-05, "loss": 0.0349, "step": 27380 }, { "grad_norm": 0.3410123884677887, "learning_rate": 7.839577875449704e-05, "loss": 0.0333, "step": 27390 }, { "grad_norm": 0.41221335530281067, "learning_rate": 7.837876446394489e-05, "loss": 0.0303, "step": 27400 }, { "grad_norm": 0.31064727902412415, "learning_rate": 7.836174532423805e-05, "loss": 0.042, "step": 27410 }, { "grad_norm": 0.45578649640083313, "learning_rate": 7.834472133828466e-05, "loss": 0.0354, "step": 27420 }, { "grad_norm": 0.43843379616737366, "learning_rate": 7.832769250899359e-05, "loss": 0.0429, "step": 27430 }, { "grad_norm": 0.4786359965801239, "learning_rate": 7.831065883927464e-05, "loss": 0.0475, "step": 27440 }, { "grad_norm": 0.36015650629997253, "learning_rate": 7.829362033203841e-05, "loss": 0.0369, "step": 27450 }, { "grad_norm": 0.3214558959007263, "learning_rate": 7.827657699019628e-05, "loss": 0.0347, "step": 27460 }, { "grad_norm": 0.4103115200996399, "learning_rate": 7.825952881666052e-05, "loss": 0.0527, "step": 27470 }, { "grad_norm": 0.3625074028968811, "learning_rate": 7.824247581434418e-05, "loss": 0.033, "step": 27480 }, { "grad_norm": 0.3174270689487457, "learning_rate": 7.822541798616116e-05, "loss": 0.0378, "step": 27490 }, { "grad_norm": 0.34872156381607056, "learning_rate": 7.820835533502617e-05, "loss": 0.0433, "step": 27500 }, { "grad_norm": 0.4182727634906769, "learning_rate": 7.819128786385475e-05, "loss": 0.0388, "step": 27510 }, { "grad_norm": 0.32628345489501953, "learning_rate": 7.817421557556329e-05, "loss": 0.0391, "step": 27520 }, { "grad_norm": 0.38081076741218567, "learning_rate": 7.815713847306893e-05, "loss": 0.039, "step": 27530 }, { "grad_norm": 0.4362550675868988, "learning_rate": 7.81400565592897e-05, "loss": 0.0333, "step": 27540 }, { "grad_norm": 0.3294679820537567, "learning_rate": 7.812296983714444e-05, "loss": 0.0425, "step": 27550 }, { "grad_norm": 0.27882206439971924, "learning_rate": 7.810587830955281e-05, "loss": 0.0281, "step": 27560 }, { "grad_norm": 0.3783169090747833, "learning_rate": 7.808878197943528e-05, "loss": 0.0386, "step": 27570 }, { "grad_norm": 0.28871285915374756, "learning_rate": 7.807168084971312e-05, "loss": 0.0338, "step": 27580 }, { "grad_norm": 0.25289520621299744, "learning_rate": 7.805457492330849e-05, "loss": 0.0291, "step": 27590 }, { "grad_norm": 0.2858397662639618, "learning_rate": 7.803746420314428e-05, "loss": 0.0325, "step": 27600 }, { "grad_norm": 0.33809465169906616, "learning_rate": 7.802034869214428e-05, "loss": 0.0425, "step": 27610 }, { "grad_norm": 0.5180492997169495, "learning_rate": 7.800322839323303e-05, "loss": 0.0321, "step": 27620 }, { "grad_norm": 0.19653497636318207, "learning_rate": 7.798610330933593e-05, "loss": 0.0258, "step": 27630 }, { "grad_norm": 0.2550888955593109, "learning_rate": 7.796897344337922e-05, "loss": 0.033, "step": 27640 }, { "grad_norm": 0.3227608799934387, "learning_rate": 7.795183879828989e-05, "loss": 0.0412, "step": 27650 }, { "grad_norm": 0.3608383238315582, "learning_rate": 7.793469937699579e-05, "loss": 0.043, "step": 27660 }, { "grad_norm": 0.2714347839355469, "learning_rate": 7.791755518242558e-05, "loss": 0.0417, "step": 27670 }, { "grad_norm": 0.4581933915615082, "learning_rate": 7.790040621750876e-05, "loss": 0.0344, "step": 27680 }, { "grad_norm": 0.31366831064224243, "learning_rate": 7.788325248517558e-05, "loss": 0.0285, "step": 27690 }, { "grad_norm": 0.5491034984588623, "learning_rate": 7.786609398835715e-05, "loss": 0.0357, "step": 27700 }, { "grad_norm": 0.3137940764427185, "learning_rate": 7.784893072998541e-05, "loss": 0.0249, "step": 27710 }, { "grad_norm": 0.29296448826789856, "learning_rate": 7.783176271299306e-05, "loss": 0.0251, "step": 27720 }, { "grad_norm": 0.3740236759185791, "learning_rate": 7.781458994031368e-05, "loss": 0.0293, "step": 27730 }, { "grad_norm": 0.28131502866744995, "learning_rate": 7.779741241488161e-05, "loss": 0.0307, "step": 27740 }, { "grad_norm": 0.34045466780662537, "learning_rate": 7.7780230139632e-05, "loss": 0.0307, "step": 27750 }, { "grad_norm": 0.30445191264152527, "learning_rate": 7.776304311750087e-05, "loss": 0.0261, "step": 27760 }, { "grad_norm": 0.4608837962150574, "learning_rate": 7.7745851351425e-05, "loss": 0.0402, "step": 27770 }, { "grad_norm": 0.3589542806148529, "learning_rate": 7.772865484434197e-05, "loss": 0.0285, "step": 27780 }, { "grad_norm": 0.332383930683136, "learning_rate": 7.77114535991902e-05, "loss": 0.0388, "step": 27790 }, { "grad_norm": 0.3171994984149933, "learning_rate": 7.769424761890893e-05, "loss": 0.0318, "step": 27800 }, { "grad_norm": 0.4992240071296692, "learning_rate": 7.767703690643817e-05, "loss": 0.0402, "step": 27810 }, { "grad_norm": 0.24242807924747467, "learning_rate": 7.76598214647188e-05, "loss": 0.0275, "step": 27820 }, { "grad_norm": 0.272283136844635, "learning_rate": 7.764260129669241e-05, "loss": 0.0271, "step": 27830 }, { "grad_norm": 0.3485232889652252, "learning_rate": 7.76253764053015e-05, "loss": 0.0409, "step": 27840 }, { "grad_norm": 0.24079284071922302, "learning_rate": 7.760814679348932e-05, "loss": 0.0296, "step": 27850 }, { "grad_norm": 0.3503090441226959, "learning_rate": 7.759091246419992e-05, "loss": 0.0322, "step": 27860 }, { "grad_norm": 0.3077537417411804, "learning_rate": 7.75736734203782e-05, "loss": 0.0271, "step": 27870 }, { "grad_norm": 0.35417640209198, "learning_rate": 7.755642966496985e-05, "loss": 0.0502, "step": 27880 }, { "grad_norm": 0.3068143129348755, "learning_rate": 7.753918120092132e-05, "loss": 0.0324, "step": 27890 }, { "grad_norm": 0.2517128586769104, "learning_rate": 7.752192803117993e-05, "loss": 0.0311, "step": 27900 }, { "grad_norm": 0.2818782925605774, "learning_rate": 7.750467015869377e-05, "loss": 0.0287, "step": 27910 }, { "grad_norm": 0.34418484568595886, "learning_rate": 7.748740758641174e-05, "loss": 0.0379, "step": 27920 }, { "grad_norm": 0.3455619215965271, "learning_rate": 7.74701403172835e-05, "loss": 0.039, "step": 27930 }, { "grad_norm": 0.26610055565834045, "learning_rate": 7.745286835425962e-05, "loss": 0.0379, "step": 27940 }, { "grad_norm": 0.4986259937286377, "learning_rate": 7.743559170029138e-05, "loss": 0.0406, "step": 27950 }, { "grad_norm": 0.25906068086624146, "learning_rate": 7.741831035833087e-05, "loss": 0.0454, "step": 27960 }, { "grad_norm": 0.35983341932296753, "learning_rate": 7.740102433133102e-05, "loss": 0.0362, "step": 27970 }, { "grad_norm": 0.2734425365924835, "learning_rate": 7.738373362224553e-05, "loss": 0.0302, "step": 27980 }, { "grad_norm": 0.46552079916000366, "learning_rate": 7.73664382340289e-05, "loss": 0.0361, "step": 27990 }, { "grad_norm": 0.333962082862854, "learning_rate": 7.734913816963647e-05, "loss": 0.0427, "step": 28000 }, { "grad_norm": 0.3621683418750763, "learning_rate": 7.73318334320243e-05, "loss": 0.0314, "step": 28010 }, { "grad_norm": 0.3527536690235138, "learning_rate": 7.731452402414934e-05, "loss": 0.0386, "step": 28020 }, { "grad_norm": 0.37970250844955444, "learning_rate": 7.729720994896928e-05, "loss": 0.0322, "step": 28030 }, { "grad_norm": 0.27625858783721924, "learning_rate": 7.727989120944262e-05, "loss": 0.0289, "step": 28040 }, { "grad_norm": 0.2141437530517578, "learning_rate": 7.726256780852865e-05, "loss": 0.0261, "step": 28050 }, { "grad_norm": 0.3304905891418457, "learning_rate": 7.724523974918749e-05, "loss": 0.0296, "step": 28060 }, { "grad_norm": 0.324662983417511, "learning_rate": 7.722790703438002e-05, "loss": 0.0265, "step": 28070 }, { "grad_norm": 0.35457858443260193, "learning_rate": 7.72105696670679e-05, "loss": 0.0258, "step": 28080 }, { "grad_norm": 0.4894339144229889, "learning_rate": 7.719322765021364e-05, "loss": 0.0375, "step": 28090 }, { "grad_norm": 0.5662849545478821, "learning_rate": 7.717588098678051e-05, "loss": 0.0325, "step": 28100 }, { "grad_norm": 0.3982728123664856, "learning_rate": 7.715852967973258e-05, "loss": 0.0333, "step": 28110 }, { "grad_norm": 0.4194806218147278, "learning_rate": 7.714117373203474e-05, "loss": 0.0359, "step": 28120 }, { "grad_norm": 0.3282265365123749, "learning_rate": 7.712381314665259e-05, "loss": 0.0316, "step": 28130 }, { "grad_norm": 0.2895919382572174, "learning_rate": 7.710644792655261e-05, "loss": 0.034, "step": 28140 }, { "grad_norm": 0.4469287097454071, "learning_rate": 7.708907807470207e-05, "loss": 0.0475, "step": 28150 }, { "grad_norm": 0.26559293270111084, "learning_rate": 7.707170359406896e-05, "loss": 0.0368, "step": 28160 }, { "grad_norm": 0.3235374093055725, "learning_rate": 7.705432448762213e-05, "loss": 0.0289, "step": 28170 }, { "grad_norm": 0.33609265089035034, "learning_rate": 7.703694075833117e-05, "loss": 0.0464, "step": 28180 }, { "grad_norm": 0.4137428402900696, "learning_rate": 7.70195524091665e-05, "loss": 0.0579, "step": 28190 }, { "grad_norm": 0.4413910210132599, "learning_rate": 7.70021594430993e-05, "loss": 0.0415, "step": 28200 }, { "grad_norm": 0.31561967730522156, "learning_rate": 7.698476186310157e-05, "loss": 0.0344, "step": 28210 }, { "grad_norm": 0.43413859605789185, "learning_rate": 7.696735967214608e-05, "loss": 0.0304, "step": 28220 }, { "grad_norm": 0.3705666959285736, "learning_rate": 7.694995287320636e-05, "loss": 0.0412, "step": 28230 }, { "grad_norm": 0.34413763880729675, "learning_rate": 7.693254146925679e-05, "loss": 0.0367, "step": 28240 }, { "grad_norm": 0.36643579602241516, "learning_rate": 7.691512546327251e-05, "loss": 0.0435, "step": 28250 }, { "grad_norm": 0.4436851739883423, "learning_rate": 7.689770485822939e-05, "loss": 0.0381, "step": 28260 }, { "grad_norm": 0.3076631426811218, "learning_rate": 7.688027965710416e-05, "loss": 0.0294, "step": 28270 }, { "grad_norm": 0.3668132424354553, "learning_rate": 7.686284986287433e-05, "loss": 0.0443, "step": 28280 }, { "grad_norm": 0.4234139621257782, "learning_rate": 7.684541547851817e-05, "loss": 0.0336, "step": 28290 }, { "grad_norm": 0.2390568107366562, "learning_rate": 7.68279765070147e-05, "loss": 0.0326, "step": 28300 }, { "grad_norm": 0.3530780076980591, "learning_rate": 7.68105329513438e-05, "loss": 0.0317, "step": 28310 }, { "grad_norm": 0.35991421341896057, "learning_rate": 7.67930848144861e-05, "loss": 0.0337, "step": 28320 }, { "grad_norm": 0.3467405140399933, "learning_rate": 7.6775632099423e-05, "loss": 0.0386, "step": 28330 }, { "grad_norm": 0.34789615869522095, "learning_rate": 7.675817480913667e-05, "loss": 0.0363, "step": 28340 }, { "grad_norm": 0.5023091435432434, "learning_rate": 7.674071294661011e-05, "loss": 0.0309, "step": 28350 }, { "grad_norm": 0.24054928123950958, "learning_rate": 7.672324651482707e-05, "loss": 0.0265, "step": 28360 }, { "grad_norm": 0.5340259075164795, "learning_rate": 7.670577551677209e-05, "loss": 0.0425, "step": 28370 }, { "grad_norm": 0.3241019546985626, "learning_rate": 7.668829995543047e-05, "loss": 0.0319, "step": 28380 }, { "grad_norm": 0.3748045265674591, "learning_rate": 7.667081983378832e-05, "loss": 0.0401, "step": 28390 }, { "grad_norm": 0.4802176356315613, "learning_rate": 7.66533351548325e-05, "loss": 0.0309, "step": 28400 }, { "grad_norm": 0.5524996519088745, "learning_rate": 7.663584592155069e-05, "loss": 0.0342, "step": 28410 }, { "grad_norm": 0.40176165103912354, "learning_rate": 7.661835213693129e-05, "loss": 0.0424, "step": 28420 }, { "grad_norm": 0.412074476480484, "learning_rate": 7.660085380396353e-05, "loss": 0.0337, "step": 28430 }, { "grad_norm": 0.273101270198822, "learning_rate": 7.658335092563738e-05, "loss": 0.0337, "step": 28440 }, { "grad_norm": 0.37089723348617554, "learning_rate": 7.656584350494362e-05, "loss": 0.0324, "step": 28450 }, { "grad_norm": 0.4992944300174713, "learning_rate": 7.654833154487378e-05, "loss": 0.0376, "step": 28460 }, { "grad_norm": 0.34736964106559753, "learning_rate": 7.653081504842017e-05, "loss": 0.032, "step": 28470 }, { "grad_norm": 0.31093692779541016, "learning_rate": 7.65132940185759e-05, "loss": 0.0398, "step": 28480 }, { "grad_norm": 0.43093910813331604, "learning_rate": 7.649576845833481e-05, "loss": 0.033, "step": 28490 }, { "grad_norm": 0.40543434023857117, "learning_rate": 7.647823837069156e-05, "loss": 0.0336, "step": 28500 }, { "grad_norm": 0.31557050347328186, "learning_rate": 7.646070375864156e-05, "loss": 0.032, "step": 28510 }, { "grad_norm": 0.2508668303489685, "learning_rate": 7.644316462518099e-05, "loss": 0.0387, "step": 28520 }, { "grad_norm": 0.4832836985588074, "learning_rate": 7.642562097330679e-05, "loss": 0.0398, "step": 28530 }, { "grad_norm": 0.3358461558818817, "learning_rate": 7.640807280601671e-05, "loss": 0.0298, "step": 28540 }, { "grad_norm": 0.40052178502082825, "learning_rate": 7.639052012630927e-05, "loss": 0.0391, "step": 28550 }, { "grad_norm": 0.387079119682312, "learning_rate": 7.63729629371837e-05, "loss": 0.0411, "step": 28560 }, { "grad_norm": 0.38626953959465027, "learning_rate": 7.635540124164009e-05, "loss": 0.0464, "step": 28570 }, { "grad_norm": 0.3676629066467285, "learning_rate": 7.633783504267922e-05, "loss": 0.0378, "step": 28580 }, { "grad_norm": 0.32372748851776123, "learning_rate": 7.632026434330269e-05, "loss": 0.0373, "step": 28590 }, { "grad_norm": 0.2786182761192322, "learning_rate": 7.630268914651282e-05, "loss": 0.0285, "step": 28600 }, { "grad_norm": 0.3749789595603943, "learning_rate": 7.628510945531278e-05, "loss": 0.0509, "step": 28610 }, { "grad_norm": 0.36116909980773926, "learning_rate": 7.626752527270641e-05, "loss": 0.0382, "step": 28620 }, { "grad_norm": 0.2727866470813751, "learning_rate": 7.62499366016984e-05, "loss": 0.0319, "step": 28630 }, { "grad_norm": 0.29334554076194763, "learning_rate": 7.623234344529416e-05, "loss": 0.0342, "step": 28640 }, { "grad_norm": 0.37576913833618164, "learning_rate": 7.62147458064999e-05, "loss": 0.0393, "step": 28650 }, { "grad_norm": 0.4612075686454773, "learning_rate": 7.619714368832254e-05, "loss": 0.0409, "step": 28660 }, { "grad_norm": 0.3375016152858734, "learning_rate": 7.61795370937698e-05, "loss": 0.0353, "step": 28670 }, { "grad_norm": 0.2650165557861328, "learning_rate": 7.61619260258502e-05, "loss": 0.0376, "step": 28680 }, { "grad_norm": 0.3162626326084137, "learning_rate": 7.614431048757298e-05, "loss": 0.0318, "step": 28690 }, { "grad_norm": 0.3447328209877014, "learning_rate": 7.612669048194814e-05, "loss": 0.0326, "step": 28700 }, { "grad_norm": 0.34412840008735657, "learning_rate": 7.610906601198646e-05, "loss": 0.0342, "step": 28710 }, { "grad_norm": 0.27688390016555786, "learning_rate": 7.60914370806995e-05, "loss": 0.0389, "step": 28720 }, { "grad_norm": 0.30574148893356323, "learning_rate": 7.607380369109953e-05, "loss": 0.0524, "step": 28730 }, { "grad_norm": 0.24846796691417694, "learning_rate": 7.605616584619961e-05, "loss": 0.0304, "step": 28740 }, { "grad_norm": 0.4109937250614166, "learning_rate": 7.603852354901362e-05, "loss": 0.0293, "step": 28750 }, { "grad_norm": 0.5274378657341003, "learning_rate": 7.602087680255609e-05, "loss": 0.0481, "step": 28760 }, { "grad_norm": 0.26941734552383423, "learning_rate": 7.600322560984238e-05, "loss": 0.0308, "step": 28770 }, { "grad_norm": 0.3552314341068268, "learning_rate": 7.598556997388863e-05, "loss": 0.0331, "step": 28780 }, { "grad_norm": 0.28768590092658997, "learning_rate": 7.596790989771166e-05, "loss": 0.0384, "step": 28790 }, { "grad_norm": 0.34772735834121704, "learning_rate": 7.595024538432914e-05, "loss": 0.0368, "step": 28800 }, { "grad_norm": 0.25874063372612, "learning_rate": 7.59325764367594e-05, "loss": 0.0285, "step": 28810 }, { "grad_norm": 0.5053936243057251, "learning_rate": 7.59149030580216e-05, "loss": 0.0355, "step": 28820 }, { "grad_norm": 0.3322874903678894, "learning_rate": 7.589722525113562e-05, "loss": 0.031, "step": 28830 }, { "grad_norm": 0.4138982594013214, "learning_rate": 7.587954301912216e-05, "loss": 0.03, "step": 28840 }, { "grad_norm": 0.493630051612854, "learning_rate": 7.586185636500263e-05, "loss": 0.0428, "step": 28850 }, { "grad_norm": 0.30932000279426575, "learning_rate": 7.584416529179914e-05, "loss": 0.0363, "step": 28860 }, { "grad_norm": 0.3199576735496521, "learning_rate": 7.582646980253465e-05, "loss": 0.0346, "step": 28870 }, { "grad_norm": 0.34793588519096375, "learning_rate": 7.580876990023282e-05, "loss": 0.0492, "step": 28880 }, { "grad_norm": 0.3768908977508545, "learning_rate": 7.579106558791809e-05, "loss": 0.0314, "step": 28890 }, { "grad_norm": 0.2711675763130188, "learning_rate": 7.577335686861565e-05, "loss": 0.0318, "step": 28900 }, { "grad_norm": 0.35693004727363586, "learning_rate": 7.575564374535141e-05, "loss": 0.0377, "step": 28910 }, { "grad_norm": 0.3723639249801636, "learning_rate": 7.573792622115207e-05, "loss": 0.0297, "step": 28920 }, { "grad_norm": 0.4058425724506378, "learning_rate": 7.572020429904507e-05, "loss": 0.0291, "step": 28930 }, { "grad_norm": 0.3894508481025696, "learning_rate": 7.570247798205861e-05, "loss": 0.031, "step": 28940 }, { "grad_norm": 0.27055126428604126, "learning_rate": 7.568474727322164e-05, "loss": 0.0416, "step": 28950 }, { "grad_norm": 0.3559167683124542, "learning_rate": 7.566701217556384e-05, "loss": 0.0322, "step": 28960 }, { "grad_norm": 0.3276084065437317, "learning_rate": 7.564927269211564e-05, "loss": 0.035, "step": 28970 }, { "grad_norm": 0.22893750667572021, "learning_rate": 7.563152882590824e-05, "loss": 0.0304, "step": 28980 }, { "grad_norm": 0.35040923953056335, "learning_rate": 7.56137805799736e-05, "loss": 0.0313, "step": 28990 }, { "grad_norm": 0.4420996606349945, "learning_rate": 7.559602795734439e-05, "loss": 0.0397, "step": 29000 }, { "grad_norm": 0.6146987676620483, "learning_rate": 7.557827096105408e-05, "loss": 0.0335, "step": 29010 }, { "grad_norm": 0.45977136492729187, "learning_rate": 7.55605095941368e-05, "loss": 0.0393, "step": 29020 }, { "grad_norm": 0.19380886852741241, "learning_rate": 7.554274385962752e-05, "loss": 0.0344, "step": 29030 }, { "grad_norm": 0.5033217072486877, "learning_rate": 7.552497376056191e-05, "loss": 0.0439, "step": 29040 }, { "grad_norm": 0.3312060236930847, "learning_rate": 7.550719929997639e-05, "loss": 0.039, "step": 29050 }, { "grad_norm": 0.38827842473983765, "learning_rate": 7.548942048090813e-05, "loss": 0.0389, "step": 29060 }, { "grad_norm": 0.3622697591781616, "learning_rate": 7.547163730639506e-05, "loss": 0.0374, "step": 29070 }, { "grad_norm": 0.2686944305896759, "learning_rate": 7.545384977947583e-05, "loss": 0.0338, "step": 29080 }, { "grad_norm": 0.3094998300075531, "learning_rate": 7.543605790318981e-05, "loss": 0.0343, "step": 29090 }, { "grad_norm": 0.41405072808265686, "learning_rate": 7.54182616805772e-05, "loss": 0.0405, "step": 29100 }, { "grad_norm": 0.33776283264160156, "learning_rate": 7.540046111467885e-05, "loss": 0.0334, "step": 29110 }, { "grad_norm": 0.41228020191192627, "learning_rate": 7.53826562085364e-05, "loss": 0.0402, "step": 29120 }, { "grad_norm": 0.4607851803302765, "learning_rate": 7.536484696519221e-05, "loss": 0.0321, "step": 29130 }, { "grad_norm": 0.41756942868232727, "learning_rate": 7.534703338768942e-05, "loss": 0.0396, "step": 29140 }, { "grad_norm": 0.3453037440776825, "learning_rate": 7.532921547907185e-05, "loss": 0.0387, "step": 29150 }, { "grad_norm": 0.23620784282684326, "learning_rate": 7.531139324238412e-05, "loss": 0.0468, "step": 29160 }, { "grad_norm": 0.5008851885795593, "learning_rate": 7.529356668067157e-05, "loss": 0.0475, "step": 29170 }, { "grad_norm": 0.44552692770957947, "learning_rate": 7.527573579698023e-05, "loss": 0.0448, "step": 29180 }, { "grad_norm": 0.35555529594421387, "learning_rate": 7.525790059435693e-05, "loss": 0.0334, "step": 29190 }, { "grad_norm": 0.3781869411468506, "learning_rate": 7.524006107584926e-05, "loss": 0.0392, "step": 29200 }, { "grad_norm": 0.3939588665962219, "learning_rate": 7.522221724450544e-05, "loss": 0.0362, "step": 29210 }, { "grad_norm": 0.3756147623062134, "learning_rate": 7.520436910337451e-05, "loss": 0.0473, "step": 29220 }, { "grad_norm": 0.3002183139324188, "learning_rate": 7.518651665550627e-05, "loss": 0.0319, "step": 29230 }, { "grad_norm": 0.26243144273757935, "learning_rate": 7.516865990395117e-05, "loss": 0.042, "step": 29240 }, { "grad_norm": 0.311149001121521, "learning_rate": 7.515079885176047e-05, "loss": 0.0307, "step": 29250 }, { "grad_norm": 0.2334628701210022, "learning_rate": 7.513293350198612e-05, "loss": 0.0383, "step": 29260 }, { "grad_norm": 0.2352788895368576, "learning_rate": 7.511506385768081e-05, "loss": 0.0306, "step": 29270 }, { "grad_norm": 0.41397035121917725, "learning_rate": 7.509718992189801e-05, "loss": 0.0392, "step": 29280 }, { "grad_norm": 0.4309038817882538, "learning_rate": 7.507931169769182e-05, "loss": 0.0335, "step": 29290 }, { "grad_norm": 0.47375163435935974, "learning_rate": 7.506142918811722e-05, "loss": 0.0329, "step": 29300 }, { "grad_norm": 0.30449795722961426, "learning_rate": 7.504354239622978e-05, "loss": 0.037, "step": 29310 }, { "grad_norm": 0.3660223186016083, "learning_rate": 7.50256513250859e-05, "loss": 0.0354, "step": 29320 }, { "grad_norm": 0.4837057590484619, "learning_rate": 7.500775597774265e-05, "loss": 0.0336, "step": 29330 }, { "grad_norm": 0.4266793727874756, "learning_rate": 7.498985635725788e-05, "loss": 0.0322, "step": 29340 }, { "grad_norm": 0.43279990553855896, "learning_rate": 7.497195246669012e-05, "loss": 0.0394, "step": 29350 }, { "grad_norm": 0.3631435036659241, "learning_rate": 7.495404430909868e-05, "loss": 0.032, "step": 29360 }, { "grad_norm": 0.4506160318851471, "learning_rate": 7.493613188754356e-05, "loss": 0.0399, "step": 29370 }, { "grad_norm": 0.305023193359375, "learning_rate": 7.49182152050855e-05, "loss": 0.0352, "step": 29380 }, { "grad_norm": 0.3007987141609192, "learning_rate": 7.490029426478598e-05, "loss": 0.0273, "step": 29390 }, { "grad_norm": 0.31043851375579834, "learning_rate": 7.488236906970719e-05, "loss": 0.032, "step": 29400 }, { "grad_norm": 0.4052787125110626, "learning_rate": 7.486443962291207e-05, "loss": 0.0344, "step": 29410 }, { "grad_norm": 0.27084487676620483, "learning_rate": 7.484650592746424e-05, "loss": 0.0272, "step": 29420 }, { "grad_norm": 0.3401007652282715, "learning_rate": 7.482856798642811e-05, "loss": 0.0274, "step": 29430 }, { "grad_norm": 0.30369797348976135, "learning_rate": 7.481062580286878e-05, "loss": 0.0308, "step": 29440 }, { "grad_norm": 0.3885611295700073, "learning_rate": 7.479267937985208e-05, "loss": 0.0306, "step": 29450 }, { "grad_norm": 0.2959013879299164, "learning_rate": 7.477472872044456e-05, "loss": 0.0261, "step": 29460 }, { "grad_norm": 0.3350154757499695, "learning_rate": 7.475677382771347e-05, "loss": 0.0279, "step": 29470 }, { "grad_norm": 0.2935747802257538, "learning_rate": 7.473881470472683e-05, "loss": 0.0405, "step": 29480 }, { "grad_norm": 0.5096431970596313, "learning_rate": 7.472085135455337e-05, "loss": 0.0374, "step": 29490 }, { "grad_norm": 0.5782259106636047, "learning_rate": 7.470288378026256e-05, "loss": 0.0392, "step": 29500 }, { "grad_norm": 0.4395737946033478, "learning_rate": 7.468491198492451e-05, "loss": 0.0316, "step": 29510 }, { "grad_norm": 0.3453959822654724, "learning_rate": 7.466693597161013e-05, "loss": 0.0306, "step": 29520 }, { "grad_norm": 0.3191542327404022, "learning_rate": 7.464895574339104e-05, "loss": 0.0364, "step": 29530 }, { "grad_norm": 0.33620449900627136, "learning_rate": 7.463097130333958e-05, "loss": 0.0435, "step": 29540 }, { "grad_norm": 0.3418984115123749, "learning_rate": 7.461298265452876e-05, "loss": 0.0312, "step": 29550 }, { "grad_norm": 0.27312546968460083, "learning_rate": 7.459498980003239e-05, "loss": 0.0276, "step": 29560 }, { "grad_norm": 0.3974238932132721, "learning_rate": 7.457699274292493e-05, "loss": 0.0332, "step": 29570 }, { "grad_norm": 0.3265877366065979, "learning_rate": 7.455899148628159e-05, "loss": 0.0389, "step": 29580 }, { "grad_norm": 0.33644425868988037, "learning_rate": 7.45409860331783e-05, "loss": 0.0326, "step": 29590 }, { "grad_norm": 0.33252954483032227, "learning_rate": 7.452297638669169e-05, "loss": 0.0268, "step": 29600 }, { "grad_norm": 0.31108635663986206, "learning_rate": 7.450496254989911e-05, "loss": 0.0277, "step": 29610 }, { "grad_norm": 0.3439798951148987, "learning_rate": 7.448694452587866e-05, "loss": 0.0308, "step": 29620 }, { "grad_norm": 0.33589184284210205, "learning_rate": 7.44689223177091e-05, "loss": 0.0391, "step": 29630 }, { "grad_norm": 0.46752116084098816, "learning_rate": 7.445089592846994e-05, "loss": 0.032, "step": 29640 }, { "grad_norm": 0.246892049908638, "learning_rate": 7.443286536124141e-05, "loss": 0.0369, "step": 29650 }, { "grad_norm": 0.23014284670352936, "learning_rate": 7.441483061910443e-05, "loss": 0.0364, "step": 29660 }, { "grad_norm": 0.5576774477958679, "learning_rate": 7.439679170514064e-05, "loss": 0.0418, "step": 29670 }, { "grad_norm": 0.38814371824264526, "learning_rate": 7.43787486224324e-05, "loss": 0.0383, "step": 29680 }, { "grad_norm": 0.3299621641635895, "learning_rate": 7.436070137406276e-05, "loss": 0.0366, "step": 29690 }, { "grad_norm": 0.3847931921482086, "learning_rate": 7.434264996311556e-05, "loss": 0.0338, "step": 29700 }, { "grad_norm": 0.36109277606010437, "learning_rate": 7.432459439267525e-05, "loss": 0.0371, "step": 29710 }, { "grad_norm": 0.3610721826553345, "learning_rate": 7.430653466582701e-05, "loss": 0.0381, "step": 29720 }, { "grad_norm": 0.32221177220344543, "learning_rate": 7.42884707856568e-05, "loss": 0.044, "step": 29730 }, { "grad_norm": 0.31020575761795044, "learning_rate": 7.427040275525122e-05, "loss": 0.0309, "step": 29740 }, { "grad_norm": 0.283563494682312, "learning_rate": 7.42523305776976e-05, "loss": 0.0344, "step": 29750 }, { "grad_norm": 0.325982928276062, "learning_rate": 7.4234254256084e-05, "loss": 0.0287, "step": 29760 }, { "grad_norm": 0.25027209520339966, "learning_rate": 7.421617379349915e-05, "loss": 0.0342, "step": 29770 }, { "grad_norm": 0.45647621154785156, "learning_rate": 7.41980891930325e-05, "loss": 0.0348, "step": 29780 }, { "grad_norm": 0.28400129079818726, "learning_rate": 7.418000045777425e-05, "loss": 0.0271, "step": 29790 }, { "grad_norm": 0.3616170585155487, "learning_rate": 7.416190759081523e-05, "loss": 0.0368, "step": 29800 }, { "grad_norm": 0.3452478349208832, "learning_rate": 7.414381059524704e-05, "loss": 0.0243, "step": 29810 }, { "grad_norm": 0.25450679659843445, "learning_rate": 7.412570947416195e-05, "loss": 0.0237, "step": 29820 }, { "grad_norm": 0.5764988660812378, "learning_rate": 7.410760423065295e-05, "loss": 0.0445, "step": 29830 }, { "grad_norm": 0.3523467183113098, "learning_rate": 7.408949486781372e-05, "loss": 0.0387, "step": 29840 }, { "grad_norm": 0.45996904373168945, "learning_rate": 7.407138138873868e-05, "loss": 0.0379, "step": 29850 }, { "grad_norm": 0.7081389427185059, "learning_rate": 7.405326379652292e-05, "loss": 0.0407, "step": 29860 }, { "grad_norm": 0.40001270174980164, "learning_rate": 7.403514209426222e-05, "loss": 0.0294, "step": 29870 }, { "grad_norm": 0.40340155363082886, "learning_rate": 7.40170162850531e-05, "loss": 0.0354, "step": 29880 }, { "grad_norm": 0.4107131361961365, "learning_rate": 7.399888637199278e-05, "loss": 0.03, "step": 29890 }, { "grad_norm": 0.34652620553970337, "learning_rate": 7.398075235817914e-05, "loss": 0.0515, "step": 29900 }, { "grad_norm": 0.4369105398654938, "learning_rate": 7.39626142467108e-05, "loss": 0.0285, "step": 29910 }, { "grad_norm": 0.3894273638725281, "learning_rate": 7.394447204068706e-05, "loss": 0.0318, "step": 29920 }, { "grad_norm": 0.3474145531654358, "learning_rate": 7.392632574320793e-05, "loss": 0.0405, "step": 29930 }, { "grad_norm": 0.36610475182533264, "learning_rate": 7.390817535737411e-05, "loss": 0.0337, "step": 29940 }, { "grad_norm": 0.3377203047275543, "learning_rate": 7.389002088628703e-05, "loss": 0.028, "step": 29950 }, { "grad_norm": 0.37345951795578003, "learning_rate": 7.387186233304877e-05, "loss": 0.039, "step": 29960 }, { "grad_norm": 0.2863219380378723, "learning_rate": 7.385369970076212e-05, "loss": 0.0293, "step": 29970 }, { "grad_norm": 0.31916946172714233, "learning_rate": 7.38355329925306e-05, "loss": 0.0403, "step": 29980 }, { "grad_norm": 0.27191361784935, "learning_rate": 7.381736221145838e-05, "loss": 0.0333, "step": 29990 }, { "grad_norm": 0.40218469500541687, "learning_rate": 7.37991873606504e-05, "loss": 0.0326, "step": 30000 }, { "grad_norm": 0.3066421151161194, "learning_rate": 7.378100844321218e-05, "loss": 0.0297, "step": 30010 }, { "grad_norm": 0.26529666781425476, "learning_rate": 7.376282546225004e-05, "loss": 0.0256, "step": 30020 }, { "grad_norm": 0.37119001150131226, "learning_rate": 7.374463842087094e-05, "loss": 0.0277, "step": 30030 }, { "grad_norm": 0.39197012782096863, "learning_rate": 7.372644732218254e-05, "loss": 0.0275, "step": 30040 }, { "grad_norm": 0.32198387384414673, "learning_rate": 7.370825216929322e-05, "loss": 0.0437, "step": 30050 }, { "grad_norm": 0.5117150545120239, "learning_rate": 7.369005296531205e-05, "loss": 0.0375, "step": 30060 }, { "grad_norm": 0.2812100946903229, "learning_rate": 7.367184971334873e-05, "loss": 0.03, "step": 30070 }, { "grad_norm": 0.33395543694496155, "learning_rate": 7.365364241651371e-05, "loss": 0.0371, "step": 30080 }, { "grad_norm": 0.4956393837928772, "learning_rate": 7.363543107791815e-05, "loss": 0.0309, "step": 30090 }, { "grad_norm": 0.4124906361103058, "learning_rate": 7.361721570067384e-05, "loss": 0.0311, "step": 30100 }, { "grad_norm": 0.22847172617912292, "learning_rate": 7.359899628789331e-05, "loss": 0.042, "step": 30110 }, { "grad_norm": 0.40619957447052, "learning_rate": 7.358077284268974e-05, "loss": 0.0409, "step": 30120 }, { "grad_norm": 0.2648240327835083, "learning_rate": 7.356254536817702e-05, "loss": 0.0344, "step": 30130 }, { "grad_norm": 0.3601340651512146, "learning_rate": 7.354431386746973e-05, "loss": 0.0352, "step": 30140 }, { "grad_norm": 0.46885964274406433, "learning_rate": 7.352607834368316e-05, "loss": 0.0318, "step": 30150 }, { "grad_norm": 0.22025109827518463, "learning_rate": 7.350783879993324e-05, "loss": 0.0293, "step": 30160 }, { "grad_norm": 0.2793053686618805, "learning_rate": 7.348959523933658e-05, "loss": 0.0357, "step": 30170 }, { "grad_norm": 0.32497045397758484, "learning_rate": 7.347134766501057e-05, "loss": 0.0284, "step": 30180 }, { "grad_norm": 0.4124326705932617, "learning_rate": 7.345309608007315e-05, "loss": 0.0224, "step": 30190 }, { "grad_norm": 0.40135958790779114, "learning_rate": 7.343484048764308e-05, "loss": 0.0304, "step": 30200 }, { "grad_norm": 0.33584779500961304, "learning_rate": 7.341658089083972e-05, "loss": 0.0285, "step": 30210 }, { "grad_norm": 0.29895588755607605, "learning_rate": 7.339831729278313e-05, "loss": 0.0265, "step": 30220 }, { "grad_norm": 0.3391129970550537, "learning_rate": 7.338004969659404e-05, "loss": 0.0297, "step": 30230 }, { "grad_norm": 0.48219403624534607, "learning_rate": 7.336177810539391e-05, "loss": 0.0398, "step": 30240 }, { "grad_norm": 0.38251179456710815, "learning_rate": 7.334350252230485e-05, "loss": 0.0279, "step": 30250 }, { "grad_norm": 0.360059916973114, "learning_rate": 7.332522295044965e-05, "loss": 0.0246, "step": 30260 }, { "grad_norm": 0.383321613073349, "learning_rate": 7.33069393929518e-05, "loss": 0.0391, "step": 30270 }, { "grad_norm": 0.465757817029953, "learning_rate": 7.328865185293545e-05, "loss": 0.0454, "step": 30280 }, { "grad_norm": 0.2972668707370758, "learning_rate": 7.327036033352546e-05, "loss": 0.0331, "step": 30290 }, { "grad_norm": 0.4802784025669098, "learning_rate": 7.325206483784733e-05, "loss": 0.0298, "step": 30300 }, { "grad_norm": 0.46134835481643677, "learning_rate": 7.323376536902724e-05, "loss": 0.0464, "step": 30310 }, { "grad_norm": 0.3600975275039673, "learning_rate": 7.321546193019213e-05, "loss": 0.0398, "step": 30320 }, { "grad_norm": 0.2453923374414444, "learning_rate": 7.31971545244695e-05, "loss": 0.0343, "step": 30330 }, { "grad_norm": 0.2988850474357605, "learning_rate": 7.31788431549876e-05, "loss": 0.0363, "step": 30340 }, { "grad_norm": 0.32467660307884216, "learning_rate": 7.316052782487534e-05, "loss": 0.0333, "step": 30350 }, { "grad_norm": 0.4385451078414917, "learning_rate": 7.314220853726234e-05, "loss": 0.0438, "step": 30360 }, { "grad_norm": 0.40850192308425903, "learning_rate": 7.312388529527884e-05, "loss": 0.0365, "step": 30370 }, { "grad_norm": 0.31577858328819275, "learning_rate": 7.310555810205577e-05, "loss": 0.0355, "step": 30380 }, { "grad_norm": 0.22665643692016602, "learning_rate": 7.308722696072476e-05, "loss": 0.0359, "step": 30390 }, { "grad_norm": 0.2014240324497223, "learning_rate": 7.306889187441811e-05, "loss": 0.0328, "step": 30400 }, { "grad_norm": 0.2718612253665924, "learning_rate": 7.305055284626876e-05, "loss": 0.0283, "step": 30410 }, { "grad_norm": 0.3494410812854767, "learning_rate": 7.303220987941037e-05, "loss": 0.0343, "step": 30420 }, { "grad_norm": 0.3234085738658905, "learning_rate": 7.301386297697726e-05, "loss": 0.0307, "step": 30430 }, { "grad_norm": 0.4369049072265625, "learning_rate": 7.299551214210438e-05, "loss": 0.0248, "step": 30440 }, { "grad_norm": 0.2841491401195526, "learning_rate": 7.297715737792738e-05, "loss": 0.0315, "step": 30450 }, { "grad_norm": 0.34427717328071594, "learning_rate": 7.295879868758265e-05, "loss": 0.0331, "step": 30460 }, { "grad_norm": 0.2899985909461975, "learning_rate": 7.294043607420713e-05, "loss": 0.0309, "step": 30470 }, { "grad_norm": 0.3347655236721039, "learning_rate": 7.292206954093852e-05, "loss": 0.0294, "step": 30480 }, { "grad_norm": 0.373430997133255, "learning_rate": 7.290369909091515e-05, "loss": 0.0337, "step": 30490 }, { "grad_norm": 0.30412912368774414, "learning_rate": 7.2885324727276e-05, "loss": 0.0325, "step": 30500 }, { "grad_norm": 0.3648809790611267, "learning_rate": 7.286694645316076e-05, "loss": 0.0293, "step": 30510 }, { "grad_norm": 0.32596179842948914, "learning_rate": 7.284856427170982e-05, "loss": 0.0295, "step": 30520 }, { "grad_norm": 0.35216835141181946, "learning_rate": 7.283017818606414e-05, "loss": 0.0364, "step": 30530 }, { "grad_norm": 0.5119812488555908, "learning_rate": 7.28117881993654e-05, "loss": 0.042, "step": 30540 }, { "grad_norm": 0.4638897478580475, "learning_rate": 7.279339431475598e-05, "loss": 0.0292, "step": 30550 }, { "grad_norm": 0.31085342168807983, "learning_rate": 7.277499653537887e-05, "loss": 0.0289, "step": 30560 }, { "grad_norm": 0.4851141571998596, "learning_rate": 7.275659486437776e-05, "loss": 0.0373, "step": 30570 }, { "grad_norm": 0.33807429671287537, "learning_rate": 7.273818930489695e-05, "loss": 0.0403, "step": 30580 }, { "grad_norm": 0.22917556762695312, "learning_rate": 7.271977986008151e-05, "loss": 0.034, "step": 30590 }, { "grad_norm": 0.40721502900123596, "learning_rate": 7.270136653307705e-05, "loss": 0.0391, "step": 30600 }, { "grad_norm": 0.3726303279399872, "learning_rate": 7.268294932702994e-05, "loss": 0.0383, "step": 30610 }, { "grad_norm": 0.3456633388996124, "learning_rate": 7.266452824508719e-05, "loss": 0.0352, "step": 30620 }, { "grad_norm": 0.2836572229862213, "learning_rate": 7.264610329039643e-05, "loss": 0.0335, "step": 30630 }, { "grad_norm": 0.3096032738685608, "learning_rate": 7.262767446610599e-05, "loss": 0.0299, "step": 30640 }, { "grad_norm": 0.4715510308742523, "learning_rate": 7.260924177536485e-05, "loss": 0.0406, "step": 30650 }, { "grad_norm": 0.30910634994506836, "learning_rate": 7.259080522132265e-05, "loss": 0.0327, "step": 30660 }, { "grad_norm": 0.3354969322681427, "learning_rate": 7.257236480712972e-05, "loss": 0.0335, "step": 30670 }, { "grad_norm": 0.32193735241889954, "learning_rate": 7.255392053593697e-05, "loss": 0.0285, "step": 30680 }, { "grad_norm": 0.3460315763950348, "learning_rate": 7.253547241089607e-05, "loss": 0.0259, "step": 30690 }, { "grad_norm": 0.3685179352760315, "learning_rate": 7.251702043515927e-05, "loss": 0.0348, "step": 30700 }, { "grad_norm": 0.2603670060634613, "learning_rate": 7.249856461187952e-05, "loss": 0.0327, "step": 30710 }, { "grad_norm": 0.41436314582824707, "learning_rate": 7.248010494421042e-05, "loss": 0.0299, "step": 30720 }, { "grad_norm": 0.2241179198026657, "learning_rate": 7.246164143530622e-05, "loss": 0.0402, "step": 30730 }, { "grad_norm": 0.3293612599372864, "learning_rate": 7.244317408832181e-05, "loss": 0.0247, "step": 30740 }, { "grad_norm": 0.48946329951286316, "learning_rate": 7.242470290641279e-05, "loss": 0.0331, "step": 30750 }, { "grad_norm": 0.25486502051353455, "learning_rate": 7.240622789273536e-05, "loss": 0.0282, "step": 30760 }, { "grad_norm": 0.39178040623664856, "learning_rate": 7.238774905044638e-05, "loss": 0.0338, "step": 30770 }, { "grad_norm": 0.326900452375412, "learning_rate": 7.236926638270341e-05, "loss": 0.0312, "step": 30780 }, { "grad_norm": 0.36236506700515747, "learning_rate": 7.23507798926646e-05, "loss": 0.0307, "step": 30790 }, { "grad_norm": 0.46881505846977234, "learning_rate": 7.23322895834888e-05, "loss": 0.031, "step": 30800 }, { "grad_norm": 0.3256478011608124, "learning_rate": 7.231379545833552e-05, "loss": 0.0329, "step": 30810 }, { "grad_norm": 0.24092918634414673, "learning_rate": 7.229529752036487e-05, "loss": 0.0328, "step": 30820 }, { "grad_norm": 0.3847922086715698, "learning_rate": 7.227679577273765e-05, "loss": 0.0331, "step": 30830 }, { "grad_norm": 0.4230874478816986, "learning_rate": 7.225829021861529e-05, "loss": 0.0386, "step": 30840 }, { "grad_norm": 0.3088169991970062, "learning_rate": 7.223978086115992e-05, "loss": 0.025, "step": 30850 }, { "grad_norm": 0.31122949719429016, "learning_rate": 7.222126770353425e-05, "loss": 0.0308, "step": 30860 }, { "grad_norm": 0.456034779548645, "learning_rate": 7.22027507489017e-05, "loss": 0.0304, "step": 30870 }, { "grad_norm": 0.3101060688495636, "learning_rate": 7.218423000042627e-05, "loss": 0.0277, "step": 30880 }, { "grad_norm": 0.23124143481254578, "learning_rate": 7.216570546127268e-05, "loss": 0.0284, "step": 30890 }, { "grad_norm": 0.2602880299091339, "learning_rate": 7.214717713460626e-05, "loss": 0.0388, "step": 30900 }, { "grad_norm": 0.36959198117256165, "learning_rate": 7.2128645023593e-05, "loss": 0.0315, "step": 30910 }, { "grad_norm": 0.2633839547634125, "learning_rate": 7.211010913139951e-05, "loss": 0.0321, "step": 30920 }, { "grad_norm": 0.46824511885643005, "learning_rate": 7.209156946119308e-05, "loss": 0.0285, "step": 30930 }, { "grad_norm": 0.33190077543258667, "learning_rate": 7.207302601614166e-05, "loss": 0.0299, "step": 30940 }, { "grad_norm": 0.37035638093948364, "learning_rate": 7.205447879941378e-05, "loss": 0.0333, "step": 30950 }, { "grad_norm": 0.43253645300865173, "learning_rate": 7.203592781417866e-05, "loss": 0.0301, "step": 30960 }, { "grad_norm": 0.42020273208618164, "learning_rate": 7.201737306360617e-05, "loss": 0.0309, "step": 30970 }, { "grad_norm": 0.3700869381427765, "learning_rate": 7.19988145508668e-05, "loss": 0.0273, "step": 30980 }, { "grad_norm": 0.3953740894794464, "learning_rate": 7.198025227913168e-05, "loss": 0.0323, "step": 30990 }, { "grad_norm": 0.3493745028972626, "learning_rate": 7.196168625157261e-05, "loss": 0.0303, "step": 31000 }, { "grad_norm": 0.4652794897556305, "learning_rate": 7.194311647136201e-05, "loss": 0.03, "step": 31010 }, { "grad_norm": 0.4972611963748932, "learning_rate": 7.192454294167297e-05, "loss": 0.027, "step": 31020 }, { "grad_norm": 0.5083319544792175, "learning_rate": 7.190596566567917e-05, "loss": 0.0353, "step": 31030 }, { "grad_norm": 0.33622604608535767, "learning_rate": 7.188738464655496e-05, "loss": 0.039, "step": 31040 }, { "grad_norm": 0.48772385716438293, "learning_rate": 7.186879988747533e-05, "loss": 0.031, "step": 31050 }, { "grad_norm": 0.3009481132030487, "learning_rate": 7.185021139161592e-05, "loss": 0.0242, "step": 31060 }, { "grad_norm": 0.3231717050075531, "learning_rate": 7.1831619162153e-05, "loss": 0.0333, "step": 31070 }, { "grad_norm": 0.32893452048301697, "learning_rate": 7.181302320226345e-05, "loss": 0.0311, "step": 31080 }, { "grad_norm": 0.2513285279273987, "learning_rate": 7.179442351512482e-05, "loss": 0.0372, "step": 31090 }, { "grad_norm": 0.34045344591140747, "learning_rate": 7.177582010391528e-05, "loss": 0.0341, "step": 31100 }, { "grad_norm": 0.4014431834220886, "learning_rate": 7.175721297181366e-05, "loss": 0.0391, "step": 31110 }, { "grad_norm": 0.467827707529068, "learning_rate": 7.173860212199942e-05, "loss": 0.0305, "step": 31120 }, { "grad_norm": 0.38077226281166077, "learning_rate": 7.171998755765263e-05, "loss": 0.0337, "step": 31130 }, { "grad_norm": 0.2958994209766388, "learning_rate": 7.170136928195398e-05, "loss": 0.0286, "step": 31140 }, { "grad_norm": 0.2933282256126404, "learning_rate": 7.168274729808489e-05, "loss": 0.0283, "step": 31150 }, { "grad_norm": 0.36427539587020874, "learning_rate": 7.166412160922728e-05, "loss": 0.032, "step": 31160 }, { "grad_norm": 0.3788098096847534, "learning_rate": 7.164549221856382e-05, "loss": 0.0312, "step": 31170 }, { "grad_norm": 0.17244933545589447, "learning_rate": 7.162685912927775e-05, "loss": 0.042, "step": 31180 }, { "grad_norm": 0.2735154628753662, "learning_rate": 7.160822234455294e-05, "loss": 0.0258, "step": 31190 }, { "grad_norm": 0.3421221375465393, "learning_rate": 7.158958186757391e-05, "loss": 0.0264, "step": 31200 }, { "grad_norm": 0.39140447974205017, "learning_rate": 7.157093770152582e-05, "loss": 0.0306, "step": 31210 }, { "grad_norm": 0.25799620151519775, "learning_rate": 7.155228984959446e-05, "loss": 0.0307, "step": 31220 }, { "grad_norm": 0.5056716799736023, "learning_rate": 7.153363831496621e-05, "loss": 0.0328, "step": 31230 }, { "grad_norm": 0.3121665418148041, "learning_rate": 7.151498310082811e-05, "loss": 0.037, "step": 31240 }, { "grad_norm": 0.2759324014186859, "learning_rate": 7.149632421036784e-05, "loss": 0.0323, "step": 31250 }, { "grad_norm": 0.4189710319042206, "learning_rate": 7.147766164677369e-05, "loss": 0.0527, "step": 31260 }, { "grad_norm": 0.2742319703102112, "learning_rate": 7.145899541323459e-05, "loss": 0.02, "step": 31270 }, { "grad_norm": 0.4245151877403259, "learning_rate": 7.144032551294007e-05, "loss": 0.0342, "step": 31280 }, { "grad_norm": 0.3056701421737671, "learning_rate": 7.14216519490803e-05, "loss": 0.0336, "step": 31290 }, { "grad_norm": 0.4025845527648926, "learning_rate": 7.140297472484609e-05, "loss": 0.0444, "step": 31300 }, { "grad_norm": 0.2507457435131073, "learning_rate": 7.138429384342891e-05, "loss": 0.0514, "step": 31310 }, { "grad_norm": 0.299541175365448, "learning_rate": 7.136560930802074e-05, "loss": 0.0308, "step": 31320 }, { "grad_norm": 0.3755943477153778, "learning_rate": 7.134692112181431e-05, "loss": 0.0316, "step": 31330 }, { "grad_norm": 0.3107064962387085, "learning_rate": 7.13282292880029e-05, "loss": 0.0323, "step": 31340 }, { "grad_norm": 0.34434184432029724, "learning_rate": 7.130953380978043e-05, "loss": 0.0336, "step": 31350 }, { "grad_norm": 0.2759600281715393, "learning_rate": 7.129083469034144e-05, "loss": 0.0378, "step": 31360 }, { "grad_norm": 0.35655999183654785, "learning_rate": 7.127213193288112e-05, "loss": 0.037, "step": 31370 }, { "grad_norm": 0.26378941535949707, "learning_rate": 7.125342554059522e-05, "loss": 0.0277, "step": 31380 }, { "grad_norm": 0.38880863785743713, "learning_rate": 7.12347155166802e-05, "loss": 0.0341, "step": 31390 }, { "grad_norm": 0.35309433937072754, "learning_rate": 7.121600186433306e-05, "loss": 0.0273, "step": 31400 }, { "grad_norm": 0.30945441126823425, "learning_rate": 7.119728458675148e-05, "loss": 0.0294, "step": 31410 }, { "grad_norm": 0.26759642362594604, "learning_rate": 7.117856368713369e-05, "loss": 0.0217, "step": 31420 }, { "grad_norm": 0.3768475353717804, "learning_rate": 7.115983916867861e-05, "loss": 0.0303, "step": 31430 }, { "grad_norm": 0.6186339259147644, "learning_rate": 7.114111103458574e-05, "loss": 0.0274, "step": 31440 }, { "grad_norm": 0.3234468102455139, "learning_rate": 7.11223792880552e-05, "loss": 0.0332, "step": 31450 }, { "grad_norm": 0.4526195526123047, "learning_rate": 7.110364393228773e-05, "loss": 0.0315, "step": 31460 }, { "grad_norm": 0.290327787399292, "learning_rate": 7.108490497048471e-05, "loss": 0.0287, "step": 31470 }, { "grad_norm": 0.434426486492157, "learning_rate": 7.10661624058481e-05, "loss": 0.0365, "step": 31480 }, { "grad_norm": 0.3342244625091553, "learning_rate": 7.10474162415805e-05, "loss": 0.0315, "step": 31490 }, { "grad_norm": 0.25017186999320984, "learning_rate": 7.102866648088511e-05, "loss": 0.0373, "step": 31500 }, { "grad_norm": 0.4644787907600403, "learning_rate": 7.100991312696576e-05, "loss": 0.0299, "step": 31510 }, { "grad_norm": 0.2713409662246704, "learning_rate": 7.099115618302686e-05, "loss": 0.0297, "step": 31520 }, { "grad_norm": 0.328180730342865, "learning_rate": 7.097239565227349e-05, "loss": 0.0353, "step": 31530 }, { "grad_norm": 0.21648967266082764, "learning_rate": 7.09536315379113e-05, "loss": 0.0275, "step": 31540 }, { "grad_norm": 0.3316892981529236, "learning_rate": 7.093486384314656e-05, "loss": 0.0291, "step": 31550 }, { "grad_norm": 0.3592073619365692, "learning_rate": 7.091609257118616e-05, "loss": 0.0339, "step": 31560 }, { "grad_norm": 0.5639441609382629, "learning_rate": 7.08973177252376e-05, "loss": 0.0383, "step": 31570 }, { "grad_norm": 0.31567829847335815, "learning_rate": 7.087853930850898e-05, "loss": 0.0314, "step": 31580 }, { "grad_norm": 0.4517826735973358, "learning_rate": 7.085975732420903e-05, "loss": 0.0333, "step": 31590 }, { "grad_norm": 0.2728775143623352, "learning_rate": 7.084097177554706e-05, "loss": 0.0278, "step": 31600 }, { "grad_norm": 0.3393813371658325, "learning_rate": 7.082218266573301e-05, "loss": 0.0381, "step": 31610 }, { "grad_norm": 0.5566453337669373, "learning_rate": 7.080338999797743e-05, "loss": 0.0376, "step": 31620 }, { "grad_norm": 0.2663707137107849, "learning_rate": 7.07845937754915e-05, "loss": 0.0323, "step": 31630 }, { "grad_norm": 0.36581188440322876, "learning_rate": 7.076579400148693e-05, "loss": 0.0418, "step": 31640 }, { "grad_norm": 0.3111031651496887, "learning_rate": 7.074699067917611e-05, "loss": 0.0379, "step": 31650 }, { "grad_norm": 0.23608888685703278, "learning_rate": 7.072818381177201e-05, "loss": 0.0282, "step": 31660 }, { "grad_norm": 0.23473547399044037, "learning_rate": 7.070937340248823e-05, "loss": 0.031, "step": 31670 }, { "grad_norm": 0.34573543071746826, "learning_rate": 7.069055945453893e-05, "loss": 0.0332, "step": 31680 }, { "grad_norm": 0.3287443518638611, "learning_rate": 7.067174197113892e-05, "loss": 0.0247, "step": 31690 }, { "grad_norm": 0.3451549708843231, "learning_rate": 7.065292095550355e-05, "loss": 0.0307, "step": 31700 }, { "grad_norm": 0.3101724684238434, "learning_rate": 7.063409641084887e-05, "loss": 0.0306, "step": 31710 }, { "grad_norm": 0.32047995924949646, "learning_rate": 7.061526834039145e-05, "loss": 0.0338, "step": 31720 }, { "grad_norm": 0.31989383697509766, "learning_rate": 7.05964367473485e-05, "loss": 0.0224, "step": 31730 }, { "grad_norm": 0.3799598217010498, "learning_rate": 7.057760163493783e-05, "loss": 0.0334, "step": 31740 }, { "grad_norm": 0.38736796379089355, "learning_rate": 7.055876300637783e-05, "loss": 0.0279, "step": 31750 }, { "grad_norm": 0.329391747713089, "learning_rate": 7.053992086488753e-05, "loss": 0.0364, "step": 31760 }, { "grad_norm": 0.25541967153549194, "learning_rate": 7.052107521368651e-05, "loss": 0.0376, "step": 31770 }, { "grad_norm": 0.3155290484428406, "learning_rate": 7.0502226055995e-05, "loss": 0.0393, "step": 31780 }, { "grad_norm": 0.22483527660369873, "learning_rate": 7.048337339503379e-05, "loss": 0.0411, "step": 31790 }, { "grad_norm": 0.3234565258026123, "learning_rate": 7.046451723402427e-05, "loss": 0.0296, "step": 31800 }, { "grad_norm": 0.38195982575416565, "learning_rate": 7.044565757618848e-05, "loss": 0.0432, "step": 31810 }, { "grad_norm": 0.3270483911037445, "learning_rate": 7.042679442474899e-05, "loss": 0.026, "step": 31820 }, { "grad_norm": 0.24284520745277405, "learning_rate": 7.040792778292902e-05, "loss": 0.0403, "step": 31830 }, { "grad_norm": 0.30756884813308716, "learning_rate": 7.038905765395234e-05, "loss": 0.0396, "step": 31840 }, { "grad_norm": 0.31431344151496887, "learning_rate": 7.037018404104334e-05, "loss": 0.0381, "step": 31850 }, { "grad_norm": 0.37173742055892944, "learning_rate": 7.035130694742702e-05, "loss": 0.0265, "step": 31860 }, { "grad_norm": 0.3557199537754059, "learning_rate": 7.033242637632897e-05, "loss": 0.0313, "step": 31870 }, { "grad_norm": 0.4008415937423706, "learning_rate": 7.031354233097534e-05, "loss": 0.031, "step": 31880 }, { "grad_norm": 0.29884180426597595, "learning_rate": 7.029465481459289e-05, "loss": 0.0426, "step": 31890 }, { "grad_norm": 0.33168187737464905, "learning_rate": 7.027576383040898e-05, "loss": 0.0288, "step": 31900 }, { "grad_norm": 0.33176445960998535, "learning_rate": 7.025686938165159e-05, "loss": 0.0387, "step": 31910 }, { "grad_norm": 0.4785069227218628, "learning_rate": 7.023797147154924e-05, "loss": 0.0454, "step": 31920 }, { "grad_norm": 0.275242418050766, "learning_rate": 7.021907010333111e-05, "loss": 0.0278, "step": 31930 }, { "grad_norm": 0.2792867124080658, "learning_rate": 7.020016528022685e-05, "loss": 0.0237, "step": 31940 }, { "grad_norm": 0.3361912965774536, "learning_rate": 7.018125700546683e-05, "loss": 0.0341, "step": 31950 }, { "grad_norm": 0.47828131914138794, "learning_rate": 7.016234528228196e-05, "loss": 0.0333, "step": 31960 }, { "grad_norm": 0.431959867477417, "learning_rate": 7.014343011390372e-05, "loss": 0.0299, "step": 31970 }, { "grad_norm": 0.28426286578178406, "learning_rate": 7.01245115035642e-05, "loss": 0.0234, "step": 31980 }, { "grad_norm": 0.3838920593261719, "learning_rate": 7.010558945449606e-05, "loss": 0.0394, "step": 31990 }, { "grad_norm": 0.4048043191432953, "learning_rate": 7.008666396993258e-05, "loss": 0.0271, "step": 32000 }, { "grad_norm": 0.3370746374130249, "learning_rate": 7.006773505310759e-05, "loss": 0.0293, "step": 32010 }, { "grad_norm": 0.3680848479270935, "learning_rate": 7.004880270725553e-05, "loss": 0.0304, "step": 32020 }, { "grad_norm": 0.2906339168548584, "learning_rate": 7.002986693561144e-05, "loss": 0.0264, "step": 32030 }, { "grad_norm": 0.2672676146030426, "learning_rate": 7.001092774141089e-05, "loss": 0.0324, "step": 32040 }, { "grad_norm": 0.24742786586284637, "learning_rate": 6.999198512789009e-05, "loss": 0.0296, "step": 32050 }, { "grad_norm": 0.5766381025314331, "learning_rate": 6.997303909828584e-05, "loss": 0.0453, "step": 32060 }, { "grad_norm": 0.35499295592308044, "learning_rate": 6.995408965583544e-05, "loss": 0.0361, "step": 32070 }, { "grad_norm": 0.19947387278079987, "learning_rate": 6.993513680377688e-05, "loss": 0.0236, "step": 32080 }, { "grad_norm": 0.30470728874206543, "learning_rate": 6.991618054534868e-05, "loss": 0.0356, "step": 32090 }, { "grad_norm": 0.37927675247192383, "learning_rate": 6.989722088378991e-05, "loss": 0.0271, "step": 32100 }, { "grad_norm": 0.3206133544445038, "learning_rate": 6.987825782234027e-05, "loss": 0.0298, "step": 32110 }, { "grad_norm": 0.42667675018310547, "learning_rate": 6.985929136424006e-05, "loss": 0.0299, "step": 32120 }, { "grad_norm": 0.4147922694683075, "learning_rate": 6.984032151273012e-05, "loss": 0.031, "step": 32130 }, { "grad_norm": 0.3755325675010681, "learning_rate": 6.982134827105186e-05, "loss": 0.0258, "step": 32140 }, { "grad_norm": 0.5248706340789795, "learning_rate": 6.980237164244729e-05, "loss": 0.0267, "step": 32150 }, { "grad_norm": 0.4636741876602173, "learning_rate": 6.9783391630159e-05, "loss": 0.0282, "step": 32160 }, { "grad_norm": 0.40960854291915894, "learning_rate": 6.976440823743015e-05, "loss": 0.0335, "step": 32170 }, { "grad_norm": 0.20566613972187042, "learning_rate": 6.974542146750451e-05, "loss": 0.0263, "step": 32180 }, { "grad_norm": 0.4348551630973816, "learning_rate": 6.972643132362637e-05, "loss": 0.0277, "step": 32190 }, { "grad_norm": 0.27964872121810913, "learning_rate": 6.970743780904064e-05, "loss": 0.0311, "step": 32200 }, { "grad_norm": 0.3309416174888611, "learning_rate": 6.968844092699277e-05, "loss": 0.0352, "step": 32210 }, { "grad_norm": 0.2753322720527649, "learning_rate": 6.966944068072883e-05, "loss": 0.0308, "step": 32220 }, { "grad_norm": 0.3113926649093628, "learning_rate": 6.965043707349545e-05, "loss": 0.0233, "step": 32230 }, { "grad_norm": 0.35878536105155945, "learning_rate": 6.963143010853982e-05, "loss": 0.0353, "step": 32240 }, { "grad_norm": 0.4036201238632202, "learning_rate": 6.961241978910971e-05, "loss": 0.0355, "step": 32250 }, { "grad_norm": 0.4251430034637451, "learning_rate": 6.959340611845344e-05, "loss": 0.0329, "step": 32260 }, { "grad_norm": 0.4441140294075012, "learning_rate": 6.957438909981995e-05, "loss": 0.0291, "step": 32270 }, { "grad_norm": 0.3813125789165497, "learning_rate": 6.955536873645872e-05, "loss": 0.0232, "step": 32280 }, { "grad_norm": 0.322194904088974, "learning_rate": 6.953634503161982e-05, "loss": 0.0323, "step": 32290 }, { "grad_norm": 0.38056617975234985, "learning_rate": 6.951731798855387e-05, "loss": 0.0285, "step": 32300 }, { "grad_norm": 0.3388384282588959, "learning_rate": 6.949828761051208e-05, "loss": 0.0398, "step": 32310 }, { "grad_norm": 0.3240244686603546, "learning_rate": 6.947925390074622e-05, "loss": 0.0275, "step": 32320 }, { "grad_norm": 0.3040470480918884, "learning_rate": 6.946021686250863e-05, "loss": 0.0218, "step": 32330 }, { "grad_norm": 0.3185479938983917, "learning_rate": 6.94411764990522e-05, "loss": 0.0319, "step": 32340 }, { "grad_norm": 0.3568972051143646, "learning_rate": 6.942213281363044e-05, "loss": 0.0225, "step": 32350 }, { "grad_norm": 0.2814798057079315, "learning_rate": 6.940308580949737e-05, "loss": 0.0271, "step": 32360 }, { "grad_norm": 0.2705220878124237, "learning_rate": 6.93840354899076e-05, "loss": 0.0247, "step": 32370 }, { "grad_norm": 0.29374995827674866, "learning_rate": 6.936498185811633e-05, "loss": 0.0293, "step": 32380 }, { "grad_norm": 0.28175753355026245, "learning_rate": 6.93459249173793e-05, "loss": 0.02, "step": 32390 }, { "grad_norm": 0.45566895604133606, "learning_rate": 6.932686467095279e-05, "loss": 0.0289, "step": 32400 }, { "grad_norm": 0.3237261474132538, "learning_rate": 6.930780112209373e-05, "loss": 0.0324, "step": 32410 }, { "grad_norm": 0.26875874400138855, "learning_rate": 6.92887342740595e-05, "loss": 0.0308, "step": 32420 }, { "grad_norm": 0.3022359311580658, "learning_rate": 6.926966413010816e-05, "loss": 0.0379, "step": 32430 }, { "grad_norm": 0.28874555230140686, "learning_rate": 6.925059069349824e-05, "loss": 0.0298, "step": 32440 }, { "grad_norm": 0.3589833676815033, "learning_rate": 6.923151396748886e-05, "loss": 0.0335, "step": 32450 }, { "grad_norm": 0.19028973579406738, "learning_rate": 6.921243395533974e-05, "loss": 0.0285, "step": 32460 }, { "grad_norm": 0.3775513470172882, "learning_rate": 6.919335066031109e-05, "loss": 0.0364, "step": 32470 }, { "grad_norm": 0.45251211524009705, "learning_rate": 6.917426408566379e-05, "loss": 0.0337, "step": 32480 }, { "grad_norm": 0.34061136841773987, "learning_rate": 6.915517423465916e-05, "loss": 0.0341, "step": 32490 }, { "grad_norm": 0.2676742672920227, "learning_rate": 6.913608111055914e-05, "loss": 0.0326, "step": 32500 }, { "grad_norm": 0.3336775600910187, "learning_rate": 6.911698471662623e-05, "loss": 0.0321, "step": 32510 }, { "grad_norm": 0.3625970184803009, "learning_rate": 6.90978850561235e-05, "loss": 0.0268, "step": 32520 }, { "grad_norm": 0.27100870013237, "learning_rate": 6.907878213231454e-05, "loss": 0.0276, "step": 32530 }, { "grad_norm": 0.3493269979953766, "learning_rate": 6.90596759484635e-05, "loss": 0.0306, "step": 32540 }, { "grad_norm": 0.28498584032058716, "learning_rate": 6.904056650783514e-05, "loss": 0.0397, "step": 32550 }, { "grad_norm": 0.23258182406425476, "learning_rate": 6.902145381369471e-05, "loss": 0.0376, "step": 32560 }, { "grad_norm": 0.4762793183326721, "learning_rate": 6.900233786930808e-05, "loss": 0.0375, "step": 32570 }, { "grad_norm": 0.37589237093925476, "learning_rate": 6.898321867794161e-05, "loss": 0.0322, "step": 32580 }, { "grad_norm": 0.3432908356189728, "learning_rate": 6.896409624286226e-05, "loss": 0.0356, "step": 32590 }, { "grad_norm": 0.32102108001708984, "learning_rate": 6.894497056733754e-05, "loss": 0.0379, "step": 32600 }, { "grad_norm": 0.3016708493232727, "learning_rate": 6.89258416546355e-05, "loss": 0.035, "step": 32610 }, { "grad_norm": 0.2554391920566559, "learning_rate": 6.890670950802474e-05, "loss": 0.0348, "step": 32620 }, { "grad_norm": 0.21853460371494293, "learning_rate": 6.88875741307744e-05, "loss": 0.0268, "step": 32630 }, { "grad_norm": 0.1886567622423172, "learning_rate": 6.886843552615425e-05, "loss": 0.0279, "step": 32640 }, { "grad_norm": 0.4027649462223053, "learning_rate": 6.884929369743451e-05, "loss": 0.0349, "step": 32650 }, { "grad_norm": 0.4735437035560608, "learning_rate": 6.8830148647886e-05, "loss": 0.0304, "step": 32660 }, { "grad_norm": 0.3552495241165161, "learning_rate": 6.88110003807801e-05, "loss": 0.0374, "step": 32670 }, { "grad_norm": 0.32985201478004456, "learning_rate": 6.87918488993887e-05, "loss": 0.0358, "step": 32680 }, { "grad_norm": 0.467688649892807, "learning_rate": 6.877269420698431e-05, "loss": 0.038, "step": 32690 }, { "grad_norm": 0.33645981550216675, "learning_rate": 6.875353630683989e-05, "loss": 0.0352, "step": 32700 }, { "grad_norm": 0.44669732451438904, "learning_rate": 6.873437520222905e-05, "loss": 0.0404, "step": 32710 }, { "grad_norm": 0.3914763629436493, "learning_rate": 6.871521089642585e-05, "loss": 0.0323, "step": 32720 }, { "grad_norm": 0.2502870559692383, "learning_rate": 6.869604339270498e-05, "loss": 0.0304, "step": 32730 }, { "grad_norm": 0.2910063862800598, "learning_rate": 6.867687269434164e-05, "loss": 0.0372, "step": 32740 }, { "grad_norm": 0.29884010553359985, "learning_rate": 6.865769880461156e-05, "loss": 0.0323, "step": 32750 }, { "grad_norm": 0.2723960876464844, "learning_rate": 6.863852172679104e-05, "loss": 0.0338, "step": 32760 }, { "grad_norm": 0.31563037633895874, "learning_rate": 6.861934146415693e-05, "loss": 0.0305, "step": 32770 }, { "grad_norm": 0.32172855734825134, "learning_rate": 6.86001580199866e-05, "loss": 0.0416, "step": 32780 }, { "grad_norm": 0.21980607509613037, "learning_rate": 6.858097139755798e-05, "loss": 0.0239, "step": 32790 }, { "grad_norm": 0.36172622442245483, "learning_rate": 6.856178160014955e-05, "loss": 0.0357, "step": 32800 }, { "grad_norm": 0.2572788894176483, "learning_rate": 6.85425886310403e-05, "loss": 0.0288, "step": 32810 }, { "grad_norm": 0.31073835492134094, "learning_rate": 6.852339249350979e-05, "loss": 0.0301, "step": 32820 }, { "grad_norm": 0.2762887179851532, "learning_rate": 6.850419319083812e-05, "loss": 0.0357, "step": 32830 }, { "grad_norm": 0.40670937299728394, "learning_rate": 6.848499072630592e-05, "loss": 0.0314, "step": 32840 }, { "grad_norm": 0.37687087059020996, "learning_rate": 6.846578510319439e-05, "loss": 0.0357, "step": 32850 }, { "grad_norm": 0.3243304193019867, "learning_rate": 6.844657632478519e-05, "loss": 0.0353, "step": 32860 }, { "grad_norm": 0.2322005182504654, "learning_rate": 6.842736439436063e-05, "loss": 0.0243, "step": 32870 }, { "grad_norm": 0.2607971131801605, "learning_rate": 6.84081493152035e-05, "loss": 0.0321, "step": 32880 }, { "grad_norm": 0.380619078874588, "learning_rate": 6.83889310905971e-05, "loss": 0.0279, "step": 32890 }, { "grad_norm": 0.31094786524772644, "learning_rate": 6.836970972382533e-05, "loss": 0.0362, "step": 32900 }, { "grad_norm": 0.2888157069683075, "learning_rate": 6.835048521817257e-05, "loss": 0.0298, "step": 32910 }, { "grad_norm": 0.3542643189430237, "learning_rate": 6.833125757692379e-05, "loss": 0.0324, "step": 32920 }, { "grad_norm": 0.3525804281234741, "learning_rate": 6.831202680336441e-05, "loss": 0.0313, "step": 32930 }, { "grad_norm": 0.210499107837677, "learning_rate": 6.829279290078052e-05, "loss": 0.0289, "step": 32940 }, { "grad_norm": 0.35258641839027405, "learning_rate": 6.827355587245863e-05, "loss": 0.0347, "step": 32950 }, { "grad_norm": 0.2658860385417938, "learning_rate": 6.82543157216858e-05, "loss": 0.0292, "step": 32960 }, { "grad_norm": 0.2550126016139984, "learning_rate": 6.823507245174969e-05, "loss": 0.0286, "step": 32970 }, { "grad_norm": 0.3051714301109314, "learning_rate": 6.821582606593841e-05, "loss": 0.0248, "step": 32980 }, { "grad_norm": 0.3429614305496216, "learning_rate": 6.81965765675407e-05, "loss": 0.0314, "step": 32990 }, { "grad_norm": 0.6217733025550842, "learning_rate": 6.81773239598457e-05, "loss": 0.0296, "step": 33000 }, { "grad_norm": 0.4234807789325714, "learning_rate": 6.815806824614319e-05, "loss": 0.037, "step": 33010 }, { "grad_norm": 0.20912721753120422, "learning_rate": 6.813880942972343e-05, "loss": 0.0238, "step": 33020 }, { "grad_norm": 0.37777021527290344, "learning_rate": 6.811954751387726e-05, "loss": 0.0302, "step": 33030 }, { "grad_norm": 0.32999321818351746, "learning_rate": 6.810028250189598e-05, "loss": 0.0266, "step": 33040 }, { "grad_norm": 0.5161471366882324, "learning_rate": 6.808101439707147e-05, "loss": 0.0393, "step": 33050 }, { "grad_norm": 0.3018495738506317, "learning_rate": 6.806174320269609e-05, "loss": 0.0334, "step": 33060 }, { "grad_norm": 0.37556445598602295, "learning_rate": 6.804246892206281e-05, "loss": 0.0338, "step": 33070 }, { "grad_norm": 0.423198401927948, "learning_rate": 6.802319155846506e-05, "loss": 0.0472, "step": 33080 }, { "grad_norm": 0.34373608231544495, "learning_rate": 6.800391111519679e-05, "loss": 0.0259, "step": 33090 }, { "grad_norm": 0.25223055481910706, "learning_rate": 6.798462759555253e-05, "loss": 0.0362, "step": 33100 }, { "grad_norm": 0.5899559259414673, "learning_rate": 6.79653410028273e-05, "loss": 0.029, "step": 33110 }, { "grad_norm": 0.30447283387184143, "learning_rate": 6.794605134031663e-05, "loss": 0.0313, "step": 33120 }, { "grad_norm": 0.30041638016700745, "learning_rate": 6.792675861131661e-05, "loss": 0.0304, "step": 33130 }, { "grad_norm": 0.3989008665084839, "learning_rate": 6.790746281912386e-05, "loss": 0.0347, "step": 33140 }, { "grad_norm": 0.3153613209724426, "learning_rate": 6.788816396703546e-05, "loss": 0.0211, "step": 33150 }, { "grad_norm": 0.33808791637420654, "learning_rate": 6.78688620583491e-05, "loss": 0.0282, "step": 33160 }, { "grad_norm": 0.27956336736679077, "learning_rate": 6.784955709636292e-05, "loss": 0.0266, "step": 33170 }, { "grad_norm": 0.37028032541275024, "learning_rate": 6.783024908437564e-05, "loss": 0.0278, "step": 33180 }, { "grad_norm": 0.25049859285354614, "learning_rate": 6.781093802568641e-05, "loss": 0.0208, "step": 33190 }, { "grad_norm": 0.3038049638271332, "learning_rate": 6.779162392359504e-05, "loss": 0.0199, "step": 33200 }, { "grad_norm": 0.30663546919822693, "learning_rate": 6.777230678140172e-05, "loss": 0.0339, "step": 33210 }, { "grad_norm": 0.3202745020389557, "learning_rate": 6.775298660240726e-05, "loss": 0.0309, "step": 33220 }, { "grad_norm": 0.2684898376464844, "learning_rate": 6.773366338991292e-05, "loss": 0.0474, "step": 33230 }, { "grad_norm": 0.3081757724285126, "learning_rate": 6.771433714722052e-05, "loss": 0.0304, "step": 33240 }, { "grad_norm": 0.5525970458984375, "learning_rate": 6.769500787763239e-05, "loss": 0.0357, "step": 33250 }, { "grad_norm": 0.2846370041370392, "learning_rate": 6.76756755844514e-05, "loss": 0.0291, "step": 33260 }, { "grad_norm": 0.4353986978530884, "learning_rate": 6.765634027098087e-05, "loss": 0.0347, "step": 33270 }, { "grad_norm": 0.40161848068237305, "learning_rate": 6.763700194052468e-05, "loss": 0.0302, "step": 33280 }, { "grad_norm": 0.3223072290420532, "learning_rate": 6.761766059638723e-05, "loss": 0.0279, "step": 33290 }, { "grad_norm": 0.36896610260009766, "learning_rate": 6.759831624187345e-05, "loss": 0.0314, "step": 33300 }, { "grad_norm": 0.32988232374191284, "learning_rate": 6.757896888028871e-05, "loss": 0.0354, "step": 33310 }, { "grad_norm": 0.26651981472969055, "learning_rate": 6.7559618514939e-05, "loss": 0.0293, "step": 33320 }, { "grad_norm": 0.3227904438972473, "learning_rate": 6.754026514913073e-05, "loss": 0.0285, "step": 33330 }, { "grad_norm": 0.27699509263038635, "learning_rate": 6.752090878617087e-05, "loss": 0.033, "step": 33340 }, { "grad_norm": 0.35150888562202454, "learning_rate": 6.75015494293669e-05, "loss": 0.0264, "step": 33350 }, { "grad_norm": 0.3758482038974762, "learning_rate": 6.74821870820268e-05, "loss": 0.0252, "step": 33360 }, { "grad_norm": 0.20571452379226685, "learning_rate": 6.746282174745907e-05, "loss": 0.0319, "step": 33370 }, { "grad_norm": 0.39171305298805237, "learning_rate": 6.744345342897271e-05, "loss": 0.0375, "step": 33380 }, { "grad_norm": 0.3149794936180115, "learning_rate": 6.742408212987724e-05, "loss": 0.0353, "step": 33390 }, { "grad_norm": 0.28672197461128235, "learning_rate": 6.740470785348269e-05, "loss": 0.0236, "step": 33400 }, { "grad_norm": 0.3383709192276001, "learning_rate": 6.738533060309958e-05, "loss": 0.0311, "step": 33410 }, { "grad_norm": 0.3244625926017761, "learning_rate": 6.736595038203894e-05, "loss": 0.0251, "step": 33420 }, { "grad_norm": 0.24384956061840057, "learning_rate": 6.734656719361236e-05, "loss": 0.03, "step": 33430 }, { "grad_norm": 0.2945652902126312, "learning_rate": 6.732718104113189e-05, "loss": 0.0305, "step": 33440 }, { "grad_norm": 0.3824161887168884, "learning_rate": 6.730779192791006e-05, "loss": 0.0248, "step": 33450 }, { "grad_norm": 0.2656119763851166, "learning_rate": 6.728839985725997e-05, "loss": 0.0269, "step": 33460 }, { "grad_norm": 0.30501794815063477, "learning_rate": 6.726900483249517e-05, "loss": 0.0295, "step": 33470 }, { "grad_norm": 0.3112584948539734, "learning_rate": 6.724960685692976e-05, "loss": 0.0255, "step": 33480 }, { "grad_norm": 0.4809025228023529, "learning_rate": 6.723020593387833e-05, "loss": 0.0279, "step": 33490 }, { "grad_norm": 0.26472046971321106, "learning_rate": 6.721080206665593e-05, "loss": 0.0306, "step": 33500 }, { "grad_norm": 0.30327239632606506, "learning_rate": 6.719139525857819e-05, "loss": 0.0385, "step": 33510 }, { "grad_norm": 0.30450209975242615, "learning_rate": 6.717198551296117e-05, "loss": 0.0283, "step": 33520 }, { "grad_norm": 0.3490113914012909, "learning_rate": 6.715257283312148e-05, "loss": 0.0339, "step": 33530 }, { "grad_norm": 0.3603193759918213, "learning_rate": 6.713315722237623e-05, "loss": 0.0323, "step": 33540 }, { "grad_norm": 0.32892826199531555, "learning_rate": 6.7113738684043e-05, "loss": 0.0331, "step": 33550 }, { "grad_norm": 0.3533191978931427, "learning_rate": 6.709431722143989e-05, "loss": 0.0299, "step": 33560 }, { "grad_norm": 0.27263325452804565, "learning_rate": 6.70748928378855e-05, "loss": 0.0373, "step": 33570 }, { "grad_norm": 0.1812848597764969, "learning_rate": 6.705546553669891e-05, "loss": 0.0192, "step": 33580 }, { "grad_norm": 0.19852256774902344, "learning_rate": 6.703603532119974e-05, "loss": 0.0208, "step": 33590 }, { "grad_norm": 0.2818899154663086, "learning_rate": 6.701660219470808e-05, "loss": 0.0275, "step": 33600 }, { "grad_norm": 0.36501312255859375, "learning_rate": 6.69971661605445e-05, "loss": 0.0249, "step": 33610 }, { "grad_norm": 0.2753349840641022, "learning_rate": 6.697772722203008e-05, "loss": 0.0329, "step": 33620 }, { "grad_norm": 0.3438108563423157, "learning_rate": 6.695828538248643e-05, "loss": 0.0309, "step": 33630 }, { "grad_norm": 0.390380322933197, "learning_rate": 6.693884064523563e-05, "loss": 0.0323, "step": 33640 }, { "grad_norm": 0.32630446553230286, "learning_rate": 6.691939301360023e-05, "loss": 0.0324, "step": 33650 }, { "grad_norm": 0.26451361179351807, "learning_rate": 6.689994249090333e-05, "loss": 0.0302, "step": 33660 }, { "grad_norm": 0.32913535833358765, "learning_rate": 6.688048908046845e-05, "loss": 0.0345, "step": 33670 }, { "grad_norm": 0.19069364666938782, "learning_rate": 6.686103278561969e-05, "loss": 0.0222, "step": 33680 }, { "grad_norm": 0.22885921597480774, "learning_rate": 6.684157360968156e-05, "loss": 0.0275, "step": 33690 }, { "grad_norm": 0.26067832112312317, "learning_rate": 6.682211155597911e-05, "loss": 0.0266, "step": 33700 }, { "grad_norm": 0.32740622758865356, "learning_rate": 6.680264662783789e-05, "loss": 0.03, "step": 33710 }, { "grad_norm": 0.36949002742767334, "learning_rate": 6.678317882858391e-05, "loss": 0.0324, "step": 33720 }, { "grad_norm": 0.23999856412410736, "learning_rate": 6.67637081615437e-05, "loss": 0.0297, "step": 33730 }, { "grad_norm": 0.369790256023407, "learning_rate": 6.674423463004427e-05, "loss": 0.0227, "step": 33740 }, { "grad_norm": 0.30869877338409424, "learning_rate": 6.672475823741308e-05, "loss": 0.0268, "step": 33750 }, { "grad_norm": 0.2626858949661255, "learning_rate": 6.670527898697811e-05, "loss": 0.0318, "step": 33760 }, { "grad_norm": 0.3562728762626648, "learning_rate": 6.668579688206788e-05, "loss": 0.0282, "step": 33770 }, { "grad_norm": 0.2740049958229065, "learning_rate": 6.666631192601131e-05, "loss": 0.0272, "step": 33780 }, { "grad_norm": 0.3311575949192047, "learning_rate": 6.664682412213785e-05, "loss": 0.0296, "step": 33790 }, { "grad_norm": 0.360919713973999, "learning_rate": 6.662733347377745e-05, "loss": 0.029, "step": 33800 }, { "grad_norm": 0.2735474109649658, "learning_rate": 6.660783998426051e-05, "loss": 0.0217, "step": 33810 }, { "grad_norm": 0.2964324355125427, "learning_rate": 6.658834365691794e-05, "loss": 0.0281, "step": 33820 }, { "grad_norm": 0.2356255203485489, "learning_rate": 6.656884449508115e-05, "loss": 0.0279, "step": 33830 }, { "grad_norm": 0.3745068907737732, "learning_rate": 6.654934250208198e-05, "loss": 0.027, "step": 33840 }, { "grad_norm": 0.30489373207092285, "learning_rate": 6.65298376812528e-05, "loss": 0.0305, "step": 33850 }, { "grad_norm": 0.29194220900535583, "learning_rate": 6.651033003592646e-05, "loss": 0.0314, "step": 33860 }, { "grad_norm": 0.31999143958091736, "learning_rate": 6.649081956943626e-05, "loss": 0.0234, "step": 33870 }, { "grad_norm": 0.3587128221988678, "learning_rate": 6.647130628511604e-05, "loss": 0.0306, "step": 33880 }, { "grad_norm": 0.3172418475151062, "learning_rate": 6.645179018630005e-05, "loss": 0.0304, "step": 33890 }, { "grad_norm": 0.3250369131565094, "learning_rate": 6.643227127632309e-05, "loss": 0.0275, "step": 33900 }, { "grad_norm": 0.40686485171318054, "learning_rate": 6.641274955852038e-05, "loss": 0.0323, "step": 33910 }, { "grad_norm": 0.5079012513160706, "learning_rate": 6.639322503622768e-05, "loss": 0.0566, "step": 33920 }, { "grad_norm": 0.2699950933456421, "learning_rate": 6.637369771278116e-05, "loss": 0.0235, "step": 33930 }, { "grad_norm": 0.36649182438850403, "learning_rate": 6.635416759151751e-05, "loss": 0.0253, "step": 33940 }, { "grad_norm": 0.40757113695144653, "learning_rate": 6.633463467577394e-05, "loss": 0.0517, "step": 33950 }, { "grad_norm": 0.30324798822402954, "learning_rate": 6.631509896888803e-05, "loss": 0.0291, "step": 33960 }, { "grad_norm": 0.3182865381240845, "learning_rate": 6.629556047419794e-05, "loss": 0.0237, "step": 33970 }, { "grad_norm": 0.30566689372062683, "learning_rate": 6.627601919504223e-05, "loss": 0.0238, "step": 33980 }, { "grad_norm": 0.27174317836761475, "learning_rate": 6.625647513476001e-05, "loss": 0.0253, "step": 33990 }, { "grad_norm": 0.37970414757728577, "learning_rate": 6.62369282966908e-05, "loss": 0.0342, "step": 34000 }, { "grad_norm": 0.2989344000816345, "learning_rate": 6.621737868417464e-05, "loss": 0.0333, "step": 34010 }, { "grad_norm": 0.22993850708007812, "learning_rate": 6.619782630055198e-05, "loss": 0.0308, "step": 34020 }, { "grad_norm": 0.35457301139831543, "learning_rate": 6.617827114916382e-05, "loss": 0.0301, "step": 34030 }, { "grad_norm": 0.29229238629341125, "learning_rate": 6.615871323335161e-05, "loss": 0.0388, "step": 34040 }, { "grad_norm": 0.24465838074684143, "learning_rate": 6.613915255645725e-05, "loss": 0.0301, "step": 34050 }, { "grad_norm": 0.31797152757644653, "learning_rate": 6.611958912182312e-05, "loss": 0.0292, "step": 34060 }, { "grad_norm": 0.3647797703742981, "learning_rate": 6.610002293279207e-05, "loss": 0.0315, "step": 34070 }, { "grad_norm": 0.3399507403373718, "learning_rate": 6.608045399270746e-05, "loss": 0.0309, "step": 34080 }, { "grad_norm": 0.26419779658317566, "learning_rate": 6.606088230491304e-05, "loss": 0.0291, "step": 34090 }, { "grad_norm": 0.3535561263561249, "learning_rate": 6.604130787275312e-05, "loss": 0.0337, "step": 34100 }, { "grad_norm": 0.3511500358581543, "learning_rate": 6.602173069957242e-05, "loss": 0.027, "step": 34110 }, { "grad_norm": 0.440780907869339, "learning_rate": 6.600215078871612e-05, "loss": 0.0284, "step": 34120 }, { "grad_norm": 0.26568302512168884, "learning_rate": 6.598256814352992e-05, "loss": 0.0316, "step": 34130 }, { "grad_norm": 0.3440282344818115, "learning_rate": 6.596298276735995e-05, "loss": 0.02, "step": 34140 }, { "grad_norm": 0.386436402797699, "learning_rate": 6.594339466355282e-05, "loss": 0.0313, "step": 34150 }, { "grad_norm": 0.2828430235385895, "learning_rate": 6.592380383545558e-05, "loss": 0.0282, "step": 34160 }, { "grad_norm": 0.304823100566864, "learning_rate": 6.590421028641577e-05, "loss": 0.0345, "step": 34170 }, { "grad_norm": 0.35130953788757324, "learning_rate": 6.588461401978143e-05, "loss": 0.032, "step": 34180 }, { "grad_norm": 0.3932972252368927, "learning_rate": 6.586501503890099e-05, "loss": 0.0291, "step": 34190 }, { "grad_norm": 0.2878785729408264, "learning_rate": 6.584541334712338e-05, "loss": 0.0285, "step": 34200 }, { "grad_norm": 0.3457110524177551, "learning_rate": 6.582580894779802e-05, "loss": 0.0313, "step": 34210 }, { "grad_norm": 0.26295793056488037, "learning_rate": 6.580620184427473e-05, "loss": 0.0308, "step": 34220 }, { "grad_norm": 0.257441908121109, "learning_rate": 6.578659203990385e-05, "loss": 0.0261, "step": 34230 }, { "grad_norm": 0.3153543174266815, "learning_rate": 6.576697953803615e-05, "loss": 0.0247, "step": 34240 }, { "grad_norm": 0.3318617045879364, "learning_rate": 6.57473643420229e-05, "loss": 0.029, "step": 34250 }, { "grad_norm": 0.34606796503067017, "learning_rate": 6.572774645521574e-05, "loss": 0.0329, "step": 34260 }, { "grad_norm": 0.40807101130485535, "learning_rate": 6.570812588096688e-05, "loss": 0.0353, "step": 34270 }, { "grad_norm": 0.3973495364189148, "learning_rate": 6.568850262262893e-05, "loss": 0.0336, "step": 34280 }, { "grad_norm": 0.41236212849617004, "learning_rate": 6.566887668355497e-05, "loss": 0.0313, "step": 34290 }, { "grad_norm": 0.31139636039733887, "learning_rate": 6.564924806709851e-05, "loss": 0.0261, "step": 34300 }, { "grad_norm": 0.43112310767173767, "learning_rate": 6.562961677661359e-05, "loss": 0.0348, "step": 34310 }, { "grad_norm": 0.3897986114025116, "learning_rate": 6.56099828154546e-05, "loss": 0.0258, "step": 34320 }, { "grad_norm": 0.359455943107605, "learning_rate": 6.55903461869765e-05, "loss": 0.0257, "step": 34330 }, { "grad_norm": 0.25176748633384705, "learning_rate": 6.557070689453465e-05, "loss": 0.0217, "step": 34340 }, { "grad_norm": 0.22219392657279968, "learning_rate": 6.555106494148482e-05, "loss": 0.0358, "step": 34350 }, { "grad_norm": 0.34097713232040405, "learning_rate": 6.553142033118333e-05, "loss": 0.03, "step": 34360 }, { "grad_norm": 0.3702402114868164, "learning_rate": 6.551177306698688e-05, "loss": 0.0351, "step": 34370 }, { "grad_norm": 0.23352685570716858, "learning_rate": 6.549212315225267e-05, "loss": 0.0346, "step": 34380 }, { "grad_norm": 0.27118808031082153, "learning_rate": 6.547247059033833e-05, "loss": 0.034, "step": 34390 }, { "grad_norm": 0.34230324625968933, "learning_rate": 6.545281538460193e-05, "loss": 0.0292, "step": 34400 }, { "grad_norm": 0.23693102598190308, "learning_rate": 6.543315753840202e-05, "loss": 0.0292, "step": 34410 }, { "grad_norm": 0.37841495871543884, "learning_rate": 6.541349705509758e-05, "loss": 0.0309, "step": 34420 }, { "grad_norm": 0.30338940024375916, "learning_rate": 6.539383393804805e-05, "loss": 0.0317, "step": 34430 }, { "grad_norm": 0.3357982635498047, "learning_rate": 6.537416819061333e-05, "loss": 0.0343, "step": 34440 }, { "grad_norm": 0.27799880504608154, "learning_rate": 6.535449981615375e-05, "loss": 0.0327, "step": 34450 }, { "grad_norm": 0.2604098320007324, "learning_rate": 6.53348288180301e-05, "loss": 0.0251, "step": 34460 }, { "grad_norm": 0.48930490016937256, "learning_rate": 6.531515519960361e-05, "loss": 0.0364, "step": 34470 }, { "grad_norm": 0.27466800808906555, "learning_rate": 6.529547896423597e-05, "loss": 0.031, "step": 34480 }, { "grad_norm": 0.3543141186237335, "learning_rate": 6.52758001152893e-05, "loss": 0.0314, "step": 34490 }, { "grad_norm": 0.27580389380455017, "learning_rate": 6.525611865612618e-05, "loss": 0.0313, "step": 34500 }, { "grad_norm": 0.3374340832233429, "learning_rate": 6.523643459010966e-05, "loss": 0.0255, "step": 34510 }, { "grad_norm": 0.3178052008152008, "learning_rate": 6.521674792060317e-05, "loss": 0.0267, "step": 34520 }, { "grad_norm": 0.29517805576324463, "learning_rate": 6.519705865097063e-05, "loss": 0.0359, "step": 34530 }, { "grad_norm": 0.32997962832450867, "learning_rate": 6.517736678457641e-05, "loss": 0.0226, "step": 34540 }, { "grad_norm": 0.3360059857368469, "learning_rate": 6.515767232478534e-05, "loss": 0.0317, "step": 34550 }, { "grad_norm": 0.30815568566322327, "learning_rate": 6.51379752749626e-05, "loss": 0.0247, "step": 34560 }, { "grad_norm": 0.37588322162628174, "learning_rate": 6.511827563847393e-05, "loss": 0.0322, "step": 34570 }, { "grad_norm": 0.49504464864730835, "learning_rate": 6.509857341868542e-05, "loss": 0.0351, "step": 34580 }, { "grad_norm": 0.30576762557029724, "learning_rate": 6.507886861896367e-05, "loss": 0.0234, "step": 34590 }, { "grad_norm": 0.41121789813041687, "learning_rate": 6.505916124267567e-05, "loss": 0.0341, "step": 34600 }, { "grad_norm": 0.28894442319869995, "learning_rate": 6.503945129318891e-05, "loss": 0.0274, "step": 34610 }, { "grad_norm": 0.4076763689517975, "learning_rate": 6.501973877387122e-05, "loss": 0.0264, "step": 34620 }, { "grad_norm": 0.3950258791446686, "learning_rate": 6.500002368809098e-05, "loss": 0.0274, "step": 34630 }, { "grad_norm": 0.36422550678253174, "learning_rate": 6.498030603921694e-05, "loss": 0.0315, "step": 34640 }, { "grad_norm": 0.3567795157432556, "learning_rate": 6.496058583061832e-05, "loss": 0.0304, "step": 34650 }, { "grad_norm": 0.2207404226064682, "learning_rate": 6.494086306566475e-05, "loss": 0.0212, "step": 34660 }, { "grad_norm": 0.2946997582912445, "learning_rate": 6.492113774772632e-05, "loss": 0.0267, "step": 34670 }, { "grad_norm": 0.4183695316314697, "learning_rate": 6.490140988017354e-05, "loss": 0.0315, "step": 34680 }, { "grad_norm": 0.27985239028930664, "learning_rate": 6.488167946637736e-05, "loss": 0.0314, "step": 34690 }, { "grad_norm": 0.4341282248497009, "learning_rate": 6.486194650970915e-05, "loss": 0.033, "step": 34700 }, { "grad_norm": 0.3201943337917328, "learning_rate": 6.48422110135408e-05, "loss": 0.0308, "step": 34710 }, { "grad_norm": 0.49095842242240906, "learning_rate": 6.482247298124451e-05, "loss": 0.0437, "step": 34720 }, { "grad_norm": 0.26480481028556824, "learning_rate": 6.480273241619297e-05, "loss": 0.0292, "step": 34730 }, { "grad_norm": 0.4817121922969818, "learning_rate": 6.478298932175933e-05, "loss": 0.0312, "step": 34740 }, { "grad_norm": 0.3558942973613739, "learning_rate": 6.476324370131712e-05, "loss": 0.0278, "step": 34750 }, { "grad_norm": 0.32247650623321533, "learning_rate": 6.474349555824036e-05, "loss": 0.0361, "step": 34760 }, { "grad_norm": 0.3145099878311157, "learning_rate": 6.472374489590342e-05, "loss": 0.0289, "step": 34770 }, { "grad_norm": 0.21432755887508392, "learning_rate": 6.470399171768118e-05, "loss": 0.0399, "step": 34780 }, { "grad_norm": 0.25122979283332825, "learning_rate": 6.468423602694891e-05, "loss": 0.0257, "step": 34790 }, { "grad_norm": 0.2740298807621002, "learning_rate": 6.466447782708232e-05, "loss": 0.0219, "step": 34800 }, { "grad_norm": 0.3517705798149109, "learning_rate": 6.464471712145754e-05, "loss": 0.0274, "step": 34810 }, { "grad_norm": 0.4280552566051483, "learning_rate": 6.462495391345114e-05, "loss": 0.0294, "step": 34820 }, { "grad_norm": 0.3146255314350128, "learning_rate": 6.46051882064401e-05, "loss": 0.0225, "step": 34830 }, { "grad_norm": 0.33227720856666565, "learning_rate": 6.458542000380186e-05, "loss": 0.0256, "step": 34840 }, { "grad_norm": 0.30885058641433716, "learning_rate": 6.456564930891424e-05, "loss": 0.0301, "step": 34850 }, { "grad_norm": 0.3390328288078308, "learning_rate": 6.454587612515555e-05, "loss": 0.0345, "step": 34860 }, { "grad_norm": 0.375132292509079, "learning_rate": 6.452610045590444e-05, "loss": 0.032, "step": 34870 }, { "grad_norm": 0.34749719500541687, "learning_rate": 6.450632230454005e-05, "loss": 0.0288, "step": 34880 }, { "grad_norm": 0.29646050930023193, "learning_rate": 6.448654167444195e-05, "loss": 0.0346, "step": 34890 }, { "grad_norm": 0.28383123874664307, "learning_rate": 6.446675856899005e-05, "loss": 0.0263, "step": 34900 }, { "grad_norm": 0.20585252344608307, "learning_rate": 6.444697299156481e-05, "loss": 0.0289, "step": 34910 }, { "grad_norm": 0.31249526143074036, "learning_rate": 6.442718494554701e-05, "loss": 0.0271, "step": 34920 }, { "grad_norm": 0.3902466893196106, "learning_rate": 6.440739443431787e-05, "loss": 0.0269, "step": 34930 }, { "grad_norm": 0.2134762555360794, "learning_rate": 6.438760146125906e-05, "loss": 0.0267, "step": 34940 }, { "grad_norm": 0.2888185381889343, "learning_rate": 6.436780602975267e-05, "loss": 0.0313, "step": 34950 }, { "grad_norm": 0.3802764415740967, "learning_rate": 6.43480081431812e-05, "loss": 0.0323, "step": 34960 }, { "grad_norm": 0.34701448678970337, "learning_rate": 6.432820780492756e-05, "loss": 0.0333, "step": 34970 }, { "grad_norm": 0.37268519401550293, "learning_rate": 6.430840501837506e-05, "loss": 0.0338, "step": 34980 }, { "grad_norm": 0.33944252133369446, "learning_rate": 6.428859978690748e-05, "loss": 0.0283, "step": 34990 }, { "grad_norm": 0.38295188546180725, "learning_rate": 6.426879211390901e-05, "loss": 0.0323, "step": 35000 }, { "grad_norm": 0.2993597686290741, "learning_rate": 6.424898200276422e-05, "loss": 0.0323, "step": 35010 }, { "grad_norm": 0.2964978814125061, "learning_rate": 6.42291694568581e-05, "loss": 0.0253, "step": 35020 }, { "grad_norm": 0.3999641239643097, "learning_rate": 6.42093544795761e-05, "loss": 0.0351, "step": 35030 }, { "grad_norm": 0.2577328085899353, "learning_rate": 6.418953707430403e-05, "loss": 0.0364, "step": 35040 }, { "grad_norm": 0.24218867719173431, "learning_rate": 6.416971724442819e-05, "loss": 0.026, "step": 35050 }, { "grad_norm": 0.2157934159040451, "learning_rate": 6.414989499333519e-05, "loss": 0.0278, "step": 35060 }, { "grad_norm": 0.30225276947021484, "learning_rate": 6.413007032441214e-05, "loss": 0.0309, "step": 35070 }, { "grad_norm": 0.29013627767562866, "learning_rate": 6.411024324104653e-05, "loss": 0.0322, "step": 35080 }, { "grad_norm": 0.31283700466156006, "learning_rate": 6.409041374662628e-05, "loss": 0.027, "step": 35090 }, { "grad_norm": 0.4542400538921356, "learning_rate": 6.407058184453967e-05, "loss": 0.0412, "step": 35100 }, { "grad_norm": 0.2777858078479767, "learning_rate": 6.405074753817548e-05, "loss": 0.0305, "step": 35110 }, { "grad_norm": 0.5081503987312317, "learning_rate": 6.40309108309228e-05, "loss": 0.0302, "step": 35120 }, { "grad_norm": 0.3533255159854889, "learning_rate": 6.401107172617122e-05, "loss": 0.0351, "step": 35130 }, { "grad_norm": 0.38289764523506165, "learning_rate": 6.399123022731068e-05, "loss": 0.0503, "step": 35140 }, { "grad_norm": 0.3835667669773102, "learning_rate": 6.397138633773157e-05, "loss": 0.0338, "step": 35150 }, { "grad_norm": 0.42722445726394653, "learning_rate": 6.395154006082463e-05, "loss": 0.0283, "step": 35160 }, { "grad_norm": 0.40837615728378296, "learning_rate": 6.393169139998109e-05, "loss": 0.0282, "step": 35170 }, { "grad_norm": 0.3113228678703308, "learning_rate": 6.39118403585925e-05, "loss": 0.0277, "step": 35180 }, { "grad_norm": 0.3282149136066437, "learning_rate": 6.38919869400509e-05, "loss": 0.0288, "step": 35190 }, { "grad_norm": 0.3145189881324768, "learning_rate": 6.387213114774865e-05, "loss": 0.0336, "step": 35200 }, { "grad_norm": 0.4444306194782257, "learning_rate": 6.385227298507863e-05, "loss": 0.053, "step": 35210 }, { "grad_norm": 0.4447457492351532, "learning_rate": 6.3832412455434e-05, "loss": 0.0333, "step": 35220 }, { "grad_norm": 0.23332709074020386, "learning_rate": 6.381254956220841e-05, "loss": 0.0285, "step": 35230 }, { "grad_norm": 0.32931238412857056, "learning_rate": 6.379268430879586e-05, "loss": 0.0298, "step": 35240 }, { "grad_norm": 0.302214652299881, "learning_rate": 6.37728166985908e-05, "loss": 0.0261, "step": 35250 }, { "grad_norm": 0.19757208228111267, "learning_rate": 6.375294673498804e-05, "loss": 0.0258, "step": 35260 }, { "grad_norm": 0.3479123115539551, "learning_rate": 6.373307442138284e-05, "loss": 0.035, "step": 35270 }, { "grad_norm": 0.3140723407268524, "learning_rate": 6.371319976117081e-05, "loss": 0.0409, "step": 35280 }, { "grad_norm": 0.3476649820804596, "learning_rate": 6.3693322757748e-05, "loss": 0.0303, "step": 35290 }, { "grad_norm": 0.31588509678840637, "learning_rate": 6.367344341451086e-05, "loss": 0.0325, "step": 35300 }, { "grad_norm": 0.3526175916194916, "learning_rate": 6.36535617348562e-05, "loss": 0.0303, "step": 35310 }, { "grad_norm": 0.28908953070640564, "learning_rate": 6.363367772218128e-05, "loss": 0.0317, "step": 35320 }, { "grad_norm": 0.331636518239975, "learning_rate": 6.36137913798837e-05, "loss": 0.0277, "step": 35330 }, { "grad_norm": 0.4198959767818451, "learning_rate": 6.359390271136151e-05, "loss": 0.0241, "step": 35340 }, { "grad_norm": 0.38493427634239197, "learning_rate": 6.357401172001314e-05, "loss": 0.0289, "step": 35350 }, { "grad_norm": 0.24111467599868774, "learning_rate": 6.355411840923742e-05, "loss": 0.0297, "step": 35360 }, { "grad_norm": 0.34255585074424744, "learning_rate": 6.353422278243358e-05, "loss": 0.0282, "step": 35370 }, { "grad_norm": 0.3239433467388153, "learning_rate": 6.351432484300121e-05, "loss": 0.0321, "step": 35380 }, { "grad_norm": 0.29789549112319946, "learning_rate": 6.349442459434036e-05, "loss": 0.0326, "step": 35390 }, { "grad_norm": 0.2148335576057434, "learning_rate": 6.34745220398514e-05, "loss": 0.0313, "step": 35400 }, { "grad_norm": 0.40558940172195435, "learning_rate": 6.345461718293518e-05, "loss": 0.0272, "step": 35410 }, { "grad_norm": 0.2966856360435486, "learning_rate": 6.343471002699286e-05, "loss": 0.0309, "step": 35420 }, { "grad_norm": 0.46328210830688477, "learning_rate": 6.341480057542602e-05, "loss": 0.0309, "step": 35430 }, { "grad_norm": 0.3721269369125366, "learning_rate": 6.339488883163667e-05, "loss": 0.0298, "step": 35440 }, { "grad_norm": 0.348336398601532, "learning_rate": 6.337497479902716e-05, "loss": 0.0286, "step": 35450 }, { "grad_norm": 0.32172760367393494, "learning_rate": 6.335505848100027e-05, "loss": 0.0274, "step": 35460 }, { "grad_norm": 0.2134677618741989, "learning_rate": 6.333513988095915e-05, "loss": 0.0224, "step": 35470 }, { "grad_norm": 0.35449814796447754, "learning_rate": 6.331521900230735e-05, "loss": 0.0246, "step": 35480 }, { "grad_norm": 0.19358956813812256, "learning_rate": 6.329529584844878e-05, "loss": 0.0259, "step": 35490 }, { "grad_norm": 0.3323582708835602, "learning_rate": 6.327537042278777e-05, "loss": 0.0288, "step": 35500 }, { "grad_norm": 0.46215465664863586, "learning_rate": 6.325544272872905e-05, "loss": 0.0308, "step": 35510 }, { "grad_norm": 0.29729315638542175, "learning_rate": 6.323551276967771e-05, "loss": 0.0345, "step": 35520 }, { "grad_norm": 0.37955960631370544, "learning_rate": 6.321558054903922e-05, "loss": 0.0276, "step": 35530 }, { "grad_norm": 0.34257686138153076, "learning_rate": 6.319564607021947e-05, "loss": 0.0281, "step": 35540 }, { "grad_norm": 0.4183367192745209, "learning_rate": 6.31757093366247e-05, "loss": 0.0266, "step": 35550 }, { "grad_norm": 0.39620548486709595, "learning_rate": 6.315577035166154e-05, "loss": 0.0271, "step": 35560 }, { "grad_norm": 0.46922022104263306, "learning_rate": 6.313582911873708e-05, "loss": 0.0315, "step": 35570 }, { "grad_norm": 0.2645760774612427, "learning_rate": 6.311588564125865e-05, "loss": 0.0273, "step": 35580 }, { "grad_norm": 0.484537810087204, "learning_rate": 6.30959399226341e-05, "loss": 0.041, "step": 35590 }, { "grad_norm": 0.30071747303009033, "learning_rate": 6.30759919662716e-05, "loss": 0.0404, "step": 35600 }, { "grad_norm": 0.351480633020401, "learning_rate": 6.30560417755797e-05, "loss": 0.0266, "step": 35610 }, { "grad_norm": 0.28156304359436035, "learning_rate": 6.303608935396735e-05, "loss": 0.0245, "step": 35620 }, { "grad_norm": 0.2622629702091217, "learning_rate": 6.301613470484386e-05, "loss": 0.0297, "step": 35630 }, { "grad_norm": 0.30884042382240295, "learning_rate": 6.299617783161893e-05, "loss": 0.0263, "step": 35640 }, { "grad_norm": 0.27523547410964966, "learning_rate": 6.297621873770266e-05, "loss": 0.0281, "step": 35650 }, { "grad_norm": 0.32307863235473633, "learning_rate": 6.29562574265055e-05, "loss": 0.0284, "step": 35660 }, { "grad_norm": 0.34381476044654846, "learning_rate": 6.293629390143834e-05, "loss": 0.0348, "step": 35670 }, { "grad_norm": 0.3208024203777313, "learning_rate": 6.291632816591232e-05, "loss": 0.0306, "step": 35680 }, { "grad_norm": 0.27273449301719666, "learning_rate": 6.28963602233391e-05, "loss": 0.0282, "step": 35690 }, { "grad_norm": 0.28080400824546814, "learning_rate": 6.287639007713062e-05, "loss": 0.0296, "step": 35700 }, { "grad_norm": 0.34996065497398376, "learning_rate": 6.285641773069926e-05, "loss": 0.0279, "step": 35710 }, { "grad_norm": 0.29383817315101624, "learning_rate": 6.283644318745773e-05, "loss": 0.0241, "step": 35720 }, { "grad_norm": 0.30001577734947205, "learning_rate": 6.281646645081912e-05, "loss": 0.0325, "step": 35730 }, { "grad_norm": 0.346221923828125, "learning_rate": 6.279648752419693e-05, "loss": 0.0248, "step": 35740 }, { "grad_norm": 0.3568514585494995, "learning_rate": 6.2776506411005e-05, "loss": 0.0268, "step": 35750 }, { "grad_norm": 0.3791641294956207, "learning_rate": 6.275652311465758e-05, "loss": 0.0289, "step": 35760 }, { "grad_norm": 0.2530185878276825, "learning_rate": 6.273653763856926e-05, "loss": 0.0304, "step": 35770 }, { "grad_norm": 0.32048946619033813, "learning_rate": 6.271654998615501e-05, "loss": 0.0326, "step": 35780 }, { "grad_norm": 0.3305789828300476, "learning_rate": 6.269656016083013e-05, "loss": 0.0263, "step": 35790 }, { "grad_norm": 0.2572552561759949, "learning_rate": 6.267656816601038e-05, "loss": 0.032, "step": 35800 }, { "grad_norm": 0.34699398279190063, "learning_rate": 6.265657400511185e-05, "loss": 0.0337, "step": 35810 }, { "grad_norm": 0.29160404205322266, "learning_rate": 6.263657768155098e-05, "loss": 0.0279, "step": 35820 }, { "grad_norm": 0.23755882680416107, "learning_rate": 6.261657919874457e-05, "loss": 0.0272, "step": 35830 }, { "grad_norm": 0.31684356927871704, "learning_rate": 6.259657856010986e-05, "loss": 0.0337, "step": 35840 }, { "grad_norm": 0.31535205245018005, "learning_rate": 6.257657576906439e-05, "loss": 0.0276, "step": 35850 }, { "grad_norm": 0.47957196831703186, "learning_rate": 6.255657082902609e-05, "loss": 0.0249, "step": 35860 }, { "grad_norm": 0.30594369769096375, "learning_rate": 6.253656374341325e-05, "loss": 0.0222, "step": 35870 }, { "grad_norm": 0.3608192801475525, "learning_rate": 6.251655451564457e-05, "loss": 0.0288, "step": 35880 }, { "grad_norm": 0.22194914519786835, "learning_rate": 6.249654314913902e-05, "loss": 0.0244, "step": 35890 }, { "grad_norm": 0.20553582906723022, "learning_rate": 6.247652964731604e-05, "loss": 0.024, "step": 35900 }, { "grad_norm": 0.3073943853378296, "learning_rate": 6.245651401359537e-05, "loss": 0.0387, "step": 35910 }, { "grad_norm": 0.3890179395675659, "learning_rate": 6.243649625139715e-05, "loss": 0.029, "step": 35920 }, { "grad_norm": 0.37708795070648193, "learning_rate": 6.241647636414185e-05, "loss": 0.0255, "step": 35930 }, { "grad_norm": 0.5104573369026184, "learning_rate": 6.239645435525034e-05, "loss": 0.0332, "step": 35940 }, { "grad_norm": 0.2824173867702484, "learning_rate": 6.237643022814381e-05, "loss": 0.0284, "step": 35950 }, { "grad_norm": 0.32968658208847046, "learning_rate": 6.235640398624386e-05, "loss": 0.0296, "step": 35960 }, { "grad_norm": 0.25064340233802795, "learning_rate": 6.233637563297243e-05, "loss": 0.0362, "step": 35970 }, { "grad_norm": 0.45310166478157043, "learning_rate": 6.23163451717518e-05, "loss": 0.0303, "step": 35980 }, { "grad_norm": 0.24228529632091522, "learning_rate": 6.229631260600463e-05, "loss": 0.0306, "step": 35990 }, { "grad_norm": 0.33618593215942383, "learning_rate": 6.227627793915392e-05, "loss": 0.0305, "step": 36000 }, { "grad_norm": 0.44276732206344604, "learning_rate": 6.225624117462309e-05, "loss": 0.0299, "step": 36010 }, { "grad_norm": 0.40365880727767944, "learning_rate": 6.223620231583586e-05, "loss": 0.0291, "step": 36020 }, { "grad_norm": 0.43379175662994385, "learning_rate": 6.221616136621629e-05, "loss": 0.0225, "step": 36030 }, { "grad_norm": 0.3553803563117981, "learning_rate": 6.219611832918887e-05, "loss": 0.0279, "step": 36040 }, { "grad_norm": 0.2189980000257492, "learning_rate": 6.217607320817838e-05, "loss": 0.0229, "step": 36050 }, { "grad_norm": 0.3604696989059448, "learning_rate": 6.215602600661001e-05, "loss": 0.0256, "step": 36060 }, { "grad_norm": 0.3809962570667267, "learning_rate": 6.213597672790925e-05, "loss": 0.0266, "step": 36070 }, { "grad_norm": 0.35393813252449036, "learning_rate": 6.2115925375502e-05, "loss": 0.0267, "step": 36080 }, { "grad_norm": 0.3062933385372162, "learning_rate": 6.209587195281447e-05, "loss": 0.0308, "step": 36090 }, { "grad_norm": 0.3799043297767639, "learning_rate": 6.207581646327324e-05, "loss": 0.033, "step": 36100 }, { "grad_norm": 0.5007234811782837, "learning_rate": 6.205575891030526e-05, "loss": 0.0231, "step": 36110 }, { "grad_norm": 0.23790505528450012, "learning_rate": 6.203569929733781e-05, "loss": 0.025, "step": 36120 }, { "grad_norm": 0.3789195120334625, "learning_rate": 6.201563762779852e-05, "loss": 0.0292, "step": 36130 }, { "grad_norm": 0.3244292140007019, "learning_rate": 6.199557390511538e-05, "loss": 0.024, "step": 36140 }, { "grad_norm": 0.23383410274982452, "learning_rate": 6.197550813271675e-05, "loss": 0.0221, "step": 36150 }, { "grad_norm": 0.2472129911184311, "learning_rate": 6.195544031403131e-05, "loss": 0.0284, "step": 36160 }, { "grad_norm": 0.5324993133544922, "learning_rate": 6.19353704524881e-05, "loss": 0.027, "step": 36170 }, { "grad_norm": 0.41535037755966187, "learning_rate": 6.191529855151652e-05, "loss": 0.0253, "step": 36180 }, { "grad_norm": 0.32861778140068054, "learning_rate": 6.189522461454629e-05, "loss": 0.0321, "step": 36190 }, { "grad_norm": 0.39483746886253357, "learning_rate": 6.187514864500752e-05, "loss": 0.0279, "step": 36200 }, { "grad_norm": 0.4639918804168701, "learning_rate": 6.185507064633062e-05, "loss": 0.0304, "step": 36210 }, { "grad_norm": 0.3679129481315613, "learning_rate": 6.18349906219464e-05, "loss": 0.0349, "step": 36220 }, { "grad_norm": 0.37455177307128906, "learning_rate": 6.181490857528596e-05, "loss": 0.0277, "step": 36230 }, { "grad_norm": 0.27270978689193726, "learning_rate": 6.179482450978077e-05, "loss": 0.0231, "step": 36240 }, { "grad_norm": 0.4719718396663666, "learning_rate": 6.177473842886269e-05, "loss": 0.0268, "step": 36250 }, { "grad_norm": 0.25340986251831055, "learning_rate": 6.175465033596382e-05, "loss": 0.0258, "step": 36260 }, { "grad_norm": 0.43084779381752014, "learning_rate": 6.173456023451671e-05, "loss": 0.0273, "step": 36270 }, { "grad_norm": 0.35217010974884033, "learning_rate": 6.171446812795422e-05, "loss": 0.0279, "step": 36280 }, { "grad_norm": 0.33540141582489014, "learning_rate": 6.169437401970949e-05, "loss": 0.0366, "step": 36290 }, { "grad_norm": 0.2902522087097168, "learning_rate": 6.16742779132161e-05, "loss": 0.0259, "step": 36300 }, { "grad_norm": 0.25259947776794434, "learning_rate": 6.165417981190789e-05, "loss": 0.0289, "step": 36310 }, { "grad_norm": 0.2126581370830536, "learning_rate": 6.16340797192191e-05, "loss": 0.0212, "step": 36320 }, { "grad_norm": 0.3089945614337921, "learning_rate": 6.161397763858427e-05, "loss": 0.0249, "step": 36330 }, { "grad_norm": 0.3062695264816284, "learning_rate": 6.159387357343834e-05, "loss": 0.0186, "step": 36340 }, { "grad_norm": 0.25928571820259094, "learning_rate": 6.157376752721648e-05, "loss": 0.0255, "step": 36350 }, { "grad_norm": 0.3297683894634247, "learning_rate": 6.155365950335428e-05, "loss": 0.0294, "step": 36360 }, { "grad_norm": 0.4230065941810608, "learning_rate": 6.153354950528768e-05, "loss": 0.0286, "step": 36370 }, { "grad_norm": 0.34814849495887756, "learning_rate": 6.151343753645293e-05, "loss": 0.0313, "step": 36380 }, { "grad_norm": 0.371071457862854, "learning_rate": 6.149332360028657e-05, "loss": 0.0329, "step": 36390 }, { "grad_norm": 0.36160436272621155, "learning_rate": 6.147320770022555e-05, "loss": 0.0288, "step": 36400 }, { "grad_norm": 0.4231102466583252, "learning_rate": 6.145308983970715e-05, "loss": 0.0288, "step": 36410 }, { "grad_norm": 0.3055754601955414, "learning_rate": 6.143297002216892e-05, "loss": 0.0234, "step": 36420 }, { "grad_norm": 0.377127081155777, "learning_rate": 6.141284825104882e-05, "loss": 0.0301, "step": 36430 }, { "grad_norm": 0.3757885694503784, "learning_rate": 6.13927245297851e-05, "loss": 0.0344, "step": 36440 }, { "grad_norm": 0.22498194873332977, "learning_rate": 6.137259886181633e-05, "loss": 0.0313, "step": 36450 }, { "grad_norm": 0.30146750807762146, "learning_rate": 6.135247125058145e-05, "loss": 0.0315, "step": 36460 }, { "grad_norm": 0.20221024751663208, "learning_rate": 6.133234169951974e-05, "loss": 0.0236, "step": 36470 }, { "grad_norm": 0.3882438838481903, "learning_rate": 6.131221021207078e-05, "loss": 0.0281, "step": 36480 }, { "grad_norm": 0.2755313515663147, "learning_rate": 6.129207679167448e-05, "loss": 0.0275, "step": 36490 }, { "grad_norm": 0.2993893623352051, "learning_rate": 6.127194144177109e-05, "loss": 0.0225, "step": 36500 }, { "grad_norm": 0.2983550727367401, "learning_rate": 6.125180416580118e-05, "loss": 0.0235, "step": 36510 }, { "grad_norm": 0.28385862708091736, "learning_rate": 6.123166496720571e-05, "loss": 0.0268, "step": 36520 }, { "grad_norm": 0.3927823305130005, "learning_rate": 6.121152384942588e-05, "loss": 0.0314, "step": 36530 }, { "grad_norm": 0.3414541184902191, "learning_rate": 6.119138081590324e-05, "loss": 0.0286, "step": 36540 }, { "grad_norm": 0.23305846750736237, "learning_rate": 6.117123587007971e-05, "loss": 0.0312, "step": 36550 }, { "grad_norm": 0.2788001000881195, "learning_rate": 6.11510890153975e-05, "loss": 0.0405, "step": 36560 }, { "grad_norm": 0.33940431475639343, "learning_rate": 6.113094025529916e-05, "loss": 0.0245, "step": 36570 }, { "grad_norm": 0.40844252705574036, "learning_rate": 6.111078959322757e-05, "loss": 0.0478, "step": 36580 }, { "grad_norm": 0.35232505202293396, "learning_rate": 6.109063703262592e-05, "loss": 0.0339, "step": 36590 }, { "grad_norm": 0.3090684711933136, "learning_rate": 6.107048257693772e-05, "loss": 0.0346, "step": 36600 }, { "grad_norm": 0.26215294003486633, "learning_rate": 6.105032622960683e-05, "loss": 0.0304, "step": 36610 }, { "grad_norm": 0.34955376386642456, "learning_rate": 6.103016799407743e-05, "loss": 0.0294, "step": 36620 }, { "grad_norm": 0.32430076599121094, "learning_rate": 6.1010007873793984e-05, "loss": 0.0324, "step": 36630 }, { "grad_norm": 0.236144557595253, "learning_rate": 6.098984587220131e-05, "loss": 0.025, "step": 36640 }, { "grad_norm": 0.23098906874656677, "learning_rate": 6.096968199274456e-05, "loss": 0.0295, "step": 36650 }, { "grad_norm": 0.30879655480384827, "learning_rate": 6.0949516238869166e-05, "loss": 0.0268, "step": 36660 }, { "grad_norm": 0.27699729800224304, "learning_rate": 6.092934861402092e-05, "loss": 0.0368, "step": 36670 }, { "grad_norm": 0.426439106464386, "learning_rate": 6.0909179121645924e-05, "loss": 0.0366, "step": 36680 }, { "grad_norm": 0.3353533148765564, "learning_rate": 6.0889007765190576e-05, "loss": 0.0361, "step": 36690 }, { "grad_norm": 0.3021993637084961, "learning_rate": 6.086883454810162e-05, "loss": 0.0268, "step": 36700 }, { "grad_norm": 0.315593421459198, "learning_rate": 6.0848659473826084e-05, "loss": 0.034, "step": 36710 }, { "grad_norm": 0.8420658707618713, "learning_rate": 6.082848254581138e-05, "loss": 0.0289, "step": 36720 }, { "grad_norm": 0.35497185587882996, "learning_rate": 6.080830376750517e-05, "loss": 0.0297, "step": 36730 }, { "grad_norm": 0.31888118386268616, "learning_rate": 6.0788123142355445e-05, "loss": 0.0235, "step": 36740 }, { "grad_norm": 0.43615609407424927, "learning_rate": 6.076794067381052e-05, "loss": 0.0307, "step": 36750 }, { "grad_norm": 0.2918499708175659, "learning_rate": 6.074775636531905e-05, "loss": 0.0289, "step": 36760 }, { "grad_norm": 0.38415902853012085, "learning_rate": 6.072757022032997e-05, "loss": 0.0344, "step": 36770 }, { "grad_norm": 0.2538168132305145, "learning_rate": 6.070738224229253e-05, "loss": 0.036, "step": 36780 }, { "grad_norm": 0.3271588683128357, "learning_rate": 6.0687192434656314e-05, "loss": 0.0295, "step": 36790 }, { "grad_norm": 0.3245992660522461, "learning_rate": 6.066700080087121e-05, "loss": 0.0251, "step": 36800 }, { "grad_norm": 0.4294223189353943, "learning_rate": 6.0646807344387424e-05, "loss": 0.0302, "step": 36810 }, { "grad_norm": 0.2849973142147064, "learning_rate": 6.062661206865543e-05, "loss": 0.0332, "step": 36820 }, { "grad_norm": 0.32527855038642883, "learning_rate": 6.06064149771261e-05, "loss": 0.0289, "step": 36830 }, { "grad_norm": 0.2694505453109741, "learning_rate": 6.058621607325051e-05, "loss": 0.0325, "step": 36840 }, { "grad_norm": 0.33215275406837463, "learning_rate": 6.056601536048014e-05, "loss": 0.0345, "step": 36850 }, { "grad_norm": 0.3779591917991638, "learning_rate": 6.0545812842266725e-05, "loss": 0.0358, "step": 36860 }, { "grad_norm": 0.3028695285320282, "learning_rate": 6.052560852206232e-05, "loss": 0.0305, "step": 36870 }, { "grad_norm": 0.332357794046402, "learning_rate": 6.05054024033193e-05, "loss": 0.0357, "step": 36880 }, { "grad_norm": 0.2982681095600128, "learning_rate": 6.048519448949032e-05, "loss": 0.0271, "step": 36890 }, { "grad_norm": 0.2190476953983307, "learning_rate": 6.046498478402839e-05, "loss": 0.03, "step": 36900 }, { "grad_norm": 0.3968721330165863, "learning_rate": 6.044477329038677e-05, "loss": 0.0327, "step": 36910 }, { "grad_norm": 0.4216737747192383, "learning_rate": 6.042456001201906e-05, "loss": 0.0304, "step": 36920 }, { "grad_norm": 0.23074756562709808, "learning_rate": 6.040434495237917e-05, "loss": 0.0305, "step": 36930 }, { "grad_norm": 0.3388216197490692, "learning_rate": 6.0384128114921256e-05, "loss": 0.0251, "step": 36940 }, { "grad_norm": 0.46199744939804077, "learning_rate": 6.036390950309987e-05, "loss": 0.0218, "step": 36950 }, { "grad_norm": 0.368252694606781, "learning_rate": 6.0343689120369805e-05, "loss": 0.0317, "step": 36960 }, { "grad_norm": 0.4224070906639099, "learning_rate": 6.032346697018616e-05, "loss": 0.0304, "step": 36970 }, { "grad_norm": 0.24407413601875305, "learning_rate": 6.0303243056004375e-05, "loss": 0.0417, "step": 36980 }, { "grad_norm": 0.3077678978443146, "learning_rate": 6.0283017381280136e-05, "loss": 0.0277, "step": 36990 }, { "grad_norm": 0.3305642008781433, "learning_rate": 6.026278994946948e-05, "loss": 0.0279, "step": 37000 }, { "grad_norm": 0.1898428201675415, "learning_rate": 6.02425607640287e-05, "loss": 0.03, "step": 37010 }, { "grad_norm": 0.5732061266899109, "learning_rate": 6.022232982841441e-05, "loss": 0.028, "step": 37020 }, { "grad_norm": 0.24983741343021393, "learning_rate": 6.020209714608355e-05, "loss": 0.0248, "step": 37030 }, { "grad_norm": 0.4406345784664154, "learning_rate": 6.018186272049331e-05, "loss": 0.0256, "step": 37040 }, { "grad_norm": 0.4699452519416809, "learning_rate": 6.01616265551012e-05, "loss": 0.0318, "step": 37050 }, { "grad_norm": 0.4326050579547882, "learning_rate": 6.014138865336503e-05, "loss": 0.0268, "step": 37060 }, { "grad_norm": 0.3744455873966217, "learning_rate": 6.0121149018742905e-05, "loss": 0.0275, "step": 37070 }, { "grad_norm": 0.33719491958618164, "learning_rate": 6.010090765469325e-05, "loss": 0.0304, "step": 37080 }, { "grad_norm": 0.23619921505451202, "learning_rate": 6.008066456467473e-05, "loss": 0.0244, "step": 37090 }, { "grad_norm": 0.25259485840797424, "learning_rate": 6.0060419752146335e-05, "loss": 0.0283, "step": 37100 }, { "grad_norm": 0.30665504932403564, "learning_rate": 6.0040173220567353e-05, "loss": 0.026, "step": 37110 }, { "grad_norm": 0.3392440378665924, "learning_rate": 6.001992497339737e-05, "loss": 0.0257, "step": 37120 }, { "grad_norm": 0.22543871402740479, "learning_rate": 5.999967501409626e-05, "loss": 0.0299, "step": 37130 }, { "grad_norm": 0.3250083029270172, "learning_rate": 5.997942334612418e-05, "loss": 0.0247, "step": 37140 }, { "grad_norm": 0.3407641053199768, "learning_rate": 5.995916997294158e-05, "loss": 0.0216, "step": 37150 }, { "grad_norm": 0.3146149516105652, "learning_rate": 5.9938914898009214e-05, "loss": 0.0281, "step": 37160 }, { "grad_norm": 0.30403828620910645, "learning_rate": 5.991865812478813e-05, "loss": 0.0216, "step": 37170 }, { "grad_norm": 0.443088173866272, "learning_rate": 5.989839965673964e-05, "loss": 0.0399, "step": 37180 }, { "grad_norm": 0.3291073143482208, "learning_rate": 5.987813949732539e-05, "loss": 0.0274, "step": 37190 }, { "grad_norm": 0.46256202459335327, "learning_rate": 5.9857877650007255e-05, "loss": 0.0277, "step": 37200 }, { "grad_norm": 0.2468244582414627, "learning_rate": 5.983761411824744e-05, "loss": 0.0245, "step": 37210 }, { "grad_norm": 0.3658480644226074, "learning_rate": 5.981734890550844e-05, "loss": 0.0303, "step": 37220 }, { "grad_norm": 0.2977968454360962, "learning_rate": 5.979708201525301e-05, "loss": 0.0338, "step": 37230 }, { "grad_norm": 0.35778501629829407, "learning_rate": 5.977681345094422e-05, "loss": 0.0333, "step": 37240 }, { "grad_norm": 0.31026318669319153, "learning_rate": 5.97565432160454e-05, "loss": 0.0295, "step": 37250 }, { "grad_norm": 0.26853054761886597, "learning_rate": 5.9736271314020186e-05, "loss": 0.0293, "step": 37260 }, { "grad_norm": 0.3803733289241791, "learning_rate": 5.971599774833251e-05, "loss": 0.0319, "step": 37270 }, { "grad_norm": 0.27343595027923584, "learning_rate": 5.9695722522446525e-05, "loss": 0.0349, "step": 37280 }, { "grad_norm": 0.35173484683036804, "learning_rate": 5.9675445639826765e-05, "loss": 0.0277, "step": 37290 }, { "grad_norm": 0.2564821243286133, "learning_rate": 5.965516710393796e-05, "loss": 0.0288, "step": 37300 }, { "grad_norm": 0.46713128685951233, "learning_rate": 5.963488691824516e-05, "loss": 0.0377, "step": 37310 }, { "grad_norm": 0.3038281202316284, "learning_rate": 5.96146050862137e-05, "loss": 0.0382, "step": 37320 }, { "grad_norm": 0.39362093806266785, "learning_rate": 5.959432161130919e-05, "loss": 0.0309, "step": 37330 }, { "grad_norm": 0.31374502182006836, "learning_rate": 5.9574036496997545e-05, "loss": 0.0316, "step": 37340 }, { "grad_norm": 0.4318157136440277, "learning_rate": 5.955374974674488e-05, "loss": 0.0313, "step": 37350 }, { "grad_norm": 0.3234250843524933, "learning_rate": 5.9533461364017696e-05, "loss": 0.0333, "step": 37360 }, { "grad_norm": 0.29276949167251587, "learning_rate": 5.9513171352282716e-05, "loss": 0.0304, "step": 37370 }, { "grad_norm": 0.2727774679660797, "learning_rate": 5.949287971500692e-05, "loss": 0.0278, "step": 37380 }, { "grad_norm": 0.2238992601633072, "learning_rate": 5.947258645565762e-05, "loss": 0.0266, "step": 37390 }, { "grad_norm": 0.17367321252822876, "learning_rate": 5.945229157770237e-05, "loss": 0.0254, "step": 37400 }, { "grad_norm": 0.27920982241630554, "learning_rate": 5.9431995084609006e-05, "loss": 0.0317, "step": 37410 }, { "grad_norm": 0.5342788696289062, "learning_rate": 5.941169697984564e-05, "loss": 0.0278, "step": 37420 }, { "grad_norm": 0.569366455078125, "learning_rate": 5.9391397266880675e-05, "loss": 0.0275, "step": 37430 }, { "grad_norm": 0.4302216172218323, "learning_rate": 5.937109594918279e-05, "loss": 0.0451, "step": 37440 }, { "grad_norm": 0.322499543428421, "learning_rate": 5.9350793030220884e-05, "loss": 0.0496, "step": 37450 }, { "grad_norm": 0.3731062412261963, "learning_rate": 5.933048851346421e-05, "loss": 0.0313, "step": 37460 }, { "grad_norm": 0.2822629511356354, "learning_rate": 5.931018240238224e-05, "loss": 0.0401, "step": 37470 }, { "grad_norm": 0.30723243951797485, "learning_rate": 5.928987470044471e-05, "loss": 0.0254, "step": 37480 }, { "grad_norm": 0.3098335266113281, "learning_rate": 5.9269565411121695e-05, "loss": 0.0263, "step": 37490 }, { "grad_norm": 0.361075758934021, "learning_rate": 5.924925453788347e-05, "loss": 0.0262, "step": 37500 }, { "grad_norm": 0.2642127275466919, "learning_rate": 5.92289420842006e-05, "loss": 0.0319, "step": 37510 }, { "grad_norm": 0.4070005714893341, "learning_rate": 5.9208628053543945e-05, "loss": 0.044, "step": 37520 }, { "grad_norm": 0.5890128016471863, "learning_rate": 5.918831244938462e-05, "loss": 0.0322, "step": 37530 }, { "grad_norm": 0.37369149923324585, "learning_rate": 5.9167995275194e-05, "loss": 0.0348, "step": 37540 }, { "grad_norm": 0.2708825170993805, "learning_rate": 5.914767653444373e-05, "loss": 0.0308, "step": 37550 }, { "grad_norm": 0.42929425835609436, "learning_rate": 5.912735623060572e-05, "loss": 0.0313, "step": 37560 }, { "grad_norm": 0.2861383557319641, "learning_rate": 5.910703436715217e-05, "loss": 0.0255, "step": 37570 }, { "grad_norm": 0.25392502546310425, "learning_rate": 5.908671094755552e-05, "loss": 0.0291, "step": 37580 }, { "grad_norm": 0.33898720145225525, "learning_rate": 5.906638597528851e-05, "loss": 0.0265, "step": 37590 }, { "grad_norm": 0.3480910360813141, "learning_rate": 5.9046059453824076e-05, "loss": 0.0281, "step": 37600 }, { "grad_norm": 0.2906879484653473, "learning_rate": 5.9025731386635505e-05, "loss": 0.0249, "step": 37610 }, { "grad_norm": 0.42605865001678467, "learning_rate": 5.900540177719629e-05, "loss": 0.0274, "step": 37620 }, { "grad_norm": 0.3492317795753479, "learning_rate": 5.898507062898021e-05, "loss": 0.0301, "step": 37630 }, { "grad_norm": 0.39334776997566223, "learning_rate": 5.8964737945461316e-05, "loss": 0.0261, "step": 37640 }, { "grad_norm": 0.2819443643093109, "learning_rate": 5.8944403730113885e-05, "loss": 0.0262, "step": 37650 }, { "grad_norm": 0.40716928243637085, "learning_rate": 5.892406798641248e-05, "loss": 0.0399, "step": 37660 }, { "grad_norm": 0.2855798006057739, "learning_rate": 5.890373071783193e-05, "loss": 0.0256, "step": 37670 }, { "grad_norm": 0.387432724237442, "learning_rate": 5.888339192784732e-05, "loss": 0.0324, "step": 37680 }, { "grad_norm": 0.4028191864490509, "learning_rate": 5.8863051619934003e-05, "loss": 0.0278, "step": 37690 }, { "grad_norm": 0.3677807152271271, "learning_rate": 5.8842709797567554e-05, "loss": 0.0253, "step": 37700 }, { "grad_norm": 0.3091357946395874, "learning_rate": 5.8822366464223855e-05, "loss": 0.0296, "step": 37710 }, { "grad_norm": 0.25543874502182007, "learning_rate": 5.880202162337901e-05, "loss": 0.035, "step": 37720 }, { "grad_norm": 0.4059304893016815, "learning_rate": 5.8781675278509405e-05, "loss": 0.0238, "step": 37730 }, { "grad_norm": 0.30719465017318726, "learning_rate": 5.8761327433091696e-05, "loss": 0.0267, "step": 37740 }, { "grad_norm": 0.4152972102165222, "learning_rate": 5.874097809060275e-05, "loss": 0.0363, "step": 37750 }, { "grad_norm": 0.36020392179489136, "learning_rate": 5.87206272545197e-05, "loss": 0.0452, "step": 37760 }, { "grad_norm": 0.2898147404193878, "learning_rate": 5.8700274928319955e-05, "loss": 0.0302, "step": 37770 }, { "grad_norm": 0.5596878528594971, "learning_rate": 5.867992111548118e-05, "loss": 0.0293, "step": 37780 }, { "grad_norm": 0.3190706670284271, "learning_rate": 5.865956581948131e-05, "loss": 0.025, "step": 37790 }, { "grad_norm": 0.23104733228683472, "learning_rate": 5.863920904379845e-05, "loss": 0.025, "step": 37800 }, { "grad_norm": 0.30541807413101196, "learning_rate": 5.8618850791911064e-05, "loss": 0.0303, "step": 37810 }, { "grad_norm": 0.286121129989624, "learning_rate": 5.859849106729779e-05, "loss": 0.0317, "step": 37820 }, { "grad_norm": 0.3457632064819336, "learning_rate": 5.857812987343758e-05, "loss": 0.0285, "step": 37830 }, { "grad_norm": 0.29501277208328247, "learning_rate": 5.855776721380957e-05, "loss": 0.0303, "step": 37840 }, { "grad_norm": 0.2726703882217407, "learning_rate": 5.8537403091893217e-05, "loss": 0.0292, "step": 37850 }, { "grad_norm": 0.33496421575546265, "learning_rate": 5.851703751116816e-05, "loss": 0.0337, "step": 37860 }, { "grad_norm": 0.2581999897956848, "learning_rate": 5.8496670475114336e-05, "loss": 0.0268, "step": 37870 }, { "grad_norm": 0.27832725644111633, "learning_rate": 5.84763019872119e-05, "loss": 0.0362, "step": 37880 }, { "grad_norm": 0.39235907793045044, "learning_rate": 5.845593205094131e-05, "loss": 0.0262, "step": 37890 }, { "grad_norm": 0.34074166417121887, "learning_rate": 5.843556066978318e-05, "loss": 0.0251, "step": 37900 }, { "grad_norm": 0.2863534092903137, "learning_rate": 5.8415187847218455e-05, "loss": 0.0268, "step": 37910 }, { "grad_norm": 0.22944094240665436, "learning_rate": 5.839481358672827e-05, "loss": 0.0237, "step": 37920 }, { "grad_norm": 0.350963294506073, "learning_rate": 5.837443789179407e-05, "loss": 0.0352, "step": 37930 }, { "grad_norm": 0.3079375922679901, "learning_rate": 5.8354060765897445e-05, "loss": 0.032, "step": 37940 }, { "grad_norm": 0.36971616744995117, "learning_rate": 5.8333682212520334e-05, "loss": 0.0394, "step": 37950 }, { "grad_norm": 0.31702759861946106, "learning_rate": 5.831330223514486e-05, "loss": 0.0315, "step": 37960 }, { "grad_norm": 0.2953803241252899, "learning_rate": 5.8292920837253396e-05, "loss": 0.0318, "step": 37970 }, { "grad_norm": 0.34483104944229126, "learning_rate": 5.827253802232857e-05, "loss": 0.0254, "step": 37980 }, { "grad_norm": 0.4207150936126709, "learning_rate": 5.825215379385327e-05, "loss": 0.0224, "step": 37990 }, { "grad_norm": 0.27706092596054077, "learning_rate": 5.823176815531057e-05, "loss": 0.0431, "step": 38000 }, { "grad_norm": 0.2661052644252777, "learning_rate": 5.8211381110183826e-05, "loss": 0.0359, "step": 38010 }, { "grad_norm": 0.32531100511550903, "learning_rate": 5.8190992661956645e-05, "loss": 0.0232, "step": 38020 }, { "grad_norm": 0.23010551929473877, "learning_rate": 5.817060281411284e-05, "loss": 0.0277, "step": 38030 }, { "grad_norm": 0.21503190696239471, "learning_rate": 5.815021157013647e-05, "loss": 0.0231, "step": 38040 }, { "grad_norm": 0.4049590826034546, "learning_rate": 5.8129818933511856e-05, "loss": 0.04, "step": 38050 }, { "grad_norm": 0.3963305950164795, "learning_rate": 5.8109424907723544e-05, "loss": 0.0282, "step": 38060 }, { "grad_norm": 0.29637762904167175, "learning_rate": 5.80890294962563e-05, "loss": 0.0246, "step": 38070 }, { "grad_norm": 0.3690702021121979, "learning_rate": 5.806863270259515e-05, "loss": 0.037, "step": 38080 }, { "grad_norm": 0.28135716915130615, "learning_rate": 5.804823453022536e-05, "loss": 0.0284, "step": 38090 }, { "grad_norm": 0.3562660813331604, "learning_rate": 5.80278349826324e-05, "loss": 0.0298, "step": 38100 }, { "grad_norm": 0.4184704124927521, "learning_rate": 5.8007434063302014e-05, "loss": 0.0281, "step": 38110 }, { "grad_norm": 0.3674294650554657, "learning_rate": 5.7987031775720136e-05, "loss": 0.0336, "step": 38120 }, { "grad_norm": 0.23318471014499664, "learning_rate": 5.7966628123372976e-05, "loss": 0.0349, "step": 38130 }, { "grad_norm": 0.42515259981155396, "learning_rate": 5.7946223109746956e-05, "loss": 0.0359, "step": 38140 }, { "grad_norm": 0.3612125813961029, "learning_rate": 5.7925816738328754e-05, "loss": 0.0275, "step": 38150 }, { "grad_norm": 0.39523637294769287, "learning_rate": 5.790540901260521e-05, "loss": 0.0263, "step": 38160 }, { "grad_norm": 0.2563115954399109, "learning_rate": 5.788499993606351e-05, "loss": 0.0291, "step": 38170 }, { "grad_norm": 0.2194790542125702, "learning_rate": 5.786458951219096e-05, "loss": 0.0288, "step": 38180 }, { "grad_norm": 0.2871936857700348, "learning_rate": 5.784417774447517e-05, "loss": 0.0373, "step": 38190 }, { "grad_norm": 0.3566823899745941, "learning_rate": 5.782376463640393e-05, "loss": 0.0312, "step": 38200 }, { "grad_norm": 0.35908186435699463, "learning_rate": 5.780335019146531e-05, "loss": 0.0274, "step": 38210 }, { "grad_norm": 0.2827844023704529, "learning_rate": 5.778293441314755e-05, "loss": 0.0294, "step": 38220 }, { "grad_norm": 0.3614049553871155, "learning_rate": 5.776251730493917e-05, "loss": 0.0357, "step": 38230 }, { "grad_norm": 0.3324192762374878, "learning_rate": 5.774209887032887e-05, "loss": 0.0272, "step": 38240 }, { "grad_norm": 0.2865576148033142, "learning_rate": 5.772167911280565e-05, "loss": 0.0265, "step": 38250 }, { "grad_norm": 0.28298380970954895, "learning_rate": 5.770125803585864e-05, "loss": 0.027, "step": 38260 }, { "grad_norm": 0.3759949207305908, "learning_rate": 5.768083564297726e-05, "loss": 0.0364, "step": 38270 }, { "grad_norm": 0.2943289279937744, "learning_rate": 5.766041193765114e-05, "loss": 0.0253, "step": 38280 }, { "grad_norm": 0.24801377952098846, "learning_rate": 5.763998692337015e-05, "loss": 0.0303, "step": 38290 }, { "grad_norm": 0.36219167709350586, "learning_rate": 5.761956060362433e-05, "loss": 0.0299, "step": 38300 }, { "grad_norm": 0.34961700439453125, "learning_rate": 5.7599132981904e-05, "loss": 0.0287, "step": 38310 }, { "grad_norm": 0.27892249822616577, "learning_rate": 5.75787040616997e-05, "loss": 0.0327, "step": 38320 }, { "grad_norm": 0.2872087061405182, "learning_rate": 5.755827384650212e-05, "loss": 0.0233, "step": 38330 }, { "grad_norm": 0.29571276903152466, "learning_rate": 5.753784233980228e-05, "loss": 0.028, "step": 38340 }, { "grad_norm": 0.3450956344604492, "learning_rate": 5.751740954509135e-05, "loss": 0.024, "step": 38350 }, { "grad_norm": 0.23924057185649872, "learning_rate": 5.7496975465860715e-05, "loss": 0.022, "step": 38360 }, { "grad_norm": 0.22632203996181488, "learning_rate": 5.747654010560202e-05, "loss": 0.0217, "step": 38370 }, { "grad_norm": 0.23680391907691956, "learning_rate": 5.7456103467807097e-05, "loss": 0.0291, "step": 38380 }, { "grad_norm": 0.38656193017959595, "learning_rate": 5.7435665555968046e-05, "loss": 0.0251, "step": 38390 }, { "grad_norm": 0.2051021158695221, "learning_rate": 5.74152263735771e-05, "loss": 0.0276, "step": 38400 }, { "grad_norm": 0.5042263865470886, "learning_rate": 5.739478592412677e-05, "loss": 0.0275, "step": 38410 }, { "grad_norm": 0.6130473017692566, "learning_rate": 5.7374344211109766e-05, "loss": 0.0411, "step": 38420 }, { "grad_norm": 0.38564950227737427, "learning_rate": 5.735390123801904e-05, "loss": 0.0291, "step": 38430 }, { "grad_norm": 0.44409996271133423, "learning_rate": 5.7333457008347704e-05, "loss": 0.0397, "step": 38440 }, { "grad_norm": 0.4027993679046631, "learning_rate": 5.7313011525589156e-05, "loss": 0.033, "step": 38450 }, { "grad_norm": 0.28675442934036255, "learning_rate": 5.729256479323694e-05, "loss": 0.0254, "step": 38460 }, { "grad_norm": 0.32879874110221863, "learning_rate": 5.727211681478485e-05, "loss": 0.0298, "step": 38470 }, { "grad_norm": 0.35876893997192383, "learning_rate": 5.7251667593726886e-05, "loss": 0.0333, "step": 38480 }, { "grad_norm": 0.28813469409942627, "learning_rate": 5.723121713355728e-05, "loss": 0.0433, "step": 38490 }, { "grad_norm": 0.23477233946323395, "learning_rate": 5.721076543777044e-05, "loss": 0.03, "step": 38500 }, { "grad_norm": 0.3066933751106262, "learning_rate": 5.7190312509860986e-05, "loss": 0.0242, "step": 38510 }, { "grad_norm": 0.22220341861248016, "learning_rate": 5.716985835332379e-05, "loss": 0.0333, "step": 38520 }, { "grad_norm": 0.21105845272541046, "learning_rate": 5.714940297165389e-05, "loss": 0.0243, "step": 38530 }, { "grad_norm": 0.2681308388710022, "learning_rate": 5.712894636834656e-05, "loss": 0.0282, "step": 38540 }, { "grad_norm": 0.3376646339893341, "learning_rate": 5.7108488546897276e-05, "loss": 0.0311, "step": 38550 }, { "grad_norm": 0.24255062639713287, "learning_rate": 5.708802951080172e-05, "loss": 0.0242, "step": 38560 }, { "grad_norm": 0.31043317914009094, "learning_rate": 5.706756926355576e-05, "loss": 0.035, "step": 38570 }, { "grad_norm": 0.3319458067417145, "learning_rate": 5.704710780865554e-05, "loss": 0.0315, "step": 38580 }, { "grad_norm": 0.31584978103637695, "learning_rate": 5.7026645149597325e-05, "loss": 0.0291, "step": 38590 }, { "grad_norm": 0.5672577619552612, "learning_rate": 5.700618128987764e-05, "loss": 0.036, "step": 38600 }, { "grad_norm": 0.2309836894273758, "learning_rate": 5.698571623299317e-05, "loss": 0.0246, "step": 38610 }, { "grad_norm": 0.19002081453800201, "learning_rate": 5.696524998244086e-05, "loss": 0.0239, "step": 38620 }, { "grad_norm": 0.2998516261577606, "learning_rate": 5.6944782541717836e-05, "loss": 0.0333, "step": 38630 }, { "grad_norm": 0.29089510440826416, "learning_rate": 5.69243139143214e-05, "loss": 0.028, "step": 38640 }, { "grad_norm": 0.27219322323799133, "learning_rate": 5.6903844103749125e-05, "loss": 0.0358, "step": 38650 }, { "grad_norm": 0.22437472641468048, "learning_rate": 5.688337311349869e-05, "loss": 0.0206, "step": 38660 }, { "grad_norm": 0.27760833501815796, "learning_rate": 5.6862900947068074e-05, "loss": 0.0292, "step": 38670 }, { "grad_norm": 0.43002018332481384, "learning_rate": 5.6842427607955374e-05, "loss": 0.033, "step": 38680 }, { "grad_norm": 0.21988777816295624, "learning_rate": 5.682195309965893e-05, "loss": 0.0267, "step": 38690 }, { "grad_norm": 0.3045696020126343, "learning_rate": 5.6801477425677294e-05, "loss": 0.0314, "step": 38700 }, { "grad_norm": 0.3595489263534546, "learning_rate": 5.678100058950917e-05, "loss": 0.0302, "step": 38710 }, { "grad_norm": 0.38589954376220703, "learning_rate": 5.676052259465352e-05, "loss": 0.0303, "step": 38720 }, { "grad_norm": 0.18412871658802032, "learning_rate": 5.674004344460945e-05, "loss": 0.0256, "step": 38730 }, { "grad_norm": 0.28587859869003296, "learning_rate": 5.6719563142876295e-05, "loss": 0.0258, "step": 38740 }, { "grad_norm": 0.34680885076522827, "learning_rate": 5.669908169295359e-05, "loss": 0.0258, "step": 38750 }, { "grad_norm": 0.38012826442718506, "learning_rate": 5.667859909834105e-05, "loss": 0.0293, "step": 38760 }, { "grad_norm": 0.3115202486515045, "learning_rate": 5.6658115362538565e-05, "loss": 0.0309, "step": 38770 }, { "grad_norm": 0.38131386041641235, "learning_rate": 5.663763048904628e-05, "loss": 0.0326, "step": 38780 }, { "grad_norm": 0.4297114908695221, "learning_rate": 5.661714448136447e-05, "loss": 0.0242, "step": 38790 }, { "grad_norm": 0.29620784521102905, "learning_rate": 5.659665734299366e-05, "loss": 0.0223, "step": 38800 }, { "grad_norm": 0.37351876497268677, "learning_rate": 5.6576169077434516e-05, "loss": 0.0316, "step": 38810 }, { "grad_norm": 0.5006489157676697, "learning_rate": 5.6555679688187944e-05, "loss": 0.0288, "step": 38820 }, { "grad_norm": 0.47792744636535645, "learning_rate": 5.6535189178755e-05, "loss": 0.0316, "step": 38830 }, { "grad_norm": 0.2945295572280884, "learning_rate": 5.651469755263698e-05, "loss": 0.0286, "step": 38840 }, { "grad_norm": 0.2179105579853058, "learning_rate": 5.6494204813335316e-05, "loss": 0.0345, "step": 38850 }, { "grad_norm": 0.3670286536216736, "learning_rate": 5.647371096435168e-05, "loss": 0.0308, "step": 38860 }, { "grad_norm": 0.2914901375770569, "learning_rate": 5.645321600918788e-05, "loss": 0.0251, "step": 38870 }, { "grad_norm": 0.28924593329429626, "learning_rate": 5.643271995134597e-05, "loss": 0.031, "step": 38880 }, { "grad_norm": 0.27792441844940186, "learning_rate": 5.641222279432814e-05, "loss": 0.0239, "step": 38890 }, { "grad_norm": 0.3525048792362213, "learning_rate": 5.6391724541636834e-05, "loss": 0.03, "step": 38900 }, { "grad_norm": 0.43191730976104736, "learning_rate": 5.6371225196774605e-05, "loss": 0.0291, "step": 38910 }, { "grad_norm": 0.3631560504436493, "learning_rate": 5.635072476324423e-05, "loss": 0.0306, "step": 38920 }, { "grad_norm": 0.3827269673347473, "learning_rate": 5.63302232445487e-05, "loss": 0.0265, "step": 38930 }, { "grad_norm": 0.29403066635131836, "learning_rate": 5.6309720644191144e-05, "loss": 0.0309, "step": 38940 }, { "grad_norm": 0.2724177837371826, "learning_rate": 5.628921696567491e-05, "loss": 0.0235, "step": 38950 }, { "grad_norm": 0.2648388147354126, "learning_rate": 5.62687122125035e-05, "loss": 0.0224, "step": 38960 }, { "grad_norm": 0.32511547207832336, "learning_rate": 5.624820638818062e-05, "loss": 0.0232, "step": 38970 }, { "grad_norm": 0.27796801924705505, "learning_rate": 5.6227699496210164e-05, "loss": 0.0336, "step": 38980 }, { "grad_norm": 0.23568597435951233, "learning_rate": 5.6207191540096195e-05, "loss": 0.0243, "step": 38990 }, { "grad_norm": 0.2854265570640564, "learning_rate": 5.618668252334296e-05, "loss": 0.0238, "step": 39000 }, { "grad_norm": 0.37558814883232117, "learning_rate": 5.616617244945488e-05, "loss": 0.0294, "step": 39010 }, { "grad_norm": 0.30270466208457947, "learning_rate": 5.614566132193656e-05, "loss": 0.0266, "step": 39020 }, { "grad_norm": 0.2675607204437256, "learning_rate": 5.612514914429282e-05, "loss": 0.0286, "step": 39030 }, { "grad_norm": 0.388992577791214, "learning_rate": 5.610463592002863e-05, "loss": 0.0248, "step": 39040 }, { "grad_norm": 0.3554891347885132, "learning_rate": 5.608412165264909e-05, "loss": 0.0305, "step": 39050 }, { "grad_norm": 0.2867237627506256, "learning_rate": 5.606360634565959e-05, "loss": 0.0201, "step": 39060 }, { "grad_norm": 0.27642226219177246, "learning_rate": 5.604309000256559e-05, "loss": 0.0246, "step": 39070 }, { "grad_norm": 0.2667423486709595, "learning_rate": 5.6022572626872785e-05, "loss": 0.0247, "step": 39080 }, { "grad_norm": 0.32223203778266907, "learning_rate": 5.600205422208704e-05, "loss": 0.0277, "step": 39090 }, { "grad_norm": 0.35541123151779175, "learning_rate": 5.59815347917144e-05, "loss": 0.0315, "step": 39100 }, { "grad_norm": 0.3857019245624542, "learning_rate": 5.596101433926103e-05, "loss": 0.0301, "step": 39110 }, { "grad_norm": 0.36876654624938965, "learning_rate": 5.5940492868233364e-05, "loss": 0.0226, "step": 39120 }, { "grad_norm": 0.31946951150894165, "learning_rate": 5.591997038213793e-05, "loss": 0.0272, "step": 39130 }, { "grad_norm": 0.3076338469982147, "learning_rate": 5.5899446884481475e-05, "loss": 0.0223, "step": 39140 }, { "grad_norm": 0.35329470038414, "learning_rate": 5.5878922378770906e-05, "loss": 0.0429, "step": 39150 }, { "grad_norm": 0.2528274655342102, "learning_rate": 5.5858396868513285e-05, "loss": 0.026, "step": 39160 }, { "grad_norm": 0.3058774471282959, "learning_rate": 5.583787035721586e-05, "loss": 0.025, "step": 39170 }, { "grad_norm": 0.22937451303005219, "learning_rate": 5.581734284838606e-05, "loss": 0.024, "step": 39180 }, { "grad_norm": 0.35726046562194824, "learning_rate": 5.579681434553147e-05, "loss": 0.0273, "step": 39190 }, { "grad_norm": 0.2973663806915283, "learning_rate": 5.5776284852159854e-05, "loss": 0.0226, "step": 39200 }, { "grad_norm": 0.2692084014415741, "learning_rate": 5.575575437177913e-05, "loss": 0.0307, "step": 39210 }, { "grad_norm": 0.3258095979690552, "learning_rate": 5.573522290789742e-05, "loss": 0.027, "step": 39220 }, { "grad_norm": 0.25515809655189514, "learning_rate": 5.571469046402298e-05, "loss": 0.0344, "step": 39230 }, { "grad_norm": 0.359831303358078, "learning_rate": 5.5694157043664205e-05, "loss": 0.0302, "step": 39240 }, { "grad_norm": 0.29494911432266235, "learning_rate": 5.567362265032975e-05, "loss": 0.0217, "step": 39250 }, { "grad_norm": 0.2607616186141968, "learning_rate": 5.565308728752836e-05, "loss": 0.0245, "step": 39260 }, { "grad_norm": 0.32055607438087463, "learning_rate": 5.5632550958768956e-05, "loss": 0.0263, "step": 39270 }, { "grad_norm": 0.3491227626800537, "learning_rate": 5.5612013667560636e-05, "loss": 0.0312, "step": 39280 }, { "grad_norm": 0.334012508392334, "learning_rate": 5.5591475417412676e-05, "loss": 0.0433, "step": 39290 }, { "grad_norm": 0.26741525530815125, "learning_rate": 5.557093621183451e-05, "loss": 0.0278, "step": 39300 }, { "grad_norm": 0.31935247778892517, "learning_rate": 5.55503960543357e-05, "loss": 0.0263, "step": 39310 }, { "grad_norm": 0.32356345653533936, "learning_rate": 5.552985494842601e-05, "loss": 0.0241, "step": 39320 }, { "grad_norm": 0.16334974765777588, "learning_rate": 5.550931289761534e-05, "loss": 0.0244, "step": 39330 }, { "grad_norm": 0.3008478581905365, "learning_rate": 5.548876990541378e-05, "loss": 0.0271, "step": 39340 }, { "grad_norm": 0.4935770332813263, "learning_rate": 5.5468225975331564e-05, "loss": 0.0312, "step": 39350 }, { "grad_norm": 0.4123554229736328, "learning_rate": 5.544768111087909e-05, "loss": 0.0274, "step": 39360 }, { "grad_norm": 0.2767738401889801, "learning_rate": 5.54271353155669e-05, "loss": 0.0312, "step": 39370 }, { "grad_norm": 0.20081104338169098, "learning_rate": 5.5406588592905715e-05, "loss": 0.0242, "step": 39380 }, { "grad_norm": 0.4529109597206116, "learning_rate": 5.5386040946406416e-05, "loss": 0.0329, "step": 39390 }, { "grad_norm": 0.4195348620414734, "learning_rate": 5.536549237958004e-05, "loss": 0.038, "step": 39400 }, { "grad_norm": 0.3573198616504669, "learning_rate": 5.5344942895937744e-05, "loss": 0.0338, "step": 39410 }, { "grad_norm": 0.2776550054550171, "learning_rate": 5.5324392498990904e-05, "loss": 0.0263, "step": 39420 }, { "grad_norm": 0.3894125819206238, "learning_rate": 5.5303841192251e-05, "loss": 0.0251, "step": 39430 }, { "grad_norm": 0.29004910588264465, "learning_rate": 5.52832889792297e-05, "loss": 0.0306, "step": 39440 }, { "grad_norm": 0.32309895753860474, "learning_rate": 5.526273586343881e-05, "loss": 0.0205, "step": 39450 }, { "grad_norm": 0.33577442169189453, "learning_rate": 5.5242181848390306e-05, "loss": 0.0237, "step": 39460 }, { "grad_norm": 0.36894479393959045, "learning_rate": 5.5221626937596285e-05, "loss": 0.0217, "step": 39470 }, { "grad_norm": 0.23166438937187195, "learning_rate": 5.520107113456903e-05, "loss": 0.0289, "step": 39480 }, { "grad_norm": 0.3740004599094391, "learning_rate": 5.5180514442820974e-05, "loss": 0.0403, "step": 39490 }, { "grad_norm": 0.2186223864555359, "learning_rate": 5.515995686586469e-05, "loss": 0.0203, "step": 39500 }, { "grad_norm": 0.3332778513431549, "learning_rate": 5.5139398407212916e-05, "loss": 0.0242, "step": 39510 }, { "grad_norm": 0.4996870160102844, "learning_rate": 5.511883907037849e-05, "loss": 0.0258, "step": 39520 }, { "grad_norm": 0.31508204340934753, "learning_rate": 5.509827885887449e-05, "loss": 0.0265, "step": 39530 }, { "grad_norm": 0.4692125618457794, "learning_rate": 5.507771777621406e-05, "loss": 0.0253, "step": 39540 }, { "grad_norm": 0.26056990027427673, "learning_rate": 5.505715582591052e-05, "loss": 0.0344, "step": 39550 }, { "grad_norm": 0.40180063247680664, "learning_rate": 5.50365930114774e-05, "loss": 0.0316, "step": 39560 }, { "grad_norm": 0.23673738539218903, "learning_rate": 5.5016029336428255e-05, "loss": 0.0218, "step": 39570 }, { "grad_norm": 0.3198263347148895, "learning_rate": 5.49954648042769e-05, "loss": 0.0299, "step": 39580 }, { "grad_norm": 0.30453750491142273, "learning_rate": 5.4974899418537226e-05, "loss": 0.0388, "step": 39590 }, { "grad_norm": 0.24750006198883057, "learning_rate": 5.4954333182723316e-05, "loss": 0.0285, "step": 39600 }, { "grad_norm": 0.3348338007926941, "learning_rate": 5.493376610034937e-05, "loss": 0.0233, "step": 39610 }, { "grad_norm": 0.3620821535587311, "learning_rate": 5.4913198174929735e-05, "loss": 0.0249, "step": 39620 }, { "grad_norm": 0.23702438175678253, "learning_rate": 5.489262940997889e-05, "loss": 0.0272, "step": 39630 }, { "grad_norm": 0.29045891761779785, "learning_rate": 5.487205980901151e-05, "loss": 0.0361, "step": 39640 }, { "grad_norm": 0.3193206787109375, "learning_rate": 5.485148937554234e-05, "loss": 0.0435, "step": 39650 }, { "grad_norm": 0.2698415219783783, "learning_rate": 5.483091811308635e-05, "loss": 0.0225, "step": 39660 }, { "grad_norm": 0.26875796914100647, "learning_rate": 5.4810346025158564e-05, "loss": 0.0244, "step": 39670 }, { "grad_norm": 0.3143197000026703, "learning_rate": 5.478977311527421e-05, "loss": 0.0253, "step": 39680 }, { "grad_norm": 0.2441483438014984, "learning_rate": 5.476919938694863e-05, "loss": 0.0228, "step": 39690 }, { "grad_norm": 0.2173449546098709, "learning_rate": 5.474862484369733e-05, "loss": 0.0169, "step": 39700 }, { "grad_norm": 0.36282041668891907, "learning_rate": 5.472804948903589e-05, "loss": 0.0275, "step": 39710 }, { "grad_norm": 0.27141672372817993, "learning_rate": 5.470747332648013e-05, "loss": 0.0206, "step": 39720 }, { "grad_norm": 0.40244460105895996, "learning_rate": 5.468689635954591e-05, "loss": 0.0355, "step": 39730 }, { "grad_norm": 0.31546497344970703, "learning_rate": 5.46663185917493e-05, "loss": 0.0264, "step": 39740 }, { "grad_norm": 0.37021663784980774, "learning_rate": 5.464574002660645e-05, "loss": 0.0321, "step": 39750 }, { "grad_norm": 0.35373732447624207, "learning_rate": 5.4625160667633724e-05, "loss": 0.0295, "step": 39760 }, { "grad_norm": 0.20010091364383698, "learning_rate": 5.4604580518347505e-05, "loss": 0.0252, "step": 39770 }, { "grad_norm": 0.3869308531284332, "learning_rate": 5.458399958226442e-05, "loss": 0.0265, "step": 39780 }, { "grad_norm": 0.23916202783584595, "learning_rate": 5.456341786290119e-05, "loss": 0.0287, "step": 39790 }, { "grad_norm": 0.3972839415073395, "learning_rate": 5.454283536377465e-05, "loss": 0.0299, "step": 39800 }, { "grad_norm": 0.33316275477409363, "learning_rate": 5.452225208840179e-05, "loss": 0.026, "step": 39810 }, { "grad_norm": 0.43414056301116943, "learning_rate": 5.450166804029975e-05, "loss": 0.035, "step": 39820 }, { "grad_norm": 0.29144152998924255, "learning_rate": 5.448108322298574e-05, "loss": 0.0243, "step": 39830 }, { "grad_norm": 0.2606659531593323, "learning_rate": 5.446049763997717e-05, "loss": 0.0261, "step": 39840 }, { "grad_norm": 0.3239496350288391, "learning_rate": 5.4439911294791546e-05, "loss": 0.0223, "step": 39850 }, { "grad_norm": 0.36482277512550354, "learning_rate": 5.441932419094652e-05, "loss": 0.0247, "step": 39860 }, { "grad_norm": 0.34644320607185364, "learning_rate": 5.439873633195985e-05, "loss": 0.026, "step": 39870 }, { "grad_norm": 0.18894532322883606, "learning_rate": 5.437814772134947e-05, "loss": 0.0305, "step": 39880 }, { "grad_norm": 0.4550608694553375, "learning_rate": 5.4357558362633366e-05, "loss": 0.0288, "step": 39890 }, { "grad_norm": 0.2675139605998993, "learning_rate": 5.4336968259329726e-05, "loss": 0.0229, "step": 39900 }, { "grad_norm": 0.22283755242824554, "learning_rate": 5.431637741495681e-05, "loss": 0.0276, "step": 39910 }, { "grad_norm": 0.38823047280311584, "learning_rate": 5.429578583303307e-05, "loss": 0.0313, "step": 39920 }, { "grad_norm": 0.26733019948005676, "learning_rate": 5.427519351707701e-05, "loss": 0.032, "step": 39930 }, { "grad_norm": 0.2690189480781555, "learning_rate": 5.42546004706073e-05, "loss": 0.0265, "step": 39940 }, { "grad_norm": 0.4105515480041504, "learning_rate": 5.4234006697142735e-05, "loss": 0.0265, "step": 39950 }, { "grad_norm": 0.21326260268688202, "learning_rate": 5.421341220020224e-05, "loss": 0.0294, "step": 39960 }, { "grad_norm": 0.21838535368442535, "learning_rate": 5.419281698330482e-05, "loss": 0.03, "step": 39970 }, { "grad_norm": 0.22344394028186798, "learning_rate": 5.4172221049969665e-05, "loss": 0.0264, "step": 39980 }, { "grad_norm": 0.29230213165283203, "learning_rate": 5.415162440371604e-05, "loss": 0.0243, "step": 39990 }, { "grad_norm": 0.3356602191925049, "learning_rate": 5.413102704806334e-05, "loss": 0.0331, "step": 40000 }, { "grad_norm": 0.3163367211818695, "learning_rate": 5.41104289865311e-05, "loss": 0.0289, "step": 40010 }, { "grad_norm": 0.2966891825199127, "learning_rate": 5.408983022263898e-05, "loss": 0.0243, "step": 40020 }, { "grad_norm": 0.2818072438240051, "learning_rate": 5.406923075990673e-05, "loss": 0.0287, "step": 40030 }, { "grad_norm": 0.22710372507572174, "learning_rate": 5.404863060185423e-05, "loss": 0.0269, "step": 40040 }, { "grad_norm": 0.26511675119400024, "learning_rate": 5.40280297520015e-05, "loss": 0.0305, "step": 40050 }, { "grad_norm": 0.24942854046821594, "learning_rate": 5.400742821386865e-05, "loss": 0.0255, "step": 40060 }, { "grad_norm": 0.27737054228782654, "learning_rate": 5.398682599097592e-05, "loss": 0.0317, "step": 40070 }, { "grad_norm": 0.2813319265842438, "learning_rate": 5.396622308684367e-05, "loss": 0.0251, "step": 40080 }, { "grad_norm": 0.3438456058502197, "learning_rate": 5.394561950499236e-05, "loss": 0.029, "step": 40090 }, { "grad_norm": 0.44765305519104004, "learning_rate": 5.39250152489426e-05, "loss": 0.0285, "step": 40100 }, { "grad_norm": 0.8602872490882874, "learning_rate": 5.390441032221507e-05, "loss": 0.0265, "step": 40110 }, { "grad_norm": 0.35779696702957153, "learning_rate": 5.388380472833062e-05, "loss": 0.0272, "step": 40120 }, { "grad_norm": 0.31392890214920044, "learning_rate": 5.386319847081016e-05, "loss": 0.0291, "step": 40130 }, { "grad_norm": 0.4779713749885559, "learning_rate": 5.384259155317473e-05, "loss": 0.0305, "step": 40140 }, { "grad_norm": 0.2982940971851349, "learning_rate": 5.38219839789455e-05, "loss": 0.0313, "step": 40150 }, { "grad_norm": 0.34877610206604004, "learning_rate": 5.380137575164376e-05, "loss": 0.0326, "step": 40160 }, { "grad_norm": 0.3695119023323059, "learning_rate": 5.378076687479085e-05, "loss": 0.0277, "step": 40170 }, { "grad_norm": 0.2793993353843689, "learning_rate": 5.3760157351908305e-05, "loss": 0.027, "step": 40180 }, { "grad_norm": 0.40117281675338745, "learning_rate": 5.373954718651768e-05, "loss": 0.0297, "step": 40190 }, { "grad_norm": 0.2754353880882263, "learning_rate": 5.371893638214074e-05, "loss": 0.0229, "step": 40200 }, { "grad_norm": 0.4931115210056305, "learning_rate": 5.369832494229927e-05, "loss": 0.0296, "step": 40210 }, { "grad_norm": 0.3924010694026947, "learning_rate": 5.367771287051524e-05, "loss": 0.0331, "step": 40220 }, { "grad_norm": 0.34667161107063293, "learning_rate": 5.365710017031065e-05, "loss": 0.0292, "step": 40230 }, { "grad_norm": 0.28955143690109253, "learning_rate": 5.3636486845207654e-05, "loss": 0.0291, "step": 40240 }, { "grad_norm": 0.32101649045944214, "learning_rate": 5.3615872898728526e-05, "loss": 0.0246, "step": 40250 }, { "grad_norm": 0.3862840533256531, "learning_rate": 5.3595258334395614e-05, "loss": 0.0282, "step": 40260 }, { "grad_norm": 0.2968849837779999, "learning_rate": 5.3574643155731396e-05, "loss": 0.0299, "step": 40270 }, { "grad_norm": 0.3046218156814575, "learning_rate": 5.3554027366258405e-05, "loss": 0.0248, "step": 40280 }, { "grad_norm": 0.3165713846683502, "learning_rate": 5.3533410969499355e-05, "loss": 0.0285, "step": 40290 }, { "grad_norm": 0.30201780796051025, "learning_rate": 5.3512793968977e-05, "loss": 0.0248, "step": 40300 }, { "grad_norm": 0.2998168468475342, "learning_rate": 5.349217636821423e-05, "loss": 0.0252, "step": 40310 }, { "grad_norm": 0.20138679444789886, "learning_rate": 5.347155817073405e-05, "loss": 0.0196, "step": 40320 }, { "grad_norm": 0.28320443630218506, "learning_rate": 5.34509393800595e-05, "loss": 0.0207, "step": 40330 }, { "grad_norm": 0.40541964769363403, "learning_rate": 5.343031999971381e-05, "loss": 0.0265, "step": 40340 }, { "grad_norm": 0.2613144814968109, "learning_rate": 5.340970003322026e-05, "loss": 0.0243, "step": 40350 }, { "grad_norm": 0.26426374912261963, "learning_rate": 5.33890794841022e-05, "loss": 0.0246, "step": 40360 }, { "grad_norm": 0.26679304242134094, "learning_rate": 5.336845835588318e-05, "loss": 0.0248, "step": 40370 }, { "grad_norm": 0.32943326234817505, "learning_rate": 5.334783665208674e-05, "loss": 0.0252, "step": 40380 }, { "grad_norm": 0.2644769549369812, "learning_rate": 5.332721437623657e-05, "loss": 0.0228, "step": 40390 }, { "grad_norm": 0.3195379376411438, "learning_rate": 5.3306591531856464e-05, "loss": 0.0242, "step": 40400 }, { "grad_norm": 0.2211904078722, "learning_rate": 5.3285968122470295e-05, "loss": 0.0191, "step": 40410 }, { "grad_norm": 0.26282766461372375, "learning_rate": 5.326534415160207e-05, "loss": 0.0279, "step": 40420 }, { "grad_norm": 0.19748948514461517, "learning_rate": 5.324471962277582e-05, "loss": 0.0272, "step": 40430 }, { "grad_norm": 0.5327901244163513, "learning_rate": 5.3224094539515746e-05, "loss": 0.0286, "step": 40440 }, { "grad_norm": 0.26615235209465027, "learning_rate": 5.3203468905346076e-05, "loss": 0.0231, "step": 40450 }, { "grad_norm": 0.27748245000839233, "learning_rate": 5.3182842723791195e-05, "loss": 0.0282, "step": 40460 }, { "grad_norm": 0.2689853012561798, "learning_rate": 5.316221599837554e-05, "loss": 0.0192, "step": 40470 }, { "grad_norm": 0.3880312144756317, "learning_rate": 5.314158873262366e-05, "loss": 0.0311, "step": 40480 }, { "grad_norm": 0.20982439815998077, "learning_rate": 5.312096093006018e-05, "loss": 0.0189, "step": 40490 }, { "grad_norm": 0.27616044878959656, "learning_rate": 5.3100332594209846e-05, "loss": 0.0242, "step": 40500 }, { "grad_norm": 0.32138675451278687, "learning_rate": 5.3079703728597454e-05, "loss": 0.0264, "step": 40510 }, { "grad_norm": 0.5618691444396973, "learning_rate": 5.305907433674794e-05, "loss": 0.0251, "step": 40520 }, { "grad_norm": 0.26559334993362427, "learning_rate": 5.3038444422186284e-05, "loss": 0.023, "step": 40530 }, { "grad_norm": 0.268612265586853, "learning_rate": 5.301781398843757e-05, "loss": 0.0227, "step": 40540 }, { "grad_norm": 0.2532086670398712, "learning_rate": 5.299718303902699e-05, "loss": 0.0258, "step": 40550 }, { "grad_norm": 0.21364736557006836, "learning_rate": 5.297655157747979e-05, "loss": 0.0277, "step": 40560 }, { "grad_norm": 0.2553573250770569, "learning_rate": 5.295591960732136e-05, "loss": 0.0221, "step": 40570 }, { "grad_norm": 0.2785101532936096, "learning_rate": 5.293528713207708e-05, "loss": 0.0248, "step": 40580 }, { "grad_norm": 0.23678845167160034, "learning_rate": 5.291465415527253e-05, "loss": 0.0245, "step": 40590 }, { "grad_norm": 0.31405743956565857, "learning_rate": 5.2894020680433296e-05, "loss": 0.024, "step": 40600 }, { "grad_norm": 0.27629488706588745, "learning_rate": 5.287338671108507e-05, "loss": 0.0232, "step": 40610 }, { "grad_norm": 0.36777257919311523, "learning_rate": 5.285275225075367e-05, "loss": 0.0214, "step": 40620 }, { "grad_norm": 0.3295373022556305, "learning_rate": 5.283211730296492e-05, "loss": 0.0261, "step": 40630 }, { "grad_norm": 0.20547597110271454, "learning_rate": 5.281148187124477e-05, "loss": 0.0267, "step": 40640 }, { "grad_norm": 0.28332075476646423, "learning_rate": 5.279084595911927e-05, "loss": 0.0233, "step": 40650 }, { "grad_norm": 0.33721160888671875, "learning_rate": 5.277020957011453e-05, "loss": 0.0346, "step": 40660 }, { "grad_norm": 0.4705507755279541, "learning_rate": 5.274957270775673e-05, "loss": 0.0313, "step": 40670 }, { "grad_norm": 0.40763282775878906, "learning_rate": 5.2728935375572164e-05, "loss": 0.0211, "step": 40680 }, { "grad_norm": 0.2851496934890747, "learning_rate": 5.2708297577087176e-05, "loss": 0.0256, "step": 40690 }, { "grad_norm": 0.3773028552532196, "learning_rate": 5.2687659315828196e-05, "loss": 0.0288, "step": 40700 }, { "grad_norm": 0.2298918515443802, "learning_rate": 5.266702059532175e-05, "loss": 0.0247, "step": 40710 }, { "grad_norm": 0.360293447971344, "learning_rate": 5.264638141909444e-05, "loss": 0.0311, "step": 40720 }, { "grad_norm": 0.2655559480190277, "learning_rate": 5.2625741790672925e-05, "loss": 0.026, "step": 40730 }, { "grad_norm": 0.27665582299232483, "learning_rate": 5.2605101713583935e-05, "loss": 0.0265, "step": 40740 }, { "grad_norm": 0.3512047529220581, "learning_rate": 5.2584461191354315e-05, "loss": 0.0337, "step": 40750 }, { "grad_norm": 0.19443970918655396, "learning_rate": 5.2563820227510964e-05, "loss": 0.0246, "step": 40760 }, { "grad_norm": 0.3595316708087921, "learning_rate": 5.2543178825580864e-05, "loss": 0.0267, "step": 40770 }, { "grad_norm": 0.3679657578468323, "learning_rate": 5.2522536989091054e-05, "loss": 0.0322, "step": 40780 }, { "grad_norm": 0.406604528427124, "learning_rate": 5.2501894721568655e-05, "loss": 0.0248, "step": 40790 }, { "grad_norm": 0.2831105887889862, "learning_rate": 5.248125202654089e-05, "loss": 0.0241, "step": 40800 }, { "grad_norm": 0.3204568326473236, "learning_rate": 5.246060890753501e-05, "loss": 0.0374, "step": 40810 }, { "grad_norm": 0.2281641811132431, "learning_rate": 5.243996536807837e-05, "loss": 0.0248, "step": 40820 }, { "grad_norm": 0.3221447765827179, "learning_rate": 5.2419321411698384e-05, "loss": 0.0248, "step": 40830 }, { "grad_norm": 0.31819948554039, "learning_rate": 5.239867704192253e-05, "loss": 0.0282, "step": 40840 }, { "grad_norm": 0.21809948980808258, "learning_rate": 5.237803226227838e-05, "loss": 0.0206, "step": 40850 }, { "grad_norm": 0.4487459659576416, "learning_rate": 5.235738707629354e-05, "loss": 0.0282, "step": 40860 }, { "grad_norm": 0.2247646003961563, "learning_rate": 5.233674148749575e-05, "loss": 0.0288, "step": 40870 }, { "grad_norm": 0.4502328038215637, "learning_rate": 5.231609549941272e-05, "loss": 0.0297, "step": 40880 }, { "grad_norm": 0.3325689733028412, "learning_rate": 5.229544911557233e-05, "loss": 0.0268, "step": 40890 }, { "grad_norm": 0.3086658716201782, "learning_rate": 5.227480233950246e-05, "loss": 0.0299, "step": 40900 }, { "grad_norm": 0.3690568804740906, "learning_rate": 5.22541551747311e-05, "loss": 0.0322, "step": 40910 }, { "grad_norm": 0.3093624711036682, "learning_rate": 5.223350762478626e-05, "loss": 0.0243, "step": 40920 }, { "grad_norm": 0.30424055457115173, "learning_rate": 5.221285969319606e-05, "loss": 0.0266, "step": 40930 }, { "grad_norm": 0.31010958552360535, "learning_rate": 5.219221138348865e-05, "loss": 0.0314, "step": 40940 }, { "grad_norm": 0.6625386476516724, "learning_rate": 5.217156269919228e-05, "loss": 0.0304, "step": 40950 }, { "grad_norm": 0.1820317804813385, "learning_rate": 5.215091364383523e-05, "loss": 0.0252, "step": 40960 }, { "grad_norm": 0.38464558124542236, "learning_rate": 5.213026422094588e-05, "loss": 0.0327, "step": 40970 }, { "grad_norm": 0.36512187123298645, "learning_rate": 5.210961443405262e-05, "loss": 0.0251, "step": 40980 }, { "grad_norm": 0.34901902079582214, "learning_rate": 5.208896428668396e-05, "loss": 0.0269, "step": 40990 }, { "grad_norm": 0.3676213026046753, "learning_rate": 5.206831378236845e-05, "loss": 0.0319, "step": 41000 }, { "grad_norm": 0.2630783021450043, "learning_rate": 5.2047662924634666e-05, "loss": 0.0235, "step": 41010 }, { "grad_norm": 0.29558250308036804, "learning_rate": 5.20270117170113e-05, "loss": 0.0218, "step": 41020 }, { "grad_norm": 0.3530506491661072, "learning_rate": 5.200636016302707e-05, "loss": 0.0339, "step": 41030 }, { "grad_norm": 0.4744938015937805, "learning_rate": 5.1985708266210754e-05, "loss": 0.0422, "step": 41040 }, { "grad_norm": 0.2772565186023712, "learning_rate": 5.1965056030091206e-05, "loss": 0.0353, "step": 41050 }, { "grad_norm": 0.2668367028236389, "learning_rate": 5.194440345819732e-05, "loss": 0.0221, "step": 41060 }, { "grad_norm": 0.326097309589386, "learning_rate": 5.1923750554058084e-05, "loss": 0.0323, "step": 41070 }, { "grad_norm": 0.23859776556491852, "learning_rate": 5.1903097321202476e-05, "loss": 0.0213, "step": 41080 }, { "grad_norm": 0.39173659682273865, "learning_rate": 5.18824437631596e-05, "loss": 0.0421, "step": 41090 }, { "grad_norm": 0.24368804693222046, "learning_rate": 5.186178988345856e-05, "loss": 0.0218, "step": 41100 }, { "grad_norm": 0.4325285851955414, "learning_rate": 5.184113568562855e-05, "loss": 0.0321, "step": 41110 }, { "grad_norm": 0.3772478401660919, "learning_rate": 5.18204811731988e-05, "loss": 0.0353, "step": 41120 }, { "grad_norm": 0.3488560616970062, "learning_rate": 5.179982634969862e-05, "loss": 0.0285, "step": 41130 }, { "grad_norm": 0.3648497760295868, "learning_rate": 5.177917121865734e-05, "loss": 0.032, "step": 41140 }, { "grad_norm": 0.25543516874313354, "learning_rate": 5.1758515783604346e-05, "loss": 0.0252, "step": 41150 }, { "grad_norm": 0.4217635691165924, "learning_rate": 5.173786004806911e-05, "loss": 0.0318, "step": 41160 }, { "grad_norm": 0.4520794749259949, "learning_rate": 5.1717204015581135e-05, "loss": 0.0305, "step": 41170 }, { "grad_norm": 0.3243190348148346, "learning_rate": 5.169654768966994e-05, "loss": 0.0262, "step": 41180 }, { "grad_norm": 0.4284961521625519, "learning_rate": 5.1675891073865156e-05, "loss": 0.0268, "step": 41190 }, { "grad_norm": 0.22636482119560242, "learning_rate": 5.1655234171696424e-05, "loss": 0.0196, "step": 41200 }, { "grad_norm": 0.2746266722679138, "learning_rate": 5.163457698669343e-05, "loss": 0.0207, "step": 41210 }, { "grad_norm": 0.21355165541172028, "learning_rate": 5.1613919522385936e-05, "loss": 0.0215, "step": 41220 }, { "grad_norm": 0.298751562833786, "learning_rate": 5.1593261782303746e-05, "loss": 0.0353, "step": 41230 }, { "grad_norm": 0.2902316749095917, "learning_rate": 5.157260376997669e-05, "loss": 0.0234, "step": 41240 }, { "grad_norm": 0.2591992914676666, "learning_rate": 5.155194548893466e-05, "loss": 0.0283, "step": 41250 }, { "grad_norm": 0.27280229330062866, "learning_rate": 5.153128694270759e-05, "loss": 0.0378, "step": 41260 }, { "grad_norm": 0.3693796694278717, "learning_rate": 5.151062813482548e-05, "loss": 0.0354, "step": 41270 }, { "grad_norm": 0.3376827538013458, "learning_rate": 5.148996906881832e-05, "loss": 0.023, "step": 41280 }, { "grad_norm": 0.30352646112442017, "learning_rate": 5.1469309748216196e-05, "loss": 0.024, "step": 41290 }, { "grad_norm": 0.3454643189907074, "learning_rate": 5.144865017654923e-05, "loss": 0.0258, "step": 41300 }, { "grad_norm": 0.40418148040771484, "learning_rate": 5.1427990357347556e-05, "loss": 0.0358, "step": 41310 }, { "grad_norm": 0.2523251473903656, "learning_rate": 5.140733029414139e-05, "loss": 0.0268, "step": 41320 }, { "grad_norm": 0.31149742007255554, "learning_rate": 5.138666999046098e-05, "loss": 0.0225, "step": 41330 }, { "grad_norm": 0.35917866230010986, "learning_rate": 5.136600944983658e-05, "loss": 0.0239, "step": 41340 }, { "grad_norm": 0.29850849509239197, "learning_rate": 5.134534867579853e-05, "loss": 0.0287, "step": 41350 }, { "grad_norm": 0.45279327034950256, "learning_rate": 5.1324687671877184e-05, "loss": 0.0251, "step": 41360 }, { "grad_norm": 0.25888702273368835, "learning_rate": 5.130402644160296e-05, "loss": 0.0295, "step": 41370 }, { "grad_norm": 0.34287771582603455, "learning_rate": 5.128336498850628e-05, "loss": 0.0298, "step": 41380 }, { "grad_norm": 0.3788878917694092, "learning_rate": 5.126270331611761e-05, "loss": 0.026, "step": 41390 }, { "grad_norm": 0.3895569145679474, "learning_rate": 5.124204142796748e-05, "loss": 0.0256, "step": 41400 }, { "grad_norm": 0.20175960659980774, "learning_rate": 5.122137932758644e-05, "loss": 0.0242, "step": 41410 }, { "grad_norm": 0.300035297870636, "learning_rate": 5.1200717018505074e-05, "loss": 0.0264, "step": 41420 }, { "grad_norm": 0.2607087790966034, "learning_rate": 5.118005450425403e-05, "loss": 0.0246, "step": 41430 }, { "grad_norm": 0.22605927288532257, "learning_rate": 5.115939178836392e-05, "loss": 0.0212, "step": 41440 }, { "grad_norm": 0.22748278081417084, "learning_rate": 5.113872887436547e-05, "loss": 0.0296, "step": 41450 }, { "grad_norm": 0.18190687894821167, "learning_rate": 5.11180657657894e-05, "loss": 0.0272, "step": 41460 }, { "grad_norm": 0.32012397050857544, "learning_rate": 5.109740246616649e-05, "loss": 0.0254, "step": 41470 }, { "grad_norm": 0.46340593695640564, "learning_rate": 5.10767389790275e-05, "loss": 0.035, "step": 41480 }, { "grad_norm": 0.3395310044288635, "learning_rate": 5.1056075307903265e-05, "loss": 0.0276, "step": 41490 }, { "grad_norm": 0.316762775182724, "learning_rate": 5.1035411456324634e-05, "loss": 0.0235, "step": 41500 }, { "grad_norm": 0.2590913772583008, "learning_rate": 5.1014747427822504e-05, "loss": 0.0288, "step": 41510 }, { "grad_norm": 0.4632296562194824, "learning_rate": 5.0994083225927804e-05, "loss": 0.0333, "step": 41520 }, { "grad_norm": 0.41024157404899597, "learning_rate": 5.0973418854171475e-05, "loss": 0.029, "step": 41530 }, { "grad_norm": 0.2761724293231964, "learning_rate": 5.095275431608447e-05, "loss": 0.0236, "step": 41540 }, { "grad_norm": 0.2931906580924988, "learning_rate": 5.093208961519782e-05, "loss": 0.0242, "step": 41550 }, { "grad_norm": 0.29308655858039856, "learning_rate": 5.091142475504255e-05, "loss": 0.0301, "step": 41560 }, { "grad_norm": 0.3941382169723511, "learning_rate": 5.089075973914971e-05, "loss": 0.0334, "step": 41570 }, { "grad_norm": 0.4001002013683319, "learning_rate": 5.08700945710504e-05, "loss": 0.0377, "step": 41580 }, { "grad_norm": 0.3697991669178009, "learning_rate": 5.0849429254275714e-05, "loss": 0.0265, "step": 41590 }, { "grad_norm": 0.20044180750846863, "learning_rate": 5.0828763792356804e-05, "loss": 0.0243, "step": 41600 }, { "grad_norm": 0.2585298418998718, "learning_rate": 5.080809818882483e-05, "loss": 0.0275, "step": 41610 }, { "grad_norm": 0.39618799090385437, "learning_rate": 5.078743244721097e-05, "loss": 0.0247, "step": 41620 }, { "grad_norm": 0.2931278347969055, "learning_rate": 5.0766766571046455e-05, "loss": 0.0275, "step": 41630 }, { "grad_norm": 0.37303426861763, "learning_rate": 5.07461005638625e-05, "loss": 0.0218, "step": 41640 }, { "grad_norm": 0.2851966917514801, "learning_rate": 5.072543442919037e-05, "loss": 0.0282, "step": 41650 }, { "grad_norm": 0.3693010210990906, "learning_rate": 5.070476817056132e-05, "loss": 0.0357, "step": 41660 }, { "grad_norm": 0.33512571454048157, "learning_rate": 5.068410179150668e-05, "loss": 0.0251, "step": 41670 }, { "grad_norm": 0.2417929619550705, "learning_rate": 5.066343529555775e-05, "loss": 0.0258, "step": 41680 }, { "grad_norm": 0.22716060280799866, "learning_rate": 5.064276868624587e-05, "loss": 0.0221, "step": 41690 }, { "grad_norm": 0.23150065541267395, "learning_rate": 5.0622101967102396e-05, "loss": 0.0242, "step": 41700 }, { "grad_norm": 0.22984279692173004, "learning_rate": 5.0601435141658714e-05, "loss": 0.0218, "step": 41710 }, { "grad_norm": 0.3237409293651581, "learning_rate": 5.058076821344622e-05, "loss": 0.0288, "step": 41720 }, { "grad_norm": 0.2755570411682129, "learning_rate": 5.056010118599632e-05, "loss": 0.039, "step": 41730 }, { "grad_norm": 0.32653263211250305, "learning_rate": 5.053943406284044e-05, "loss": 0.0242, "step": 41740 }, { "grad_norm": 0.33237752318382263, "learning_rate": 5.051876684751006e-05, "loss": 0.0269, "step": 41750 }, { "grad_norm": 0.2196044623851776, "learning_rate": 5.0498099543536584e-05, "loss": 0.0277, "step": 41760 }, { "grad_norm": 0.37502795457839966, "learning_rate": 5.047743215445152e-05, "loss": 0.029, "step": 41770 }, { "grad_norm": 0.3355187475681305, "learning_rate": 5.045676468378637e-05, "loss": 0.0249, "step": 41780 }, { "grad_norm": 0.3542466461658478, "learning_rate": 5.0436097135072626e-05, "loss": 0.0314, "step": 41790 }, { "grad_norm": 0.2613717019557953, "learning_rate": 5.041542951184181e-05, "loss": 0.0269, "step": 41800 }, { "grad_norm": 0.23770178854465485, "learning_rate": 5.039476181762545e-05, "loss": 0.024, "step": 41810 }, { "grad_norm": 0.220929816365242, "learning_rate": 5.037409405595508e-05, "loss": 0.021, "step": 41820 }, { "grad_norm": 0.4217451512813568, "learning_rate": 5.035342623036229e-05, "loss": 0.0255, "step": 41830 }, { "grad_norm": 0.34591376781463623, "learning_rate": 5.033275834437862e-05, "loss": 0.0303, "step": 41840 }, { "grad_norm": 0.3688986301422119, "learning_rate": 5.031209040153564e-05, "loss": 0.0255, "step": 41850 }, { "grad_norm": 0.31852445006370544, "learning_rate": 5.0291422405364955e-05, "loss": 0.0398, "step": 41860 }, { "grad_norm": 0.29997026920318604, "learning_rate": 5.0270754359398133e-05, "loss": 0.0264, "step": 41870 }, { "grad_norm": 0.2805119752883911, "learning_rate": 5.025008626716682e-05, "loss": 0.0222, "step": 41880 }, { "grad_norm": 0.26457470655441284, "learning_rate": 5.0229418132202585e-05, "loss": 0.0231, "step": 41890 }, { "grad_norm": 0.2673797607421875, "learning_rate": 5.020874995803707e-05, "loss": 0.0265, "step": 41900 }, { "grad_norm": 0.33975571393966675, "learning_rate": 5.01880817482019e-05, "loss": 0.0255, "step": 41910 }, { "grad_norm": 0.2746519446372986, "learning_rate": 5.01674135062287e-05, "loss": 0.0279, "step": 41920 }, { "grad_norm": 0.40599507093429565, "learning_rate": 5.0146745235649115e-05, "loss": 0.0318, "step": 41930 }, { "grad_norm": 0.42603954672813416, "learning_rate": 5.0126076939994795e-05, "loss": 0.0251, "step": 41940 }, { "grad_norm": 0.23284456133842468, "learning_rate": 5.010540862279736e-05, "loss": 0.0313, "step": 41950 }, { "grad_norm": 0.2831239402294159, "learning_rate": 5.008474028758846e-05, "loss": 0.0281, "step": 41960 }, { "grad_norm": 0.3220471143722534, "learning_rate": 5.0064071937899784e-05, "loss": 0.0315, "step": 41970 }, { "grad_norm": 0.4816947281360626, "learning_rate": 5.004340357726296e-05, "loss": 0.0257, "step": 41980 }, { "grad_norm": 0.4621039628982544, "learning_rate": 5.002273520920965e-05, "loss": 0.0318, "step": 41990 }, { "grad_norm": 0.27491286396980286, "learning_rate": 5.000206683727151e-05, "loss": 0.0212, "step": 42000 }, { "grad_norm": 0.2812165915966034, "learning_rate": 4.998139846498021e-05, "loss": 0.0203, "step": 42010 }, { "grad_norm": 0.30758556723594666, "learning_rate": 4.9960730095867405e-05, "loss": 0.0319, "step": 42020 }, { "grad_norm": 0.21380415558815002, "learning_rate": 4.9940061733464755e-05, "loss": 0.0306, "step": 42030 }, { "grad_norm": 0.25289294123649597, "learning_rate": 4.991939338130392e-05, "loss": 0.0352, "step": 42040 }, { "grad_norm": 0.3609459102153778, "learning_rate": 4.989872504291653e-05, "loss": 0.0321, "step": 42050 }, { "grad_norm": 0.31753405928611755, "learning_rate": 4.9878056721834273e-05, "loss": 0.0264, "step": 42060 }, { "grad_norm": 0.3832649886608124, "learning_rate": 4.98573884215888e-05, "loss": 0.0267, "step": 42070 }, { "grad_norm": 0.31557679176330566, "learning_rate": 4.9836720145711715e-05, "loss": 0.0274, "step": 42080 }, { "grad_norm": 0.5179601311683655, "learning_rate": 4.9816051897734725e-05, "loss": 0.0294, "step": 42090 }, { "grad_norm": 0.32422560453414917, "learning_rate": 4.979538368118941e-05, "loss": 0.0316, "step": 42100 }, { "grad_norm": 0.244246706366539, "learning_rate": 4.9774715499607446e-05, "loss": 0.0251, "step": 42110 }, { "grad_norm": 0.23458951711654663, "learning_rate": 4.9754047356520416e-05, "loss": 0.0381, "step": 42120 }, { "grad_norm": 0.18947724997997284, "learning_rate": 4.973337925545999e-05, "loss": 0.0267, "step": 42130 }, { "grad_norm": 0.30158406496047974, "learning_rate": 4.9712711199957746e-05, "loss": 0.0224, "step": 42140 }, { "grad_norm": 0.30379170179367065, "learning_rate": 4.969204319354531e-05, "loss": 0.0254, "step": 42150 }, { "grad_norm": 0.3145346939563751, "learning_rate": 4.9671375239754267e-05, "loss": 0.0242, "step": 42160 }, { "grad_norm": 0.25754913687705994, "learning_rate": 4.9650707342116195e-05, "loss": 0.0249, "step": 42170 }, { "grad_norm": 0.2101435661315918, "learning_rate": 4.96300395041627e-05, "loss": 0.0239, "step": 42180 }, { "grad_norm": 0.3943336308002472, "learning_rate": 4.960937172942532e-05, "loss": 0.0226, "step": 42190 }, { "grad_norm": 0.32458630204200745, "learning_rate": 4.958870402143565e-05, "loss": 0.0212, "step": 42200 }, { "grad_norm": 0.2343602180480957, "learning_rate": 4.9568036383725186e-05, "loss": 0.0248, "step": 42210 }, { "grad_norm": 0.47267085313796997, "learning_rate": 4.9547368819825516e-05, "loss": 0.0247, "step": 42220 }, { "grad_norm": 0.29416608810424805, "learning_rate": 4.952670133326812e-05, "loss": 0.0281, "step": 42230 }, { "grad_norm": 0.332828164100647, "learning_rate": 4.950603392758453e-05, "loss": 0.0262, "step": 42240 }, { "grad_norm": 0.2740955352783203, "learning_rate": 4.948536660630621e-05, "loss": 0.022, "step": 42250 }, { "grad_norm": 0.35655534267425537, "learning_rate": 4.9464699372964685e-05, "loss": 0.029, "step": 42260 }, { "grad_norm": 0.3147779405117035, "learning_rate": 4.9444032231091395e-05, "loss": 0.0253, "step": 42270 }, { "grad_norm": 0.25503015518188477, "learning_rate": 4.9423365184217776e-05, "loss": 0.0254, "step": 42280 }, { "grad_norm": 0.31890740990638733, "learning_rate": 4.940269823587529e-05, "loss": 0.0231, "step": 42290 }, { "grad_norm": 0.20400486886501312, "learning_rate": 4.938203138959533e-05, "loss": 0.0185, "step": 42300 }, { "grad_norm": 0.31182730197906494, "learning_rate": 4.9361364648909325e-05, "loss": 0.0251, "step": 42310 }, { "grad_norm": 0.35789695382118225, "learning_rate": 4.9340698017348605e-05, "loss": 0.0227, "step": 42320 }, { "grad_norm": 0.34599700570106506, "learning_rate": 4.932003149844458e-05, "loss": 0.024, "step": 42330 }, { "grad_norm": 0.33702847361564636, "learning_rate": 4.929936509572857e-05, "loss": 0.0351, "step": 42340 }, { "grad_norm": 0.22876641154289246, "learning_rate": 4.927869881273191e-05, "loss": 0.0268, "step": 42350 }, { "grad_norm": 0.3009509742259979, "learning_rate": 4.9258032652985894e-05, "loss": 0.0262, "step": 42360 }, { "grad_norm": 0.24878689646720886, "learning_rate": 4.9237366620021786e-05, "loss": 0.0279, "step": 42370 }, { "grad_norm": 0.26669368147850037, "learning_rate": 4.921670071737089e-05, "loss": 0.0227, "step": 42380 }, { "grad_norm": 0.285422682762146, "learning_rate": 4.91960349485644e-05, "loss": 0.0289, "step": 42390 }, { "grad_norm": 0.2674351632595062, "learning_rate": 4.9175369317133566e-05, "loss": 0.0223, "step": 42400 }, { "grad_norm": 0.24435238540172577, "learning_rate": 4.915470382660953e-05, "loss": 0.0254, "step": 42410 }, { "grad_norm": 0.21336059272289276, "learning_rate": 4.9134038480523524e-05, "loss": 0.0236, "step": 42420 }, { "grad_norm": 0.2396959662437439, "learning_rate": 4.911337328240664e-05, "loss": 0.0271, "step": 42430 }, { "grad_norm": 0.31267258524894714, "learning_rate": 4.909270823579003e-05, "loss": 0.0261, "step": 42440 }, { "grad_norm": 0.3784463405609131, "learning_rate": 4.907204334420476e-05, "loss": 0.022, "step": 42450 }, { "grad_norm": 0.317329466342926, "learning_rate": 4.9051378611181884e-05, "loss": 0.0295, "step": 42460 }, { "grad_norm": 0.28937003016471863, "learning_rate": 4.9030714040252486e-05, "loss": 0.0234, "step": 42470 }, { "grad_norm": 0.19318196177482605, "learning_rate": 4.901004963494752e-05, "loss": 0.0182, "step": 42480 }, { "grad_norm": 0.3602297902107239, "learning_rate": 4.898938539879802e-05, "loss": 0.0273, "step": 42490 }, { "grad_norm": 0.30657079815864563, "learning_rate": 4.8968721335334904e-05, "loss": 0.0238, "step": 42500 }, { "grad_norm": 0.32904982566833496, "learning_rate": 4.894805744808912e-05, "loss": 0.0292, "step": 42510 }, { "grad_norm": 0.18729440867900848, "learning_rate": 4.8927393740591524e-05, "loss": 0.0258, "step": 42520 }, { "grad_norm": 0.2840012311935425, "learning_rate": 4.890673021637302e-05, "loss": 0.0235, "step": 42530 }, { "grad_norm": 0.22043728828430176, "learning_rate": 4.888606687896442e-05, "loss": 0.0226, "step": 42540 }, { "grad_norm": 0.335780531167984, "learning_rate": 4.886540373189652e-05, "loss": 0.0249, "step": 42550 }, { "grad_norm": 0.3784685730934143, "learning_rate": 4.88447407787001e-05, "loss": 0.0221, "step": 42560 }, { "grad_norm": 0.3791496157646179, "learning_rate": 4.8824078022905864e-05, "loss": 0.0222, "step": 42570 }, { "grad_norm": 0.2675478160381317, "learning_rate": 4.880341546804455e-05, "loss": 0.0211, "step": 42580 }, { "grad_norm": 0.3840921223163605, "learning_rate": 4.878275311764679e-05, "loss": 0.0257, "step": 42590 }, { "grad_norm": 0.37959060072898865, "learning_rate": 4.8762090975243254e-05, "loss": 0.0247, "step": 42600 }, { "grad_norm": 0.3071611225605011, "learning_rate": 4.874142904436448e-05, "loss": 0.0216, "step": 42610 }, { "grad_norm": 0.349852591753006, "learning_rate": 4.8720767328541094e-05, "loss": 0.0391, "step": 42620 }, { "grad_norm": 0.3171502947807312, "learning_rate": 4.870010583130358e-05, "loss": 0.029, "step": 42630 }, { "grad_norm": 0.19182239472866058, "learning_rate": 4.867944455618241e-05, "loss": 0.0207, "step": 42640 }, { "grad_norm": 0.7946368455886841, "learning_rate": 4.865878350670807e-05, "loss": 0.0312, "step": 42650 }, { "grad_norm": 0.4093213677406311, "learning_rate": 4.8638122686410914e-05, "loss": 0.028, "step": 42660 }, { "grad_norm": 0.28048044443130493, "learning_rate": 4.861746209882137e-05, "loss": 0.0227, "step": 42670 }, { "grad_norm": 0.2780359983444214, "learning_rate": 4.859680174746972e-05, "loss": 0.0332, "step": 42680 }, { "grad_norm": 0.40622296929359436, "learning_rate": 4.857614163588629e-05, "loss": 0.0289, "step": 42690 }, { "grad_norm": 0.27060577273368835, "learning_rate": 4.855548176760131e-05, "loss": 0.0219, "step": 42700 }, { "grad_norm": 0.2877148985862732, "learning_rate": 4.8534822146144986e-05, "loss": 0.0289, "step": 42710 }, { "grad_norm": 0.4078803062438965, "learning_rate": 4.851416277504749e-05, "loss": 0.0327, "step": 42720 }, { "grad_norm": 0.22458210587501526, "learning_rate": 4.8493503657838923e-05, "loss": 0.0189, "step": 42730 }, { "grad_norm": 0.18800541758537292, "learning_rate": 4.84728447980494e-05, "loss": 0.0229, "step": 42740 }, { "grad_norm": 0.2769237756729126, "learning_rate": 4.8452186199208914e-05, "loss": 0.0225, "step": 42750 }, { "grad_norm": 0.3137410581111908, "learning_rate": 4.843152786484749e-05, "loss": 0.0224, "step": 42760 }, { "grad_norm": 0.46988731622695923, "learning_rate": 4.8410869798495035e-05, "loss": 0.0231, "step": 42770 }, { "grad_norm": 0.18701085448265076, "learning_rate": 4.8390212003681486e-05, "loss": 0.0239, "step": 42780 }, { "grad_norm": 0.3365234136581421, "learning_rate": 4.836955448393667e-05, "loss": 0.03, "step": 42790 }, { "grad_norm": 0.322988361120224, "learning_rate": 4.8348897242790406e-05, "loss": 0.0207, "step": 42800 }, { "grad_norm": 0.18783622980117798, "learning_rate": 4.832824028377243e-05, "loss": 0.0199, "step": 42810 }, { "grad_norm": 0.2909359037876129, "learning_rate": 4.830758361041249e-05, "loss": 0.0216, "step": 42820 }, { "grad_norm": 0.42842787504196167, "learning_rate": 4.828692722624022e-05, "loss": 0.0269, "step": 42830 }, { "grad_norm": 0.2666766345500946, "learning_rate": 4.826627113478522e-05, "loss": 0.0434, "step": 42840 }, { "grad_norm": 0.2894212603569031, "learning_rate": 4.824561533957708e-05, "loss": 0.0193, "step": 42850 }, { "grad_norm": 0.32998543977737427, "learning_rate": 4.822495984414527e-05, "loss": 0.0189, "step": 42860 }, { "grad_norm": 0.30979159474372864, "learning_rate": 4.8204304652019304e-05, "loss": 0.0265, "step": 42870 }, { "grad_norm": 0.2343204915523529, "learning_rate": 4.8183649766728537e-05, "loss": 0.0213, "step": 42880 }, { "grad_norm": 0.4454139173030853, "learning_rate": 4.816299519180237e-05, "loss": 0.0262, "step": 42890 }, { "grad_norm": 0.3149837851524353, "learning_rate": 4.8142340930770066e-05, "loss": 0.023, "step": 42900 }, { "grad_norm": 0.28818386793136597, "learning_rate": 4.81216869871609e-05, "loss": 0.0209, "step": 42910 }, { "grad_norm": 0.3888249397277832, "learning_rate": 4.8101033364504064e-05, "loss": 0.0216, "step": 42920 }, { "grad_norm": 0.2725582718849182, "learning_rate": 4.808038006632866e-05, "loss": 0.0338, "step": 42930 }, { "grad_norm": 0.35073599219322205, "learning_rate": 4.805972709616383e-05, "loss": 0.027, "step": 42940 }, { "grad_norm": 0.33361953496932983, "learning_rate": 4.8039074457538556e-05, "loss": 0.0199, "step": 42950 }, { "grad_norm": 0.18023274838924408, "learning_rate": 4.801842215398184e-05, "loss": 0.0273, "step": 42960 }, { "grad_norm": 0.28215306997299194, "learning_rate": 4.799777018902256e-05, "loss": 0.0261, "step": 42970 }, { "grad_norm": 0.3017377257347107, "learning_rate": 4.797711856618961e-05, "loss": 0.0218, "step": 42980 }, { "grad_norm": 0.3279024362564087, "learning_rate": 4.7956467289011765e-05, "loss": 0.0303, "step": 42990 }, { "grad_norm": 0.48643937706947327, "learning_rate": 4.793581636101778e-05, "loss": 0.0245, "step": 43000 }, { "grad_norm": 0.3599279224872589, "learning_rate": 4.7915165785736326e-05, "loss": 0.0242, "step": 43010 }, { "grad_norm": 0.2778525650501251, "learning_rate": 4.789451556669599e-05, "loss": 0.0223, "step": 43020 }, { "grad_norm": 0.2504437267780304, "learning_rate": 4.7873865707425384e-05, "loss": 0.0291, "step": 43030 }, { "grad_norm": 0.46562936902046204, "learning_rate": 4.7853216211452975e-05, "loss": 0.0296, "step": 43040 }, { "grad_norm": 0.39921846985816956, "learning_rate": 4.78325670823072e-05, "loss": 0.0248, "step": 43050 }, { "grad_norm": 0.22874397039413452, "learning_rate": 4.781191832351641e-05, "loss": 0.023, "step": 43060 }, { "grad_norm": 0.27662691473960876, "learning_rate": 4.7791269938608965e-05, "loss": 0.0257, "step": 43070 }, { "grad_norm": 0.2585393190383911, "learning_rate": 4.777062193111305e-05, "loss": 0.0231, "step": 43080 }, { "grad_norm": 0.314008504152298, "learning_rate": 4.77499743045569e-05, "loss": 0.0255, "step": 43090 }, { "grad_norm": 0.3664350211620331, "learning_rate": 4.7729327062468596e-05, "loss": 0.0274, "step": 43100 }, { "grad_norm": 0.4140094518661499, "learning_rate": 4.770868020837619e-05, "loss": 0.0325, "step": 43110 }, { "grad_norm": 0.29218846559524536, "learning_rate": 4.768803374580768e-05, "loss": 0.0181, "step": 43120 }, { "grad_norm": 0.35062193870544434, "learning_rate": 4.766738767829094e-05, "loss": 0.0243, "step": 43130 }, { "grad_norm": 0.30773472785949707, "learning_rate": 4.764674200935388e-05, "loss": 0.0208, "step": 43140 }, { "grad_norm": 0.36914005875587463, "learning_rate": 4.762609674252424e-05, "loss": 0.0217, "step": 43150 }, { "grad_norm": 0.2583218812942505, "learning_rate": 4.760545188132974e-05, "loss": 0.0332, "step": 43160 }, { "grad_norm": 0.23994110524654388, "learning_rate": 4.758480742929801e-05, "loss": 0.0245, "step": 43170 }, { "grad_norm": 0.29745200276374817, "learning_rate": 4.756416338995664e-05, "loss": 0.0257, "step": 43180 }, { "grad_norm": 0.28574612736701965, "learning_rate": 4.7543519766833145e-05, "loss": 0.0226, "step": 43190 }, { "grad_norm": 0.36471956968307495, "learning_rate": 4.752287656345492e-05, "loss": 0.0264, "step": 43200 }, { "grad_norm": 0.30473655462265015, "learning_rate": 4.7502233783349355e-05, "loss": 0.0243, "step": 43210 }, { "grad_norm": 0.39277350902557373, "learning_rate": 4.7481591430043694e-05, "loss": 0.0211, "step": 43220 }, { "grad_norm": 0.2976706922054291, "learning_rate": 4.746094950706521e-05, "loss": 0.0286, "step": 43230 }, { "grad_norm": 0.30431535840034485, "learning_rate": 4.744030801794099e-05, "loss": 0.0258, "step": 43240 }, { "grad_norm": 0.373739093542099, "learning_rate": 4.741966696619813e-05, "loss": 0.0309, "step": 43250 }, { "grad_norm": 0.20889364182949066, "learning_rate": 4.739902635536361e-05, "loss": 0.0344, "step": 43260 }, { "grad_norm": 0.24837364256381989, "learning_rate": 4.737838618896436e-05, "loss": 0.0245, "step": 43270 }, { "grad_norm": 0.2687264084815979, "learning_rate": 4.7357746470527203e-05, "loss": 0.0266, "step": 43280 }, { "grad_norm": 0.30416789650917053, "learning_rate": 4.73371072035789e-05, "loss": 0.028, "step": 43290 }, { "grad_norm": 0.4066213369369507, "learning_rate": 4.731646839164616e-05, "loss": 0.0258, "step": 43300 }, { "grad_norm": 0.4686112403869629, "learning_rate": 4.729583003825556e-05, "loss": 0.028, "step": 43310 }, { "grad_norm": 0.27607157826423645, "learning_rate": 4.727519214693367e-05, "loss": 0.03, "step": 43320 }, { "grad_norm": 0.2675568759441376, "learning_rate": 4.725455472120689e-05, "loss": 0.025, "step": 43330 }, { "grad_norm": 0.27129364013671875, "learning_rate": 4.723391776460164e-05, "loss": 0.0319, "step": 43340 }, { "grad_norm": 0.30883002281188965, "learning_rate": 4.7213281280644186e-05, "loss": 0.0249, "step": 43350 }, { "grad_norm": 0.30315032601356506, "learning_rate": 4.719264527286075e-05, "loss": 0.0264, "step": 43360 }, { "grad_norm": 0.34053629636764526, "learning_rate": 4.717200974477744e-05, "loss": 0.0312, "step": 43370 }, { "grad_norm": 0.37321367859840393, "learning_rate": 4.715137469992034e-05, "loss": 0.029, "step": 43380 }, { "grad_norm": 0.3465414345264435, "learning_rate": 4.713074014181539e-05, "loss": 0.02, "step": 43390 }, { "grad_norm": 0.2255929857492447, "learning_rate": 4.711010607398846e-05, "loss": 0.0184, "step": 43400 }, { "grad_norm": 0.3090037703514099, "learning_rate": 4.708947249996538e-05, "loss": 0.0213, "step": 43410 }, { "grad_norm": 0.2234322726726532, "learning_rate": 4.706883942327183e-05, "loss": 0.0188, "step": 43420 }, { "grad_norm": 0.2864609956741333, "learning_rate": 4.704820684743347e-05, "loss": 0.0253, "step": 43430 }, { "grad_norm": 0.34597378969192505, "learning_rate": 4.702757477597581e-05, "loss": 0.0248, "step": 43440 }, { "grad_norm": 0.27000153064727783, "learning_rate": 4.700694321242435e-05, "loss": 0.0296, "step": 43450 }, { "grad_norm": 0.37240344285964966, "learning_rate": 4.698631216030441e-05, "loss": 0.0255, "step": 43460 }, { "grad_norm": 0.26471906900405884, "learning_rate": 4.6965681623141314e-05, "loss": 0.0271, "step": 43470 }, { "grad_norm": 0.3068387508392334, "learning_rate": 4.694505160446024e-05, "loss": 0.0214, "step": 43480 }, { "grad_norm": 0.25807324051856995, "learning_rate": 4.692442210778628e-05, "loss": 0.0243, "step": 43490 }, { "grad_norm": 0.40090957283973694, "learning_rate": 4.690379313664448e-05, "loss": 0.0248, "step": 43500 }, { "grad_norm": 0.30000799894332886, "learning_rate": 4.688316469455973e-05, "loss": 0.0284, "step": 43510 }, { "grad_norm": 0.29401376843452454, "learning_rate": 4.6862536785056914e-05, "loss": 0.0279, "step": 43520 }, { "grad_norm": 0.4677225947380066, "learning_rate": 4.684190941166073e-05, "loss": 0.0245, "step": 43530 }, { "grad_norm": 0.360179603099823, "learning_rate": 4.682128257789587e-05, "loss": 0.0225, "step": 43540 }, { "grad_norm": 0.3343961834907532, "learning_rate": 4.680065628728687e-05, "loss": 0.0209, "step": 43550 }, { "grad_norm": 0.3506662845611572, "learning_rate": 4.678003054335822e-05, "loss": 0.0252, "step": 43560 }, { "grad_norm": 0.344731867313385, "learning_rate": 4.675940534963428e-05, "loss": 0.0246, "step": 43570 }, { "grad_norm": 0.3522107005119324, "learning_rate": 4.673878070963931e-05, "loss": 0.0223, "step": 43580 }, { "grad_norm": 0.269758015871048, "learning_rate": 4.671815662689756e-05, "loss": 0.0172, "step": 43590 }, { "grad_norm": 0.36103886365890503, "learning_rate": 4.669753310493306e-05, "loss": 0.0221, "step": 43600 }, { "grad_norm": 0.32324889302253723, "learning_rate": 4.6676910147269845e-05, "loss": 0.0247, "step": 43610 }, { "grad_norm": 0.38036832213401794, "learning_rate": 4.665628775743177e-05, "loss": 0.0226, "step": 43620 }, { "grad_norm": 0.32066357135772705, "learning_rate": 4.66356659389427e-05, "loss": 0.0277, "step": 43630 }, { "grad_norm": 0.2495562583208084, "learning_rate": 4.661504469532629e-05, "loss": 0.022, "step": 43640 }, { "grad_norm": 0.2750486433506012, "learning_rate": 4.659442403010617e-05, "loss": 0.0206, "step": 43650 }, { "grad_norm": 0.28443172574043274, "learning_rate": 4.6573803946805845e-05, "loss": 0.0185, "step": 43660 }, { "grad_norm": 0.24002091586589813, "learning_rate": 4.6553184448948695e-05, "loss": 0.0276, "step": 43670 }, { "grad_norm": 0.2456234246492386, "learning_rate": 4.653256554005807e-05, "loss": 0.0169, "step": 43680 }, { "grad_norm": 0.17467935383319855, "learning_rate": 4.6511947223657145e-05, "loss": 0.0286, "step": 43690 }, { "grad_norm": 0.3044104278087616, "learning_rate": 4.649132950326906e-05, "loss": 0.0278, "step": 43700 }, { "grad_norm": 0.25149908661842346, "learning_rate": 4.647071238241679e-05, "loss": 0.025, "step": 43710 }, { "grad_norm": 0.29238080978393555, "learning_rate": 4.6450095864623264e-05, "loss": 0.0211, "step": 43720 }, { "grad_norm": 0.2953144609928131, "learning_rate": 4.642947995341123e-05, "loss": 0.0189, "step": 43730 }, { "grad_norm": 0.24816730618476868, "learning_rate": 4.640886465230345e-05, "loss": 0.0261, "step": 43740 }, { "grad_norm": 0.23897100985050201, "learning_rate": 4.6388249964822485e-05, "loss": 0.0275, "step": 43750 }, { "grad_norm": 0.3328101933002472, "learning_rate": 4.63676358944908e-05, "loss": 0.0242, "step": 43760 }, { "grad_norm": 0.6622969508171082, "learning_rate": 4.634702244483081e-05, "loss": 0.0308, "step": 43770 }, { "grad_norm": 0.36645257472991943, "learning_rate": 4.632640961936475e-05, "loss": 0.0227, "step": 43780 }, { "grad_norm": 0.21685664355754852, "learning_rate": 4.6305797421614835e-05, "loss": 0.0231, "step": 43790 }, { "grad_norm": 0.1923460066318512, "learning_rate": 4.6285185855103105e-05, "loss": 0.0263, "step": 43800 }, { "grad_norm": 0.3536278009414673, "learning_rate": 4.626457492335151e-05, "loss": 0.0238, "step": 43810 }, { "grad_norm": 0.24205662310123444, "learning_rate": 4.624396462988188e-05, "loss": 0.0173, "step": 43820 }, { "grad_norm": 0.33113572001457214, "learning_rate": 4.6223354978215985e-05, "loss": 0.0241, "step": 43830 }, { "grad_norm": 0.24007651209831238, "learning_rate": 4.620274597187544e-05, "loss": 0.022, "step": 43840 }, { "grad_norm": 0.4516037404537201, "learning_rate": 4.6182137614381736e-05, "loss": 0.0244, "step": 43850 }, { "grad_norm": 0.28822967410087585, "learning_rate": 4.616152990925631e-05, "loss": 0.0207, "step": 43860 }, { "grad_norm": 0.35360997915267944, "learning_rate": 4.614092286002041e-05, "loss": 0.0225, "step": 43870 }, { "grad_norm": 0.30341479182243347, "learning_rate": 4.612031647019528e-05, "loss": 0.0251, "step": 43880 }, { "grad_norm": 0.41976794600486755, "learning_rate": 4.609971074330193e-05, "loss": 0.0231, "step": 43890 }, { "grad_norm": 0.4441498816013336, "learning_rate": 4.607910568286136e-05, "loss": 0.0279, "step": 43900 }, { "grad_norm": 0.2462780475616455, "learning_rate": 4.605850129239438e-05, "loss": 0.0246, "step": 43910 }, { "grad_norm": 0.3257734775543213, "learning_rate": 4.603789757542174e-05, "loss": 0.0209, "step": 43920 }, { "grad_norm": 0.3330494165420532, "learning_rate": 4.601729453546404e-05, "loss": 0.0243, "step": 43930 }, { "grad_norm": 0.24688401818275452, "learning_rate": 4.599669217604177e-05, "loss": 0.0218, "step": 43940 }, { "grad_norm": 0.23755502700805664, "learning_rate": 4.597609050067532e-05, "loss": 0.0271, "step": 43950 }, { "grad_norm": 0.24506106972694397, "learning_rate": 4.595548951288495e-05, "loss": 0.0261, "step": 43960 }, { "grad_norm": 0.2965695559978485, "learning_rate": 4.593488921619081e-05, "loss": 0.0317, "step": 43970 }, { "grad_norm": 0.3055723011493683, "learning_rate": 4.591428961411289e-05, "loss": 0.0247, "step": 43980 }, { "grad_norm": 0.4024428725242615, "learning_rate": 4.589369071017117e-05, "loss": 0.0308, "step": 43990 }, { "grad_norm": 0.370498925447464, "learning_rate": 4.587309250788538e-05, "loss": 0.028, "step": 44000 }, { "grad_norm": 0.18593117594718933, "learning_rate": 4.585249501077522e-05, "loss": 0.0232, "step": 44010 }, { "grad_norm": 0.2314170002937317, "learning_rate": 4.583189822236021e-05, "loss": 0.0277, "step": 44020 }, { "grad_norm": 0.3138795495033264, "learning_rate": 4.5811302146159816e-05, "loss": 0.0222, "step": 44030 }, { "grad_norm": 0.28650930523872375, "learning_rate": 4.579070678569332e-05, "loss": 0.0234, "step": 44040 }, { "grad_norm": 0.25703164935112, "learning_rate": 4.5770112144479904e-05, "loss": 0.026, "step": 44050 }, { "grad_norm": 0.2919495403766632, "learning_rate": 4.5749518226038645e-05, "loss": 0.0191, "step": 44060 }, { "grad_norm": 0.340896874666214, "learning_rate": 4.572892503388845e-05, "loss": 0.0248, "step": 44070 }, { "grad_norm": 0.3079904615879059, "learning_rate": 4.570833257154817e-05, "loss": 0.023, "step": 44080 }, { "grad_norm": 0.2567722201347351, "learning_rate": 4.568774084253646e-05, "loss": 0.0298, "step": 44090 }, { "grad_norm": 0.3542255759239197, "learning_rate": 4.566714985037191e-05, "loss": 0.0318, "step": 44100 }, { "grad_norm": 0.47008463740348816, "learning_rate": 4.564655959857295e-05, "loss": 0.0252, "step": 44110 }, { "grad_norm": 0.3434602618217468, "learning_rate": 4.5625970090657893e-05, "loss": 0.0275, "step": 44120 }, { "grad_norm": 0.2529603838920593, "learning_rate": 4.560538133014491e-05, "loss": 0.0232, "step": 44130 }, { "grad_norm": 0.2344667613506317, "learning_rate": 4.5584793320552055e-05, "loss": 0.024, "step": 44140 }, { "grad_norm": 0.3170844316482544, "learning_rate": 4.556420606539728e-05, "loss": 0.0226, "step": 44150 }, { "grad_norm": 0.3707510232925415, "learning_rate": 4.554361956819836e-05, "loss": 0.0206, "step": 44160 }, { "grad_norm": 0.35830262303352356, "learning_rate": 4.552303383247299e-05, "loss": 0.0208, "step": 44170 }, { "grad_norm": 0.3962932527065277, "learning_rate": 4.5502448861738664e-05, "loss": 0.0204, "step": 44180 }, { "grad_norm": 0.30411210656166077, "learning_rate": 4.548186465951284e-05, "loss": 0.0388, "step": 44190 }, { "grad_norm": 0.40696802735328674, "learning_rate": 4.546128122931277e-05, "loss": 0.0274, "step": 44200 }, { "grad_norm": 0.2903398871421814, "learning_rate": 4.544069857465561e-05, "loss": 0.0227, "step": 44210 }, { "grad_norm": 0.30662989616394043, "learning_rate": 4.542011669905837e-05, "loss": 0.0179, "step": 44220 }, { "grad_norm": 0.3324110805988312, "learning_rate": 4.53995356060379e-05, "loss": 0.0229, "step": 44230 }, { "grad_norm": 0.2511032819747925, "learning_rate": 4.5378955299110994e-05, "loss": 0.0179, "step": 44240 }, { "grad_norm": 0.32718122005462646, "learning_rate": 4.5358375781794233e-05, "loss": 0.0286, "step": 44250 }, { "grad_norm": 0.3137694299221039, "learning_rate": 4.533779705760411e-05, "loss": 0.0233, "step": 44260 }, { "grad_norm": 0.5137515664100647, "learning_rate": 4.5317219130056934e-05, "loss": 0.0257, "step": 44270 }, { "grad_norm": 0.3201408386230469, "learning_rate": 4.5296642002668964e-05, "loss": 0.0301, "step": 44280 }, { "grad_norm": 0.30386850237846375, "learning_rate": 4.5276065678956216e-05, "loss": 0.0237, "step": 44290 }, { "grad_norm": 0.4085293114185333, "learning_rate": 4.525549016243466e-05, "loss": 0.0258, "step": 44300 }, { "grad_norm": 0.19407086074352264, "learning_rate": 4.523491545662008e-05, "loss": 0.0186, "step": 44310 }, { "grad_norm": 0.2913292944431305, "learning_rate": 4.5214341565028104e-05, "loss": 0.0287, "step": 44320 }, { "grad_norm": 0.37765243649482727, "learning_rate": 4.519376849117428e-05, "loss": 0.021, "step": 44330 }, { "grad_norm": 0.12530100345611572, "learning_rate": 4.517319623857395e-05, "loss": 0.031, "step": 44340 }, { "grad_norm": 0.1998419612646103, "learning_rate": 4.5152624810742394e-05, "loss": 0.0258, "step": 44350 }, { "grad_norm": 0.2310793101787567, "learning_rate": 4.5132054211194664e-05, "loss": 0.0208, "step": 44360 }, { "grad_norm": 0.3051217496395111, "learning_rate": 4.511148444344574e-05, "loss": 0.0184, "step": 44370 }, { "grad_norm": 0.29894211888313293, "learning_rate": 4.509091551101041e-05, "loss": 0.0293, "step": 44380 }, { "grad_norm": 0.24856138229370117, "learning_rate": 4.507034741740338e-05, "loss": 0.0199, "step": 44390 }, { "grad_norm": 0.3212547302246094, "learning_rate": 4.5049780166139145e-05, "loss": 0.0303, "step": 44400 }, { "grad_norm": 0.35207599401474, "learning_rate": 4.5029213760732075e-05, "loss": 0.0402, "step": 44410 }, { "grad_norm": 0.3685966432094574, "learning_rate": 4.5008648204696434e-05, "loss": 0.0333, "step": 44420 }, { "grad_norm": 0.326994389295578, "learning_rate": 4.4988083501546284e-05, "loss": 0.0282, "step": 44430 }, { "grad_norm": 0.22585918009281158, "learning_rate": 4.4967519654795606e-05, "loss": 0.0243, "step": 44440 }, { "grad_norm": 0.34779244661331177, "learning_rate": 4.494695666795816e-05, "loss": 0.0234, "step": 44450 }, { "grad_norm": 0.37820005416870117, "learning_rate": 4.4926394544547644e-05, "loss": 0.0272, "step": 44460 }, { "grad_norm": 0.23561139404773712, "learning_rate": 4.490583328807752e-05, "loss": 0.0236, "step": 44470 }, { "grad_norm": 0.19762049615383148, "learning_rate": 4.488527290206117e-05, "loss": 0.0235, "step": 44480 }, { "grad_norm": 0.2950964868068695, "learning_rate": 4.48647133900118e-05, "loss": 0.0207, "step": 44490 }, { "grad_norm": 0.3509270250797272, "learning_rate": 4.484415475544244e-05, "loss": 0.0262, "step": 44500 }, { "grad_norm": 0.2294931262731552, "learning_rate": 4.4823597001866035e-05, "loss": 0.0225, "step": 44510 }, { "grad_norm": 0.27267009019851685, "learning_rate": 4.480304013279532e-05, "loss": 0.024, "step": 44520 }, { "grad_norm": 0.2646942734718323, "learning_rate": 4.478248415174292e-05, "loss": 0.0202, "step": 44530 }, { "grad_norm": 0.25913211703300476, "learning_rate": 4.476192906222126e-05, "loss": 0.0275, "step": 44540 }, { "grad_norm": 0.3957403302192688, "learning_rate": 4.474137486774268e-05, "loss": 0.0356, "step": 44550 }, { "grad_norm": 0.28656935691833496, "learning_rate": 4.4720821571819296e-05, "loss": 0.0237, "step": 44560 }, { "grad_norm": 0.28710678219795227, "learning_rate": 4.470026917796314e-05, "loss": 0.0306, "step": 44570 }, { "grad_norm": 0.42682531476020813, "learning_rate": 4.4679717689686005e-05, "loss": 0.0438, "step": 44580 }, { "grad_norm": 0.3232438266277313, "learning_rate": 4.465916711049963e-05, "loss": 0.0242, "step": 44590 }, { "grad_norm": 0.25445765256881714, "learning_rate": 4.4638617443915524e-05, "loss": 0.0231, "step": 44600 }, { "grad_norm": 0.36879345774650574, "learning_rate": 4.4618068693445055e-05, "loss": 0.0272, "step": 44610 }, { "grad_norm": 0.29068002104759216, "learning_rate": 4.459752086259946e-05, "loss": 0.0232, "step": 44620 }, { "grad_norm": 0.3850894272327423, "learning_rate": 4.457697395488977e-05, "loss": 0.0297, "step": 44630 }, { "grad_norm": 0.2201663851737976, "learning_rate": 4.455642797382693e-05, "loss": 0.0289, "step": 44640 }, { "grad_norm": 0.29473569989204407, "learning_rate": 4.4535882922921643e-05, "loss": 0.0262, "step": 44650 }, { "grad_norm": 0.2933155596256256, "learning_rate": 4.451533880568455e-05, "loss": 0.0276, "step": 44660 }, { "grad_norm": 0.3400036096572876, "learning_rate": 4.449479562562603e-05, "loss": 0.0286, "step": 44670 }, { "grad_norm": 0.31301942467689514, "learning_rate": 4.447425338625639e-05, "loss": 0.0288, "step": 44680 }, { "grad_norm": 0.26969942450523376, "learning_rate": 4.4453712091085705e-05, "loss": 0.0228, "step": 44690 }, { "grad_norm": 0.2928808629512787, "learning_rate": 4.443317174362392e-05, "loss": 0.0273, "step": 44700 }, { "grad_norm": 0.3746318519115448, "learning_rate": 4.441263234738085e-05, "loss": 0.0244, "step": 44710 }, { "grad_norm": 0.2728104889392853, "learning_rate": 4.4392093905866086e-05, "loss": 0.016, "step": 44720 }, { "grad_norm": 0.4208295941352844, "learning_rate": 4.437155642258911e-05, "loss": 0.025, "step": 44730 }, { "grad_norm": 0.2065831422805786, "learning_rate": 4.4351019901059177e-05, "loss": 0.0292, "step": 44740 }, { "grad_norm": 0.3452509939670563, "learning_rate": 4.433048434478547e-05, "loss": 0.0363, "step": 44750 }, { "grad_norm": 0.4708053767681122, "learning_rate": 4.4309949757276906e-05, "loss": 0.0296, "step": 44760 }, { "grad_norm": 0.2531954050064087, "learning_rate": 4.428941614204233e-05, "loss": 0.0268, "step": 44770 }, { "grad_norm": 0.2646729350090027, "learning_rate": 4.426888350259034e-05, "loss": 0.0255, "step": 44780 }, { "grad_norm": 0.25933218002319336, "learning_rate": 4.424835184242939e-05, "loss": 0.0369, "step": 44790 }, { "grad_norm": 0.32523155212402344, "learning_rate": 4.422782116506784e-05, "loss": 0.0232, "step": 44800 }, { "grad_norm": 0.40904808044433594, "learning_rate": 4.4207291474013754e-05, "loss": 0.0254, "step": 44810 }, { "grad_norm": 0.35646572709083557, "learning_rate": 4.418676277277514e-05, "loss": 0.0318, "step": 44820 }, { "grad_norm": 0.22894273698329926, "learning_rate": 4.416623506485976e-05, "loss": 0.0312, "step": 44830 }, { "grad_norm": 0.23928824067115784, "learning_rate": 4.414570835377527e-05, "loss": 0.0223, "step": 44840 }, { "grad_norm": 0.2513490617275238, "learning_rate": 4.412518264302909e-05, "loss": 0.02, "step": 44850 }, { "grad_norm": 0.2627162039279938, "learning_rate": 4.4104657936128535e-05, "loss": 0.0185, "step": 44860 }, { "grad_norm": 0.32174110412597656, "learning_rate": 4.408413423658071e-05, "loss": 0.0258, "step": 44870 }, { "grad_norm": 0.3557426333427429, "learning_rate": 4.406361154789253e-05, "loss": 0.0233, "step": 44880 }, { "grad_norm": 0.4331205189228058, "learning_rate": 4.4043089873570776e-05, "loss": 0.0267, "step": 44890 }, { "grad_norm": 0.4076271057128906, "learning_rate": 4.4022569217122025e-05, "loss": 0.0257, "step": 44900 }, { "grad_norm": 0.37447893619537354, "learning_rate": 4.400204958205274e-05, "loss": 0.0251, "step": 44910 }, { "grad_norm": 0.315834641456604, "learning_rate": 4.3981530971869125e-05, "loss": 0.0301, "step": 44920 }, { "grad_norm": 0.23806507885456085, "learning_rate": 4.396101339007727e-05, "loss": 0.021, "step": 44930 }, { "grad_norm": 0.2140141874551773, "learning_rate": 4.394049684018304e-05, "loss": 0.0224, "step": 44940 }, { "grad_norm": 0.46072882413864136, "learning_rate": 4.3919981325692186e-05, "loss": 0.0278, "step": 44950 }, { "grad_norm": 0.2737088203430176, "learning_rate": 4.389946685011024e-05, "loss": 0.0181, "step": 44960 }, { "grad_norm": 0.2980045974254608, "learning_rate": 4.387895341694255e-05, "loss": 0.0306, "step": 44970 }, { "grad_norm": 0.2884047329425812, "learning_rate": 4.3858441029694324e-05, "loss": 0.0305, "step": 44980 }, { "grad_norm": 0.24653764069080353, "learning_rate": 4.3837929691870527e-05, "loss": 0.0211, "step": 44990 }, { "grad_norm": 0.24529506266117096, "learning_rate": 4.381741940697604e-05, "loss": 0.0274, "step": 45000 }, { "grad_norm": 0.3255265951156616, "learning_rate": 4.379691017851547e-05, "loss": 0.0198, "step": 45010 }, { "grad_norm": 0.25156086683273315, "learning_rate": 4.3776402009993304e-05, "loss": 0.027, "step": 45020 }, { "grad_norm": 0.3211059868335724, "learning_rate": 4.3755894904913794e-05, "loss": 0.0238, "step": 45030 }, { "grad_norm": 0.44128990173339844, "learning_rate": 4.373538886678109e-05, "loss": 0.0207, "step": 45040 }, { "grad_norm": 0.26096683740615845, "learning_rate": 4.371488389909909e-05, "loss": 0.0239, "step": 45050 }, { "grad_norm": 0.26768559217453003, "learning_rate": 4.3694380005371515e-05, "loss": 0.0231, "step": 45060 }, { "grad_norm": 0.2493806779384613, "learning_rate": 4.367387718910196e-05, "loss": 0.0241, "step": 45070 }, { "grad_norm": 0.27917808294296265, "learning_rate": 4.3653375453793764e-05, "loss": 0.0215, "step": 45080 }, { "grad_norm": 0.19005359709262848, "learning_rate": 4.3632874802950136e-05, "loss": 0.0192, "step": 45090 }, { "grad_norm": 0.2713112533092499, "learning_rate": 4.3612375240074034e-05, "loss": 0.0232, "step": 45100 }, { "grad_norm": 0.46830299496650696, "learning_rate": 4.3591876768668325e-05, "loss": 0.0243, "step": 45110 }, { "grad_norm": 0.22643469274044037, "learning_rate": 4.3571379392235605e-05, "loss": 0.026, "step": 45120 }, { "grad_norm": 0.42491719126701355, "learning_rate": 4.3550883114278335e-05, "loss": 0.0276, "step": 45130 }, { "grad_norm": 0.28085625171661377, "learning_rate": 4.353038793829876e-05, "loss": 0.0193, "step": 45140 }, { "grad_norm": 0.39308443665504456, "learning_rate": 4.350989386779891e-05, "loss": 0.0254, "step": 45150 }, { "grad_norm": 0.351581871509552, "learning_rate": 4.3489400906280724e-05, "loss": 0.032, "step": 45160 }, { "grad_norm": 0.33669403195381165, "learning_rate": 4.3468909057245845e-05, "loss": 0.0264, "step": 45170 }, { "grad_norm": 0.22864271700382233, "learning_rate": 4.3448418324195794e-05, "loss": 0.0247, "step": 45180 }, { "grad_norm": 0.20269103348255157, "learning_rate": 4.342792871063184e-05, "loss": 0.0263, "step": 45190 }, { "grad_norm": 0.21983669698238373, "learning_rate": 4.3407440220055145e-05, "loss": 0.0232, "step": 45200 }, { "grad_norm": 0.3438045382499695, "learning_rate": 4.33869528559666e-05, "loss": 0.0235, "step": 45210 }, { "grad_norm": 0.44732001423835754, "learning_rate": 4.336646662186696e-05, "loss": 0.0248, "step": 45220 }, { "grad_norm": 0.27769383788108826, "learning_rate": 4.334598152125672e-05, "loss": 0.0235, "step": 45230 }, { "grad_norm": 0.23639917373657227, "learning_rate": 4.3325497557636276e-05, "loss": 0.0284, "step": 45240 }, { "grad_norm": 0.17668305337429047, "learning_rate": 4.330501473450574e-05, "loss": 0.0194, "step": 45250 }, { "grad_norm": 0.34386110305786133, "learning_rate": 4.328453305536507e-05, "loss": 0.0226, "step": 45260 }, { "grad_norm": 0.351509153842926, "learning_rate": 4.326405252371404e-05, "loss": 0.0264, "step": 45270 }, { "grad_norm": 0.38225823640823364, "learning_rate": 4.324357314305221e-05, "loss": 0.0251, "step": 45280 }, { "grad_norm": 0.24029043316841125, "learning_rate": 4.3223094916878945e-05, "loss": 0.0235, "step": 45290 }, { "grad_norm": 0.20544728636741638, "learning_rate": 4.320261784869338e-05, "loss": 0.0259, "step": 45300 }, { "grad_norm": 0.19117431342601776, "learning_rate": 4.318214194199455e-05, "loss": 0.0301, "step": 45310 }, { "grad_norm": 0.31656399369239807, "learning_rate": 4.316166720028118e-05, "loss": 0.0192, "step": 45320 }, { "grad_norm": 0.2504903972148895, "learning_rate": 4.3141193627051864e-05, "loss": 0.0258, "step": 45330 }, { "grad_norm": 0.4001651108264923, "learning_rate": 4.312072122580496e-05, "loss": 0.0266, "step": 45340 }, { "grad_norm": 0.15117450058460236, "learning_rate": 4.3100250000038646e-05, "loss": 0.0223, "step": 45350 }, { "grad_norm": 0.39831119775772095, "learning_rate": 4.307977995325091e-05, "loss": 0.0372, "step": 45360 }, { "grad_norm": 0.36206918954849243, "learning_rate": 4.30593110889395e-05, "loss": 0.0209, "step": 45370 }, { "grad_norm": 0.4160236418247223, "learning_rate": 4.3038843410602016e-05, "loss": 0.0234, "step": 45380 }, { "grad_norm": 0.2540568709373474, "learning_rate": 4.3018376921735774e-05, "loss": 0.0271, "step": 45390 }, { "grad_norm": 0.23795369267463684, "learning_rate": 4.299791162583799e-05, "loss": 0.0198, "step": 45400 }, { "grad_norm": 0.2690275013446808, "learning_rate": 4.29774475264056e-05, "loss": 0.0251, "step": 45410 }, { "grad_norm": 0.3694800138473511, "learning_rate": 4.2956984626935365e-05, "loss": 0.0219, "step": 45420 }, { "grad_norm": 0.34320661425590515, "learning_rate": 4.293652293092383e-05, "loss": 0.0216, "step": 45430 }, { "grad_norm": 0.2824220657348633, "learning_rate": 4.2916062441867324e-05, "loss": 0.0235, "step": 45440 }, { "grad_norm": 0.3056287169456482, "learning_rate": 4.289560316326201e-05, "loss": 0.024, "step": 45450 }, { "grad_norm": 0.4782887399196625, "learning_rate": 4.28751450986038e-05, "loss": 0.0345, "step": 45460 }, { "grad_norm": 0.2880133390426636, "learning_rate": 4.2854688251388444e-05, "loss": 0.0242, "step": 45470 }, { "grad_norm": 0.3396446406841278, "learning_rate": 4.2834232625111425e-05, "loss": 0.026, "step": 45480 }, { "grad_norm": 0.26572737097740173, "learning_rate": 4.2813778223268086e-05, "loss": 0.0205, "step": 45490 }, { "grad_norm": 0.4003857970237732, "learning_rate": 4.2793325049353477e-05, "loss": 0.0233, "step": 45500 }, { "grad_norm": 0.293119341135025, "learning_rate": 4.2772873106862535e-05, "loss": 0.0224, "step": 45510 }, { "grad_norm": 0.3652534782886505, "learning_rate": 4.275242239928993e-05, "loss": 0.0206, "step": 45520 }, { "grad_norm": 0.2071252316236496, "learning_rate": 4.273197293013009e-05, "loss": 0.0176, "step": 45530 }, { "grad_norm": 0.21089398860931396, "learning_rate": 4.271152470287731e-05, "loss": 0.02, "step": 45540 }, { "grad_norm": 0.2740284204483032, "learning_rate": 4.2691077721025594e-05, "loss": 0.021, "step": 45550 }, { "grad_norm": 0.2170359343290329, "learning_rate": 4.267063198806883e-05, "loss": 0.0214, "step": 45560 }, { "grad_norm": 0.2694702446460724, "learning_rate": 4.2650187507500574e-05, "loss": 0.0165, "step": 45570 }, { "grad_norm": 0.32480859756469727, "learning_rate": 4.2629744282814275e-05, "loss": 0.0256, "step": 45580 }, { "grad_norm": 0.3098178803920746, "learning_rate": 4.2609302317503074e-05, "loss": 0.0207, "step": 45590 }, { "grad_norm": 0.3292342722415924, "learning_rate": 4.258886161505999e-05, "loss": 0.0234, "step": 45600 }, { "grad_norm": 0.37876927852630615, "learning_rate": 4.2568422178977775e-05, "loss": 0.0228, "step": 45610 }, { "grad_norm": 0.3371129333972931, "learning_rate": 4.254798401274894e-05, "loss": 0.0307, "step": 45620 }, { "grad_norm": 0.31838271021842957, "learning_rate": 4.252754711986583e-05, "loss": 0.0226, "step": 45630 }, { "grad_norm": 0.2123676985502243, "learning_rate": 4.250711150382052e-05, "loss": 0.0214, "step": 45640 }, { "grad_norm": 0.24647390842437744, "learning_rate": 4.248667716810495e-05, "loss": 0.0241, "step": 45650 }, { "grad_norm": 0.21153448522090912, "learning_rate": 4.246624411621074e-05, "loss": 0.0197, "step": 45660 }, { "grad_norm": 0.34801021218299866, "learning_rate": 4.244581235162938e-05, "loss": 0.0206, "step": 45670 }, { "grad_norm": 0.2158217579126358, "learning_rate": 4.2425381877852075e-05, "loss": 0.0231, "step": 45680 }, { "grad_norm": 0.33381009101867676, "learning_rate": 4.2404952698369856e-05, "loss": 0.0196, "step": 45690 }, { "grad_norm": 0.235011488199234, "learning_rate": 4.238452481667349e-05, "loss": 0.0236, "step": 45700 }, { "grad_norm": 0.2175016701221466, "learning_rate": 4.2364098236253526e-05, "loss": 0.0286, "step": 45710 }, { "grad_norm": 0.3284439146518707, "learning_rate": 4.2343672960600356e-05, "loss": 0.0229, "step": 45720 }, { "grad_norm": 0.3394656479358673, "learning_rate": 4.232324899320406e-05, "loss": 0.0218, "step": 45730 }, { "grad_norm": 0.23226158320903778, "learning_rate": 4.230282633755457e-05, "loss": 0.0243, "step": 45740 }, { "grad_norm": 0.2861132025718689, "learning_rate": 4.2282404997141515e-05, "loss": 0.0269, "step": 45750 }, { "grad_norm": 0.4543817639350891, "learning_rate": 4.2261984975454397e-05, "loss": 0.0263, "step": 45760 }, { "grad_norm": 0.33137696981430054, "learning_rate": 4.224156627598239e-05, "loss": 0.0221, "step": 45770 }, { "grad_norm": 0.30014708638191223, "learning_rate": 4.222114890221453e-05, "loss": 0.025, "step": 45780 }, { "grad_norm": 0.32215869426727295, "learning_rate": 4.2200732857639546e-05, "loss": 0.0295, "step": 45790 }, { "grad_norm": 0.23243527114391327, "learning_rate": 4.2180318145746035e-05, "loss": 0.0229, "step": 45800 }, { "grad_norm": 0.33870723843574524, "learning_rate": 4.215990477002227e-05, "loss": 0.0292, "step": 45810 }, { "grad_norm": 0.3459489941596985, "learning_rate": 4.2139492733956356e-05, "loss": 0.0194, "step": 45820 }, { "grad_norm": 0.28872594237327576, "learning_rate": 4.211908204103615e-05, "loss": 0.0335, "step": 45830 }, { "grad_norm": 0.24283456802368164, "learning_rate": 4.2098672694749265e-05, "loss": 0.0356, "step": 45840 }, { "grad_norm": 0.25008460879325867, "learning_rate": 4.2078264698583133e-05, "loss": 0.0296, "step": 45850 }, { "grad_norm": 0.18246343731880188, "learning_rate": 4.205785805602488e-05, "loss": 0.0213, "step": 45860 }, { "grad_norm": 0.24995532631874084, "learning_rate": 4.203745277056149e-05, "loss": 0.0259, "step": 45870 }, { "grad_norm": 0.17445802688598633, "learning_rate": 4.201704884567964e-05, "loss": 0.0203, "step": 45880 }, { "grad_norm": 0.4337257146835327, "learning_rate": 4.1996646284865816e-05, "loss": 0.0303, "step": 45890 }, { "grad_norm": 0.42376965284347534, "learning_rate": 4.197624509160625e-05, "loss": 0.0263, "step": 45900 }, { "grad_norm": 0.25962793827056885, "learning_rate": 4.195584526938692e-05, "loss": 0.0216, "step": 45910 }, { "grad_norm": 0.35532015562057495, "learning_rate": 4.193544682169365e-05, "loss": 0.0225, "step": 45920 }, { "grad_norm": 0.42676234245300293, "learning_rate": 4.1915049752011946e-05, "loss": 0.0229, "step": 45930 }, { "grad_norm": 0.3330972194671631, "learning_rate": 4.189465406382712e-05, "loss": 0.019, "step": 45940 }, { "grad_norm": 0.3737477958202362, "learning_rate": 4.187425976062422e-05, "loss": 0.0271, "step": 45950 }, { "grad_norm": 0.28897982835769653, "learning_rate": 4.18538668458881e-05, "loss": 0.0181, "step": 45960 }, { "grad_norm": 0.2312711477279663, "learning_rate": 4.183347532310333e-05, "loss": 0.0248, "step": 45970 }, { "grad_norm": 0.26747918128967285, "learning_rate": 4.181308519575429e-05, "loss": 0.0223, "step": 45980 }, { "grad_norm": 0.22424158453941345, "learning_rate": 4.179269646732507e-05, "loss": 0.0314, "step": 45990 }, { "grad_norm": 0.21783144772052765, "learning_rate": 4.177230914129954e-05, "loss": 0.0248, "step": 46000 }, { "grad_norm": 0.3456493020057678, "learning_rate": 4.175192322116136e-05, "loss": 0.0211, "step": 46010 }, { "grad_norm": 0.20795880258083344, "learning_rate": 4.173153871039391e-05, "loss": 0.0191, "step": 46020 }, { "grad_norm": 0.38984259963035583, "learning_rate": 4.171115561248036e-05, "loss": 0.0222, "step": 46030 }, { "grad_norm": 0.2824731767177582, "learning_rate": 4.16907739309036e-05, "loss": 0.0231, "step": 46040 }, { "grad_norm": 0.23706910014152527, "learning_rate": 4.167039366914633e-05, "loss": 0.016, "step": 46050 }, { "grad_norm": 0.3814046084880829, "learning_rate": 4.165001483069096e-05, "loss": 0.0219, "step": 46060 }, { "grad_norm": 0.3507574498653412, "learning_rate": 4.16296374190197e-05, "loss": 0.0311, "step": 46070 }, { "grad_norm": 0.2774868309497833, "learning_rate": 4.1609261437614464e-05, "loss": 0.022, "step": 46080 }, { "grad_norm": 0.39699164032936096, "learning_rate": 4.158888688995696e-05, "loss": 0.0262, "step": 46090 }, { "grad_norm": 0.32169800996780396, "learning_rate": 4.1568513779528645e-05, "loss": 0.0192, "step": 46100 }, { "grad_norm": 0.3685661852359772, "learning_rate": 4.1548142109810704e-05, "loss": 0.0361, "step": 46110 }, { "grad_norm": 0.3278943598270416, "learning_rate": 4.152777188428414e-05, "loss": 0.0212, "step": 46120 }, { "grad_norm": 0.38593193888664246, "learning_rate": 4.1507403106429646e-05, "loss": 0.0244, "step": 46130 }, { "grad_norm": 0.28430992364883423, "learning_rate": 4.148703577972768e-05, "loss": 0.0193, "step": 46140 }, { "grad_norm": 0.2752138376235962, "learning_rate": 4.146666990765846e-05, "loss": 0.0218, "step": 46150 }, { "grad_norm": 0.25285816192626953, "learning_rate": 4.1446305493701986e-05, "loss": 0.0226, "step": 46160 }, { "grad_norm": 0.27378585934638977, "learning_rate": 4.142594254133796e-05, "loss": 0.0212, "step": 46170 }, { "grad_norm": 0.31534385681152344, "learning_rate": 4.1405581054045835e-05, "loss": 0.0257, "step": 46180 }, { "grad_norm": 0.33779847621917725, "learning_rate": 4.1385221035304864e-05, "loss": 0.0249, "step": 46190 }, { "grad_norm": 0.38754427433013916, "learning_rate": 4.136486248859398e-05, "loss": 0.027, "step": 46200 }, { "grad_norm": 0.18256960809230804, "learning_rate": 4.1344505417391955e-05, "loss": 0.0241, "step": 46210 }, { "grad_norm": 0.3169498145580292, "learning_rate": 4.132414982517721e-05, "loss": 0.0212, "step": 46220 }, { "grad_norm": 0.31019115447998047, "learning_rate": 4.130379571542798e-05, "loss": 0.0264, "step": 46230 }, { "grad_norm": 0.37282273173332214, "learning_rate": 4.12834430916222e-05, "loss": 0.0262, "step": 46240 }, { "grad_norm": 0.4624139666557312, "learning_rate": 4.126309195723763e-05, "loss": 0.0265, "step": 46250 }, { "grad_norm": 0.2840980887413025, "learning_rate": 4.124274231575168e-05, "loss": 0.0255, "step": 46260 }, { "grad_norm": 0.2613815367221832, "learning_rate": 4.122239417064154e-05, "loss": 0.0203, "step": 46270 }, { "grad_norm": 0.22756502032279968, "learning_rate": 4.1202047525384184e-05, "loss": 0.0181, "step": 46280 }, { "grad_norm": 0.26016512513160706, "learning_rate": 4.118170238345627e-05, "loss": 0.0187, "step": 46290 }, { "grad_norm": 0.1896827071905136, "learning_rate": 4.1161358748334256e-05, "loss": 0.0205, "step": 46300 }, { "grad_norm": 0.34062254428863525, "learning_rate": 4.1141016623494266e-05, "loss": 0.0202, "step": 46310 }, { "grad_norm": 0.36904528737068176, "learning_rate": 4.112067601241227e-05, "loss": 0.0209, "step": 46320 }, { "grad_norm": 0.14568114280700684, "learning_rate": 4.110033691856387e-05, "loss": 0.0215, "step": 46330 }, { "grad_norm": 0.3902808427810669, "learning_rate": 4.107999934542451e-05, "loss": 0.0299, "step": 46340 }, { "grad_norm": 0.20977389812469482, "learning_rate": 4.105966329646928e-05, "loss": 0.0227, "step": 46350 }, { "grad_norm": 0.23681411147117615, "learning_rate": 4.103932877517308e-05, "loss": 0.0199, "step": 46360 }, { "grad_norm": 0.27492690086364746, "learning_rate": 4.101899578501052e-05, "loss": 0.0238, "step": 46370 }, { "grad_norm": 0.3800300657749176, "learning_rate": 4.099866432945595e-05, "loss": 0.0244, "step": 46380 }, { "grad_norm": 0.3731664717197418, "learning_rate": 4.097833441198346e-05, "loss": 0.0251, "step": 46390 }, { "grad_norm": 0.19698624312877655, "learning_rate": 4.095800603606685e-05, "loss": 0.0187, "step": 46400 }, { "grad_norm": 0.28434091806411743, "learning_rate": 4.093767920517975e-05, "loss": 0.0239, "step": 46410 }, { "grad_norm": 0.39680299162864685, "learning_rate": 4.091735392279539e-05, "loss": 0.0283, "step": 46420 }, { "grad_norm": 0.2424388825893402, "learning_rate": 4.089703019238685e-05, "loss": 0.0184, "step": 46430 }, { "grad_norm": 0.2136172205209732, "learning_rate": 4.0876708017426866e-05, "loss": 0.0227, "step": 46440 }, { "grad_norm": 0.30589842796325684, "learning_rate": 4.085638740138798e-05, "loss": 0.0182, "step": 46450 }, { "grad_norm": 0.32434186339378357, "learning_rate": 4.08360683477424e-05, "loss": 0.0234, "step": 46460 }, { "grad_norm": 0.25061988830566406, "learning_rate": 4.0815750859962085e-05, "loss": 0.0244, "step": 46470 }, { "grad_norm": 0.3019554913043976, "learning_rate": 4.079543494151879e-05, "loss": 0.0257, "step": 46480 }, { "grad_norm": 0.2884865701198578, "learning_rate": 4.07751205958839e-05, "loss": 0.0215, "step": 46490 }, { "grad_norm": 0.20994485914707184, "learning_rate": 4.0754807826528615e-05, "loss": 0.0187, "step": 46500 }, { "grad_norm": 0.34368669986724854, "learning_rate": 4.073449663692379e-05, "loss": 0.0272, "step": 46510 }, { "grad_norm": 0.2832493185997009, "learning_rate": 4.07141870305401e-05, "loss": 0.0206, "step": 46520 }, { "grad_norm": 0.3599371612071991, "learning_rate": 4.0693879010847866e-05, "loss": 0.0368, "step": 46530 }, { "grad_norm": 0.3392722010612488, "learning_rate": 4.06735725813172e-05, "loss": 0.0236, "step": 46540 }, { "grad_norm": 0.2022201120853424, "learning_rate": 4.0653267745417903e-05, "loss": 0.0149, "step": 46550 }, { "grad_norm": 0.27698513865470886, "learning_rate": 4.063296450661949e-05, "loss": 0.019, "step": 46560 }, { "grad_norm": 0.2187720686197281, "learning_rate": 4.061266286839128e-05, "loss": 0.0186, "step": 46570 }, { "grad_norm": 0.3092033267021179, "learning_rate": 4.0592362834202225e-05, "loss": 0.0281, "step": 46580 }, { "grad_norm": 0.2571254372596741, "learning_rate": 4.057206440752107e-05, "loss": 0.0179, "step": 46590 }, { "grad_norm": 0.2516120970249176, "learning_rate": 4.0551767591816245e-05, "loss": 0.0194, "step": 46600 }, { "grad_norm": 0.17532962560653687, "learning_rate": 4.0531472390555935e-05, "loss": 0.0277, "step": 46610 }, { "grad_norm": 0.5876168608665466, "learning_rate": 4.051117880720802e-05, "loss": 0.0204, "step": 46620 }, { "grad_norm": 0.24234862625598907, "learning_rate": 4.049088684524015e-05, "loss": 0.0194, "step": 46630 }, { "grad_norm": 0.39655354619026184, "learning_rate": 4.0470596508119636e-05, "loss": 0.0235, "step": 46640 }, { "grad_norm": 0.18564891815185547, "learning_rate": 4.0450307799313524e-05, "loss": 0.0267, "step": 46650 }, { "grad_norm": 0.3445281982421875, "learning_rate": 4.0430020722288656e-05, "loss": 0.0238, "step": 46660 }, { "grad_norm": 0.4318860173225403, "learning_rate": 4.040973528051148e-05, "loss": 0.0216, "step": 46670 }, { "grad_norm": 0.30288422107696533, "learning_rate": 4.038945147744827e-05, "loss": 0.0224, "step": 46680 }, { "grad_norm": 0.15233217179775238, "learning_rate": 4.0369169316564945e-05, "loss": 0.0216, "step": 46690 }, { "grad_norm": 0.29078209400177, "learning_rate": 4.034888880132718e-05, "loss": 0.017, "step": 46700 }, { "grad_norm": 0.37923553586006165, "learning_rate": 4.032860993520035e-05, "loss": 0.0251, "step": 46710 }, { "grad_norm": 0.3019581437110901, "learning_rate": 4.030833272164959e-05, "loss": 0.0168, "step": 46720 }, { "grad_norm": 0.2602010667324066, "learning_rate": 4.0288057164139705e-05, "loss": 0.0251, "step": 46730 }, { "grad_norm": 0.23188534379005432, "learning_rate": 4.0267783266135205e-05, "loss": 0.0192, "step": 46740 }, { "grad_norm": 0.16970747709274292, "learning_rate": 4.024751103110039e-05, "loss": 0.0234, "step": 46750 }, { "grad_norm": 0.18962892889976501, "learning_rate": 4.0227240462499176e-05, "loss": 0.0282, "step": 46760 }, { "grad_norm": 0.23533107340335846, "learning_rate": 4.020697156379531e-05, "loss": 0.0272, "step": 46770 }, { "grad_norm": 0.22553375363349915, "learning_rate": 4.018670433845215e-05, "loss": 0.0219, "step": 46780 }, { "grad_norm": 0.16150103509426117, "learning_rate": 4.016643878993284e-05, "loss": 0.0175, "step": 46790 }, { "grad_norm": 0.526844322681427, "learning_rate": 4.014617492170017e-05, "loss": 0.0271, "step": 46800 }, { "grad_norm": 0.29191938042640686, "learning_rate": 4.0125912737216726e-05, "loss": 0.0382, "step": 46810 }, { "grad_norm": 0.3382125794887543, "learning_rate": 4.0105652239944735e-05, "loss": 0.0285, "step": 46820 }, { "grad_norm": 0.27014800906181335, "learning_rate": 4.0085393433346144e-05, "loss": 0.0239, "step": 46830 }, { "grad_norm": 0.28690263628959656, "learning_rate": 4.006513632088268e-05, "loss": 0.0186, "step": 46840 }, { "grad_norm": 0.2109595537185669, "learning_rate": 4.004488090601567e-05, "loss": 0.0187, "step": 46850 }, { "grad_norm": 0.2724420726299286, "learning_rate": 4.002462719220626e-05, "loss": 0.0185, "step": 46860 }, { "grad_norm": 0.21111327409744263, "learning_rate": 4.000437518291522e-05, "loss": 0.0231, "step": 46870 }, { "grad_norm": 0.2749210596084595, "learning_rate": 3.9984124881603094e-05, "loss": 0.0232, "step": 46880 }, { "grad_norm": 0.26433831453323364, "learning_rate": 3.9963876291730086e-05, "loss": 0.0188, "step": 46890 }, { "grad_norm": 0.28925806283950806, "learning_rate": 3.994362941675614e-05, "loss": 0.0205, "step": 46900 }, { "grad_norm": 0.26469671726226807, "learning_rate": 3.992338426014088e-05, "loss": 0.0254, "step": 46910 }, { "grad_norm": 0.34102416038513184, "learning_rate": 3.9903140825343636e-05, "loss": 0.0212, "step": 46920 }, { "grad_norm": 0.26942333579063416, "learning_rate": 3.98828991158235e-05, "loss": 0.0209, "step": 46930 }, { "grad_norm": 0.23008441925048828, "learning_rate": 3.9862659135039185e-05, "loss": 0.0257, "step": 46940 }, { "grad_norm": 0.26398804783821106, "learning_rate": 3.984242088644918e-05, "loss": 0.0228, "step": 46950 }, { "grad_norm": 0.26420384645462036, "learning_rate": 3.9822184373511615e-05, "loss": 0.0208, "step": 46960 }, { "grad_norm": 0.28271347284317017, "learning_rate": 3.98019495996844e-05, "loss": 0.0244, "step": 46970 }, { "grad_norm": 0.26913273334503174, "learning_rate": 3.978171656842507e-05, "loss": 0.0208, "step": 46980 }, { "grad_norm": 0.38938724994659424, "learning_rate": 3.976148528319091e-05, "loss": 0.0214, "step": 46990 }, { "grad_norm": 0.30042654275894165, "learning_rate": 3.974125574743888e-05, "loss": 0.0231, "step": 47000 }, { "grad_norm": 0.43268921971321106, "learning_rate": 3.9721027964625686e-05, "loss": 0.0245, "step": 47010 }, { "grad_norm": 0.3095177710056305, "learning_rate": 3.9700801938207676e-05, "loss": 0.0214, "step": 47020 }, { "grad_norm": 0.251695841550827, "learning_rate": 3.9680577671640916e-05, "loss": 0.0171, "step": 47030 }, { "grad_norm": 0.2930164039134979, "learning_rate": 3.966035516838121e-05, "loss": 0.0269, "step": 47040 }, { "grad_norm": 0.25976696610450745, "learning_rate": 3.9640134431884014e-05, "loss": 0.0182, "step": 47050 }, { "grad_norm": 0.3545687794685364, "learning_rate": 3.961991546560451e-05, "loss": 0.0175, "step": 47060 }, { "grad_norm": 0.263517290353775, "learning_rate": 3.959969827299753e-05, "loss": 0.0218, "step": 47070 }, { "grad_norm": 0.3291858732700348, "learning_rate": 3.9579482857517684e-05, "loss": 0.0245, "step": 47080 }, { "grad_norm": 0.34343862533569336, "learning_rate": 3.955926922261921e-05, "loss": 0.0191, "step": 47090 }, { "grad_norm": 0.24798519909381866, "learning_rate": 3.9539057371756084e-05, "loss": 0.0211, "step": 47100 }, { "grad_norm": 0.3306889832019806, "learning_rate": 3.951884730838195e-05, "loss": 0.0152, "step": 47110 }, { "grad_norm": 0.25190815329551697, "learning_rate": 3.949863903595012e-05, "loss": 0.0183, "step": 47120 }, { "grad_norm": 0.32540932297706604, "learning_rate": 3.947843255791369e-05, "loss": 0.0198, "step": 47130 }, { "grad_norm": 0.3936443626880646, "learning_rate": 3.9458227877725364e-05, "loss": 0.024, "step": 47140 }, { "grad_norm": 0.2329408824443817, "learning_rate": 3.943802499883758e-05, "loss": 0.02, "step": 47150 }, { "grad_norm": 0.357953816652298, "learning_rate": 3.9417823924702437e-05, "loss": 0.0211, "step": 47160 }, { "grad_norm": 0.3790017068386078, "learning_rate": 3.939762465877178e-05, "loss": 0.0261, "step": 47170 }, { "grad_norm": 0.3156682848930359, "learning_rate": 3.937742720449708e-05, "loss": 0.0219, "step": 47180 }, { "grad_norm": 0.25383758544921875, "learning_rate": 3.9357231565329563e-05, "loss": 0.0192, "step": 47190 }, { "grad_norm": 0.2026115208864212, "learning_rate": 3.933703774472008e-05, "loss": 0.0165, "step": 47200 }, { "grad_norm": 0.14997629821300507, "learning_rate": 3.93168457461192e-05, "loss": 0.0158, "step": 47210 }, { "grad_norm": 0.28846773505210876, "learning_rate": 3.9296655572977216e-05, "loss": 0.0212, "step": 47220 }, { "grad_norm": 0.35917362570762634, "learning_rate": 3.927646722874404e-05, "loss": 0.0264, "step": 47230 }, { "grad_norm": 0.35652783513069153, "learning_rate": 3.925628071686934e-05, "loss": 0.0237, "step": 47240 }, { "grad_norm": 0.2946647107601166, "learning_rate": 3.9236096040802415e-05, "loss": 0.0218, "step": 47250 }, { "grad_norm": 0.46895062923431396, "learning_rate": 3.9215913203992294e-05, "loss": 0.0319, "step": 47260 }, { "grad_norm": 0.21683433651924133, "learning_rate": 3.9195732209887645e-05, "loss": 0.0199, "step": 47270 }, { "grad_norm": 0.25822699069976807, "learning_rate": 3.9175553061936875e-05, "loss": 0.0248, "step": 47280 }, { "grad_norm": 0.23962348699569702, "learning_rate": 3.9155375763588045e-05, "loss": 0.0273, "step": 47290 }, { "grad_norm": 0.24835850298404694, "learning_rate": 3.913520031828889e-05, "loss": 0.0197, "step": 47300 }, { "grad_norm": 0.4568828046321869, "learning_rate": 3.911502672948685e-05, "loss": 0.0257, "step": 47310 }, { "grad_norm": 0.37536385655403137, "learning_rate": 3.9094855000629014e-05, "loss": 0.0252, "step": 47320 }, { "grad_norm": 0.30294719338417053, "learning_rate": 3.907468513516223e-05, "loss": 0.0196, "step": 47330 }, { "grad_norm": 0.31910407543182373, "learning_rate": 3.905451713653294e-05, "loss": 0.0187, "step": 47340 }, { "grad_norm": 0.1983843296766281, "learning_rate": 3.903435100818731e-05, "loss": 0.0232, "step": 47350 }, { "grad_norm": 0.3711545467376709, "learning_rate": 3.901418675357117e-05, "loss": 0.0234, "step": 47360 }, { "grad_norm": 0.29337239265441895, "learning_rate": 3.8994024376130075e-05, "loss": 0.0285, "step": 47370 }, { "grad_norm": 0.5131166577339172, "learning_rate": 3.8973863879309194e-05, "loss": 0.0236, "step": 47380 }, { "grad_norm": 0.3093138337135315, "learning_rate": 3.8953705266553394e-05, "loss": 0.0167, "step": 47390 }, { "grad_norm": 0.32662200927734375, "learning_rate": 3.893354854130727e-05, "loss": 0.0241, "step": 47400 }, { "grad_norm": 0.2958597242832184, "learning_rate": 3.8913393707015006e-05, "loss": 0.0204, "step": 47410 }, { "grad_norm": 0.3120728135108948, "learning_rate": 3.889324076712056e-05, "loss": 0.0264, "step": 47420 }, { "grad_norm": 0.31204962730407715, "learning_rate": 3.8873089725067476e-05, "loss": 0.0265, "step": 47430 }, { "grad_norm": 0.2932623326778412, "learning_rate": 3.885294058429905e-05, "loss": 0.0335, "step": 47440 }, { "grad_norm": 0.2054392695426941, "learning_rate": 3.8832793348258206e-05, "loss": 0.0229, "step": 47450 }, { "grad_norm": 0.3673732280731201, "learning_rate": 3.881264802038756e-05, "loss": 0.0368, "step": 47460 }, { "grad_norm": 0.4018521308898926, "learning_rate": 3.879250460412939e-05, "loss": 0.0284, "step": 47470 }, { "grad_norm": 0.29849886894226074, "learning_rate": 3.8772363102925644e-05, "loss": 0.0267, "step": 47480 }, { "grad_norm": 0.2974206507205963, "learning_rate": 3.875222352021798e-05, "loss": 0.0182, "step": 47490 }, { "grad_norm": 2.4658007621765137, "learning_rate": 3.8732085859447686e-05, "loss": 0.0259, "step": 47500 }, { "grad_norm": 0.3479096591472626, "learning_rate": 3.871195012405575e-05, "loss": 0.0221, "step": 47510 }, { "grad_norm": 0.2721080780029297, "learning_rate": 3.869181631748278e-05, "loss": 0.0269, "step": 47520 }, { "grad_norm": 0.4228963255882263, "learning_rate": 3.867168444316915e-05, "loss": 0.0212, "step": 47530 }, { "grad_norm": 0.1534755378961563, "learning_rate": 3.865155450455481e-05, "loss": 0.0186, "step": 47540 }, { "grad_norm": 0.31606051325798035, "learning_rate": 3.8631426505079426e-05, "loss": 0.0207, "step": 47550 }, { "grad_norm": 0.33198392391204834, "learning_rate": 3.8611300448182304e-05, "loss": 0.0329, "step": 47560 }, { "grad_norm": 0.2618497312068939, "learning_rate": 3.859117633730248e-05, "loss": 0.0222, "step": 47570 }, { "grad_norm": 0.3673458397388458, "learning_rate": 3.857105417587858e-05, "loss": 0.0253, "step": 47580 }, { "grad_norm": 0.3053150177001953, "learning_rate": 3.855093396734894e-05, "loss": 0.0203, "step": 47590 }, { "grad_norm": 0.27541881799697876, "learning_rate": 3.8530815715151545e-05, "loss": 0.0327, "step": 47600 }, { "grad_norm": 0.22218018770217896, "learning_rate": 3.851069942272405e-05, "loss": 0.019, "step": 47610 }, { "grad_norm": 0.33050602674484253, "learning_rate": 3.849058509350382e-05, "loss": 0.0261, "step": 47620 }, { "grad_norm": 0.2555914521217346, "learning_rate": 3.8470472730927783e-05, "loss": 0.0307, "step": 47630 }, { "grad_norm": 0.2711887061595917, "learning_rate": 3.845036233843264e-05, "loss": 0.0171, "step": 47640 }, { "grad_norm": 0.2239241898059845, "learning_rate": 3.843025391945469e-05, "loss": 0.0219, "step": 47650 }, { "grad_norm": 0.2881619930267334, "learning_rate": 3.841014747742992e-05, "loss": 0.0209, "step": 47660 }, { "grad_norm": 0.20424020290374756, "learning_rate": 3.839004301579397e-05, "loss": 0.0183, "step": 47670 }, { "grad_norm": 0.19991369545459747, "learning_rate": 3.8369940537982097e-05, "loss": 0.0259, "step": 47680 }, { "grad_norm": 0.2717697322368622, "learning_rate": 3.834984004742933e-05, "loss": 0.0173, "step": 47690 }, { "grad_norm": 0.2241794466972351, "learning_rate": 3.832974154757026e-05, "loss": 0.02, "step": 47700 }, { "grad_norm": 0.3248586654663086, "learning_rate": 3.830964504183919e-05, "loss": 0.0211, "step": 47710 }, { "grad_norm": 0.3545602560043335, "learning_rate": 3.828955053367003e-05, "loss": 0.0221, "step": 47720 }, { "grad_norm": 0.4555385410785675, "learning_rate": 3.826945802649642e-05, "loss": 0.0246, "step": 47730 }, { "grad_norm": 0.22813618183135986, "learning_rate": 3.824936752375159e-05, "loss": 0.0345, "step": 47740 }, { "grad_norm": 0.27401432394981384, "learning_rate": 3.822927902886848e-05, "loss": 0.0216, "step": 47750 }, { "grad_norm": 0.2617412805557251, "learning_rate": 3.8209192545279653e-05, "loss": 0.0206, "step": 47760 }, { "grad_norm": 0.1471312940120697, "learning_rate": 3.8189108076417326e-05, "loss": 0.0214, "step": 47770 }, { "grad_norm": 0.2610427141189575, "learning_rate": 3.816902562571342e-05, "loss": 0.027, "step": 47780 }, { "grad_norm": 0.28422123193740845, "learning_rate": 3.814894519659944e-05, "loss": 0.0196, "step": 47790 }, { "grad_norm": 0.2580568194389343, "learning_rate": 3.812886679250661e-05, "loss": 0.0182, "step": 47800 }, { "grad_norm": 0.3604111075401306, "learning_rate": 3.810879041686575e-05, "loss": 0.0206, "step": 47810 }, { "grad_norm": 0.3965678811073303, "learning_rate": 3.808871607310741e-05, "loss": 0.0266, "step": 47820 }, { "grad_norm": 0.2777230739593506, "learning_rate": 3.806864376466169e-05, "loss": 0.0234, "step": 47830 }, { "grad_norm": 0.17248007655143738, "learning_rate": 3.804857349495845e-05, "loss": 0.0176, "step": 47840 }, { "grad_norm": 0.3690606355667114, "learning_rate": 3.802850526742713e-05, "loss": 0.0369, "step": 47850 }, { "grad_norm": 0.26469936966896057, "learning_rate": 3.800843908549683e-05, "loss": 0.0197, "step": 47860 }, { "grad_norm": 0.34836581349372864, "learning_rate": 3.7988374952596325e-05, "loss": 0.0203, "step": 47870 }, { "grad_norm": 0.3594931662082672, "learning_rate": 3.7968312872154e-05, "loss": 0.0218, "step": 47880 }, { "grad_norm": 0.2943393588066101, "learning_rate": 3.7948252847597965e-05, "loss": 0.0225, "step": 47890 }, { "grad_norm": 0.250360906124115, "learning_rate": 3.7928194882355885e-05, "loss": 0.0239, "step": 47900 }, { "grad_norm": 0.22940552234649658, "learning_rate": 3.790813897985515e-05, "loss": 0.0226, "step": 47910 }, { "grad_norm": 0.3430441915988922, "learning_rate": 3.7888085143522726e-05, "loss": 0.0226, "step": 47920 }, { "grad_norm": 0.28641682863235474, "learning_rate": 3.7868033376785314e-05, "loss": 0.0237, "step": 47930 }, { "grad_norm": 0.367031067609787, "learning_rate": 3.784798368306919e-05, "loss": 0.0218, "step": 47940 }, { "grad_norm": 0.2015407383441925, "learning_rate": 3.782793606580029e-05, "loss": 0.0213, "step": 47950 }, { "grad_norm": 0.1843338906764984, "learning_rate": 3.7807890528404205e-05, "loss": 0.0267, "step": 47960 }, { "grad_norm": 0.26803284883499146, "learning_rate": 3.778784707430616e-05, "loss": 0.0189, "step": 47970 }, { "grad_norm": 0.3221442699432373, "learning_rate": 3.776780570693107e-05, "loss": 0.0183, "step": 47980 }, { "grad_norm": 0.25919783115386963, "learning_rate": 3.774776642970342e-05, "loss": 0.0268, "step": 47990 }, { "grad_norm": 0.23558972775936127, "learning_rate": 3.77277292460474e-05, "loss": 0.0181, "step": 48000 }, { "grad_norm": 0.31064510345458984, "learning_rate": 3.770769415938678e-05, "loss": 0.0207, "step": 48010 }, { "grad_norm": 0.2685403525829315, "learning_rate": 3.768766117314506e-05, "loss": 0.0288, "step": 48020 }, { "grad_norm": 0.2474498301744461, "learning_rate": 3.76676302907453e-05, "loss": 0.0197, "step": 48030 }, { "grad_norm": 0.358025461435318, "learning_rate": 3.764760151561021e-05, "loss": 0.0276, "step": 48040 }, { "grad_norm": 0.3407188951969147, "learning_rate": 3.76275748511622e-05, "loss": 0.0206, "step": 48050 }, { "grad_norm": 0.23531390726566315, "learning_rate": 3.7607550300823255e-05, "loss": 0.022, "step": 48060 }, { "grad_norm": 0.4113655388355255, "learning_rate": 3.7587527868015044e-05, "loss": 0.0393, "step": 48070 }, { "grad_norm": 0.3507985472679138, "learning_rate": 3.756750755615881e-05, "loss": 0.0282, "step": 48080 }, { "grad_norm": 0.1650751680135727, "learning_rate": 3.754748936867553e-05, "loss": 0.0213, "step": 48090 }, { "grad_norm": 0.37003859877586365, "learning_rate": 3.752747330898573e-05, "loss": 0.0217, "step": 48100 }, { "grad_norm": 0.2119390368461609, "learning_rate": 3.750745938050962e-05, "loss": 0.0232, "step": 48110 }, { "grad_norm": 0.19956359267234802, "learning_rate": 3.7487447586667025e-05, "loss": 0.0217, "step": 48120 }, { "grad_norm": 0.3031868636608124, "learning_rate": 3.7467437930877426e-05, "loss": 0.0183, "step": 48130 }, { "grad_norm": 0.20353718101978302, "learning_rate": 3.744743041655992e-05, "loss": 0.0183, "step": 48140 }, { "grad_norm": 0.27036070823669434, "learning_rate": 3.742742504713324e-05, "loss": 0.0221, "step": 48150 }, { "grad_norm": 0.2923623025417328, "learning_rate": 3.740742182601576e-05, "loss": 0.0242, "step": 48160 }, { "grad_norm": 0.21619100868701935, "learning_rate": 3.7387420756625464e-05, "loss": 0.0261, "step": 48170 }, { "grad_norm": 0.2742885649204254, "learning_rate": 3.736742184238002e-05, "loss": 0.0251, "step": 48180 }, { "grad_norm": 0.24223898351192474, "learning_rate": 3.7347425086696684e-05, "loss": 0.0233, "step": 48190 }, { "grad_norm": 0.38439831137657166, "learning_rate": 3.732743049299235e-05, "loss": 0.026, "step": 48200 }, { "grad_norm": 0.2933253347873688, "learning_rate": 3.730743806468354e-05, "loss": 0.017, "step": 48210 }, { "grad_norm": 0.2877168357372284, "learning_rate": 3.7287447805186436e-05, "loss": 0.0166, "step": 48220 }, { "grad_norm": 0.396244078874588, "learning_rate": 3.726745971791682e-05, "loss": 0.0309, "step": 48230 }, { "grad_norm": 0.32216861844062805, "learning_rate": 3.724747380629008e-05, "loss": 0.0335, "step": 48240 }, { "grad_norm": 0.31365519762039185, "learning_rate": 3.72274900737213e-05, "loss": 0.0215, "step": 48250 }, { "grad_norm": 0.2730153203010559, "learning_rate": 3.7207508523625123e-05, "loss": 0.0212, "step": 48260 }, { "grad_norm": 0.24762362241744995, "learning_rate": 3.718752915941588e-05, "loss": 0.0187, "step": 48270 }, { "grad_norm": 0.4366445243358612, "learning_rate": 3.7167551984507464e-05, "loss": 0.0237, "step": 48280 }, { "grad_norm": 0.29231148958206177, "learning_rate": 3.714757700231346e-05, "loss": 0.0223, "step": 48290 }, { "grad_norm": 0.2776114344596863, "learning_rate": 3.712760421624703e-05, "loss": 0.0218, "step": 48300 }, { "grad_norm": 0.19899360835552216, "learning_rate": 3.710763362972099e-05, "loss": 0.0219, "step": 48310 }, { "grad_norm": 0.24481485784053802, "learning_rate": 3.708766524614774e-05, "loss": 0.0188, "step": 48320 }, { "grad_norm": 0.2351902574300766, "learning_rate": 3.7067699068939335e-05, "loss": 0.017, "step": 48330 }, { "grad_norm": 0.25080806016921997, "learning_rate": 3.704773510150748e-05, "loss": 0.0212, "step": 48340 }, { "grad_norm": 0.32836779952049255, "learning_rate": 3.702777334726344e-05, "loss": 0.0242, "step": 48350 }, { "grad_norm": 0.23645159602165222, "learning_rate": 3.7007813809618164e-05, "loss": 0.0157, "step": 48360 }, { "grad_norm": 0.19315826892852783, "learning_rate": 3.6987856491982145e-05, "loss": 0.0265, "step": 48370 }, { "grad_norm": 0.26376017928123474, "learning_rate": 3.696790139776558e-05, "loss": 0.0239, "step": 48380 }, { "grad_norm": 0.19302091002464294, "learning_rate": 3.6947948530378235e-05, "loss": 0.0193, "step": 48390 }, { "grad_norm": 0.3777017891407013, "learning_rate": 3.6927997893229516e-05, "loss": 0.0212, "step": 48400 }, { "grad_norm": 0.30563539266586304, "learning_rate": 3.690804948972845e-05, "loss": 0.018, "step": 48410 }, { "grad_norm": 0.17830199003219604, "learning_rate": 3.688810332328363e-05, "loss": 0.0206, "step": 48420 }, { "grad_norm": 0.2773894667625427, "learning_rate": 3.686815939730336e-05, "loss": 0.0275, "step": 48430 }, { "grad_norm": 0.2702781856060028, "learning_rate": 3.684821771519548e-05, "loss": 0.0155, "step": 48440 }, { "grad_norm": 0.2962709069252014, "learning_rate": 3.68282782803675e-05, "loss": 0.0176, "step": 48450 }, { "grad_norm": 0.3157820403575897, "learning_rate": 3.6808341096226504e-05, "loss": 0.024, "step": 48460 }, { "grad_norm": 0.3706531822681427, "learning_rate": 3.678840616617924e-05, "loss": 0.0229, "step": 48470 }, { "grad_norm": 0.3081062436103821, "learning_rate": 3.6768473493632e-05, "loss": 0.0213, "step": 48480 }, { "grad_norm": 0.21434856951236725, "learning_rate": 3.6748543081990783e-05, "loss": 0.0184, "step": 48490 }, { "grad_norm": 0.24063856899738312, "learning_rate": 3.672861493466112e-05, "loss": 0.0214, "step": 48500 }, { "grad_norm": 0.2944867014884949, "learning_rate": 3.670868905504818e-05, "loss": 0.0187, "step": 48510 }, { "grad_norm": 0.28492963314056396, "learning_rate": 3.6688765446556784e-05, "loss": 0.0185, "step": 48520 }, { "grad_norm": 0.2512575089931488, "learning_rate": 3.6668844112591276e-05, "loss": 0.0196, "step": 48530 }, { "grad_norm": 0.2576994299888611, "learning_rate": 3.664892505655573e-05, "loss": 0.0189, "step": 48540 }, { "grad_norm": 0.3723800778388977, "learning_rate": 3.662900828185373e-05, "loss": 0.0231, "step": 48550 }, { "grad_norm": 0.24340149760246277, "learning_rate": 3.6609093791888516e-05, "loss": 0.0199, "step": 48560 }, { "grad_norm": 0.28157293796539307, "learning_rate": 3.658918159006292e-05, "loss": 0.0272, "step": 48570 }, { "grad_norm": 0.23568095266819, "learning_rate": 3.656927167977942e-05, "loss": 0.0218, "step": 48580 }, { "grad_norm": 0.24957183003425598, "learning_rate": 3.654936406444006e-05, "loss": 0.0205, "step": 48590 }, { "grad_norm": 0.28642722964286804, "learning_rate": 3.65294587474465e-05, "loss": 0.0203, "step": 48600 }, { "grad_norm": 0.3246328830718994, "learning_rate": 3.650955573220002e-05, "loss": 0.0499, "step": 48610 }, { "grad_norm": 0.3639640808105469, "learning_rate": 3.648965502210149e-05, "loss": 0.0213, "step": 48620 }, { "grad_norm": 0.4581015706062317, "learning_rate": 3.646975662055142e-05, "loss": 0.0217, "step": 48630 }, { "grad_norm": 0.17246831953525543, "learning_rate": 3.644986053094987e-05, "loss": 0.0194, "step": 48640 }, { "grad_norm": 0.19303227961063385, "learning_rate": 3.642996675669659e-05, "loss": 0.0224, "step": 48650 }, { "grad_norm": 0.1847536414861679, "learning_rate": 3.641007530119083e-05, "loss": 0.0171, "step": 48660 }, { "grad_norm": 0.30275169014930725, "learning_rate": 3.639018616783153e-05, "loss": 0.0261, "step": 48670 }, { "grad_norm": 0.2623416781425476, "learning_rate": 3.637029936001719e-05, "loss": 0.0168, "step": 48680 }, { "grad_norm": 0.2753441333770752, "learning_rate": 3.6350414881145886e-05, "loss": 0.0275, "step": 48690 }, { "grad_norm": 0.1693595051765442, "learning_rate": 3.6330532734615386e-05, "loss": 0.0211, "step": 48700 }, { "grad_norm": 0.316977322101593, "learning_rate": 3.6310652923822975e-05, "loss": 0.0243, "step": 48710 }, { "grad_norm": 0.20587317645549774, "learning_rate": 3.629077545216558e-05, "loss": 0.0213, "step": 48720 }, { "grad_norm": 0.2984674572944641, "learning_rate": 3.627090032303969e-05, "loss": 0.0299, "step": 48730 }, { "grad_norm": 0.3287566602230072, "learning_rate": 3.625102753984146e-05, "loss": 0.0239, "step": 48740 }, { "grad_norm": 0.2817630469799042, "learning_rate": 3.623115710596659e-05, "loss": 0.0165, "step": 48750 }, { "grad_norm": 0.4280964732170105, "learning_rate": 3.6211289024810395e-05, "loss": 0.0275, "step": 48760 }, { "grad_norm": 0.22215993702411652, "learning_rate": 3.619142329976777e-05, "loss": 0.0174, "step": 48770 }, { "grad_norm": 0.29095202684402466, "learning_rate": 3.6171559934233247e-05, "loss": 0.0157, "step": 48780 }, { "grad_norm": 0.2740655541419983, "learning_rate": 3.615169893160093e-05, "loss": 0.0226, "step": 48790 }, { "grad_norm": 0.27272409200668335, "learning_rate": 3.61318402952645e-05, "loss": 0.0178, "step": 48800 }, { "grad_norm": 0.2693926692008972, "learning_rate": 3.6111984028617285e-05, "loss": 0.0214, "step": 48810 }, { "grad_norm": 0.35581204295158386, "learning_rate": 3.6092130135052134e-05, "loss": 0.0189, "step": 48820 }, { "grad_norm": 0.2482585906982422, "learning_rate": 3.6072278617961584e-05, "loss": 0.0278, "step": 48830 }, { "grad_norm": 0.35528862476348877, "learning_rate": 3.605242948073767e-05, "loss": 0.0236, "step": 48840 }, { "grad_norm": 0.39342960715293884, "learning_rate": 3.603258272677212e-05, "loss": 0.0208, "step": 48850 }, { "grad_norm": 0.28251487016677856, "learning_rate": 3.601273835945616e-05, "loss": 0.0227, "step": 48860 }, { "grad_norm": 0.22929637134075165, "learning_rate": 3.5992896382180664e-05, "loss": 0.0235, "step": 48870 }, { "grad_norm": 0.2831234931945801, "learning_rate": 3.597305679833609e-05, "loss": 0.0257, "step": 48880 }, { "grad_norm": 0.25330057740211487, "learning_rate": 3.595321961131245e-05, "loss": 0.0241, "step": 48890 }, { "grad_norm": 0.3774259388446808, "learning_rate": 3.593338482449942e-05, "loss": 0.027, "step": 48900 }, { "grad_norm": 0.27362877130508423, "learning_rate": 3.591355244128618e-05, "loss": 0.0318, "step": 48910 }, { "grad_norm": 0.3372771739959717, "learning_rate": 3.589372246506158e-05, "loss": 0.0267, "step": 48920 }, { "grad_norm": 0.19465133547782898, "learning_rate": 3.5873894899213984e-05, "loss": 0.0187, "step": 48930 }, { "grad_norm": 0.2567785680294037, "learning_rate": 3.5854069747131416e-05, "loss": 0.0171, "step": 48940 }, { "grad_norm": 0.34313517808914185, "learning_rate": 3.583424701220143e-05, "loss": 0.0223, "step": 48950 }, { "grad_norm": 0.3018667697906494, "learning_rate": 3.581442669781121e-05, "loss": 0.0194, "step": 48960 }, { "grad_norm": 0.35776039958000183, "learning_rate": 3.579460880734749e-05, "loss": 0.0225, "step": 48970 }, { "grad_norm": 0.27872753143310547, "learning_rate": 3.577479334419657e-05, "loss": 0.0193, "step": 48980 }, { "grad_norm": 0.3362092673778534, "learning_rate": 3.575498031174444e-05, "loss": 0.0229, "step": 48990 }, { "grad_norm": 0.27504241466522217, "learning_rate": 3.573516971337657e-05, "loss": 0.0216, "step": 49000 }, { "grad_norm": 0.33757275342941284, "learning_rate": 3.5715361552478046e-05, "loss": 0.032, "step": 49010 }, { "grad_norm": 0.22168105840682983, "learning_rate": 3.5695555832433536e-05, "loss": 0.0205, "step": 49020 }, { "grad_norm": 0.39095187187194824, "learning_rate": 3.5675752556627325e-05, "loss": 0.0208, "step": 49030 }, { "grad_norm": 0.3441425561904907, "learning_rate": 3.565595172844322e-05, "loss": 0.0229, "step": 49040 }, { "grad_norm": 0.2544497549533844, "learning_rate": 3.5636153351264666e-05, "loss": 0.0243, "step": 49050 }, { "grad_norm": 0.3155495226383209, "learning_rate": 3.5616357428474655e-05, "loss": 0.0172, "step": 49060 }, { "grad_norm": 0.31049782037734985, "learning_rate": 3.559656396345575e-05, "loss": 0.0242, "step": 49070 }, { "grad_norm": 0.45050331950187683, "learning_rate": 3.5576772959590146e-05, "loss": 0.026, "step": 49080 }, { "grad_norm": 0.40486788749694824, "learning_rate": 3.5556984420259545e-05, "loss": 0.0209, "step": 49090 }, { "grad_norm": 0.32313549518585205, "learning_rate": 3.5537198348845305e-05, "loss": 0.0212, "step": 49100 }, { "grad_norm": 0.18756932020187378, "learning_rate": 3.551741474872831e-05, "loss": 0.0211, "step": 49110 }, { "grad_norm": 0.24619893729686737, "learning_rate": 3.549763362328903e-05, "loss": 0.0222, "step": 49120 }, { "grad_norm": 0.2343413531780243, "learning_rate": 3.5477854975907515e-05, "loss": 0.0213, "step": 49130 }, { "grad_norm": 0.20828621089458466, "learning_rate": 3.5458078809963416e-05, "loss": 0.024, "step": 49140 }, { "grad_norm": 0.36299562454223633, "learning_rate": 3.543830512883594e-05, "loss": 0.0173, "step": 49150 }, { "grad_norm": 0.27720382809638977, "learning_rate": 3.5418533935903824e-05, "loss": 0.0194, "step": 49160 }, { "grad_norm": 0.38875526189804077, "learning_rate": 3.539876523454547e-05, "loss": 0.0232, "step": 49170 }, { "grad_norm": 0.27881762385368347, "learning_rate": 3.537899902813878e-05, "loss": 0.0187, "step": 49180 }, { "grad_norm": 0.25333163142204285, "learning_rate": 3.5359235320061293e-05, "loss": 0.0191, "step": 49190 }, { "grad_norm": 0.22632309794425964, "learning_rate": 3.533947411369003e-05, "loss": 0.0172, "step": 49200 }, { "grad_norm": 0.25333064794540405, "learning_rate": 3.53197154124017e-05, "loss": 0.0202, "step": 49210 }, { "grad_norm": 0.22178608179092407, "learning_rate": 3.52999592195725e-05, "loss": 0.0235, "step": 49220 }, { "grad_norm": 0.2974521815776825, "learning_rate": 3.5280205538578224e-05, "loss": 0.0205, "step": 49230 }, { "grad_norm": 0.3687826097011566, "learning_rate": 3.5260454372794236e-05, "loss": 0.021, "step": 49240 }, { "grad_norm": 0.2654675841331482, "learning_rate": 3.524070572559545e-05, "loss": 0.0226, "step": 49250 }, { "grad_norm": 0.27564895153045654, "learning_rate": 3.5220959600356395e-05, "loss": 0.0193, "step": 49260 }, { "grad_norm": 0.21957902610301971, "learning_rate": 3.5201216000451145e-05, "loss": 0.0197, "step": 49270 }, { "grad_norm": 0.3038572072982788, "learning_rate": 3.5181474929253335e-05, "loss": 0.0214, "step": 49280 }, { "grad_norm": 0.15476417541503906, "learning_rate": 3.516173639013615e-05, "loss": 0.0168, "step": 49290 }, { "grad_norm": 0.260944128036499, "learning_rate": 3.5142000386472406e-05, "loss": 0.0158, "step": 49300 }, { "grad_norm": 0.299479603767395, "learning_rate": 3.5122266921634427e-05, "loss": 0.0207, "step": 49310 }, { "grad_norm": 0.3627201020717621, "learning_rate": 3.510253599899413e-05, "loss": 0.022, "step": 49320 }, { "grad_norm": 0.2522619962692261, "learning_rate": 3.5082807621922965e-05, "loss": 0.0205, "step": 49330 }, { "grad_norm": 0.41757482290267944, "learning_rate": 3.506308179379201e-05, "loss": 0.0265, "step": 49340 }, { "grad_norm": 0.2385452836751938, "learning_rate": 3.5043358517971844e-05, "loss": 0.019, "step": 49350 }, { "grad_norm": 0.2571069002151489, "learning_rate": 3.502363779783264e-05, "loss": 0.0184, "step": 49360 }, { "grad_norm": 0.2527503967285156, "learning_rate": 3.500391963674415e-05, "loss": 0.0193, "step": 49370 }, { "grad_norm": 0.25816071033477783, "learning_rate": 3.4984204038075615e-05, "loss": 0.0245, "step": 49380 }, { "grad_norm": 0.2455865740776062, "learning_rate": 3.496449100519595e-05, "loss": 0.0276, "step": 49390 }, { "grad_norm": 0.21322113275527954, "learning_rate": 3.494478054147354e-05, "loss": 0.0185, "step": 49400 }, { "grad_norm": 0.3264799416065216, "learning_rate": 3.4925072650276395e-05, "loss": 0.0201, "step": 49410 }, { "grad_norm": 0.1988651305437088, "learning_rate": 3.4905367334972016e-05, "loss": 0.016, "step": 49420 }, { "grad_norm": 0.2780072093009949, "learning_rate": 3.488566459892752e-05, "loss": 0.0208, "step": 49430 }, { "grad_norm": 0.31079646944999695, "learning_rate": 3.4865964445509585e-05, "loss": 0.022, "step": 49440 }, { "grad_norm": 0.311672568321228, "learning_rate": 3.484626687808438e-05, "loss": 0.0198, "step": 49450 }, { "grad_norm": 0.3140128552913666, "learning_rate": 3.4826571900017735e-05, "loss": 0.0161, "step": 49460 }, { "grad_norm": 0.2104760855436325, "learning_rate": 3.480687951467495e-05, "loss": 0.0165, "step": 49470 }, { "grad_norm": 0.25801968574523926, "learning_rate": 3.4787189725420925e-05, "loss": 0.0214, "step": 49480 }, { "grad_norm": 0.20508979260921478, "learning_rate": 3.4767502535620086e-05, "loss": 0.0197, "step": 49490 }, { "grad_norm": 0.26177993416786194, "learning_rate": 3.474781794863648e-05, "loss": 0.0355, "step": 49500 }, { "grad_norm": 0.30667486786842346, "learning_rate": 3.472813596783363e-05, "loss": 0.0273, "step": 49510 }, { "grad_norm": 0.259531170129776, "learning_rate": 3.470845659657466e-05, "loss": 0.0232, "step": 49520 }, { "grad_norm": 0.250119149684906, "learning_rate": 3.468877983822223e-05, "loss": 0.0243, "step": 49530 }, { "grad_norm": 0.2664264142513275, "learning_rate": 3.466910569613855e-05, "loss": 0.0191, "step": 49540 }, { "grad_norm": 0.42799580097198486, "learning_rate": 3.464943417368542e-05, "loss": 0.0262, "step": 49550 }, { "grad_norm": 0.22634752094745636, "learning_rate": 3.462976527422415e-05, "loss": 0.0165, "step": 49560 }, { "grad_norm": 0.3465370237827301, "learning_rate": 3.461009900111562e-05, "loss": 0.0227, "step": 49570 }, { "grad_norm": 0.32960939407348633, "learning_rate": 3.459043535772023e-05, "loss": 0.0211, "step": 49580 }, { "grad_norm": 0.3818696439266205, "learning_rate": 3.4570774347398014e-05, "loss": 0.0242, "step": 49590 }, { "grad_norm": 0.390296071767807, "learning_rate": 3.4551115973508454e-05, "loss": 0.0243, "step": 49600 }, { "grad_norm": 0.4840227961540222, "learning_rate": 3.453146023941066e-05, "loss": 0.0278, "step": 49610 }, { "grad_norm": 0.29055431485176086, "learning_rate": 3.451180714846325e-05, "loss": 0.0236, "step": 49620 }, { "grad_norm": 0.34036388993263245, "learning_rate": 3.449215670402438e-05, "loss": 0.023, "step": 49630 }, { "grad_norm": 0.33027833700180054, "learning_rate": 3.447250890945181e-05, "loss": 0.0185, "step": 49640 }, { "grad_norm": 0.22598353028297424, "learning_rate": 3.4452863768102754e-05, "loss": 0.0247, "step": 49650 }, { "grad_norm": 0.2831927537918091, "learning_rate": 3.443322128333409e-05, "loss": 0.0266, "step": 49660 }, { "grad_norm": 0.28846287727355957, "learning_rate": 3.441358145850215e-05, "loss": 0.0264, "step": 49670 }, { "grad_norm": 0.2346785068511963, "learning_rate": 3.439394429696286e-05, "loss": 0.0221, "step": 49680 }, { "grad_norm": 0.22192664444446564, "learning_rate": 3.4374309802071644e-05, "loss": 0.0197, "step": 49690 }, { "grad_norm": 0.304168164730072, "learning_rate": 3.435467797718353e-05, "loss": 0.0243, "step": 49700 }, { "grad_norm": 0.22598357498645782, "learning_rate": 3.433504882565306e-05, "loss": 0.0213, "step": 49710 }, { "grad_norm": 0.3702962398529053, "learning_rate": 3.43154223508343e-05, "loss": 0.0269, "step": 49720 }, { "grad_norm": 0.19443650543689728, "learning_rate": 3.429579855608089e-05, "loss": 0.0229, "step": 49730 }, { "grad_norm": 0.3252553343772888, "learning_rate": 3.427617744474597e-05, "loss": 0.0199, "step": 49740 }, { "grad_norm": 0.25573232769966125, "learning_rate": 3.425655902018231e-05, "loss": 0.0231, "step": 49750 }, { "grad_norm": 0.2598764896392822, "learning_rate": 3.423694328574211e-05, "loss": 0.0199, "step": 49760 }, { "grad_norm": 0.24961604177951813, "learning_rate": 3.42173302447772e-05, "loss": 0.0192, "step": 49770 }, { "grad_norm": 0.2404632866382599, "learning_rate": 3.419771990063886e-05, "loss": 0.0163, "step": 49780 }, { "grad_norm": 0.31827300786972046, "learning_rate": 3.417811225667803e-05, "loss": 0.0172, "step": 49790 }, { "grad_norm": 0.2286568582057953, "learning_rate": 3.415850731624508e-05, "loss": 0.0212, "step": 49800 }, { "grad_norm": 0.32462626695632935, "learning_rate": 3.4138905082689945e-05, "loss": 0.0258, "step": 49810 }, { "grad_norm": 0.19047503173351288, "learning_rate": 3.4119305559362145e-05, "loss": 0.0276, "step": 49820 }, { "grad_norm": 0.29141202569007874, "learning_rate": 3.4099708749610684e-05, "loss": 0.018, "step": 49830 }, { "grad_norm": 0.3149436116218567, "learning_rate": 3.408011465678413e-05, "loss": 0.0249, "step": 49840 }, { "grad_norm": 0.3852950632572174, "learning_rate": 3.406052328423055e-05, "loss": 0.0267, "step": 49850 }, { "grad_norm": 0.29859814047813416, "learning_rate": 3.4040934635297615e-05, "loss": 0.0201, "step": 49860 }, { "grad_norm": 0.22420218586921692, "learning_rate": 3.4021348713332466e-05, "loss": 0.0178, "step": 49870 }, { "grad_norm": 0.2667616307735443, "learning_rate": 3.4001765521681807e-05, "loss": 0.0207, "step": 49880 }, { "grad_norm": 0.2922937870025635, "learning_rate": 3.398218506369188e-05, "loss": 0.0306, "step": 49890 }, { "grad_norm": 0.23557889461517334, "learning_rate": 3.3962607342708404e-05, "loss": 0.0217, "step": 49900 }, { "grad_norm": 0.3465516269207001, "learning_rate": 3.394303236207673e-05, "loss": 0.0192, "step": 49910 }, { "grad_norm": 0.3338271677494049, "learning_rate": 3.392346012514166e-05, "loss": 0.0223, "step": 49920 }, { "grad_norm": 0.3421711325645447, "learning_rate": 3.390389063524757e-05, "loss": 0.0231, "step": 49930 }, { "grad_norm": 0.2502405643463135, "learning_rate": 3.3884323895738324e-05, "loss": 0.0194, "step": 49940 }, { "grad_norm": 0.3599887788295746, "learning_rate": 3.386475990995738e-05, "loss": 0.0246, "step": 49950 }, { "grad_norm": 0.3517838418483734, "learning_rate": 3.384519868124765e-05, "loss": 0.0207, "step": 49960 }, { "grad_norm": 0.24992337822914124, "learning_rate": 3.3825640212951645e-05, "loss": 0.0201, "step": 49970 }, { "grad_norm": 0.2489318549633026, "learning_rate": 3.380608450841134e-05, "loss": 0.0195, "step": 49980 }, { "grad_norm": 0.43608564138412476, "learning_rate": 3.3786531570968305e-05, "loss": 0.0211, "step": 49990 }, { "grad_norm": 0.23580871522426605, "learning_rate": 3.3766981403963584e-05, "loss": 0.0273, "step": 50000 }, { "grad_norm": 0.23518122732639313, "learning_rate": 3.374743401073775e-05, "loss": 0.0226, "step": 50010 }, { "grad_norm": 0.33753645420074463, "learning_rate": 3.372788939463095e-05, "loss": 0.0167, "step": 50020 }, { "grad_norm": 0.27592265605926514, "learning_rate": 3.370834755898281e-05, "loss": 0.0205, "step": 50030 }, { "grad_norm": 0.3465802073478699, "learning_rate": 3.3688808507132493e-05, "loss": 0.0244, "step": 50040 }, { "grad_norm": 0.1922816038131714, "learning_rate": 3.3669272242418685e-05, "loss": 0.0243, "step": 50050 }, { "grad_norm": 0.28778547048568726, "learning_rate": 3.364973876817961e-05, "loss": 0.0227, "step": 50060 }, { "grad_norm": 0.30660074949264526, "learning_rate": 3.363020808775299e-05, "loss": 0.0175, "step": 50070 }, { "grad_norm": 0.3238946497440338, "learning_rate": 3.361068020447611e-05, "loss": 0.0208, "step": 50080 }, { "grad_norm": 0.25909900665283203, "learning_rate": 3.3591155121685724e-05, "loss": 0.0214, "step": 50090 }, { "grad_norm": 0.3409917652606964, "learning_rate": 3.357163284271814e-05, "loss": 0.0244, "step": 50100 }, { "grad_norm": 0.4167574644088745, "learning_rate": 3.355211337090919e-05, "loss": 0.0177, "step": 50110 }, { "grad_norm": 0.2403954714536667, "learning_rate": 3.353259670959421e-05, "loss": 0.0212, "step": 50120 }, { "grad_norm": 0.19930948317050934, "learning_rate": 3.351308286210808e-05, "loss": 0.0192, "step": 50130 }, { "grad_norm": 0.23754066228866577, "learning_rate": 3.3493571831785156e-05, "loss": 0.0208, "step": 50140 }, { "grad_norm": 0.40973398089408875, "learning_rate": 3.347406362195936e-05, "loss": 0.0182, "step": 50150 }, { "grad_norm": 0.34180402755737305, "learning_rate": 3.345455823596411e-05, "loss": 0.0202, "step": 50160 }, { "grad_norm": 0.2590610384941101, "learning_rate": 3.3435055677132346e-05, "loss": 0.0205, "step": 50170 }, { "grad_norm": 0.3498656153678894, "learning_rate": 3.3415555948796505e-05, "loss": 0.0216, "step": 50180 }, { "grad_norm": 0.30570727586746216, "learning_rate": 3.3396059054288556e-05, "loss": 0.0196, "step": 50190 }, { "grad_norm": 0.2471766173839569, "learning_rate": 3.3376564996940015e-05, "loss": 0.0318, "step": 50200 }, { "grad_norm": 0.2832709848880768, "learning_rate": 3.3357073780081836e-05, "loss": 0.0228, "step": 50210 }, { "grad_norm": 0.22756317257881165, "learning_rate": 3.333758540704459e-05, "loss": 0.018, "step": 50220 }, { "grad_norm": 0.2361152023077011, "learning_rate": 3.331809988115827e-05, "loss": 0.0193, "step": 50230 }, { "grad_norm": 0.26276594400405884, "learning_rate": 3.329861720575244e-05, "loss": 0.0234, "step": 50240 }, { "grad_norm": 0.22251245379447937, "learning_rate": 3.3279137384156126e-05, "loss": 0.0198, "step": 50250 }, { "grad_norm": 0.3072737157344818, "learning_rate": 3.3259660419697934e-05, "loss": 0.0248, "step": 50260 }, { "grad_norm": 0.290511816740036, "learning_rate": 3.3240186315705926e-05, "loss": 0.0262, "step": 50270 }, { "grad_norm": 0.28526705503463745, "learning_rate": 3.322071507550769e-05, "loss": 0.0296, "step": 50280 }, { "grad_norm": 0.25864794850349426, "learning_rate": 3.320124670243033e-05, "loss": 0.0205, "step": 50290 }, { "grad_norm": 0.1524956375360489, "learning_rate": 3.318178119980045e-05, "loss": 0.0204, "step": 50300 }, { "grad_norm": 0.3022781014442444, "learning_rate": 3.316231857094421e-05, "loss": 0.0221, "step": 50310 }, { "grad_norm": 0.22989341616630554, "learning_rate": 3.31428588191872e-05, "loss": 0.0206, "step": 50320 }, { "grad_norm": 0.3243195116519928, "learning_rate": 3.312340194785458e-05, "loss": 0.0182, "step": 50330 }, { "grad_norm": 0.3479572534561157, "learning_rate": 3.3103947960270975e-05, "loss": 0.0222, "step": 50340 }, { "grad_norm": 0.35594403743743896, "learning_rate": 3.308449685976058e-05, "loss": 0.0185, "step": 50350 }, { "grad_norm": 0.34071245789527893, "learning_rate": 3.3065048649647024e-05, "loss": 0.0213, "step": 50360 }, { "grad_norm": 0.25228816270828247, "learning_rate": 3.304560333325348e-05, "loss": 0.019, "step": 50370 }, { "grad_norm": 0.29522809386253357, "learning_rate": 3.3026160913902624e-05, "loss": 0.0238, "step": 50380 }, { "grad_norm": 0.425584077835083, "learning_rate": 3.300672139491662e-05, "loss": 0.0252, "step": 50390 }, { "grad_norm": 0.18099091947078705, "learning_rate": 3.298728477961717e-05, "loss": 0.0182, "step": 50400 }, { "grad_norm": 0.233952596783638, "learning_rate": 3.296785107132544e-05, "loss": 0.0232, "step": 50410 }, { "grad_norm": 0.24619479477405548, "learning_rate": 3.2948420273362156e-05, "loss": 0.0203, "step": 50420 }, { "grad_norm": 0.23008669912815094, "learning_rate": 3.292899238904747e-05, "loss": 0.021, "step": 50430 }, { "grad_norm": 0.2547648847103119, "learning_rate": 3.29095674217011e-05, "loss": 0.0188, "step": 50440 }, { "grad_norm": 0.2762756943702698, "learning_rate": 3.289014537464224e-05, "loss": 0.0229, "step": 50450 }, { "grad_norm": 0.14714674651622772, "learning_rate": 3.287072625118955e-05, "loss": 0.0199, "step": 50460 }, { "grad_norm": 0.2766597867012024, "learning_rate": 3.285131005466129e-05, "loss": 0.0194, "step": 50470 }, { "grad_norm": 0.2442505955696106, "learning_rate": 3.2831896788375105e-05, "loss": 0.0148, "step": 50480 }, { "grad_norm": 0.22105570137500763, "learning_rate": 3.281248645564822e-05, "loss": 0.0168, "step": 50490 }, { "grad_norm": 0.3960030674934387, "learning_rate": 3.2793079059797306e-05, "loss": 0.0227, "step": 50500 }, { "grad_norm": 0.24373216927051544, "learning_rate": 3.277367460413859e-05, "loss": 0.0219, "step": 50510 }, { "grad_norm": 0.2971193790435791, "learning_rate": 3.275427309198773e-05, "loss": 0.0292, "step": 50520 }, { "grad_norm": 0.2562364637851715, "learning_rate": 3.273487452665993e-05, "loss": 0.0223, "step": 50530 }, { "grad_norm": 0.3740476965904236, "learning_rate": 3.271547891146986e-05, "loss": 0.0248, "step": 50540 }, { "grad_norm": 0.28941917419433594, "learning_rate": 3.269608624973173e-05, "loss": 0.0199, "step": 50550 }, { "grad_norm": 0.1918330192565918, "learning_rate": 3.26766965447592e-05, "loss": 0.0227, "step": 50560 }, { "grad_norm": 0.2655545175075531, "learning_rate": 3.2657309799865424e-05, "loss": 0.0221, "step": 50570 }, { "grad_norm": 0.3959464132785797, "learning_rate": 3.2637926018363084e-05, "loss": 0.0252, "step": 50580 }, { "grad_norm": 0.32267993688583374, "learning_rate": 3.261854520356432e-05, "loss": 0.0234, "step": 50590 }, { "grad_norm": 0.23683127760887146, "learning_rate": 3.2599167358780816e-05, "loss": 0.022, "step": 50600 }, { "grad_norm": 0.257548451423645, "learning_rate": 3.2579792487323676e-05, "loss": 0.0159, "step": 50610 }, { "grad_norm": 0.22636720538139343, "learning_rate": 3.256042059250358e-05, "loss": 0.0241, "step": 50620 }, { "grad_norm": 0.42404335737228394, "learning_rate": 3.254105167763062e-05, "loss": 0.0249, "step": 50630 }, { "grad_norm": 0.324850469827652, "learning_rate": 3.252168574601443e-05, "loss": 0.0222, "step": 50640 }, { "grad_norm": 0.22002756595611572, "learning_rate": 3.2502322800964115e-05, "loss": 0.0183, "step": 50650 }, { "grad_norm": 0.3442106246948242, "learning_rate": 3.248296284578825e-05, "loss": 0.0194, "step": 50660 }, { "grad_norm": 0.5703299641609192, "learning_rate": 3.246360588379497e-05, "loss": 0.0277, "step": 50670 }, { "grad_norm": 0.4716506600379944, "learning_rate": 3.24442519182918e-05, "loss": 0.0287, "step": 50680 }, { "grad_norm": 0.23273542523384094, "learning_rate": 3.2424900952585845e-05, "loss": 0.0179, "step": 50690 }, { "grad_norm": 0.20138877630233765, "learning_rate": 3.240555298998362e-05, "loss": 0.0206, "step": 50700 }, { "grad_norm": 0.27076834440231323, "learning_rate": 3.2386208033791194e-05, "loss": 0.0186, "step": 50710 }, { "grad_norm": 0.4035496115684509, "learning_rate": 3.2366866087314064e-05, "loss": 0.0193, "step": 50720 }, { "grad_norm": 0.24623960256576538, "learning_rate": 3.234752715385727e-05, "loss": 0.0211, "step": 50730 }, { "grad_norm": 0.30992332100868225, "learning_rate": 3.232819123672529e-05, "loss": 0.0273, "step": 50740 }, { "grad_norm": 0.47307512164115906, "learning_rate": 3.2308858339222076e-05, "loss": 0.0251, "step": 50750 }, { "grad_norm": 0.37730515003204346, "learning_rate": 3.2289528464651144e-05, "loss": 0.0257, "step": 50760 }, { "grad_norm": 0.3095604181289673, "learning_rate": 3.22702016163154e-05, "loss": 0.0211, "step": 50770 }, { "grad_norm": 0.2207927405834198, "learning_rate": 3.2250877797517295e-05, "loss": 0.0235, "step": 50780 }, { "grad_norm": 0.22310277819633484, "learning_rate": 3.223155701155872e-05, "loss": 0.0233, "step": 50790 }, { "grad_norm": 0.46976563334465027, "learning_rate": 3.2212239261741105e-05, "loss": 0.0258, "step": 50800 }, { "grad_norm": 0.2568058967590332, "learning_rate": 3.219292455136528e-05, "loss": 0.0176, "step": 50810 }, { "grad_norm": 0.21342045068740845, "learning_rate": 3.217361288373165e-05, "loss": 0.0144, "step": 50820 }, { "grad_norm": 0.2709153890609741, "learning_rate": 3.215430426214002e-05, "loss": 0.0178, "step": 50830 }, { "grad_norm": 0.24813111126422882, "learning_rate": 3.2134998689889705e-05, "loss": 0.0173, "step": 50840 }, { "grad_norm": 0.28475669026374817, "learning_rate": 3.211569617027952e-05, "loss": 0.0225, "step": 50850 }, { "grad_norm": 0.16945196688175201, "learning_rate": 3.209639670660769e-05, "loss": 0.0233, "step": 50860 }, { "grad_norm": 0.4979739189147949, "learning_rate": 3.207710030217202e-05, "loss": 0.0236, "step": 50870 }, { "grad_norm": 0.49884161353111267, "learning_rate": 3.205780696026972e-05, "loss": 0.0187, "step": 50880 }, { "grad_norm": 0.1615084707736969, "learning_rate": 3.203851668419749e-05, "loss": 0.0168, "step": 50890 }, { "grad_norm": 0.19967572391033173, "learning_rate": 3.201922947725149e-05, "loss": 0.0209, "step": 50900 }, { "grad_norm": 0.2936621606349945, "learning_rate": 3.199994534272742e-05, "loss": 0.0202, "step": 50910 }, { "grad_norm": 0.31017354130744934, "learning_rate": 3.1980664283920394e-05, "loss": 0.0221, "step": 50920 }, { "grad_norm": 0.24570681154727936, "learning_rate": 3.196138630412499e-05, "loss": 0.0191, "step": 50930 }, { "grad_norm": 0.2797727882862091, "learning_rate": 3.194211140663532e-05, "loss": 0.0151, "step": 50940 }, { "grad_norm": 0.2207183688879013, "learning_rate": 3.1922839594744914e-05, "loss": 0.0226, "step": 50950 }, { "grad_norm": 0.34587010741233826, "learning_rate": 3.190357087174683e-05, "loss": 0.0217, "step": 50960 }, { "grad_norm": 0.19716677069664001, "learning_rate": 3.188430524093353e-05, "loss": 0.0153, "step": 50970 }, { "grad_norm": 0.3043725788593292, "learning_rate": 3.1865042705597014e-05, "loss": 0.0211, "step": 50980 }, { "grad_norm": 0.33521801233291626, "learning_rate": 3.184578326902868e-05, "loss": 0.0198, "step": 50990 }, { "grad_norm": 0.20729570090770721, "learning_rate": 3.182652693451949e-05, "loss": 0.0208, "step": 51000 }, { "grad_norm": 0.3486770987510681, "learning_rate": 3.180727370535981e-05, "loss": 0.02, "step": 51010 }, { "grad_norm": 0.29018908739089966, "learning_rate": 3.178802358483945e-05, "loss": 0.0187, "step": 51020 }, { "grad_norm": 0.21504077315330505, "learning_rate": 3.17687765762478e-05, "loss": 0.0188, "step": 51030 }, { "grad_norm": 0.16371318697929382, "learning_rate": 3.174953268287357e-05, "loss": 0.0254, "step": 51040 }, { "grad_norm": 0.41259831190109253, "learning_rate": 3.173029190800508e-05, "loss": 0.0178, "step": 51050 }, { "grad_norm": 0.41634759306907654, "learning_rate": 3.171105425493e-05, "loss": 0.0245, "step": 51060 }, { "grad_norm": 0.24393339455127716, "learning_rate": 3.169181972693557e-05, "loss": 0.0138, "step": 51070 }, { "grad_norm": 0.18447613716125488, "learning_rate": 3.1672588327308404e-05, "loss": 0.0196, "step": 51080 }, { "grad_norm": 0.34996283054351807, "learning_rate": 3.1653360059334646e-05, "loss": 0.022, "step": 51090 }, { "grad_norm": 0.4140549302101135, "learning_rate": 3.163413492629985e-05, "loss": 0.0345, "step": 51100 }, { "grad_norm": 0.2889138162136078, "learning_rate": 3.161491293148912e-05, "loss": 0.0201, "step": 51110 }, { "grad_norm": 0.33799564838409424, "learning_rate": 3.1595694078186925e-05, "loss": 0.0228, "step": 51120 }, { "grad_norm": 0.2390437126159668, "learning_rate": 3.157647836967724e-05, "loss": 0.0174, "step": 51130 }, { "grad_norm": 0.26203590631484985, "learning_rate": 3.155726580924354e-05, "loss": 0.0186, "step": 51140 }, { "grad_norm": 0.3249180316925049, "learning_rate": 3.1538056400168676e-05, "loss": 0.0309, "step": 51150 }, { "grad_norm": 0.210271954536438, "learning_rate": 3.151885014573506e-05, "loss": 0.0246, "step": 51160 }, { "grad_norm": 0.30874279141426086, "learning_rate": 3.149964704922448e-05, "loss": 0.0307, "step": 51170 }, { "grad_norm": 0.27266809344291687, "learning_rate": 3.148044711391825e-05, "loss": 0.019, "step": 51180 }, { "grad_norm": 0.29029443860054016, "learning_rate": 3.146125034309707e-05, "loss": 0.0195, "step": 51190 }, { "grad_norm": 0.30684694647789, "learning_rate": 3.1442056740041195e-05, "loss": 0.0242, "step": 51200 }, { "grad_norm": 0.21406729519367218, "learning_rate": 3.142286630803026e-05, "loss": 0.031, "step": 51210 }, { "grad_norm": 0.35375866293907166, "learning_rate": 3.140367905034337e-05, "loss": 0.0181, "step": 51220 }, { "grad_norm": 0.2862853407859802, "learning_rate": 3.138449497025914e-05, "loss": 0.0209, "step": 51230 }, { "grad_norm": 0.3432901203632355, "learning_rate": 3.136531407105557e-05, "loss": 0.0216, "step": 51240 }, { "grad_norm": 0.2706763744354248, "learning_rate": 3.1346136356010184e-05, "loss": 0.0245, "step": 51250 }, { "grad_norm": 0.29476815462112427, "learning_rate": 3.132696182839988e-05, "loss": 0.0196, "step": 51260 }, { "grad_norm": 0.32484114170074463, "learning_rate": 3.130779049150111e-05, "loss": 0.0224, "step": 51270 }, { "grad_norm": 0.34139811992645264, "learning_rate": 3.128862234858971e-05, "loss": 0.0178, "step": 51280 }, { "grad_norm": 0.3612024188041687, "learning_rate": 3.1269457402941005e-05, "loss": 0.0231, "step": 51290 }, { "grad_norm": 0.4033359885215759, "learning_rate": 3.125029565782974e-05, "loss": 0.026, "step": 51300 }, { "grad_norm": 0.3993130624294281, "learning_rate": 3.123113711653013e-05, "loss": 0.0225, "step": 51310 }, { "grad_norm": 0.2146792858839035, "learning_rate": 3.121198178231587e-05, "loss": 0.0251, "step": 51320 }, { "grad_norm": 0.38551095128059387, "learning_rate": 3.119282965846006e-05, "loss": 0.0222, "step": 51330 }, { "grad_norm": 0.3743003308773041, "learning_rate": 3.1173680748235303e-05, "loss": 0.0254, "step": 51340 }, { "grad_norm": 0.27918297052383423, "learning_rate": 3.115453505491358e-05, "loss": 0.0249, "step": 51350 }, { "grad_norm": 0.2827973961830139, "learning_rate": 3.1135392581766404e-05, "loss": 0.025, "step": 51360 }, { "grad_norm": 0.2618171274662018, "learning_rate": 3.111625333206467e-05, "loss": 0.0192, "step": 51370 }, { "grad_norm": 0.31075412034988403, "learning_rate": 3.109711730907881e-05, "loss": 0.0166, "step": 51380 }, { "grad_norm": 0.25739234685897827, "learning_rate": 3.107798451607859e-05, "loss": 0.0233, "step": 51390 }, { "grad_norm": 0.3492089807987213, "learning_rate": 3.105885495633328e-05, "loss": 0.0229, "step": 51400 }, { "grad_norm": 0.2153223603963852, "learning_rate": 3.103972863311163e-05, "loss": 0.0194, "step": 51410 }, { "grad_norm": 0.37095892429351807, "learning_rate": 3.102060554968178e-05, "loss": 0.0218, "step": 51420 }, { "grad_norm": 0.20588988065719604, "learning_rate": 3.100148570931137e-05, "loss": 0.0226, "step": 51430 }, { "grad_norm": 0.33796828985214233, "learning_rate": 3.098236911526744e-05, "loss": 0.024, "step": 51440 }, { "grad_norm": 0.2230650633573532, "learning_rate": 3.0963255770816495e-05, "loss": 0.0182, "step": 51450 }, { "grad_norm": 0.35726019740104675, "learning_rate": 3.094414567922447e-05, "loss": 0.0223, "step": 51460 }, { "grad_norm": 0.26313263177871704, "learning_rate": 3.092503884375678e-05, "loss": 0.0235, "step": 51470 }, { "grad_norm": 0.26925021409988403, "learning_rate": 3.0905935267678255e-05, "loss": 0.0225, "step": 51480 }, { "grad_norm": 0.3039238452911377, "learning_rate": 3.0886834954253166e-05, "loss": 0.0224, "step": 51490 }, { "grad_norm": 0.20674967765808105, "learning_rate": 3.086773790674524e-05, "loss": 0.0219, "step": 51500 }, { "grad_norm": 0.3428414463996887, "learning_rate": 3.0848644128417617e-05, "loss": 0.0247, "step": 51510 }, { "grad_norm": 0.268081933259964, "learning_rate": 3.082955362253294e-05, "loss": 0.0272, "step": 51520 }, { "grad_norm": 0.22959427535533905, "learning_rate": 3.081046639235323e-05, "loss": 0.0243, "step": 51530 }, { "grad_norm": 0.357727974653244, "learning_rate": 3.0791382441139986e-05, "loss": 0.02, "step": 51540 }, { "grad_norm": 0.37167537212371826, "learning_rate": 3.077230177215411e-05, "loss": 0.0275, "step": 51550 }, { "grad_norm": 0.16772127151489258, "learning_rate": 3.0753224388656e-05, "loss": 0.0209, "step": 51560 }, { "grad_norm": 0.23894858360290527, "learning_rate": 3.073415029390544e-05, "loss": 0.0257, "step": 51570 }, { "grad_norm": 0.2752065360546112, "learning_rate": 3.071507949116166e-05, "loss": 0.0201, "step": 51580 }, { "grad_norm": 0.2394440919160843, "learning_rate": 3.069601198368337e-05, "loss": 0.0171, "step": 51590 }, { "grad_norm": 0.2412358671426773, "learning_rate": 3.067694777472864e-05, "loss": 0.0241, "step": 51600 }, { "grad_norm": 0.24055615067481995, "learning_rate": 3.065788686755508e-05, "loss": 0.0186, "step": 51610 }, { "grad_norm": 0.2913733720779419, "learning_rate": 3.063882926541961e-05, "loss": 0.0267, "step": 51620 }, { "grad_norm": 0.35271021723747253, "learning_rate": 3.061977497157872e-05, "loss": 0.0232, "step": 51630 }, { "grad_norm": 0.34146609902381897, "learning_rate": 3.0600723989288224e-05, "loss": 0.0198, "step": 51640 }, { "grad_norm": 0.2753088176250458, "learning_rate": 3.058167632180343e-05, "loss": 0.0228, "step": 51650 }, { "grad_norm": 0.282877117395401, "learning_rate": 3.0562631972379074e-05, "loss": 0.0271, "step": 51660 }, { "grad_norm": 0.22498317062854767, "learning_rate": 3.0543590944269276e-05, "loss": 0.0199, "step": 51670 }, { "grad_norm": 0.431680828332901, "learning_rate": 3.052455324072766e-05, "loss": 0.0313, "step": 51680 }, { "grad_norm": 0.43041932582855225, "learning_rate": 3.0505518865007244e-05, "loss": 0.025, "step": 51690 }, { "grad_norm": 0.2332422435283661, "learning_rate": 3.048648782036048e-05, "loss": 0.0252, "step": 51700 }, { "grad_norm": 0.18338268995285034, "learning_rate": 3.0467460110039236e-05, "loss": 0.0193, "step": 51710 }, { "grad_norm": 0.36113491654396057, "learning_rate": 3.0448435737294866e-05, "loss": 0.0233, "step": 51720 }, { "grad_norm": 0.17432543635368347, "learning_rate": 3.042941470537808e-05, "loss": 0.0193, "step": 51730 }, { "grad_norm": 0.3224011957645416, "learning_rate": 3.0410397017539072e-05, "loss": 0.0187, "step": 51740 }, { "grad_norm": 0.2760820984840393, "learning_rate": 3.039138267702742e-05, "loss": 0.0184, "step": 51750 }, { "grad_norm": 0.2701491117477417, "learning_rate": 3.0372371687092193e-05, "loss": 0.0196, "step": 51760 }, { "grad_norm": 0.26078182458877563, "learning_rate": 3.0353364050981823e-05, "loss": 0.0215, "step": 51770 }, { "grad_norm": 0.24323886632919312, "learning_rate": 3.033435977194418e-05, "loss": 0.024, "step": 51780 }, { "grad_norm": 0.21824848651885986, "learning_rate": 3.0315358853226618e-05, "loss": 0.024, "step": 51790 }, { "grad_norm": 0.26552388072013855, "learning_rate": 3.0296361298075844e-05, "loss": 0.0206, "step": 51800 }, { "grad_norm": 0.20267699658870697, "learning_rate": 3.027736710973803e-05, "loss": 0.0247, "step": 51810 }, { "grad_norm": 0.2061607539653778, "learning_rate": 3.025837629145875e-05, "loss": 0.0167, "step": 51820 }, { "grad_norm": 0.24447378516197205, "learning_rate": 3.0239388846483048e-05, "loss": 0.0183, "step": 51830 }, { "grad_norm": 0.36403539776802063, "learning_rate": 3.022040477805532e-05, "loss": 0.0217, "step": 51840 }, { "grad_norm": 0.2988021969795227, "learning_rate": 3.020142408941946e-05, "loss": 0.0199, "step": 51850 }, { "grad_norm": 0.29653504490852356, "learning_rate": 3.018244678381873e-05, "loss": 0.0214, "step": 51860 }, { "grad_norm": 0.2910669147968292, "learning_rate": 3.0163472864495812e-05, "loss": 0.0244, "step": 51870 }, { "grad_norm": 0.2073761224746704, "learning_rate": 3.014450233469287e-05, "loss": 0.0228, "step": 51880 }, { "grad_norm": 0.48125314712524414, "learning_rate": 3.0125535197651422e-05, "loss": 0.0304, "step": 51890 }, { "grad_norm": 0.3307420313358307, "learning_rate": 3.0106571456612448e-05, "loss": 0.0181, "step": 51900 }, { "grad_norm": 0.2613823711872101, "learning_rate": 3.0087611114816305e-05, "loss": 0.0279, "step": 51910 }, { "grad_norm": 0.24830904603004456, "learning_rate": 3.006865417550284e-05, "loss": 0.0181, "step": 51920 }, { "grad_norm": 0.30174657702445984, "learning_rate": 3.0049700641911242e-05, "loss": 0.0225, "step": 51930 }, { "grad_norm": 0.37996599078178406, "learning_rate": 3.0030750517280175e-05, "loss": 0.0184, "step": 51940 }, { "grad_norm": 0.2800562083721161, "learning_rate": 3.0011803804847682e-05, "loss": 0.0233, "step": 51950 }, { "grad_norm": 0.3381738066673279, "learning_rate": 2.9992860507851224e-05, "loss": 0.0218, "step": 51960 }, { "grad_norm": 0.2995864748954773, "learning_rate": 2.9973920629527723e-05, "loss": 0.0168, "step": 51970 }, { "grad_norm": 0.2747781276702881, "learning_rate": 2.9954984173113453e-05, "loss": 0.0211, "step": 51980 }, { "grad_norm": 0.2788437306880951, "learning_rate": 2.993605114184418e-05, "loss": 0.0232, "step": 51990 }, { "grad_norm": 0.38984599709510803, "learning_rate": 2.9917121538955005e-05, "loss": 0.0298, "step": 52000 }, { "grad_norm": 0.3462652266025543, "learning_rate": 2.98981953676805e-05, "loss": 0.0281, "step": 52010 }, { "grad_norm": 0.30951303243637085, "learning_rate": 2.9879272631254594e-05, "loss": 0.0184, "step": 52020 }, { "grad_norm": 0.23114977777004242, "learning_rate": 2.986035333291073e-05, "loss": 0.0244, "step": 52030 }, { "grad_norm": 0.2409086674451828, "learning_rate": 2.9841437475881646e-05, "loss": 0.027, "step": 52040 }, { "grad_norm": 0.3043532073497772, "learning_rate": 2.9822525063399555e-05, "loss": 0.0214, "step": 52050 }, { "grad_norm": 0.348283588886261, "learning_rate": 2.9803616098696087e-05, "loss": 0.0256, "step": 52060 }, { "grad_norm": 0.36524972319602966, "learning_rate": 2.9784710585002228e-05, "loss": 0.0268, "step": 52070 }, { "grad_norm": 0.3727952837944031, "learning_rate": 2.9765808525548467e-05, "loss": 0.0186, "step": 52080 }, { "grad_norm": 0.2720358669757843, "learning_rate": 2.974690992356461e-05, "loss": 0.0183, "step": 52090 }, { "grad_norm": 0.20327474176883698, "learning_rate": 2.9728014782279934e-05, "loss": 0.0271, "step": 52100 }, { "grad_norm": 0.48849472403526306, "learning_rate": 2.970912310492307e-05, "loss": 0.0224, "step": 52110 }, { "grad_norm": 0.23660583794116974, "learning_rate": 2.9690234894722134e-05, "loss": 0.0232, "step": 52120 }, { "grad_norm": 0.25829043984413147, "learning_rate": 2.9671350154904577e-05, "loss": 0.0277, "step": 52130 }, { "grad_norm": 0.25796228647232056, "learning_rate": 2.9652468888697282e-05, "loss": 0.022, "step": 52140 }, { "grad_norm": 0.24632664024829865, "learning_rate": 2.9633591099326562e-05, "loss": 0.0161, "step": 52150 }, { "grad_norm": 0.26158249378204346, "learning_rate": 2.9614716790018072e-05, "loss": 0.0193, "step": 52160 }, { "grad_norm": 0.2803332507610321, "learning_rate": 2.959584596399697e-05, "loss": 0.0181, "step": 52170 }, { "grad_norm": 0.36087146401405334, "learning_rate": 2.9576978624487717e-05, "loss": 0.0257, "step": 52180 }, { "grad_norm": 0.2741893231868744, "learning_rate": 2.9558114774714263e-05, "loss": 0.0184, "step": 52190 }, { "grad_norm": 0.31794312596321106, "learning_rate": 2.9539254417899897e-05, "loss": 0.0233, "step": 52200 }, { "grad_norm": 0.25560158491134644, "learning_rate": 2.9520397557267365e-05, "loss": 0.0174, "step": 52210 }, { "grad_norm": 0.3093714416027069, "learning_rate": 2.9501544196038765e-05, "loss": 0.0182, "step": 52220 }, { "grad_norm": 0.25475597381591797, "learning_rate": 2.9482694337435614e-05, "loss": 0.0246, "step": 52230 }, { "grad_norm": 0.31003037095069885, "learning_rate": 2.946384798467887e-05, "loss": 0.017, "step": 52240 }, { "grad_norm": 0.2175293266773224, "learning_rate": 2.9445005140988824e-05, "loss": 0.0111, "step": 52250 }, { "grad_norm": 0.24047422409057617, "learning_rate": 2.942616580958524e-05, "loss": 0.0187, "step": 52260 }, { "grad_norm": 0.22113387286663055, "learning_rate": 2.9407329993687193e-05, "loss": 0.017, "step": 52270 }, { "grad_norm": 0.3296876847743988, "learning_rate": 2.938849769651326e-05, "loss": 0.0164, "step": 52280 }, { "grad_norm": 0.2728898227214813, "learning_rate": 2.9369668921281345e-05, "loss": 0.0211, "step": 52290 }, { "grad_norm": 0.22921328246593475, "learning_rate": 2.9350843671208773e-05, "loss": 0.018, "step": 52300 }, { "grad_norm": 0.2810388505458832, "learning_rate": 2.9332021949512245e-05, "loss": 0.0212, "step": 52310 }, { "grad_norm": 0.30610209703445435, "learning_rate": 2.9313203759407908e-05, "loss": 0.0212, "step": 52320 }, { "grad_norm": 0.20173388719558716, "learning_rate": 2.929438910411127e-05, "loss": 0.0339, "step": 52330 }, { "grad_norm": 0.3945330083370209, "learning_rate": 2.927557798683722e-05, "loss": 0.0247, "step": 52340 }, { "grad_norm": 0.2976433336734772, "learning_rate": 2.9256770410800095e-05, "loss": 0.0171, "step": 52350 }, { "grad_norm": 0.25264298915863037, "learning_rate": 2.9237966379213554e-05, "loss": 0.0156, "step": 52360 }, { "grad_norm": 0.20570164918899536, "learning_rate": 2.9219165895290736e-05, "loss": 0.0203, "step": 52370 }, { "grad_norm": 0.24759283661842346, "learning_rate": 2.9200368962244083e-05, "loss": 0.0247, "step": 52380 }, { "grad_norm": 0.3481157422065735, "learning_rate": 2.9181575583285536e-05, "loss": 0.0191, "step": 52390 }, { "grad_norm": 0.24145977199077606, "learning_rate": 2.916278576162632e-05, "loss": 0.0211, "step": 52400 }, { "grad_norm": 0.23495928943157196, "learning_rate": 2.9143999500477116e-05, "loss": 0.0194, "step": 52410 }, { "grad_norm": 0.2948795557022095, "learning_rate": 2.9125216803048004e-05, "loss": 0.0234, "step": 52420 }, { "grad_norm": 0.30505692958831787, "learning_rate": 2.9106437672548403e-05, "loss": 0.0204, "step": 52430 }, { "grad_norm": 0.19118477404117584, "learning_rate": 2.9087662112187154e-05, "loss": 0.0171, "step": 52440 }, { "grad_norm": 0.19289368391036987, "learning_rate": 2.9068890125172498e-05, "loss": 0.0124, "step": 52450 }, { "grad_norm": 0.22217532992362976, "learning_rate": 2.9050121714712058e-05, "loss": 0.0305, "step": 52460 }, { "grad_norm": 0.372934490442276, "learning_rate": 2.9031356884012835e-05, "loss": 0.0222, "step": 52470 }, { "grad_norm": 0.275324285030365, "learning_rate": 2.9012595636281247e-05, "loss": 0.0214, "step": 52480 }, { "grad_norm": 0.4029109477996826, "learning_rate": 2.8993837974723016e-05, "loss": 0.0337, "step": 52490 }, { "grad_norm": 0.23274967074394226, "learning_rate": 2.8975083902543394e-05, "loss": 0.0179, "step": 52500 }, { "grad_norm": 0.35213321447372437, "learning_rate": 2.895633342294688e-05, "loss": 0.0198, "step": 52510 }, { "grad_norm": 0.23471398651599884, "learning_rate": 2.893758653913744e-05, "loss": 0.0194, "step": 52520 }, { "grad_norm": 0.2340496927499771, "learning_rate": 2.8918843254318423e-05, "loss": 0.018, "step": 52530 }, { "grad_norm": 0.3788285255432129, "learning_rate": 2.8900103571692483e-05, "loss": 0.0201, "step": 52540 }, { "grad_norm": 0.24160628020763397, "learning_rate": 2.8881367494461796e-05, "loss": 0.0163, "step": 52550 }, { "grad_norm": 0.19769851863384247, "learning_rate": 2.886263502582779e-05, "loss": 0.0182, "step": 52560 }, { "grad_norm": 0.37122681736946106, "learning_rate": 2.8843906168991353e-05, "loss": 0.0186, "step": 52570 }, { "grad_norm": 0.303423136472702, "learning_rate": 2.8825180927152728e-05, "loss": 0.0239, "step": 52580 }, { "grad_norm": 0.2874569892883301, "learning_rate": 2.8806459303511547e-05, "loss": 0.0168, "step": 52590 }, { "grad_norm": 0.3408852219581604, "learning_rate": 2.8787741301266852e-05, "loss": 0.0175, "step": 52600 }, { "grad_norm": 0.24639521539211273, "learning_rate": 2.876902692361697e-05, "loss": 0.0137, "step": 52610 }, { "grad_norm": 0.27335843443870544, "learning_rate": 2.875031617375976e-05, "loss": 0.013, "step": 52620 }, { "grad_norm": 0.2834213972091675, "learning_rate": 2.8731609054892318e-05, "loss": 0.0181, "step": 52630 }, { "grad_norm": 0.25104910135269165, "learning_rate": 2.871290557021119e-05, "loss": 0.016, "step": 52640 }, { "grad_norm": 0.24385227262973785, "learning_rate": 2.8694205722912298e-05, "loss": 0.0204, "step": 52650 }, { "grad_norm": 0.23380425572395325, "learning_rate": 2.8675509516190936e-05, "loss": 0.0222, "step": 52660 }, { "grad_norm": 0.2722335159778595, "learning_rate": 2.8656816953241772e-05, "loss": 0.0228, "step": 52670 }, { "grad_norm": 0.4485841989517212, "learning_rate": 2.863812803725887e-05, "loss": 0.0188, "step": 52680 }, { "grad_norm": 0.24967126548290253, "learning_rate": 2.8619442771435623e-05, "loss": 0.0137, "step": 52690 }, { "grad_norm": 0.1845681220293045, "learning_rate": 2.860076115896484e-05, "loss": 0.0232, "step": 52700 }, { "grad_norm": 0.27312734723091125, "learning_rate": 2.8582083203038712e-05, "loss": 0.0193, "step": 52710 }, { "grad_norm": 0.26854196190834045, "learning_rate": 2.8563408906848778e-05, "loss": 0.017, "step": 52720 }, { "grad_norm": 0.4882086515426636, "learning_rate": 2.8544738273585993e-05, "loss": 0.0149, "step": 52730 }, { "grad_norm": 0.24664883315563202, "learning_rate": 2.8526071306440595e-05, "loss": 0.0133, "step": 52740 }, { "grad_norm": 0.278361976146698, "learning_rate": 2.850740800860233e-05, "loss": 0.0205, "step": 52750 }, { "grad_norm": 0.2554566264152527, "learning_rate": 2.8488748383260178e-05, "loss": 0.0179, "step": 52760 }, { "grad_norm": 0.1627199351787567, "learning_rate": 2.8470092433602635e-05, "loss": 0.0223, "step": 52770 }, { "grad_norm": 0.30185171961784363, "learning_rate": 2.845144016281745e-05, "loss": 0.0213, "step": 52780 }, { "grad_norm": 0.27902188897132874, "learning_rate": 2.8432791574091744e-05, "loss": 0.0283, "step": 52790 }, { "grad_norm": 0.35589873790740967, "learning_rate": 2.8414146670612134e-05, "loss": 0.0195, "step": 52800 }, { "grad_norm": 0.2633219361305237, "learning_rate": 2.8395505455564446e-05, "loss": 0.0197, "step": 52810 }, { "grad_norm": 0.16644684970378876, "learning_rate": 2.837686793213403e-05, "loss": 0.02, "step": 52820 }, { "grad_norm": 0.3375222682952881, "learning_rate": 2.8358234103505477e-05, "loss": 0.0203, "step": 52830 }, { "grad_norm": 0.3833443522453308, "learning_rate": 2.8339603972862806e-05, "loss": 0.0162, "step": 52840 }, { "grad_norm": 0.24220189452171326, "learning_rate": 2.83209775433894e-05, "loss": 0.0191, "step": 52850 }, { "grad_norm": 0.1852225363254547, "learning_rate": 2.8302354818268013e-05, "loss": 0.0195, "step": 52860 }, { "grad_norm": 0.3436907231807709, "learning_rate": 2.8283735800680754e-05, "loss": 0.0234, "step": 52870 }, { "grad_norm": 0.20586615800857544, "learning_rate": 2.826512049380913e-05, "loss": 0.0195, "step": 52880 }, { "grad_norm": 0.2674957811832428, "learning_rate": 2.8246508900833935e-05, "loss": 0.0271, "step": 52890 }, { "grad_norm": 0.2015347182750702, "learning_rate": 2.822790102493541e-05, "loss": 0.0202, "step": 52900 }, { "grad_norm": 0.2750607132911682, "learning_rate": 2.820929686929314e-05, "loss": 0.0198, "step": 52910 }, { "grad_norm": 0.22796964645385742, "learning_rate": 2.819069643708605e-05, "loss": 0.0207, "step": 52920 }, { "grad_norm": 0.2770877778530121, "learning_rate": 2.8172099731492484e-05, "loss": 0.02, "step": 52930 }, { "grad_norm": 0.2527632713317871, "learning_rate": 2.815350675569003e-05, "loss": 0.0256, "step": 52940 }, { "grad_norm": 0.22037620842456818, "learning_rate": 2.8134917512855807e-05, "loss": 0.0246, "step": 52950 }, { "grad_norm": 0.2694461941719055, "learning_rate": 2.8116332006166134e-05, "loss": 0.0134, "step": 52960 }, { "grad_norm": 0.31563684344291687, "learning_rate": 2.809775023879685e-05, "loss": 0.0297, "step": 52970 }, { "grad_norm": 0.2115899920463562, "learning_rate": 2.807917221392299e-05, "loss": 0.0191, "step": 52980 }, { "grad_norm": 0.23237168788909912, "learning_rate": 2.8060597934719067e-05, "loss": 0.0184, "step": 52990 }, { "grad_norm": 0.23763880133628845, "learning_rate": 2.8042027404358927e-05, "loss": 0.0219, "step": 53000 }, { "grad_norm": 0.3199527859687805, "learning_rate": 2.802346062601572e-05, "loss": 0.0201, "step": 53010 }, { "grad_norm": 0.38216879963874817, "learning_rate": 2.800489760286207e-05, "loss": 0.0286, "step": 53020 }, { "grad_norm": 0.3435550928115845, "learning_rate": 2.7986338338069834e-05, "loss": 0.0228, "step": 53030 }, { "grad_norm": 0.37589940428733826, "learning_rate": 2.7967782834810297e-05, "loss": 0.0229, "step": 53040 }, { "grad_norm": 0.222707137465477, "learning_rate": 2.794923109625409e-05, "loss": 0.0175, "step": 53050 }, { "grad_norm": 0.3081243336200714, "learning_rate": 2.7930683125571205e-05, "loss": 0.0193, "step": 53060 }, { "grad_norm": 0.4299439787864685, "learning_rate": 2.7912138925930997e-05, "loss": 0.0229, "step": 53070 }, { "grad_norm": 0.2348068505525589, "learning_rate": 2.7893598500502117e-05, "loss": 0.0302, "step": 53080 }, { "grad_norm": 0.3311437666416168, "learning_rate": 2.7875061852452644e-05, "loss": 0.0199, "step": 53090 }, { "grad_norm": 0.361443430185318, "learning_rate": 2.7856528984949982e-05, "loss": 0.0216, "step": 53100 }, { "grad_norm": 0.3123660385608673, "learning_rate": 2.7837999901160888e-05, "loss": 0.0192, "step": 53110 }, { "grad_norm": 0.13748560845851898, "learning_rate": 2.7819474604251484e-05, "loss": 0.0165, "step": 53120 }, { "grad_norm": 0.2687397301197052, "learning_rate": 2.780095309738725e-05, "loss": 0.0197, "step": 53130 }, { "grad_norm": 0.24769051373004913, "learning_rate": 2.778243538373294e-05, "loss": 0.017, "step": 53140 }, { "grad_norm": 0.11918815225362778, "learning_rate": 2.7763921466452826e-05, "loss": 0.0189, "step": 53150 }, { "grad_norm": 0.2524813711643219, "learning_rate": 2.7745411348710336e-05, "loss": 0.0289, "step": 53160 }, { "grad_norm": 0.4179994463920593, "learning_rate": 2.7726905033668395e-05, "loss": 0.0256, "step": 53170 }, { "grad_norm": 0.27192091941833496, "learning_rate": 2.7708402524489214e-05, "loss": 0.0214, "step": 53180 }, { "grad_norm": 0.30334168672561646, "learning_rate": 2.7689903824334364e-05, "loss": 0.0246, "step": 53190 }, { "grad_norm": 0.32940664887428284, "learning_rate": 2.7671408936364785e-05, "loss": 0.0208, "step": 53200 }, { "grad_norm": 0.19803833961486816, "learning_rate": 2.765291786374069e-05, "loss": 0.0163, "step": 53210 }, { "grad_norm": 0.13464872539043427, "learning_rate": 2.7634430609621775e-05, "loss": 0.0143, "step": 53220 }, { "grad_norm": 0.3194074332714081, "learning_rate": 2.7615947177166956e-05, "loss": 0.0195, "step": 53230 }, { "grad_norm": 0.6118699312210083, "learning_rate": 2.7597467569534553e-05, "loss": 0.0162, "step": 53240 }, { "grad_norm": 0.19868211448192596, "learning_rate": 2.757899178988226e-05, "loss": 0.0156, "step": 53250 }, { "grad_norm": 0.13754522800445557, "learning_rate": 2.7560519841367005e-05, "loss": 0.0192, "step": 53260 }, { "grad_norm": 0.33913201093673706, "learning_rate": 2.7542051727145237e-05, "loss": 0.0191, "step": 53270 }, { "grad_norm": 0.21721051633358002, "learning_rate": 2.7523587450372578e-05, "loss": 0.0223, "step": 53280 }, { "grad_norm": 0.28477150201797485, "learning_rate": 2.750512701420409e-05, "loss": 0.0182, "step": 53290 }, { "grad_norm": 0.3308212459087372, "learning_rate": 2.7486670421794158e-05, "loss": 0.0175, "step": 53300 }, { "grad_norm": 0.26775139570236206, "learning_rate": 2.7468217676296515e-05, "loss": 0.0247, "step": 53310 }, { "grad_norm": 0.30226200819015503, "learning_rate": 2.7449768780864226e-05, "loss": 0.0177, "step": 53320 }, { "grad_norm": 0.18541933596134186, "learning_rate": 2.7431323738649724e-05, "loss": 0.0177, "step": 53330 }, { "grad_norm": 0.341795951128006, "learning_rate": 2.7412882552804713e-05, "loss": 0.0234, "step": 53340 }, { "grad_norm": 0.21622076630592346, "learning_rate": 2.739444522648032e-05, "loss": 0.0135, "step": 53350 }, { "grad_norm": 0.19672194123268127, "learning_rate": 2.737601176282697e-05, "loss": 0.0229, "step": 53360 }, { "grad_norm": 0.3220633566379547, "learning_rate": 2.735758216499445e-05, "loss": 0.0185, "step": 53370 }, { "grad_norm": 0.2599659860134125, "learning_rate": 2.7339156436131864e-05, "loss": 0.0146, "step": 53380 }, { "grad_norm": 0.5028589367866516, "learning_rate": 2.7320734579387663e-05, "loss": 0.0194, "step": 53390 }, { "grad_norm": 0.25694361329078674, "learning_rate": 2.730231659790966e-05, "loss": 0.015, "step": 53400 }, { "grad_norm": 0.23034822940826416, "learning_rate": 2.728390249484494e-05, "loss": 0.0264, "step": 53410 }, { "grad_norm": 0.22576025128364563, "learning_rate": 2.7265492273340032e-05, "loss": 0.0188, "step": 53420 }, { "grad_norm": 0.22448086738586426, "learning_rate": 2.7247085936540678e-05, "loss": 0.0165, "step": 53430 }, { "grad_norm": 0.2151932269334793, "learning_rate": 2.722868348759205e-05, "loss": 0.021, "step": 53440 }, { "grad_norm": 0.23789666593074799, "learning_rate": 2.7210284929638635e-05, "loss": 0.02, "step": 53450 }, { "grad_norm": 0.26946330070495605, "learning_rate": 2.7191890265824183e-05, "loss": 0.02, "step": 53460 }, { "grad_norm": 0.2427995353937149, "learning_rate": 2.7173499499291926e-05, "loss": 0.0193, "step": 53470 }, { "grad_norm": 0.23149894177913666, "learning_rate": 2.7155112633184277e-05, "loss": 0.0156, "step": 53480 }, { "grad_norm": 0.3702005445957184, "learning_rate": 2.713672967064307e-05, "loss": 0.0218, "step": 53490 }, { "grad_norm": 0.36685940623283386, "learning_rate": 2.711835061480945e-05, "loss": 0.0171, "step": 53500 }, { "grad_norm": 0.23366491496562958, "learning_rate": 2.7099975468823896e-05, "loss": 0.0245, "step": 53510 }, { "grad_norm": 0.2661653459072113, "learning_rate": 2.708160423582622e-05, "loss": 0.0174, "step": 53520 }, { "grad_norm": 0.367301344871521, "learning_rate": 2.706323691895557e-05, "loss": 0.023, "step": 53530 }, { "grad_norm": 0.1650090217590332, "learning_rate": 2.70448735213504e-05, "loss": 0.0178, "step": 53540 }, { "grad_norm": 0.28106826543807983, "learning_rate": 2.702651404614852e-05, "loss": 0.0205, "step": 53550 }, { "grad_norm": 0.19882959127426147, "learning_rate": 2.7008158496487056e-05, "loss": 0.0225, "step": 53560 }, { "grad_norm": 0.2701888978481293, "learning_rate": 2.6989806875502487e-05, "loss": 0.017, "step": 53570 }, { "grad_norm": 0.19805142283439636, "learning_rate": 2.6971459186330584e-05, "loss": 0.0164, "step": 53580 }, { "grad_norm": 0.2594984173774719, "learning_rate": 2.695311543210648e-05, "loss": 0.0154, "step": 53590 }, { "grad_norm": 0.3396434485912323, "learning_rate": 2.693477561596463e-05, "loss": 0.021, "step": 53600 }, { "grad_norm": 0.35068562626838684, "learning_rate": 2.6916439741038756e-05, "loss": 0.0219, "step": 53610 }, { "grad_norm": 0.26134562492370605, "learning_rate": 2.689810781046203e-05, "loss": 0.0213, "step": 53620 }, { "grad_norm": 0.23169530928134918, "learning_rate": 2.6879779827366823e-05, "loss": 0.0185, "step": 53630 }, { "grad_norm": 0.23758564889431, "learning_rate": 2.6861455794884904e-05, "loss": 0.0203, "step": 53640 }, { "grad_norm": 0.4380934536457062, "learning_rate": 2.6843135716147373e-05, "loss": 0.0214, "step": 53650 }, { "grad_norm": 0.32591673731803894, "learning_rate": 2.6824819594284556e-05, "loss": 0.0231, "step": 53660 }, { "grad_norm": 0.29544952511787415, "learning_rate": 2.6806507432426275e-05, "loss": 0.0209, "step": 53670 }, { "grad_norm": 0.24192790687084198, "learning_rate": 2.6788199233701512e-05, "loss": 0.0251, "step": 53680 }, { "grad_norm": 0.21628622710704803, "learning_rate": 2.6769895001238652e-05, "loss": 0.0219, "step": 53690 }, { "grad_norm": 0.43998393416404724, "learning_rate": 2.67515947381654e-05, "loss": 0.021, "step": 53700 }, { "grad_norm": 0.23156802356243134, "learning_rate": 2.6733298447608745e-05, "loss": 0.018, "step": 53710 }, { "grad_norm": 0.19727209210395813, "learning_rate": 2.6715006132695074e-05, "loss": 0.0211, "step": 53720 }, { "grad_norm": 0.27957209944725037, "learning_rate": 2.6696717796549985e-05, "loss": 0.0288, "step": 53730 }, { "grad_norm": 0.35295894742012024, "learning_rate": 2.667843344229848e-05, "loss": 0.0223, "step": 53740 }, { "grad_norm": 0.42251327633857727, "learning_rate": 2.6660153073064846e-05, "loss": 0.0224, "step": 53750 }, { "grad_norm": 0.23887862265110016, "learning_rate": 2.664187669197271e-05, "loss": 0.0234, "step": 53760 }, { "grad_norm": 0.2974654734134674, "learning_rate": 2.6623604302145005e-05, "loss": 0.0191, "step": 53770 }, { "grad_norm": 0.3594464063644409, "learning_rate": 2.6605335906703975e-05, "loss": 0.023, "step": 53780 }, { "grad_norm": 0.2119125872850418, "learning_rate": 2.65870715087712e-05, "loss": 0.0235, "step": 53790 }, { "grad_norm": 0.22968412935733795, "learning_rate": 2.6568811111467573e-05, "loss": 0.015, "step": 53800 }, { "grad_norm": 0.3040134310722351, "learning_rate": 2.6550554717913258e-05, "loss": 0.0216, "step": 53810 }, { "grad_norm": 0.29138824343681335, "learning_rate": 2.6532302331227805e-05, "loss": 0.0203, "step": 53820 }, { "grad_norm": 0.23642700910568237, "learning_rate": 2.651405395453004e-05, "loss": 0.0177, "step": 53830 }, { "grad_norm": 0.2692227065563202, "learning_rate": 2.6495809590938115e-05, "loss": 0.0252, "step": 53840 }, { "grad_norm": 0.13191916048526764, "learning_rate": 2.64775692435695e-05, "loss": 0.0201, "step": 53850 }, { "grad_norm": 0.2572711408138275, "learning_rate": 2.6459332915540928e-05, "loss": 0.015, "step": 53860 }, { "grad_norm": 0.293323814868927, "learning_rate": 2.644110060996856e-05, "loss": 0.0269, "step": 53870 }, { "grad_norm": 0.3053918778896332, "learning_rate": 2.642287232996774e-05, "loss": 0.017, "step": 53880 }, { "grad_norm": 0.40225571393966675, "learning_rate": 2.6404648078653205e-05, "loss": 0.0198, "step": 53890 }, { "grad_norm": 0.32385194301605225, "learning_rate": 2.6386427859139002e-05, "loss": 0.0177, "step": 53900 }, { "grad_norm": 0.44698449969291687, "learning_rate": 2.63682116745384e-05, "loss": 0.0263, "step": 53910 }, { "grad_norm": 0.2810054123401642, "learning_rate": 2.6349999527964138e-05, "loss": 0.0216, "step": 53920 }, { "grad_norm": 0.24126900732517242, "learning_rate": 2.63317914225281e-05, "loss": 0.0278, "step": 53930 }, { "grad_norm": 0.26456573605537415, "learning_rate": 2.6313587361341585e-05, "loss": 0.0254, "step": 53940 }, { "grad_norm": 0.46402281522750854, "learning_rate": 2.6295387347515165e-05, "loss": 0.0255, "step": 53950 }, { "grad_norm": 0.17335984110832214, "learning_rate": 2.6277191384158727e-05, "loss": 0.0252, "step": 53960 }, { "grad_norm": 0.3937638998031616, "learning_rate": 2.6258999474381457e-05, "loss": 0.0246, "step": 53970 }, { "grad_norm": 0.322516530752182, "learning_rate": 2.624081162129186e-05, "loss": 0.0201, "step": 53980 }, { "grad_norm": 0.2221968173980713, "learning_rate": 2.6222627827997765e-05, "loss": 0.0314, "step": 53990 }, { "grad_norm": 0.3184608221054077, "learning_rate": 2.6204448097606236e-05, "loss": 0.0186, "step": 54000 }, { "grad_norm": 0.34378892183303833, "learning_rate": 2.6186272433223726e-05, "loss": 0.0248, "step": 54010 }, { "grad_norm": 0.28692471981048584, "learning_rate": 2.6168100837955943e-05, "loss": 0.0234, "step": 54020 }, { "grad_norm": 0.7033337950706482, "learning_rate": 2.6149933314907926e-05, "loss": 0.0307, "step": 54030 }, { "grad_norm": 0.28987789154052734, "learning_rate": 2.6131769867184e-05, "loss": 0.0282, "step": 54040 }, { "grad_norm": 0.216096431016922, "learning_rate": 2.611361049788783e-05, "loss": 0.0186, "step": 54050 }, { "grad_norm": 0.45318886637687683, "learning_rate": 2.6095455210122292e-05, "loss": 0.0219, "step": 54060 }, { "grad_norm": 0.2266804426908493, "learning_rate": 2.6077304006989712e-05, "loss": 0.0145, "step": 54070 }, { "grad_norm": 0.22271838784217834, "learning_rate": 2.6059156891591562e-05, "loss": 0.0213, "step": 54080 }, { "grad_norm": 0.16993746161460876, "learning_rate": 2.6041013867028718e-05, "loss": 0.0182, "step": 54090 }, { "grad_norm": 0.39746662974357605, "learning_rate": 2.6022874936401347e-05, "loss": 0.0158, "step": 54100 }, { "grad_norm": 0.2680805027484894, "learning_rate": 2.6004740102808832e-05, "loss": 0.0258, "step": 54110 }, { "grad_norm": 0.18296697735786438, "learning_rate": 2.598660936935e-05, "loss": 0.0153, "step": 54120 }, { "grad_norm": 0.31224942207336426, "learning_rate": 2.5968482739122845e-05, "loss": 0.023, "step": 54130 }, { "grad_norm": 0.31552526354789734, "learning_rate": 2.595036021522472e-05, "loss": 0.0192, "step": 54140 }, { "grad_norm": 0.20661115646362305, "learning_rate": 2.5932241800752278e-05, "loss": 0.02, "step": 54150 }, { "grad_norm": 0.2975567877292633, "learning_rate": 2.5914127498801453e-05, "loss": 0.0275, "step": 54160 }, { "grad_norm": 0.2889992892742157, "learning_rate": 2.5896017312467497e-05, "loss": 0.0213, "step": 54170 }, { "grad_norm": 0.3317245841026306, "learning_rate": 2.587791124484493e-05, "loss": 0.0165, "step": 54180 }, { "grad_norm": 0.17938712239265442, "learning_rate": 2.5859809299027615e-05, "loss": 0.0217, "step": 54190 }, { "grad_norm": 0.3402295410633087, "learning_rate": 2.5841711478108632e-05, "loss": 0.024, "step": 54200 }, { "grad_norm": 0.3258368670940399, "learning_rate": 2.582361778518043e-05, "loss": 0.0222, "step": 54210 }, { "grad_norm": 0.2861771285533905, "learning_rate": 2.580552822333472e-05, "loss": 0.0249, "step": 54220 }, { "grad_norm": 0.2795950174331665, "learning_rate": 2.578744279566252e-05, "loss": 0.0228, "step": 54230 }, { "grad_norm": 0.2585671544075012, "learning_rate": 2.576936150525413e-05, "loss": 0.0247, "step": 54240 }, { "grad_norm": 0.2305120825767517, "learning_rate": 2.5751284355199168e-05, "loss": 0.0154, "step": 54250 }, { "grad_norm": 0.23075740039348602, "learning_rate": 2.573321134858646e-05, "loss": 0.0164, "step": 54260 }, { "grad_norm": 0.29367393255233765, "learning_rate": 2.5715142488504286e-05, "loss": 0.0179, "step": 54270 }, { "grad_norm": 0.16168802976608276, "learning_rate": 2.5697077778040042e-05, "loss": 0.0185, "step": 54280 }, { "grad_norm": 0.3928499221801758, "learning_rate": 2.5679017220280522e-05, "loss": 0.0288, "step": 54290 }, { "grad_norm": 0.13503305613994598, "learning_rate": 2.5660960818311796e-05, "loss": 0.0217, "step": 54300 }, { "grad_norm": 0.2935873866081238, "learning_rate": 2.564290857521915e-05, "loss": 0.0224, "step": 54310 }, { "grad_norm": 0.22187313437461853, "learning_rate": 2.5624860494087298e-05, "loss": 0.0189, "step": 54320 }, { "grad_norm": 0.23295940458774567, "learning_rate": 2.5606816578000115e-05, "loss": 0.0223, "step": 54330 }, { "grad_norm": 0.19349399209022522, "learning_rate": 2.558877683004082e-05, "loss": 0.0209, "step": 54340 }, { "grad_norm": 0.201091930270195, "learning_rate": 2.557074125329192e-05, "loss": 0.0141, "step": 54350 }, { "grad_norm": 0.2011241465806961, "learning_rate": 2.5552709850835195e-05, "loss": 0.0178, "step": 54360 }, { "grad_norm": 0.28280776739120483, "learning_rate": 2.5534682625751738e-05, "loss": 0.0166, "step": 54370 }, { "grad_norm": 0.24587437510490417, "learning_rate": 2.551665958112186e-05, "loss": 0.0269, "step": 54380 }, { "grad_norm": 0.37859031558036804, "learning_rate": 2.549864072002527e-05, "loss": 0.0159, "step": 54390 }, { "grad_norm": 0.26091212034225464, "learning_rate": 2.5480626045540858e-05, "loss": 0.0205, "step": 54400 }, { "grad_norm": 0.32003581523895264, "learning_rate": 2.546261556074684e-05, "loss": 0.019, "step": 54410 }, { "grad_norm": 0.29124563932418823, "learning_rate": 2.5444609268720726e-05, "loss": 0.0198, "step": 54420 }, { "grad_norm": 0.2270756959915161, "learning_rate": 2.5426607172539297e-05, "loss": 0.0177, "step": 54430 }, { "grad_norm": 0.16716358065605164, "learning_rate": 2.5408609275278617e-05, "loss": 0.0272, "step": 54440 }, { "grad_norm": 0.1944500058889389, "learning_rate": 2.5390615580014055e-05, "loss": 0.0175, "step": 54450 }, { "grad_norm": 0.2068471610546112, "learning_rate": 2.5372626089820207e-05, "loss": 0.0252, "step": 54460 }, { "grad_norm": 0.1993086040019989, "learning_rate": 2.5354640807770997e-05, "loss": 0.0138, "step": 54470 }, { "grad_norm": 0.3297176659107208, "learning_rate": 2.5336659736939622e-05, "loss": 0.0172, "step": 54480 }, { "grad_norm": 0.28092673420906067, "learning_rate": 2.5318682880398554e-05, "loss": 0.0223, "step": 54490 }, { "grad_norm": 0.224690243601799, "learning_rate": 2.530071024121956e-05, "loss": 0.0191, "step": 54500 }, { "grad_norm": 0.3961135149002075, "learning_rate": 2.5282741822473627e-05, "loss": 0.0186, "step": 54510 }, { "grad_norm": 0.3225138783454895, "learning_rate": 2.526477762723114e-05, "loss": 0.0178, "step": 54520 }, { "grad_norm": 0.17967791855335236, "learning_rate": 2.5246817658561618e-05, "loss": 0.0207, "step": 54530 }, { "grad_norm": 0.17963343858718872, "learning_rate": 2.5228861919533965e-05, "loss": 0.0185, "step": 54540 }, { "grad_norm": 0.20243936777114868, "learning_rate": 2.5210910413216326e-05, "loss": 0.0277, "step": 54550 }, { "grad_norm": 0.23326683044433594, "learning_rate": 2.5192963142676086e-05, "loss": 0.0142, "step": 54560 }, { "grad_norm": 0.2891051769256592, "learning_rate": 2.517502011098001e-05, "loss": 0.018, "step": 54570 }, { "grad_norm": 0.2363438755273819, "learning_rate": 2.5157081321193987e-05, "loss": 0.0181, "step": 54580 }, { "grad_norm": 0.18243660032749176, "learning_rate": 2.5139146776383356e-05, "loss": 0.0141, "step": 54590 }, { "grad_norm": 0.34789952635765076, "learning_rate": 2.5121216479612575e-05, "loss": 0.0228, "step": 54600 }, { "grad_norm": 0.20485791563987732, "learning_rate": 2.510329043394546e-05, "loss": 0.0253, "step": 54610 }, { "grad_norm": 0.31812646985054016, "learning_rate": 2.508536864244508e-05, "loss": 0.0233, "step": 54620 }, { "grad_norm": 0.28107312321662903, "learning_rate": 2.5067451108173778e-05, "loss": 0.0184, "step": 54630 }, { "grad_norm": 0.2243795245885849, "learning_rate": 2.5049537834193204e-05, "loss": 0.0178, "step": 54640 }, { "grad_norm": 0.16635914146900177, "learning_rate": 2.5031628823564194e-05, "loss": 0.0176, "step": 54650 }, { "grad_norm": 0.27675941586494446, "learning_rate": 2.5013724079346933e-05, "loss": 0.0156, "step": 54660 }, { "grad_norm": 0.2144007831811905, "learning_rate": 2.4995823604600854e-05, "loss": 0.0143, "step": 54670 }, { "grad_norm": 0.178584486246109, "learning_rate": 2.497792740238465e-05, "loss": 0.0185, "step": 54680 }, { "grad_norm": 0.33932721614837646, "learning_rate": 2.49600354757563e-05, "loss": 0.0224, "step": 54690 }, { "grad_norm": 0.16659115254878998, "learning_rate": 2.494214782777306e-05, "loss": 0.0208, "step": 54700 }, { "grad_norm": 0.2302631437778473, "learning_rate": 2.4924264461491386e-05, "loss": 0.0203, "step": 54710 }, { "grad_norm": 0.35703369975090027, "learning_rate": 2.4906385379967133e-05, "loss": 0.0219, "step": 54720 }, { "grad_norm": 0.25384020805358887, "learning_rate": 2.4888510586255285e-05, "loss": 0.0181, "step": 54730 }, { "grad_norm": 0.28811463713645935, "learning_rate": 2.487064008341018e-05, "loss": 0.0191, "step": 54740 }, { "grad_norm": 0.2955230474472046, "learning_rate": 2.4852773874485407e-05, "loss": 0.0173, "step": 54750 }, { "grad_norm": 0.27734047174453735, "learning_rate": 2.483491196253377e-05, "loss": 0.0186, "step": 54760 }, { "grad_norm": 0.3104941248893738, "learning_rate": 2.4817054350607443e-05, "loss": 0.0238, "step": 54770 }, { "grad_norm": 0.2455163449048996, "learning_rate": 2.4799201041757742e-05, "loss": 0.0151, "step": 54780 }, { "grad_norm": 0.39975598454475403, "learning_rate": 2.4781352039035373e-05, "loss": 0.0174, "step": 54790 }, { "grad_norm": 0.29198041558265686, "learning_rate": 2.4763507345490194e-05, "loss": 0.0182, "step": 54800 }, { "grad_norm": 0.17933273315429688, "learning_rate": 2.4745666964171386e-05, "loss": 0.0183, "step": 54810 }, { "grad_norm": 0.35522937774658203, "learning_rate": 2.4727830898127387e-05, "loss": 0.0214, "step": 54820 }, { "grad_norm": 0.17921271920204163, "learning_rate": 2.4709999150405895e-05, "loss": 0.0244, "step": 54830 }, { "grad_norm": 0.30680254101753235, "learning_rate": 2.469217172405388e-05, "loss": 0.0196, "step": 54840 }, { "grad_norm": 0.17870421707630157, "learning_rate": 2.4674348622117527e-05, "loss": 0.014, "step": 54850 }, { "grad_norm": 0.2640298902988434, "learning_rate": 2.465652984764234e-05, "loss": 0.0157, "step": 54860 }, { "grad_norm": 0.21864916384220123, "learning_rate": 2.4638715403673056e-05, "loss": 0.0166, "step": 54870 }, { "grad_norm": 0.4384072422981262, "learning_rate": 2.4620905293253676e-05, "loss": 0.0214, "step": 54880 }, { "grad_norm": 0.2863263189792633, "learning_rate": 2.4603099519427463e-05, "loss": 0.0221, "step": 54890 }, { "grad_norm": 0.2695430815219879, "learning_rate": 2.458529808523695e-05, "loss": 0.016, "step": 54900 }, { "grad_norm": 0.24595649540424347, "learning_rate": 2.456750099372387e-05, "loss": 0.0147, "step": 54910 }, { "grad_norm": 0.23298875987529755, "learning_rate": 2.4549708247929327e-05, "loss": 0.0277, "step": 54920 }, { "grad_norm": 0.2725452184677124, "learning_rate": 2.4531919850893554e-05, "loss": 0.015, "step": 54930 }, { "grad_norm": 0.20676326751708984, "learning_rate": 2.4514135805656125e-05, "loss": 0.015, "step": 54940 }, { "grad_norm": 0.35125043988227844, "learning_rate": 2.449635611525587e-05, "loss": 0.0279, "step": 54950 }, { "grad_norm": 0.23396700620651245, "learning_rate": 2.447858078273079e-05, "loss": 0.0201, "step": 54960 }, { "grad_norm": 0.2060713917016983, "learning_rate": 2.4460809811118275e-05, "loss": 0.0245, "step": 54970 }, { "grad_norm": 0.18681791424751282, "learning_rate": 2.444304320345483e-05, "loss": 0.0168, "step": 54980 }, { "grad_norm": 0.24325475096702576, "learning_rate": 2.4425280962776354e-05, "loss": 0.0233, "step": 54990 }, { "grad_norm": 0.29390379786491394, "learning_rate": 2.4407523092117875e-05, "loss": 0.0195, "step": 55000 }, { "grad_norm": 0.2732887268066406, "learning_rate": 2.438976959451374e-05, "loss": 0.0183, "step": 55010 }, { "grad_norm": 0.31114134192466736, "learning_rate": 2.4372020472997565e-05, "loss": 0.0175, "step": 55020 }, { "grad_norm": 0.2914748191833496, "learning_rate": 2.4354275730602122e-05, "loss": 0.0184, "step": 55030 }, { "grad_norm": 0.27935317158699036, "learning_rate": 2.433653537035958e-05, "loss": 0.0184, "step": 55040 }, { "grad_norm": 0.32375138998031616, "learning_rate": 2.431879939530123e-05, "loss": 0.0217, "step": 55050 }, { "grad_norm": 0.2585068941116333, "learning_rate": 2.4301067808457684e-05, "loss": 0.0178, "step": 55060 }, { "grad_norm": 0.22875399887561798, "learning_rate": 2.428334061285878e-05, "loss": 0.0198, "step": 55070 }, { "grad_norm": 0.41638872027397156, "learning_rate": 2.426561781153361e-05, "loss": 0.0181, "step": 55080 }, { "grad_norm": 0.31079116463661194, "learning_rate": 2.424789940751052e-05, "loss": 0.0139, "step": 55090 }, { "grad_norm": 0.18078115582466125, "learning_rate": 2.423018540381712e-05, "loss": 0.0177, "step": 55100 }, { "grad_norm": 0.24468109011650085, "learning_rate": 2.421247580348021e-05, "loss": 0.0202, "step": 55110 }, { "grad_norm": 0.3683489263057709, "learning_rate": 2.4194770609525897e-05, "loss": 0.0179, "step": 55120 }, { "grad_norm": 0.2339937537908554, "learning_rate": 2.417706982497951e-05, "loss": 0.0187, "step": 55130 }, { "grad_norm": 0.31104499101638794, "learning_rate": 2.415937345286563e-05, "loss": 0.0229, "step": 55140 }, { "grad_norm": 0.2709771394729614, "learning_rate": 2.4141681496208087e-05, "loss": 0.016, "step": 55150 }, { "grad_norm": 0.2159743309020996, "learning_rate": 2.4123993958029946e-05, "loss": 0.0145, "step": 55160 }, { "grad_norm": 0.3933449983596802, "learning_rate": 2.4106310841353548e-05, "loss": 0.0161, "step": 55170 }, { "grad_norm": 0.26991498470306396, "learning_rate": 2.4088632149200398e-05, "loss": 0.0231, "step": 55180 }, { "grad_norm": 0.2722584009170532, "learning_rate": 2.4070957884591367e-05, "loss": 0.0218, "step": 55190 }, { "grad_norm": 0.5229691863059998, "learning_rate": 2.4053288050546464e-05, "loss": 0.0268, "step": 55200 }, { "grad_norm": 0.23256322741508484, "learning_rate": 2.403562265008498e-05, "loss": 0.0245, "step": 55210 }, { "grad_norm": 0.4448957145214081, "learning_rate": 2.4017961686225483e-05, "loss": 0.0244, "step": 55220 }, { "grad_norm": 0.18863597512245178, "learning_rate": 2.400030516198568e-05, "loss": 0.0271, "step": 55230 }, { "grad_norm": 0.29723304510116577, "learning_rate": 2.3982653080382673e-05, "loss": 0.018, "step": 55240 }, { "grad_norm": 0.2428157776594162, "learning_rate": 2.396500544443266e-05, "loss": 0.0148, "step": 55250 }, { "grad_norm": 0.31559494137763977, "learning_rate": 2.3947362257151156e-05, "loss": 0.017, "step": 55260 }, { "grad_norm": 0.4035339653491974, "learning_rate": 2.39297235215529e-05, "loss": 0.0249, "step": 55270 }, { "grad_norm": 0.295274019241333, "learning_rate": 2.3912089240651873e-05, "loss": 0.02, "step": 55280 }, { "grad_norm": 0.2690828740596771, "learning_rate": 2.389445941746129e-05, "loss": 0.0282, "step": 55290 }, { "grad_norm": 0.3492707908153534, "learning_rate": 2.3876834054993625e-05, "loss": 0.0184, "step": 55300 }, { "grad_norm": 0.24738620221614838, "learning_rate": 2.3859213156260522e-05, "loss": 0.0185, "step": 55310 }, { "grad_norm": 0.1985957771539688, "learning_rate": 2.3841596724272948e-05, "loss": 0.0206, "step": 55320 }, { "grad_norm": 0.16761890053749084, "learning_rate": 2.382398476204106e-05, "loss": 0.0211, "step": 55330 }, { "grad_norm": 0.24159160256385803, "learning_rate": 2.3806377272574254e-05, "loss": 0.0199, "step": 55340 }, { "grad_norm": 0.38635796308517456, "learning_rate": 2.3788774258881174e-05, "loss": 0.0173, "step": 55350 }, { "grad_norm": 0.34735769033432007, "learning_rate": 2.37711757239697e-05, "loss": 0.0208, "step": 55360 }, { "grad_norm": 0.35966238379478455, "learning_rate": 2.3753581670846954e-05, "loss": 0.018, "step": 55370 }, { "grad_norm": 0.1981518417596817, "learning_rate": 2.3735992102519216e-05, "loss": 0.0181, "step": 55380 }, { "grad_norm": 0.3054204285144806, "learning_rate": 2.371840702199215e-05, "loss": 0.0175, "step": 55390 }, { "grad_norm": 0.28671252727508545, "learning_rate": 2.3700826432270494e-05, "loss": 0.0177, "step": 55400 }, { "grad_norm": 0.23021307587623596, "learning_rate": 2.3683250336358326e-05, "loss": 0.0291, "step": 55410 }, { "grad_norm": 0.2969226837158203, "learning_rate": 2.3665678737258923e-05, "loss": 0.0212, "step": 55420 }, { "grad_norm": 0.24895459413528442, "learning_rate": 2.3648111637974745e-05, "loss": 0.0161, "step": 55430 }, { "grad_norm": 0.2501356601715088, "learning_rate": 2.36305490415076e-05, "loss": 0.028, "step": 55440 }, { "grad_norm": 0.32806044816970825, "learning_rate": 2.3612990950858392e-05, "loss": 0.0173, "step": 55450 }, { "grad_norm": 0.10900106281042099, "learning_rate": 2.359543736902735e-05, "loss": 0.0167, "step": 55460 }, { "grad_norm": 0.41718050837516785, "learning_rate": 2.3577888299013896e-05, "loss": 0.0178, "step": 55470 }, { "grad_norm": 0.3432043492794037, "learning_rate": 2.3560343743816683e-05, "loss": 0.017, "step": 55480 }, { "grad_norm": 0.20135079324245453, "learning_rate": 2.354280370643362e-05, "loss": 0.0237, "step": 55490 }, { "grad_norm": 0.20622242987155914, "learning_rate": 2.3525268189861777e-05, "loss": 0.0174, "step": 55500 }, { "grad_norm": 0.1817127913236618, "learning_rate": 2.3507737197097513e-05, "loss": 0.0164, "step": 55510 }, { "grad_norm": 0.29298582673072815, "learning_rate": 2.34902107311364e-05, "loss": 0.0199, "step": 55520 }, { "grad_norm": 0.29643765091896057, "learning_rate": 2.3472688794973225e-05, "loss": 0.0213, "step": 55530 }, { "grad_norm": 0.283123642206192, "learning_rate": 2.3455171391602016e-05, "loss": 0.0212, "step": 55540 }, { "grad_norm": 0.38085514307022095, "learning_rate": 2.343765852401601e-05, "loss": 0.0205, "step": 55550 }, { "grad_norm": 0.1794070601463318, "learning_rate": 2.342015019520768e-05, "loss": 0.029, "step": 55560 }, { "grad_norm": 0.1778079718351364, "learning_rate": 2.3402646408168742e-05, "loss": 0.0153, "step": 55570 }, { "grad_norm": 0.37951740622520447, "learning_rate": 2.3385147165890074e-05, "loss": 0.0223, "step": 55580 }, { "grad_norm": 0.2822319567203522, "learning_rate": 2.336765247136184e-05, "loss": 0.0195, "step": 55590 }, { "grad_norm": 0.37494930624961853, "learning_rate": 2.33501623275734e-05, "loss": 0.0264, "step": 55600 }, { "grad_norm": 0.21456287801265717, "learning_rate": 2.333267673751334e-05, "loss": 0.0142, "step": 55610 }, { "grad_norm": 0.2310762107372284, "learning_rate": 2.33151957041695e-05, "loss": 0.0171, "step": 55620 }, { "grad_norm": 0.2945444881916046, "learning_rate": 2.329771923052884e-05, "loss": 0.0211, "step": 55630 }, { "grad_norm": 0.32483920454978943, "learning_rate": 2.3280247319577697e-05, "loss": 0.0184, "step": 55640 }, { "grad_norm": 0.27883508801460266, "learning_rate": 2.3262779974301473e-05, "loss": 0.0247, "step": 55650 }, { "grad_norm": 0.3394421935081482, "learning_rate": 2.3245317197684895e-05, "loss": 0.0183, "step": 55660 }, { "grad_norm": 0.21120266616344452, "learning_rate": 2.322785899271188e-05, "loss": 0.0185, "step": 55670 }, { "grad_norm": 0.2901175320148468, "learning_rate": 2.3210405362365507e-05, "loss": 0.0284, "step": 55680 }, { "grad_norm": 0.28452441096305847, "learning_rate": 2.31929563096282e-05, "loss": 0.0207, "step": 55690 }, { "grad_norm": 0.20243021845817566, "learning_rate": 2.317551183748146e-05, "loss": 0.0202, "step": 55700 }, { "grad_norm": 0.2344704121351242, "learning_rate": 2.3158071948906103e-05, "loss": 0.0166, "step": 55710 }, { "grad_norm": 0.48341917991638184, "learning_rate": 2.314063664688212e-05, "loss": 0.0185, "step": 55720 }, { "grad_norm": 0.2970225214958191, "learning_rate": 2.3123205934388725e-05, "loss": 0.0161, "step": 55730 }, { "grad_norm": 0.32869917154312134, "learning_rate": 2.310577981440436e-05, "loss": 0.0382, "step": 55740 }, { "grad_norm": 0.28745874762535095, "learning_rate": 2.3088358289906665e-05, "loss": 0.0185, "step": 55750 }, { "grad_norm": 0.3200635313987732, "learning_rate": 2.307094136387252e-05, "loss": 0.0173, "step": 55760 }, { "grad_norm": 0.3112211525440216, "learning_rate": 2.305352903927796e-05, "loss": 0.0252, "step": 55770 }, { "grad_norm": 0.2033843845129013, "learning_rate": 2.303612131909831e-05, "loss": 0.0153, "step": 55780 }, { "grad_norm": 0.2838995158672333, "learning_rate": 2.3018718206308054e-05, "loss": 0.0233, "step": 55790 }, { "grad_norm": 0.449760764837265, "learning_rate": 2.3001319703880925e-05, "loss": 0.0224, "step": 55800 }, { "grad_norm": 0.22248475253582, "learning_rate": 2.2983925814789835e-05, "loss": 0.0181, "step": 55810 }, { "grad_norm": 0.2888225317001343, "learning_rate": 2.2966536542006957e-05, "loss": 0.0209, "step": 55820 }, { "grad_norm": 0.22166694700717926, "learning_rate": 2.294915188850358e-05, "loss": 0.0175, "step": 55830 }, { "grad_norm": 0.3257429003715515, "learning_rate": 2.2931771857250333e-05, "loss": 0.0223, "step": 55840 }, { "grad_norm": 0.1833931803703308, "learning_rate": 2.2914396451216946e-05, "loss": 0.0166, "step": 55850 }, { "grad_norm": 0.35428038239479065, "learning_rate": 2.2897025673372412e-05, "loss": 0.0164, "step": 55860 }, { "grad_norm": 0.20594732463359833, "learning_rate": 2.287965952668494e-05, "loss": 0.0189, "step": 55870 }, { "grad_norm": 0.25720247626304626, "learning_rate": 2.2862298014121873e-05, "loss": 0.0185, "step": 55880 }, { "grad_norm": 0.3213188946247101, "learning_rate": 2.28449411386499e-05, "loss": 0.0243, "step": 55890 }, { "grad_norm": 0.2253555804491043, "learning_rate": 2.282758890323477e-05, "loss": 0.0179, "step": 55900 }, { "grad_norm": 0.22175505757331848, "learning_rate": 2.2810241310841528e-05, "loss": 0.0179, "step": 55910 }, { "grad_norm": 0.233973428606987, "learning_rate": 2.2792898364434412e-05, "loss": 0.0161, "step": 55920 }, { "grad_norm": 0.2688196897506714, "learning_rate": 2.2775560066976843e-05, "loss": 0.019, "step": 55930 }, { "grad_norm": 0.24146100878715515, "learning_rate": 2.275822642143147e-05, "loss": 0.0209, "step": 55940 }, { "grad_norm": 0.23961572349071503, "learning_rate": 2.274089743076014e-05, "loss": 0.0146, "step": 55950 }, { "grad_norm": 0.37146294116973877, "learning_rate": 2.2723573097923917e-05, "loss": 0.0203, "step": 55960 }, { "grad_norm": 0.2383880913257599, "learning_rate": 2.2706253425883024e-05, "loss": 0.0129, "step": 55970 }, { "grad_norm": 0.4534856379032135, "learning_rate": 2.2688938417596933e-05, "loss": 0.0225, "step": 55980 }, { "grad_norm": 0.213136225938797, "learning_rate": 2.2671628076024305e-05, "loss": 0.0179, "step": 55990 }, { "grad_norm": 0.2238943725824356, "learning_rate": 2.2654322404123008e-05, "loss": 0.0163, "step": 56000 }, { "grad_norm": 0.19329243898391724, "learning_rate": 2.2637021404850105e-05, "loss": 0.0136, "step": 56010 }, { "grad_norm": 0.29049429297447205, "learning_rate": 2.2619725081161885e-05, "loss": 0.0231, "step": 56020 }, { "grad_norm": 0.24048109352588654, "learning_rate": 2.260243343601376e-05, "loss": 0.0213, "step": 56030 }, { "grad_norm": 0.21893949806690216, "learning_rate": 2.2585146472360473e-05, "loss": 0.0202, "step": 56040 }, { "grad_norm": 0.31021827459335327, "learning_rate": 2.2567864193155834e-05, "loss": 0.0197, "step": 56050 }, { "grad_norm": 0.3474430441856384, "learning_rate": 2.255058660135294e-05, "loss": 0.0222, "step": 56060 }, { "grad_norm": 0.512179970741272, "learning_rate": 2.253331369990407e-05, "loss": 0.0204, "step": 56070 }, { "grad_norm": 0.264718621969223, "learning_rate": 2.251604549176063e-05, "loss": 0.0198, "step": 56080 }, { "grad_norm": 0.1524503231048584, "learning_rate": 2.249878197987337e-05, "loss": 0.0158, "step": 56090 }, { "grad_norm": 0.2788522243499756, "learning_rate": 2.2481523167192087e-05, "loss": 0.016, "step": 56100 }, { "grad_norm": 0.17145974934101105, "learning_rate": 2.2464269056665867e-05, "loss": 0.0128, "step": 56110 }, { "grad_norm": 0.29121994972229004, "learning_rate": 2.2447019651242958e-05, "loss": 0.0198, "step": 56120 }, { "grad_norm": 0.25654715299606323, "learning_rate": 2.2429774953870818e-05, "loss": 0.0136, "step": 56130 }, { "grad_norm": 0.3564276695251465, "learning_rate": 2.241253496749611e-05, "loss": 0.0239, "step": 56140 }, { "grad_norm": 0.2560678720474243, "learning_rate": 2.2395299695064614e-05, "loss": 0.0166, "step": 56150 }, { "grad_norm": 0.23767057061195374, "learning_rate": 2.237806913952145e-05, "loss": 0.0161, "step": 56160 }, { "grad_norm": 0.3309471309185028, "learning_rate": 2.2360843303810798e-05, "loss": 0.0147, "step": 56170 }, { "grad_norm": 0.21262486279010773, "learning_rate": 2.2343622190876084e-05, "loss": 0.0165, "step": 56180 }, { "grad_norm": 0.2228023260831833, "learning_rate": 2.2326405803659935e-05, "loss": 0.017, "step": 56190 }, { "grad_norm": 0.4641167223453522, "learning_rate": 2.230919414510416e-05, "loss": 0.0223, "step": 56200 }, { "grad_norm": 0.45158135890960693, "learning_rate": 2.229198721814976e-05, "loss": 0.021, "step": 56210 }, { "grad_norm": 0.303934246301651, "learning_rate": 2.2274785025736948e-05, "loss": 0.0142, "step": 56220 }, { "grad_norm": 0.299439936876297, "learning_rate": 2.225758757080507e-05, "loss": 0.0204, "step": 56230 }, { "grad_norm": 0.40221238136291504, "learning_rate": 2.2240394856292723e-05, "loss": 0.0212, "step": 56240 }, { "grad_norm": 0.19879992306232452, "learning_rate": 2.2223206885137664e-05, "loss": 0.0144, "step": 56250 }, { "grad_norm": 0.278847873210907, "learning_rate": 2.2206023660276853e-05, "loss": 0.0194, "step": 56260 }, { "grad_norm": 0.20596787333488464, "learning_rate": 2.218884518464645e-05, "loss": 0.0183, "step": 56270 }, { "grad_norm": 0.4436582922935486, "learning_rate": 2.2171671461181732e-05, "loss": 0.02, "step": 56280 }, { "grad_norm": 0.2098793387413025, "learning_rate": 2.2154502492817292e-05, "loss": 0.021, "step": 56290 }, { "grad_norm": 0.28320032358169556, "learning_rate": 2.2137338282486782e-05, "loss": 0.0141, "step": 56300 }, { "grad_norm": 0.13066881895065308, "learning_rate": 2.2120178833123113e-05, "loss": 0.017, "step": 56310 }, { "grad_norm": 0.2555553913116455, "learning_rate": 2.210302414765838e-05, "loss": 0.0143, "step": 56320 }, { "grad_norm": 0.3429144024848938, "learning_rate": 2.2085874229023808e-05, "loss": 0.0132, "step": 56330 }, { "grad_norm": 0.37489670515060425, "learning_rate": 2.2068729080149907e-05, "loss": 0.0149, "step": 56340 }, { "grad_norm": 0.29707062244415283, "learning_rate": 2.205158870396625e-05, "loss": 0.02, "step": 56350 }, { "grad_norm": 0.28204652667045593, "learning_rate": 2.2034453103401732e-05, "loss": 0.0169, "step": 56360 }, { "grad_norm": 0.2789798080921173, "learning_rate": 2.20173222813843e-05, "loss": 0.0143, "step": 56370 }, { "grad_norm": 0.24459844827651978, "learning_rate": 2.200019624084116e-05, "loss": 0.0192, "step": 56380 }, { "grad_norm": 0.55711430311203, "learning_rate": 2.1983074984698687e-05, "loss": 0.0174, "step": 56390 }, { "grad_norm": 0.7671307921409607, "learning_rate": 2.1965958515882433e-05, "loss": 0.0262, "step": 56400 }, { "grad_norm": 0.2111169397830963, "learning_rate": 2.1948846837317162e-05, "loss": 0.0183, "step": 56410 }, { "grad_norm": 0.3485848903656006, "learning_rate": 2.1931739951926738e-05, "loss": 0.0152, "step": 56420 }, { "grad_norm": 0.1638270616531372, "learning_rate": 2.191463786263429e-05, "loss": 0.0145, "step": 56430 }, { "grad_norm": 0.26430296897888184, "learning_rate": 2.1897540572362095e-05, "loss": 0.0206, "step": 56440 }, { "grad_norm": 0.2939148545265198, "learning_rate": 2.1880448084031614e-05, "loss": 0.0164, "step": 56450 }, { "grad_norm": 0.20097118616104126, "learning_rate": 2.1863360400563482e-05, "loss": 0.0176, "step": 56460 }, { "grad_norm": 0.33556053042411804, "learning_rate": 2.184627752487754e-05, "loss": 0.0213, "step": 56470 }, { "grad_norm": 0.22285036742687225, "learning_rate": 2.1829199459892725e-05, "loss": 0.0194, "step": 56480 }, { "grad_norm": 0.2823503613471985, "learning_rate": 2.1812126208527282e-05, "loss": 0.0187, "step": 56490 }, { "grad_norm": 0.17643027007579803, "learning_rate": 2.179505777369852e-05, "loss": 0.0177, "step": 56500 }, { "grad_norm": 0.15853111445903778, "learning_rate": 2.1777994158322974e-05, "loss": 0.0174, "step": 56510 }, { "grad_norm": 0.32758665084838867, "learning_rate": 2.176093536531637e-05, "loss": 0.029, "step": 56520 }, { "grad_norm": 0.17683525383472443, "learning_rate": 2.1743881397593537e-05, "loss": 0.0169, "step": 56530 }, { "grad_norm": 0.3420083224773407, "learning_rate": 2.1726832258068595e-05, "loss": 0.0271, "step": 56540 }, { "grad_norm": 0.40792617201805115, "learning_rate": 2.170978794965472e-05, "loss": 0.0218, "step": 56550 }, { "grad_norm": 0.26991599798202515, "learning_rate": 2.169274847526438e-05, "loss": 0.0179, "step": 56560 }, { "grad_norm": 0.2769797444343567, "learning_rate": 2.1675713837809103e-05, "loss": 0.0199, "step": 56570 }, { "grad_norm": 0.25650250911712646, "learning_rate": 2.1658684040199655e-05, "loss": 0.0235, "step": 56580 }, { "grad_norm": 0.25935354828834534, "learning_rate": 2.1641659085345974e-05, "loss": 0.0181, "step": 56590 }, { "grad_norm": 0.24995990097522736, "learning_rate": 2.1624638976157154e-05, "loss": 0.0185, "step": 56600 }, { "grad_norm": 0.4399595856666565, "learning_rate": 2.1607623715541476e-05, "loss": 0.0279, "step": 56610 }, { "grad_norm": 0.23230469226837158, "learning_rate": 2.159061330640636e-05, "loss": 0.0131, "step": 56620 }, { "grad_norm": 0.23645929992198944, "learning_rate": 2.1573607751658425e-05, "loss": 0.0154, "step": 56630 }, { "grad_norm": 0.18629096448421478, "learning_rate": 2.155660705420347e-05, "loss": 0.0177, "step": 56640 }, { "grad_norm": 0.21869422495365143, "learning_rate": 2.153961121694644e-05, "loss": 0.0163, "step": 56650 }, { "grad_norm": 0.214909166097641, "learning_rate": 2.152262024279145e-05, "loss": 0.0306, "step": 56660 }, { "grad_norm": 0.4092141091823578, "learning_rate": 2.150563413464183e-05, "loss": 0.0187, "step": 56670 }, { "grad_norm": 0.2714375853538513, "learning_rate": 2.1488652895399974e-05, "loss": 0.0132, "step": 56680 }, { "grad_norm": 0.2836550772190094, "learning_rate": 2.1471676527967587e-05, "loss": 0.0201, "step": 56690 }, { "grad_norm": 0.31448549032211304, "learning_rate": 2.1454705035245414e-05, "loss": 0.026, "step": 56700 }, { "grad_norm": 0.16634351015090942, "learning_rate": 2.143773842013343e-05, "loss": 0.0144, "step": 56710 }, { "grad_norm": 0.22151058912277222, "learning_rate": 2.1420776685530796e-05, "loss": 0.0174, "step": 56720 }, { "grad_norm": 0.27330344915390015, "learning_rate": 2.1403819834335742e-05, "loss": 0.0148, "step": 56730 }, { "grad_norm": 0.12789852917194366, "learning_rate": 2.1386867869445808e-05, "loss": 0.0203, "step": 56740 }, { "grad_norm": 0.24548111855983734, "learning_rate": 2.1369920793757548e-05, "loss": 0.0224, "step": 56750 }, { "grad_norm": 0.19677165150642395, "learning_rate": 2.1352978610166828e-05, "loss": 0.0167, "step": 56760 }, { "grad_norm": 0.39995384216308594, "learning_rate": 2.1336041321568546e-05, "loss": 0.0201, "step": 56770 }, { "grad_norm": 0.23079469799995422, "learning_rate": 2.131910893085684e-05, "loss": 0.0158, "step": 56780 }, { "grad_norm": 0.1825895607471466, "learning_rate": 2.1302181440925012e-05, "loss": 0.0204, "step": 56790 }, { "grad_norm": 0.1576451063156128, "learning_rate": 2.128525885466546e-05, "loss": 0.018, "step": 56800 }, { "grad_norm": 0.40008631348609924, "learning_rate": 2.1268341174969847e-05, "loss": 0.0223, "step": 56810 }, { "grad_norm": 0.36498597264289856, "learning_rate": 2.1251428404728894e-05, "loss": 0.0166, "step": 56820 }, { "grad_norm": 0.20864973962306976, "learning_rate": 2.123452054683256e-05, "loss": 0.015, "step": 56830 }, { "grad_norm": 0.271248996257782, "learning_rate": 2.1217617604169927e-05, "loss": 0.0215, "step": 56840 }, { "grad_norm": 0.2914033532142639, "learning_rate": 2.120071957962924e-05, "loss": 0.0193, "step": 56850 }, { "grad_norm": 0.23655128479003906, "learning_rate": 2.1183826476097917e-05, "loss": 0.0224, "step": 56860 }, { "grad_norm": 0.3378297686576843, "learning_rate": 2.116693829646254e-05, "loss": 0.0193, "step": 56870 }, { "grad_norm": 0.40594133734703064, "learning_rate": 2.1150055043608806e-05, "loss": 0.0184, "step": 56880 }, { "grad_norm": 0.45829376578330994, "learning_rate": 2.113317672042162e-05, "loss": 0.0257, "step": 56890 }, { "grad_norm": 0.23314650356769562, "learning_rate": 2.1116303329785025e-05, "loss": 0.0155, "step": 56900 }, { "grad_norm": 0.23927085101604462, "learning_rate": 2.109943487458222e-05, "loss": 0.0243, "step": 56910 }, { "grad_norm": 0.253182977437973, "learning_rate": 2.1082571357695574e-05, "loss": 0.0177, "step": 56920 }, { "grad_norm": 0.23528821766376495, "learning_rate": 2.1065712782006557e-05, "loss": 0.0175, "step": 56930 }, { "grad_norm": 0.20257975161075592, "learning_rate": 2.104885915039591e-05, "loss": 0.0155, "step": 56940 }, { "grad_norm": 0.3209477663040161, "learning_rate": 2.103201046574338e-05, "loss": 0.018, "step": 56950 }, { "grad_norm": 0.34489187598228455, "learning_rate": 2.1015166730928022e-05, "loss": 0.0172, "step": 56960 }, { "grad_norm": 0.3124829828739166, "learning_rate": 2.0998327948827912e-05, "loss": 0.0207, "step": 56970 }, { "grad_norm": 0.3547350764274597, "learning_rate": 2.0981494122320363e-05, "loss": 0.0211, "step": 56980 }, { "grad_norm": 0.21715596318244934, "learning_rate": 2.0964665254281822e-05, "loss": 0.0227, "step": 56990 }, { "grad_norm": 0.20795907080173492, "learning_rate": 2.094784134758784e-05, "loss": 0.0181, "step": 57000 }, { "grad_norm": 0.3448745608329773, "learning_rate": 2.0931022405113226e-05, "loss": 0.0306, "step": 57010 }, { "grad_norm": 0.22178156673908234, "learning_rate": 2.091420842973183e-05, "loss": 0.0198, "step": 57020 }, { "grad_norm": 0.24924473464488983, "learning_rate": 2.0897399424316715e-05, "loss": 0.0154, "step": 57030 }, { "grad_norm": 0.3538764715194702, "learning_rate": 2.0880595391740078e-05, "loss": 0.0187, "step": 57040 }, { "grad_norm": 0.19486960768699646, "learning_rate": 2.0863796334873277e-05, "loss": 0.0235, "step": 57050 }, { "grad_norm": 0.1428605616092682, "learning_rate": 2.08470022565868e-05, "loss": 0.0168, "step": 57060 }, { "grad_norm": 0.22694802284240723, "learning_rate": 2.0830213159750317e-05, "loss": 0.0177, "step": 57070 }, { "grad_norm": 0.18454138934612274, "learning_rate": 2.0813429047232596e-05, "loss": 0.0219, "step": 57080 }, { "grad_norm": 0.24777327477931976, "learning_rate": 2.0796649921901594e-05, "loss": 0.0214, "step": 57090 }, { "grad_norm": 0.2795179486274719, "learning_rate": 2.077987578662441e-05, "loss": 0.0166, "step": 57100 }, { "grad_norm": 0.35765641927719116, "learning_rate": 2.0763106644267277e-05, "loss": 0.0227, "step": 57110 }, { "grad_norm": 0.20836785435676575, "learning_rate": 2.0746342497695607e-05, "loss": 0.0147, "step": 57120 }, { "grad_norm": 0.22362400591373444, "learning_rate": 2.0729583349773886e-05, "loss": 0.0156, "step": 57130 }, { "grad_norm": 0.3275515139102936, "learning_rate": 2.0712829203365853e-05, "loss": 0.0162, "step": 57140 }, { "grad_norm": 0.17795948684215546, "learning_rate": 2.0696080061334267e-05, "loss": 0.0164, "step": 57150 }, { "grad_norm": 0.2518918812274933, "learning_rate": 2.067933592654117e-05, "loss": 0.0166, "step": 57160 }, { "grad_norm": 0.2993772625923157, "learning_rate": 2.066259680184763e-05, "loss": 0.0187, "step": 57170 }, { "grad_norm": 0.2675286531448364, "learning_rate": 2.0645862690113908e-05, "loss": 0.018, "step": 57180 }, { "grad_norm": 0.3302907645702362, "learning_rate": 2.0629133594199436e-05, "loss": 0.0237, "step": 57190 }, { "grad_norm": 0.21227267384529114, "learning_rate": 2.0612409516962704e-05, "loss": 0.0188, "step": 57200 }, { "grad_norm": 0.25832417607307434, "learning_rate": 2.0595690461261467e-05, "loss": 0.0195, "step": 57210 }, { "grad_norm": 0.28368932008743286, "learning_rate": 2.0578976429952503e-05, "loss": 0.0144, "step": 57220 }, { "grad_norm": 0.2294219732284546, "learning_rate": 2.0562267425891802e-05, "loss": 0.018, "step": 57230 }, { "grad_norm": 0.34080037474632263, "learning_rate": 2.0545563451934467e-05, "loss": 0.0144, "step": 57240 }, { "grad_norm": 0.2569804787635803, "learning_rate": 2.0528864510934764e-05, "loss": 0.0196, "step": 57250 }, { "grad_norm": 0.1791010946035385, "learning_rate": 2.0512170605746096e-05, "loss": 0.0269, "step": 57260 }, { "grad_norm": 0.38821181654930115, "learning_rate": 2.0495481739220963e-05, "loss": 0.019, "step": 57270 }, { "grad_norm": 0.14855919778347015, "learning_rate": 2.0478797914211045e-05, "loss": 0.0186, "step": 57280 }, { "grad_norm": 0.2969185709953308, "learning_rate": 2.046211913356716e-05, "loss": 0.0186, "step": 57290 }, { "grad_norm": 0.21397468447685242, "learning_rate": 2.0445445400139247e-05, "loss": 0.0226, "step": 57300 }, { "grad_norm": 0.42229917645454407, "learning_rate": 2.0428776716776405e-05, "loss": 0.0187, "step": 57310 }, { "grad_norm": 0.4463784396648407, "learning_rate": 2.0412113086326856e-05, "loss": 0.0221, "step": 57320 }, { "grad_norm": 0.23178048431873322, "learning_rate": 2.0395454511637918e-05, "loss": 0.016, "step": 57330 }, { "grad_norm": 0.3338053822517395, "learning_rate": 2.037880099555616e-05, "loss": 0.0186, "step": 57340 }, { "grad_norm": 0.2374492883682251, "learning_rate": 2.0362152540927144e-05, "loss": 0.0163, "step": 57350 }, { "grad_norm": 0.2535077631473541, "learning_rate": 2.0345509150595666e-05, "loss": 0.0298, "step": 57360 }, { "grad_norm": 0.19866299629211426, "learning_rate": 2.0328870827405617e-05, "loss": 0.0231, "step": 57370 }, { "grad_norm": 0.2836088240146637, "learning_rate": 2.0312237574200043e-05, "loss": 0.0156, "step": 57380 }, { "grad_norm": 0.36830762028694153, "learning_rate": 2.029560939382112e-05, "loss": 0.0213, "step": 57390 }, { "grad_norm": 0.24335689842700958, "learning_rate": 2.0278986289110097e-05, "loss": 0.0184, "step": 57400 }, { "grad_norm": 0.25242775678634644, "learning_rate": 2.0262368262907484e-05, "loss": 0.0185, "step": 57410 }, { "grad_norm": 0.37061983346939087, "learning_rate": 2.024575531805279e-05, "loss": 0.0218, "step": 57420 }, { "grad_norm": 0.3387817442417145, "learning_rate": 2.0229147457384735e-05, "loss": 0.0148, "step": 57430 }, { "grad_norm": 0.3064365088939667, "learning_rate": 2.0212544683741157e-05, "loss": 0.0294, "step": 57440 }, { "grad_norm": 0.32531285285949707, "learning_rate": 2.0195946999958976e-05, "loss": 0.0257, "step": 57450 }, { "grad_norm": 0.23612578213214874, "learning_rate": 2.017935440887434e-05, "loss": 0.0199, "step": 57460 }, { "grad_norm": 0.35971885919570923, "learning_rate": 2.0162766913322423e-05, "loss": 0.0174, "step": 57470 }, { "grad_norm": 0.27418386936187744, "learning_rate": 2.0146184516137588e-05, "loss": 0.0191, "step": 57480 }, { "grad_norm": 0.12041937559843063, "learning_rate": 2.012960722015332e-05, "loss": 0.0154, "step": 57490 }, { "grad_norm": 0.24134992063045502, "learning_rate": 2.0113035028202214e-05, "loss": 0.0179, "step": 57500 }, { "grad_norm": 0.22862626612186432, "learning_rate": 2.009646794311602e-05, "loss": 0.0256, "step": 57510 }, { "grad_norm": 0.1836833953857422, "learning_rate": 2.007990596772559e-05, "loss": 0.0205, "step": 57520 }, { "grad_norm": 0.30530014634132385, "learning_rate": 2.0063349104860923e-05, "loss": 0.0195, "step": 57530 }, { "grad_norm": 0.30763694643974304, "learning_rate": 2.0046797357351116e-05, "loss": 0.0276, "step": 57540 }, { "grad_norm": 0.24697145819664001, "learning_rate": 2.0030250728024412e-05, "loss": 0.0157, "step": 57550 }, { "grad_norm": 0.19956786930561066, "learning_rate": 2.001370921970819e-05, "loss": 0.0132, "step": 57560 }, { "grad_norm": 0.16162851452827454, "learning_rate": 1.9997172835228932e-05, "loss": 0.0148, "step": 57570 }, { "grad_norm": 0.3872242271900177, "learning_rate": 1.9980641577412262e-05, "loss": 0.0186, "step": 57580 }, { "grad_norm": 0.3899165689945221, "learning_rate": 1.9964115449082925e-05, "loss": 0.0206, "step": 57590 }, { "grad_norm": 0.23956654965877533, "learning_rate": 1.9947594453064742e-05, "loss": 0.023, "step": 57600 }, { "grad_norm": 0.2380211055278778, "learning_rate": 1.9931078592180774e-05, "loss": 0.0186, "step": 57610 }, { "grad_norm": 0.26265186071395874, "learning_rate": 1.9914567869253065e-05, "loss": 0.0186, "step": 57620 }, { "grad_norm": 0.20741966366767883, "learning_rate": 1.989806228710287e-05, "loss": 0.0269, "step": 57630 }, { "grad_norm": 0.20802626013755798, "learning_rate": 1.9881561848550555e-05, "loss": 0.0149, "step": 57640 }, { "grad_norm": 0.28103846311569214, "learning_rate": 1.9865066556415544e-05, "loss": 0.022, "step": 57650 }, { "grad_norm": 0.24043874442577362, "learning_rate": 1.98485764135165e-05, "loss": 0.0162, "step": 57660 }, { "grad_norm": 0.2751789391040802, "learning_rate": 1.983209142267109e-05, "loss": 0.0162, "step": 57670 }, { "grad_norm": 0.33653688430786133, "learning_rate": 1.9815611586696165e-05, "loss": 0.0181, "step": 57680 }, { "grad_norm": 0.20819486677646637, "learning_rate": 1.9799136908407667e-05, "loss": 0.0169, "step": 57690 }, { "grad_norm": 0.20914693176746368, "learning_rate": 1.9782667390620678e-05, "loss": 0.0146, "step": 57700 }, { "grad_norm": 0.30469074845314026, "learning_rate": 1.976620303614939e-05, "loss": 0.0162, "step": 57710 }, { "grad_norm": 0.2704017460346222, "learning_rate": 1.9749743847807108e-05, "loss": 0.0174, "step": 57720 }, { "grad_norm": 0.34880155324935913, "learning_rate": 1.9733289828406272e-05, "loss": 0.0129, "step": 57730 }, { "grad_norm": 0.537647545337677, "learning_rate": 1.9716840980758382e-05, "loss": 0.0179, "step": 57740 }, { "grad_norm": 0.24801237881183624, "learning_rate": 1.9700397307674134e-05, "loss": 0.0179, "step": 57750 }, { "grad_norm": 0.2018640637397766, "learning_rate": 1.968395881196328e-05, "loss": 0.0151, "step": 57760 }, { "grad_norm": 0.3588026165962219, "learning_rate": 1.966752549643473e-05, "loss": 0.0229, "step": 57770 }, { "grad_norm": 0.21004018187522888, "learning_rate": 1.965109736389647e-05, "loss": 0.0153, "step": 57780 }, { "grad_norm": 0.2617485821247101, "learning_rate": 1.9634674417155645e-05, "loss": 0.0176, "step": 57790 }, { "grad_norm": 0.2333524078130722, "learning_rate": 1.9618256659018434e-05, "loss": 0.0391, "step": 57800 }, { "grad_norm": 0.34096550941467285, "learning_rate": 1.9601844092290257e-05, "loss": 0.0193, "step": 57810 }, { "grad_norm": 0.2948092818260193, "learning_rate": 1.9585436719775512e-05, "loss": 0.0133, "step": 57820 }, { "grad_norm": 0.14786162972450256, "learning_rate": 1.9569034544277793e-05, "loss": 0.0207, "step": 57830 }, { "grad_norm": 0.2733878493309021, "learning_rate": 1.9552637568599798e-05, "loss": 0.0192, "step": 57840 }, { "grad_norm": 0.2871216833591461, "learning_rate": 1.953624579554327e-05, "loss": 0.0268, "step": 57850 }, { "grad_norm": 0.25924915075302124, "learning_rate": 1.951985922790918e-05, "loss": 0.0226, "step": 57860 }, { "grad_norm": 0.28707683086395264, "learning_rate": 1.9503477868497505e-05, "loss": 0.0167, "step": 57870 }, { "grad_norm": 0.29261502623558044, "learning_rate": 1.9487101720107375e-05, "loss": 0.02, "step": 57880 }, { "grad_norm": 0.25055983662605286, "learning_rate": 1.9470730785537032e-05, "loss": 0.0226, "step": 57890 }, { "grad_norm": 0.2801913619041443, "learning_rate": 1.9454365067583823e-05, "loss": 0.0166, "step": 57900 }, { "grad_norm": 0.17945784330368042, "learning_rate": 1.9438004569044215e-05, "loss": 0.0133, "step": 57910 }, { "grad_norm": 0.24313434958457947, "learning_rate": 1.9421649292713724e-05, "loss": 0.0341, "step": 57920 }, { "grad_norm": 0.32647213339805603, "learning_rate": 1.9405299241387076e-05, "loss": 0.0171, "step": 57930 }, { "grad_norm": 0.28674402832984924, "learning_rate": 1.9388954417858007e-05, "loss": 0.0157, "step": 57940 }, { "grad_norm": 0.35097530484199524, "learning_rate": 1.9372614824919417e-05, "loss": 0.0185, "step": 57950 }, { "grad_norm": 0.33530279994010925, "learning_rate": 1.9356280465363284e-05, "loss": 0.0187, "step": 57960 }, { "grad_norm": 0.26502135396003723, "learning_rate": 1.9339951341980723e-05, "loss": 0.014, "step": 57970 }, { "grad_norm": 0.23151929676532745, "learning_rate": 1.9323627457561916e-05, "loss": 0.023, "step": 57980 }, { "grad_norm": 0.27681300044059753, "learning_rate": 1.9307308814896198e-05, "loss": 0.016, "step": 57990 }, { "grad_norm": 0.1460207849740982, "learning_rate": 1.9290995416771935e-05, "loss": 0.0161, "step": 58000 }, { "grad_norm": 0.20094072818756104, "learning_rate": 1.9274687265976665e-05, "loss": 0.0174, "step": 58010 }, { "grad_norm": 0.30723509192466736, "learning_rate": 1.9258384365297e-05, "loss": 0.0193, "step": 58020 }, { "grad_norm": 0.20576556026935577, "learning_rate": 1.924208671751866e-05, "loss": 0.0162, "step": 58030 }, { "grad_norm": 0.31328150629997253, "learning_rate": 1.9225794325426492e-05, "loss": 0.0202, "step": 58040 }, { "grad_norm": 0.1606147587299347, "learning_rate": 1.920950719180436e-05, "loss": 0.0144, "step": 58050 }, { "grad_norm": 0.20621635019779205, "learning_rate": 1.919322531943536e-05, "loss": 0.0145, "step": 58060 }, { "grad_norm": 0.5052081942558289, "learning_rate": 1.917694871110157e-05, "loss": 0.0199, "step": 58070 }, { "grad_norm": 0.21278220415115356, "learning_rate": 1.9160677369584234e-05, "loss": 0.0172, "step": 58080 }, { "grad_norm": 0.1969645470380783, "learning_rate": 1.9144411297663694e-05, "loss": 0.019, "step": 58090 }, { "grad_norm": 0.13824152946472168, "learning_rate": 1.9128150498119328e-05, "loss": 0.0226, "step": 58100 }, { "grad_norm": 0.3050776720046997, "learning_rate": 1.9111894973729726e-05, "loss": 0.0145, "step": 58110 }, { "grad_norm": 0.4967048168182373, "learning_rate": 1.9095644727272454e-05, "loss": 0.0211, "step": 58120 }, { "grad_norm": 0.41279083490371704, "learning_rate": 1.907939976152429e-05, "loss": 0.0248, "step": 58130 }, { "grad_norm": 0.3209875226020813, "learning_rate": 1.906316007926101e-05, "loss": 0.0206, "step": 58140 }, { "grad_norm": 0.26247963309288025, "learning_rate": 1.904692568325755e-05, "loss": 0.0174, "step": 58150 }, { "grad_norm": 0.2373977154493332, "learning_rate": 1.9030696576287925e-05, "loss": 0.0223, "step": 58160 }, { "grad_norm": 0.2173503190279007, "learning_rate": 1.9014472761125242e-05, "loss": 0.0137, "step": 58170 }, { "grad_norm": 0.15241165459156036, "learning_rate": 1.899825424054172e-05, "loss": 0.0186, "step": 58180 }, { "grad_norm": 0.3774428963661194, "learning_rate": 1.898204101730863e-05, "loss": 0.0187, "step": 58190 }, { "grad_norm": 0.24636009335517883, "learning_rate": 1.8965833094196394e-05, "loss": 0.0338, "step": 58200 }, { "grad_norm": 0.24199238419532776, "learning_rate": 1.8949630473974495e-05, "loss": 0.0183, "step": 58210 }, { "grad_norm": 0.21786454319953918, "learning_rate": 1.8933433159411517e-05, "loss": 0.0209, "step": 58220 }, { "grad_norm": 0.2507513165473938, "learning_rate": 1.891724115327514e-05, "loss": 0.0164, "step": 58230 }, { "grad_norm": 0.17033571004867554, "learning_rate": 1.8901054458332152e-05, "loss": 0.0191, "step": 58240 }, { "grad_norm": 0.27291443943977356, "learning_rate": 1.8884873077348364e-05, "loss": 0.0232, "step": 58250 }, { "grad_norm": 0.2591400146484375, "learning_rate": 1.88686970130888e-05, "loss": 0.023, "step": 58260 }, { "grad_norm": 0.25309327244758606, "learning_rate": 1.8852526268317456e-05, "loss": 0.0192, "step": 58270 }, { "grad_norm": 0.2936573922634125, "learning_rate": 1.883636084579749e-05, "loss": 0.0177, "step": 58280 }, { "grad_norm": 0.3754004240036011, "learning_rate": 1.8820200748291145e-05, "loss": 0.0211, "step": 58290 }, { "grad_norm": 0.2065703421831131, "learning_rate": 1.8804045978559686e-05, "loss": 0.0183, "step": 58300 }, { "grad_norm": 0.2774864733219147, "learning_rate": 1.8787896539363594e-05, "loss": 0.0213, "step": 58310 }, { "grad_norm": 0.18642549216747284, "learning_rate": 1.8771752433462298e-05, "loss": 0.0163, "step": 58320 }, { "grad_norm": 0.2405889481306076, "learning_rate": 1.8755613663614447e-05, "loss": 0.0206, "step": 58330 }, { "grad_norm": 0.250028520822525, "learning_rate": 1.873948023257767e-05, "loss": 0.0169, "step": 58340 }, { "grad_norm": 0.3021043539047241, "learning_rate": 1.8723352143108747e-05, "loss": 0.0174, "step": 58350 }, { "grad_norm": 0.2223542481660843, "learning_rate": 1.870722939796352e-05, "loss": 0.0177, "step": 58360 }, { "grad_norm": 0.251890629529953, "learning_rate": 1.869111199989693e-05, "loss": 0.0201, "step": 58370 }, { "grad_norm": 0.27428820729255676, "learning_rate": 1.8674999951663018e-05, "loss": 0.0199, "step": 58380 }, { "grad_norm": 0.23961114883422852, "learning_rate": 1.8658893256014857e-05, "loss": 0.0245, "step": 58390 }, { "grad_norm": 0.21224689483642578, "learning_rate": 1.8642791915704655e-05, "loss": 0.0182, "step": 58400 }, { "grad_norm": 0.22579853236675262, "learning_rate": 1.8626695933483697e-05, "loss": 0.0162, "step": 58410 }, { "grad_norm": 0.24047070741653442, "learning_rate": 1.8610605312102335e-05, "loss": 0.0205, "step": 58420 }, { "grad_norm": 0.3889438807964325, "learning_rate": 1.8594520054310032e-05, "loss": 0.0242, "step": 58430 }, { "grad_norm": 0.21201378107070923, "learning_rate": 1.8578440162855326e-05, "loss": 0.0196, "step": 58440 }, { "grad_norm": 0.16036953032016754, "learning_rate": 1.8562365640485774e-05, "loss": 0.021, "step": 58450 }, { "grad_norm": 0.17436476051807404, "learning_rate": 1.854629648994815e-05, "loss": 0.0153, "step": 58460 }, { "grad_norm": 0.22924305498600006, "learning_rate": 1.8530232713988183e-05, "loss": 0.0222, "step": 58470 }, { "grad_norm": 0.38382700085639954, "learning_rate": 1.851417431535074e-05, "loss": 0.0197, "step": 58480 }, { "grad_norm": 0.4547300636768341, "learning_rate": 1.8498121296779785e-05, "loss": 0.0344, "step": 58490 }, { "grad_norm": 0.372056782245636, "learning_rate": 1.848207366101829e-05, "loss": 0.0232, "step": 58500 }, { "grad_norm": 0.14351248741149902, "learning_rate": 1.8466031410808422e-05, "loss": 0.0147, "step": 58510 }, { "grad_norm": 0.25873175263404846, "learning_rate": 1.844999454889129e-05, "loss": 0.0233, "step": 58520 }, { "grad_norm": 0.2668687105178833, "learning_rate": 1.8433963078007237e-05, "loss": 0.0213, "step": 58530 }, { "grad_norm": 0.25823986530303955, "learning_rate": 1.841793700089554e-05, "loss": 0.0207, "step": 58540 }, { "grad_norm": 0.4095439910888672, "learning_rate": 1.8401916320294644e-05, "loss": 0.0177, "step": 58550 }, { "grad_norm": 0.2665375769138336, "learning_rate": 1.838590103894205e-05, "loss": 0.0171, "step": 58560 }, { "grad_norm": 0.1898697465658188, "learning_rate": 1.8369891159574292e-05, "loss": 0.0151, "step": 58570 }, { "grad_norm": 0.253150075674057, "learning_rate": 1.835388668492708e-05, "loss": 0.0201, "step": 58580 }, { "grad_norm": 0.14938050508499146, "learning_rate": 1.8337887617735095e-05, "loss": 0.015, "step": 58590 }, { "grad_norm": 0.14845746755599976, "learning_rate": 1.8321893960732157e-05, "loss": 0.0136, "step": 58600 }, { "grad_norm": 0.3029593229293823, "learning_rate": 1.8305905716651138e-05, "loss": 0.0184, "step": 58610 }, { "grad_norm": 0.18718399107456207, "learning_rate": 1.8289922888224e-05, "loss": 0.0177, "step": 58620 }, { "grad_norm": 0.20048615336418152, "learning_rate": 1.8273945478181765e-05, "loss": 0.0187, "step": 58630 }, { "grad_norm": 0.40446022152900696, "learning_rate": 1.8257973489254558e-05, "loss": 0.0153, "step": 58640 }, { "grad_norm": 0.37069639563560486, "learning_rate": 1.824200692417152e-05, "loss": 0.0157, "step": 58650 }, { "grad_norm": 0.3406635820865631, "learning_rate": 1.8226045785660912e-05, "loss": 0.0161, "step": 58660 }, { "grad_norm": 0.327976256608963, "learning_rate": 1.821009007645006e-05, "loss": 0.0181, "step": 58670 }, { "grad_norm": 0.16866163909435272, "learning_rate": 1.8194139799265357e-05, "loss": 0.0163, "step": 58680 }, { "grad_norm": 0.2150118052959442, "learning_rate": 1.8178194956832295e-05, "loss": 0.0197, "step": 58690 }, { "grad_norm": 0.1886349469423294, "learning_rate": 1.8162255551875346e-05, "loss": 0.0154, "step": 58700 }, { "grad_norm": 0.2863081395626068, "learning_rate": 1.8146321587118194e-05, "loss": 0.0223, "step": 58710 }, { "grad_norm": 0.31338101625442505, "learning_rate": 1.8130393065283448e-05, "loss": 0.0171, "step": 58720 }, { "grad_norm": 0.19182494282722473, "learning_rate": 1.8114469989092925e-05, "loss": 0.0168, "step": 58730 }, { "grad_norm": 0.26629090309143066, "learning_rate": 1.8098552361267397e-05, "loss": 0.0158, "step": 58740 }, { "grad_norm": 0.28007715940475464, "learning_rate": 1.8082640184526763e-05, "loss": 0.014, "step": 58750 }, { "grad_norm": 0.23071499168872833, "learning_rate": 1.806673346158999e-05, "loss": 0.0158, "step": 58760 }, { "grad_norm": 0.21641011536121368, "learning_rate": 1.8050832195175067e-05, "loss": 0.0167, "step": 58770 }, { "grad_norm": 0.20475652813911438, "learning_rate": 1.8034936387999136e-05, "loss": 0.0225, "step": 58780 }, { "grad_norm": 0.28033098578453064, "learning_rate": 1.8019046042778315e-05, "loss": 0.0158, "step": 58790 }, { "grad_norm": 0.3295331597328186, "learning_rate": 1.800316116222785e-05, "loss": 0.017, "step": 58800 }, { "grad_norm": 0.14760221540927887, "learning_rate": 1.7987281749062018e-05, "loss": 0.0355, "step": 58810 }, { "grad_norm": 0.10498999804258347, "learning_rate": 1.7971407805994195e-05, "loss": 0.0265, "step": 58820 }, { "grad_norm": 0.27879008650779724, "learning_rate": 1.7955539335736787e-05, "loss": 0.0191, "step": 58830 }, { "grad_norm": 0.27414199709892273, "learning_rate": 1.7939676341001304e-05, "loss": 0.0206, "step": 58840 }, { "grad_norm": 0.17241288721561432, "learning_rate": 1.7923818824498275e-05, "loss": 0.0153, "step": 58850 }, { "grad_norm": 0.43647676706314087, "learning_rate": 1.7907966788937315e-05, "loss": 0.0222, "step": 58860 }, { "grad_norm": 0.2332063764333725, "learning_rate": 1.7892120237027116e-05, "loss": 0.0238, "step": 58870 }, { "grad_norm": 0.2481536865234375, "learning_rate": 1.7876279171475413e-05, "loss": 0.0226, "step": 58880 }, { "grad_norm": 0.4716659486293793, "learning_rate": 1.7860443594989028e-05, "loss": 0.0158, "step": 58890 }, { "grad_norm": 0.2520368695259094, "learning_rate": 1.784461351027379e-05, "loss": 0.0149, "step": 58900 }, { "grad_norm": 0.28763264417648315, "learning_rate": 1.7828788920034677e-05, "loss": 0.022, "step": 58910 }, { "grad_norm": 0.36178237199783325, "learning_rate": 1.7812969826975623e-05, "loss": 0.0178, "step": 58920 }, { "grad_norm": 0.2798948585987091, "learning_rate": 1.7797156233799738e-05, "loss": 0.0196, "step": 58930 }, { "grad_norm": 0.26315003633499146, "learning_rate": 1.7781348143209094e-05, "loss": 0.0181, "step": 58940 }, { "grad_norm": 0.21691767871379852, "learning_rate": 1.7765545557904862e-05, "loss": 0.0293, "step": 58950 }, { "grad_norm": 0.26042982935905457, "learning_rate": 1.7749748480587302e-05, "loss": 0.0174, "step": 58960 }, { "grad_norm": 0.16169236600399017, "learning_rate": 1.773395691395564e-05, "loss": 0.0172, "step": 58970 }, { "grad_norm": 0.20678608119487762, "learning_rate": 1.7718170860708305e-05, "loss": 0.0146, "step": 58980 }, { "grad_norm": 0.251596599817276, "learning_rate": 1.770239032354264e-05, "loss": 0.0151, "step": 58990 }, { "grad_norm": 0.25324130058288574, "learning_rate": 1.7686615305155124e-05, "loss": 0.0285, "step": 59000 }, { "grad_norm": 0.2564810514450073, "learning_rate": 1.767084580824128e-05, "loss": 0.0133, "step": 59010 }, { "grad_norm": 0.43911051750183105, "learning_rate": 1.765508183549569e-05, "loss": 0.0209, "step": 59020 }, { "grad_norm": 0.30942195653915405, "learning_rate": 1.763932338961199e-05, "loss": 0.0165, "step": 59030 }, { "grad_norm": 0.4106414318084717, "learning_rate": 1.762357047328284e-05, "loss": 0.0177, "step": 59040 }, { "grad_norm": 0.326922208070755, "learning_rate": 1.7607823089199997e-05, "loss": 0.0217, "step": 59050 }, { "grad_norm": 0.2562219500541687, "learning_rate": 1.7592081240054265e-05, "loss": 0.0205, "step": 59060 }, { "grad_norm": 0.27762913703918457, "learning_rate": 1.7576344928535488e-05, "loss": 0.0187, "step": 59070 }, { "grad_norm": 0.3270776867866516, "learning_rate": 1.7560614157332573e-05, "loss": 0.0192, "step": 59080 }, { "grad_norm": 0.31620270013809204, "learning_rate": 1.7544888929133495e-05, "loss": 0.0194, "step": 59090 }, { "grad_norm": 0.1852564662694931, "learning_rate": 1.752916924662522e-05, "loss": 0.0187, "step": 59100 }, { "grad_norm": 0.27949368953704834, "learning_rate": 1.751345511249387e-05, "loss": 0.0153, "step": 59110 }, { "grad_norm": 0.1733822226524353, "learning_rate": 1.749774652942452e-05, "loss": 0.016, "step": 59120 }, { "grad_norm": 0.3276377022266388, "learning_rate": 1.748204350010135e-05, "loss": 0.0231, "step": 59130 }, { "grad_norm": 0.4851854741573334, "learning_rate": 1.7466346027207574e-05, "loss": 0.0166, "step": 59140 }, { "grad_norm": 0.23678573966026306, "learning_rate": 1.745065411342547e-05, "loss": 0.017, "step": 59150 }, { "grad_norm": 0.24577385187149048, "learning_rate": 1.7434967761436366e-05, "loss": 0.0227, "step": 59160 }, { "grad_norm": 0.4101892411708832, "learning_rate": 1.741928697392058e-05, "loss": 0.0182, "step": 59170 }, { "grad_norm": 0.2287883758544922, "learning_rate": 1.7403611753557597e-05, "loss": 0.0139, "step": 59180 }, { "grad_norm": 0.3616345524787903, "learning_rate": 1.738794210302584e-05, "loss": 0.0114, "step": 59190 }, { "grad_norm": 0.19636601209640503, "learning_rate": 1.7372278025002837e-05, "loss": 0.0126, "step": 59200 }, { "grad_norm": 0.34221702814102173, "learning_rate": 1.7356619522165164e-05, "loss": 0.0183, "step": 59210 }, { "grad_norm": 0.21017782390117645, "learning_rate": 1.7340966597188378e-05, "loss": 0.017, "step": 59220 }, { "grad_norm": 0.2345375269651413, "learning_rate": 1.732531925274722e-05, "loss": 0.0173, "step": 59230 }, { "grad_norm": 0.21996209025382996, "learning_rate": 1.7309677491515318e-05, "loss": 0.0165, "step": 59240 }, { "grad_norm": 0.24560675024986267, "learning_rate": 1.7294041316165455e-05, "loss": 0.0153, "step": 59250 }, { "grad_norm": 0.21954604983329773, "learning_rate": 1.727841072936942e-05, "loss": 0.0222, "step": 59260 }, { "grad_norm": 0.30903100967407227, "learning_rate": 1.7262785733798058e-05, "loss": 0.0161, "step": 59270 }, { "grad_norm": 0.3845652639865875, "learning_rate": 1.7247166332121246e-05, "loss": 0.0188, "step": 59280 }, { "grad_norm": 0.3196919858455658, "learning_rate": 1.7231552527007933e-05, "loss": 0.0166, "step": 59290 }, { "grad_norm": 0.34788209199905396, "learning_rate": 1.721594432112606e-05, "loss": 0.0157, "step": 59300 }, { "grad_norm": 0.336734801530838, "learning_rate": 1.7200341717142655e-05, "loss": 0.0247, "step": 59310 }, { "grad_norm": 0.29445311427116394, "learning_rate": 1.718474471772377e-05, "loss": 0.0191, "step": 59320 }, { "grad_norm": 0.2536519467830658, "learning_rate": 1.7169153325534528e-05, "loss": 0.0132, "step": 59330 }, { "grad_norm": 0.28508299589157104, "learning_rate": 1.7153567543239045e-05, "loss": 0.0148, "step": 59340 }, { "grad_norm": 0.19509123265743256, "learning_rate": 1.7137987373500525e-05, "loss": 0.0173, "step": 59350 }, { "grad_norm": 0.34351229667663574, "learning_rate": 1.7122412818981198e-05, "loss": 0.0208, "step": 59360 }, { "grad_norm": 0.28121501207351685, "learning_rate": 1.7106843882342283e-05, "loss": 0.017, "step": 59370 }, { "grad_norm": 0.21042940020561218, "learning_rate": 1.709128056624415e-05, "loss": 0.0167, "step": 59380 }, { "grad_norm": 0.4747889041900635, "learning_rate": 1.70757228733461e-05, "loss": 0.0145, "step": 59390 }, { "grad_norm": 0.15936222672462463, "learning_rate": 1.706017080630653e-05, "loss": 0.0195, "step": 59400 }, { "grad_norm": 0.28113633394241333, "learning_rate": 1.7044624367782873e-05, "loss": 0.0161, "step": 59410 }, { "grad_norm": 0.2957908809185028, "learning_rate": 1.7029083560431553e-05, "loss": 0.0151, "step": 59420 }, { "grad_norm": 0.3401743173599243, "learning_rate": 1.701354838690813e-05, "loss": 0.0196, "step": 59430 }, { "grad_norm": 0.15264196693897247, "learning_rate": 1.6998018849867087e-05, "loss": 0.0168, "step": 59440 }, { "grad_norm": 0.1410665512084961, "learning_rate": 1.698249495196202e-05, "loss": 0.0153, "step": 59450 }, { "grad_norm": 0.23068778216838837, "learning_rate": 1.6966976695845528e-05, "loss": 0.0177, "step": 59460 }, { "grad_norm": 0.364907443523407, "learning_rate": 1.6951464084169268e-05, "loss": 0.016, "step": 59470 }, { "grad_norm": 0.3248556852340698, "learning_rate": 1.6935957119583916e-05, "loss": 0.0195, "step": 59480 }, { "grad_norm": 0.21409837901592255, "learning_rate": 1.6920455804739205e-05, "loss": 0.0198, "step": 59490 }, { "grad_norm": 0.3283640742301941, "learning_rate": 1.6904960142283855e-05, "loss": 0.0216, "step": 59500 }, { "grad_norm": 0.13301728665828705, "learning_rate": 1.6889470134865666e-05, "loss": 0.0161, "step": 59510 }, { "grad_norm": 0.32786887884140015, "learning_rate": 1.687398578513145e-05, "loss": 0.0207, "step": 59520 }, { "grad_norm": 0.22449560463428497, "learning_rate": 1.6858507095727066e-05, "loss": 0.0189, "step": 59530 }, { "grad_norm": 0.23691144585609436, "learning_rate": 1.6843034069297403e-05, "loss": 0.0166, "step": 59540 }, { "grad_norm": 0.37078970670700073, "learning_rate": 1.682756670848637e-05, "loss": 0.0249, "step": 59550 }, { "grad_norm": 0.3991900384426117, "learning_rate": 1.6812105015936937e-05, "loss": 0.0132, "step": 59560 }, { "grad_norm": 0.49582457542419434, "learning_rate": 1.6796648994291027e-05, "loss": 0.0285, "step": 59570 }, { "grad_norm": 0.25120264291763306, "learning_rate": 1.678119864618973e-05, "loss": 0.0167, "step": 59580 }, { "grad_norm": 0.19682084023952484, "learning_rate": 1.6765753974273023e-05, "loss": 0.0142, "step": 59590 }, { "grad_norm": 0.20638145506381989, "learning_rate": 1.675031498118001e-05, "loss": 0.015, "step": 59600 }, { "grad_norm": 0.3448302447795868, "learning_rate": 1.6734881669548804e-05, "loss": 0.0171, "step": 59610 }, { "grad_norm": 0.250027060508728, "learning_rate": 1.671945404201648e-05, "loss": 0.023, "step": 59620 }, { "grad_norm": 0.25846025347709656, "learning_rate": 1.670403210121927e-05, "loss": 0.0145, "step": 59630 }, { "grad_norm": 0.20246592164039612, "learning_rate": 1.6688615849792312e-05, "loss": 0.0144, "step": 59640 }, { "grad_norm": 0.3314398229122162, "learning_rate": 1.6673205290369832e-05, "loss": 0.0224, "step": 59650 }, { "grad_norm": 0.3630693554878235, "learning_rate": 1.6657800425585086e-05, "loss": 0.0248, "step": 59660 }, { "grad_norm": 0.27916446328163147, "learning_rate": 1.6642401258070327e-05, "loss": 0.0251, "step": 59670 }, { "grad_norm": 0.2485111802816391, "learning_rate": 1.6627007790456884e-05, "loss": 0.0251, "step": 59680 }, { "grad_norm": 0.21095678210258484, "learning_rate": 1.6611620025375037e-05, "loss": 0.0158, "step": 59690 }, { "grad_norm": 0.24110253155231476, "learning_rate": 1.6596237965454154e-05, "loss": 0.0233, "step": 59700 }, { "grad_norm": 0.4540482759475708, "learning_rate": 1.6580861613322607e-05, "loss": 0.0212, "step": 59710 }, { "grad_norm": 0.35262399911880493, "learning_rate": 1.6565490971607796e-05, "loss": 0.0134, "step": 59720 }, { "grad_norm": 0.23317833244800568, "learning_rate": 1.655012604293614e-05, "loss": 0.0214, "step": 59730 }, { "grad_norm": 0.46469441056251526, "learning_rate": 1.653476682993309e-05, "loss": 0.0229, "step": 59740 }, { "grad_norm": 0.2222125232219696, "learning_rate": 1.651941333522311e-05, "loss": 0.0167, "step": 59750 }, { "grad_norm": 0.30345678329467773, "learning_rate": 1.6504065561429715e-05, "loss": 0.0258, "step": 59760 }, { "grad_norm": 0.2962581217288971, "learning_rate": 1.6488723511175385e-05, "loss": 0.031, "step": 59770 }, { "grad_norm": 0.41999202966690063, "learning_rate": 1.6473387187081668e-05, "loss": 0.0183, "step": 59780 }, { "grad_norm": 0.19164800643920898, "learning_rate": 1.6458056591769123e-05, "loss": 0.0193, "step": 59790 }, { "grad_norm": 0.27120912075042725, "learning_rate": 1.6442731727857335e-05, "loss": 0.0171, "step": 59800 }, { "grad_norm": 0.2531563341617584, "learning_rate": 1.642741259796492e-05, "loss": 0.018, "step": 59810 }, { "grad_norm": 0.2780151665210724, "learning_rate": 1.641209920470944e-05, "loss": 0.0173, "step": 59820 }, { "grad_norm": 0.17145086824893951, "learning_rate": 1.639679155070762e-05, "loss": 0.0169, "step": 59830 }, { "grad_norm": 0.296333909034729, "learning_rate": 1.6381489638575048e-05, "loss": 0.0176, "step": 59840 }, { "grad_norm": 0.17173923552036285, "learning_rate": 1.636619347092643e-05, "loss": 0.0159, "step": 59850 }, { "grad_norm": 0.45230406522750854, "learning_rate": 1.6350903050375476e-05, "loss": 0.0174, "step": 59860 }, { "grad_norm": 0.3046189248561859, "learning_rate": 1.6335618379534856e-05, "loss": 0.0171, "step": 59870 }, { "grad_norm": 0.37725457549095154, "learning_rate": 1.6320339461016364e-05, "loss": 0.0146, "step": 59880 }, { "grad_norm": 0.24648751318454742, "learning_rate": 1.6305066297430687e-05, "loss": 0.0191, "step": 59890 }, { "grad_norm": 0.21327154338359833, "learning_rate": 1.6289798891387654e-05, "loss": 0.0209, "step": 59900 }, { "grad_norm": 0.31456539034843445, "learning_rate": 1.6274537245495995e-05, "loss": 0.016, "step": 59910 }, { "grad_norm": 0.1771523356437683, "learning_rate": 1.6259281362363527e-05, "loss": 0.0168, "step": 59920 }, { "grad_norm": 0.2935710549354553, "learning_rate": 1.6244031244597068e-05, "loss": 0.0241, "step": 59930 }, { "grad_norm": 0.3298909068107605, "learning_rate": 1.622878689480244e-05, "loss": 0.0158, "step": 59940 }, { "grad_norm": 0.3721622824668884, "learning_rate": 1.6213548315584498e-05, "loss": 0.0213, "step": 59950 }, { "grad_norm": 0.18904519081115723, "learning_rate": 1.6198315509547074e-05, "loss": 0.0174, "step": 59960 }, { "grad_norm": 0.15963010489940643, "learning_rate": 1.6183088479293056e-05, "loss": 0.0129, "step": 59970 }, { "grad_norm": 0.2578234076499939, "learning_rate": 1.6167867227424316e-05, "loss": 0.0196, "step": 59980 }, { "grad_norm": 0.18065829575061798, "learning_rate": 1.615265175654176e-05, "loss": 0.0206, "step": 59990 }, { "grad_norm": 0.27598828077316284, "learning_rate": 1.613744206924529e-05, "loss": 0.0189, "step": 60000 }, { "grad_norm": 0.3349108099937439, "learning_rate": 1.6122238168133845e-05, "loss": 0.0178, "step": 60010 }, { "grad_norm": 0.14934870600700378, "learning_rate": 1.6107040055805305e-05, "loss": 0.0158, "step": 60020 }, { "grad_norm": 0.4291553795337677, "learning_rate": 1.6091847734856673e-05, "loss": 0.016, "step": 60030 }, { "grad_norm": 0.2221851944923401, "learning_rate": 1.6076661207883865e-05, "loss": 0.0164, "step": 60040 }, { "grad_norm": 0.29824575781822205, "learning_rate": 1.6061480477481848e-05, "loss": 0.0155, "step": 60050 }, { "grad_norm": 0.2793993353843689, "learning_rate": 1.604630554624461e-05, "loss": 0.0135, "step": 60060 }, { "grad_norm": 0.30237722396850586, "learning_rate": 1.603113641676509e-05, "loss": 0.0191, "step": 60070 }, { "grad_norm": 0.1302509903907776, "learning_rate": 1.6015973091635338e-05, "loss": 0.0158, "step": 60080 }, { "grad_norm": 0.1558578759431839, "learning_rate": 1.6000815573446288e-05, "loss": 0.0157, "step": 60090 }, { "grad_norm": 0.38628336787223816, "learning_rate": 1.5985663864788002e-05, "loss": 0.0295, "step": 60100 }, { "grad_norm": 0.1585656702518463, "learning_rate": 1.597051796824946e-05, "loss": 0.0249, "step": 60110 }, { "grad_norm": 0.27636146545410156, "learning_rate": 1.5955377886418682e-05, "loss": 0.0146, "step": 60120 }, { "grad_norm": 0.2714899182319641, "learning_rate": 1.5940243621882704e-05, "loss": 0.0159, "step": 60130 }, { "grad_norm": 0.14764228463172913, "learning_rate": 1.5925115177227555e-05, "loss": 0.0177, "step": 60140 }, { "grad_norm": 0.30445751547813416, "learning_rate": 1.5909992555038288e-05, "loss": 0.0177, "step": 60150 }, { "grad_norm": 0.4263870120048523, "learning_rate": 1.5894875757898912e-05, "loss": 0.0216, "step": 60160 }, { "grad_norm": 0.29884621500968933, "learning_rate": 1.58797647883925e-05, "loss": 0.0166, "step": 60170 }, { "grad_norm": 0.31518587470054626, "learning_rate": 1.586465964910109e-05, "loss": 0.0206, "step": 60180 }, { "grad_norm": 0.19776968657970428, "learning_rate": 1.5849560342605734e-05, "loss": 0.0195, "step": 60190 }, { "grad_norm": 0.25607427954673767, "learning_rate": 1.583446687148651e-05, "loss": 0.0168, "step": 60200 }, { "grad_norm": 0.26783084869384766, "learning_rate": 1.581937923832248e-05, "loss": 0.012, "step": 60210 }, { "grad_norm": 0.1301099956035614, "learning_rate": 1.5804297445691663e-05, "loss": 0.0184, "step": 60220 }, { "grad_norm": 0.385628342628479, "learning_rate": 1.578922149617119e-05, "loss": 0.02, "step": 60230 }, { "grad_norm": 0.377873957157135, "learning_rate": 1.5774151392337084e-05, "loss": 0.0161, "step": 60240 }, { "grad_norm": 0.35750702023506165, "learning_rate": 1.575908713676442e-05, "loss": 0.0191, "step": 60250 }, { "grad_norm": 0.264486163854599, "learning_rate": 1.5744028732027293e-05, "loss": 0.0175, "step": 60260 }, { "grad_norm": 0.2610975503921509, "learning_rate": 1.5728976180698723e-05, "loss": 0.0182, "step": 60270 }, { "grad_norm": 0.23922176659107208, "learning_rate": 1.5713929485350836e-05, "loss": 0.0215, "step": 60280 }, { "grad_norm": 0.2746501564979553, "learning_rate": 1.569888864855465e-05, "loss": 0.016, "step": 60290 }, { "grad_norm": 0.2517144978046417, "learning_rate": 1.568385367288028e-05, "loss": 0.0159, "step": 60300 }, { "grad_norm": 0.18001241981983185, "learning_rate": 1.5668824560896755e-05, "loss": 0.0171, "step": 60310 }, { "grad_norm": 0.32425743341445923, "learning_rate": 1.5653801315172156e-05, "loss": 0.0158, "step": 60320 }, { "grad_norm": 0.3869035243988037, "learning_rate": 1.5638783938273554e-05, "loss": 0.0148, "step": 60330 }, { "grad_norm": 0.2598261833190918, "learning_rate": 1.5623772432766966e-05, "loss": 0.0211, "step": 60340 }, { "grad_norm": 0.19443844258785248, "learning_rate": 1.5608766801217507e-05, "loss": 0.017, "step": 60350 }, { "grad_norm": 0.2518526315689087, "learning_rate": 1.5593767046189184e-05, "loss": 0.016, "step": 60360 }, { "grad_norm": 0.28208741545677185, "learning_rate": 1.5578773170245064e-05, "loss": 0.0139, "step": 60370 }, { "grad_norm": 0.30681443214416504, "learning_rate": 1.5563785175947182e-05, "loss": 0.0225, "step": 60380 }, { "grad_norm": 0.1287735402584076, "learning_rate": 1.5548803065856582e-05, "loss": 0.0156, "step": 60390 }, { "grad_norm": 0.280067503452301, "learning_rate": 1.55338268425333e-05, "loss": 0.0184, "step": 60400 }, { "grad_norm": 0.15215545892715454, "learning_rate": 1.5518856508536373e-05, "loss": 0.0171, "step": 60410 }, { "grad_norm": 0.3315253257751465, "learning_rate": 1.5503892066423786e-05, "loss": 0.0155, "step": 60420 }, { "grad_norm": 0.2641172707080841, "learning_rate": 1.548893351875258e-05, "loss": 0.0257, "step": 60430 }, { "grad_norm": 0.29375848174095154, "learning_rate": 1.547398086807875e-05, "loss": 0.0201, "step": 60440 }, { "grad_norm": 0.348317950963974, "learning_rate": 1.5459034116957305e-05, "loss": 0.0271, "step": 60450 }, { "grad_norm": 0.2572824954986572, "learning_rate": 1.544409326794225e-05, "loss": 0.0225, "step": 60460 }, { "grad_norm": 0.19716106355190277, "learning_rate": 1.542915832358651e-05, "loss": 0.0164, "step": 60470 }, { "grad_norm": 0.20780614018440247, "learning_rate": 1.541422928644214e-05, "loss": 0.0157, "step": 60480 }, { "grad_norm": 0.4809577763080597, "learning_rate": 1.5399306159060024e-05, "loss": 0.0258, "step": 60490 }, { "grad_norm": 0.13052615523338318, "learning_rate": 1.5384388943990185e-05, "loss": 0.0124, "step": 60500 }, { "grad_norm": 0.2333332598209381, "learning_rate": 1.5369477643781526e-05, "loss": 0.0154, "step": 60510 }, { "grad_norm": 0.3051370680332184, "learning_rate": 1.5354572260981985e-05, "loss": 0.0176, "step": 60520 }, { "grad_norm": 0.25484931468963623, "learning_rate": 1.533967279813851e-05, "loss": 0.0113, "step": 60530 }, { "grad_norm": 0.29639649391174316, "learning_rate": 1.5324779257796956e-05, "loss": 0.0179, "step": 60540 }, { "grad_norm": 0.21142373979091644, "learning_rate": 1.5309891642502293e-05, "loss": 0.0141, "step": 60550 }, { "grad_norm": 0.2789153456687927, "learning_rate": 1.5295009954798357e-05, "loss": 0.0124, "step": 60560 }, { "grad_norm": 0.20333898067474365, "learning_rate": 1.5280134197228036e-05, "loss": 0.0172, "step": 60570 }, { "grad_norm": 0.12589095532894135, "learning_rate": 1.526526437233319e-05, "loss": 0.0154, "step": 60580 }, { "grad_norm": 0.3151281476020813, "learning_rate": 1.525040048265466e-05, "loss": 0.0162, "step": 60590 }, { "grad_norm": 0.1631184071302414, "learning_rate": 1.5235542530732288e-05, "loss": 0.0155, "step": 60600 }, { "grad_norm": 0.14455169439315796, "learning_rate": 1.5220690519104901e-05, "loss": 0.0161, "step": 60610 }, { "grad_norm": 0.4304125905036926, "learning_rate": 1.5205844450310275e-05, "loss": 0.0151, "step": 60620 }, { "grad_norm": 0.2911994159221649, "learning_rate": 1.519100432688521e-05, "loss": 0.0134, "step": 60630 }, { "grad_norm": 0.23288816213607788, "learning_rate": 1.5176170151365476e-05, "loss": 0.0159, "step": 60640 }, { "grad_norm": 0.22459889948368073, "learning_rate": 1.5161341926285833e-05, "loss": 0.0126, "step": 60650 }, { "grad_norm": 0.16751056909561157, "learning_rate": 1.5146519654180025e-05, "loss": 0.0098, "step": 60660 }, { "grad_norm": 0.37644800543785095, "learning_rate": 1.5131703337580739e-05, "loss": 0.0192, "step": 60670 }, { "grad_norm": 0.20137859880924225, "learning_rate": 1.5116892979019731e-05, "loss": 0.0198, "step": 60680 }, { "grad_norm": 0.24648021161556244, "learning_rate": 1.5102088581027623e-05, "loss": 0.0131, "step": 60690 }, { "grad_norm": 0.24252164363861084, "learning_rate": 1.5087290146134154e-05, "loss": 0.0164, "step": 60700 }, { "grad_norm": 0.2947123050689697, "learning_rate": 1.5072497676867915e-05, "loss": 0.0164, "step": 60710 }, { "grad_norm": 0.2174292504787445, "learning_rate": 1.505771117575655e-05, "loss": 0.015, "step": 60720 }, { "grad_norm": 0.15188314020633698, "learning_rate": 1.5042930645326691e-05, "loss": 0.0141, "step": 60730 }, { "grad_norm": 0.3344663381576538, "learning_rate": 1.5028156088103879e-05, "loss": 0.0181, "step": 60740 }, { "grad_norm": 0.4853801429271698, "learning_rate": 1.5013387506612735e-05, "loss": 0.021, "step": 60750 }, { "grad_norm": 0.21373963356018066, "learning_rate": 1.4998624903376767e-05, "loss": 0.0172, "step": 60760 }, { "grad_norm": 0.3842727243900299, "learning_rate": 1.4983868280918517e-05, "loss": 0.0181, "step": 60770 }, { "grad_norm": 0.31966432929039, "learning_rate": 1.4969117641759478e-05, "loss": 0.0252, "step": 60780 }, { "grad_norm": 0.2913316488265991, "learning_rate": 1.495437298842014e-05, "loss": 0.0183, "step": 60790 }, { "grad_norm": 0.17630991339683533, "learning_rate": 1.4939634323419976e-05, "loss": 0.0178, "step": 60800 }, { "grad_norm": 0.43130427598953247, "learning_rate": 1.4924901649277384e-05, "loss": 0.0188, "step": 60810 }, { "grad_norm": 0.25507161021232605, "learning_rate": 1.4910174968509793e-05, "loss": 0.0146, "step": 60820 }, { "grad_norm": 0.1593427211046219, "learning_rate": 1.4895454283633598e-05, "loss": 0.0131, "step": 60830 }, { "grad_norm": 0.44049182534217834, "learning_rate": 1.488073959716415e-05, "loss": 0.0152, "step": 60840 }, { "grad_norm": 0.19581682980060577, "learning_rate": 1.4866030911615791e-05, "loss": 0.023, "step": 60850 }, { "grad_norm": 0.18176251649856567, "learning_rate": 1.4851328229501849e-05, "loss": 0.0142, "step": 60860 }, { "grad_norm": 0.14332368969917297, "learning_rate": 1.4836631553334562e-05, "loss": 0.0169, "step": 60870 }, { "grad_norm": 0.3814998269081116, "learning_rate": 1.4821940885625251e-05, "loss": 0.0148, "step": 60880 }, { "grad_norm": 0.3778699040412903, "learning_rate": 1.4807256228884109e-05, "loss": 0.0142, "step": 60890 }, { "grad_norm": 0.2388676255941391, "learning_rate": 1.4792577585620353e-05, "loss": 0.0134, "step": 60900 }, { "grad_norm": 0.297717422246933, "learning_rate": 1.4777904958342164e-05, "loss": 0.0213, "step": 60910 }, { "grad_norm": 0.23180153965950012, "learning_rate": 1.4763238349556691e-05, "loss": 0.01, "step": 60920 }, { "grad_norm": 0.27601388096809387, "learning_rate": 1.4748577761770072e-05, "loss": 0.0182, "step": 60930 }, { "grad_norm": 0.2103399932384491, "learning_rate": 1.4733923197487354e-05, "loss": 0.013, "step": 60940 }, { "grad_norm": 0.22279751300811768, "learning_rate": 1.4719274659212662e-05, "loss": 0.0179, "step": 60950 }, { "grad_norm": 0.23929628729820251, "learning_rate": 1.4704632149448983e-05, "loss": 0.0164, "step": 60960 }, { "grad_norm": 0.2424565851688385, "learning_rate": 1.4689995670698343e-05, "loss": 0.0234, "step": 60970 }, { "grad_norm": 0.2441679835319519, "learning_rate": 1.4675365225461728e-05, "loss": 0.0218, "step": 60980 }, { "grad_norm": 0.29168733954429626, "learning_rate": 1.4660740816239032e-05, "loss": 0.0154, "step": 60990 }, { "grad_norm": 0.26385632157325745, "learning_rate": 1.464612244552923e-05, "loss": 0.017, "step": 61000 }, { "grad_norm": 0.4028913378715515, "learning_rate": 1.4631510115830161e-05, "loss": 0.0157, "step": 61010 }, { "grad_norm": 0.22546212375164032, "learning_rate": 1.4616903829638679e-05, "loss": 0.014, "step": 61020 }, { "grad_norm": 0.19967548549175262, "learning_rate": 1.46023035894506e-05, "loss": 0.0161, "step": 61030 }, { "grad_norm": 0.34344205260276794, "learning_rate": 1.4587709397760713e-05, "loss": 0.0183, "step": 61040 }, { "grad_norm": 0.18327829241752625, "learning_rate": 1.4573121257062755e-05, "loss": 0.0173, "step": 61050 }, { "grad_norm": 0.46928519010543823, "learning_rate": 1.4558539169849472e-05, "loss": 0.0264, "step": 61060 }, { "grad_norm": 0.32347986102104187, "learning_rate": 1.4543963138612499e-05, "loss": 0.0151, "step": 61070 }, { "grad_norm": 0.23477548360824585, "learning_rate": 1.4529393165842498e-05, "loss": 0.0184, "step": 61080 }, { "grad_norm": 0.33956560492515564, "learning_rate": 1.4514829254029084e-05, "loss": 0.0159, "step": 61090 }, { "grad_norm": 0.26035842299461365, "learning_rate": 1.450027140566083e-05, "loss": 0.0151, "step": 61100 }, { "grad_norm": 0.445085346698761, "learning_rate": 1.4485719623225268e-05, "loss": 0.0181, "step": 61110 }, { "grad_norm": 0.3003486394882202, "learning_rate": 1.4471173909208912e-05, "loss": 0.0185, "step": 61120 }, { "grad_norm": 0.2340073436498642, "learning_rate": 1.4456634266097236e-05, "loss": 0.0175, "step": 61130 }, { "grad_norm": 0.2423192858695984, "learning_rate": 1.444210069637461e-05, "loss": 0.0254, "step": 61140 }, { "grad_norm": 0.15613007545471191, "learning_rate": 1.4427573202524502e-05, "loss": 0.0136, "step": 61150 }, { "grad_norm": 0.17995749413967133, "learning_rate": 1.4413051787029208e-05, "loss": 0.0208, "step": 61160 }, { "grad_norm": 0.26002612709999084, "learning_rate": 1.439853645237006e-05, "loss": 0.0181, "step": 61170 }, { "grad_norm": 0.22429324686527252, "learning_rate": 1.4384027201027344e-05, "loss": 0.0161, "step": 61180 }, { "grad_norm": 0.18208925426006317, "learning_rate": 1.4369524035480253e-05, "loss": 0.0178, "step": 61190 }, { "grad_norm": 0.23838946223258972, "learning_rate": 1.4355026958207035e-05, "loss": 0.0151, "step": 61200 }, { "grad_norm": 0.1820717602968216, "learning_rate": 1.434053597168481e-05, "loss": 0.018, "step": 61210 }, { "grad_norm": 0.2438187301158905, "learning_rate": 1.432605107838969e-05, "loss": 0.0181, "step": 61220 }, { "grad_norm": 0.3329511880874634, "learning_rate": 1.4311572280796759e-05, "loss": 0.0186, "step": 61230 }, { "grad_norm": 0.3718806505203247, "learning_rate": 1.4297099581380047e-05, "loss": 0.023, "step": 61240 }, { "grad_norm": 0.39513006806373596, "learning_rate": 1.4282632982612538e-05, "loss": 0.0227, "step": 61250 }, { "grad_norm": 0.2692045271396637, "learning_rate": 1.4268172486966197e-05, "loss": 0.0258, "step": 61260 }, { "grad_norm": 0.21256227791309357, "learning_rate": 1.4253718096911894e-05, "loss": 0.0179, "step": 61270 }, { "grad_norm": 0.3444196283817291, "learning_rate": 1.4239269814919504e-05, "loss": 0.0231, "step": 61280 }, { "grad_norm": 0.185222327709198, "learning_rate": 1.4224827643457843e-05, "loss": 0.0251, "step": 61290 }, { "grad_norm": 0.1787158101797104, "learning_rate": 1.4210391584994686e-05, "loss": 0.014, "step": 61300 }, { "grad_norm": 0.2849966287612915, "learning_rate": 1.4195961641996763e-05, "loss": 0.0188, "step": 61310 }, { "grad_norm": 0.2638421654701233, "learning_rate": 1.4181537816929751e-05, "loss": 0.0177, "step": 61320 }, { "grad_norm": 0.3163660168647766, "learning_rate": 1.4167120112258303e-05, "loss": 0.0207, "step": 61330 }, { "grad_norm": 0.27961426973342896, "learning_rate": 1.4152708530445963e-05, "loss": 0.012, "step": 61340 }, { "grad_norm": 0.13743184506893158, "learning_rate": 1.4138303073955344e-05, "loss": 0.0147, "step": 61350 }, { "grad_norm": 0.22790174186229706, "learning_rate": 1.4123903745247897e-05, "loss": 0.0176, "step": 61360 }, { "grad_norm": 0.1519487202167511, "learning_rate": 1.4109510546784082e-05, "loss": 0.0156, "step": 61370 }, { "grad_norm": 0.18779383599758148, "learning_rate": 1.4095123481023325e-05, "loss": 0.02, "step": 61380 }, { "grad_norm": 0.23437540233135223, "learning_rate": 1.4080742550423936e-05, "loss": 0.0143, "step": 61390 }, { "grad_norm": 0.19550231099128723, "learning_rate": 1.4066367757443278e-05, "loss": 0.0148, "step": 61400 }, { "grad_norm": 0.22200129926204681, "learning_rate": 1.4051999104537572e-05, "loss": 0.0143, "step": 61410 }, { "grad_norm": 0.27978745102882385, "learning_rate": 1.4037636594162034e-05, "loss": 0.0187, "step": 61420 }, { "grad_norm": 0.20728753507137299, "learning_rate": 1.402328022877083e-05, "loss": 0.0133, "step": 61430 }, { "grad_norm": 0.24655550718307495, "learning_rate": 1.4008930010817073e-05, "loss": 0.0181, "step": 61440 }, { "grad_norm": 0.28641021251678467, "learning_rate": 1.3994585942752831e-05, "loss": 0.0147, "step": 61450 }, { "grad_norm": 0.23485876619815826, "learning_rate": 1.398024802702909e-05, "loss": 0.0109, "step": 61460 }, { "grad_norm": 0.25002020597457886, "learning_rate": 1.3965916266095819e-05, "loss": 0.0201, "step": 61470 }, { "grad_norm": 0.30944180488586426, "learning_rate": 1.395159066240192e-05, "loss": 0.0162, "step": 61480 }, { "grad_norm": 0.19206607341766357, "learning_rate": 1.3937271218395248e-05, "loss": 0.0169, "step": 61490 }, { "grad_norm": 0.31167349219322205, "learning_rate": 1.3922957936522613e-05, "loss": 0.016, "step": 61500 }, { "grad_norm": 0.2281981110572815, "learning_rate": 1.3908650819229758e-05, "loss": 0.0175, "step": 61510 }, { "grad_norm": 0.2524013817310333, "learning_rate": 1.389434986896137e-05, "loss": 0.0152, "step": 61520 }, { "grad_norm": 0.2511608898639679, "learning_rate": 1.388005508816112e-05, "loss": 0.0153, "step": 61530 }, { "grad_norm": 0.21284109354019165, "learning_rate": 1.3865766479271557e-05, "loss": 0.0188, "step": 61540 }, { "grad_norm": 0.24024763703346252, "learning_rate": 1.3851484044734225e-05, "loss": 0.0254, "step": 61550 }, { "grad_norm": 0.2823110818862915, "learning_rate": 1.3837207786989608e-05, "loss": 0.0261, "step": 61560 }, { "grad_norm": 0.22770191729068756, "learning_rate": 1.382293770847713e-05, "loss": 0.0184, "step": 61570 }, { "grad_norm": 0.2976667582988739, "learning_rate": 1.3808673811635159e-05, "loss": 0.0176, "step": 61580 }, { "grad_norm": 0.4132721722126007, "learning_rate": 1.3794416098900975e-05, "loss": 0.0224, "step": 61590 }, { "grad_norm": 0.11022941768169403, "learning_rate": 1.3780164572710886e-05, "loss": 0.0239, "step": 61600 }, { "grad_norm": 0.2160041481256485, "learning_rate": 1.3765919235500035e-05, "loss": 0.0188, "step": 61610 }, { "grad_norm": 0.2186845988035202, "learning_rate": 1.375168008970259e-05, "loss": 0.0139, "step": 61620 }, { "grad_norm": 0.181938037276268, "learning_rate": 1.3737447137751635e-05, "loss": 0.0121, "step": 61630 }, { "grad_norm": 0.26526767015457153, "learning_rate": 1.372322038207915e-05, "loss": 0.0159, "step": 61640 }, { "grad_norm": 0.3208385705947876, "learning_rate": 1.3708999825116159e-05, "loss": 0.0275, "step": 61650 }, { "grad_norm": 0.28813838958740234, "learning_rate": 1.3694785469292526e-05, "loss": 0.0242, "step": 61660 }, { "grad_norm": 0.35175207257270813, "learning_rate": 1.3680577317037101e-05, "loss": 0.0192, "step": 61670 }, { "grad_norm": 0.3564051389694214, "learning_rate": 1.3666375370777678e-05, "loss": 0.0168, "step": 61680 }, { "grad_norm": 0.16883429884910583, "learning_rate": 1.3652179632940981e-05, "loss": 0.0148, "step": 61690 }, { "grad_norm": 0.231767937541008, "learning_rate": 1.3637990105952664e-05, "loss": 0.0164, "step": 61700 }, { "grad_norm": 0.09019138664007187, "learning_rate": 1.3623806792237337e-05, "loss": 0.0184, "step": 61710 }, { "grad_norm": 0.23339110612869263, "learning_rate": 1.3609629694218551e-05, "loss": 0.0165, "step": 61720 }, { "grad_norm": 0.20442509651184082, "learning_rate": 1.3595458814318762e-05, "loss": 0.0146, "step": 61730 }, { "grad_norm": 0.3090531826019287, "learning_rate": 1.3581294154959389e-05, "loss": 0.0196, "step": 61740 }, { "grad_norm": 0.22815023362636566, "learning_rate": 1.3567135718560792e-05, "loss": 0.0144, "step": 61750 }, { "grad_norm": 0.2588937282562256, "learning_rate": 1.3552983507542261e-05, "loss": 0.0162, "step": 61760 }, { "grad_norm": 0.19506630301475525, "learning_rate": 1.3538837524322023e-05, "loss": 0.0203, "step": 61770 }, { "grad_norm": 0.1760493665933609, "learning_rate": 1.3524697771317251e-05, "loss": 0.0208, "step": 61780 }, { "grad_norm": 0.23155193030834198, "learning_rate": 1.3510564250943997e-05, "loss": 0.0132, "step": 61790 }, { "grad_norm": 0.21141937375068665, "learning_rate": 1.3496436965617353e-05, "loss": 0.0164, "step": 61800 }, { "grad_norm": 0.23455402255058289, "learning_rate": 1.3482315917751243e-05, "loss": 0.0206, "step": 61810 }, { "grad_norm": 0.2220473736524582, "learning_rate": 1.3468201109758582e-05, "loss": 0.0116, "step": 61820 }, { "grad_norm": 0.23678629100322723, "learning_rate": 1.3454092544051222e-05, "loss": 0.0174, "step": 61830 }, { "grad_norm": 0.2599135637283325, "learning_rate": 1.3439990223039878e-05, "loss": 0.0168, "step": 61840 }, { "grad_norm": 0.10518394410610199, "learning_rate": 1.3425894149134321e-05, "loss": 0.0123, "step": 61850 }, { "grad_norm": 0.26867908239364624, "learning_rate": 1.3411804324743132e-05, "loss": 0.0254, "step": 61860 }, { "grad_norm": 0.3056168556213379, "learning_rate": 1.3397720752273896e-05, "loss": 0.0196, "step": 61870 }, { "grad_norm": 0.14354337751865387, "learning_rate": 1.3383643434133108e-05, "loss": 0.0127, "step": 61880 }, { "grad_norm": 0.23286467790603638, "learning_rate": 1.3369572372726196e-05, "loss": 0.0178, "step": 61890 }, { "grad_norm": 0.15847177803516388, "learning_rate": 1.3355507570457525e-05, "loss": 0.0144, "step": 61900 }, { "grad_norm": 0.26570308208465576, "learning_rate": 1.334144902973038e-05, "loss": 0.0158, "step": 61910 }, { "grad_norm": 0.258960098028183, "learning_rate": 1.3327396752946997e-05, "loss": 0.0146, "step": 61920 }, { "grad_norm": 0.24353155493736267, "learning_rate": 1.3313350742508496e-05, "loss": 0.0158, "step": 61930 }, { "grad_norm": 0.24394768476486206, "learning_rate": 1.3299311000814973e-05, "loss": 0.0189, "step": 61940 }, { "grad_norm": 0.10854239016771317, "learning_rate": 1.3285277530265432e-05, "loss": 0.0222, "step": 61950 }, { "grad_norm": 0.3010072410106659, "learning_rate": 1.3271250333257813e-05, "loss": 0.0176, "step": 61960 }, { "grad_norm": 0.1567739099264145, "learning_rate": 1.3257229412188977e-05, "loss": 0.0141, "step": 61970 }, { "grad_norm": 0.26208147406578064, "learning_rate": 1.3243214769454726e-05, "loss": 0.0162, "step": 61980 }, { "grad_norm": 0.24486923217773438, "learning_rate": 1.3229206407449751e-05, "loss": 0.0136, "step": 61990 }, { "grad_norm": 0.32795193791389465, "learning_rate": 1.3215204328567742e-05, "loss": 0.0175, "step": 62000 }, { "grad_norm": 0.2611084282398224, "learning_rate": 1.3201208535201232e-05, "loss": 0.0164, "step": 62010 }, { "grad_norm": 0.1427011489868164, "learning_rate": 1.3187219029741737e-05, "loss": 0.0139, "step": 62020 }, { "grad_norm": 0.34381401538848877, "learning_rate": 1.3173235814579693e-05, "loss": 0.0148, "step": 62030 }, { "grad_norm": 0.3351803421974182, "learning_rate": 1.3159258892104398e-05, "loss": 0.0165, "step": 62040 }, { "grad_norm": 0.28428131341934204, "learning_rate": 1.3145288264704198e-05, "loss": 0.0179, "step": 62050 }, { "grad_norm": 0.17956726253032684, "learning_rate": 1.3131323934766237e-05, "loss": 0.0158, "step": 62060 }, { "grad_norm": 0.15167784690856934, "learning_rate": 1.311736590467666e-05, "loss": 0.0144, "step": 62070 }, { "grad_norm": 0.2028212696313858, "learning_rate": 1.3103414176820505e-05, "loss": 0.0168, "step": 62080 }, { "grad_norm": 0.29822126030921936, "learning_rate": 1.308946875358174e-05, "loss": 0.0235, "step": 62090 }, { "grad_norm": 0.2722066640853882, "learning_rate": 1.3075529637343276e-05, "loss": 0.0247, "step": 62100 }, { "grad_norm": 0.4546952247619629, "learning_rate": 1.3061596830486883e-05, "loss": 0.019, "step": 62110 }, { "grad_norm": 0.2582576274871826, "learning_rate": 1.3047670335393353e-05, "loss": 0.0115, "step": 62120 }, { "grad_norm": 0.255700021982193, "learning_rate": 1.3033750154442298e-05, "loss": 0.0216, "step": 62130 }, { "grad_norm": 0.19174683094024658, "learning_rate": 1.3019836290012316e-05, "loss": 0.0184, "step": 62140 }, { "grad_norm": 0.22830958664417267, "learning_rate": 1.30059287444809e-05, "loss": 0.018, "step": 62150 }, { "grad_norm": 0.2896154224872589, "learning_rate": 1.299202752022447e-05, "loss": 0.0261, "step": 62160 }, { "grad_norm": 0.3596101999282837, "learning_rate": 1.2978132619618371e-05, "loss": 0.0147, "step": 62170 }, { "grad_norm": 0.2606388032436371, "learning_rate": 1.2964244045036866e-05, "loss": 0.0203, "step": 62180 }, { "grad_norm": 0.16422148048877716, "learning_rate": 1.295036179885311e-05, "loss": 0.0114, "step": 62190 }, { "grad_norm": 0.186460480093956, "learning_rate": 1.293648588343922e-05, "loss": 0.0232, "step": 62200 }, { "grad_norm": 0.25633594393730164, "learning_rate": 1.2922616301166196e-05, "loss": 0.0178, "step": 62210 }, { "grad_norm": 0.33669936656951904, "learning_rate": 1.2908753054403976e-05, "loss": 0.0139, "step": 62220 }, { "grad_norm": 0.2866631746292114, "learning_rate": 1.2894896145521429e-05, "loss": 0.0133, "step": 62230 }, { "grad_norm": 0.3010401427745819, "learning_rate": 1.2881045576886275e-05, "loss": 0.0179, "step": 62240 }, { "grad_norm": 0.37922438979148865, "learning_rate": 1.2867201350865254e-05, "loss": 0.017, "step": 62250 }, { "grad_norm": 0.240675687789917, "learning_rate": 1.2853363469823914e-05, "loss": 0.0189, "step": 62260 }, { "grad_norm": 0.2885066270828247, "learning_rate": 1.283953193612682e-05, "loss": 0.0116, "step": 62270 }, { "grad_norm": 0.3617173433303833, "learning_rate": 1.2825706752137372e-05, "loss": 0.0144, "step": 62280 }, { "grad_norm": 0.34836816787719727, "learning_rate": 1.2811887920217896e-05, "loss": 0.0212, "step": 62290 }, { "grad_norm": 0.1969132423400879, "learning_rate": 1.2798075442729707e-05, "loss": 0.0165, "step": 62300 }, { "grad_norm": 0.21232564747333527, "learning_rate": 1.2784269322032922e-05, "loss": 0.0114, "step": 62310 }, { "grad_norm": 0.20493130385875702, "learning_rate": 1.2770469560486686e-05, "loss": 0.0169, "step": 62320 }, { "grad_norm": 0.20067833364009857, "learning_rate": 1.2756676160448956e-05, "loss": 0.0176, "step": 62330 }, { "grad_norm": 0.43283671140670776, "learning_rate": 1.2742889124276663e-05, "loss": 0.017, "step": 62340 }, { "grad_norm": 0.5683343410491943, "learning_rate": 1.2729108454325639e-05, "loss": 0.016, "step": 62350 }, { "grad_norm": 0.3429335057735443, "learning_rate": 1.2715334152950614e-05, "loss": 0.0182, "step": 62360 }, { "grad_norm": 0.28481319546699524, "learning_rate": 1.2701566222505246e-05, "loss": 0.0199, "step": 62370 }, { "grad_norm": 0.32112959027290344, "learning_rate": 1.2687804665342107e-05, "loss": 0.0187, "step": 62380 }, { "grad_norm": 0.2962665855884552, "learning_rate": 1.267404948381265e-05, "loss": 0.0184, "step": 62390 }, { "grad_norm": 0.2534969747066498, "learning_rate": 1.2660300680267267e-05, "loss": 0.0139, "step": 62400 }, { "grad_norm": 0.37596654891967773, "learning_rate": 1.2646558257055257e-05, "loss": 0.0143, "step": 62410 }, { "grad_norm": 0.2530955970287323, "learning_rate": 1.263282221652482e-05, "loss": 0.0144, "step": 62420 }, { "grad_norm": 0.21550582349300385, "learning_rate": 1.2619092561023088e-05, "loss": 0.0202, "step": 62430 }, { "grad_norm": 0.299261212348938, "learning_rate": 1.2605369292896036e-05, "loss": 0.0193, "step": 62440 }, { "grad_norm": 0.23424379527568817, "learning_rate": 1.2591652414488658e-05, "loss": 0.016, "step": 62450 }, { "grad_norm": 0.31938228011131287, "learning_rate": 1.2577941928144732e-05, "loss": 0.016, "step": 62460 }, { "grad_norm": 0.36283236742019653, "learning_rate": 1.256423783620706e-05, "loss": 0.0199, "step": 62470 }, { "grad_norm": 0.2658807933330536, "learning_rate": 1.2550540141017264e-05, "loss": 0.014, "step": 62480 }, { "grad_norm": 0.29725509881973267, "learning_rate": 1.253684884491591e-05, "loss": 0.0141, "step": 62490 }, { "grad_norm": 0.3463105857372284, "learning_rate": 1.2523163950242483e-05, "loss": 0.013, "step": 62500 }, { "grad_norm": 0.23486356437206268, "learning_rate": 1.2509485459335313e-05, "loss": 0.0139, "step": 62510 }, { "grad_norm": 0.13708825409412384, "learning_rate": 1.2495813374531739e-05, "loss": 0.0148, "step": 62520 }, { "grad_norm": 0.29098519682884216, "learning_rate": 1.2482147698167907e-05, "loss": 0.0129, "step": 62530 }, { "grad_norm": 0.41514575481414795, "learning_rate": 1.2468488432578911e-05, "loss": 0.0165, "step": 62540 }, { "grad_norm": 0.2812367379665375, "learning_rate": 1.2454835580098761e-05, "loss": 0.0199, "step": 62550 }, { "grad_norm": 0.20274750888347626, "learning_rate": 1.2441189143060338e-05, "loss": 0.0184, "step": 62560 }, { "grad_norm": 0.1482653170824051, "learning_rate": 1.242754912379548e-05, "loss": 0.0151, "step": 62570 }, { "grad_norm": 0.21426142752170563, "learning_rate": 1.2413915524634844e-05, "loss": 0.0312, "step": 62580 }, { "grad_norm": 0.2798326015472412, "learning_rate": 1.2400288347908073e-05, "loss": 0.0184, "step": 62590 }, { "grad_norm": 0.3217341899871826, "learning_rate": 1.2386667595943663e-05, "loss": 0.0183, "step": 62600 }, { "grad_norm": 0.2801341414451599, "learning_rate": 1.2373053271069035e-05, "loss": 0.0144, "step": 62610 }, { "grad_norm": 0.17745693027973175, "learning_rate": 1.2359445375610501e-05, "loss": 0.0173, "step": 62620 }, { "grad_norm": 0.17136095464229584, "learning_rate": 1.2345843911893301e-05, "loss": 0.0151, "step": 62630 }, { "grad_norm": 0.2171856015920639, "learning_rate": 1.2332248882241498e-05, "loss": 0.0259, "step": 62640 }, { "grad_norm": 0.33266058564186096, "learning_rate": 1.2318660288978178e-05, "loss": 0.0168, "step": 62650 }, { "grad_norm": 0.2650280296802521, "learning_rate": 1.2305078134425213e-05, "loss": 0.0186, "step": 62660 }, { "grad_norm": 0.14821243286132812, "learning_rate": 1.2291502420903434e-05, "loss": 0.0175, "step": 62670 }, { "grad_norm": 0.31375518441200256, "learning_rate": 1.2277933150732567e-05, "loss": 0.0177, "step": 62680 }, { "grad_norm": 0.12116488069295883, "learning_rate": 1.2264370326231216e-05, "loss": 0.0154, "step": 62690 }, { "grad_norm": 0.22166360914707184, "learning_rate": 1.2250813949716927e-05, "loss": 0.0145, "step": 62700 }, { "grad_norm": 0.19570498168468475, "learning_rate": 1.2237264023506063e-05, "loss": 0.0198, "step": 62710 }, { "grad_norm": 0.3482266068458557, "learning_rate": 1.2223720549913987e-05, "loss": 0.016, "step": 62720 }, { "grad_norm": 0.25412559509277344, "learning_rate": 1.221018353125487e-05, "loss": 0.0141, "step": 62730 }, { "grad_norm": 0.20855259895324707, "learning_rate": 1.2196652969841837e-05, "loss": 0.0159, "step": 62740 }, { "grad_norm": 0.24642042815685272, "learning_rate": 1.2183128867986904e-05, "loss": 0.0189, "step": 62750 }, { "grad_norm": 0.2940819263458252, "learning_rate": 1.2169611228000927e-05, "loss": 0.0201, "step": 62760 }, { "grad_norm": 0.22132955491542816, "learning_rate": 1.2156100052193752e-05, "loss": 0.0194, "step": 62770 }, { "grad_norm": 0.25882238149642944, "learning_rate": 1.214259534287403e-05, "loss": 0.0161, "step": 62780 }, { "grad_norm": 0.19869373738765717, "learning_rate": 1.2129097102349363e-05, "loss": 0.0144, "step": 62790 }, { "grad_norm": 0.2044093906879425, "learning_rate": 1.2115605332926227e-05, "loss": 0.0233, "step": 62800 }, { "grad_norm": 0.22752073407173157, "learning_rate": 1.2102120036910003e-05, "loss": 0.0207, "step": 62810 }, { "grad_norm": 0.2842884063720703, "learning_rate": 1.2088641216604956e-05, "loss": 0.0165, "step": 62820 }, { "grad_norm": 0.20624308288097382, "learning_rate": 1.2075168874314264e-05, "loss": 0.0172, "step": 62830 }, { "grad_norm": 0.21279498934745789, "learning_rate": 1.2061703012339942e-05, "loss": 0.014, "step": 62840 }, { "grad_norm": 0.3789997398853302, "learning_rate": 1.204824363298297e-05, "loss": 0.0203, "step": 62850 }, { "grad_norm": 0.20836327970027924, "learning_rate": 1.2034790738543173e-05, "loss": 0.0192, "step": 62860 }, { "grad_norm": 0.19712191820144653, "learning_rate": 1.2021344331319284e-05, "loss": 0.0154, "step": 62870 }, { "grad_norm": 0.25296565890312195, "learning_rate": 1.200790441360894e-05, "loss": 0.0197, "step": 62880 }, { "grad_norm": 0.2707403600215912, "learning_rate": 1.199447098770864e-05, "loss": 0.0153, "step": 62890 }, { "grad_norm": 0.25637370347976685, "learning_rate": 1.198104405591381e-05, "loss": 0.018, "step": 62900 }, { "grad_norm": 0.28356242179870605, "learning_rate": 1.1967623620518697e-05, "loss": 0.0141, "step": 62910 }, { "grad_norm": 0.30871376395225525, "learning_rate": 1.1954209683816554e-05, "loss": 0.0178, "step": 62920 }, { "grad_norm": 0.12045294791460037, "learning_rate": 1.1940802248099402e-05, "loss": 0.0168, "step": 62930 }, { "grad_norm": 0.20791493356227875, "learning_rate": 1.1927401315658232e-05, "loss": 0.0167, "step": 62940 }, { "grad_norm": 0.1793607622385025, "learning_rate": 1.1914006888782898e-05, "loss": 0.0172, "step": 62950 }, { "grad_norm": 0.23074235022068024, "learning_rate": 1.1900618969762106e-05, "loss": 0.0177, "step": 62960 }, { "grad_norm": 0.32175004482269287, "learning_rate": 1.188723756088354e-05, "loss": 0.0156, "step": 62970 }, { "grad_norm": 0.296256422996521, "learning_rate": 1.1873862664433672e-05, "loss": 0.0202, "step": 62980 }, { "grad_norm": 0.3275686800479889, "learning_rate": 1.1860494282697927e-05, "loss": 0.0206, "step": 62990 }, { "grad_norm": 0.3581310510635376, "learning_rate": 1.1847132417960588e-05, "loss": 0.0163, "step": 63000 }, { "grad_norm": 0.2028733193874359, "learning_rate": 1.1833777072504832e-05, "loss": 0.0186, "step": 63010 }, { "grad_norm": 0.2162131518125534, "learning_rate": 1.1820428248612731e-05, "loss": 0.018, "step": 63020 }, { "grad_norm": 0.44784262776374817, "learning_rate": 1.1807085948565245e-05, "loss": 0.0224, "step": 63030 }, { "grad_norm": 0.16965387761592865, "learning_rate": 1.1793750174642172e-05, "loss": 0.0167, "step": 63040 }, { "grad_norm": 0.1333680897951126, "learning_rate": 1.1780420929122254e-05, "loss": 0.016, "step": 63050 }, { "grad_norm": 0.3598059117794037, "learning_rate": 1.176709821428309e-05, "loss": 0.0169, "step": 63060 }, { "grad_norm": 0.28104549646377563, "learning_rate": 1.1753782032401173e-05, "loss": 0.0198, "step": 63070 }, { "grad_norm": 0.3617607355117798, "learning_rate": 1.1740472385751866e-05, "loss": 0.0175, "step": 63080 }, { "grad_norm": 0.14725767076015472, "learning_rate": 1.172716927660943e-05, "loss": 0.0139, "step": 63090 }, { "grad_norm": 0.3116176724433899, "learning_rate": 1.1713872707247015e-05, "loss": 0.0185, "step": 63100 }, { "grad_norm": 0.25603416562080383, "learning_rate": 1.1700582679936595e-05, "loss": 0.0155, "step": 63110 }, { "grad_norm": 0.2244308888912201, "learning_rate": 1.1687299196949136e-05, "loss": 0.0198, "step": 63120 }, { "grad_norm": 0.33118900656700134, "learning_rate": 1.1674022260554374e-05, "loss": 0.0157, "step": 63130 }, { "grad_norm": 0.21586103737354279, "learning_rate": 1.1660751873020987e-05, "loss": 0.0164, "step": 63140 }, { "grad_norm": 0.3165450990200043, "learning_rate": 1.1647488036616538e-05, "loss": 0.0172, "step": 63150 }, { "grad_norm": 0.21362224221229553, "learning_rate": 1.1634230753607417e-05, "loss": 0.0203, "step": 63160 }, { "grad_norm": 0.347455769777298, "learning_rate": 1.1620980026258982e-05, "loss": 0.0178, "step": 63170 }, { "grad_norm": 0.24290424585342407, "learning_rate": 1.1607735856835373e-05, "loss": 0.0136, "step": 63180 }, { "grad_norm": 0.15426838397979736, "learning_rate": 1.1594498247599677e-05, "loss": 0.0163, "step": 63190 }, { "grad_norm": 0.20373404026031494, "learning_rate": 1.158126720081384e-05, "loss": 0.0238, "step": 63200 }, { "grad_norm": 0.3289007842540741, "learning_rate": 1.156804271873868e-05, "loss": 0.0144, "step": 63210 }, { "grad_norm": 0.28194865584373474, "learning_rate": 1.1554824803633924e-05, "loss": 0.0156, "step": 63220 }, { "grad_norm": 0.33623823523521423, "learning_rate": 1.154161345775811e-05, "loss": 0.0154, "step": 63230 }, { "grad_norm": 0.26878196001052856, "learning_rate": 1.1528408683368724e-05, "loss": 0.0184, "step": 63240 }, { "grad_norm": 0.47434481978416443, "learning_rate": 1.1515210482722088e-05, "loss": 0.0244, "step": 63250 }, { "grad_norm": 0.1636837273836136, "learning_rate": 1.150201885807342e-05, "loss": 0.0208, "step": 63260 }, { "grad_norm": 0.16664206981658936, "learning_rate": 1.1488833811676807e-05, "loss": 0.0119, "step": 63270 }, { "grad_norm": 0.18819795548915863, "learning_rate": 1.1475655345785213e-05, "loss": 0.0168, "step": 63280 }, { "grad_norm": 0.1418418139219284, "learning_rate": 1.1462483462650481e-05, "loss": 0.0145, "step": 63290 }, { "grad_norm": 0.34633710980415344, "learning_rate": 1.1449318164523331e-05, "loss": 0.018, "step": 63300 }, { "grad_norm": 0.2653059661388397, "learning_rate": 1.1436159453653334e-05, "loss": 0.0132, "step": 63310 }, { "grad_norm": 0.2006516009569168, "learning_rate": 1.1423007332288955e-05, "loss": 0.022, "step": 63320 }, { "grad_norm": 0.26169538497924805, "learning_rate": 1.1409861802677546e-05, "loss": 0.0168, "step": 63330 }, { "grad_norm": 0.22218692302703857, "learning_rate": 1.1396722867065313e-05, "loss": 0.0148, "step": 63340 }, { "grad_norm": 0.28910312056541443, "learning_rate": 1.1383590527697352e-05, "loss": 0.018, "step": 63350 }, { "grad_norm": 0.11335166543722153, "learning_rate": 1.137046478681758e-05, "loss": 0.0147, "step": 63360 }, { "grad_norm": 0.22211699187755585, "learning_rate": 1.1357345646668888e-05, "loss": 0.0342, "step": 63370 }, { "grad_norm": 0.1934509128332138, "learning_rate": 1.1344233109492924e-05, "loss": 0.0173, "step": 63380 }, { "grad_norm": 0.2230106145143509, "learning_rate": 1.1331127177530292e-05, "loss": 0.0116, "step": 63390 }, { "grad_norm": 0.24580740928649902, "learning_rate": 1.1318027853020441e-05, "loss": 0.0138, "step": 63400 }, { "grad_norm": 0.22891022264957428, "learning_rate": 1.1304935138201645e-05, "loss": 0.0145, "step": 63410 }, { "grad_norm": 0.2888292670249939, "learning_rate": 1.1291849035311153e-05, "loss": 0.0164, "step": 63420 }, { "grad_norm": 0.3049249053001404, "learning_rate": 1.1278769546584972e-05, "loss": 0.0137, "step": 63430 }, { "grad_norm": 0.2135397046804428, "learning_rate": 1.1265696674258052e-05, "loss": 0.0148, "step": 63440 }, { "grad_norm": 0.22337599098682404, "learning_rate": 1.1252630420564186e-05, "loss": 0.0125, "step": 63450 }, { "grad_norm": 0.3386506140232086, "learning_rate": 1.1239570787736036e-05, "loss": 0.0188, "step": 63460 }, { "grad_norm": 0.2518215775489807, "learning_rate": 1.1226517778005135e-05, "loss": 0.0154, "step": 63470 }, { "grad_norm": 0.15010550618171692, "learning_rate": 1.1213471393601893e-05, "loss": 0.0167, "step": 63480 }, { "grad_norm": 0.16425085067749023, "learning_rate": 1.1200431636755587e-05, "loss": 0.0223, "step": 63490 }, { "grad_norm": 0.3485952615737915, "learning_rate": 1.1187398509694336e-05, "loss": 0.0203, "step": 63500 }, { "grad_norm": 0.32299289107322693, "learning_rate": 1.1174372014645146e-05, "loss": 0.0178, "step": 63510 }, { "grad_norm": 0.254686564207077, "learning_rate": 1.1161352153833899e-05, "loss": 0.0217, "step": 63520 }, { "grad_norm": 0.2788718640804291, "learning_rate": 1.1148338929485325e-05, "loss": 0.0137, "step": 63530 }, { "grad_norm": 0.14491993188858032, "learning_rate": 1.113533234382304e-05, "loss": 0.0129, "step": 63540 }, { "grad_norm": 0.25701722502708435, "learning_rate": 1.1122332399069513e-05, "loss": 0.0139, "step": 63550 }, { "grad_norm": 0.34906694293022156, "learning_rate": 1.1109339097446047e-05, "loss": 0.0199, "step": 63560 }, { "grad_norm": 0.3793167471885681, "learning_rate": 1.1096352441172897e-05, "loss": 0.0252, "step": 63570 }, { "grad_norm": 0.3535051643848419, "learning_rate": 1.1083372432469086e-05, "loss": 0.0151, "step": 63580 }, { "grad_norm": 0.3908144235610962, "learning_rate": 1.107039907355255e-05, "loss": 0.0227, "step": 63590 }, { "grad_norm": 0.2785133123397827, "learning_rate": 1.1057432366640103e-05, "loss": 0.0162, "step": 63600 }, { "grad_norm": 0.3697751462459564, "learning_rate": 1.1044472313947352e-05, "loss": 0.0222, "step": 63610 }, { "grad_norm": 0.227377787232399, "learning_rate": 1.1031518917688877e-05, "loss": 0.0167, "step": 63620 }, { "grad_norm": 0.1721784472465515, "learning_rate": 1.1018572180078007e-05, "loss": 0.0228, "step": 63630 }, { "grad_norm": 0.19399651885032654, "learning_rate": 1.1005632103327018e-05, "loss": 0.0136, "step": 63640 }, { "grad_norm": 0.1801277995109558, "learning_rate": 1.0992698689646996e-05, "loss": 0.02, "step": 63650 }, { "grad_norm": 0.19342787563800812, "learning_rate": 1.0979771941247919e-05, "loss": 0.0158, "step": 63660 }, { "grad_norm": 0.24944838881492615, "learning_rate": 1.0966851860338611e-05, "loss": 0.0177, "step": 63670 }, { "grad_norm": 0.15188805758953094, "learning_rate": 1.0953938449126766e-05, "loss": 0.0142, "step": 63680 }, { "grad_norm": 0.3288954496383667, "learning_rate": 1.0941031709818933e-05, "loss": 0.0289, "step": 63690 }, { "grad_norm": 0.4697727859020233, "learning_rate": 1.0928131644620509e-05, "loss": 0.0306, "step": 63700 }, { "grad_norm": 0.3081066310405731, "learning_rate": 1.0915238255735766e-05, "loss": 0.0186, "step": 63710 }, { "grad_norm": 0.15314941108226776, "learning_rate": 1.0902351545367833e-05, "loss": 0.0179, "step": 63720 }, { "grad_norm": 0.2382155954837799, "learning_rate": 1.0889471515718702e-05, "loss": 0.0132, "step": 63730 }, { "grad_norm": 0.18180599808692932, "learning_rate": 1.087659816898921e-05, "loss": 0.01, "step": 63740 }, { "grad_norm": 0.3964841067790985, "learning_rate": 1.0863731507379082e-05, "loss": 0.015, "step": 63750 }, { "grad_norm": 0.3087758719921112, "learning_rate": 1.0850871533086827e-05, "loss": 0.0185, "step": 63760 }, { "grad_norm": 0.30894747376441956, "learning_rate": 1.0838018248309927e-05, "loss": 0.0146, "step": 63770 }, { "grad_norm": 0.26668885350227356, "learning_rate": 1.0825171655244615e-05, "loss": 0.0205, "step": 63780 }, { "grad_norm": 0.40305694937705994, "learning_rate": 1.0812331756086025e-05, "loss": 0.0168, "step": 63790 }, { "grad_norm": 0.26511648297309875, "learning_rate": 1.079949855302817e-05, "loss": 0.0153, "step": 63800 }, { "grad_norm": 0.3494492471218109, "learning_rate": 1.0786672048263852e-05, "loss": 0.0185, "step": 63810 }, { "grad_norm": 0.1956225484609604, "learning_rate": 1.0773852243984817e-05, "loss": 0.0158, "step": 63820 }, { "grad_norm": 0.21822819113731384, "learning_rate": 1.0761039142381586e-05, "loss": 0.0199, "step": 63830 }, { "grad_norm": 0.40107372403144836, "learning_rate": 1.0748232745643577e-05, "loss": 0.0194, "step": 63840 }, { "grad_norm": 0.14731475710868835, "learning_rate": 1.0735433055959055e-05, "loss": 0.0156, "step": 63850 }, { "grad_norm": 0.24956123530864716, "learning_rate": 1.0722640075515133e-05, "loss": 0.0178, "step": 63860 }, { "grad_norm": 0.18357452750205994, "learning_rate": 1.0709853806497795e-05, "loss": 0.0126, "step": 63870 }, { "grad_norm": 0.2995738089084625, "learning_rate": 1.0697074251091831e-05, "loss": 0.0158, "step": 63880 }, { "grad_norm": 0.42291712760925293, "learning_rate": 1.0684301411480962e-05, "loss": 0.0201, "step": 63890 }, { "grad_norm": 0.2640400826931, "learning_rate": 1.067153528984769e-05, "loss": 0.0131, "step": 63900 }, { "grad_norm": 0.19736038148403168, "learning_rate": 1.0658775888373395e-05, "loss": 0.0133, "step": 63910 }, { "grad_norm": 0.29122138023376465, "learning_rate": 1.0646023209238314e-05, "loss": 0.0213, "step": 63920 }, { "grad_norm": 0.24029400944709778, "learning_rate": 1.0633277254621537e-05, "loss": 0.0165, "step": 63930 }, { "grad_norm": 0.27390336990356445, "learning_rate": 1.0620538026700994e-05, "loss": 0.0109, "step": 63940 }, { "grad_norm": 0.24075451493263245, "learning_rate": 1.0607805527653486e-05, "loss": 0.0184, "step": 63950 }, { "grad_norm": 0.38356050848960876, "learning_rate": 1.059507975965462e-05, "loss": 0.0142, "step": 63960 }, { "grad_norm": 0.2301391363143921, "learning_rate": 1.0582360724878898e-05, "loss": 0.0127, "step": 63970 }, { "grad_norm": 0.3273409605026245, "learning_rate": 1.0569648425499651e-05, "loss": 0.0187, "step": 63980 }, { "grad_norm": 0.19588860869407654, "learning_rate": 1.0556942863689063e-05, "loss": 0.0178, "step": 63990 }, { "grad_norm": 0.2526707947254181, "learning_rate": 1.054424404161819e-05, "loss": 0.0151, "step": 64000 }, { "grad_norm": 0.13899777829647064, "learning_rate": 1.053155196145686e-05, "loss": 0.0117, "step": 64010 }, { "grad_norm": 0.2229880839586258, "learning_rate": 1.0518866625373863e-05, "loss": 0.0124, "step": 64020 }, { "grad_norm": 0.09932492673397064, "learning_rate": 1.0506188035536735e-05, "loss": 0.0126, "step": 64030 }, { "grad_norm": 0.24559389054775238, "learning_rate": 1.0493516194111919e-05, "loss": 0.0169, "step": 64040 }, { "grad_norm": 0.1700114905834198, "learning_rate": 1.0480851103264688e-05, "loss": 0.0124, "step": 64050 }, { "grad_norm": 0.18992730975151062, "learning_rate": 1.046819276515913e-05, "loss": 0.0128, "step": 64060 }, { "grad_norm": 0.45201656222343445, "learning_rate": 1.0455541181958256e-05, "loss": 0.0133, "step": 64070 }, { "grad_norm": 0.1584899127483368, "learning_rate": 1.0442896355823822e-05, "loss": 0.0149, "step": 64080 }, { "grad_norm": 0.21977326273918152, "learning_rate": 1.0430258288916539e-05, "loss": 0.0141, "step": 64090 }, { "grad_norm": 0.30298569798469543, "learning_rate": 1.0417626983395868e-05, "loss": 0.0141, "step": 64100 }, { "grad_norm": 0.2201080322265625, "learning_rate": 1.0405002441420165e-05, "loss": 0.0128, "step": 64110 }, { "grad_norm": 0.20137432217597961, "learning_rate": 1.039238466514662e-05, "loss": 0.013, "step": 64120 }, { "grad_norm": 0.3342224657535553, "learning_rate": 1.0379773656731262e-05, "loss": 0.0195, "step": 64130 }, { "grad_norm": 0.2668117582798004, "learning_rate": 1.0367169418328986e-05, "loss": 0.0143, "step": 64140 }, { "grad_norm": 0.2689618468284607, "learning_rate": 1.0354571952093484e-05, "loss": 0.0185, "step": 64150 }, { "grad_norm": 0.17871293425559998, "learning_rate": 1.034198126017733e-05, "loss": 0.0137, "step": 64160 }, { "grad_norm": 0.23796699941158295, "learning_rate": 1.0329397344731928e-05, "loss": 0.0189, "step": 64170 }, { "grad_norm": 0.3633840084075928, "learning_rate": 1.0316820207907524e-05, "loss": 0.0183, "step": 64180 }, { "grad_norm": 0.20137643814086914, "learning_rate": 1.0304249851853215e-05, "loss": 0.0181, "step": 64190 }, { "grad_norm": 0.3337784707546234, "learning_rate": 1.0291686278716933e-05, "loss": 0.0188, "step": 64200 }, { "grad_norm": 0.20256569981575012, "learning_rate": 1.0279129490645418e-05, "loss": 0.0148, "step": 64210 }, { "grad_norm": 0.17614424228668213, "learning_rate": 1.0266579489784328e-05, "loss": 0.0134, "step": 64220 }, { "grad_norm": 0.26289480924606323, "learning_rate": 1.0254036278278084e-05, "loss": 0.0185, "step": 64230 }, { "grad_norm": 0.3685893416404724, "learning_rate": 1.0241499858269982e-05, "loss": 0.0145, "step": 64240 }, { "grad_norm": 0.34515875577926636, "learning_rate": 1.0228970231902169e-05, "loss": 0.0143, "step": 64250 }, { "grad_norm": 0.4530719220638275, "learning_rate": 1.0216447401315582e-05, "loss": 0.0162, "step": 64260 }, { "grad_norm": 0.3340018093585968, "learning_rate": 1.020393136865007e-05, "loss": 0.0133, "step": 64270 }, { "grad_norm": 0.3684247136116028, "learning_rate": 1.0191422136044242e-05, "loss": 0.0243, "step": 64280 }, { "grad_norm": 0.14616544544696808, "learning_rate": 1.017891970563563e-05, "loss": 0.0169, "step": 64290 }, { "grad_norm": 0.2539767622947693, "learning_rate": 1.0166424079560516e-05, "loss": 0.0213, "step": 64300 }, { "grad_norm": 0.19512756168842316, "learning_rate": 1.0153935259954078e-05, "loss": 0.0199, "step": 64310 }, { "grad_norm": 0.3025159239768982, "learning_rate": 1.0141453248950311e-05, "loss": 0.022, "step": 64320 }, { "grad_norm": 0.320188969373703, "learning_rate": 1.0128978048682054e-05, "loss": 0.0188, "step": 64330 }, { "grad_norm": 0.266641765832901, "learning_rate": 1.0116509661280982e-05, "loss": 0.0193, "step": 64340 }, { "grad_norm": 0.29168248176574707, "learning_rate": 1.0104048088877576e-05, "loss": 0.015, "step": 64350 }, { "grad_norm": 0.14158682525157928, "learning_rate": 1.0091593333601201e-05, "loss": 0.0142, "step": 64360 }, { "grad_norm": 0.22124847769737244, "learning_rate": 1.0079145397580031e-05, "loss": 0.0146, "step": 64370 }, { "grad_norm": 0.2951129972934723, "learning_rate": 1.006670428294107e-05, "loss": 0.023, "step": 64380 }, { "grad_norm": 0.31074264645576477, "learning_rate": 1.0054269991810166e-05, "loss": 0.0126, "step": 64390 }, { "grad_norm": 0.2685198485851288, "learning_rate": 1.0041842526312024e-05, "loss": 0.015, "step": 64400 }, { "grad_norm": 0.5270754098892212, "learning_rate": 1.0029421888570101e-05, "loss": 0.0156, "step": 64410 }, { "grad_norm": 0.2026955634355545, "learning_rate": 1.0017008080706813e-05, "loss": 0.022, "step": 64420 }, { "grad_norm": 0.14223681390285492, "learning_rate": 1.0004601104843287e-05, "loss": 0.0145, "step": 64430 }, { "grad_norm": 0.25185534358024597, "learning_rate": 9.992200963099562e-06, "loss": 0.017, "step": 64440 }, { "grad_norm": 0.23506252467632294, "learning_rate": 9.979807657594486e-06, "loss": 0.0181, "step": 64450 }, { "grad_norm": 0.33437591791152954, "learning_rate": 9.967421190445703e-06, "loss": 0.0162, "step": 64460 }, { "grad_norm": 0.2981414496898651, "learning_rate": 9.955041563769769e-06, "loss": 0.0186, "step": 64470 }, { "grad_norm": 0.19662471115589142, "learning_rate": 9.942668779681974e-06, "loss": 0.0224, "step": 64480 }, { "grad_norm": 0.2708006501197815, "learning_rate": 9.930302840296541e-06, "loss": 0.0132, "step": 64490 }, { "grad_norm": 0.16117659211158752, "learning_rate": 9.917943747726426e-06, "loss": 0.0137, "step": 64500 }, { "grad_norm": 0.23614071309566498, "learning_rate": 9.905591504083484e-06, "loss": 0.0171, "step": 64510 }, { "grad_norm": 0.33024170994758606, "learning_rate": 9.893246111478382e-06, "loss": 0.0149, "step": 64520 }, { "grad_norm": 0.40161994099617004, "learning_rate": 9.88090757202057e-06, "loss": 0.0158, "step": 64530 }, { "grad_norm": 0.5226632952690125, "learning_rate": 9.868575887818421e-06, "loss": 0.0204, "step": 64540 }, { "grad_norm": 0.21384654939174652, "learning_rate": 9.856251060979044e-06, "loss": 0.0154, "step": 64550 }, { "grad_norm": 0.2536109983921051, "learning_rate": 9.843933093608426e-06, "loss": 0.021, "step": 64560 }, { "grad_norm": 0.25625401735305786, "learning_rate": 9.831621987811368e-06, "loss": 0.0212, "step": 64570 }, { "grad_norm": 0.16886357963085175, "learning_rate": 9.819317745691509e-06, "loss": 0.016, "step": 64580 }, { "grad_norm": 0.3542744815349579, "learning_rate": 9.8070203693513e-06, "loss": 0.0135, "step": 64590 }, { "grad_norm": 0.24008066952228546, "learning_rate": 9.794729860892048e-06, "loss": 0.0154, "step": 64600 }, { "grad_norm": 0.35113322734832764, "learning_rate": 9.782446222413827e-06, "loss": 0.0244, "step": 64610 }, { "grad_norm": 0.537289023399353, "learning_rate": 9.770169456015598e-06, "loss": 0.0205, "step": 64620 }, { "grad_norm": 0.21254698932170868, "learning_rate": 9.75789956379512e-06, "loss": 0.0183, "step": 64630 }, { "grad_norm": 0.29980140924453735, "learning_rate": 9.74563654784898e-06, "loss": 0.0192, "step": 64640 }, { "grad_norm": 0.18911617994308472, "learning_rate": 9.733380410272596e-06, "loss": 0.0142, "step": 64650 }, { "grad_norm": 0.2754842936992645, "learning_rate": 9.721131153160207e-06, "loss": 0.0149, "step": 64660 }, { "grad_norm": 0.3535687327384949, "learning_rate": 9.708888778604886e-06, "loss": 0.0151, "step": 64670 }, { "grad_norm": 0.21261174976825714, "learning_rate": 9.696653288698488e-06, "loss": 0.0158, "step": 64680 }, { "grad_norm": 0.37038663029670715, "learning_rate": 9.684424685531763e-06, "loss": 0.025, "step": 64690 }, { "grad_norm": 0.12717078626155853, "learning_rate": 9.672202971194216e-06, "loss": 0.0172, "step": 64700 }, { "grad_norm": 0.22036778926849365, "learning_rate": 9.659988147774213e-06, "loss": 0.0166, "step": 64710 }, { "grad_norm": 0.19057197868824005, "learning_rate": 9.647780217358942e-06, "loss": 0.0179, "step": 64720 }, { "grad_norm": 0.266711950302124, "learning_rate": 9.635579182034376e-06, "loss": 0.0147, "step": 64730 }, { "grad_norm": 0.2220999002456665, "learning_rate": 9.623385043885386e-06, "loss": 0.0212, "step": 64740 }, { "grad_norm": 0.29210329055786133, "learning_rate": 9.61119780499557e-06, "loss": 0.0123, "step": 64750 }, { "grad_norm": 0.19874659180641174, "learning_rate": 9.599017467447418e-06, "loss": 0.0131, "step": 64760 }, { "grad_norm": 0.24731765687465668, "learning_rate": 9.586844033322206e-06, "loss": 0.0157, "step": 64770 }, { "grad_norm": 0.3083144426345825, "learning_rate": 9.574677504700052e-06, "loss": 0.018, "step": 64780 }, { "grad_norm": 0.22822433710098267, "learning_rate": 9.562517883659877e-06, "loss": 0.0248, "step": 64790 }, { "grad_norm": 0.23233336210250854, "learning_rate": 9.55036517227944e-06, "loss": 0.0131, "step": 64800 }, { "grad_norm": 0.13052552938461304, "learning_rate": 9.538219372635282e-06, "loss": 0.0132, "step": 64810 }, { "grad_norm": 0.2445419430732727, "learning_rate": 9.526080486802802e-06, "loss": 0.0149, "step": 64820 }, { "grad_norm": 0.29436126351356506, "learning_rate": 9.513948516856203e-06, "loss": 0.0182, "step": 64830 }, { "grad_norm": 0.34348195791244507, "learning_rate": 9.501823464868504e-06, "loss": 0.0146, "step": 64840 }, { "grad_norm": 0.304047167301178, "learning_rate": 9.489705332911547e-06, "loss": 0.0137, "step": 64850 }, { "grad_norm": 0.2685416042804718, "learning_rate": 9.477594123055994e-06, "loss": 0.0169, "step": 64860 }, { "grad_norm": 0.448388010263443, "learning_rate": 9.465489837371321e-06, "loss": 0.0249, "step": 64870 }, { "grad_norm": 0.2567824423313141, "learning_rate": 9.453392477925794e-06, "loss": 0.0149, "step": 64880 }, { "grad_norm": 0.1693904995918274, "learning_rate": 9.441302046786566e-06, "loss": 0.0146, "step": 64890 }, { "grad_norm": 0.21014411747455597, "learning_rate": 9.429218546019519e-06, "loss": 0.0119, "step": 64900 }, { "grad_norm": 0.2613658308982849, "learning_rate": 9.41714197768941e-06, "loss": 0.0185, "step": 64910 }, { "grad_norm": 0.2943083345890045, "learning_rate": 9.405072343859805e-06, "loss": 0.0185, "step": 64920 }, { "grad_norm": 0.2524414360523224, "learning_rate": 9.393009646593043e-06, "loss": 0.0153, "step": 64930 }, { "grad_norm": 0.23585118353366852, "learning_rate": 9.38095388795035e-06, "loss": 0.0172, "step": 64940 }, { "grad_norm": 0.330546498298645, "learning_rate": 9.36890506999169e-06, "loss": 0.0202, "step": 64950 }, { "grad_norm": 0.24544639885425568, "learning_rate": 9.356863194775894e-06, "loss": 0.0163, "step": 64960 }, { "grad_norm": 0.281572550535202, "learning_rate": 9.344828264360583e-06, "loss": 0.017, "step": 64970 }, { "grad_norm": 0.16480819880962372, "learning_rate": 9.332800280802201e-06, "loss": 0.0181, "step": 64980 }, { "grad_norm": 0.15762819349765778, "learning_rate": 9.32077924615602e-06, "loss": 0.0148, "step": 64990 }, { "grad_norm": 0.27615290880203247, "learning_rate": 9.308765162476063e-06, "loss": 0.0141, "step": 65000 }, { "grad_norm": 0.21905812621116638, "learning_rate": 9.296758031815239e-06, "loss": 0.0202, "step": 65010 }, { "grad_norm": 0.320273220539093, "learning_rate": 9.284757856225229e-06, "loss": 0.0162, "step": 65020 }, { "grad_norm": 0.19406193494796753, "learning_rate": 9.272764637756538e-06, "loss": 0.0178, "step": 65030 }, { "grad_norm": 0.27949491143226624, "learning_rate": 9.260778378458479e-06, "loss": 0.0187, "step": 65040 }, { "grad_norm": 0.16911540925502777, "learning_rate": 9.248799080379172e-06, "loss": 0.0168, "step": 65050 }, { "grad_norm": 0.2753351628780365, "learning_rate": 9.236826745565558e-06, "loss": 0.0191, "step": 65060 }, { "grad_norm": 0.15612372756004333, "learning_rate": 9.224861376063388e-06, "loss": 0.0198, "step": 65070 }, { "grad_norm": 0.18449747562408447, "learning_rate": 9.212902973917192e-06, "loss": 0.012, "step": 65080 }, { "grad_norm": 0.25593245029449463, "learning_rate": 9.20095154117035e-06, "loss": 0.0165, "step": 65090 }, { "grad_norm": 0.261046826839447, "learning_rate": 9.189007079865036e-06, "loss": 0.0158, "step": 65100 }, { "grad_norm": 0.18237486481666565, "learning_rate": 9.177069592042226e-06, "loss": 0.0164, "step": 65110 }, { "grad_norm": 0.262572318315506, "learning_rate": 9.165139079741724e-06, "loss": 0.0195, "step": 65120 }, { "grad_norm": 0.4260689616203308, "learning_rate": 9.153215545002098e-06, "loss": 0.0218, "step": 65130 }, { "grad_norm": 0.2607114315032959, "learning_rate": 9.141298989860798e-06, "loss": 0.0244, "step": 65140 }, { "grad_norm": 0.19049261510372162, "learning_rate": 9.129389416353994e-06, "loss": 0.0153, "step": 65150 }, { "grad_norm": 0.17436805367469788, "learning_rate": 9.11748682651673e-06, "loss": 0.0147, "step": 65160 }, { "grad_norm": 0.31467363238334656, "learning_rate": 9.105591222382837e-06, "loss": 0.0199, "step": 65170 }, { "grad_norm": 0.20319464802742004, "learning_rate": 9.093702605984915e-06, "loss": 0.0168, "step": 65180 }, { "grad_norm": 0.3158746659755707, "learning_rate": 9.081820979354455e-06, "loss": 0.019, "step": 65190 }, { "grad_norm": 0.3305072486400604, "learning_rate": 9.069946344521663e-06, "loss": 0.0217, "step": 65200 }, { "grad_norm": 0.18726558983325958, "learning_rate": 9.058078703515598e-06, "loss": 0.0223, "step": 65210 }, { "grad_norm": 0.31469979882240295, "learning_rate": 9.046218058364125e-06, "loss": 0.0213, "step": 65220 }, { "grad_norm": 0.3774934709072113, "learning_rate": 9.034364411093893e-06, "loss": 0.015, "step": 65230 }, { "grad_norm": 0.23482288420200348, "learning_rate": 9.022517763730371e-06, "loss": 0.0191, "step": 65240 }, { "grad_norm": 0.26084068417549133, "learning_rate": 9.010678118297827e-06, "loss": 0.0149, "step": 65250 }, { "grad_norm": 0.38455671072006226, "learning_rate": 8.998845476819345e-06, "loss": 0.0197, "step": 65260 }, { "grad_norm": 0.19439350068569183, "learning_rate": 8.987019841316773e-06, "loss": 0.0187, "step": 65270 }, { "grad_norm": 0.26918476819992065, "learning_rate": 8.975201213810802e-06, "loss": 0.014, "step": 65280 }, { "grad_norm": 0.28773272037506104, "learning_rate": 8.963389596320915e-06, "loss": 0.0171, "step": 65290 }, { "grad_norm": 0.34029993414878845, "learning_rate": 8.951584990865391e-06, "loss": 0.0163, "step": 65300 }, { "grad_norm": 0.24190649390220642, "learning_rate": 8.939787399461319e-06, "loss": 0.0209, "step": 65310 }, { "grad_norm": 0.2975378930568695, "learning_rate": 8.927996824124591e-06, "loss": 0.0233, "step": 65320 }, { "grad_norm": 0.37012341618537903, "learning_rate": 8.916213266869854e-06, "loss": 0.0193, "step": 65330 }, { "grad_norm": 0.24859687685966492, "learning_rate": 8.904436729710658e-06, "loss": 0.0176, "step": 65340 }, { "grad_norm": 0.34307894110679626, "learning_rate": 8.892667214659245e-06, "loss": 0.0123, "step": 65350 }, { "grad_norm": 0.22702427208423615, "learning_rate": 8.880904723726713e-06, "loss": 0.0179, "step": 65360 }, { "grad_norm": 0.5626084804534912, "learning_rate": 8.869149258922971e-06, "loss": 0.0143, "step": 65370 }, { "grad_norm": 0.1860520988702774, "learning_rate": 8.857400822256662e-06, "loss": 0.0148, "step": 65380 }, { "grad_norm": 0.3059089481830597, "learning_rate": 8.845659415735324e-06, "loss": 0.0244, "step": 65390 }, { "grad_norm": 0.22843757271766663, "learning_rate": 8.83392504136521e-06, "loss": 0.0161, "step": 65400 }, { "grad_norm": 0.16447511315345764, "learning_rate": 8.822197701151407e-06, "loss": 0.0181, "step": 65410 }, { "grad_norm": 0.19990725815296173, "learning_rate": 8.810477397097804e-06, "loss": 0.0157, "step": 65420 }, { "grad_norm": 0.15178117156028748, "learning_rate": 8.798764131207077e-06, "loss": 0.0212, "step": 65430 }, { "grad_norm": 0.29737380146980286, "learning_rate": 8.787057905480706e-06, "loss": 0.018, "step": 65440 }, { "grad_norm": 0.34881576895713806, "learning_rate": 8.775358721918958e-06, "loss": 0.0207, "step": 65450 }, { "grad_norm": 0.25759392976760864, "learning_rate": 8.763666582520923e-06, "loss": 0.0165, "step": 65460 }, { "grad_norm": 0.13344697654247284, "learning_rate": 8.751981489284445e-06, "loss": 0.0126, "step": 65470 }, { "grad_norm": 0.27114585041999817, "learning_rate": 8.740303444206188e-06, "loss": 0.02, "step": 65480 }, { "grad_norm": 0.8097192049026489, "learning_rate": 8.72863244928162e-06, "loss": 0.0136, "step": 65490 }, { "grad_norm": 0.31356969475746155, "learning_rate": 8.716968506504991e-06, "loss": 0.021, "step": 65500 }, { "grad_norm": 0.24707281589508057, "learning_rate": 8.70531161786935e-06, "loss": 0.015, "step": 65510 }, { "grad_norm": 0.18940255045890808, "learning_rate": 8.693661785366558e-06, "loss": 0.0224, "step": 65520 }, { "grad_norm": 0.16619090735912323, "learning_rate": 8.682019010987208e-06, "loss": 0.0189, "step": 65530 }, { "grad_norm": 0.2838725447654724, "learning_rate": 8.670383296720786e-06, "loss": 0.0148, "step": 65540 }, { "grad_norm": 0.36885866522789, "learning_rate": 8.658754644555478e-06, "loss": 0.0159, "step": 65550 }, { "grad_norm": 0.28293484449386597, "learning_rate": 8.647133056478313e-06, "loss": 0.0166, "step": 65560 }, { "grad_norm": 0.29799410700798035, "learning_rate": 8.635518534475123e-06, "loss": 0.0154, "step": 65570 }, { "grad_norm": 0.1943131685256958, "learning_rate": 8.623911080530467e-06, "loss": 0.0168, "step": 65580 }, { "grad_norm": 0.1527998447418213, "learning_rate": 8.61231069662779e-06, "loss": 0.0134, "step": 65590 }, { "grad_norm": 0.2890580892562866, "learning_rate": 8.600717384749252e-06, "loss": 0.0193, "step": 65600 }, { "grad_norm": 0.2957732379436493, "learning_rate": 8.58913114687584e-06, "loss": 0.0142, "step": 65610 }, { "grad_norm": 0.31560832262039185, "learning_rate": 8.57755198498732e-06, "loss": 0.0186, "step": 65620 }, { "grad_norm": 0.17360229790210724, "learning_rate": 8.565979901062265e-06, "loss": 0.0197, "step": 65630 }, { "grad_norm": 0.19923143088817596, "learning_rate": 8.554414897078033e-06, "loss": 0.0139, "step": 65640 }, { "grad_norm": 0.16302873194217682, "learning_rate": 8.542856975010727e-06, "loss": 0.0116, "step": 65650 }, { "grad_norm": 0.18174675107002258, "learning_rate": 8.531306136835337e-06, "loss": 0.013, "step": 65660 }, { "grad_norm": 0.21583740413188934, "learning_rate": 8.519762384525548e-06, "loss": 0.016, "step": 65670 }, { "grad_norm": 0.27372345328330994, "learning_rate": 8.508225720053875e-06, "loss": 0.0154, "step": 65680 }, { "grad_norm": 0.2982471287250519, "learning_rate": 8.496696145391625e-06, "loss": 0.0136, "step": 65690 }, { "grad_norm": 0.18070007860660553, "learning_rate": 8.485173662508889e-06, "loss": 0.0113, "step": 65700 }, { "grad_norm": 0.19495084881782532, "learning_rate": 8.473658273374536e-06, "loss": 0.0156, "step": 65710 }, { "grad_norm": 0.270386666059494, "learning_rate": 8.462149979956253e-06, "loss": 0.0119, "step": 65720 }, { "grad_norm": 0.26496633887290955, "learning_rate": 8.450648784220461e-06, "loss": 0.0177, "step": 65730 }, { "grad_norm": 0.3990924656391144, "learning_rate": 8.439154688132417e-06, "loss": 0.0142, "step": 65740 }, { "grad_norm": 0.3107410967350006, "learning_rate": 8.427667693656143e-06, "loss": 0.0157, "step": 65750 }, { "grad_norm": 0.3489760458469391, "learning_rate": 8.416187802754454e-06, "loss": 0.0181, "step": 65760 }, { "grad_norm": 0.22390015423297882, "learning_rate": 8.404715017388965e-06, "loss": 0.0127, "step": 65770 }, { "grad_norm": 0.2601460814476013, "learning_rate": 8.393249339520015e-06, "loss": 0.0191, "step": 65780 }, { "grad_norm": 0.31538793444633484, "learning_rate": 8.381790771106834e-06, "loss": 0.0102, "step": 65790 }, { "grad_norm": 0.2302991896867752, "learning_rate": 8.370339314107339e-06, "loss": 0.0108, "step": 65800 }, { "grad_norm": 0.31638649106025696, "learning_rate": 8.358894970478281e-06, "loss": 0.0163, "step": 65810 }, { "grad_norm": 0.28122052550315857, "learning_rate": 8.347457742175196e-06, "loss": 0.0148, "step": 65820 }, { "grad_norm": 0.3294798731803894, "learning_rate": 8.33602763115236e-06, "loss": 0.0142, "step": 65830 }, { "grad_norm": 0.24010427296161652, "learning_rate": 8.324604639362916e-06, "loss": 0.0274, "step": 65840 }, { "grad_norm": 0.20061436295509338, "learning_rate": 8.31318876875869e-06, "loss": 0.0169, "step": 65850 }, { "grad_norm": 0.24319393932819366, "learning_rate": 8.30178002129039e-06, "loss": 0.0115, "step": 65860 }, { "grad_norm": 0.2554348409175873, "learning_rate": 8.290378398907423e-06, "loss": 0.0122, "step": 65870 }, { "grad_norm": 0.18747101724147797, "learning_rate": 8.278983903558029e-06, "loss": 0.02, "step": 65880 }, { "grad_norm": 0.31304696202278137, "learning_rate": 8.26759653718921e-06, "loss": 0.0166, "step": 65890 }, { "grad_norm": 0.18160799145698547, "learning_rate": 8.25621630174676e-06, "loss": 0.0153, "step": 65900 }, { "grad_norm": 0.19469285011291504, "learning_rate": 8.244843199175261e-06, "loss": 0.0211, "step": 65910 }, { "grad_norm": 0.21608062088489532, "learning_rate": 8.233477231418046e-06, "loss": 0.0115, "step": 65920 }, { "grad_norm": 0.19159117341041565, "learning_rate": 8.22211840041725e-06, "loss": 0.0173, "step": 65930 }, { "grad_norm": 0.28824689984321594, "learning_rate": 8.210766708113792e-06, "loss": 0.0218, "step": 65940 }, { "grad_norm": 0.32487937808036804, "learning_rate": 8.199422156447367e-06, "loss": 0.0188, "step": 65950 }, { "grad_norm": 0.20427757501602173, "learning_rate": 8.188084747356451e-06, "loss": 0.0162, "step": 65960 }, { "grad_norm": 0.1945946216583252, "learning_rate": 8.176754482778299e-06, "loss": 0.0162, "step": 65970 }, { "grad_norm": 0.27089184522628784, "learning_rate": 8.165431364648918e-06, "loss": 0.0176, "step": 65980 }, { "grad_norm": 0.19227132201194763, "learning_rate": 8.154115394903162e-06, "loss": 0.0116, "step": 65990 }, { "grad_norm": 0.23457080125808716, "learning_rate": 8.142806575474582e-06, "loss": 0.0169, "step": 66000 }, { "grad_norm": 0.20858636498451233, "learning_rate": 8.131504908295562e-06, "loss": 0.0166, "step": 66010 }, { "grad_norm": 0.3089175224304199, "learning_rate": 8.120210395297262e-06, "loss": 0.0134, "step": 66020 }, { "grad_norm": 0.198329359292984, "learning_rate": 8.108923038409565e-06, "loss": 0.0158, "step": 66030 }, { "grad_norm": 0.37243568897247314, "learning_rate": 8.097642839561226e-06, "loss": 0.0199, "step": 66040 }, { "grad_norm": 0.2362244576215744, "learning_rate": 8.08636980067966e-06, "loss": 0.0216, "step": 66050 }, { "grad_norm": 0.18138080835342407, "learning_rate": 8.075103923691186e-06, "loss": 0.0134, "step": 66060 }, { "grad_norm": 0.21584083139896393, "learning_rate": 8.063845210520793e-06, "loss": 0.0125, "step": 66070 }, { "grad_norm": 0.16705162823200226, "learning_rate": 8.052593663092295e-06, "loss": 0.0172, "step": 66080 }, { "grad_norm": 0.23894397914409637, "learning_rate": 8.04134928332827e-06, "loss": 0.0185, "step": 66090 }, { "grad_norm": 0.23633074760437012, "learning_rate": 8.030112073150086e-06, "loss": 0.0198, "step": 66100 }, { "grad_norm": 0.242910698056221, "learning_rate": 8.018882034477881e-06, "loss": 0.0115, "step": 66110 }, { "grad_norm": 0.14557678997516632, "learning_rate": 8.007659169230541e-06, "loss": 0.0103, "step": 66120 }, { "grad_norm": 0.32219845056533813, "learning_rate": 7.996443479325755e-06, "loss": 0.0151, "step": 66130 }, { "grad_norm": 0.22260943055152893, "learning_rate": 7.985234966679977e-06, "loss": 0.0203, "step": 66140 }, { "grad_norm": 0.24303293228149414, "learning_rate": 7.974033633208438e-06, "loss": 0.0147, "step": 66150 }, { "grad_norm": 0.150890052318573, "learning_rate": 7.962839480825135e-06, "loss": 0.0134, "step": 66160 }, { "grad_norm": 0.3582894802093506, "learning_rate": 7.951652511442858e-06, "loss": 0.0194, "step": 66170 }, { "grad_norm": 0.396406888961792, "learning_rate": 7.940472726973125e-06, "loss": 0.0195, "step": 66180 }, { "grad_norm": 0.20409110188484192, "learning_rate": 7.929300129326289e-06, "loss": 0.0141, "step": 66190 }, { "grad_norm": 0.3170006275177002, "learning_rate": 7.91813472041142e-06, "loss": 0.0178, "step": 66200 }, { "grad_norm": 0.15748965740203857, "learning_rate": 7.906976502136376e-06, "loss": 0.0137, "step": 66210 }, { "grad_norm": 0.3635634481906891, "learning_rate": 7.89582547640782e-06, "loss": 0.0175, "step": 66220 }, { "grad_norm": 0.19412069022655487, "learning_rate": 7.884681645131115e-06, "loss": 0.014, "step": 66230 }, { "grad_norm": 0.40611422061920166, "learning_rate": 7.87354501021048e-06, "loss": 0.0132, "step": 66240 }, { "grad_norm": 0.30174896121025085, "learning_rate": 7.86241557354882e-06, "loss": 0.0158, "step": 66250 }, { "grad_norm": 0.23063702881336212, "learning_rate": 7.8512933370479e-06, "loss": 0.0147, "step": 66260 }, { "grad_norm": 0.33729997277259827, "learning_rate": 7.840178302608158e-06, "loss": 0.016, "step": 66270 }, { "grad_norm": 0.17801009118556976, "learning_rate": 7.82907047212888e-06, "loss": 0.015, "step": 66280 }, { "grad_norm": 0.45829227566719055, "learning_rate": 7.81796984750809e-06, "loss": 0.0152, "step": 66290 }, { "grad_norm": 0.27233824133872986, "learning_rate": 7.806876430642546e-06, "loss": 0.0183, "step": 66300 }, { "grad_norm": 0.17622044682502747, "learning_rate": 7.795790223427862e-06, "loss": 0.0155, "step": 66310 }, { "grad_norm": 0.28617385029792786, "learning_rate": 7.784711227758324e-06, "loss": 0.0199, "step": 66320 }, { "grad_norm": 0.21296969056129456, "learning_rate": 7.773639445527053e-06, "loss": 0.0212, "step": 66330 }, { "grad_norm": 0.18747097253799438, "learning_rate": 7.762574878625905e-06, "loss": 0.016, "step": 66340 }, { "grad_norm": 0.27791061997413635, "learning_rate": 7.751517528945513e-06, "loss": 0.0125, "step": 66350 }, { "grad_norm": 0.2756810486316681, "learning_rate": 7.740467398375278e-06, "loss": 0.0133, "step": 66360 }, { "grad_norm": 0.2312324196100235, "learning_rate": 7.729424488803378e-06, "loss": 0.0142, "step": 66370 }, { "grad_norm": 0.23749734461307526, "learning_rate": 7.71838880211671e-06, "loss": 0.0229, "step": 66380 }, { "grad_norm": 0.22407129406929016, "learning_rate": 7.707360340200997e-06, "loss": 0.0149, "step": 66390 }, { "grad_norm": 0.2243606597185135, "learning_rate": 7.696339104940697e-06, "loss": 0.017, "step": 66400 }, { "grad_norm": 0.202268585562706, "learning_rate": 7.685325098219038e-06, "loss": 0.0175, "step": 66410 }, { "grad_norm": 0.28239336609840393, "learning_rate": 7.674318321918017e-06, "loss": 0.0198, "step": 66420 }, { "grad_norm": 0.26979294419288635, "learning_rate": 7.663318777918366e-06, "loss": 0.0181, "step": 66430 }, { "grad_norm": 0.24623408913612366, "learning_rate": 7.652326468099647e-06, "loss": 0.0189, "step": 66440 }, { "grad_norm": 0.2108106017112732, "learning_rate": 7.641341394340096e-06, "loss": 0.0132, "step": 66450 }, { "grad_norm": 0.39539140462875366, "learning_rate": 7.630363558516818e-06, "loss": 0.0173, "step": 66460 }, { "grad_norm": 0.35302335023880005, "learning_rate": 7.619392962505578e-06, "loss": 0.0181, "step": 66470 }, { "grad_norm": 0.2469061315059662, "learning_rate": 7.6084296081809725e-06, "loss": 0.021, "step": 66480 }, { "grad_norm": 0.4210306406021118, "learning_rate": 7.597473497416347e-06, "loss": 0.018, "step": 66490 }, { "grad_norm": 0.12377691268920898, "learning_rate": 7.586524632083764e-06, "loss": 0.0136, "step": 66500 }, { "grad_norm": 0.3321004807949066, "learning_rate": 7.5755830140541326e-06, "loss": 0.0127, "step": 66510 }, { "grad_norm": 0.23627205193042755, "learning_rate": 7.5646486451970475e-06, "loss": 0.015, "step": 66520 }, { "grad_norm": 0.14567631483078003, "learning_rate": 7.553721527380897e-06, "loss": 0.0196, "step": 66530 }, { "grad_norm": 0.14763474464416504, "learning_rate": 7.542801662472826e-06, "loss": 0.0137, "step": 66540 }, { "grad_norm": 0.25613799691200256, "learning_rate": 7.5318890523387475e-06, "loss": 0.0176, "step": 66550 }, { "grad_norm": 0.24665500223636627, "learning_rate": 7.52098369884332e-06, "loss": 0.0161, "step": 66560 }, { "grad_norm": 0.21356193721294403, "learning_rate": 7.510085603849992e-06, "loss": 0.0188, "step": 66570 }, { "grad_norm": 0.7444515824317932, "learning_rate": 7.499194769220918e-06, "loss": 0.0143, "step": 66580 }, { "grad_norm": 0.2820734977722168, "learning_rate": 7.48831119681706e-06, "loss": 0.0241, "step": 66590 }, { "grad_norm": 0.22269032895565033, "learning_rate": 7.477434888498119e-06, "loss": 0.0128, "step": 66600 }, { "grad_norm": 0.2242581695318222, "learning_rate": 7.466565846122564e-06, "loss": 0.0202, "step": 66610 }, { "grad_norm": 0.35994723439216614, "learning_rate": 7.455704071547626e-06, "loss": 0.0175, "step": 66620 }, { "grad_norm": 0.28344884514808655, "learning_rate": 7.444849566629247e-06, "loss": 0.0135, "step": 66630 }, { "grad_norm": 0.43921515345573425, "learning_rate": 7.434002333222212e-06, "loss": 0.0117, "step": 66640 }, { "grad_norm": 0.3476817309856415, "learning_rate": 7.423162373179976e-06, "loss": 0.0171, "step": 66650 }, { "grad_norm": 0.2562595307826996, "learning_rate": 7.412329688354835e-06, "loss": 0.0159, "step": 66660 }, { "grad_norm": 0.3471391499042511, "learning_rate": 7.40150428059776e-06, "loss": 0.0133, "step": 66670 }, { "grad_norm": 0.2930501699447632, "learning_rate": 7.3906861517585354e-06, "loss": 0.019, "step": 66680 }, { "grad_norm": 0.2469138652086258, "learning_rate": 7.37987530368569e-06, "loss": 0.0141, "step": 66690 }, { "grad_norm": 0.16232486069202423, "learning_rate": 7.369071738226474e-06, "loss": 0.0153, "step": 66700 }, { "grad_norm": 0.22258996963500977, "learning_rate": 7.358275457226954e-06, "loss": 0.0152, "step": 66710 }, { "grad_norm": 0.27953895926475525, "learning_rate": 7.347486462531899e-06, "loss": 0.0144, "step": 66720 }, { "grad_norm": 0.31432265043258667, "learning_rate": 7.336704755984858e-06, "loss": 0.0126, "step": 66730 }, { "grad_norm": 0.1658588945865631, "learning_rate": 7.325930339428133e-06, "loss": 0.0176, "step": 66740 }, { "grad_norm": 0.2840154767036438, "learning_rate": 7.315163214702769e-06, "loss": 0.0199, "step": 66750 }, { "grad_norm": 0.30272650718688965, "learning_rate": 7.304403383648595e-06, "loss": 0.0182, "step": 66760 }, { "grad_norm": 0.16891013085842133, "learning_rate": 7.293650848104139e-06, "loss": 0.0155, "step": 66770 }, { "grad_norm": 0.14044907689094543, "learning_rate": 7.2829056099067374e-06, "loss": 0.017, "step": 66780 }, { "grad_norm": 0.210571750998497, "learning_rate": 7.2721676708924494e-06, "loss": 0.0179, "step": 66790 }, { "grad_norm": 0.3619746267795563, "learning_rate": 7.261437032896096e-06, "loss": 0.0319, "step": 66800 }, { "grad_norm": 0.3712809383869171, "learning_rate": 7.250713697751255e-06, "loss": 0.0175, "step": 66810 }, { "grad_norm": 0.19688554108142853, "learning_rate": 7.239997667290255e-06, "loss": 0.015, "step": 66820 }, { "grad_norm": 0.12495502084493637, "learning_rate": 7.2292889433441425e-06, "loss": 0.0152, "step": 66830 }, { "grad_norm": 0.501284658908844, "learning_rate": 7.218587527742793e-06, "loss": 0.013, "step": 66840 }, { "grad_norm": 0.21194100379943848, "learning_rate": 7.207893422314749e-06, "loss": 0.0144, "step": 66850 }, { "grad_norm": 0.16713368892669678, "learning_rate": 7.1972066288873545e-06, "loss": 0.0142, "step": 66860 }, { "grad_norm": 0.3086245059967041, "learning_rate": 7.186527149286687e-06, "loss": 0.0336, "step": 66870 }, { "grad_norm": 0.2606171667575836, "learning_rate": 7.175854985337576e-06, "loss": 0.0204, "step": 66880 }, { "grad_norm": 0.32611340284347534, "learning_rate": 7.16519013886362e-06, "loss": 0.0217, "step": 66890 }, { "grad_norm": 0.18398234248161316, "learning_rate": 7.154532611687109e-06, "loss": 0.0172, "step": 66900 }, { "grad_norm": 0.22564351558685303, "learning_rate": 7.143882405629177e-06, "loss": 0.0148, "step": 66910 }, { "grad_norm": 0.20335835218429565, "learning_rate": 7.133239522509605e-06, "loss": 0.0151, "step": 66920 }, { "grad_norm": 0.13447041809558868, "learning_rate": 7.1226039641469956e-06, "loss": 0.0118, "step": 66930 }, { "grad_norm": 0.3024234175682068, "learning_rate": 7.111975732358678e-06, "loss": 0.0163, "step": 66940 }, { "grad_norm": 0.21106040477752686, "learning_rate": 7.101354828960693e-06, "loss": 0.0135, "step": 66950 }, { "grad_norm": 0.3268478810787201, "learning_rate": 7.090741255767918e-06, "loss": 0.0155, "step": 66960 }, { "grad_norm": 0.1669449508190155, "learning_rate": 7.080135014593875e-06, "loss": 0.0154, "step": 66970 }, { "grad_norm": 0.23628953099250793, "learning_rate": 7.069536107250896e-06, "loss": 0.0128, "step": 66980 }, { "grad_norm": 0.252828449010849, "learning_rate": 7.058944535550049e-06, "loss": 0.014, "step": 66990 }, { "grad_norm": 0.25095149874687195, "learning_rate": 7.048360301301138e-06, "loss": 0.0162, "step": 67000 }, { "grad_norm": 0.2587868571281433, "learning_rate": 7.03778340631272e-06, "loss": 0.0154, "step": 67010 }, { "grad_norm": 0.42081815004348755, "learning_rate": 7.0272138523921e-06, "loss": 0.0184, "step": 67020 }, { "grad_norm": 0.24517686665058136, "learning_rate": 7.016651641345334e-06, "loss": 0.0199, "step": 67030 }, { "grad_norm": 0.276355117559433, "learning_rate": 7.0060967749772e-06, "loss": 0.0155, "step": 67040 }, { "grad_norm": 0.45135265588760376, "learning_rate": 6.995549255091238e-06, "loss": 0.0234, "step": 67050 }, { "grad_norm": 0.18991531431674957, "learning_rate": 6.98500908348973e-06, "loss": 0.0184, "step": 67060 }, { "grad_norm": 0.2171792834997177, "learning_rate": 6.974476261973711e-06, "loss": 0.0161, "step": 67070 }, { "grad_norm": 0.12844370305538177, "learning_rate": 6.963950792342949e-06, "loss": 0.0153, "step": 67080 }, { "grad_norm": 0.16695080697536469, "learning_rate": 6.9534326763959715e-06, "loss": 0.0126, "step": 67090 }, { "grad_norm": 0.385847270488739, "learning_rate": 6.942921915929995e-06, "loss": 0.0157, "step": 67100 }, { "grad_norm": 0.189004048705101, "learning_rate": 6.9324185127410735e-06, "loss": 0.0168, "step": 67110 }, { "grad_norm": 0.2629829943180084, "learning_rate": 6.921922468623921e-06, "loss": 0.0184, "step": 67120 }, { "grad_norm": 0.2533285617828369, "learning_rate": 6.911433785372023e-06, "loss": 0.017, "step": 67130 }, { "grad_norm": 0.1554577648639679, "learning_rate": 6.900952464777632e-06, "loss": 0.0195, "step": 67140 }, { "grad_norm": 0.326067179441452, "learning_rate": 6.8904785086316815e-06, "loss": 0.0182, "step": 67150 }, { "grad_norm": 0.38697242736816406, "learning_rate": 6.880011918723927e-06, "loss": 0.0132, "step": 67160 }, { "grad_norm": 0.20433443784713745, "learning_rate": 6.86955269684279e-06, "loss": 0.015, "step": 67170 }, { "grad_norm": 0.3501136898994446, "learning_rate": 6.859100844775473e-06, "loss": 0.024, "step": 67180 }, { "grad_norm": 0.1690545380115509, "learning_rate": 6.84865636430792e-06, "loss": 0.0158, "step": 67190 }, { "grad_norm": 0.29324445128440857, "learning_rate": 6.838219257224804e-06, "loss": 0.0084, "step": 67200 }, { "grad_norm": 0.16575109958648682, "learning_rate": 6.827789525309536e-06, "loss": 0.0151, "step": 67210 }, { "grad_norm": 0.44724148511886597, "learning_rate": 6.81736717034428e-06, "loss": 0.0194, "step": 67220 }, { "grad_norm": 0.3100182116031647, "learning_rate": 6.806952194109933e-06, "loss": 0.0233, "step": 67230 }, { "grad_norm": 0.24377451837062836, "learning_rate": 6.796544598386112e-06, "loss": 0.0133, "step": 67240 }, { "grad_norm": 0.2949589192867279, "learning_rate": 6.786144384951204e-06, "loss": 0.0204, "step": 67250 }, { "grad_norm": 0.21599625051021576, "learning_rate": 6.775751555582322e-06, "loss": 0.015, "step": 67260 }, { "grad_norm": 0.24021746218204498, "learning_rate": 6.76536611205531e-06, "loss": 0.0151, "step": 67270 }, { "grad_norm": 0.2283381223678589, "learning_rate": 6.754988056144762e-06, "loss": 0.0171, "step": 67280 }, { "grad_norm": 0.22276601195335388, "learning_rate": 6.744617389624014e-06, "loss": 0.0144, "step": 67290 }, { "grad_norm": 0.40475231409072876, "learning_rate": 6.734254114265087e-06, "loss": 0.0172, "step": 67300 }, { "grad_norm": 0.3965282738208771, "learning_rate": 6.723898231838843e-06, "loss": 0.0141, "step": 67310 }, { "grad_norm": 0.1931147575378418, "learning_rate": 6.713549744114766e-06, "loss": 0.0189, "step": 67320 }, { "grad_norm": 0.12597380578517914, "learning_rate": 6.703208652861159e-06, "loss": 0.0172, "step": 67330 }, { "grad_norm": 0.29981642961502075, "learning_rate": 6.69287495984503e-06, "loss": 0.0194, "step": 67340 }, { "grad_norm": 0.27733516693115234, "learning_rate": 6.6825486668321e-06, "loss": 0.0123, "step": 67350 }, { "grad_norm": 0.306416779756546, "learning_rate": 6.672229775586886e-06, "loss": 0.0179, "step": 67360 }, { "grad_norm": 0.21203693747520447, "learning_rate": 6.661918287872576e-06, "loss": 0.0126, "step": 67370 }, { "grad_norm": 0.32299405336380005, "learning_rate": 6.6516142054511346e-06, "loss": 0.0199, "step": 67380 }, { "grad_norm": 0.2869791090488434, "learning_rate": 6.641317530083241e-06, "loss": 0.0119, "step": 67390 }, { "grad_norm": 0.1606636494398117, "learning_rate": 6.631028263528322e-06, "loss": 0.016, "step": 67400 }, { "grad_norm": 0.2695554494857788, "learning_rate": 6.620746407544537e-06, "loss": 0.0139, "step": 67410 }, { "grad_norm": 0.2570819556713104, "learning_rate": 6.610471963888742e-06, "loss": 0.0164, "step": 67420 }, { "grad_norm": 0.2967960238456726, "learning_rate": 6.600204934316606e-06, "loss": 0.014, "step": 67430 }, { "grad_norm": 0.2567601501941681, "learning_rate": 6.589945320582452e-06, "loss": 0.013, "step": 67440 }, { "grad_norm": 0.24564073979854584, "learning_rate": 6.579693124439374e-06, "loss": 0.0147, "step": 67450 }, { "grad_norm": 0.3097344934940338, "learning_rate": 6.56944834763919e-06, "loss": 0.0142, "step": 67460 }, { "grad_norm": 0.5146655440330505, "learning_rate": 6.5592109919324575e-06, "loss": 0.014, "step": 67470 }, { "grad_norm": 0.21447189152240753, "learning_rate": 6.54898105906846e-06, "loss": 0.0151, "step": 67480 }, { "grad_norm": 0.28193914890289307, "learning_rate": 6.5387585507952155e-06, "loss": 0.0233, "step": 67490 }, { "grad_norm": 0.24052180349826813, "learning_rate": 6.528543468859461e-06, "loss": 0.014, "step": 67500 }, { "grad_norm": 0.3075425922870636, "learning_rate": 6.518335815006682e-06, "loss": 0.0194, "step": 67510 }, { "grad_norm": 0.27181389927864075, "learning_rate": 6.5081355909810845e-06, "loss": 0.0163, "step": 67520 }, { "grad_norm": 0.3502528667449951, "learning_rate": 6.497942798525608e-06, "loss": 0.0222, "step": 67530 }, { "grad_norm": 0.34288954734802246, "learning_rate": 6.487757439381936e-06, "loss": 0.0184, "step": 67540 }, { "grad_norm": 0.2669098973274231, "learning_rate": 6.477579515290433e-06, "loss": 0.0167, "step": 67550 }, { "grad_norm": 0.2651002109050751, "learning_rate": 6.467409027990273e-06, "loss": 0.0207, "step": 67560 }, { "grad_norm": 0.2822919487953186, "learning_rate": 6.457245979219279e-06, "loss": 0.0179, "step": 67570 }, { "grad_norm": 0.4098212420940399, "learning_rate": 6.447090370714054e-06, "loss": 0.0224, "step": 67580 }, { "grad_norm": 0.2896406054496765, "learning_rate": 6.436942204209917e-06, "loss": 0.0177, "step": 67590 }, { "grad_norm": 0.19835695624351501, "learning_rate": 6.4268014814408804e-06, "loss": 0.0237, "step": 67600 }, { "grad_norm": 0.2960352301597595, "learning_rate": 6.41666820413977e-06, "loss": 0.015, "step": 67610 }, { "grad_norm": 0.4109996259212494, "learning_rate": 6.406542374038033e-06, "loss": 0.0159, "step": 67620 }, { "grad_norm": 0.3007234036922455, "learning_rate": 6.396423992865935e-06, "loss": 0.012, "step": 67630 }, { "grad_norm": 0.18805943429470062, "learning_rate": 6.386313062352412e-06, "loss": 0.0123, "step": 67640 }, { "grad_norm": 0.20834626257419586, "learning_rate": 6.376209584225152e-06, "loss": 0.0122, "step": 67650 }, { "grad_norm": 0.3049582242965698, "learning_rate": 6.366113560210557e-06, "loss": 0.0158, "step": 67660 }, { "grad_norm": 0.4187300205230713, "learning_rate": 6.356024992033766e-06, "loss": 0.0199, "step": 67670 }, { "grad_norm": 0.14840929210186005, "learning_rate": 6.345943881418648e-06, "loss": 0.0145, "step": 67680 }, { "grad_norm": 0.24475856125354767, "learning_rate": 6.33587023008777e-06, "loss": 0.0124, "step": 67690 }, { "grad_norm": 0.1917579025030136, "learning_rate": 6.32580403976245e-06, "loss": 0.0106, "step": 67700 }, { "grad_norm": 0.2533906400203705, "learning_rate": 6.31574531216273e-06, "loss": 0.0154, "step": 67710 }, { "grad_norm": 0.37698447704315186, "learning_rate": 6.305694049007371e-06, "loss": 0.0201, "step": 67720 }, { "grad_norm": 0.3466252088546753, "learning_rate": 6.2956502520138575e-06, "loss": 0.0174, "step": 67730 }, { "grad_norm": 0.2568723261356354, "learning_rate": 6.285613922898409e-06, "loss": 0.0131, "step": 67740 }, { "grad_norm": 0.22981560230255127, "learning_rate": 6.275585063375927e-06, "loss": 0.0174, "step": 67750 }, { "grad_norm": 0.506217360496521, "learning_rate": 6.265563675160113e-06, "loss": 0.0161, "step": 67760 }, { "grad_norm": 0.17650648951530457, "learning_rate": 6.25554975996332e-06, "loss": 0.0137, "step": 67770 }, { "grad_norm": 0.3310600519180298, "learning_rate": 6.245543319496661e-06, "loss": 0.0135, "step": 67780 }, { "grad_norm": 0.2334151268005371, "learning_rate": 6.2355443554699685e-06, "loss": 0.0178, "step": 67790 }, { "grad_norm": 0.2597412168979645, "learning_rate": 6.225552869591766e-06, "loss": 0.0142, "step": 67800 }, { "grad_norm": 0.3137704133987427, "learning_rate": 6.215568863569365e-06, "loss": 0.0189, "step": 67810 }, { "grad_norm": 0.3856235444545746, "learning_rate": 6.205592339108712e-06, "loss": 0.0229, "step": 67820 }, { "grad_norm": 0.2903292179107666, "learning_rate": 6.195623297914577e-06, "loss": 0.0151, "step": 67830 }, { "grad_norm": 0.41928207874298096, "learning_rate": 6.185661741690357e-06, "loss": 0.0177, "step": 67840 }, { "grad_norm": 0.19964225590229034, "learning_rate": 6.1757076721382145e-06, "loss": 0.0178, "step": 67850 }, { "grad_norm": 0.2681286633014679, "learning_rate": 6.165761090959038e-06, "loss": 0.0233, "step": 67860 }, { "grad_norm": 0.2253863513469696, "learning_rate": 6.155821999852424e-06, "loss": 0.0141, "step": 67870 }, { "grad_norm": 0.24760600924491882, "learning_rate": 6.145890400516696e-06, "loss": 0.0145, "step": 67880 }, { "grad_norm": 0.30700886249542236, "learning_rate": 6.1359662946488816e-06, "loss": 0.0132, "step": 67890 }, { "grad_norm": 0.1911010593175888, "learning_rate": 6.12604968394474e-06, "loss": 0.011, "step": 67900 }, { "grad_norm": 0.18586209416389465, "learning_rate": 6.11614057009875e-06, "loss": 0.0163, "step": 67910 }, { "grad_norm": 0.2249421626329422, "learning_rate": 6.106238954804111e-06, "loss": 0.0167, "step": 67920 }, { "grad_norm": 0.20813818275928497, "learning_rate": 6.096344839752738e-06, "loss": 0.0187, "step": 67930 }, { "grad_norm": 0.1809910237789154, "learning_rate": 6.086458226635278e-06, "loss": 0.0185, "step": 67940 }, { "grad_norm": 0.2967163026332855, "learning_rate": 6.076579117141046e-06, "loss": 0.0218, "step": 67950 }, { "grad_norm": 0.351769357919693, "learning_rate": 6.066707512958153e-06, "loss": 0.0173, "step": 67960 }, { "grad_norm": 0.16876094043254852, "learning_rate": 6.056843415773361e-06, "loss": 0.0149, "step": 67970 }, { "grad_norm": 0.1896231770515442, "learning_rate": 6.0469868272721776e-06, "loss": 0.0144, "step": 67980 }, { "grad_norm": 0.2703166604042053, "learning_rate": 6.037137749138844e-06, "loss": 0.021, "step": 67990 }, { "grad_norm": 0.5031847953796387, "learning_rate": 6.027296183056252e-06, "loss": 0.0202, "step": 68000 }, { "grad_norm": 0.3397424519062042, "learning_rate": 6.017462130706114e-06, "loss": 0.015, "step": 68010 }, { "grad_norm": 0.26605525612831116, "learning_rate": 6.0076355937687516e-06, "loss": 0.0151, "step": 68020 }, { "grad_norm": 0.20485520362854004, "learning_rate": 5.9978165739232925e-06, "loss": 0.0152, "step": 68030 }, { "grad_norm": 0.17312319576740265, "learning_rate": 5.988005072847508e-06, "loss": 0.0134, "step": 68040 }, { "grad_norm": 0.2567366361618042, "learning_rate": 5.978201092217928e-06, "loss": 0.0223, "step": 68050 }, { "grad_norm": 0.16489851474761963, "learning_rate": 5.9684046337097895e-06, "loss": 0.0131, "step": 68060 }, { "grad_norm": 0.21230340003967285, "learning_rate": 5.958615698997016e-06, "loss": 0.0181, "step": 68070 }, { "grad_norm": 0.19605664908885956, "learning_rate": 5.948834289752303e-06, "loss": 0.0132, "step": 68080 }, { "grad_norm": 0.26013410091400146, "learning_rate": 5.939060407646996e-06, "loss": 0.0206, "step": 68090 }, { "grad_norm": 0.1098376214504242, "learning_rate": 5.929294054351198e-06, "loss": 0.0145, "step": 68100 }, { "grad_norm": 0.1678335964679718, "learning_rate": 5.919535231533707e-06, "loss": 0.0126, "step": 68110 }, { "grad_norm": 0.1883605271577835, "learning_rate": 5.909783940862046e-06, "loss": 0.0129, "step": 68120 }, { "grad_norm": 0.2119694948196411, "learning_rate": 5.900040184002436e-06, "loss": 0.0127, "step": 68130 }, { "grad_norm": 0.23167136311531067, "learning_rate": 5.890303962619831e-06, "loss": 0.0119, "step": 68140 }, { "grad_norm": 0.2518170177936554, "learning_rate": 5.88057527837787e-06, "loss": 0.0158, "step": 68150 }, { "grad_norm": 0.17851628363132477, "learning_rate": 5.870854132938924e-06, "loss": 0.0131, "step": 68160 }, { "grad_norm": 0.28774774074554443, "learning_rate": 5.861140527964071e-06, "loss": 0.0156, "step": 68170 }, { "grad_norm": 0.19797956943511963, "learning_rate": 5.851434465113098e-06, "loss": 0.0104, "step": 68180 }, { "grad_norm": 0.27683743834495544, "learning_rate": 5.841735946044524e-06, "loss": 0.0156, "step": 68190 }, { "grad_norm": 0.2629181444644928, "learning_rate": 5.832044972415523e-06, "loss": 0.0143, "step": 68200 }, { "grad_norm": 0.11817918717861176, "learning_rate": 5.8223615458820655e-06, "loss": 0.0132, "step": 68210 }, { "grad_norm": 0.20195157825946808, "learning_rate": 5.812685668098733e-06, "loss": 0.0168, "step": 68220 }, { "grad_norm": 0.2376376837491989, "learning_rate": 5.803017340718913e-06, "loss": 0.0158, "step": 68230 }, { "grad_norm": 0.25647059082984924, "learning_rate": 5.793356565394636e-06, "loss": 0.015, "step": 68240 }, { "grad_norm": 0.395406574010849, "learning_rate": 5.783703343776658e-06, "loss": 0.013, "step": 68250 }, { "grad_norm": 0.165385901927948, "learning_rate": 5.774057677514477e-06, "loss": 0.0111, "step": 68260 }, { "grad_norm": 0.1975088268518448, "learning_rate": 5.764419568256235e-06, "loss": 0.0175, "step": 68270 }, { "grad_norm": 0.34362122416496277, "learning_rate": 5.754789017648865e-06, "loss": 0.0306, "step": 68280 }, { "grad_norm": 0.16539542376995087, "learning_rate": 5.745166027337934e-06, "loss": 0.0199, "step": 68290 }, { "grad_norm": 0.29312992095947266, "learning_rate": 5.735550598967754e-06, "loss": 0.018, "step": 68300 }, { "grad_norm": 0.2550654709339142, "learning_rate": 5.725942734181339e-06, "loss": 0.0182, "step": 68310 }, { "grad_norm": 0.15808305144309998, "learning_rate": 5.7163424346204146e-06, "loss": 0.0133, "step": 68320 }, { "grad_norm": 0.3229910135269165, "learning_rate": 5.706749701925407e-06, "loss": 0.0153, "step": 68330 }, { "grad_norm": 0.2160472869873047, "learning_rate": 5.697164537735461e-06, "loss": 0.0115, "step": 68340 }, { "grad_norm": 0.14528805017471313, "learning_rate": 5.687586943688406e-06, "loss": 0.0113, "step": 68350 }, { "grad_norm": 0.20698274672031403, "learning_rate": 5.678016921420792e-06, "loss": 0.0149, "step": 68360 }, { "grad_norm": 0.26282548904418945, "learning_rate": 5.668454472567875e-06, "loss": 0.0192, "step": 68370 }, { "grad_norm": 0.37840721011161804, "learning_rate": 5.658899598763617e-06, "loss": 0.0227, "step": 68380 }, { "grad_norm": 0.2888320982456207, "learning_rate": 5.6493523016407035e-06, "loss": 0.0099, "step": 68390 }, { "grad_norm": 0.3386988341808319, "learning_rate": 5.6398125828304615e-06, "loss": 0.0165, "step": 68400 }, { "grad_norm": 0.2304912954568863, "learning_rate": 5.630280443963015e-06, "loss": 0.0147, "step": 68410 }, { "grad_norm": 0.25478804111480713, "learning_rate": 5.620755886667112e-06, "loss": 0.0157, "step": 68420 }, { "grad_norm": 0.5575988292694092, "learning_rate": 5.611238912570271e-06, "loss": 0.016, "step": 68430 }, { "grad_norm": 0.411265105009079, "learning_rate": 5.601729523298649e-06, "loss": 0.025, "step": 68440 }, { "grad_norm": 0.3488572835922241, "learning_rate": 5.592227720477161e-06, "loss": 0.0157, "step": 68450 }, { "grad_norm": 0.2601339519023895, "learning_rate": 5.582733505729415e-06, "loss": 0.0156, "step": 68460 }, { "grad_norm": 0.3036210536956787, "learning_rate": 5.573246880677668e-06, "loss": 0.0143, "step": 68470 }, { "grad_norm": 0.39158034324645996, "learning_rate": 5.56376784694298e-06, "loss": 0.0185, "step": 68480 }, { "grad_norm": 0.3034204840660095, "learning_rate": 5.554296406145027e-06, "loss": 0.0202, "step": 68490 }, { "grad_norm": 0.23987889289855957, "learning_rate": 5.544832559902219e-06, "loss": 0.0143, "step": 68500 }, { "grad_norm": 0.36815372109413147, "learning_rate": 5.53537630983168e-06, "loss": 0.0205, "step": 68510 }, { "grad_norm": 0.3169839382171631, "learning_rate": 5.5259276575492125e-06, "loss": 0.0216, "step": 68520 }, { "grad_norm": 0.21100358664989471, "learning_rate": 5.516486604669357e-06, "loss": 0.0213, "step": 68530 }, { "grad_norm": 0.23483140766620636, "learning_rate": 5.507053152805303e-06, "loss": 0.0212, "step": 68540 }, { "grad_norm": 0.22161820530891418, "learning_rate": 5.497627303568975e-06, "loss": 0.0219, "step": 68550 }, { "grad_norm": 0.16667121648788452, "learning_rate": 5.488209058571003e-06, "loss": 0.0099, "step": 68560 }, { "grad_norm": 0.19176709651947021, "learning_rate": 5.4787984194207054e-06, "loss": 0.0142, "step": 68570 }, { "grad_norm": 0.288753479719162, "learning_rate": 5.469395387726095e-06, "loss": 0.014, "step": 68580 }, { "grad_norm": 0.22278298437595367, "learning_rate": 5.45999996509391e-06, "loss": 0.0167, "step": 68590 }, { "grad_norm": 0.2372516244649887, "learning_rate": 5.450612153129536e-06, "loss": 0.0174, "step": 68600 }, { "grad_norm": 0.17427875101566315, "learning_rate": 5.4412319534371426e-06, "loss": 0.0171, "step": 68610 }, { "grad_norm": 0.35451096296310425, "learning_rate": 5.431859367619513e-06, "loss": 0.0141, "step": 68620 }, { "grad_norm": 0.3484896421432495, "learning_rate": 5.422494397278172e-06, "loss": 0.0147, "step": 68630 }, { "grad_norm": 0.2701221704483032, "learning_rate": 5.413137044013344e-06, "loss": 0.0189, "step": 68640 }, { "grad_norm": 0.3500915467739105, "learning_rate": 5.403787309423941e-06, "loss": 0.0177, "step": 68650 }, { "grad_norm": 0.21374854445457458, "learning_rate": 5.3944451951075835e-06, "loss": 0.0132, "step": 68660 }, { "grad_norm": 0.2919476330280304, "learning_rate": 5.385110702660562e-06, "loss": 0.0134, "step": 68670 }, { "grad_norm": 0.14011076092720032, "learning_rate": 5.375783833677922e-06, "loss": 0.0178, "step": 68680 }, { "grad_norm": 0.31020134687423706, "learning_rate": 5.36646458975334e-06, "loss": 0.0182, "step": 68690 }, { "grad_norm": 0.43138113617897034, "learning_rate": 5.3571529724792294e-06, "loss": 0.0158, "step": 68700 }, { "grad_norm": 0.4390152394771576, "learning_rate": 5.347848983446702e-06, "loss": 0.018, "step": 68710 }, { "grad_norm": 0.24382983148097992, "learning_rate": 5.3385526242455185e-06, "loss": 0.0163, "step": 68720 }, { "grad_norm": 0.20644232630729675, "learning_rate": 5.329263896464226e-06, "loss": 0.0161, "step": 68730 }, { "grad_norm": 0.21122188866138458, "learning_rate": 5.3199828016899715e-06, "loss": 0.0166, "step": 68740 }, { "grad_norm": 0.2673951983451843, "learning_rate": 5.310709341508657e-06, "loss": 0.0137, "step": 68750 }, { "grad_norm": 0.18302322924137115, "learning_rate": 5.301443517504861e-06, "loss": 0.014, "step": 68760 }, { "grad_norm": 0.3031088411808014, "learning_rate": 5.292185331261862e-06, "loss": 0.014, "step": 68770 }, { "grad_norm": 0.2162351757287979, "learning_rate": 5.28293478436162e-06, "loss": 0.0139, "step": 68780 }, { "grad_norm": 0.20276445150375366, "learning_rate": 5.273691878384829e-06, "loss": 0.0204, "step": 68790 }, { "grad_norm": 0.24985501170158386, "learning_rate": 5.264456614910812e-06, "loss": 0.017, "step": 68800 }, { "grad_norm": 0.38659578561782837, "learning_rate": 5.255228995517647e-06, "loss": 0.0267, "step": 68810 }, { "grad_norm": 0.2678135335445404, "learning_rate": 5.246009021782067e-06, "loss": 0.0147, "step": 68820 }, { "grad_norm": 0.14864838123321533, "learning_rate": 5.2367966952795225e-06, "loss": 0.0151, "step": 68830 }, { "grad_norm": 0.3725906014442444, "learning_rate": 5.2275920175841485e-06, "loss": 0.0135, "step": 68840 }, { "grad_norm": 0.1892402321100235, "learning_rate": 5.218394990268771e-06, "loss": 0.0143, "step": 68850 }, { "grad_norm": 0.37452054023742676, "learning_rate": 5.209205614904916e-06, "loss": 0.0188, "step": 68860 }, { "grad_norm": 0.326233834028244, "learning_rate": 5.200023893062777e-06, "loss": 0.0183, "step": 68870 }, { "grad_norm": 0.2693529427051544, "learning_rate": 5.190849826311289e-06, "loss": 0.0165, "step": 68880 }, { "grad_norm": 0.38960808515548706, "learning_rate": 5.181683416218025e-06, "loss": 0.0183, "step": 68890 }, { "grad_norm": 0.3333480954170227, "learning_rate": 5.172524664349276e-06, "loss": 0.015, "step": 68900 }, { "grad_norm": 0.1527310311794281, "learning_rate": 5.1633735722700416e-06, "loss": 0.0134, "step": 68910 }, { "grad_norm": 0.12124904245138168, "learning_rate": 5.154230141543958e-06, "loss": 0.0168, "step": 68920 }, { "grad_norm": 0.24810481071472168, "learning_rate": 5.145094373733433e-06, "loss": 0.0167, "step": 68930 }, { "grad_norm": 0.10802315175533295, "learning_rate": 5.135966270399478e-06, "loss": 0.0112, "step": 68940 }, { "grad_norm": 0.46826496720314026, "learning_rate": 5.126845833101857e-06, "loss": 0.0219, "step": 68950 }, { "grad_norm": 0.21658426523208618, "learning_rate": 5.1177330633989994e-06, "loss": 0.0132, "step": 68960 }, { "grad_norm": 0.2939557731151581, "learning_rate": 5.108627962848033e-06, "loss": 0.0278, "step": 68970 }, { "grad_norm": 0.17344065010547638, "learning_rate": 5.099530533004759e-06, "loss": 0.0215, "step": 68980 }, { "grad_norm": 0.1661597192287445, "learning_rate": 5.090440775423699e-06, "loss": 0.0109, "step": 68990 }, { "grad_norm": 0.18537692725658417, "learning_rate": 5.081358691658022e-06, "loss": 0.014, "step": 69000 }, { "grad_norm": 0.42805373668670654, "learning_rate": 5.072284283259621e-06, "loss": 0.0195, "step": 69010 }, { "grad_norm": 0.12924250960350037, "learning_rate": 5.063217551779053e-06, "loss": 0.0191, "step": 69020 }, { "grad_norm": 0.4382362365722656, "learning_rate": 5.054158498765588e-06, "loss": 0.0159, "step": 69030 }, { "grad_norm": 0.2575695216655731, "learning_rate": 5.0451071257671625e-06, "loss": 0.0233, "step": 69040 }, { "grad_norm": 0.2578377425670624, "learning_rate": 5.036063434330407e-06, "loss": 0.0204, "step": 69050 }, { "grad_norm": 0.13583967089653015, "learning_rate": 5.027027426000652e-06, "loss": 0.0154, "step": 69060 }, { "grad_norm": 0.20387113094329834, "learning_rate": 5.017999102321886e-06, "loss": 0.0163, "step": 69070 }, { "grad_norm": 0.43138086795806885, "learning_rate": 5.0089784648368224e-06, "loss": 0.0173, "step": 69080 }, { "grad_norm": 0.22649678587913513, "learning_rate": 4.999965515086829e-06, "loss": 0.0159, "step": 69090 }, { "grad_norm": 0.15962186455726624, "learning_rate": 4.990960254611976e-06, "loss": 0.0125, "step": 69100 }, { "grad_norm": 0.38863709568977356, "learning_rate": 4.9819626849510194e-06, "loss": 0.0165, "step": 69110 }, { "grad_norm": 0.2733607292175293, "learning_rate": 4.97297280764138e-06, "loss": 0.0228, "step": 69120 }, { "grad_norm": 0.19915223121643066, "learning_rate": 4.963990624219211e-06, "loss": 0.0204, "step": 69130 }, { "grad_norm": 0.23773673176765442, "learning_rate": 4.955016136219298e-06, "loss": 0.013, "step": 69140 }, { "grad_norm": 0.34307000041007996, "learning_rate": 4.946049345175146e-06, "loss": 0.01, "step": 69150 }, { "grad_norm": 0.3211008310317993, "learning_rate": 4.937090252618937e-06, "loss": 0.0161, "step": 69160 }, { "grad_norm": 0.2324097603559494, "learning_rate": 4.928138860081521e-06, "loss": 0.022, "step": 69170 }, { "grad_norm": 0.1402025818824768, "learning_rate": 4.919195169092472e-06, "loss": 0.0169, "step": 69180 }, { "grad_norm": 0.23850466310977936, "learning_rate": 4.910259181179994e-06, "loss": 0.0173, "step": 69190 }, { "grad_norm": 0.2787415087223053, "learning_rate": 4.901330897871015e-06, "loss": 0.0191, "step": 69200 }, { "grad_norm": 0.16533231735229492, "learning_rate": 4.8924103206911375e-06, "loss": 0.0127, "step": 69210 }, { "grad_norm": 0.287110835313797, "learning_rate": 4.883497451164637e-06, "loss": 0.0207, "step": 69220 }, { "grad_norm": 0.2982155382633209, "learning_rate": 4.874592290814484e-06, "loss": 0.0136, "step": 69230 }, { "grad_norm": 0.44412264227867126, "learning_rate": 4.865694841162327e-06, "loss": 0.0231, "step": 69240 }, { "grad_norm": 0.2157384157180786, "learning_rate": 4.856805103728496e-06, "loss": 0.017, "step": 69250 }, { "grad_norm": 0.2804131507873535, "learning_rate": 4.847923080032007e-06, "loss": 0.0137, "step": 69260 }, { "grad_norm": 0.20825889706611633, "learning_rate": 4.8390487715905486e-06, "loss": 0.0136, "step": 69270 }, { "grad_norm": 0.13502703607082367, "learning_rate": 4.8301821799205e-06, "loss": 0.012, "step": 69280 }, { "grad_norm": 0.14442172646522522, "learning_rate": 4.821323306536918e-06, "loss": 0.0132, "step": 69290 }, { "grad_norm": 0.27429068088531494, "learning_rate": 4.8124721529535455e-06, "loss": 0.0156, "step": 69300 }, { "grad_norm": 0.25971323251724243, "learning_rate": 4.803628720682807e-06, "loss": 0.0211, "step": 69310 }, { "grad_norm": 0.3471425175666809, "learning_rate": 4.794793011235776e-06, "loss": 0.017, "step": 69320 }, { "grad_norm": 0.3214307427406311, "learning_rate": 4.78596502612228e-06, "loss": 0.0179, "step": 69330 }, { "grad_norm": 0.2845475673675537, "learning_rate": 4.777144766850738e-06, "loss": 0.0182, "step": 69340 }, { "grad_norm": 0.19171491265296936, "learning_rate": 4.768332234928313e-06, "loss": 0.0182, "step": 69350 }, { "grad_norm": 0.1573980301618576, "learning_rate": 4.759527431860828e-06, "loss": 0.0107, "step": 69360 }, { "grad_norm": 0.305301696062088, "learning_rate": 4.750730359152755e-06, "loss": 0.0198, "step": 69370 }, { "grad_norm": 0.16311301290988922, "learning_rate": 4.741941018307311e-06, "loss": 0.0122, "step": 69380 }, { "grad_norm": 0.1257106214761734, "learning_rate": 4.733159410826321e-06, "loss": 0.0186, "step": 69390 }, { "grad_norm": 0.15756823122501373, "learning_rate": 4.724385538210357e-06, "loss": 0.0135, "step": 69400 }, { "grad_norm": 0.13465233147144318, "learning_rate": 4.715619401958599e-06, "loss": 0.0152, "step": 69410 }, { "grad_norm": 0.4030759334564209, "learning_rate": 4.706861003568958e-06, "loss": 0.0224, "step": 69420 }, { "grad_norm": 0.2310808300971985, "learning_rate": 4.698110344538003e-06, "loss": 0.0153, "step": 69430 }, { "grad_norm": 0.4889182150363922, "learning_rate": 4.689367426360975e-06, "loss": 0.0153, "step": 69440 }, { "grad_norm": 0.3656935691833496, "learning_rate": 4.680632250531819e-06, "loss": 0.0163, "step": 69450 }, { "grad_norm": 0.1840905249118805, "learning_rate": 4.671904818543115e-06, "loss": 0.0119, "step": 69460 }, { "grad_norm": 0.2814893126487732, "learning_rate": 4.66318513188615e-06, "loss": 0.0144, "step": 69470 }, { "grad_norm": 0.09443849325180054, "learning_rate": 4.654473192050884e-06, "loss": 0.0134, "step": 69480 }, { "grad_norm": 0.18127407133579254, "learning_rate": 4.64576900052594e-06, "loss": 0.0148, "step": 69490 }, { "grad_norm": 0.26627039909362793, "learning_rate": 4.637072558798638e-06, "loss": 0.0129, "step": 69500 }, { "grad_norm": 0.22627028822898865, "learning_rate": 4.628383868354969e-06, "loss": 0.0178, "step": 69510 }, { "grad_norm": 0.17496809363365173, "learning_rate": 4.6197029306795595e-06, "loss": 0.0112, "step": 69520 }, { "grad_norm": 0.1923024207353592, "learning_rate": 4.611029747255779e-06, "loss": 0.0122, "step": 69530 }, { "grad_norm": 0.26203349232673645, "learning_rate": 4.6023643195656164e-06, "loss": 0.0146, "step": 69540 }, { "grad_norm": 0.14087170362472534, "learning_rate": 4.593706649089768e-06, "loss": 0.0112, "step": 69550 }, { "grad_norm": 0.2844771146774292, "learning_rate": 4.585056737307597e-06, "loss": 0.0143, "step": 69560 }, { "grad_norm": 0.26202476024627686, "learning_rate": 4.576414585697103e-06, "loss": 0.0122, "step": 69570 }, { "grad_norm": 0.3277914524078369, "learning_rate": 4.567780195735044e-06, "loss": 0.0181, "step": 69580 }, { "grad_norm": 0.21768762171268463, "learning_rate": 4.559153568896757e-06, "loss": 0.0115, "step": 69590 }, { "grad_norm": 0.35537096858024597, "learning_rate": 4.550534706656329e-06, "loss": 0.0139, "step": 69600 }, { "grad_norm": 0.14731021225452423, "learning_rate": 4.541923610486465e-06, "loss": 0.0152, "step": 69610 }, { "grad_norm": 0.39493367075920105, "learning_rate": 4.533320281858578e-06, "loss": 0.0133, "step": 69620 }, { "grad_norm": 0.11986140161752701, "learning_rate": 4.52472472224274e-06, "loss": 0.017, "step": 69630 }, { "grad_norm": 0.3643089532852173, "learning_rate": 4.51613693310769e-06, "loss": 0.017, "step": 69640 }, { "grad_norm": 0.2835071384906769, "learning_rate": 4.507556915920868e-06, "loss": 0.0164, "step": 69650 }, { "grad_norm": 0.26195672154426575, "learning_rate": 4.498984672148332e-06, "loss": 0.018, "step": 69660 }, { "grad_norm": 0.17318230867385864, "learning_rate": 4.490420203254864e-06, "loss": 0.0157, "step": 69670 }, { "grad_norm": 0.51932692527771, "learning_rate": 4.481863510703893e-06, "loss": 0.0189, "step": 69680 }, { "grad_norm": 0.16600963473320007, "learning_rate": 4.473314595957523e-06, "loss": 0.0142, "step": 69690 }, { "grad_norm": 0.4557359218597412, "learning_rate": 4.464773460476535e-06, "loss": 0.0176, "step": 69700 }, { "grad_norm": 0.14920702576637268, "learning_rate": 4.456240105720372e-06, "loss": 0.0156, "step": 69710 }, { "grad_norm": 0.15488140285015106, "learning_rate": 4.4477145331471405e-06, "loss": 0.0227, "step": 69720 }, { "grad_norm": 0.3600446581840515, "learning_rate": 4.439196744213653e-06, "loss": 0.0153, "step": 69730 }, { "grad_norm": 0.24303454160690308, "learning_rate": 4.430686740375339e-06, "loss": 0.0159, "step": 69740 }, { "grad_norm": 0.2758345901966095, "learning_rate": 4.422184523086342e-06, "loss": 0.0182, "step": 69750 }, { "grad_norm": 0.2293042093515396, "learning_rate": 4.41369009379946e-06, "loss": 0.0115, "step": 69760 }, { "grad_norm": 0.1909051239490509, "learning_rate": 4.405203453966139e-06, "loss": 0.0149, "step": 69770 }, { "grad_norm": 0.31010380387306213, "learning_rate": 4.396724605036539e-06, "loss": 0.0173, "step": 69780 }, { "grad_norm": 0.3035873472690582, "learning_rate": 4.388253548459437e-06, "loss": 0.014, "step": 69790 }, { "grad_norm": 0.2623118758201599, "learning_rate": 4.3797902856823395e-06, "loss": 0.0143, "step": 69800 }, { "grad_norm": 0.26861411333084106, "learning_rate": 4.371334818151357e-06, "loss": 0.0153, "step": 69810 }, { "grad_norm": 0.19992201030254364, "learning_rate": 4.362887147311306e-06, "loss": 0.0138, "step": 69820 }, { "grad_norm": 0.2548406422138214, "learning_rate": 4.354447274605672e-06, "loss": 0.0138, "step": 69830 }, { "grad_norm": 0.27158963680267334, "learning_rate": 4.346015201476572e-06, "loss": 0.0243, "step": 69840 }, { "grad_norm": 0.32163670659065247, "learning_rate": 4.337590929364855e-06, "loss": 0.0147, "step": 69850 }, { "grad_norm": 0.171713188290596, "learning_rate": 4.329174459709973e-06, "loss": 0.0137, "step": 69860 }, { "grad_norm": 0.16927945613861084, "learning_rate": 4.320765793950071e-06, "loss": 0.0143, "step": 69870 }, { "grad_norm": 0.2372470200061798, "learning_rate": 4.312364933521962e-06, "loss": 0.0194, "step": 69880 }, { "grad_norm": 0.2683662176132202, "learning_rate": 4.303971879861129e-06, "loss": 0.0144, "step": 69890 }, { "grad_norm": 0.28745394945144653, "learning_rate": 4.295586634401716e-06, "loss": 0.018, "step": 69900 }, { "grad_norm": 0.22657231986522675, "learning_rate": 4.287209198576536e-06, "loss": 0.0115, "step": 69910 }, { "grad_norm": 0.2315933257341385, "learning_rate": 4.278839573817045e-06, "loss": 0.0162, "step": 69920 }, { "grad_norm": 0.33910703659057617, "learning_rate": 4.270477761553399e-06, "loss": 0.0172, "step": 69930 }, { "grad_norm": 0.24205781519412994, "learning_rate": 4.262123763214393e-06, "loss": 0.011, "step": 69940 }, { "grad_norm": 0.19966328144073486, "learning_rate": 4.253777580227508e-06, "loss": 0.0132, "step": 69950 }, { "grad_norm": 0.20876996219158173, "learning_rate": 4.2454392140188755e-06, "loss": 0.0147, "step": 69960 }, { "grad_norm": 0.2029406875371933, "learning_rate": 4.2371086660132785e-06, "loss": 0.0156, "step": 69970 }, { "grad_norm": 0.207806795835495, "learning_rate": 4.228785937634205e-06, "loss": 0.0124, "step": 69980 }, { "grad_norm": 0.22590027749538422, "learning_rate": 4.220471030303758e-06, "loss": 0.0132, "step": 69990 }, { "grad_norm": 0.2147855907678604, "learning_rate": 4.212163945442754e-06, "loss": 0.0123, "step": 70000 }, { "grad_norm": 0.3476477265357971, "learning_rate": 4.203864684470621e-06, "loss": 0.0135, "step": 70010 }, { "grad_norm": 0.36012038588523865, "learning_rate": 4.19557324880549e-06, "loss": 0.0146, "step": 70020 }, { "grad_norm": 0.24182191491127014, "learning_rate": 4.187289639864145e-06, "loss": 0.0161, "step": 70030 }, { "grad_norm": 0.23208126425743103, "learning_rate": 4.1790138590619974e-06, "loss": 0.0121, "step": 70040 }, { "grad_norm": 0.28387701511383057, "learning_rate": 4.170745907813195e-06, "loss": 0.022, "step": 70050 }, { "grad_norm": 0.23047180473804474, "learning_rate": 4.162485787530479e-06, "loss": 0.0148, "step": 70060 }, { "grad_norm": 0.26727911829948425, "learning_rate": 4.154233499625282e-06, "loss": 0.0116, "step": 70070 }, { "grad_norm": 0.2105945497751236, "learning_rate": 4.145989045507692e-06, "loss": 0.0156, "step": 70080 }, { "grad_norm": 0.1971326619386673, "learning_rate": 4.1377524265864666e-06, "loss": 0.0189, "step": 70090 }, { "grad_norm": 0.1270320862531662, "learning_rate": 4.1295236442690175e-06, "loss": 0.013, "step": 70100 }, { "grad_norm": 0.27162057161331177, "learning_rate": 4.121302699961421e-06, "loss": 0.0131, "step": 70110 }, { "grad_norm": 0.29656660556793213, "learning_rate": 4.113089595068403e-06, "loss": 0.0202, "step": 70120 }, { "grad_norm": 0.25561854243278503, "learning_rate": 4.104884330993364e-06, "loss": 0.0171, "step": 70130 }, { "grad_norm": 0.20890671014785767, "learning_rate": 4.0966869091383585e-06, "loss": 0.0261, "step": 70140 }, { "grad_norm": 0.18504394590854645, "learning_rate": 4.088497330904101e-06, "loss": 0.0168, "step": 70150 }, { "grad_norm": 0.22376473248004913, "learning_rate": 4.080315597689976e-06, "loss": 0.0122, "step": 70160 }, { "grad_norm": 0.15031848847866058, "learning_rate": 4.072141710893995e-06, "loss": 0.0178, "step": 70170 }, { "grad_norm": 0.14317020773887634, "learning_rate": 4.063975671912879e-06, "loss": 0.023, "step": 70180 }, { "grad_norm": 0.2545466423034668, "learning_rate": 4.055817482141949e-06, "loss": 0.0095, "step": 70190 }, { "grad_norm": 0.3069535195827484, "learning_rate": 4.047667142975259e-06, "loss": 0.0158, "step": 70200 }, { "grad_norm": 0.19558501243591309, "learning_rate": 4.039524655805443e-06, "loss": 0.0215, "step": 70210 }, { "grad_norm": 0.21106606721878052, "learning_rate": 4.03139002202384e-06, "loss": 0.0121, "step": 70220 }, { "grad_norm": 0.3359844386577606, "learning_rate": 4.023263243020447e-06, "loss": 0.0178, "step": 70230 }, { "grad_norm": 0.26291555166244507, "learning_rate": 4.015144320183884e-06, "loss": 0.0166, "step": 70240 }, { "grad_norm": 0.18474863469600677, "learning_rate": 4.007033254901482e-06, "loss": 0.0111, "step": 70250 }, { "grad_norm": 0.35184574127197266, "learning_rate": 3.9989300485591795e-06, "loss": 0.0131, "step": 70260 }, { "grad_norm": 0.2785131633281708, "learning_rate": 3.990834702541601e-06, "loss": 0.0214, "step": 70270 }, { "grad_norm": 0.1939736157655716, "learning_rate": 3.982747218232019e-06, "loss": 0.015, "step": 70280 }, { "grad_norm": 0.3393644094467163, "learning_rate": 3.97466759701236e-06, "loss": 0.0183, "step": 70290 }, { "grad_norm": 0.2609576880931854, "learning_rate": 3.966595840263226e-06, "loss": 0.0131, "step": 70300 }, { "grad_norm": 0.35768571496009827, "learning_rate": 3.958531949363836e-06, "loss": 0.0149, "step": 70310 }, { "grad_norm": 0.1610487401485443, "learning_rate": 3.950475925692098e-06, "loss": 0.0174, "step": 70320 }, { "grad_norm": 0.27463528513908386, "learning_rate": 3.9424277706245685e-06, "loss": 0.0141, "step": 70330 }, { "grad_norm": 0.24232187867164612, "learning_rate": 3.934387485536451e-06, "loss": 0.0141, "step": 70340 }, { "grad_norm": 0.2363923341035843, "learning_rate": 3.926355071801619e-06, "loss": 0.0128, "step": 70350 }, { "grad_norm": 0.23880723118782043, "learning_rate": 3.9183305307925965e-06, "loss": 0.0148, "step": 70360 }, { "grad_norm": 0.1779807060956955, "learning_rate": 3.910313863880533e-06, "loss": 0.0131, "step": 70370 }, { "grad_norm": 0.14298351109027863, "learning_rate": 3.902305072435292e-06, "loss": 0.0098, "step": 70380 }, { "grad_norm": 0.272275447845459, "learning_rate": 3.894304157825329e-06, "loss": 0.0172, "step": 70390 }, { "grad_norm": 0.12456566095352173, "learning_rate": 3.886311121417791e-06, "loss": 0.0127, "step": 70400 }, { "grad_norm": 0.1922362893819809, "learning_rate": 3.878325964578472e-06, "loss": 0.0131, "step": 70410 }, { "grad_norm": 0.137167289853096, "learning_rate": 3.870348688671815e-06, "loss": 0.013, "step": 70420 }, { "grad_norm": 0.31979379057884216, "learning_rate": 3.862379295060931e-06, "loss": 0.0134, "step": 70430 }, { "grad_norm": 0.1762218326330185, "learning_rate": 3.854417785107534e-06, "loss": 0.0196, "step": 70440 }, { "grad_norm": 0.22674104571342468, "learning_rate": 3.8464641601720755e-06, "loss": 0.0125, "step": 70450 }, { "grad_norm": 0.26572978496551514, "learning_rate": 3.838518421613579e-06, "loss": 0.0159, "step": 70460 }, { "grad_norm": 0.34218117594718933, "learning_rate": 3.830580570789766e-06, "loss": 0.0167, "step": 70470 }, { "grad_norm": 0.14470382034778595, "learning_rate": 3.822650609057005e-06, "loss": 0.0123, "step": 70480 }, { "grad_norm": 0.23112082481384277, "learning_rate": 3.814728537770285e-06, "loss": 0.014, "step": 70490 }, { "grad_norm": 0.3647233247756958, "learning_rate": 3.806814358283306e-06, "loss": 0.0175, "step": 70500 }, { "grad_norm": 0.353193074464798, "learning_rate": 3.7989080719483596e-06, "loss": 0.0139, "step": 70510 }, { "grad_norm": 0.2872089743614197, "learning_rate": 3.7910096801164143e-06, "loss": 0.0195, "step": 70520 }, { "grad_norm": 0.200949564576149, "learning_rate": 3.7831191841371016e-06, "loss": 0.0147, "step": 70530 }, { "grad_norm": 0.2810545861721039, "learning_rate": 3.775236585358688e-06, "loss": 0.0152, "step": 70540 }, { "grad_norm": 0.2667272984981537, "learning_rate": 3.7673618851280843e-06, "loss": 0.0163, "step": 70550 }, { "grad_norm": 0.3445201814174652, "learning_rate": 3.759495084790887e-06, "loss": 0.0155, "step": 70560 }, { "grad_norm": 0.24579055607318878, "learning_rate": 3.751636185691282e-06, "loss": 0.0136, "step": 70570 }, { "grad_norm": 0.18409046530723572, "learning_rate": 3.7437851891721607e-06, "loss": 0.0129, "step": 70580 }, { "grad_norm": 0.24981169402599335, "learning_rate": 3.7359420965750404e-06, "loss": 0.0193, "step": 70590 }, { "grad_norm": 0.12686772644519806, "learning_rate": 3.7281069092400922e-06, "loss": 0.0125, "step": 70600 }, { "grad_norm": 0.36378213763237, "learning_rate": 3.7202796285061348e-06, "loss": 0.0144, "step": 70610 }, { "grad_norm": 0.21543551981449127, "learning_rate": 3.712460255710637e-06, "loss": 0.0201, "step": 70620 }, { "grad_norm": 0.3033676743507385, "learning_rate": 3.704648792189719e-06, "loss": 0.0139, "step": 70630 }, { "grad_norm": 0.4742696285247803, "learning_rate": 3.696845239278124e-06, "loss": 0.016, "step": 70640 }, { "grad_norm": 0.3635319769382477, "learning_rate": 3.689049598309302e-06, "loss": 0.0189, "step": 70650 }, { "grad_norm": 0.23392470180988312, "learning_rate": 3.681261870615288e-06, "loss": 0.0256, "step": 70660 }, { "grad_norm": 0.14386968314647675, "learning_rate": 3.6734820575268004e-06, "loss": 0.0201, "step": 70670 }, { "grad_norm": 0.24683168530464172, "learning_rate": 3.665710160373209e-06, "loss": 0.0142, "step": 70680 }, { "grad_norm": 0.15206101536750793, "learning_rate": 3.65794618048248e-06, "loss": 0.0139, "step": 70690 }, { "grad_norm": 0.21725234389305115, "learning_rate": 3.6501901191813125e-06, "loss": 0.0113, "step": 70700 }, { "grad_norm": 0.3177042007446289, "learning_rate": 3.642441977794975e-06, "loss": 0.0179, "step": 70710 }, { "grad_norm": 0.24566981196403503, "learning_rate": 3.634701757647424e-06, "loss": 0.014, "step": 70720 }, { "grad_norm": 0.4058896601200104, "learning_rate": 3.6269694600612468e-06, "loss": 0.0112, "step": 70730 }, { "grad_norm": 0.20273394882678986, "learning_rate": 3.619245086357681e-06, "loss": 0.0145, "step": 70740 }, { "grad_norm": 0.32815808057785034, "learning_rate": 3.611528637856615e-06, "loss": 0.0183, "step": 70750 }, { "grad_norm": 0.3437960743904114, "learning_rate": 3.6038201158765884e-06, "loss": 0.011, "step": 70760 }, { "grad_norm": 0.5386282801628113, "learning_rate": 3.5961195217347534e-06, "loss": 0.0237, "step": 70770 }, { "grad_norm": 0.24764202535152435, "learning_rate": 3.588426856746946e-06, "loss": 0.0161, "step": 70780 }, { "grad_norm": 0.224964901804924, "learning_rate": 3.5807421222276316e-06, "loss": 0.026, "step": 70790 }, { "grad_norm": 0.1468157023191452, "learning_rate": 3.5730653194899156e-06, "loss": 0.0143, "step": 70800 }, { "grad_norm": 0.2998543381690979, "learning_rate": 3.565396449845554e-06, "loss": 0.0131, "step": 70810 }, { "grad_norm": 0.32703477144241333, "learning_rate": 3.557735514604954e-06, "loss": 0.0194, "step": 70820 }, { "grad_norm": 0.13916365802288055, "learning_rate": 3.5500825150771633e-06, "loss": 0.0129, "step": 70830 }, { "grad_norm": 0.19377610087394714, "learning_rate": 3.5424374525698466e-06, "loss": 0.0141, "step": 70840 }, { "grad_norm": 0.24488228559494019, "learning_rate": 3.5348003283893704e-06, "loss": 0.0147, "step": 70850 }, { "grad_norm": 0.19599024951457977, "learning_rate": 3.527171143840685e-06, "loss": 0.0159, "step": 70860 }, { "grad_norm": 0.3248295485973358, "learning_rate": 3.519549900227409e-06, "loss": 0.024, "step": 70870 }, { "grad_norm": 0.28370824456214905, "learning_rate": 3.511936598851828e-06, "loss": 0.0196, "step": 70880 }, { "grad_norm": 0.17786024510860443, "learning_rate": 3.504331241014813e-06, "loss": 0.0143, "step": 70890 }, { "grad_norm": 0.3591499924659729, "learning_rate": 3.4967338280159414e-06, "loss": 0.0115, "step": 70900 }, { "grad_norm": 0.24115124344825745, "learning_rate": 3.4891443611533846e-06, "loss": 0.0224, "step": 70910 }, { "grad_norm": 0.1855691522359848, "learning_rate": 3.481562841723984e-06, "loss": 0.0199, "step": 70920 }, { "grad_norm": 0.2904166579246521, "learning_rate": 3.473989271023215e-06, "loss": 0.0151, "step": 70930 }, { "grad_norm": 0.23144365847110748, "learning_rate": 3.4664236503451854e-06, "loss": 0.0151, "step": 70940 }, { "grad_norm": 0.268226683139801, "learning_rate": 3.4588659809826674e-06, "loss": 0.0142, "step": 70950 }, { "grad_norm": 0.20719344913959503, "learning_rate": 3.451316264227039e-06, "loss": 0.0208, "step": 70960 }, { "grad_norm": 0.20325709879398346, "learning_rate": 3.4437745013683574e-06, "loss": 0.0183, "step": 70970 }, { "grad_norm": 0.21547333896160126, "learning_rate": 3.4362406936952916e-06, "loss": 0.0127, "step": 70980 }, { "grad_norm": 0.22566168010234833, "learning_rate": 3.4287148424951786e-06, "loss": 0.0124, "step": 70990 }, { "grad_norm": 0.17039723694324493, "learning_rate": 3.4211969490539617e-06, "loss": 0.0124, "step": 71000 }, { "grad_norm": 0.2358579784631729, "learning_rate": 3.4136870146562584e-06, "loss": 0.0121, "step": 71010 }, { "grad_norm": 0.1893051713705063, "learning_rate": 3.406185040585308e-06, "loss": 0.0174, "step": 71020 }, { "grad_norm": 0.2145748883485794, "learning_rate": 3.3986910281229966e-06, "loss": 0.0167, "step": 71030 }, { "grad_norm": 0.23897744715213776, "learning_rate": 3.391204978549828e-06, "loss": 0.0182, "step": 71040 }, { "grad_norm": 0.32533636689186096, "learning_rate": 3.3837268931449785e-06, "loss": 0.0232, "step": 71050 }, { "grad_norm": 0.18283295631408691, "learning_rate": 3.376256773186248e-06, "loss": 0.0186, "step": 71060 }, { "grad_norm": 0.26865246891975403, "learning_rate": 3.3687946199500664e-06, "loss": 0.0186, "step": 71070 }, { "grad_norm": 0.43022704124450684, "learning_rate": 3.3613404347115295e-06, "loss": 0.0111, "step": 71080 }, { "grad_norm": 0.22726409137248993, "learning_rate": 3.35389421874433e-06, "loss": 0.011, "step": 71090 }, { "grad_norm": 0.14598990976810455, "learning_rate": 3.3464559733208446e-06, "loss": 0.0128, "step": 71100 }, { "grad_norm": 0.2895301580429077, "learning_rate": 3.3390256997120505e-06, "loss": 0.0257, "step": 71110 }, { "grad_norm": 0.19624854624271393, "learning_rate": 3.331603399187583e-06, "loss": 0.0182, "step": 71120 }, { "grad_norm": 0.18222588300704956, "learning_rate": 3.324189073015721e-06, "loss": 0.02, "step": 71130 }, { "grad_norm": 0.23821350932121277, "learning_rate": 3.3167827224633465e-06, "loss": 0.0158, "step": 71140 }, { "grad_norm": 0.18540728092193604, "learning_rate": 3.309384348796024e-06, "loss": 0.0157, "step": 71150 }, { "grad_norm": 0.28378841280937195, "learning_rate": 3.301993953277921e-06, "loss": 0.0149, "step": 71160 }, { "grad_norm": 0.2918597459793091, "learning_rate": 3.29461153717186e-06, "loss": 0.0129, "step": 71170 }, { "grad_norm": 0.3138251006603241, "learning_rate": 3.287237101739293e-06, "loss": 0.0239, "step": 71180 }, { "grad_norm": 0.2577381730079651, "learning_rate": 3.2798706482403075e-06, "loss": 0.0229, "step": 71190 }, { "grad_norm": 0.23305872082710266, "learning_rate": 3.2725121779336285e-06, "loss": 0.0178, "step": 71200 }, { "grad_norm": 0.31750980019569397, "learning_rate": 3.265161692076618e-06, "loss": 0.018, "step": 71210 }, { "grad_norm": 0.2639026641845703, "learning_rate": 3.257819191925282e-06, "loss": 0.0136, "step": 71220 }, { "grad_norm": 0.22563037276268005, "learning_rate": 3.2504846787342392e-06, "loss": 0.015, "step": 71230 }, { "grad_norm": 0.27535027265548706, "learning_rate": 3.243158153756759e-06, "loss": 0.0198, "step": 71240 }, { "grad_norm": 0.2589913308620453, "learning_rate": 3.2358396182447514e-06, "loss": 0.0116, "step": 71250 }, { "grad_norm": 0.22304224967956543, "learning_rate": 3.2285290734487496e-06, "loss": 0.0125, "step": 71260 }, { "grad_norm": 0.18108640611171722, "learning_rate": 3.221226520617926e-06, "loss": 0.0125, "step": 71270 }, { "grad_norm": 0.2841857671737671, "learning_rate": 3.213931961000094e-06, "loss": 0.0198, "step": 71280 }, { "grad_norm": 0.3435736298561096, "learning_rate": 3.2066453958416733e-06, "loss": 0.0145, "step": 71290 }, { "grad_norm": 0.18200036883354187, "learning_rate": 3.199366826387773e-06, "loss": 0.0154, "step": 71300 }, { "grad_norm": 0.21995316445827484, "learning_rate": 3.192096253882071e-06, "loss": 0.0182, "step": 71310 }, { "grad_norm": 0.21346339583396912, "learning_rate": 3.1848336795669177e-06, "loss": 0.0152, "step": 71320 }, { "grad_norm": 0.25810351967811584, "learning_rate": 3.1775791046833035e-06, "loss": 0.0182, "step": 71330 }, { "grad_norm": 0.16186267137527466, "learning_rate": 3.170332530470804e-06, "loss": 0.0158, "step": 71340 }, { "grad_norm": 0.3107032775878906, "learning_rate": 3.1630939581677012e-06, "loss": 0.0194, "step": 71350 }, { "grad_norm": 0.19814148545265198, "learning_rate": 3.155863389010838e-06, "loss": 0.0174, "step": 71360 }, { "grad_norm": 0.21080191433429718, "learning_rate": 3.1486408242357323e-06, "loss": 0.0214, "step": 71370 }, { "grad_norm": 0.38550034165382385, "learning_rate": 3.1414262650765248e-06, "loss": 0.0387, "step": 71380 }, { "grad_norm": 0.3418893814086914, "learning_rate": 3.134219712765979e-06, "loss": 0.0227, "step": 71390 }, { "grad_norm": 0.23501302301883698, "learning_rate": 3.1270211685355044e-06, "loss": 0.02, "step": 71400 }, { "grad_norm": 0.19179488718509674, "learning_rate": 3.1198306336151338e-06, "loss": 0.0131, "step": 71410 }, { "grad_norm": 0.14774349331855774, "learning_rate": 3.11264810923354e-06, "loss": 0.0176, "step": 71420 }, { "grad_norm": 0.19109292328357697, "learning_rate": 3.105473596618008e-06, "loss": 0.0098, "step": 71430 }, { "grad_norm": 0.2486005425453186, "learning_rate": 3.0983070969944683e-06, "loss": 0.0147, "step": 71440 }, { "grad_norm": 0.2642764747142792, "learning_rate": 3.09114861158748e-06, "loss": 0.0101, "step": 71450 }, { "grad_norm": 0.2094971388578415, "learning_rate": 3.0839981416202314e-06, "loss": 0.0168, "step": 71460 }, { "grad_norm": 0.18058207631111145, "learning_rate": 3.0768556883145507e-06, "loss": 0.0122, "step": 71470 }, { "grad_norm": 0.20207881927490234, "learning_rate": 3.0697212528908834e-06, "loss": 0.0146, "step": 71480 }, { "grad_norm": 0.23180288076400757, "learning_rate": 3.0625948365683e-06, "loss": 0.0175, "step": 71490 }, { "grad_norm": 0.27309727668762207, "learning_rate": 3.055476440564531e-06, "loss": 0.0182, "step": 71500 }, { "grad_norm": 0.1563166230916977, "learning_rate": 3.0483660660958924e-06, "loss": 0.0162, "step": 71510 }, { "grad_norm": 0.23020285367965698, "learning_rate": 3.0412637143773624e-06, "loss": 0.0252, "step": 71520 }, { "grad_norm": 0.21545575559139252, "learning_rate": 3.034169386622554e-06, "loss": 0.0144, "step": 71530 }, { "grad_norm": 0.31063786149024963, "learning_rate": 3.027083084043658e-06, "loss": 0.0136, "step": 71540 }, { "grad_norm": 0.2697797119617462, "learning_rate": 3.0200048078515676e-06, "loss": 0.0169, "step": 71550 }, { "grad_norm": 0.2711721658706665, "learning_rate": 3.012934559255742e-06, "loss": 0.0163, "step": 71560 }, { "grad_norm": 0.27978020906448364, "learning_rate": 3.0058723394642994e-06, "loss": 0.0163, "step": 71570 }, { "grad_norm": 0.30634891986846924, "learning_rate": 2.998818149683985e-06, "loss": 0.0232, "step": 71580 }, { "grad_norm": 0.24075531959533691, "learning_rate": 2.9917719911201627e-06, "loss": 0.0162, "step": 71590 }, { "grad_norm": 0.2466227263212204, "learning_rate": 2.9847338649768352e-06, "loss": 0.0135, "step": 71600 }, { "grad_norm": 0.17403370141983032, "learning_rate": 2.977703772456608e-06, "loss": 0.0138, "step": 71610 }, { "grad_norm": 0.265999972820282, "learning_rate": 2.9706817147607535e-06, "loss": 0.0154, "step": 71620 }, { "grad_norm": 0.2842404544353485, "learning_rate": 2.9636676930891393e-06, "loss": 0.0175, "step": 71630 }, { "grad_norm": 0.1438155621290207, "learning_rate": 2.9566617086402625e-06, "loss": 0.0159, "step": 71640 }, { "grad_norm": 0.18484292924404144, "learning_rate": 2.9496637626112655e-06, "loss": 0.0144, "step": 71650 }, { "grad_norm": 0.32122910022735596, "learning_rate": 2.9426738561979027e-06, "loss": 0.0149, "step": 71660 }, { "grad_norm": 0.26214155554771423, "learning_rate": 2.9356919905945524e-06, "loss": 0.032, "step": 71670 }, { "grad_norm": 0.23275989294052124, "learning_rate": 2.928718166994243e-06, "loss": 0.0198, "step": 71680 }, { "grad_norm": 0.38050779700279236, "learning_rate": 2.9217523865885833e-06, "loss": 0.0208, "step": 71690 }, { "grad_norm": 0.1870974749326706, "learning_rate": 2.9147946505678537e-06, "loss": 0.0133, "step": 71700 }, { "grad_norm": 0.19313907623291016, "learning_rate": 2.9078449601209313e-06, "loss": 0.0132, "step": 71710 }, { "grad_norm": 0.34619563817977905, "learning_rate": 2.900903316435333e-06, "loss": 0.017, "step": 71720 }, { "grad_norm": 0.3565541207790375, "learning_rate": 2.8939697206971983e-06, "loss": 0.019, "step": 71730 }, { "grad_norm": 0.2168617695569992, "learning_rate": 2.8870441740912746e-06, "loss": 0.0134, "step": 71740 }, { "grad_norm": 0.3410702049732208, "learning_rate": 2.8801266778009762e-06, "loss": 0.0157, "step": 71750 }, { "grad_norm": 0.22993698716163635, "learning_rate": 2.8732172330082798e-06, "loss": 0.0195, "step": 71760 }, { "grad_norm": 0.14568021893501282, "learning_rate": 2.8663158408938517e-06, "loss": 0.0158, "step": 71770 }, { "grad_norm": 0.17955553531646729, "learning_rate": 2.859422502636938e-06, "loss": 0.0129, "step": 71780 }, { "grad_norm": 0.23194634914398193, "learning_rate": 2.8525372194154076e-06, "loss": 0.0122, "step": 71790 }, { "grad_norm": 0.19392992556095123, "learning_rate": 2.8456599924057913e-06, "loss": 0.0131, "step": 71800 }, { "grad_norm": 0.18028725683689117, "learning_rate": 2.8387908227831995e-06, "loss": 0.0187, "step": 71810 }, { "grad_norm": 0.17365705966949463, "learning_rate": 2.831929711721404e-06, "loss": 0.0179, "step": 71820 }, { "grad_norm": 0.18092267215251923, "learning_rate": 2.825076660392767e-06, "loss": 0.0128, "step": 71830 }, { "grad_norm": 0.17551398277282715, "learning_rate": 2.8182316699682908e-06, "loss": 0.0191, "step": 71840 }, { "grad_norm": 0.26495853066444397, "learning_rate": 2.811394741617601e-06, "loss": 0.0152, "step": 71850 }, { "grad_norm": 0.29722705483436584, "learning_rate": 2.8045658765089356e-06, "loss": 0.0157, "step": 71860 }, { "grad_norm": 0.29590708017349243, "learning_rate": 2.7977450758091605e-06, "loss": 0.0238, "step": 71870 }, { "grad_norm": 0.28148385882377625, "learning_rate": 2.790932340683783e-06, "loss": 0.019, "step": 71880 }, { "grad_norm": 0.20325928926467896, "learning_rate": 2.7841276722968823e-06, "loss": 0.0171, "step": 71890 }, { "grad_norm": 0.20720124244689941, "learning_rate": 2.7773310718112067e-06, "loss": 0.0151, "step": 71900 }, { "grad_norm": 0.23269912600517273, "learning_rate": 2.7705425403881102e-06, "loss": 0.014, "step": 71910 }, { "grad_norm": 0.11885742843151093, "learning_rate": 2.763762079187565e-06, "loss": 0.0123, "step": 71920 }, { "grad_norm": 0.18314430117607117, "learning_rate": 2.756989689368178e-06, "loss": 0.0188, "step": 71930 }, { "grad_norm": 0.2633041441440582, "learning_rate": 2.750225372087134e-06, "loss": 0.0153, "step": 71940 }, { "grad_norm": 0.1976548731327057, "learning_rate": 2.7434691285003033e-06, "loss": 0.013, "step": 71950 }, { "grad_norm": 0.2697567045688629, "learning_rate": 2.7367209597621178e-06, "loss": 0.0149, "step": 71960 }, { "grad_norm": 0.20464885234832764, "learning_rate": 2.729980867025683e-06, "loss": 0.0146, "step": 71970 }, { "grad_norm": 0.27703970670700073, "learning_rate": 2.7232488514426724e-06, "loss": 0.0123, "step": 71980 }, { "grad_norm": 0.2331671118736267, "learning_rate": 2.7165249141634097e-06, "loss": 0.0141, "step": 71990 }, { "grad_norm": 0.2187308371067047, "learning_rate": 2.709809056336837e-06, "loss": 0.0204, "step": 72000 }, { "grad_norm": 0.19245745241641998, "learning_rate": 2.703101279110498e-06, "loss": 0.015, "step": 72010 }, { "grad_norm": 0.35779765248298645, "learning_rate": 2.696401583630587e-06, "loss": 0.0225, "step": 72020 }, { "grad_norm": 0.18995530903339386, "learning_rate": 2.689709971041887e-06, "loss": 0.0136, "step": 72030 }, { "grad_norm": 0.46741411089897156, "learning_rate": 2.6830264424878126e-06, "loss": 0.0186, "step": 72040 }, { "grad_norm": 0.25978925824165344, "learning_rate": 2.676350999110394e-06, "loss": 0.0267, "step": 72050 }, { "grad_norm": 0.2736106514930725, "learning_rate": 2.6696836420502856e-06, "loss": 0.0207, "step": 72060 }, { "grad_norm": 0.2075016349554062, "learning_rate": 2.663024372446765e-06, "loss": 0.0117, "step": 72070 }, { "grad_norm": 0.18175113201141357, "learning_rate": 2.6563731914377055e-06, "loss": 0.0127, "step": 72080 }, { "grad_norm": 0.14348693192005157, "learning_rate": 2.6497301001596087e-06, "loss": 0.0207, "step": 72090 }, { "grad_norm": 0.23270125687122345, "learning_rate": 2.643095099747611e-06, "loss": 0.0175, "step": 72100 }, { "grad_norm": 0.19361676275730133, "learning_rate": 2.636468191335445e-06, "loss": 0.0157, "step": 72110 }, { "grad_norm": 0.26717594265937805, "learning_rate": 2.629849376055471e-06, "loss": 0.017, "step": 72120 }, { "grad_norm": 0.19129836559295654, "learning_rate": 2.6232386550386678e-06, "loss": 0.0155, "step": 72130 }, { "grad_norm": 0.20164011418819427, "learning_rate": 2.6166360294146097e-06, "loss": 0.0135, "step": 72140 }, { "grad_norm": 0.1743135303258896, "learning_rate": 2.6100415003115275e-06, "loss": 0.0202, "step": 72150 }, { "grad_norm": 0.2262740582227707, "learning_rate": 2.603455068856225e-06, "loss": 0.0166, "step": 72160 }, { "grad_norm": 0.21078987419605255, "learning_rate": 2.5968767361741584e-06, "loss": 0.015, "step": 72170 }, { "grad_norm": 0.2939801514148712, "learning_rate": 2.590306503389378e-06, "loss": 0.0177, "step": 72180 }, { "grad_norm": 0.2721458077430725, "learning_rate": 2.583744371624558e-06, "loss": 0.0261, "step": 72190 }, { "grad_norm": 0.2605692446231842, "learning_rate": 2.57719034200099e-06, "loss": 0.0152, "step": 72200 }, { "grad_norm": 0.3419889509677887, "learning_rate": 2.570644415638568e-06, "loss": 0.0143, "step": 72210 }, { "grad_norm": 0.19831009209156036, "learning_rate": 2.5641065936558295e-06, "loss": 0.0195, "step": 72220 }, { "grad_norm": 0.23934942483901978, "learning_rate": 2.557576877169898e-06, "loss": 0.0212, "step": 72230 }, { "grad_norm": 0.370322048664093, "learning_rate": 2.5510552672965205e-06, "loss": 0.0142, "step": 72240 }, { "grad_norm": 0.33505669236183167, "learning_rate": 2.544541765150077e-06, "loss": 0.0137, "step": 72250 }, { "grad_norm": 0.20624472200870514, "learning_rate": 2.5380363718435165e-06, "loss": 0.018, "step": 72260 }, { "grad_norm": 0.2574249804019928, "learning_rate": 2.5315390884884714e-06, "loss": 0.0108, "step": 72270 }, { "grad_norm": 0.3014138638973236, "learning_rate": 2.5250499161951214e-06, "loss": 0.0184, "step": 72280 }, { "grad_norm": 0.18731306493282318, "learning_rate": 2.518568856072301e-06, "loss": 0.0093, "step": 72290 }, { "grad_norm": 0.2808803617954254, "learning_rate": 2.512095909227441e-06, "loss": 0.017, "step": 72300 }, { "grad_norm": 0.29180145263671875, "learning_rate": 2.50563107676659e-06, "loss": 0.0187, "step": 72310 }, { "grad_norm": 0.36232277750968933, "learning_rate": 2.4991743597944152e-06, "loss": 0.0155, "step": 72320 }, { "grad_norm": 0.1471160501241684, "learning_rate": 2.492725759414205e-06, "loss": 0.0155, "step": 72330 }, { "grad_norm": 0.1761508285999298, "learning_rate": 2.4862852767278234e-06, "loss": 0.009, "step": 72340 }, { "grad_norm": 0.11389870196580887, "learning_rate": 2.479852912835784e-06, "loss": 0.0104, "step": 72350 }, { "grad_norm": 0.29798346757888794, "learning_rate": 2.4734286688372075e-06, "loss": 0.0176, "step": 72360 }, { "grad_norm": 0.38641127943992615, "learning_rate": 2.467012545829811e-06, "loss": 0.0225, "step": 72370 }, { "grad_norm": 0.25339391827583313, "learning_rate": 2.460604544909945e-06, "loss": 0.0137, "step": 72380 }, { "grad_norm": 0.24519971013069153, "learning_rate": 2.454204667172555e-06, "loss": 0.0098, "step": 72390 }, { "grad_norm": 0.19716966152191162, "learning_rate": 2.447812913711217e-06, "loss": 0.0152, "step": 72400 }, { "grad_norm": 0.10993792116641998, "learning_rate": 2.44142928561808e-06, "loss": 0.0182, "step": 72410 }, { "grad_norm": 0.19439499080181122, "learning_rate": 2.435053783983965e-06, "loss": 0.0149, "step": 72420 }, { "grad_norm": 0.33250269293785095, "learning_rate": 2.4286864098982453e-06, "loss": 0.0154, "step": 72430 }, { "grad_norm": 0.11927396059036255, "learning_rate": 2.4223271644489397e-06, "loss": 0.0135, "step": 72440 }, { "grad_norm": 0.5576775670051575, "learning_rate": 2.415976048722679e-06, "loss": 0.0128, "step": 72450 }, { "grad_norm": 0.32134249806404114, "learning_rate": 2.4096330638046673e-06, "loss": 0.0147, "step": 72460 }, { "grad_norm": 0.2050008326768875, "learning_rate": 2.4032982107787816e-06, "loss": 0.0183, "step": 72470 }, { "grad_norm": 0.1765158325433731, "learning_rate": 2.396971490727451e-06, "loss": 0.0194, "step": 72480 }, { "grad_norm": 0.2694711685180664, "learning_rate": 2.3906529047317493e-06, "loss": 0.0127, "step": 72490 }, { "grad_norm": 0.22251802682876587, "learning_rate": 2.3843424538713465e-06, "loss": 0.0145, "step": 72500 }, { "grad_norm": 0.39523619413375854, "learning_rate": 2.3780401392245298e-06, "loss": 0.0177, "step": 72510 }, { "grad_norm": 0.1693686544895172, "learning_rate": 2.3717459618681883e-06, "loss": 0.0107, "step": 72520 }, { "grad_norm": 0.15788021683692932, "learning_rate": 2.3654599228778274e-06, "loss": 0.0122, "step": 72530 }, { "grad_norm": 0.3393179476261139, "learning_rate": 2.3591820233275607e-06, "loss": 0.0173, "step": 72540 }, { "grad_norm": 0.08432936668395996, "learning_rate": 2.3529122642901024e-06, "loss": 0.0136, "step": 72550 }, { "grad_norm": 0.20438839495182037, "learning_rate": 2.3466506468367845e-06, "loss": 0.0246, "step": 72560 }, { "grad_norm": 0.15185187757015228, "learning_rate": 2.3403971720375506e-06, "loss": 0.0145, "step": 72570 }, { "grad_norm": 0.2599256634712219, "learning_rate": 2.334151840960952e-06, "loss": 0.0131, "step": 72580 }, { "grad_norm": 0.14860497415065765, "learning_rate": 2.3279146546741347e-06, "loss": 0.0119, "step": 72590 }, { "grad_norm": 0.4694006145000458, "learning_rate": 2.32168561424288e-06, "loss": 0.0136, "step": 72600 }, { "grad_norm": 0.11604578047990799, "learning_rate": 2.3154647207315307e-06, "loss": 0.0167, "step": 72610 }, { "grad_norm": 0.38261064887046814, "learning_rate": 2.309251975203103e-06, "loss": 0.0171, "step": 72620 }, { "grad_norm": 0.2399720400571823, "learning_rate": 2.303047378719159e-06, "loss": 0.0202, "step": 72630 }, { "grad_norm": 0.1894955188035965, "learning_rate": 2.2968509323399e-06, "loss": 0.0173, "step": 72640 }, { "grad_norm": 0.1341005265712738, "learning_rate": 2.290662637124147e-06, "loss": 0.0113, "step": 72650 }, { "grad_norm": 0.1469821035861969, "learning_rate": 2.2844824941292807e-06, "loss": 0.0092, "step": 72660 }, { "grad_norm": 0.2500680088996887, "learning_rate": 2.2783105044113406e-06, "loss": 0.0184, "step": 72670 }, { "grad_norm": 0.2169220745563507, "learning_rate": 2.272146669024944e-06, "loss": 0.013, "step": 72680 }, { "grad_norm": 0.27226579189300537, "learning_rate": 2.26599098902332e-06, "loss": 0.0143, "step": 72690 }, { "grad_norm": 0.24393032491207123, "learning_rate": 2.2598434654583113e-06, "loss": 0.0169, "step": 72700 }, { "grad_norm": 0.3065067231655121, "learning_rate": 2.253704099380355e-06, "loss": 0.0221, "step": 72710 }, { "grad_norm": 0.22804668545722961, "learning_rate": 2.247572891838512e-06, "loss": 0.0173, "step": 72720 }, { "grad_norm": 0.41458868980407715, "learning_rate": 2.241449843880422e-06, "loss": 0.0183, "step": 72730 }, { "grad_norm": 0.20546552538871765, "learning_rate": 2.235334956552354e-06, "loss": 0.0174, "step": 72740 }, { "grad_norm": 0.18594792485237122, "learning_rate": 2.2292282308991775e-06, "loss": 0.0178, "step": 72750 }, { "grad_norm": 0.2797333896160126, "learning_rate": 2.223129667964363e-06, "loss": 0.0155, "step": 72760 }, { "grad_norm": 0.28988468647003174, "learning_rate": 2.2170392687899834e-06, "loss": 0.0112, "step": 72770 }, { "grad_norm": 0.11616716533899307, "learning_rate": 2.210957034416733e-06, "loss": 0.0117, "step": 72780 }, { "grad_norm": 0.23408173024654388, "learning_rate": 2.2048829658838867e-06, "loss": 0.0146, "step": 72790 }, { "grad_norm": 0.12575377523899078, "learning_rate": 2.1988170642293525e-06, "loss": 0.0127, "step": 72800 }, { "grad_norm": 0.13337326049804688, "learning_rate": 2.1927593304896075e-06, "loss": 0.0106, "step": 72810 }, { "grad_norm": 0.20425474643707275, "learning_rate": 2.1867097656997626e-06, "loss": 0.0136, "step": 72820 }, { "grad_norm": 0.16752766072750092, "learning_rate": 2.180668370893524e-06, "loss": 0.0138, "step": 72830 }, { "grad_norm": 0.34123891592025757, "learning_rate": 2.174635147103199e-06, "loss": 0.0139, "step": 72840 }, { "grad_norm": 0.241019144654274, "learning_rate": 2.1686100953597075e-06, "loss": 0.0148, "step": 72850 }, { "grad_norm": 0.2948845624923706, "learning_rate": 2.1625932166925433e-06, "loss": 0.0145, "step": 72860 }, { "grad_norm": 0.3847816288471222, "learning_rate": 2.1565845121298556e-06, "loss": 0.0126, "step": 72870 }, { "grad_norm": 0.19925647974014282, "learning_rate": 2.150583982698351e-06, "loss": 0.0173, "step": 72880 }, { "grad_norm": 0.449684739112854, "learning_rate": 2.144591629423359e-06, "loss": 0.0138, "step": 72890 }, { "grad_norm": 0.20248928666114807, "learning_rate": 2.138607453328817e-06, "loss": 0.0161, "step": 72900 }, { "grad_norm": 0.2903115451335907, "learning_rate": 2.132631455437234e-06, "loss": 0.0184, "step": 72910 }, { "grad_norm": 0.2142850160598755, "learning_rate": 2.1266636367697714e-06, "loss": 0.0231, "step": 72920 }, { "grad_norm": 0.35964906215667725, "learning_rate": 2.120703998346152e-06, "loss": 0.0158, "step": 72930 }, { "grad_norm": 0.2850298285484314, "learning_rate": 2.1147525411847114e-06, "loss": 0.0171, "step": 72940 }, { "grad_norm": 0.16409914195537567, "learning_rate": 2.108809266302403e-06, "loss": 0.0135, "step": 72950 }, { "grad_norm": 0.44783055782318115, "learning_rate": 2.102874174714764e-06, "loss": 0.0169, "step": 72960 }, { "grad_norm": 0.24399937689304352, "learning_rate": 2.0969472674359393e-06, "loss": 0.0153, "step": 72970 }, { "grad_norm": 0.13752922415733337, "learning_rate": 2.091028545478668e-06, "loss": 0.0152, "step": 72980 }, { "grad_norm": 0.37343335151672363, "learning_rate": 2.08511800985432e-06, "loss": 0.0135, "step": 72990 }, { "grad_norm": 0.17911292612552643, "learning_rate": 2.07921566157282e-06, "loss": 0.021, "step": 73000 }, { "grad_norm": 0.1834980547428131, "learning_rate": 2.073321501642728e-06, "loss": 0.0147, "step": 73010 }, { "grad_norm": 0.1925290822982788, "learning_rate": 2.0674355310711937e-06, "loss": 0.0202, "step": 73020 }, { "grad_norm": 0.23721130192279816, "learning_rate": 2.0615577508639682e-06, "loss": 0.01, "step": 73030 }, { "grad_norm": 0.10061311721801758, "learning_rate": 2.055688162025399e-06, "loss": 0.0123, "step": 73040 }, { "grad_norm": 0.15089592337608337, "learning_rate": 2.0498267655584546e-06, "loss": 0.0194, "step": 73050 }, { "grad_norm": 0.19244463741779327, "learning_rate": 2.0439735624646626e-06, "loss": 0.0154, "step": 73060 }, { "grad_norm": 0.2163478583097458, "learning_rate": 2.0381285537442008e-06, "loss": 0.0182, "step": 73070 }, { "grad_norm": 0.27456581592559814, "learning_rate": 2.032291740395803e-06, "loss": 0.0141, "step": 73080 }, { "grad_norm": 0.22075976431369781, "learning_rate": 2.0264631234168276e-06, "loss": 0.0157, "step": 73090 }, { "grad_norm": 0.28004294633865356, "learning_rate": 2.0206427038032326e-06, "loss": 0.0166, "step": 73100 }, { "grad_norm": 0.25283920764923096, "learning_rate": 2.0148304825495457e-06, "loss": 0.0124, "step": 73110 }, { "grad_norm": 0.18172143399715424, "learning_rate": 2.0090264606489496e-06, "loss": 0.0142, "step": 73120 }, { "grad_norm": 0.17868240177631378, "learning_rate": 2.003230639093162e-06, "loss": 0.0129, "step": 73130 }, { "grad_norm": 0.11732418090105057, "learning_rate": 1.9974430188725525e-06, "loss": 0.0159, "step": 73140 }, { "grad_norm": 0.26225677132606506, "learning_rate": 1.9916636009760513e-06, "loss": 0.0148, "step": 73150 }, { "grad_norm": 0.20979458093643188, "learning_rate": 1.985892386391208e-06, "loss": 0.0165, "step": 73160 }, { "grad_norm": 0.2271103858947754, "learning_rate": 1.980129376104173e-06, "loss": 0.0194, "step": 73170 }, { "grad_norm": 0.16681216657161713, "learning_rate": 1.9743745710996796e-06, "loss": 0.0157, "step": 73180 }, { "grad_norm": 0.08228758722543716, "learning_rate": 1.9686279723610757e-06, "loss": 0.0173, "step": 73190 }, { "grad_norm": 0.5146677494049072, "learning_rate": 1.962889580870281e-06, "loss": 0.0167, "step": 73200 }, { "grad_norm": 0.40217652916908264, "learning_rate": 1.9571593976078385e-06, "loss": 0.0129, "step": 73210 }, { "grad_norm": 0.18238967657089233, "learning_rate": 1.9514374235528876e-06, "loss": 0.0157, "step": 73220 }, { "grad_norm": 0.2913033366203308, "learning_rate": 1.9457236596831408e-06, "loss": 0.0137, "step": 73230 }, { "grad_norm": 0.16338029503822327, "learning_rate": 1.940018106974939e-06, "loss": 0.0192, "step": 73240 }, { "grad_norm": 0.15580716729164124, "learning_rate": 1.9343207664032025e-06, "loss": 0.0143, "step": 73250 }, { "grad_norm": 0.10087677836418152, "learning_rate": 1.928631638941436e-06, "loss": 0.0113, "step": 73260 }, { "grad_norm": 0.16078665852546692, "learning_rate": 1.922950725561784e-06, "loss": 0.0147, "step": 73270 }, { "grad_norm": 0.13552816212177277, "learning_rate": 1.9172780272349312e-06, "loss": 0.013, "step": 73280 }, { "grad_norm": 0.2624725103378296, "learning_rate": 1.911613544930202e-06, "loss": 0.0146, "step": 73290 }, { "grad_norm": 0.14958052337169647, "learning_rate": 1.9059572796155e-06, "loss": 0.0167, "step": 73300 }, { "grad_norm": 0.31230077147483826, "learning_rate": 1.9003092322573192e-06, "loss": 0.013, "step": 73310 }, { "grad_norm": 0.3103550374507904, "learning_rate": 1.8946694038207647e-06, "loss": 0.0237, "step": 73320 }, { "grad_norm": 0.28848573565483093, "learning_rate": 1.8890377952695215e-06, "loss": 0.0139, "step": 73330 }, { "grad_norm": 0.21828575432300568, "learning_rate": 1.8834144075658865e-06, "loss": 0.0154, "step": 73340 }, { "grad_norm": 0.2182784229516983, "learning_rate": 1.8777992416707357e-06, "loss": 0.0199, "step": 73350 }, { "grad_norm": 0.19914501905441284, "learning_rate": 1.8721922985435458e-06, "loss": 0.0109, "step": 73360 }, { "grad_norm": 0.21789789199829102, "learning_rate": 1.8665935791424006e-06, "loss": 0.017, "step": 73370 }, { "grad_norm": 0.2387203425168991, "learning_rate": 1.8610030844239512e-06, "loss": 0.0161, "step": 73380 }, { "grad_norm": 0.19327186048030853, "learning_rate": 1.8554208153434838e-06, "loss": 0.0114, "step": 73390 }, { "grad_norm": 0.3456512689590454, "learning_rate": 1.8498467728548296e-06, "loss": 0.01, "step": 73400 }, { "grad_norm": 0.24019946157932281, "learning_rate": 1.8442809579104547e-06, "loss": 0.0149, "step": 73410 }, { "grad_norm": 0.34733450412750244, "learning_rate": 1.8387233714614039e-06, "loss": 0.0165, "step": 73420 }, { "grad_norm": 0.19407552480697632, "learning_rate": 1.8331740144573174e-06, "loss": 0.0211, "step": 73430 }, { "grad_norm": 0.23327775299549103, "learning_rate": 1.8276328878464255e-06, "loss": 0.0129, "step": 73440 }, { "grad_norm": 0.3468153476715088, "learning_rate": 1.82209999257556e-06, "loss": 0.0155, "step": 73450 }, { "grad_norm": 0.1905720829963684, "learning_rate": 1.8165753295901311e-06, "loss": 0.0172, "step": 73460 }, { "grad_norm": 0.25449901819229126, "learning_rate": 1.8110588998341616e-06, "loss": 0.0168, "step": 73470 }, { "grad_norm": 0.17311134934425354, "learning_rate": 1.805550704250253e-06, "loss": 0.0264, "step": 73480 }, { "grad_norm": 0.2720947861671448, "learning_rate": 1.8000507437796077e-06, "loss": 0.0152, "step": 73490 }, { "grad_norm": 0.20674844086170197, "learning_rate": 1.7945590193620242e-06, "loss": 0.0151, "step": 73500 }, { "grad_norm": 0.34605929255485535, "learning_rate": 1.7890755319358742e-06, "loss": 0.0142, "step": 73510 }, { "grad_norm": 0.36347365379333496, "learning_rate": 1.7836002824381525e-06, "loss": 0.0173, "step": 73520 }, { "grad_norm": 0.2736019194126129, "learning_rate": 1.7781332718044165e-06, "loss": 0.0092, "step": 73530 }, { "grad_norm": 0.2664792537689209, "learning_rate": 1.772674500968835e-06, "loss": 0.0159, "step": 73540 }, { "grad_norm": 0.31945642828941345, "learning_rate": 1.7672239708641624e-06, "loss": 0.0181, "step": 73550 }, { "grad_norm": 0.30801430344581604, "learning_rate": 1.7617816824217371e-06, "loss": 0.0105, "step": 73560 }, { "grad_norm": 0.27610936760902405, "learning_rate": 1.7563476365715148e-06, "loss": 0.0233, "step": 73570 }, { "grad_norm": 0.20748871564865112, "learning_rate": 1.750921834242003e-06, "loss": 0.027, "step": 73580 }, { "grad_norm": 0.16762091219425201, "learning_rate": 1.7455042763603436e-06, "loss": 0.0161, "step": 73590 }, { "grad_norm": 0.1813977062702179, "learning_rate": 1.7400949638522345e-06, "loss": 0.0102, "step": 73600 }, { "grad_norm": 0.44176602363586426, "learning_rate": 1.734693897641987e-06, "loss": 0.0319, "step": 73610 }, { "grad_norm": 0.16359874606132507, "learning_rate": 1.7293010786524955e-06, "loss": 0.0108, "step": 73620 }, { "grad_norm": 0.34666287899017334, "learning_rate": 1.72391650780524e-06, "loss": 0.0187, "step": 73630 }, { "grad_norm": 0.2583019733428955, "learning_rate": 1.718540186020301e-06, "loss": 0.0139, "step": 73640 }, { "grad_norm": 0.252300500869751, "learning_rate": 1.7131721142163437e-06, "loss": 0.0145, "step": 73650 }, { "grad_norm": 0.22345639765262604, "learning_rate": 1.7078122933106233e-06, "loss": 0.019, "step": 73660 }, { "grad_norm": 0.21521954238414764, "learning_rate": 1.7024607242189905e-06, "loss": 0.0181, "step": 73670 }, { "grad_norm": 0.1754511594772339, "learning_rate": 1.6971174078558749e-06, "loss": 0.012, "step": 73680 }, { "grad_norm": 0.22486305236816406, "learning_rate": 1.6917823451343073e-06, "loss": 0.0168, "step": 73690 }, { "grad_norm": 0.24194231629371643, "learning_rate": 1.6864555369659142e-06, "loss": 0.0152, "step": 73700 }, { "grad_norm": 0.33447736501693726, "learning_rate": 1.681136984260878e-06, "loss": 0.0202, "step": 73710 }, { "grad_norm": 0.21312522888183594, "learning_rate": 1.6758266879280172e-06, "loss": 0.0124, "step": 73720 }, { "grad_norm": 0.23236840963363647, "learning_rate": 1.6705246488746996e-06, "loss": 0.012, "step": 73730 }, { "grad_norm": 0.1928640753030777, "learning_rate": 1.6652308680069062e-06, "loss": 0.0157, "step": 73740 }, { "grad_norm": 0.2672840654850006, "learning_rate": 1.6599453462292081e-06, "loss": 0.0145, "step": 73750 }, { "grad_norm": 0.37766918540000916, "learning_rate": 1.6546680844447326e-06, "loss": 0.0181, "step": 73760 }, { "grad_norm": 0.23932276666164398, "learning_rate": 1.6493990835552475e-06, "loss": 0.0177, "step": 73770 }, { "grad_norm": 0.2129199057817459, "learning_rate": 1.6441383444610547e-06, "loss": 0.013, "step": 73780 }, { "grad_norm": 0.16509540379047394, "learning_rate": 1.6388858680610908e-06, "loss": 0.0176, "step": 73790 }, { "grad_norm": 0.21824796497821808, "learning_rate": 1.6336416552528543e-06, "loss": 0.0125, "step": 73800 }, { "grad_norm": 0.13298963010311127, "learning_rate": 1.6284057069324288e-06, "loss": 0.0164, "step": 73810 }, { "grad_norm": 0.21941988170146942, "learning_rate": 1.623178023994504e-06, "loss": 0.0193, "step": 73820 }, { "grad_norm": 0.2486545443534851, "learning_rate": 1.6179586073323483e-06, "loss": 0.0153, "step": 73830 }, { "grad_norm": 0.33138835430145264, "learning_rate": 1.612747457837821e-06, "loss": 0.0135, "step": 73840 }, { "grad_norm": 0.25092723965644836, "learning_rate": 1.607544576401354e-06, "loss": 0.0177, "step": 73850 }, { "grad_norm": 0.24779003858566284, "learning_rate": 1.6023499639119754e-06, "loss": 0.0159, "step": 73860 }, { "grad_norm": 0.23972943425178528, "learning_rate": 1.5971636212573138e-06, "loss": 0.0166, "step": 73870 }, { "grad_norm": 0.22352376580238342, "learning_rate": 1.5919855493235714e-06, "loss": 0.014, "step": 73880 }, { "grad_norm": 0.18956172466278076, "learning_rate": 1.586815748995535e-06, "loss": 0.0125, "step": 73890 }, { "grad_norm": 0.1688580960035324, "learning_rate": 1.5816542211565866e-06, "loss": 0.0109, "step": 73900 }, { "grad_norm": 0.15853239595890045, "learning_rate": 1.5765009666886765e-06, "loss": 0.0136, "step": 73910 }, { "grad_norm": 0.2564862370491028, "learning_rate": 1.5713559864723781e-06, "loss": 0.0203, "step": 73920 }, { "grad_norm": 0.2932562232017517, "learning_rate": 1.5662192813868104e-06, "loss": 0.0135, "step": 73930 }, { "grad_norm": 0.2689657211303711, "learning_rate": 1.561090852309699e-06, "loss": 0.0189, "step": 73940 }, { "grad_norm": 0.3528340458869934, "learning_rate": 1.5559707001173651e-06, "loss": 0.0125, "step": 73950 }, { "grad_norm": 0.2117641419172287, "learning_rate": 1.5508588256846757e-06, "loss": 0.0176, "step": 73960 }, { "grad_norm": 0.14885786175727844, "learning_rate": 1.5457552298851319e-06, "loss": 0.0134, "step": 73970 }, { "grad_norm": 0.3101990818977356, "learning_rate": 1.5406599135907918e-06, "loss": 0.015, "step": 73980 }, { "grad_norm": 0.185090109705925, "learning_rate": 1.5355728776723088e-06, "loss": 0.018, "step": 73990 }, { "grad_norm": 0.17061622440814972, "learning_rate": 1.5304941229989155e-06, "loss": 0.0124, "step": 74000 }, { "grad_norm": 0.26684918999671936, "learning_rate": 1.5254236504384345e-06, "loss": 0.0181, "step": 74010 }, { "grad_norm": 0.14673109352588654, "learning_rate": 1.5203614608572726e-06, "loss": 0.0173, "step": 74020 }, { "grad_norm": 0.3175637423992157, "learning_rate": 1.5153075551204044e-06, "loss": 0.013, "step": 74030 }, { "grad_norm": 0.24887822568416595, "learning_rate": 1.5102619340914225e-06, "loss": 0.0275, "step": 74040 }, { "grad_norm": 0.22532129287719727, "learning_rate": 1.505224598632482e-06, "loss": 0.0128, "step": 74050 }, { "grad_norm": 0.10432735085487366, "learning_rate": 1.5001955496043164e-06, "loss": 0.0183, "step": 74060 }, { "grad_norm": 0.4023880362510681, "learning_rate": 1.4951747878662604e-06, "loss": 0.0137, "step": 74070 }, { "grad_norm": 0.15947359800338745, "learning_rate": 1.4901623142762221e-06, "loss": 0.0161, "step": 74080 }, { "grad_norm": 0.16904056072235107, "learning_rate": 1.4851581296907001e-06, "loss": 0.0147, "step": 74090 }, { "grad_norm": 0.17048844695091248, "learning_rate": 1.4801622349647714e-06, "loss": 0.0118, "step": 74100 }, { "grad_norm": 0.3791833221912384, "learning_rate": 1.4751746309520976e-06, "loss": 0.017, "step": 74110 }, { "grad_norm": 0.18874657154083252, "learning_rate": 1.4701953185049132e-06, "loss": 0.0177, "step": 74120 }, { "grad_norm": 0.18950825929641724, "learning_rate": 1.4652242984740661e-06, "loss": 0.0179, "step": 74130 }, { "grad_norm": 0.14251655340194702, "learning_rate": 1.4602615717089484e-06, "loss": 0.0217, "step": 74140 }, { "grad_norm": 0.224446102976799, "learning_rate": 1.4553071390575657e-06, "loss": 0.0141, "step": 74150 }, { "grad_norm": 0.19961011409759521, "learning_rate": 1.450361001366496e-06, "loss": 0.0204, "step": 74160 }, { "grad_norm": 0.2038348764181137, "learning_rate": 1.4454231594809021e-06, "loss": 0.0163, "step": 74170 }, { "grad_norm": 0.19370394945144653, "learning_rate": 1.4404936142445036e-06, "loss": 0.0155, "step": 74180 }, { "grad_norm": 0.1556333601474762, "learning_rate": 1.4355723664996546e-06, "loss": 0.0125, "step": 74190 }, { "grad_norm": 0.29422661662101746, "learning_rate": 1.4306594170872433e-06, "loss": 0.0126, "step": 74200 }, { "grad_norm": 0.3571937382221222, "learning_rate": 1.4257547668467598e-06, "loss": 0.0237, "step": 74210 }, { "grad_norm": 0.1762944608926773, "learning_rate": 1.4208584166162886e-06, "loss": 0.018, "step": 74220 }, { "grad_norm": 0.2616738975048065, "learning_rate": 1.4159703672324553e-06, "loss": 0.0202, "step": 74230 }, { "grad_norm": 0.24904708564281464, "learning_rate": 1.4110906195305252e-06, "loss": 0.0149, "step": 74240 }, { "grad_norm": 0.28774598240852356, "learning_rate": 1.406219174344292e-06, "loss": 0.0137, "step": 74250 }, { "grad_norm": 0.3194971978664398, "learning_rate": 1.401356032506157e-06, "loss": 0.0148, "step": 74260 }, { "grad_norm": 0.2754138708114624, "learning_rate": 1.3965011948471051e-06, "loss": 0.0165, "step": 74270 }, { "grad_norm": 0.223277285695076, "learning_rate": 1.3916546621966842e-06, "loss": 0.0143, "step": 74280 }, { "grad_norm": 0.14382320642471313, "learning_rate": 1.3868164353830482e-06, "loss": 0.0114, "step": 74290 }, { "grad_norm": 0.21194857358932495, "learning_rate": 1.3819865152329082e-06, "loss": 0.0087, "step": 74300 }, { "grad_norm": 0.4523277282714844, "learning_rate": 1.377164902571565e-06, "loss": 0.0236, "step": 74310 }, { "grad_norm": 0.2050638645887375, "learning_rate": 1.372351598222904e-06, "loss": 0.0144, "step": 74320 }, { "grad_norm": 0.39542901515960693, "learning_rate": 1.3675466030093843e-06, "loss": 0.0147, "step": 74330 }, { "grad_norm": 0.16156616806983948, "learning_rate": 1.3627499177520486e-06, "loss": 0.0152, "step": 74340 }, { "grad_norm": 0.13202528655529022, "learning_rate": 1.3579615432705196e-06, "loss": 0.0122, "step": 74350 }, { "grad_norm": 0.1389203667640686, "learning_rate": 1.3531814803829978e-06, "loss": 0.0185, "step": 74360 }, { "grad_norm": 0.17572703957557678, "learning_rate": 1.3484097299062803e-06, "loss": 0.0146, "step": 74370 }, { "grad_norm": 0.19599010050296783, "learning_rate": 1.343646292655698e-06, "loss": 0.015, "step": 74380 }, { "grad_norm": 0.5223817825317383, "learning_rate": 1.3388911694452277e-06, "loss": 0.0155, "step": 74390 }, { "grad_norm": 0.3901529610157013, "learning_rate": 1.3341443610873638e-06, "loss": 0.0198, "step": 74400 }, { "grad_norm": 0.2309701293706894, "learning_rate": 1.329405868393213e-06, "loss": 0.0155, "step": 74410 }, { "grad_norm": 0.21579045057296753, "learning_rate": 1.3246756921724613e-06, "loss": 0.019, "step": 74420 }, { "grad_norm": 0.2910512387752533, "learning_rate": 1.3199538332333506e-06, "loss": 0.0153, "step": 74430 }, { "grad_norm": 0.28503111004829407, "learning_rate": 1.3152402923827411e-06, "loss": 0.018, "step": 74440 }, { "grad_norm": 0.22896528244018555, "learning_rate": 1.3105350704260277e-06, "loss": 0.0143, "step": 74450 }, { "grad_norm": 0.26303112506866455, "learning_rate": 1.305838168167206e-06, "loss": 0.0101, "step": 74460 }, { "grad_norm": 0.1280703991651535, "learning_rate": 1.301149586408862e-06, "loss": 0.011, "step": 74470 }, { "grad_norm": 0.21511469781398773, "learning_rate": 1.2964693259521322e-06, "loss": 0.019, "step": 74480 }, { "grad_norm": 0.19460588693618774, "learning_rate": 1.2917973875967548e-06, "loss": 0.0146, "step": 74490 }, { "grad_norm": 0.4201350510120392, "learning_rate": 1.2871337721410249e-06, "loss": 0.02, "step": 74500 }, { "grad_norm": 0.1647162139415741, "learning_rate": 1.2824784803818379e-06, "loss": 0.0181, "step": 74510 }, { "grad_norm": 0.33573004603385925, "learning_rate": 1.2778315131146467e-06, "loss": 0.0095, "step": 74520 }, { "grad_norm": 0.41901329159736633, "learning_rate": 1.2731928711334994e-06, "loss": 0.0113, "step": 74530 }, { "grad_norm": 0.24654985964298248, "learning_rate": 1.2685625552310122e-06, "loss": 0.0127, "step": 74540 }, { "grad_norm": 0.18944689631462097, "learning_rate": 1.263940566198374e-06, "loss": 0.0141, "step": 74550 }, { "grad_norm": 0.1498774290084839, "learning_rate": 1.259326904825353e-06, "loss": 0.0138, "step": 74560 }, { "grad_norm": 0.20520083606243134, "learning_rate": 1.2547215719003137e-06, "loss": 0.0172, "step": 74570 }, { "grad_norm": 0.13469921052455902, "learning_rate": 1.2501245682101703e-06, "loss": 0.0143, "step": 74580 }, { "grad_norm": 0.40866705775260925, "learning_rate": 1.2455358945404171e-06, "loss": 0.0142, "step": 74590 }, { "grad_norm": 0.18245939910411835, "learning_rate": 1.2409555516751493e-06, "loss": 0.0104, "step": 74600 }, { "grad_norm": 0.21886025369167328, "learning_rate": 1.2363835403970125e-06, "loss": 0.0151, "step": 74610 }, { "grad_norm": 0.1411195993423462, "learning_rate": 1.2318198614872489e-06, "loss": 0.0125, "step": 74620 }, { "grad_norm": 0.3768377900123596, "learning_rate": 1.2272645157256457e-06, "loss": 0.0137, "step": 74630 }, { "grad_norm": 0.32721656560897827, "learning_rate": 1.222717503890608e-06, "loss": 0.0191, "step": 74640 }, { "grad_norm": 0.36104917526245117, "learning_rate": 1.218178826759081e-06, "loss": 0.0198, "step": 74650 }, { "grad_norm": 0.4639359712600708, "learning_rate": 1.213648485106611e-06, "loss": 0.0149, "step": 74660 }, { "grad_norm": 0.22746916115283966, "learning_rate": 1.2091264797073066e-06, "loss": 0.0257, "step": 74670 }, { "grad_norm": 0.2634337246417999, "learning_rate": 1.2046128113338494e-06, "loss": 0.0268, "step": 74680 }, { "grad_norm": 0.19621779024600983, "learning_rate": 1.200107480757512e-06, "loss": 0.0173, "step": 74690 }, { "grad_norm": 0.21957021951675415, "learning_rate": 1.1956104887481168e-06, "loss": 0.0193, "step": 74700 }, { "grad_norm": 0.23992440104484558, "learning_rate": 1.191121836074094e-06, "loss": 0.0104, "step": 74710 }, { "grad_norm": 0.320547878742218, "learning_rate": 1.1866415235024186e-06, "loss": 0.0097, "step": 74720 }, { "grad_norm": 0.287335067987442, "learning_rate": 1.182169551798662e-06, "loss": 0.0166, "step": 74730 }, { "grad_norm": 0.1842448115348816, "learning_rate": 1.177705921726957e-06, "loss": 0.011, "step": 74740 }, { "grad_norm": 0.10287800431251526, "learning_rate": 1.173250634050016e-06, "loss": 0.0143, "step": 74750 }, { "grad_norm": 0.32667264342308044, "learning_rate": 1.1688036895291354e-06, "loss": 0.0139, "step": 74760 }, { "grad_norm": 0.21861502528190613, "learning_rate": 1.1643650889241574e-06, "loss": 0.014, "step": 74770 }, { "grad_norm": 0.24794669449329376, "learning_rate": 1.1599348329935311e-06, "loss": 0.0149, "step": 74780 }, { "grad_norm": 0.212676540017128, "learning_rate": 1.1555129224942673e-06, "loss": 0.0243, "step": 74790 }, { "grad_norm": 0.193558469414711, "learning_rate": 1.1510993581819396e-06, "loss": 0.0164, "step": 74800 }, { "grad_norm": 0.14567436277866364, "learning_rate": 1.1466941408107112e-06, "loss": 0.0195, "step": 74810 }, { "grad_norm": 0.20757494866847992, "learning_rate": 1.1422972711333247e-06, "loss": 0.0153, "step": 74820 }, { "grad_norm": 0.45840004086494446, "learning_rate": 1.1379087499010565e-06, "loss": 0.0133, "step": 74830 }, { "grad_norm": 0.1482359915971756, "learning_rate": 1.1335285778638127e-06, "loss": 0.0113, "step": 74840 }, { "grad_norm": 0.23751361668109894, "learning_rate": 1.1291567557700333e-06, "loss": 0.0132, "step": 74850 }, { "grad_norm": 0.23598246276378632, "learning_rate": 1.124793284366743e-06, "loss": 0.0135, "step": 74860 }, { "grad_norm": 0.24828989803791046, "learning_rate": 1.1204381643995455e-06, "loss": 0.0128, "step": 74870 }, { "grad_norm": 0.15998463332653046, "learning_rate": 1.116091396612595e-06, "loss": 0.0127, "step": 74880 }, { "grad_norm": 0.23483906686306, "learning_rate": 1.1117529817486538e-06, "loss": 0.0138, "step": 74890 }, { "grad_norm": 0.2008865773677826, "learning_rate": 1.107422920549034e-06, "loss": 0.011, "step": 74900 }, { "grad_norm": 0.243455171585083, "learning_rate": 1.1031012137536157e-06, "loss": 0.0131, "step": 74910 }, { "grad_norm": 0.145908385515213, "learning_rate": 1.0987878621008695e-06, "loss": 0.0127, "step": 74920 }, { "grad_norm": 0.32682520151138306, "learning_rate": 1.0944828663278284e-06, "loss": 0.0184, "step": 74930 }, { "grad_norm": 0.149260476231575, "learning_rate": 1.0901862271700925e-06, "loss": 0.0149, "step": 74940 }, { "grad_norm": 0.12733492255210876, "learning_rate": 1.0858979453618467e-06, "loss": 0.0106, "step": 74950 }, { "grad_norm": 0.36714276671409607, "learning_rate": 1.0816180216358441e-06, "loss": 0.0158, "step": 74960 }, { "grad_norm": 0.2278548926115036, "learning_rate": 1.0773464567233937e-06, "loss": 0.0126, "step": 74970 }, { "grad_norm": 0.21863752603530884, "learning_rate": 1.0730832513543953e-06, "loss": 0.0129, "step": 74980 }, { "grad_norm": 0.19645123183727264, "learning_rate": 1.0688284062573218e-06, "loss": 0.0167, "step": 74990 }, { "grad_norm": 0.23733431100845337, "learning_rate": 1.0645819221591969e-06, "loss": 0.0114, "step": 75000 }, { "grad_norm": 0.34292569756507874, "learning_rate": 1.0603437997856347e-06, "loss": 0.0144, "step": 75010 }, { "grad_norm": 0.19803467392921448, "learning_rate": 1.0561140398608228e-06, "loss": 0.0124, "step": 75020 }, { "grad_norm": 0.2379046231508255, "learning_rate": 1.0518926431074937e-06, "loss": 0.0137, "step": 75030 }, { "grad_norm": 0.21201744675636292, "learning_rate": 1.0476796102469877e-06, "loss": 0.0149, "step": 75040 }, { "grad_norm": 0.18401162326335907, "learning_rate": 1.0434749419991786e-06, "loss": 0.0141, "step": 75050 }, { "grad_norm": 0.17463982105255127, "learning_rate": 1.0392786390825415e-06, "loss": 0.0134, "step": 75060 }, { "grad_norm": 0.33579257130622864, "learning_rate": 1.0350907022141087e-06, "loss": 0.014, "step": 75070 }, { "grad_norm": 0.2944130003452301, "learning_rate": 1.0309111321094744e-06, "loss": 0.0131, "step": 75080 }, { "grad_norm": 0.15132209658622742, "learning_rate": 1.026739929482834e-06, "loss": 0.011, "step": 75090 }, { "grad_norm": 0.2816718518733978, "learning_rate": 1.022577095046906e-06, "loss": 0.0158, "step": 75100 }, { "grad_norm": 0.17615486681461334, "learning_rate": 1.0184226295130217e-06, "loss": 0.0146, "step": 75110 }, { "grad_norm": 0.22629204392433167, "learning_rate": 1.0142765335910575e-06, "loss": 0.0135, "step": 75120 }, { "grad_norm": 0.27548685669898987, "learning_rate": 1.0101388079894746e-06, "loss": 0.0111, "step": 75130 }, { "grad_norm": 0.17480020225048065, "learning_rate": 1.0060094534152908e-06, "loss": 0.0158, "step": 75140 }, { "grad_norm": 0.11893580108880997, "learning_rate": 1.0018884705741028e-06, "loss": 0.0118, "step": 75150 }, { "grad_norm": 0.20087392628192902, "learning_rate": 9.977758601700805e-07, "loss": 0.0174, "step": 75160 }, { "grad_norm": 0.1961119920015335, "learning_rate": 9.93671622905945e-07, "loss": 0.0225, "step": 75170 }, { "grad_norm": 0.54056715965271, "learning_rate": 9.895757594830024e-07, "loss": 0.0127, "step": 75180 }, { "grad_norm": 0.15910425782203674, "learning_rate": 9.854882706011204e-07, "loss": 0.0118, "step": 75190 }, { "grad_norm": 0.43138396739959717, "learning_rate": 9.814091569587513e-07, "loss": 0.0117, "step": 75200 }, { "grad_norm": 0.41257718205451965, "learning_rate": 9.77338419252888e-07, "loss": 0.022, "step": 75210 }, { "grad_norm": 0.2487783282995224, "learning_rate": 9.732760581791289e-07, "loss": 0.0178, "step": 75220 }, { "grad_norm": 0.2411259263753891, "learning_rate": 9.692220744315972e-07, "loss": 0.0129, "step": 75230 }, { "grad_norm": 0.1878194957971573, "learning_rate": 9.651764687030162e-07, "loss": 0.0123, "step": 75240 }, { "grad_norm": 0.17653970420360565, "learning_rate": 9.611392416846776e-07, "loss": 0.0143, "step": 75250 }, { "grad_norm": 0.2593761384487152, "learning_rate": 9.571103940664238e-07, "loss": 0.0136, "step": 75260 }, { "grad_norm": 0.2503906488418579, "learning_rate": 9.530899265366822e-07, "loss": 0.0177, "step": 75270 }, { "grad_norm": 0.15620936453342438, "learning_rate": 9.490778397824307e-07, "loss": 0.0157, "step": 75280 }, { "grad_norm": 0.20480658113956451, "learning_rate": 9.450741344892322e-07, "loss": 0.0156, "step": 75290 }, { "grad_norm": 0.33012136816978455, "learning_rate": 9.410788113412116e-07, "loss": 0.0148, "step": 75300 }, { "grad_norm": 0.2097007781267166, "learning_rate": 9.370918710210563e-07, "loss": 0.012, "step": 75310 }, { "grad_norm": 0.24309787154197693, "learning_rate": 9.331133142100268e-07, "loss": 0.0152, "step": 75320 }, { "grad_norm": 0.32405155897140503, "learning_rate": 9.291431415879459e-07, "loss": 0.01, "step": 75330 }, { "grad_norm": 0.2715691328048706, "learning_rate": 9.251813538332155e-07, "loss": 0.0165, "step": 75340 }, { "grad_norm": 0.12458503246307373, "learning_rate": 9.212279516227884e-07, "loss": 0.0112, "step": 75350 }, { "grad_norm": 0.27908453345298767, "learning_rate": 9.172829356322021e-07, "loss": 0.0124, "step": 75360 }, { "grad_norm": 0.18541091680526733, "learning_rate": 9.133463065355452e-07, "loss": 0.0153, "step": 75370 }, { "grad_norm": 0.38411810994148254, "learning_rate": 9.094180650054796e-07, "loss": 0.0212, "step": 75380 }, { "grad_norm": 0.36620450019836426, "learning_rate": 9.054982117132404e-07, "loss": 0.0108, "step": 75390 }, { "grad_norm": 0.35347115993499756, "learning_rate": 9.015867473286143e-07, "loss": 0.0158, "step": 75400 }, { "grad_norm": 0.2659150958061218, "learning_rate": 8.976836725199778e-07, "loss": 0.0117, "step": 75410 }, { "grad_norm": 0.278532475233078, "learning_rate": 8.937889879542416e-07, "loss": 0.0142, "step": 75420 }, { "grad_norm": 0.2836959660053253, "learning_rate": 8.899026942969068e-07, "loss": 0.0138, "step": 75430 }, { "grad_norm": 0.15162304043769836, "learning_rate": 8.860247922120424e-07, "loss": 0.0131, "step": 75440 }, { "grad_norm": 0.22837303578853607, "learning_rate": 8.821552823622737e-07, "loss": 0.0125, "step": 75450 }, { "grad_norm": 0.16915397346019745, "learning_rate": 8.782941654087884e-07, "loss": 0.0102, "step": 75460 }, { "grad_norm": 0.49761343002319336, "learning_rate": 8.744414420113478e-07, "loss": 0.0173, "step": 75470 }, { "grad_norm": 0.22498486936092377, "learning_rate": 8.705971128282753e-07, "loss": 0.0126, "step": 75480 }, { "grad_norm": 0.23627914488315582, "learning_rate": 8.66761178516473e-07, "loss": 0.0117, "step": 75490 }, { "grad_norm": 0.22187648713588715, "learning_rate": 8.629336397313781e-07, "loss": 0.014, "step": 75500 }, { "grad_norm": 0.1343381702899933, "learning_rate": 8.591144971270282e-07, "loss": 0.0153, "step": 75510 }, { "grad_norm": 0.15101133286952972, "learning_rate": 8.553037513560069e-07, "loss": 0.0165, "step": 75520 }, { "grad_norm": 0.3071613013744354, "learning_rate": 8.515014030694546e-07, "loss": 0.0116, "step": 75530 }, { "grad_norm": 0.2054980993270874, "learning_rate": 8.47707452917107e-07, "loss": 0.0116, "step": 75540 }, { "grad_norm": 0.290997713804245, "learning_rate": 8.439219015472343e-07, "loss": 0.0159, "step": 75550 }, { "grad_norm": 0.2226158231496811, "learning_rate": 8.401447496066861e-07, "loss": 0.0133, "step": 75560 }, { "grad_norm": 0.3855540156364441, "learning_rate": 8.363759977408792e-07, "loss": 0.0134, "step": 75570 }, { "grad_norm": 0.4022369086742401, "learning_rate": 8.326156465937817e-07, "loss": 0.0126, "step": 75580 }, { "grad_norm": 0.40016159415245056, "learning_rate": 8.28863696807941e-07, "loss": 0.0245, "step": 75590 }, { "grad_norm": 0.12271151691675186, "learning_rate": 8.251201490244609e-07, "loss": 0.0168, "step": 75600 }, { "grad_norm": 0.37939712405204773, "learning_rate": 8.213850038830129e-07, "loss": 0.0131, "step": 75610 }, { "grad_norm": 0.29352590441703796, "learning_rate": 8.176582620218309e-07, "loss": 0.0113, "step": 75620 }, { "grad_norm": 0.21483242511749268, "learning_rate": 8.139399240777057e-07, "loss": 0.011, "step": 75630 }, { "grad_norm": 0.2800740599632263, "learning_rate": 8.102299906860122e-07, "loss": 0.0162, "step": 75640 }, { "grad_norm": 0.3833867311477661, "learning_rate": 8.065284624806657e-07, "loss": 0.0197, "step": 75650 }, { "grad_norm": 0.21739964187145233, "learning_rate": 8.028353400941601e-07, "loss": 0.0135, "step": 75660 }, { "grad_norm": 0.13892774283885956, "learning_rate": 7.991506241575575e-07, "loss": 0.0176, "step": 75670 }, { "grad_norm": 0.27817749977111816, "learning_rate": 7.954743153004541e-07, "loss": 0.015, "step": 75680 }, { "grad_norm": 0.20535744726657867, "learning_rate": 7.918064141510528e-07, "loss": 0.0127, "step": 75690 }, { "grad_norm": 0.25609514117240906, "learning_rate": 7.881469213360859e-07, "loss": 0.0196, "step": 75700 }, { "grad_norm": 0.19173189997673035, "learning_rate": 7.844958374808642e-07, "loss": 0.0109, "step": 75710 }, { "grad_norm": 0.21053215861320496, "learning_rate": 7.808531632092608e-07, "loss": 0.0113, "step": 75720 }, { "grad_norm": 0.21751989424228668, "learning_rate": 7.772188991436946e-07, "loss": 0.0165, "step": 75730 }, { "grad_norm": 0.20911948382854462, "learning_rate": 7.735930459051799e-07, "loss": 0.0158, "step": 75740 }, { "grad_norm": 0.2875312864780426, "learning_rate": 7.699756041132655e-07, "loss": 0.0143, "step": 75750 }, { "grad_norm": 0.2499992847442627, "learning_rate": 7.663665743860793e-07, "loss": 0.0106, "step": 75760 }, { "grad_norm": 0.14294949173927307, "learning_rate": 7.627659573403001e-07, "loss": 0.0103, "step": 75770 }, { "grad_norm": 0.18369702994823456, "learning_rate": 7.591737535911802e-07, "loss": 0.0135, "step": 75780 }, { "grad_norm": 0.12219326198101044, "learning_rate": 7.555899637525288e-07, "loss": 0.0128, "step": 75790 }, { "grad_norm": 0.2210167497396469, "learning_rate": 7.520145884367058e-07, "loss": 0.0105, "step": 75800 }, { "grad_norm": 0.2180289477109909, "learning_rate": 7.484476282546615e-07, "loss": 0.0172, "step": 75810 }, { "grad_norm": 0.2581161558628082, "learning_rate": 7.448890838158806e-07, "loss": 0.0177, "step": 75820 }, { "grad_norm": 0.15125501155853271, "learning_rate": 7.413389557284267e-07, "loss": 0.0141, "step": 75830 }, { "grad_norm": 0.21755456924438477, "learning_rate": 7.377972445989201e-07, "loss": 0.0174, "step": 75840 }, { "grad_norm": 0.2931934595108032, "learning_rate": 7.342639510325377e-07, "loss": 0.0172, "step": 75850 }, { "grad_norm": 0.34880584478378296, "learning_rate": 7.307390756330246e-07, "loss": 0.0129, "step": 75860 }, { "grad_norm": 0.3224204182624817, "learning_rate": 7.272226190026876e-07, "loss": 0.0116, "step": 75870 }, { "grad_norm": 0.15443925559520721, "learning_rate": 7.237145817423907e-07, "loss": 0.0128, "step": 75880 }, { "grad_norm": 0.10249390453100204, "learning_rate": 7.202149644515654e-07, "loss": 0.0125, "step": 75890 }, { "grad_norm": 0.25586026906967163, "learning_rate": 7.167237677281946e-07, "loss": 0.0158, "step": 75900 }, { "grad_norm": 0.3714282810688019, "learning_rate": 7.132409921688288e-07, "loss": 0.0136, "step": 75910 }, { "grad_norm": 0.2067984789609909, "learning_rate": 7.09766638368592e-07, "loss": 0.0132, "step": 75920 }, { "grad_norm": 0.26291948556900024, "learning_rate": 7.063007069211313e-07, "loss": 0.0106, "step": 75930 }, { "grad_norm": 0.34479159116744995, "learning_rate": 7.028431984187067e-07, "loss": 0.0154, "step": 75940 }, { "grad_norm": 0.23069845139980316, "learning_rate": 6.993941134520898e-07, "loss": 0.0126, "step": 75950 }, { "grad_norm": 0.17351821064949036, "learning_rate": 6.959534526106537e-07, "loss": 0.0146, "step": 75960 }, { "grad_norm": 0.3491973876953125, "learning_rate": 6.925212164822948e-07, "loss": 0.0136, "step": 75970 }, { "grad_norm": 0.2681771516799927, "learning_rate": 6.890974056535049e-07, "loss": 0.0137, "step": 75980 }, { "grad_norm": 0.19850052893161774, "learning_rate": 6.856820207093106e-07, "loss": 0.0148, "step": 75990 }, { "grad_norm": 0.12353724986314774, "learning_rate": 6.822750622333063e-07, "loss": 0.0113, "step": 76000 }, { "grad_norm": 0.27445462346076965, "learning_rate": 6.788765308076539e-07, "loss": 0.0175, "step": 76010 }, { "grad_norm": 0.21087948977947235, "learning_rate": 6.754864270130668e-07, "loss": 0.0099, "step": 76020 }, { "grad_norm": 0.2481634020805359, "learning_rate": 6.721047514288203e-07, "loss": 0.0123, "step": 76030 }, { "grad_norm": 0.3164518475532532, "learning_rate": 6.687315046327469e-07, "loss": 0.0172, "step": 76040 }, { "grad_norm": 0.3539390563964844, "learning_rate": 6.65366687201252e-07, "loss": 0.0141, "step": 76050 }, { "grad_norm": 0.2657627463340759, "learning_rate": 6.620102997092814e-07, "loss": 0.022, "step": 76060 }, { "grad_norm": 0.35728394985198975, "learning_rate": 6.586623427303596e-07, "loss": 0.0122, "step": 76070 }, { "grad_norm": 0.15855805575847626, "learning_rate": 6.553228168365455e-07, "loss": 0.0161, "step": 76080 }, { "grad_norm": 0.38112878799438477, "learning_rate": 6.519917225984884e-07, "loss": 0.0184, "step": 76090 }, { "grad_norm": 0.3065507113933563, "learning_rate": 6.486690605853718e-07, "loss": 0.017, "step": 76100 }, { "grad_norm": 0.4314275085926056, "learning_rate": 6.453548313649527e-07, "loss": 0.016, "step": 76110 }, { "grad_norm": 0.38915789127349854, "learning_rate": 6.420490355035446e-07, "loss": 0.0194, "step": 76120 }, { "grad_norm": 0.21805939078330994, "learning_rate": 6.387516735660071e-07, "loss": 0.0201, "step": 76130 }, { "grad_norm": 0.19053353369235992, "learning_rate": 6.35462746115778e-07, "loss": 0.0133, "step": 76140 }, { "grad_norm": 0.3351880609989166, "learning_rate": 6.321822537148358e-07, "loss": 0.0185, "step": 76150 }, { "grad_norm": 0.17846448719501495, "learning_rate": 6.289101969237432e-07, "loss": 0.0141, "step": 76160 }, { "grad_norm": 0.23906628787517548, "learning_rate": 6.256465763015918e-07, "loss": 0.0148, "step": 76170 }, { "grad_norm": 0.284351110458374, "learning_rate": 6.22391392406052e-07, "loss": 0.0115, "step": 76180 }, { "grad_norm": 0.17567101120948792, "learning_rate": 6.191446457933403e-07, "loss": 0.0121, "step": 76190 }, { "grad_norm": 0.16839657723903656, "learning_rate": 6.159063370182406e-07, "loss": 0.012, "step": 76200 }, { "grad_norm": 0.1971758008003235, "learning_rate": 6.126764666340879e-07, "loss": 0.0116, "step": 76210 }, { "grad_norm": 0.1057681292295456, "learning_rate": 6.094550351927852e-07, "loss": 0.0147, "step": 76220 }, { "grad_norm": 0.27334150671958923, "learning_rate": 6.062420432447757e-07, "loss": 0.0167, "step": 76230 }, { "grad_norm": 0.24329018592834473, "learning_rate": 6.030374913390813e-07, "loss": 0.0127, "step": 76240 }, { "grad_norm": 0.2626296579837799, "learning_rate": 5.998413800232694e-07, "loss": 0.0148, "step": 76250 }, { "grad_norm": 0.22365683317184448, "learning_rate": 5.966537098434755e-07, "loss": 0.0144, "step": 76260 }, { "grad_norm": 0.14670976996421814, "learning_rate": 5.934744813443694e-07, "loss": 0.0116, "step": 76270 }, { "grad_norm": 0.18872977793216705, "learning_rate": 5.903036950692054e-07, "loss": 0.0173, "step": 76280 }, { "grad_norm": 0.15415292978286743, "learning_rate": 5.871413515597835e-07, "loss": 0.0156, "step": 76290 }, { "grad_norm": 0.19389045238494873, "learning_rate": 5.839874513564547e-07, "loss": 0.0104, "step": 76300 }, { "grad_norm": 0.22360914945602417, "learning_rate": 5.808419949981436e-07, "loss": 0.0163, "step": 76310 }, { "grad_norm": 0.11629445850849152, "learning_rate": 5.777049830223257e-07, "loss": 0.0141, "step": 76320 }, { "grad_norm": 0.12040002644062042, "learning_rate": 5.745764159650114e-07, "loss": 0.0112, "step": 76330 }, { "grad_norm": 0.18066024780273438, "learning_rate": 5.71456294360806e-07, "loss": 0.0115, "step": 76340 }, { "grad_norm": 0.2194063812494278, "learning_rate": 5.683446187428443e-07, "loss": 0.0194, "step": 76350 }, { "grad_norm": 0.23403869569301605, "learning_rate": 5.652413896428288e-07, "loss": 0.0185, "step": 76360 }, { "grad_norm": 0.3219810426235199, "learning_rate": 5.621466075910131e-07, "loss": 0.0161, "step": 76370 }, { "grad_norm": 0.1138489618897438, "learning_rate": 5.590602731162187e-07, "loss": 0.0126, "step": 76380 }, { "grad_norm": 0.37802156805992126, "learning_rate": 5.559823867458125e-07, "loss": 0.0143, "step": 76390 }, { "grad_norm": 0.15022693574428558, "learning_rate": 5.529129490057128e-07, "loss": 0.0182, "step": 76400 }, { "grad_norm": 0.26283279061317444, "learning_rate": 5.498519604204167e-07, "loss": 0.0137, "step": 76410 }, { "grad_norm": 0.24537847936153412, "learning_rate": 5.467994215129501e-07, "loss": 0.0128, "step": 76420 }, { "grad_norm": 0.1297939568758011, "learning_rate": 5.437553328049183e-07, "loss": 0.0114, "step": 76430 }, { "grad_norm": 0.2774641811847687, "learning_rate": 5.407196948164661e-07, "loss": 0.0119, "step": 76440 }, { "grad_norm": 0.21860790252685547, "learning_rate": 5.376925080663009e-07, "loss": 0.0138, "step": 76450 }, { "grad_norm": 0.1760302037000656, "learning_rate": 5.346737730716977e-07, "loss": 0.0156, "step": 76460 }, { "grad_norm": 0.2674236297607422, "learning_rate": 5.316634903484607e-07, "loss": 0.0114, "step": 76470 }, { "grad_norm": 0.23773586750030518, "learning_rate": 5.286616604109728e-07, "loss": 0.012, "step": 76480 }, { "grad_norm": 0.26363447308540344, "learning_rate": 5.256682837721627e-07, "loss": 0.0195, "step": 76490 }, { "grad_norm": 0.20064398646354675, "learning_rate": 5.226833609435156e-07, "loss": 0.0113, "step": 76500 }, { "grad_norm": 0.18684346973896027, "learning_rate": 5.19706892435079e-07, "loss": 0.0153, "step": 76510 }, { "grad_norm": 0.20143046975135803, "learning_rate": 5.167388787554406e-07, "loss": 0.0155, "step": 76520 }, { "grad_norm": 0.15489642322063446, "learning_rate": 5.137793204117614e-07, "loss": 0.0134, "step": 76530 }, { "grad_norm": 0.21628277003765106, "learning_rate": 5.108282179097479e-07, "loss": 0.0144, "step": 76540 }, { "grad_norm": 0.21999016404151917, "learning_rate": 5.078855717536523e-07, "loss": 0.0108, "step": 76550 }, { "grad_norm": 0.2367987036705017, "learning_rate": 5.049513824463059e-07, "loss": 0.0114, "step": 76560 }, { "grad_norm": 0.24261808395385742, "learning_rate": 5.020256504890742e-07, "loss": 0.011, "step": 76570 }, { "grad_norm": 0.1371420919895172, "learning_rate": 4.991083763818849e-07, "loss": 0.017, "step": 76580 }, { "grad_norm": 0.26250404119491577, "learning_rate": 4.961995606232228e-07, "loss": 0.0172, "step": 76590 }, { "grad_norm": 0.26606285572052, "learning_rate": 4.932992037101236e-07, "loss": 0.0121, "step": 76600 }, { "grad_norm": 0.34475207328796387, "learning_rate": 4.904073061381798e-07, "loss": 0.0172, "step": 76610 }, { "grad_norm": 0.201701819896698, "learning_rate": 4.875238684015349e-07, "loss": 0.0165, "step": 76620 }, { "grad_norm": 0.1613646298646927, "learning_rate": 4.846488909928948e-07, "loss": 0.0125, "step": 76630 }, { "grad_norm": 0.2368830144405365, "learning_rate": 4.81782374403511e-07, "loss": 0.0233, "step": 76640 }, { "grad_norm": 0.18550635874271393, "learning_rate": 4.789243191231918e-07, "loss": 0.0127, "step": 76650 }, { "grad_norm": 0.18780896067619324, "learning_rate": 4.76074725640302e-07, "loss": 0.0166, "step": 76660 }, { "grad_norm": 0.45358750224113464, "learning_rate": 4.732335944417632e-07, "loss": 0.0205, "step": 76670 }, { "grad_norm": 0.5316731333732605, "learning_rate": 4.704009260130371e-07, "loss": 0.0159, "step": 76680 }, { "grad_norm": 0.09114643186330795, "learning_rate": 4.675767208381587e-07, "loss": 0.0173, "step": 76690 }, { "grad_norm": 0.2574155032634735, "learning_rate": 4.647609793997032e-07, "loss": 0.0135, "step": 76700 }, { "grad_norm": 0.18201550841331482, "learning_rate": 4.619537021788023e-07, "loss": 0.0141, "step": 76710 }, { "grad_norm": 0.2659475803375244, "learning_rate": 4.5915488965515005e-07, "loss": 0.0152, "step": 76720 }, { "grad_norm": 0.36819952726364136, "learning_rate": 4.563645423069807e-07, "loss": 0.0158, "step": 76730 }, { "grad_norm": 0.3879263401031494, "learning_rate": 4.535826606110849e-07, "loss": 0.0127, "step": 76740 }, { "grad_norm": 0.3827848732471466, "learning_rate": 4.508092450428214e-07, "loss": 0.0162, "step": 76750 }, { "grad_norm": 0.19153320789337158, "learning_rate": 4.480442960760778e-07, "loss": 0.0186, "step": 76760 }, { "grad_norm": 0.20625652372837067, "learning_rate": 4.4528781418332053e-07, "loss": 0.0121, "step": 76770 }, { "grad_norm": 0.23178452253341675, "learning_rate": 4.4253979983555073e-07, "loss": 0.0201, "step": 76780 }, { "grad_norm": 0.10894288867712021, "learning_rate": 4.3980025350233154e-07, "loss": 0.0114, "step": 76790 }, { "grad_norm": 0.4191039204597473, "learning_rate": 4.370691756517664e-07, "loss": 0.0145, "step": 76800 }, { "grad_norm": 0.20169708132743835, "learning_rate": 4.3434656675053753e-07, "loss": 0.0116, "step": 76810 }, { "grad_norm": 0.4070451855659485, "learning_rate": 4.3163242726385613e-07, "loss": 0.0135, "step": 76820 }, { "grad_norm": 0.3026162087917328, "learning_rate": 4.289267576554956e-07, "loss": 0.0173, "step": 76830 }, { "grad_norm": 0.26415491104125977, "learning_rate": 4.262295583877807e-07, "loss": 0.0113, "step": 76840 }, { "grad_norm": 0.3532002866268158, "learning_rate": 4.235408299215815e-07, "loss": 0.0178, "step": 76850 }, { "grad_norm": 0.24566981196403503, "learning_rate": 4.2086057271634173e-07, "loss": 0.0157, "step": 76860 }, { "grad_norm": 0.3421016335487366, "learning_rate": 4.181887872300394e-07, "loss": 0.0149, "step": 76870 }, { "grad_norm": 0.22607466578483582, "learning_rate": 4.155254739192038e-07, "loss": 0.0121, "step": 76880 }, { "grad_norm": 0.27526190876960754, "learning_rate": 4.128706332389265e-07, "loss": 0.0228, "step": 76890 }, { "grad_norm": 0.2571883797645569, "learning_rate": 4.1022426564284453e-07, "loss": 0.0153, "step": 76900 }, { "grad_norm": 0.30595719814300537, "learning_rate": 4.075863715831574e-07, "loss": 0.0135, "step": 76910 }, { "grad_norm": 0.20717793703079224, "learning_rate": 4.0495695151059887e-07, "loss": 0.0196, "step": 76920 }, { "grad_norm": 0.3296600878238678, "learning_rate": 4.023360058744763e-07, "loss": 0.0113, "step": 76930 }, { "grad_norm": 0.20568984746932983, "learning_rate": 3.997235351226314e-07, "loss": 0.0112, "step": 76940 }, { "grad_norm": 0.1861262321472168, "learning_rate": 3.971195397014571e-07, "loss": 0.0168, "step": 76950 }, { "grad_norm": 0.30451807379722595, "learning_rate": 3.9452402005591417e-07, "loss": 0.0194, "step": 76960 }, { "grad_norm": 0.1799151748418808, "learning_rate": 3.9193697662950334e-07, "loss": 0.0121, "step": 76970 }, { "grad_norm": 0.29338908195495605, "learning_rate": 3.893584098642822e-07, "loss": 0.019, "step": 76980 }, { "grad_norm": 0.1358087658882141, "learning_rate": 3.867883202008538e-07, "loss": 0.017, "step": 76990 }, { "grad_norm": 0.3851732015609741, "learning_rate": 3.842267080783779e-07, "loss": 0.0141, "step": 77000 }, { "grad_norm": 0.15627825260162354, "learning_rate": 3.816735739345656e-07, "loss": 0.0156, "step": 77010 }, { "grad_norm": 0.3042284846305847, "learning_rate": 3.791289182056679e-07, "loss": 0.0142, "step": 77020 }, { "grad_norm": 0.23027180135250092, "learning_rate": 3.7659274132650913e-07, "loss": 0.0139, "step": 77030 }, { "grad_norm": 0.25870776176452637, "learning_rate": 3.7406504373044826e-07, "loss": 0.0124, "step": 77040 }, { "grad_norm": 0.33251944184303284, "learning_rate": 3.7154582584939533e-07, "loss": 0.0142, "step": 77050 }, { "grad_norm": 0.36711591482162476, "learning_rate": 3.690350881138227e-07, "loss": 0.0135, "step": 77060 }, { "grad_norm": 0.3328956663608551, "learning_rate": 3.665328309527427e-07, "loss": 0.015, "step": 77070 }, { "grad_norm": 0.3760884702205658, "learning_rate": 3.640390547937245e-07, "loss": 0.0143, "step": 77080 }, { "grad_norm": 0.18840226531028748, "learning_rate": 3.615537600628882e-07, "loss": 0.0169, "step": 77090 }, { "grad_norm": 0.5752468705177307, "learning_rate": 3.5907694718489427e-07, "loss": 0.0144, "step": 77100 }, { "grad_norm": 0.13025841116905212, "learning_rate": 3.566086165829707e-07, "loss": 0.015, "step": 77110 }, { "grad_norm": 0.4825440049171448, "learning_rate": 3.5414876867888024e-07, "loss": 0.0135, "step": 77120 }, { "grad_norm": 0.18009249866008759, "learning_rate": 3.5169740389295326e-07, "loss": 0.0213, "step": 77130 }, { "grad_norm": 0.34562233090400696, "learning_rate": 3.4925452264405466e-07, "loss": 0.0139, "step": 77140 }, { "grad_norm": 0.191012442111969, "learning_rate": 3.468201253496062e-07, "loss": 0.0121, "step": 77150 }, { "grad_norm": 0.34483158588409424, "learning_rate": 3.443942124255861e-07, "loss": 0.0153, "step": 77160 }, { "grad_norm": 0.259303480386734, "learning_rate": 3.4197678428650183e-07, "loss": 0.014, "step": 77170 }, { "grad_norm": 0.2581540048122406, "learning_rate": 3.3956784134544504e-07, "loss": 0.0125, "step": 77180 }, { "grad_norm": 0.2785334885120392, "learning_rate": 3.3716738401402547e-07, "loss": 0.01, "step": 77190 }, { "grad_norm": 0.4213665723800659, "learning_rate": 3.3477541270241495e-07, "loss": 0.0119, "step": 77200 }, { "grad_norm": 0.19521166384220123, "learning_rate": 3.3239192781934215e-07, "loss": 0.0127, "step": 77210 }, { "grad_norm": 0.2239711433649063, "learning_rate": 3.3001692977207563e-07, "loss": 0.0139, "step": 77220 }, { "grad_norm": 0.11686999350786209, "learning_rate": 3.276504189664409e-07, "loss": 0.0125, "step": 77230 }, { "grad_norm": 0.33186230063438416, "learning_rate": 3.252923958068088e-07, "loss": 0.013, "step": 77240 }, { "grad_norm": 0.23270170390605927, "learning_rate": 3.2294286069609046e-07, "loss": 0.0168, "step": 77250 }, { "grad_norm": 0.13307535648345947, "learning_rate": 3.2060181403577583e-07, "loss": 0.0141, "step": 77260 }, { "grad_norm": 0.25781458616256714, "learning_rate": 3.1826925622587823e-07, "loss": 0.0135, "step": 77270 }, { "grad_norm": 0.3406505286693573, "learning_rate": 3.159451876649622e-07, "loss": 0.015, "step": 77280 }, { "grad_norm": 0.35716503858566284, "learning_rate": 3.1362960875015444e-07, "loss": 0.0111, "step": 77290 }, { "grad_norm": 0.33818450570106506, "learning_rate": 3.1132251987711637e-07, "loss": 0.018, "step": 77300 }, { "grad_norm": 0.23427684605121613, "learning_rate": 3.0902392144007695e-07, "loss": 0.0153, "step": 77310 }, { "grad_norm": 0.2384510636329651, "learning_rate": 3.0673381383179436e-07, "loss": 0.0156, "step": 77320 }, { "grad_norm": 0.2750563621520996, "learning_rate": 3.0445219744358875e-07, "loss": 0.012, "step": 77330 }, { "grad_norm": 0.30225998163223267, "learning_rate": 3.021790726653262e-07, "loss": 0.0226, "step": 77340 }, { "grad_norm": 0.2570245563983917, "learning_rate": 2.9991443988542366e-07, "loss": 0.015, "step": 77350 }, { "grad_norm": 0.17362941801548004, "learning_rate": 2.976582994908439e-07, "loss": 0.0111, "step": 77360 }, { "grad_norm": 0.2539781332015991, "learning_rate": 2.954106518671007e-07, "loss": 0.0131, "step": 77370 }, { "grad_norm": 0.19415506720542908, "learning_rate": 2.9317149739825356e-07, "loss": 0.0136, "step": 77380 }, { "grad_norm": 0.5206913352012634, "learning_rate": 2.909408364669075e-07, "loss": 0.0214, "step": 77390 }, { "grad_norm": 0.31802937388420105, "learning_rate": 2.887186694542299e-07, "loss": 0.0196, "step": 77400 }, { "grad_norm": 0.1564435064792633, "learning_rate": 2.865049967399225e-07, "loss": 0.0122, "step": 77410 }, { "grad_norm": 0.16652537882328033, "learning_rate": 2.842998187022439e-07, "loss": 0.0146, "step": 77420 }, { "grad_norm": 0.3316512405872345, "learning_rate": 2.8210313571800374e-07, "loss": 0.0171, "step": 77430 }, { "grad_norm": 0.23142468929290771, "learning_rate": 2.7991494816255184e-07, "loss": 0.0118, "step": 77440 }, { "grad_norm": 0.2894766926765442, "learning_rate": 2.777352564097779e-07, "loss": 0.0131, "step": 77450 }, { "grad_norm": 0.23494978249073029, "learning_rate": 2.755640608321508e-07, "loss": 0.0137, "step": 77460 }, { "grad_norm": 0.4623352885246277, "learning_rate": 2.73401361800657e-07, "loss": 0.0146, "step": 77470 }, { "grad_norm": 0.29688307642936707, "learning_rate": 2.7124715968484537e-07, "loss": 0.0179, "step": 77480 }, { "grad_norm": 0.207910418510437, "learning_rate": 2.691014548528104e-07, "loss": 0.017, "step": 77490 }, { "grad_norm": 0.17205798625946045, "learning_rate": 2.669642476711864e-07, "loss": 0.0133, "step": 77500 }, { "grad_norm": 0.20043595135211945, "learning_rate": 2.648355385051815e-07, "loss": 0.014, "step": 77510 }, { "grad_norm": 0.21643957495689392, "learning_rate": 2.627153277185157e-07, "loss": 0.0144, "step": 77520 }, { "grad_norm": 0.23229767382144928, "learning_rate": 2.606036156734881e-07, "loss": 0.0119, "step": 77530 }, { "grad_norm": 0.16601133346557617, "learning_rate": 2.5850040273092127e-07, "loss": 0.0136, "step": 77540 }, { "grad_norm": 0.23250097036361694, "learning_rate": 2.5640568925020536e-07, "loss": 0.0176, "step": 77550 }, { "grad_norm": 0.15225829184055328, "learning_rate": 2.5431947558927083e-07, "loss": 0.0174, "step": 77560 }, { "grad_norm": 0.16033656895160675, "learning_rate": 2.522417621045825e-07, "loss": 0.0112, "step": 77570 }, { "grad_norm": 0.29171067476272583, "learning_rate": 2.501725491511786e-07, "loss": 0.0196, "step": 77580 }, { "grad_norm": 0.1989615261554718, "learning_rate": 2.481118370826263e-07, "loss": 0.0154, "step": 77590 }, { "grad_norm": 0.2705433964729309, "learning_rate": 2.4605962625104393e-07, "loss": 0.0167, "step": 77600 }, { "grad_norm": 0.22759075462818146, "learning_rate": 2.440159170070955e-07, "loss": 0.0179, "step": 77610 }, { "grad_norm": 0.27156099677085876, "learning_rate": 2.41980709699996e-07, "loss": 0.0139, "step": 77620 }, { "grad_norm": 0.21639670431613922, "learning_rate": 2.3995400467751174e-07, "loss": 0.0156, "step": 77630 }, { "grad_norm": 0.3866531550884247, "learning_rate": 2.3793580228594902e-07, "loss": 0.0194, "step": 77640 }, { "grad_norm": 0.23163184523582458, "learning_rate": 2.3592610287015982e-07, "loss": 0.0117, "step": 77650 }, { "grad_norm": 0.2287617027759552, "learning_rate": 2.3392490677354718e-07, "loss": 0.0138, "step": 77660 }, { "grad_norm": 0.2492421269416809, "learning_rate": 2.3193221433806533e-07, "loss": 0.0134, "step": 77670 }, { "grad_norm": 0.2620326578617096, "learning_rate": 2.299480259042086e-07, "loss": 0.0121, "step": 77680 }, { "grad_norm": 0.17565405368804932, "learning_rate": 2.2797234181102244e-07, "loss": 0.0159, "step": 77690 }, { "grad_norm": 0.1784963458776474, "learning_rate": 2.2600516239609238e-07, "loss": 0.0153, "step": 77700 }, { "grad_norm": 0.14639759063720703, "learning_rate": 2.240464879955606e-07, "loss": 0.0163, "step": 77710 }, { "grad_norm": 0.29135411977767944, "learning_rate": 2.2209631894410387e-07, "loss": 0.0144, "step": 77720 }, { "grad_norm": 0.1769483983516693, "learning_rate": 2.2015465557496117e-07, "loss": 0.0126, "step": 77730 }, { "grad_norm": 0.24709191918373108, "learning_rate": 2.1822149821990602e-07, "loss": 0.0164, "step": 77740 }, { "grad_norm": 0.3120182454586029, "learning_rate": 2.1629684720926303e-07, "loss": 0.0186, "step": 77750 }, { "grad_norm": 0.25155797600746155, "learning_rate": 2.1438070287189693e-07, "loss": 0.0163, "step": 77760 }, { "grad_norm": 0.2933257520198822, "learning_rate": 2.1247306553523472e-07, "loss": 0.0172, "step": 77770 }, { "grad_norm": 0.22649024426937103, "learning_rate": 2.105739355252323e-07, "loss": 0.0154, "step": 77780 }, { "grad_norm": 0.16539302468299866, "learning_rate": 2.0868331316639678e-07, "loss": 0.0128, "step": 77790 }, { "grad_norm": 0.20993003249168396, "learning_rate": 2.0680119878179193e-07, "loss": 0.0111, "step": 77800 }, { "grad_norm": 0.4085078835487366, "learning_rate": 2.0492759269301054e-07, "loss": 0.0177, "step": 77810 }, { "grad_norm": 0.13188108801841736, "learning_rate": 2.0306249522021315e-07, "loss": 0.0101, "step": 77820 }, { "grad_norm": 0.15835833549499512, "learning_rate": 2.0120590668207816e-07, "loss": 0.0146, "step": 77830 }, { "grad_norm": 0.15798819065093994, "learning_rate": 1.993578273958574e-07, "loss": 0.0116, "step": 77840 }, { "grad_norm": 0.27605095505714417, "learning_rate": 1.975182576773371e-07, "loss": 0.0165, "step": 77850 }, { "grad_norm": 0.43953776359558105, "learning_rate": 1.9568719784083812e-07, "loss": 0.0231, "step": 77860 }, { "grad_norm": 0.4342625141143799, "learning_rate": 1.9386464819924898e-07, "loss": 0.0198, "step": 77870 }, { "grad_norm": 0.16813857853412628, "learning_rate": 1.9205060906399285e-07, "loss": 0.0137, "step": 77880 }, { "grad_norm": 0.12011654675006866, "learning_rate": 1.9024508074503845e-07, "loss": 0.0153, "step": 77890 }, { "grad_norm": 0.3688790202140808, "learning_rate": 1.8844806355089452e-07, "loss": 0.0228, "step": 77900 }, { "grad_norm": 0.2896844148635864, "learning_rate": 1.8665955778863208e-07, "loss": 0.0205, "step": 77910 }, { "grad_norm": 0.2818766236305237, "learning_rate": 1.848795637638512e-07, "loss": 0.0149, "step": 77920 }, { "grad_norm": 0.33515626192092896, "learning_rate": 1.831080817807085e-07, "loss": 0.0213, "step": 77930 }, { "grad_norm": 0.1624746173620224, "learning_rate": 1.813451121418952e-07, "loss": 0.0222, "step": 77940 }, { "grad_norm": 0.29693830013275146, "learning_rate": 1.795906551486648e-07, "loss": 0.013, "step": 77950 }, { "grad_norm": 0.22879713773727417, "learning_rate": 1.7784471110079414e-07, "loss": 0.0162, "step": 77960 }, { "grad_norm": 0.2909114956855774, "learning_rate": 1.7610728029662793e-07, "loss": 0.0142, "step": 77970 }, { "grad_norm": 0.1745731234550476, "learning_rate": 1.7437836303303979e-07, "loss": 0.0159, "step": 77980 }, { "grad_norm": 0.20688392221927643, "learning_rate": 1.726579596054545e-07, "loss": 0.0139, "step": 77990 }, { "grad_norm": 0.19881755113601685, "learning_rate": 1.7094607030784803e-07, "loss": 0.0192, "step": 78000 }, { "grad_norm": 0.2414986938238144, "learning_rate": 1.6924269543272532e-07, "loss": 0.014, "step": 78010 }, { "grad_norm": 0.1962551325559616, "learning_rate": 1.6754783527115348e-07, "loss": 0.0113, "step": 78020 }, { "grad_norm": 0.18606743216514587, "learning_rate": 1.6586149011273422e-07, "loss": 0.0143, "step": 78030 }, { "grad_norm": 0.2947212755680084, "learning_rate": 1.6418366024562038e-07, "loss": 0.0206, "step": 78040 }, { "grad_norm": 0.20081400871276855, "learning_rate": 1.6251434595651037e-07, "loss": 0.017, "step": 78050 }, { "grad_norm": 0.25879669189453125, "learning_rate": 1.608535475306372e-07, "loss": 0.0162, "step": 78060 }, { "grad_norm": 0.27059322595596313, "learning_rate": 1.5920126525179048e-07, "loss": 0.0151, "step": 78070 }, { "grad_norm": 0.22354112565517426, "learning_rate": 1.5755749940229435e-07, "loss": 0.0121, "step": 78080 }, { "grad_norm": 0.24992208182811737, "learning_rate": 1.559222502630353e-07, "loss": 0.014, "step": 78090 }, { "grad_norm": 0.19027745723724365, "learning_rate": 1.5429551811341757e-07, "loss": 0.0161, "step": 78100 }, { "grad_norm": 0.26507076621055603, "learning_rate": 1.5267730323141882e-07, "loss": 0.0195, "step": 78110 }, { "grad_norm": 0.19280977547168732, "learning_rate": 1.5106760589353454e-07, "loss": 0.0132, "step": 78120 }, { "grad_norm": 0.18989089131355286, "learning_rate": 1.4946642637483355e-07, "loss": 0.013, "step": 78130 }, { "grad_norm": 0.3979249596595764, "learning_rate": 1.4787376494889703e-07, "loss": 0.0147, "step": 78140 }, { "grad_norm": 0.20799896121025085, "learning_rate": 1.4628962188787955e-07, "loss": 0.0117, "step": 78150 }, { "grad_norm": 0.25896593928337097, "learning_rate": 1.4471399746247006e-07, "loss": 0.0165, "step": 78160 }, { "grad_norm": 0.3137790858745575, "learning_rate": 1.4314689194188103e-07, "loss": 0.0122, "step": 78170 }, { "grad_norm": 0.3894438147544861, "learning_rate": 1.4158830559390933e-07, "loss": 0.0181, "step": 78180 }, { "grad_norm": 0.14941076934337616, "learning_rate": 1.4003823868486422e-07, "loss": 0.0323, "step": 78190 }, { "grad_norm": 0.24456074833869934, "learning_rate": 1.3849669147960598e-07, "loss": 0.0184, "step": 78200 }, { "grad_norm": 0.19595983624458313, "learning_rate": 1.3696366424155726e-07, "loss": 0.014, "step": 78210 }, { "grad_norm": 0.407488614320755, "learning_rate": 1.35439157232653e-07, "loss": 0.0244, "step": 78220 }, { "grad_norm": 0.22189432382583618, "learning_rate": 1.3392317071340144e-07, "loss": 0.0158, "step": 78230 }, { "grad_norm": 0.41554009914398193, "learning_rate": 1.3241570494283984e-07, "loss": 0.0157, "step": 78240 }, { "grad_norm": 0.30746692419052124, "learning_rate": 1.3091676017855103e-07, "loss": 0.0206, "step": 78250 }, { "grad_norm": 0.2770298719406128, "learning_rate": 1.2942633667666904e-07, "loss": 0.014, "step": 78260 }, { "grad_norm": 0.11338603496551514, "learning_rate": 1.2794443469185679e-07, "loss": 0.0216, "step": 78270 }, { "grad_norm": 0.3397178053855896, "learning_rate": 1.2647105447734508e-07, "loss": 0.0162, "step": 78280 }, { "grad_norm": 0.11109776794910431, "learning_rate": 1.2500619628488254e-07, "loss": 0.0131, "step": 78290 }, { "grad_norm": 0.2569252848625183, "learning_rate": 1.2354986036477446e-07, "loss": 0.0213, "step": 78300 }, { "grad_norm": 0.2855610251426697, "learning_rate": 1.221020469658718e-07, "loss": 0.0169, "step": 78310 }, { "grad_norm": 0.22746232151985168, "learning_rate": 1.2066275633556556e-07, "loss": 0.0136, "step": 78320 }, { "grad_norm": 0.33410388231277466, "learning_rate": 1.192319887197979e-07, "loss": 0.013, "step": 78330 }, { "grad_norm": 0.18832653760910034, "learning_rate": 1.1780974436303438e-07, "loss": 0.0109, "step": 78340 }, { "grad_norm": 0.4118538498878479, "learning_rate": 1.1639602350830836e-07, "loss": 0.017, "step": 78350 }, { "grad_norm": 0.33303171396255493, "learning_rate": 1.1499082639718217e-07, "loss": 0.0165, "step": 78360 }, { "grad_norm": 0.27627506852149963, "learning_rate": 1.1359415326976374e-07, "loss": 0.0098, "step": 78370 }, { "grad_norm": 0.17733557522296906, "learning_rate": 1.1220600436470663e-07, "loss": 0.0118, "step": 78380 }, { "grad_norm": 0.26584354043006897, "learning_rate": 1.1082637991920996e-07, "loss": 0.0172, "step": 78390 }, { "grad_norm": 0.2236141413450241, "learning_rate": 1.0945528016901851e-07, "loss": 0.0148, "step": 78400 }, { "grad_norm": 0.14466789364814758, "learning_rate": 1.0809270534840599e-07, "loss": 0.0121, "step": 78410 }, { "grad_norm": 0.2759215533733368, "learning_rate": 1.067386556902028e-07, "loss": 0.0133, "step": 78420 }, { "grad_norm": 0.22647303342819214, "learning_rate": 1.05393131425785e-07, "loss": 0.0096, "step": 78430 }, { "grad_norm": 0.4814756512641907, "learning_rate": 1.0405613278505199e-07, "loss": 0.0132, "step": 78440 }, { "grad_norm": 0.33794185519218445, "learning_rate": 1.027276599964766e-07, "loss": 0.017, "step": 78450 }, { "grad_norm": 0.20342908799648285, "learning_rate": 1.0140771328704391e-07, "loss": 0.0105, "step": 78460 }, { "grad_norm": 0.169488787651062, "learning_rate": 1.0009629288231237e-07, "loss": 0.0159, "step": 78470 }, { "grad_norm": 0.1520940661430359, "learning_rate": 9.879339900635276e-08, "loss": 0.0143, "step": 78480 }, { "grad_norm": 0.436989426612854, "learning_rate": 9.74990318817981e-08, "loss": 0.0159, "step": 78490 }, { "grad_norm": 0.3816255033016205, "learning_rate": 9.621319172982701e-08, "loss": 0.0188, "step": 78500 }, { "grad_norm": 0.2731134593486786, "learning_rate": 9.493587877015264e-08, "loss": 0.0142, "step": 78510 }, { "grad_norm": 0.3729822039604187, "learning_rate": 9.366709322102263e-08, "loss": 0.0126, "step": 78520 }, { "grad_norm": 0.2744227349758148, "learning_rate": 9.240683529924688e-08, "loss": 0.0179, "step": 78530 }, { "grad_norm": 0.19702930748462677, "learning_rate": 9.115510522016979e-08, "loss": 0.0128, "step": 78540 }, { "grad_norm": 0.18282534182071686, "learning_rate": 8.991190319767583e-08, "loss": 0.0114, "step": 78550 }, { "grad_norm": 0.16174481809139252, "learning_rate": 8.867722944419509e-08, "loss": 0.014, "step": 78560 }, { "grad_norm": 0.16752247512340546, "learning_rate": 8.745108417069214e-08, "loss": 0.0116, "step": 78570 }, { "grad_norm": 0.22595492005348206, "learning_rate": 8.623346758669381e-08, "loss": 0.0109, "step": 78580 }, { "grad_norm": 0.2572394013404846, "learning_rate": 8.502437990025037e-08, "loss": 0.0145, "step": 78590 }, { "grad_norm": 0.20747262239456177, "learning_rate": 8.382382131795764e-08, "loss": 0.013, "step": 78600 }, { "grad_norm": 0.16187186539173126, "learning_rate": 8.263179204496818e-08, "loss": 0.0181, "step": 78610 }, { "grad_norm": 0.19893883168697357, "learning_rate": 8.144829228496354e-08, "loss": 0.015, "step": 78620 }, { "grad_norm": 0.47549891471862793, "learning_rate": 8.027332224016526e-08, "loss": 0.0163, "step": 78630 }, { "grad_norm": 0.18310190737247467, "learning_rate": 7.910688211135164e-08, "loss": 0.0152, "step": 78640 }, { "grad_norm": 0.187565878033638, "learning_rate": 7.794897209783546e-08, "loss": 0.018, "step": 78650 }, { "grad_norm": 0.3248983919620514, "learning_rate": 7.679959239746403e-08, "loss": 0.0175, "step": 78660 }, { "grad_norm": 0.12935663759708405, "learning_rate": 7.565874320664135e-08, "loss": 0.0187, "step": 78670 }, { "grad_norm": 0.3608725070953369, "learning_rate": 7.45264247203059e-08, "loss": 0.0168, "step": 78680 }, { "grad_norm": 0.18308471143245697, "learning_rate": 7.340263713194184e-08, "loss": 0.0119, "step": 78690 }, { "grad_norm": 0.2708888351917267, "learning_rate": 7.228738063356777e-08, "loss": 0.0128, "step": 78700 }, { "grad_norm": 0.5233723521232605, "learning_rate": 7.118065541575903e-08, "loss": 0.02, "step": 78710 }, { "grad_norm": 0.25153806805610657, "learning_rate": 7.008246166761435e-08, "loss": 0.0113, "step": 78720 }, { "grad_norm": 0.19685736298561096, "learning_rate": 6.899279957679472e-08, "loss": 0.0112, "step": 78730 }, { "grad_norm": 0.2807829976081848, "learning_rate": 6.791166932949011e-08, "loss": 0.0122, "step": 78740 }, { "grad_norm": 0.32414597272872925, "learning_rate": 6.683907111043608e-08, "loss": 0.0149, "step": 78750 }, { "grad_norm": 0.18992990255355835, "learning_rate": 6.577500510290824e-08, "loss": 0.0131, "step": 78760 }, { "grad_norm": 0.1989320069551468, "learning_rate": 6.47194714887278e-08, "loss": 0.0269, "step": 78770 }, { "grad_norm": 0.19494116306304932, "learning_rate": 6.367247044825608e-08, "loss": 0.012, "step": 78780 }, { "grad_norm": 0.15804359316825867, "learning_rate": 6.263400216039994e-08, "loss": 0.0132, "step": 78790 }, { "grad_norm": 0.27892768383026123, "learning_rate": 6.160406680260078e-08, "loss": 0.0309, "step": 78800 }, { "grad_norm": 0.15841500461101532, "learning_rate": 6.058266455084561e-08, "loss": 0.0139, "step": 78810 }, { "grad_norm": 0.22209344804286957, "learning_rate": 5.956979557967257e-08, "loss": 0.0166, "step": 78820 }, { "grad_norm": 0.3358629047870636, "learning_rate": 5.856546006214325e-08, "loss": 0.018, "step": 78830 }, { "grad_norm": 0.1333339661359787, "learning_rate": 5.7569658169881466e-08, "loss": 0.0152, "step": 78840 }, { "grad_norm": 0.12864162027835846, "learning_rate": 5.658239007303445e-08, "loss": 0.0159, "step": 78850 }, { "grad_norm": 0.26387715339660645, "learning_rate": 5.560365594030059e-08, "loss": 0.0172, "step": 78860 }, { "grad_norm": 0.15721702575683594, "learning_rate": 5.463345593891833e-08, "loss": 0.0134, "step": 78870 }, { "grad_norm": 0.17716272175312042, "learning_rate": 5.367179023467173e-08, "loss": 0.0152, "step": 78880 }, { "grad_norm": 0.24599140882492065, "learning_rate": 5.271865899187378e-08, "loss": 0.0132, "step": 78890 }, { "grad_norm": 0.19560019671916962, "learning_rate": 5.177406237340532e-08, "loss": 0.0256, "step": 78900 }, { "grad_norm": 0.17179669439792633, "learning_rate": 5.083800054065946e-08, "loss": 0.0128, "step": 78910 }, { "grad_norm": 0.35797205567359924, "learning_rate": 4.991047365358603e-08, "loss": 0.0122, "step": 78920 }, { "grad_norm": 0.3876006603240967, "learning_rate": 4.899148187067493e-08, "loss": 0.0248, "step": 78930 }, { "grad_norm": 0.14712302386760712, "learning_rate": 4.808102534895609e-08, "loss": 0.0122, "step": 78940 }, { "grad_norm": 0.22644226253032684, "learning_rate": 4.717910424400507e-08, "loss": 0.0164, "step": 78950 }, { "grad_norm": 0.5686753392219543, "learning_rate": 4.628571870993193e-08, "loss": 0.016, "step": 78960 }, { "grad_norm": 0.4105573296546936, "learning_rate": 4.5400868899392324e-08, "loss": 0.0183, "step": 78970 }, { "grad_norm": 0.3358902931213379, "learning_rate": 4.4524554963576436e-08, "loss": 0.0221, "step": 78980 }, { "grad_norm": 0.28493764996528625, "learning_rate": 4.36567770522367e-08, "loss": 0.0147, "step": 78990 }, { "grad_norm": 0.44342705607414246, "learning_rate": 4.279753531364339e-08, "loss": 0.0132, "step": 79000 }, { "grad_norm": 0.2721249759197235, "learning_rate": 4.1946829894617955e-08, "loss": 0.0185, "step": 79010 }, { "grad_norm": 0.1474376618862152, "learning_rate": 4.11046609405219e-08, "loss": 0.0217, "step": 79020 }, { "grad_norm": 0.16746602952480316, "learning_rate": 4.027102859526233e-08, "loss": 0.0105, "step": 79030 }, { "grad_norm": 0.2591817378997803, "learning_rate": 3.9445933001280856e-08, "loss": 0.0149, "step": 79040 }, { "grad_norm": 0.2146659940481186, "learning_rate": 3.8629374299564704e-08, "loss": 0.0155, "step": 79050 }, { "grad_norm": 0.43398621678352356, "learning_rate": 3.78213526296467e-08, "loss": 0.016, "step": 79060 }, { "grad_norm": 0.15709783136844635, "learning_rate": 3.702186812958308e-08, "loss": 0.0143, "step": 79070 }, { "grad_norm": 0.2172141969203949, "learning_rate": 3.623092093599789e-08, "loss": 0.0138, "step": 79080 }, { "grad_norm": 0.34572717547416687, "learning_rate": 3.544851118403303e-08, "loss": 0.0174, "step": 79090 }, { "grad_norm": 0.38825222849845886, "learning_rate": 3.46746390073871e-08, "loss": 0.0204, "step": 79100 }, { "grad_norm": 0.3645015060901642, "learning_rate": 3.390930453829322e-08, "loss": 0.0249, "step": 79110 }, { "grad_norm": 0.24157358705997467, "learning_rate": 3.3152507907519005e-08, "loss": 0.0108, "step": 79120 }, { "grad_norm": 0.10237310081720352, "learning_rate": 3.240424924438323e-08, "loss": 0.0157, "step": 79130 }, { "grad_norm": 0.14993520081043243, "learning_rate": 3.1664528676750295e-08, "loss": 0.0172, "step": 79140 }, { "grad_norm": 0.29136887192726135, "learning_rate": 3.093334633100797e-08, "loss": 0.0239, "step": 79150 }, { "grad_norm": 0.20742955803871155, "learning_rate": 3.021070233210077e-08, "loss": 0.0171, "step": 79160 }, { "grad_norm": 0.17708714306354523, "learning_rate": 2.9496596803507692e-08, "loss": 0.0147, "step": 79170 }, { "grad_norm": 0.2561282813549042, "learning_rate": 2.879102986725335e-08, "loss": 0.0135, "step": 79180 }, { "grad_norm": 0.12465126812458038, "learning_rate": 2.809400164389131e-08, "loss": 0.0132, "step": 79190 }, { "grad_norm": 0.17464381456375122, "learning_rate": 2.740551225253185e-08, "loss": 0.012, "step": 79200 }, { "grad_norm": 0.39551112055778503, "learning_rate": 2.6725561810819754e-08, "loss": 0.0181, "step": 79210 }, { "grad_norm": 0.6541927456855774, "learning_rate": 2.605415043493431e-08, "loss": 0.0136, "step": 79220 }, { "grad_norm": 0.19806857407093048, "learning_rate": 2.5391278239605965e-08, "loss": 0.013, "step": 79230 }, { "grad_norm": 0.1457970291376114, "learning_rate": 2.4736945338094118e-08, "loss": 0.0124, "step": 79240 }, { "grad_norm": 0.2517150044441223, "learning_rate": 2.4091151842214887e-08, "loss": 0.0131, "step": 79250 }, { "grad_norm": 0.16092173755168915, "learning_rate": 2.3453897862318884e-08, "loss": 0.0113, "step": 79260 }, { "grad_norm": 0.14343468844890594, "learning_rate": 2.2825183507285686e-08, "loss": 0.0142, "step": 79270 }, { "grad_norm": 0.13915885984897614, "learning_rate": 2.220500888455157e-08, "loss": 0.0192, "step": 79280 }, { "grad_norm": 0.32110339403152466, "learning_rate": 2.1593374100081777e-08, "loss": 0.0209, "step": 79290 }, { "grad_norm": 0.2566273510456085, "learning_rate": 2.09902792583927e-08, "loss": 0.0216, "step": 79300 }, { "grad_norm": 0.22688961029052734, "learning_rate": 2.0395724462540788e-08, "loss": 0.0122, "step": 79310 }, { "grad_norm": 0.12002266943454742, "learning_rate": 1.9809709814111453e-08, "loss": 0.0145, "step": 79320 }, { "grad_norm": 0.22862489521503448, "learning_rate": 1.923223541324126e-08, "loss": 0.0198, "step": 79330 }, { "grad_norm": 0.17358741164207458, "learning_rate": 1.8663301358606833e-08, "loss": 0.015, "step": 79340 }, { "grad_norm": 0.17611783742904663, "learning_rate": 1.8102907747419295e-08, "loss": 0.0198, "step": 79350 }, { "grad_norm": 0.17556904256343842, "learning_rate": 1.7551054675435385e-08, "loss": 0.015, "step": 79360 }, { "grad_norm": 0.17442092299461365, "learning_rate": 1.7007742236957447e-08, "loss": 0.0157, "step": 79370 }, { "grad_norm": 0.2524092495441437, "learning_rate": 1.647297052481678e-08, "loss": 0.0141, "step": 79380 }, { "grad_norm": 0.39932188391685486, "learning_rate": 1.5946739630390285e-08, "loss": 0.0166, "step": 79390 }, { "grad_norm": 0.17369899153709412, "learning_rate": 1.5429049643606032e-08, "loss": 0.0176, "step": 79400 }, { "grad_norm": 0.1554180085659027, "learning_rate": 1.4919900652909935e-08, "loss": 0.0189, "step": 79410 }, { "grad_norm": 0.21460169553756714, "learning_rate": 1.4419292745310175e-08, "loss": 0.0181, "step": 79420 }, { "grad_norm": 0.2015552967786789, "learning_rate": 1.3927226006343885e-08, "loss": 0.012, "step": 79430 }, { "grad_norm": 0.3756334185600281, "learning_rate": 1.3443700520093805e-08, "loss": 0.0201, "step": 79440 }, { "grad_norm": 0.16659900546073914, "learning_rate": 1.296871636917718e-08, "loss": 0.0132, "step": 79450 }, { "grad_norm": 0.2572267949581146, "learning_rate": 1.2502273634762419e-08, "loss": 0.0123, "step": 79460 }, { "grad_norm": 0.34878307580947876, "learning_rate": 1.2044372396546876e-08, "loss": 0.0135, "step": 79470 }, { "grad_norm": 0.3569834530353546, "learning_rate": 1.1595012732773524e-08, "loss": 0.0166, "step": 79480 }, { "grad_norm": 0.13204801082611084, "learning_rate": 1.1154194720225386e-08, "loss": 0.0137, "step": 79490 }, { "grad_norm": 0.23826110363006592, "learning_rate": 1.0721918434231093e-08, "loss": 0.0154, "step": 79500 }, { "grad_norm": 0.25958114862442017, "learning_rate": 1.0298183948648232e-08, "loss": 0.0164, "step": 79510 }, { "grad_norm": 0.1740562617778778, "learning_rate": 9.88299133588555e-09, "loss": 0.01, "step": 79520 }, { "grad_norm": 0.18255192041397095, "learning_rate": 9.476340666891847e-09, "loss": 0.0178, "step": 79530 }, { "grad_norm": 0.19700422883033752, "learning_rate": 9.078232011139332e-09, "loss": 0.0116, "step": 79540 }, { "grad_norm": 0.13375012576580048, "learning_rate": 8.68866543666802e-09, "loss": 0.0116, "step": 79550 }, { "grad_norm": 0.1607828140258789, "learning_rate": 8.307641010035782e-09, "loss": 0.0146, "step": 79560 }, { "grad_norm": 0.19659124314785004, "learning_rate": 7.93515879635165e-09, "loss": 0.0131, "step": 79570 }, { "grad_norm": 0.24880695343017578, "learning_rate": 7.571218859264706e-09, "loss": 0.0128, "step": 79580 }, { "grad_norm": 0.24367287755012512, "learning_rate": 7.215821260958544e-09, "loss": 0.0169, "step": 79590 }, { "grad_norm": 0.226822629570961, "learning_rate": 6.868966062162363e-09, "loss": 0.0143, "step": 79600 }, { "grad_norm": 0.20603206753730774, "learning_rate": 6.530653322145419e-09, "loss": 0.0134, "step": 79610 }, { "grad_norm": 0.19357575476169586, "learning_rate": 6.200883098717025e-09, "loss": 0.0096, "step": 79620 }, { "grad_norm": 0.2978785037994385, "learning_rate": 5.879655448226551e-09, "loss": 0.022, "step": 79630 }, { "grad_norm": 0.29744771122932434, "learning_rate": 5.566970425557872e-09, "loss": 0.0105, "step": 79640 }, { "grad_norm": 0.34560465812683105, "learning_rate": 5.26282808414047e-09, "loss": 0.0118, "step": 79650 }, { "grad_norm": 0.2500053346157074, "learning_rate": 4.967228475949437e-09, "loss": 0.0184, "step": 79660 }, { "grad_norm": 0.3895353674888611, "learning_rate": 4.680171651494369e-09, "loss": 0.0133, "step": 79670 }, { "grad_norm": 0.3962966501712799, "learning_rate": 4.40165765981937e-09, "loss": 0.0158, "step": 79680 }, { "grad_norm": 0.18979476392269135, "learning_rate": 4.131686548519698e-09, "loss": 0.0122, "step": 79690 }, { "grad_norm": 0.23886306583881378, "learning_rate": 3.8702583637251214e-09, "loss": 0.0153, "step": 79700 }, { "grad_norm": 0.16622067987918854, "learning_rate": 3.617373150105463e-09, "loss": 0.0133, "step": 79710 }, { "grad_norm": 0.24382583796977997, "learning_rate": 3.3730309508706036e-09, "loss": 0.0124, "step": 79720 }, { "grad_norm": 0.28104910254478455, "learning_rate": 3.137231807781582e-09, "loss": 0.0096, "step": 79730 }, { "grad_norm": 0.12413223832845688, "learning_rate": 2.9099757611172894e-09, "loss": 0.0147, "step": 79740 }, { "grad_norm": 0.21121245622634888, "learning_rate": 2.6912628497133275e-09, "loss": 0.014, "step": 79750 }, { "grad_norm": 0.3613011837005615, "learning_rate": 2.481093110945354e-09, "loss": 0.0157, "step": 79760 }, { "grad_norm": 0.17436926066875458, "learning_rate": 2.2794665807235327e-09, "loss": 0.0182, "step": 79770 }, { "grad_norm": 0.1763845533132553, "learning_rate": 2.0863832934980843e-09, "loss": 0.0153, "step": 79780 }, { "grad_norm": 0.15690070390701294, "learning_rate": 1.901843282264837e-09, "loss": 0.0177, "step": 79790 }, { "grad_norm": 0.28319674730300903, "learning_rate": 1.7258465785541245e-09, "loss": 0.0136, "step": 79800 }, { "grad_norm": 0.1914740651845932, "learning_rate": 1.5583932124418887e-09, "loss": 0.0117, "step": 79810 }, { "grad_norm": 0.2030181735754013, "learning_rate": 1.3994832125441282e-09, "loss": 0.0136, "step": 79820 }, { "grad_norm": 0.2555520534515381, "learning_rate": 1.2491166060057958e-09, "loss": 0.0147, "step": 79830 }, { "grad_norm": 0.2508776783943176, "learning_rate": 1.1072934185230032e-09, "loss": 0.0148, "step": 79840 }, { "grad_norm": 0.25737476348876953, "learning_rate": 9.740136743319195e-10, "loss": 0.0159, "step": 79850 }, { "grad_norm": 0.24877318739891052, "learning_rate": 8.492773962087696e-10, "loss": 0.0131, "step": 79860 }, { "grad_norm": 0.3309752643108368, "learning_rate": 7.330846054587337e-10, "loss": 0.0229, "step": 79870 }, { "grad_norm": 0.2802787125110626, "learning_rate": 6.254353219492526e-10, "loss": 0.0142, "step": 79880 }, { "grad_norm": 0.19212143123149872, "learning_rate": 5.263295640600685e-10, "loss": 0.0119, "step": 79890 }, { "grad_norm": 0.17195898294448853, "learning_rate": 4.357673487387359e-10, "loss": 0.0135, "step": 79900 }, { "grad_norm": 0.25104397535324097, "learning_rate": 3.537486914506616e-10, "loss": 0.0131, "step": 79910 }, { "grad_norm": 0.21288718283176422, "learning_rate": 2.8027360621241117e-10, "loss": 0.0155, "step": 79920 }, { "grad_norm": 0.1911114901304245, "learning_rate": 2.153421055806071e-10, "loss": 0.0121, "step": 79930 }, { "grad_norm": 0.3642168939113617, "learning_rate": 1.5895420064637733e-10, "loss": 0.015, "step": 79940 }, { "grad_norm": 0.21506117284297943, "learning_rate": 1.1110990105200891e-10, "loss": 0.0149, "step": 79950 }, { "grad_norm": 0.21517065167427063, "learning_rate": 7.180921496874326e-11, "loss": 0.0188, "step": 79960 }, { "grad_norm": 0.44228604435920715, "learning_rate": 4.10521491134297e-11, "loss": 0.0137, "step": 79970 }, { "grad_norm": 0.2547098994255066, "learning_rate": 1.883870874297422e-11, "loss": 0.0126, "step": 79980 }, { "grad_norm": 0.35104477405548096, "learning_rate": 5.1688976432373584e-12, "loss": 0.0171, "step": 79990 }, { "grad_norm": 0.5942901372909546, "learning_rate": 4.271816234080461e-14, "loss": 0.0123, "step": 80000 } ], "logging_steps": 10, "max_steps": 80000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }