{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.5, "eval_steps": 500, "global_step": 40000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 2.444127082824707, "learning_rate": 2.25e-07, "loss": 1.3287, "step": 10 }, { "grad_norm": 2.5416176319122314, "learning_rate": 4.75e-07, "loss": 1.3297, "step": 20 }, { "grad_norm": 2.3953311443328857, "learning_rate": 7.25e-07, "loss": 1.3296, "step": 30 }, { "grad_norm": 1.8786507844924927, "learning_rate": 9.75e-07, "loss": 1.3174, "step": 40 }, { "grad_norm": 1.7076902389526367, "learning_rate": 1.2250000000000001e-06, "loss": 1.2963, "step": 50 }, { "grad_norm": 1.1601673364639282, "learning_rate": 1.475e-06, "loss": 1.2438, "step": 60 }, { "grad_norm": 1.0342538356781006, "learning_rate": 1.7250000000000002e-06, "loss": 1.2045, "step": 70 }, { "grad_norm": 0.8647598028182983, "learning_rate": 1.975e-06, "loss": 1.1552, "step": 80 }, { "grad_norm": 0.7864705920219421, "learning_rate": 2.225e-06, "loss": 1.1224, "step": 90 }, { "grad_norm": 0.4195609986782074, "learning_rate": 2.4750000000000004e-06, "loss": 1.1144, "step": 100 }, { "grad_norm": 0.48813897371292114, "learning_rate": 2.725e-06, "loss": 1.0994, "step": 110 }, { "grad_norm": 0.36481109261512756, "learning_rate": 2.975e-06, "loss": 1.0946, "step": 120 }, { "grad_norm": 0.4182380437850952, "learning_rate": 3.225e-06, "loss": 1.0948, "step": 130 }, { "grad_norm": 0.36900538206100464, "learning_rate": 3.4750000000000006e-06, "loss": 1.0986, "step": 140 }, { "grad_norm": 0.36304715275764465, "learning_rate": 3.725e-06, "loss": 1.0922, "step": 150 }, { "grad_norm": 0.4019674062728882, "learning_rate": 3.975e-06, "loss": 1.0804, "step": 160 }, { "grad_norm": 0.3392553925514221, "learning_rate": 4.225e-06, "loss": 1.0676, "step": 170 }, { "grad_norm": 0.48098742961883545, "learning_rate": 4.475e-06, "loss": 1.059, "step": 180 }, { "grad_norm": 0.38152098655700684, "learning_rate": 4.7250000000000005e-06, "loss": 1.0456, "step": 190 }, { "grad_norm": 0.49172866344451904, "learning_rate": 4.975000000000001e-06, "loss": 1.0479, "step": 200 }, { "grad_norm": 0.39684605598449707, "learning_rate": 5.225e-06, "loss": 1.0438, "step": 210 }, { "grad_norm": 0.442143976688385, "learning_rate": 5.475e-06, "loss": 1.0444, "step": 220 }, { "grad_norm": 0.44135600328445435, "learning_rate": 5.725e-06, "loss": 1.0365, "step": 230 }, { "grad_norm": 0.3978268802165985, "learning_rate": 5.975e-06, "loss": 1.0335, "step": 240 }, { "grad_norm": 0.3391270339488983, "learning_rate": 6.2250000000000005e-06, "loss": 1.027, "step": 250 }, { "grad_norm": 0.4496247172355652, "learning_rate": 6.475000000000001e-06, "loss": 1.0372, "step": 260 }, { "grad_norm": 0.456939697265625, "learning_rate": 6.725000000000001e-06, "loss": 1.0304, "step": 270 }, { "grad_norm": 0.4169675409793854, "learning_rate": 6.975000000000001e-06, "loss": 1.0232, "step": 280 }, { "grad_norm": 0.5090293288230896, "learning_rate": 7.2249999999999994e-06, "loss": 1.0095, "step": 290 }, { "grad_norm": 0.6883742809295654, "learning_rate": 7.4750000000000004e-06, "loss": 1.0039, "step": 300 }, { "grad_norm": 1.1060408353805542, "learning_rate": 7.725e-06, "loss": 0.9884, "step": 310 }, { "grad_norm": 1.1030851602554321, "learning_rate": 7.975e-06, "loss": 0.9721, "step": 320 }, { "grad_norm": 0.8201788663864136, "learning_rate": 8.225e-06, "loss": 0.9456, "step": 330 }, { "grad_norm": 1.387393593788147, "learning_rate": 8.475000000000001e-06, "loss": 0.9133, "step": 340 }, { "grad_norm": 1.5607719421386719, "learning_rate": 8.725e-06, "loss": 0.895, "step": 350 }, { "grad_norm": 1.4895448684692383, "learning_rate": 8.975e-06, "loss": 0.8629, "step": 360 }, { "grad_norm": 1.8840293884277344, "learning_rate": 9.225e-06, "loss": 0.829, "step": 370 }, { "grad_norm": 1.6571836471557617, "learning_rate": 9.475e-06, "loss": 0.8115, "step": 380 }, { "grad_norm": 1.284320592880249, "learning_rate": 9.725000000000001e-06, "loss": 0.7843, "step": 390 }, { "grad_norm": 2.281913995742798, "learning_rate": 9.975e-06, "loss": 0.754, "step": 400 }, { "grad_norm": 1.9086191654205322, "learning_rate": 1.0225e-05, "loss": 0.7177, "step": 410 }, { "grad_norm": 1.6675872802734375, "learning_rate": 1.0475e-05, "loss": 0.6636, "step": 420 }, { "grad_norm": 2.505089282989502, "learning_rate": 1.0725e-05, "loss": 0.6346, "step": 430 }, { "grad_norm": 2.870072603225708, "learning_rate": 1.0975e-05, "loss": 0.6026, "step": 440 }, { "grad_norm": 2.3144166469573975, "learning_rate": 1.1225e-05, "loss": 0.5832, "step": 450 }, { "grad_norm": 2.205669641494751, "learning_rate": 1.1475000000000001e-05, "loss": 0.5393, "step": 460 }, { "grad_norm": 2.4057912826538086, "learning_rate": 1.1725e-05, "loss": 0.5165, "step": 470 }, { "grad_norm": 2.499223470687866, "learning_rate": 1.1975e-05, "loss": 0.4821, "step": 480 }, { "grad_norm": 2.657151699066162, "learning_rate": 1.2225e-05, "loss": 0.4442, "step": 490 }, { "grad_norm": 2.5828635692596436, "learning_rate": 1.2475e-05, "loss": 0.4217, "step": 500 }, { "grad_norm": 2.4686176776885986, "learning_rate": 1.2725000000000001e-05, "loss": 0.3987, "step": 510 }, { "grad_norm": 2.322493553161621, "learning_rate": 1.2975e-05, "loss": 0.3749, "step": 520 }, { "grad_norm": 2.7166411876678467, "learning_rate": 1.3225000000000001e-05, "loss": 0.3509, "step": 530 }, { "grad_norm": 2.5917770862579346, "learning_rate": 1.3475000000000002e-05, "loss": 0.3443, "step": 540 }, { "grad_norm": 2.1678125858306885, "learning_rate": 1.3725000000000002e-05, "loss": 0.322, "step": 550 }, { "grad_norm": 2.35670804977417, "learning_rate": 1.3975000000000003e-05, "loss": 0.3081, "step": 560 }, { "grad_norm": 2.6138381958007812, "learning_rate": 1.4225e-05, "loss": 0.288, "step": 570 }, { "grad_norm": 3.61136531829834, "learning_rate": 1.4475e-05, "loss": 0.2635, "step": 580 }, { "grad_norm": 3.0468008518218994, "learning_rate": 1.4725e-05, "loss": 0.2395, "step": 590 }, { "grad_norm": 2.7573180198669434, "learning_rate": 1.4975e-05, "loss": 0.2358, "step": 600 }, { "grad_norm": 2.4090518951416016, "learning_rate": 1.5225e-05, "loss": 0.2135, "step": 610 }, { "grad_norm": 2.1365246772766113, "learning_rate": 1.5475e-05, "loss": 0.2148, "step": 620 }, { "grad_norm": 2.821779727935791, "learning_rate": 1.5725e-05, "loss": 0.2058, "step": 630 }, { "grad_norm": 2.9579076766967773, "learning_rate": 1.5975000000000002e-05, "loss": 0.1896, "step": 640 }, { "grad_norm": 2.896094560623169, "learning_rate": 1.6225e-05, "loss": 0.1878, "step": 650 }, { "grad_norm": 2.6122028827667236, "learning_rate": 1.6475e-05, "loss": 0.1819, "step": 660 }, { "grad_norm": 2.5644006729125977, "learning_rate": 1.6725000000000003e-05, "loss": 0.1602, "step": 670 }, { "grad_norm": 2.188441276550293, "learning_rate": 1.6975000000000003e-05, "loss": 0.1466, "step": 680 }, { "grad_norm": 2.966672420501709, "learning_rate": 1.7225e-05, "loss": 0.1528, "step": 690 }, { "grad_norm": 1.8051854372024536, "learning_rate": 1.7475e-05, "loss": 0.1334, "step": 700 }, { "grad_norm": 3.4293932914733887, "learning_rate": 1.7725e-05, "loss": 0.1455, "step": 710 }, { "grad_norm": 2.001197576522827, "learning_rate": 1.7975e-05, "loss": 0.1349, "step": 720 }, { "grad_norm": 2.41823673248291, "learning_rate": 1.8225e-05, "loss": 0.1206, "step": 730 }, { "grad_norm": 2.5729737281799316, "learning_rate": 1.8475000000000002e-05, "loss": 0.126, "step": 740 }, { "grad_norm": 2.722001075744629, "learning_rate": 1.8725e-05, "loss": 0.1259, "step": 750 }, { "grad_norm": 2.093275308609009, "learning_rate": 1.8975e-05, "loss": 0.1401, "step": 760 }, { "grad_norm": 2.0091121196746826, "learning_rate": 1.9225e-05, "loss": 0.1237, "step": 770 }, { "grad_norm": 2.3565216064453125, "learning_rate": 1.9475000000000002e-05, "loss": 0.1193, "step": 780 }, { "grad_norm": 2.0500078201293945, "learning_rate": 1.9725000000000002e-05, "loss": 0.1156, "step": 790 }, { "grad_norm": 1.9491560459136963, "learning_rate": 1.9975e-05, "loss": 0.1169, "step": 800 }, { "grad_norm": 1.850030541419983, "learning_rate": 2.0225000000000004e-05, "loss": 0.1141, "step": 810 }, { "grad_norm": 2.5336716175079346, "learning_rate": 2.0475e-05, "loss": 0.1102, "step": 820 }, { "grad_norm": 1.7570431232452393, "learning_rate": 2.0725e-05, "loss": 0.108, "step": 830 }, { "grad_norm": 1.8658971786499023, "learning_rate": 2.0975e-05, "loss": 0.1054, "step": 840 }, { "grad_norm": 2.1174964904785156, "learning_rate": 2.1225e-05, "loss": 0.1037, "step": 850 }, { "grad_norm": 1.8886892795562744, "learning_rate": 2.1475e-05, "loss": 0.1151, "step": 860 }, { "grad_norm": 1.9222513437271118, "learning_rate": 2.1725e-05, "loss": 0.1012, "step": 870 }, { "grad_norm": 1.9165852069854736, "learning_rate": 2.1975000000000002e-05, "loss": 0.1006, "step": 880 }, { "grad_norm": 2.3876850605010986, "learning_rate": 2.2225e-05, "loss": 0.0944, "step": 890 }, { "grad_norm": 2.251217842102051, "learning_rate": 2.2475e-05, "loss": 0.0908, "step": 900 }, { "grad_norm": 1.687900185585022, "learning_rate": 2.2725000000000003e-05, "loss": 0.0857, "step": 910 }, { "grad_norm": 2.1535897254943848, "learning_rate": 2.2975000000000003e-05, "loss": 0.0868, "step": 920 }, { "grad_norm": 2.2971646785736084, "learning_rate": 2.3225000000000002e-05, "loss": 0.0879, "step": 930 }, { "grad_norm": 2.00461483001709, "learning_rate": 2.3475e-05, "loss": 0.0877, "step": 940 }, { "grad_norm": 1.5629587173461914, "learning_rate": 2.3725e-05, "loss": 0.0886, "step": 950 }, { "grad_norm": 1.7507143020629883, "learning_rate": 2.3975e-05, "loss": 0.0868, "step": 960 }, { "grad_norm": 1.9804723262786865, "learning_rate": 2.4225e-05, "loss": 0.0917, "step": 970 }, { "grad_norm": 1.7602338790893555, "learning_rate": 2.4475000000000002e-05, "loss": 0.0882, "step": 980 }, { "grad_norm": 2.0631558895111084, "learning_rate": 2.4725e-05, "loss": 0.0888, "step": 990 }, { "grad_norm": 1.9578195810317993, "learning_rate": 2.4975e-05, "loss": 0.0925, "step": 1000 }, { "grad_norm": 1.7717386484146118, "learning_rate": 2.5225e-05, "loss": 0.0859, "step": 1010 }, { "grad_norm": 1.5248647928237915, "learning_rate": 2.5475e-05, "loss": 0.0888, "step": 1020 }, { "grad_norm": 1.330062747001648, "learning_rate": 2.5725e-05, "loss": 0.0937, "step": 1030 }, { "grad_norm": 1.8169385194778442, "learning_rate": 2.5974999999999998e-05, "loss": 0.0816, "step": 1040 }, { "grad_norm": 1.9444684982299805, "learning_rate": 2.6225e-05, "loss": 0.0834, "step": 1050 }, { "grad_norm": 1.8416461944580078, "learning_rate": 2.6475e-05, "loss": 0.0805, "step": 1060 }, { "grad_norm": 1.514032244682312, "learning_rate": 2.6725e-05, "loss": 0.0781, "step": 1070 }, { "grad_norm": 1.554686188697815, "learning_rate": 2.6975000000000002e-05, "loss": 0.071, "step": 1080 }, { "grad_norm": 1.25917649269104, "learning_rate": 2.7225e-05, "loss": 0.0758, "step": 1090 }, { "grad_norm": 2.293268918991089, "learning_rate": 2.7475e-05, "loss": 0.0794, "step": 1100 }, { "grad_norm": 1.5006048679351807, "learning_rate": 2.7725e-05, "loss": 0.0701, "step": 1110 }, { "grad_norm": 1.5325112342834473, "learning_rate": 2.7975000000000002e-05, "loss": 0.0796, "step": 1120 }, { "grad_norm": 2.0761561393737793, "learning_rate": 2.8225e-05, "loss": 0.0783, "step": 1130 }, { "grad_norm": 1.9612795114517212, "learning_rate": 2.8475e-05, "loss": 0.0806, "step": 1140 }, { "grad_norm": 1.6686702966690063, "learning_rate": 2.8725e-05, "loss": 0.0792, "step": 1150 }, { "grad_norm": 1.6503490209579468, "learning_rate": 2.8975000000000003e-05, "loss": 0.0716, "step": 1160 }, { "grad_norm": 1.5445433855056763, "learning_rate": 2.9225000000000002e-05, "loss": 0.0723, "step": 1170 }, { "grad_norm": 1.9652950763702393, "learning_rate": 2.9475e-05, "loss": 0.0765, "step": 1180 }, { "grad_norm": 1.008411169052124, "learning_rate": 2.9725000000000004e-05, "loss": 0.0742, "step": 1190 }, { "grad_norm": 1.277388572692871, "learning_rate": 2.9975000000000004e-05, "loss": 0.0789, "step": 1200 }, { "grad_norm": 1.3899060487747192, "learning_rate": 3.0225000000000003e-05, "loss": 0.0735, "step": 1210 }, { "grad_norm": 1.3375567197799683, "learning_rate": 3.0475000000000002e-05, "loss": 0.0638, "step": 1220 }, { "grad_norm": 1.0741196870803833, "learning_rate": 3.0725e-05, "loss": 0.0658, "step": 1230 }, { "grad_norm": 1.3962572813034058, "learning_rate": 3.0975e-05, "loss": 0.0715, "step": 1240 }, { "grad_norm": 1.4199970960617065, "learning_rate": 3.122500000000001e-05, "loss": 0.0714, "step": 1250 }, { "grad_norm": 1.0287177562713623, "learning_rate": 3.1475e-05, "loss": 0.0606, "step": 1260 }, { "grad_norm": 1.3658097982406616, "learning_rate": 3.1725e-05, "loss": 0.0781, "step": 1270 }, { "grad_norm": 1.417392611503601, "learning_rate": 3.1975e-05, "loss": 0.0662, "step": 1280 }, { "grad_norm": 1.601388931274414, "learning_rate": 3.2225e-05, "loss": 0.0688, "step": 1290 }, { "grad_norm": 1.3638015985488892, "learning_rate": 3.2474999999999997e-05, "loss": 0.0694, "step": 1300 }, { "grad_norm": 1.3995943069458008, "learning_rate": 3.2725e-05, "loss": 0.0615, "step": 1310 }, { "grad_norm": 1.1054657697677612, "learning_rate": 3.2975e-05, "loss": 0.0784, "step": 1320 }, { "grad_norm": 1.4073281288146973, "learning_rate": 3.3225e-05, "loss": 0.0665, "step": 1330 }, { "grad_norm": 1.2477874755859375, "learning_rate": 3.3475e-05, "loss": 0.0557, "step": 1340 }, { "grad_norm": 1.1357970237731934, "learning_rate": 3.3725e-05, "loss": 0.068, "step": 1350 }, { "grad_norm": 0.9391803741455078, "learning_rate": 3.3975e-05, "loss": 0.0713, "step": 1360 }, { "grad_norm": 0.9041767716407776, "learning_rate": 3.4225e-05, "loss": 0.0608, "step": 1370 }, { "grad_norm": 1.0704375505447388, "learning_rate": 3.4475000000000005e-05, "loss": 0.063, "step": 1380 }, { "grad_norm": 1.1470776796340942, "learning_rate": 3.4725000000000004e-05, "loss": 0.0578, "step": 1390 }, { "grad_norm": 1.0744597911834717, "learning_rate": 3.4975e-05, "loss": 0.0596, "step": 1400 }, { "grad_norm": 1.4251841306686401, "learning_rate": 3.5225e-05, "loss": 0.059, "step": 1410 }, { "grad_norm": 0.9280602931976318, "learning_rate": 3.5475e-05, "loss": 0.0612, "step": 1420 }, { "grad_norm": 1.5519558191299438, "learning_rate": 3.5725e-05, "loss": 0.0681, "step": 1430 }, { "grad_norm": 1.1834940910339355, "learning_rate": 3.5975e-05, "loss": 0.059, "step": 1440 }, { "grad_norm": 1.5375288724899292, "learning_rate": 3.6225000000000006e-05, "loss": 0.0571, "step": 1450 }, { "grad_norm": 1.373563528060913, "learning_rate": 3.6475000000000006e-05, "loss": 0.0645, "step": 1460 }, { "grad_norm": 1.2224409580230713, "learning_rate": 3.6725000000000005e-05, "loss": 0.056, "step": 1470 }, { "grad_norm": 1.1786165237426758, "learning_rate": 3.6975000000000004e-05, "loss": 0.0524, "step": 1480 }, { "grad_norm": 1.3715639114379883, "learning_rate": 3.7225000000000004e-05, "loss": 0.0582, "step": 1490 }, { "grad_norm": 1.1368002891540527, "learning_rate": 3.7475e-05, "loss": 0.0571, "step": 1500 }, { "grad_norm": 1.2492544651031494, "learning_rate": 3.7725e-05, "loss": 0.0601, "step": 1510 }, { "grad_norm": 1.1304931640625, "learning_rate": 3.7975e-05, "loss": 0.0607, "step": 1520 }, { "grad_norm": 0.9530954957008362, "learning_rate": 3.8225e-05, "loss": 0.055, "step": 1530 }, { "grad_norm": 1.1582841873168945, "learning_rate": 3.8475e-05, "loss": 0.0601, "step": 1540 }, { "grad_norm": 1.039911150932312, "learning_rate": 3.8725e-05, "loss": 0.0597, "step": 1550 }, { "grad_norm": 1.2413557767868042, "learning_rate": 3.8975e-05, "loss": 0.0584, "step": 1560 }, { "grad_norm": 1.3195173740386963, "learning_rate": 3.9225e-05, "loss": 0.0581, "step": 1570 }, { "grad_norm": 0.7354806065559387, "learning_rate": 3.9475000000000004e-05, "loss": 0.0542, "step": 1580 }, { "grad_norm": 1.3607177734375, "learning_rate": 3.9725e-05, "loss": 0.0559, "step": 1590 }, { "grad_norm": 1.0201555490493774, "learning_rate": 3.9975e-05, "loss": 0.0589, "step": 1600 }, { "grad_norm": 0.9478843212127686, "learning_rate": 4.0225e-05, "loss": 0.0526, "step": 1610 }, { "grad_norm": 1.1804845333099365, "learning_rate": 4.0475e-05, "loss": 0.0508, "step": 1620 }, { "grad_norm": 1.2193561792373657, "learning_rate": 4.0725e-05, "loss": 0.0558, "step": 1630 }, { "grad_norm": 0.972760796546936, "learning_rate": 4.0975e-05, "loss": 0.0572, "step": 1640 }, { "grad_norm": 0.9105261564254761, "learning_rate": 4.1225e-05, "loss": 0.054, "step": 1650 }, { "grad_norm": 1.1124672889709473, "learning_rate": 4.1475000000000005e-05, "loss": 0.0539, "step": 1660 }, { "grad_norm": 1.294737458229065, "learning_rate": 4.1725000000000005e-05, "loss": 0.0587, "step": 1670 }, { "grad_norm": 1.2361412048339844, "learning_rate": 4.1975000000000004e-05, "loss": 0.0507, "step": 1680 }, { "grad_norm": 0.8178699612617493, "learning_rate": 4.2225e-05, "loss": 0.0522, "step": 1690 }, { "grad_norm": 1.061469316482544, "learning_rate": 4.2475e-05, "loss": 0.0649, "step": 1700 }, { "grad_norm": 0.8193264603614807, "learning_rate": 4.2725e-05, "loss": 0.0521, "step": 1710 }, { "grad_norm": 0.9072064161300659, "learning_rate": 4.2975e-05, "loss": 0.0572, "step": 1720 }, { "grad_norm": 1.2194528579711914, "learning_rate": 4.322500000000001e-05, "loss": 0.0547, "step": 1730 }, { "grad_norm": 1.0128321647644043, "learning_rate": 4.3475000000000006e-05, "loss": 0.0517, "step": 1740 }, { "grad_norm": 0.9794602394104004, "learning_rate": 4.3725000000000006e-05, "loss": 0.0488, "step": 1750 }, { "grad_norm": 0.7325600981712341, "learning_rate": 4.3975e-05, "loss": 0.0481, "step": 1760 }, { "grad_norm": 0.8593426942825317, "learning_rate": 4.4225e-05, "loss": 0.0623, "step": 1770 }, { "grad_norm": 1.235285997390747, "learning_rate": 4.4475e-05, "loss": 0.0588, "step": 1780 }, { "grad_norm": 0.8057830333709717, "learning_rate": 4.4725e-05, "loss": 0.0539, "step": 1790 }, { "grad_norm": 1.081629753112793, "learning_rate": 4.4975e-05, "loss": 0.0589, "step": 1800 }, { "grad_norm": 0.9073432087898254, "learning_rate": 4.5225e-05, "loss": 0.0533, "step": 1810 }, { "grad_norm": 0.7307747602462769, "learning_rate": 4.5475e-05, "loss": 0.0499, "step": 1820 }, { "grad_norm": 0.9531450271606445, "learning_rate": 4.5725e-05, "loss": 0.0477, "step": 1830 }, { "grad_norm": 0.9912897348403931, "learning_rate": 4.5975e-05, "loss": 0.0466, "step": 1840 }, { "grad_norm": 1.073301076889038, "learning_rate": 4.6225e-05, "loss": 0.0498, "step": 1850 }, { "grad_norm": 1.020856499671936, "learning_rate": 4.6475000000000005e-05, "loss": 0.048, "step": 1860 }, { "grad_norm": 1.09601628780365, "learning_rate": 4.6725000000000004e-05, "loss": 0.0467, "step": 1870 }, { "grad_norm": 1.0110914707183838, "learning_rate": 4.6975000000000003e-05, "loss": 0.05, "step": 1880 }, { "grad_norm": 0.7108571529388428, "learning_rate": 4.7225e-05, "loss": 0.043, "step": 1890 }, { "grad_norm": 1.2225464582443237, "learning_rate": 4.7475e-05, "loss": 0.0508, "step": 1900 }, { "grad_norm": 1.044870138168335, "learning_rate": 4.7725e-05, "loss": 0.0551, "step": 1910 }, { "grad_norm": 0.9102166295051575, "learning_rate": 4.7975e-05, "loss": 0.0472, "step": 1920 }, { "grad_norm": 1.02168607711792, "learning_rate": 4.822500000000001e-05, "loss": 0.0521, "step": 1930 }, { "grad_norm": 0.8986673355102539, "learning_rate": 4.8475000000000006e-05, "loss": 0.0579, "step": 1940 }, { "grad_norm": 0.6753257513046265, "learning_rate": 4.8725000000000005e-05, "loss": 0.0459, "step": 1950 }, { "grad_norm": 0.687861442565918, "learning_rate": 4.8975000000000005e-05, "loss": 0.0465, "step": 1960 }, { "grad_norm": 0.6262573003768921, "learning_rate": 4.9225000000000004e-05, "loss": 0.0476, "step": 1970 }, { "grad_norm": 0.942902684211731, "learning_rate": 4.9475e-05, "loss": 0.0491, "step": 1980 }, { "grad_norm": 0.7510301470756531, "learning_rate": 4.9725e-05, "loss": 0.0441, "step": 1990 }, { "grad_norm": 0.8349032402038574, "learning_rate": 4.9975e-05, "loss": 0.0469, "step": 2000 }, { "grad_norm": 0.9996463060379028, "learning_rate": 5.0225e-05, "loss": 0.0486, "step": 2010 }, { "grad_norm": 0.8656439781188965, "learning_rate": 5.047500000000001e-05, "loss": 0.0499, "step": 2020 }, { "grad_norm": 0.7092085480690002, "learning_rate": 5.0725e-05, "loss": 0.0433, "step": 2030 }, { "grad_norm": 0.7938534617424011, "learning_rate": 5.0975000000000006e-05, "loss": 0.0484, "step": 2040 }, { "grad_norm": 0.8455259203910828, "learning_rate": 5.1225e-05, "loss": 0.0487, "step": 2050 }, { "grad_norm": 0.7615934610366821, "learning_rate": 5.1475000000000004e-05, "loss": 0.0482, "step": 2060 }, { "grad_norm": 0.747049868106842, "learning_rate": 5.1725000000000004e-05, "loss": 0.0461, "step": 2070 }, { "grad_norm": 0.8067249655723572, "learning_rate": 5.197500000000001e-05, "loss": 0.0486, "step": 2080 }, { "grad_norm": 0.8234438300132751, "learning_rate": 5.2225e-05, "loss": 0.0511, "step": 2090 }, { "grad_norm": 0.8779579997062683, "learning_rate": 5.247500000000001e-05, "loss": 0.0465, "step": 2100 }, { "grad_norm": 1.170148253440857, "learning_rate": 5.2725e-05, "loss": 0.0474, "step": 2110 }, { "grad_norm": 0.7061724066734314, "learning_rate": 5.297500000000001e-05, "loss": 0.0455, "step": 2120 }, { "grad_norm": 0.5906034708023071, "learning_rate": 5.3225e-05, "loss": 0.0487, "step": 2130 }, { "grad_norm": 0.8452056050300598, "learning_rate": 5.3475e-05, "loss": 0.0458, "step": 2140 }, { "grad_norm": 0.8605411648750305, "learning_rate": 5.3725000000000005e-05, "loss": 0.0447, "step": 2150 }, { "grad_norm": 0.8327534794807434, "learning_rate": 5.3975e-05, "loss": 0.0374, "step": 2160 }, { "grad_norm": 0.9574945569038391, "learning_rate": 5.4225000000000003e-05, "loss": 0.0465, "step": 2170 }, { "grad_norm": 0.9269689321517944, "learning_rate": 5.4474999999999996e-05, "loss": 0.0468, "step": 2180 }, { "grad_norm": 0.7847471833229065, "learning_rate": 5.4725e-05, "loss": 0.0451, "step": 2190 }, { "grad_norm": 0.9259092807769775, "learning_rate": 5.4975e-05, "loss": 0.0404, "step": 2200 }, { "grad_norm": 0.7237085700035095, "learning_rate": 5.522500000000001e-05, "loss": 0.0384, "step": 2210 }, { "grad_norm": 0.7758882641792297, "learning_rate": 5.5475e-05, "loss": 0.0421, "step": 2220 }, { "grad_norm": 0.6709749102592468, "learning_rate": 5.5725000000000006e-05, "loss": 0.043, "step": 2230 }, { "grad_norm": 0.6801204681396484, "learning_rate": 5.5975e-05, "loss": 0.0425, "step": 2240 }, { "grad_norm": 0.7372690439224243, "learning_rate": 5.6225000000000005e-05, "loss": 0.0496, "step": 2250 }, { "grad_norm": 0.8312591314315796, "learning_rate": 5.6475e-05, "loss": 0.0494, "step": 2260 }, { "grad_norm": 0.8039832711219788, "learning_rate": 5.6725e-05, "loss": 0.0505, "step": 2270 }, { "grad_norm": 0.6732032895088196, "learning_rate": 5.6975e-05, "loss": 0.0422, "step": 2280 }, { "grad_norm": 0.5765743255615234, "learning_rate": 5.722500000000001e-05, "loss": 0.0388, "step": 2290 }, { "grad_norm": 0.67974454164505, "learning_rate": 5.7475e-05, "loss": 0.042, "step": 2300 }, { "grad_norm": 0.8932593464851379, "learning_rate": 5.772500000000001e-05, "loss": 0.0426, "step": 2310 }, { "grad_norm": 0.5268281698226929, "learning_rate": 5.7975e-05, "loss": 0.0416, "step": 2320 }, { "grad_norm": 0.7050196528434753, "learning_rate": 5.8225000000000006e-05, "loss": 0.0471, "step": 2330 }, { "grad_norm": 0.6511847972869873, "learning_rate": 5.8475000000000005e-05, "loss": 0.0422, "step": 2340 }, { "grad_norm": 0.8192784190177917, "learning_rate": 5.8725000000000004e-05, "loss": 0.0392, "step": 2350 }, { "grad_norm": 0.6750844717025757, "learning_rate": 5.8975000000000004e-05, "loss": 0.0448, "step": 2360 }, { "grad_norm": 0.707129716873169, "learning_rate": 5.922500000000001e-05, "loss": 0.0436, "step": 2370 }, { "grad_norm": 0.6087886095046997, "learning_rate": 5.9475e-05, "loss": 0.0426, "step": 2380 }, { "grad_norm": 0.618620753288269, "learning_rate": 5.9724999999999995e-05, "loss": 0.0411, "step": 2390 }, { "grad_norm": 0.7287383079528809, "learning_rate": 5.9975e-05, "loss": 0.0427, "step": 2400 }, { "grad_norm": 0.6331295967102051, "learning_rate": 6.0225e-05, "loss": 0.039, "step": 2410 }, { "grad_norm": 0.7426950931549072, "learning_rate": 6.0475000000000006e-05, "loss": 0.0418, "step": 2420 }, { "grad_norm": 0.5602689385414124, "learning_rate": 6.0725e-05, "loss": 0.0409, "step": 2430 }, { "grad_norm": 0.8371598124504089, "learning_rate": 6.0975000000000005e-05, "loss": 0.0415, "step": 2440 }, { "grad_norm": 0.7284623384475708, "learning_rate": 6.1225e-05, "loss": 0.0411, "step": 2450 }, { "grad_norm": 0.6360520124435425, "learning_rate": 6.1475e-05, "loss": 0.0482, "step": 2460 }, { "grad_norm": 0.7570112347602844, "learning_rate": 6.1725e-05, "loss": 0.037, "step": 2470 }, { "grad_norm": 0.7068567276000977, "learning_rate": 6.1975e-05, "loss": 0.0444, "step": 2480 }, { "grad_norm": 0.6143538951873779, "learning_rate": 6.2225e-05, "loss": 0.0399, "step": 2490 }, { "grad_norm": 0.715553879737854, "learning_rate": 6.2475e-05, "loss": 0.0423, "step": 2500 }, { "grad_norm": 0.7359753847122192, "learning_rate": 6.2725e-05, "loss": 0.0451, "step": 2510 }, { "grad_norm": 0.7386523485183716, "learning_rate": 6.297500000000001e-05, "loss": 0.0423, "step": 2520 }, { "grad_norm": 0.6692306399345398, "learning_rate": 6.3225e-05, "loss": 0.039, "step": 2530 }, { "grad_norm": 0.5462504625320435, "learning_rate": 6.347500000000001e-05, "loss": 0.038, "step": 2540 }, { "grad_norm": 0.6675525903701782, "learning_rate": 6.3725e-05, "loss": 0.0423, "step": 2550 }, { "grad_norm": 0.7169699668884277, "learning_rate": 6.397500000000001e-05, "loss": 0.0366, "step": 2560 }, { "grad_norm": 0.6262779235839844, "learning_rate": 6.4225e-05, "loss": 0.0398, "step": 2570 }, { "grad_norm": 0.6239219307899475, "learning_rate": 6.447500000000001e-05, "loss": 0.0441, "step": 2580 }, { "grad_norm": 0.6158590316772461, "learning_rate": 6.4725e-05, "loss": 0.0404, "step": 2590 }, { "grad_norm": 0.7212485074996948, "learning_rate": 6.497500000000001e-05, "loss": 0.0428, "step": 2600 }, { "grad_norm": 0.6385735273361206, "learning_rate": 6.5225e-05, "loss": 0.0487, "step": 2610 }, { "grad_norm": 0.6857914924621582, "learning_rate": 6.5475e-05, "loss": 0.0396, "step": 2620 }, { "grad_norm": 0.705255925655365, "learning_rate": 6.5725e-05, "loss": 0.0432, "step": 2630 }, { "grad_norm": 0.708908200263977, "learning_rate": 6.5975e-05, "loss": 0.0448, "step": 2640 }, { "grad_norm": 0.666921079158783, "learning_rate": 6.6225e-05, "loss": 0.0422, "step": 2650 }, { "grad_norm": 0.6128845810890198, "learning_rate": 6.6475e-05, "loss": 0.0401, "step": 2660 }, { "grad_norm": 0.7193278074264526, "learning_rate": 6.672500000000001e-05, "loss": 0.0457, "step": 2670 }, { "grad_norm": 0.790504515171051, "learning_rate": 6.6975e-05, "loss": 0.0399, "step": 2680 }, { "grad_norm": 0.7167389988899231, "learning_rate": 6.722500000000001e-05, "loss": 0.0448, "step": 2690 }, { "grad_norm": 0.6122627854347229, "learning_rate": 6.7475e-05, "loss": 0.0411, "step": 2700 }, { "grad_norm": 0.760185956954956, "learning_rate": 6.7725e-05, "loss": 0.0443, "step": 2710 }, { "grad_norm": 0.5548823475837708, "learning_rate": 6.7975e-05, "loss": 0.04, "step": 2720 }, { "grad_norm": 0.6041572093963623, "learning_rate": 6.8225e-05, "loss": 0.0413, "step": 2730 }, { "grad_norm": 0.6461893320083618, "learning_rate": 6.8475e-05, "loss": 0.041, "step": 2740 }, { "grad_norm": 0.656644880771637, "learning_rate": 6.8725e-05, "loss": 0.0404, "step": 2750 }, { "grad_norm": 0.76760333776474, "learning_rate": 6.8975e-05, "loss": 0.0405, "step": 2760 }, { "grad_norm": 0.6117868423461914, "learning_rate": 6.9225e-05, "loss": 0.035, "step": 2770 }, { "grad_norm": 0.5329188108444214, "learning_rate": 6.9475e-05, "loss": 0.0394, "step": 2780 }, { "grad_norm": 0.5637672543525696, "learning_rate": 6.9725e-05, "loss": 0.0398, "step": 2790 }, { "grad_norm": 0.601099967956543, "learning_rate": 6.997500000000001e-05, "loss": 0.0352, "step": 2800 }, { "grad_norm": 0.5599784255027771, "learning_rate": 7.022500000000001e-05, "loss": 0.0328, "step": 2810 }, { "grad_norm": 0.5154405832290649, "learning_rate": 7.0475e-05, "loss": 0.0322, "step": 2820 }, { "grad_norm": 0.570813000202179, "learning_rate": 7.072500000000001e-05, "loss": 0.0391, "step": 2830 }, { "grad_norm": 0.5124101638793945, "learning_rate": 7.0975e-05, "loss": 0.0425, "step": 2840 }, { "grad_norm": 0.7653075456619263, "learning_rate": 7.122500000000001e-05, "loss": 0.0362, "step": 2850 }, { "grad_norm": 0.5702785849571228, "learning_rate": 7.1475e-05, "loss": 0.0352, "step": 2860 }, { "grad_norm": 0.4418085813522339, "learning_rate": 7.172500000000001e-05, "loss": 0.0353, "step": 2870 }, { "grad_norm": 0.4835088849067688, "learning_rate": 7.1975e-05, "loss": 0.0347, "step": 2880 }, { "grad_norm": 0.47152000665664673, "learning_rate": 7.2225e-05, "loss": 0.0335, "step": 2890 }, { "grad_norm": 0.5707868337631226, "learning_rate": 7.2475e-05, "loss": 0.0337, "step": 2900 }, { "grad_norm": 0.7188113927841187, "learning_rate": 7.272499999999999e-05, "loss": 0.0366, "step": 2910 }, { "grad_norm": 0.5551576018333435, "learning_rate": 7.2975e-05, "loss": 0.0361, "step": 2920 }, { "grad_norm": 0.6607410907745361, "learning_rate": 7.3225e-05, "loss": 0.0394, "step": 2930 }, { "grad_norm": 0.4866035580635071, "learning_rate": 7.347500000000001e-05, "loss": 0.0434, "step": 2940 }, { "grad_norm": 0.5697668790817261, "learning_rate": 7.3725e-05, "loss": 0.0429, "step": 2950 }, { "grad_norm": 0.5855584740638733, "learning_rate": 7.397500000000001e-05, "loss": 0.0461, "step": 2960 }, { "grad_norm": 0.5771389603614807, "learning_rate": 7.4225e-05, "loss": 0.0387, "step": 2970 }, { "grad_norm": 0.6740640997886658, "learning_rate": 7.447500000000001e-05, "loss": 0.0376, "step": 2980 }, { "grad_norm": 0.57091224193573, "learning_rate": 7.4725e-05, "loss": 0.0364, "step": 2990 }, { "grad_norm": 0.5627328157424927, "learning_rate": 7.4975e-05, "loss": 0.0396, "step": 3000 }, { "grad_norm": 0.5884373188018799, "learning_rate": 7.5225e-05, "loss": 0.045, "step": 3010 }, { "grad_norm": 0.6273873448371887, "learning_rate": 7.5475e-05, "loss": 0.0388, "step": 3020 }, { "grad_norm": 0.5485155582427979, "learning_rate": 7.5725e-05, "loss": 0.0343, "step": 3030 }, { "grad_norm": 0.7143105268478394, "learning_rate": 7.5975e-05, "loss": 0.0359, "step": 3040 }, { "grad_norm": 0.5313730239868164, "learning_rate": 7.6225e-05, "loss": 0.038, "step": 3050 }, { "grad_norm": 0.5556248426437378, "learning_rate": 7.6475e-05, "loss": 0.0384, "step": 3060 }, { "grad_norm": 0.6702959537506104, "learning_rate": 7.672500000000001e-05, "loss": 0.0385, "step": 3070 }, { "grad_norm": 0.5053375363349915, "learning_rate": 7.697500000000001e-05, "loss": 0.0369, "step": 3080 }, { "grad_norm": 0.44124385714530945, "learning_rate": 7.722500000000001e-05, "loss": 0.0364, "step": 3090 }, { "grad_norm": 0.6789709329605103, "learning_rate": 7.747500000000001e-05, "loss": 0.0436, "step": 3100 }, { "grad_norm": 0.6154891848564148, "learning_rate": 7.7725e-05, "loss": 0.0456, "step": 3110 }, { "grad_norm": 0.5047471523284912, "learning_rate": 7.797500000000001e-05, "loss": 0.0337, "step": 3120 }, { "grad_norm": 0.5095394849777222, "learning_rate": 7.8225e-05, "loss": 0.0487, "step": 3130 }, { "grad_norm": 0.6081569194793701, "learning_rate": 7.8475e-05, "loss": 0.0358, "step": 3140 }, { "grad_norm": 0.5940938591957092, "learning_rate": 7.8725e-05, "loss": 0.0417, "step": 3150 }, { "grad_norm": 0.47531411051750183, "learning_rate": 7.8975e-05, "loss": 0.0352, "step": 3160 }, { "grad_norm": 0.4195316433906555, "learning_rate": 7.9225e-05, "loss": 0.0351, "step": 3170 }, { "grad_norm": 0.43882933259010315, "learning_rate": 7.9475e-05, "loss": 0.0376, "step": 3180 }, { "grad_norm": 0.6655321717262268, "learning_rate": 7.9725e-05, "loss": 0.0387, "step": 3190 }, { "grad_norm": 0.47428232431411743, "learning_rate": 7.9975e-05, "loss": 0.0348, "step": 3200 }, { "grad_norm": 0.4666725993156433, "learning_rate": 8.022500000000001e-05, "loss": 0.0336, "step": 3210 }, { "grad_norm": 0.4769156575202942, "learning_rate": 8.0475e-05, "loss": 0.0359, "step": 3220 }, { "grad_norm": 0.4183741807937622, "learning_rate": 8.072500000000001e-05, "loss": 0.0404, "step": 3230 }, { "grad_norm": 0.5680645108222961, "learning_rate": 8.0975e-05, "loss": 0.0365, "step": 3240 }, { "grad_norm": 0.45024868845939636, "learning_rate": 8.122500000000001e-05, "loss": 0.0367, "step": 3250 }, { "grad_norm": 0.6349425911903381, "learning_rate": 8.1475e-05, "loss": 0.0388, "step": 3260 }, { "grad_norm": 0.5658817291259766, "learning_rate": 8.172500000000001e-05, "loss": 0.0407, "step": 3270 }, { "grad_norm": 0.5200574994087219, "learning_rate": 8.1975e-05, "loss": 0.036, "step": 3280 }, { "grad_norm": 0.5979814529418945, "learning_rate": 8.2225e-05, "loss": 0.0359, "step": 3290 }, { "grad_norm": 0.5134994983673096, "learning_rate": 8.2475e-05, "loss": 0.037, "step": 3300 }, { "grad_norm": 0.563730776309967, "learning_rate": 8.2725e-05, "loss": 0.0359, "step": 3310 }, { "grad_norm": 0.49309366941452026, "learning_rate": 8.2975e-05, "loss": 0.0399, "step": 3320 }, { "grad_norm": 0.5473214983940125, "learning_rate": 8.3225e-05, "loss": 0.0364, "step": 3330 }, { "grad_norm": 0.5309115052223206, "learning_rate": 8.347500000000001e-05, "loss": 0.0374, "step": 3340 }, { "grad_norm": 0.7044671773910522, "learning_rate": 8.3725e-05, "loss": 0.0384, "step": 3350 }, { "grad_norm": 0.3930470049381256, "learning_rate": 8.397500000000001e-05, "loss": 0.036, "step": 3360 }, { "grad_norm": 0.35104310512542725, "learning_rate": 8.422500000000001e-05, "loss": 0.0354, "step": 3370 }, { "grad_norm": 0.5063814520835876, "learning_rate": 8.447500000000001e-05, "loss": 0.032, "step": 3380 }, { "grad_norm": 0.5145220160484314, "learning_rate": 8.4725e-05, "loss": 0.0364, "step": 3390 }, { "grad_norm": 0.39753881096839905, "learning_rate": 8.4975e-05, "loss": 0.0378, "step": 3400 }, { "grad_norm": 0.593582034111023, "learning_rate": 8.5225e-05, "loss": 0.0411, "step": 3410 }, { "grad_norm": 0.43744924664497375, "learning_rate": 8.5475e-05, "loss": 0.0342, "step": 3420 }, { "grad_norm": 0.4539826810359955, "learning_rate": 8.5725e-05, "loss": 0.031, "step": 3430 }, { "grad_norm": 0.5615444779396057, "learning_rate": 8.5975e-05, "loss": 0.0356, "step": 3440 }, { "grad_norm": 0.5211179852485657, "learning_rate": 8.6225e-05, "loss": 0.0292, "step": 3450 }, { "grad_norm": 0.38468238711357117, "learning_rate": 8.6475e-05, "loss": 0.0333, "step": 3460 }, { "grad_norm": 0.6872231960296631, "learning_rate": 8.672500000000001e-05, "loss": 0.041, "step": 3470 }, { "grad_norm": 0.5322626233100891, "learning_rate": 8.6975e-05, "loss": 0.0354, "step": 3480 }, { "grad_norm": 0.39256131649017334, "learning_rate": 8.7225e-05, "loss": 0.034, "step": 3490 }, { "grad_norm": 0.4836067855358124, "learning_rate": 8.747500000000001e-05, "loss": 0.0322, "step": 3500 }, { "grad_norm": 0.45551154017448425, "learning_rate": 8.7725e-05, "loss": 0.0376, "step": 3510 }, { "grad_norm": 0.5901736617088318, "learning_rate": 8.797500000000001e-05, "loss": 0.0336, "step": 3520 }, { "grad_norm": 0.4333467483520508, "learning_rate": 8.8225e-05, "loss": 0.0344, "step": 3530 }, { "grad_norm": 0.5899000763893127, "learning_rate": 8.847500000000001e-05, "loss": 0.0341, "step": 3540 }, { "grad_norm": 0.5923473834991455, "learning_rate": 8.8725e-05, "loss": 0.0356, "step": 3550 }, { "grad_norm": 0.5653948187828064, "learning_rate": 8.897500000000001e-05, "loss": 0.0366, "step": 3560 }, { "grad_norm": 0.593252956867218, "learning_rate": 8.9225e-05, "loss": 0.0391, "step": 3570 }, { "grad_norm": 0.47041916847229004, "learning_rate": 8.9475e-05, "loss": 0.0381, "step": 3580 }, { "grad_norm": 0.46736493706703186, "learning_rate": 8.9725e-05, "loss": 0.0326, "step": 3590 }, { "grad_norm": 0.48775869607925415, "learning_rate": 8.9975e-05, "loss": 0.0334, "step": 3600 }, { "grad_norm": 0.5731347799301147, "learning_rate": 9.0225e-05, "loss": 0.036, "step": 3610 }, { "grad_norm": 0.4869445264339447, "learning_rate": 9.0475e-05, "loss": 0.033, "step": 3620 }, { "grad_norm": 0.4496923089027405, "learning_rate": 9.072500000000001e-05, "loss": 0.0362, "step": 3630 }, { "grad_norm": 0.33777669072151184, "learning_rate": 9.0975e-05, "loss": 0.0316, "step": 3640 }, { "grad_norm": 0.5263034105300903, "learning_rate": 9.122500000000001e-05, "loss": 0.0317, "step": 3650 }, { "grad_norm": 0.49129605293273926, "learning_rate": 9.1475e-05, "loss": 0.0308, "step": 3660 }, { "grad_norm": 0.3935355246067047, "learning_rate": 9.172500000000001e-05, "loss": 0.0291, "step": 3670 }, { "grad_norm": 0.49592095613479614, "learning_rate": 9.1975e-05, "loss": 0.0308, "step": 3680 }, { "grad_norm": 0.60346519947052, "learning_rate": 9.2225e-05, "loss": 0.0382, "step": 3690 }, { "grad_norm": 0.5465278625488281, "learning_rate": 9.2475e-05, "loss": 0.0296, "step": 3700 }, { "grad_norm": 0.4588545858860016, "learning_rate": 9.2725e-05, "loss": 0.0309, "step": 3710 }, { "grad_norm": 0.35712820291519165, "learning_rate": 9.2975e-05, "loss": 0.0309, "step": 3720 }, { "grad_norm": 0.4033416211605072, "learning_rate": 9.3225e-05, "loss": 0.0344, "step": 3730 }, { "grad_norm": 0.42913585901260376, "learning_rate": 9.3475e-05, "loss": 0.0335, "step": 3740 }, { "grad_norm": 0.42991307377815247, "learning_rate": 9.3725e-05, "loss": 0.0335, "step": 3750 }, { "grad_norm": 0.5419926643371582, "learning_rate": 9.397500000000001e-05, "loss": 0.0373, "step": 3760 }, { "grad_norm": 0.5479491353034973, "learning_rate": 9.422500000000001e-05, "loss": 0.0294, "step": 3770 }, { "grad_norm": 0.6009234189987183, "learning_rate": 9.4475e-05, "loss": 0.0374, "step": 3780 }, { "grad_norm": 0.3437947630882263, "learning_rate": 9.472500000000001e-05, "loss": 0.0309, "step": 3790 }, { "grad_norm": 0.5199187994003296, "learning_rate": 9.4975e-05, "loss": 0.0441, "step": 3800 }, { "grad_norm": 0.7951790690422058, "learning_rate": 9.522500000000001e-05, "loss": 0.0481, "step": 3810 }, { "grad_norm": 0.5120330452919006, "learning_rate": 9.5475e-05, "loss": 0.047, "step": 3820 }, { "grad_norm": 0.40319350361824036, "learning_rate": 9.572500000000001e-05, "loss": 0.0372, "step": 3830 }, { "grad_norm": 0.9117347598075867, "learning_rate": 9.5975e-05, "loss": 0.0457, "step": 3840 }, { "grad_norm": 0.5029100775718689, "learning_rate": 9.622500000000001e-05, "loss": 0.0584, "step": 3850 }, { "grad_norm": 0.47153201699256897, "learning_rate": 9.6475e-05, "loss": 0.0477, "step": 3860 }, { "grad_norm": 0.394255667924881, "learning_rate": 9.6725e-05, "loss": 0.0372, "step": 3870 }, { "grad_norm": 0.5985202789306641, "learning_rate": 9.6975e-05, "loss": 0.0359, "step": 3880 }, { "grad_norm": 0.41486525535583496, "learning_rate": 9.7225e-05, "loss": 0.0386, "step": 3890 }, { "grad_norm": 0.5381838083267212, "learning_rate": 9.747500000000001e-05, "loss": 0.0412, "step": 3900 }, { "grad_norm": 0.4185585379600525, "learning_rate": 9.7725e-05, "loss": 0.0348, "step": 3910 }, { "grad_norm": 0.4898519217967987, "learning_rate": 9.797500000000001e-05, "loss": 0.0428, "step": 3920 }, { "grad_norm": 0.5101723074913025, "learning_rate": 9.8225e-05, "loss": 0.0359, "step": 3930 }, { "grad_norm": 0.5099445581436157, "learning_rate": 9.847500000000001e-05, "loss": 0.0355, "step": 3940 }, { "grad_norm": 0.5377159118652344, "learning_rate": 9.8725e-05, "loss": 0.0382, "step": 3950 }, { "grad_norm": 0.3399759531021118, "learning_rate": 9.897500000000001e-05, "loss": 0.0331, "step": 3960 }, { "grad_norm": 0.40001291036605835, "learning_rate": 9.9225e-05, "loss": 0.031, "step": 3970 }, { "grad_norm": 0.43823346495628357, "learning_rate": 9.9475e-05, "loss": 0.0313, "step": 3980 }, { "grad_norm": 0.3298901617527008, "learning_rate": 9.9725e-05, "loss": 0.0327, "step": 3990 }, { "grad_norm": 0.30759021639823914, "learning_rate": 9.9975e-05, "loss": 0.0341, "step": 4000 }, { "grad_norm": 0.48216521739959717, "learning_rate": 9.999999653982884e-05, "loss": 0.0299, "step": 4010 }, { "grad_norm": 0.4706026017665863, "learning_rate": 9.999998457874392e-05, "loss": 0.0312, "step": 4020 }, { "grad_norm": 0.5281700491905212, "learning_rate": 9.999996407402913e-05, "loss": 0.0338, "step": 4030 }, { "grad_norm": 0.4799742102622986, "learning_rate": 9.999993502568801e-05, "loss": 0.0314, "step": 4040 }, { "grad_norm": 0.4191226065158844, "learning_rate": 9.999989743372548e-05, "loss": 0.0338, "step": 4050 }, { "grad_norm": 0.4003202021121979, "learning_rate": 9.999985129814798e-05, "loss": 0.0323, "step": 4060 }, { "grad_norm": 0.41825148463249207, "learning_rate": 9.99997966189634e-05, "loss": 0.0284, "step": 4070 }, { "grad_norm": 0.3589204251766205, "learning_rate": 9.999973339618107e-05, "loss": 0.0326, "step": 4080 }, { "grad_norm": 0.4290521740913391, "learning_rate": 9.999966162981179e-05, "loss": 0.0303, "step": 4090 }, { "grad_norm": 0.3995952606201172, "learning_rate": 9.999958131986784e-05, "loss": 0.0304, "step": 4100 }, { "grad_norm": 0.49926528334617615, "learning_rate": 9.999949246636293e-05, "loss": 0.0364, "step": 4110 }, { "grad_norm": 0.3887423276901245, "learning_rate": 9.999939506931224e-05, "loss": 0.034, "step": 4120 }, { "grad_norm": 0.46500054001808167, "learning_rate": 9.999928912873243e-05, "loss": 0.0385, "step": 4130 }, { "grad_norm": 0.317947655916214, "learning_rate": 9.999917464464159e-05, "loss": 0.0326, "step": 4140 }, { "grad_norm": 0.5337103605270386, "learning_rate": 9.999905161705929e-05, "loss": 0.0337, "step": 4150 }, { "grad_norm": 0.33284932374954224, "learning_rate": 9.999892004600653e-05, "loss": 0.0325, "step": 4160 }, { "grad_norm": 0.3919568359851837, "learning_rate": 9.999877993150581e-05, "loss": 0.0355, "step": 4170 }, { "grad_norm": 0.5926645398139954, "learning_rate": 9.999863127358108e-05, "loss": 0.0337, "step": 4180 }, { "grad_norm": 0.5604860782623291, "learning_rate": 9.999847407225773e-05, "loss": 0.0384, "step": 4190 }, { "grad_norm": 0.540971577167511, "learning_rate": 9.999830832756262e-05, "loss": 0.0325, "step": 4200 }, { "grad_norm": 0.3464512526988983, "learning_rate": 9.999813403952407e-05, "loss": 0.031, "step": 4210 }, { "grad_norm": 0.48718979954719543, "learning_rate": 9.999795120817187e-05, "loss": 0.0298, "step": 4220 }, { "grad_norm": 0.3509076237678528, "learning_rate": 9.999775983353725e-05, "loss": 0.0307, "step": 4230 }, { "grad_norm": 0.40553709864616394, "learning_rate": 9.999755991565292e-05, "loss": 0.0311, "step": 4240 }, { "grad_norm": 0.4272875487804413, "learning_rate": 9.999735145455303e-05, "loss": 0.0317, "step": 4250 }, { "grad_norm": 0.41973716020584106, "learning_rate": 9.99971344502732e-05, "loss": 0.0306, "step": 4260 }, { "grad_norm": 0.3833211660385132, "learning_rate": 9.999690890285053e-05, "loss": 0.0309, "step": 4270 }, { "grad_norm": 0.5018987059593201, "learning_rate": 9.999667481232356e-05, "loss": 0.035, "step": 4280 }, { "grad_norm": 0.3636203408241272, "learning_rate": 9.999643217873225e-05, "loss": 0.0291, "step": 4290 }, { "grad_norm": 0.292222797870636, "learning_rate": 9.999618100211809e-05, "loss": 0.0256, "step": 4300 }, { "grad_norm": 0.355231910943985, "learning_rate": 9.999592128252402e-05, "loss": 0.0268, "step": 4310 }, { "grad_norm": 0.588303804397583, "learning_rate": 9.999565301999437e-05, "loss": 0.0326, "step": 4320 }, { "grad_norm": 0.4010990560054779, "learning_rate": 9.999537621457502e-05, "loss": 0.0288, "step": 4330 }, { "grad_norm": 0.4317692220211029, "learning_rate": 9.999509086631323e-05, "loss": 0.0315, "step": 4340 }, { "grad_norm": 0.41767749190330505, "learning_rate": 9.99947969752578e-05, "loss": 0.0351, "step": 4350 }, { "grad_norm": 0.3814225494861603, "learning_rate": 9.999449454145891e-05, "loss": 0.028, "step": 4360 }, { "grad_norm": 0.375503271818161, "learning_rate": 9.999418356496827e-05, "loss": 0.0302, "step": 4370 }, { "grad_norm": 0.38195741176605225, "learning_rate": 9.999386404583899e-05, "loss": 0.0324, "step": 4380 }, { "grad_norm": 0.3557645082473755, "learning_rate": 9.999353598412568e-05, "loss": 0.0368, "step": 4390 }, { "grad_norm": 0.3276844620704651, "learning_rate": 9.999319937988442e-05, "loss": 0.0308, "step": 4400 }, { "grad_norm": 0.3730248212814331, "learning_rate": 9.999285423317268e-05, "loss": 0.0313, "step": 4410 }, { "grad_norm": 0.43569743633270264, "learning_rate": 9.999250054404947e-05, "loss": 0.0333, "step": 4420 }, { "grad_norm": 0.3343174159526825, "learning_rate": 9.99921383125752e-05, "loss": 0.0288, "step": 4430 }, { "grad_norm": 0.4123683571815491, "learning_rate": 9.99917675388118e-05, "loss": 0.0309, "step": 4440 }, { "grad_norm": 0.44969648122787476, "learning_rate": 9.99913882228226e-05, "loss": 0.0321, "step": 4450 }, { "grad_norm": 0.31638529896736145, "learning_rate": 9.999100036467242e-05, "loss": 0.0291, "step": 4460 }, { "grad_norm": 0.3516879379749298, "learning_rate": 9.999060396442753e-05, "loss": 0.028, "step": 4470 }, { "grad_norm": 0.45113256573677063, "learning_rate": 9.999019902215566e-05, "loss": 0.0306, "step": 4480 }, { "grad_norm": 0.42489656805992126, "learning_rate": 9.998978553792602e-05, "loss": 0.0361, "step": 4490 }, { "grad_norm": 0.42562440037727356, "learning_rate": 9.998936351180926e-05, "loss": 0.0301, "step": 4500 }, { "grad_norm": 0.3117844760417938, "learning_rate": 9.998893294387747e-05, "loss": 0.0268, "step": 4510 }, { "grad_norm": 0.30520421266555786, "learning_rate": 9.998849383420426e-05, "loss": 0.0275, "step": 4520 }, { "grad_norm": 0.4402061402797699, "learning_rate": 9.998804618286465e-05, "loss": 0.0326, "step": 4530 }, { "grad_norm": 0.5081169009208679, "learning_rate": 9.99875899899351e-05, "loss": 0.0303, "step": 4540 }, { "grad_norm": 0.3585740029811859, "learning_rate": 9.99871252554936e-05, "loss": 0.0252, "step": 4550 }, { "grad_norm": 0.41979941725730896, "learning_rate": 9.998665197961955e-05, "loss": 0.0276, "step": 4560 }, { "grad_norm": 0.32496950030326843, "learning_rate": 9.998617016239379e-05, "loss": 0.027, "step": 4570 }, { "grad_norm": 0.3417847752571106, "learning_rate": 9.998567980389869e-05, "loss": 0.0329, "step": 4580 }, { "grad_norm": 0.2903156280517578, "learning_rate": 9.998518090421802e-05, "loss": 0.033, "step": 4590 }, { "grad_norm": 0.38048461079597473, "learning_rate": 9.998467346343703e-05, "loss": 0.0276, "step": 4600 }, { "grad_norm": 0.43306055665016174, "learning_rate": 9.998415748164243e-05, "loss": 0.0327, "step": 4610 }, { "grad_norm": 0.5074381232261658, "learning_rate": 9.998363295892238e-05, "loss": 0.0321, "step": 4620 }, { "grad_norm": 0.42333269119262695, "learning_rate": 9.998309989536652e-05, "loss": 0.0319, "step": 4630 }, { "grad_norm": 0.3368111848831177, "learning_rate": 9.998255829106593e-05, "loss": 0.0317, "step": 4640 }, { "grad_norm": 0.3365783095359802, "learning_rate": 9.998200814611316e-05, "loss": 0.0289, "step": 4650 }, { "grad_norm": 0.3627699315547943, "learning_rate": 9.998144946060219e-05, "loss": 0.0311, "step": 4660 }, { "grad_norm": 0.41448819637298584, "learning_rate": 9.998088223462852e-05, "loss": 0.029, "step": 4670 }, { "grad_norm": 0.3129977583885193, "learning_rate": 9.998030646828905e-05, "loss": 0.0263, "step": 4680 }, { "grad_norm": 0.3178776502609253, "learning_rate": 9.997972216168217e-05, "loss": 0.0277, "step": 4690 }, { "grad_norm": 0.38112378120422363, "learning_rate": 9.997912931490771e-05, "loss": 0.0393, "step": 4700 }, { "grad_norm": 0.4310973584651947, "learning_rate": 9.9978527928067e-05, "loss": 0.0317, "step": 4710 }, { "grad_norm": 0.336720734834671, "learning_rate": 9.997791800126277e-05, "loss": 0.0287, "step": 4720 }, { "grad_norm": 0.3358677327632904, "learning_rate": 9.997729953459927e-05, "loss": 0.0294, "step": 4730 }, { "grad_norm": 0.398337721824646, "learning_rate": 9.997667252818214e-05, "loss": 0.0276, "step": 4740 }, { "grad_norm": 0.38465994596481323, "learning_rate": 9.997603698211855e-05, "loss": 0.0285, "step": 4750 }, { "grad_norm": 0.36851364374160767, "learning_rate": 9.99753928965171e-05, "loss": 0.0309, "step": 4760 }, { "grad_norm": 0.34803253412246704, "learning_rate": 9.997474027148781e-05, "loss": 0.0354, "step": 4770 }, { "grad_norm": 0.360138475894928, "learning_rate": 9.997407910714223e-05, "loss": 0.0281, "step": 4780 }, { "grad_norm": 0.3004770576953888, "learning_rate": 9.997340940359332e-05, "loss": 0.0322, "step": 4790 }, { "grad_norm": 0.38833293318748474, "learning_rate": 9.997273116095552e-05, "loss": 0.0293, "step": 4800 }, { "grad_norm": 0.37431246042251587, "learning_rate": 9.997204437934473e-05, "loss": 0.0268, "step": 4810 }, { "grad_norm": 0.45171576738357544, "learning_rate": 9.997134905887829e-05, "loss": 0.0286, "step": 4820 }, { "grad_norm": 0.3620045781135559, "learning_rate": 9.997064519967501e-05, "loss": 0.0278, "step": 4830 }, { "grad_norm": 0.3968224823474884, "learning_rate": 9.996993280185517e-05, "loss": 0.0278, "step": 4840 }, { "grad_norm": 0.4062437415122986, "learning_rate": 9.99692118655405e-05, "loss": 0.0266, "step": 4850 }, { "grad_norm": 0.36219218373298645, "learning_rate": 9.996848239085417e-05, "loss": 0.025, "step": 4860 }, { "grad_norm": 0.4593404233455658, "learning_rate": 9.996774437792085e-05, "loss": 0.0299, "step": 4870 }, { "grad_norm": 0.4393312335014343, "learning_rate": 9.996699782686664e-05, "loss": 0.0279, "step": 4880 }, { "grad_norm": 0.39250463247299194, "learning_rate": 9.996624273781909e-05, "loss": 0.0275, "step": 4890 }, { "grad_norm": 0.3436809182167053, "learning_rate": 9.996547911090725e-05, "loss": 0.0269, "step": 4900 }, { "grad_norm": 0.3088705837726593, "learning_rate": 9.996470694626157e-05, "loss": 0.027, "step": 4910 }, { "grad_norm": 0.36944133043289185, "learning_rate": 9.996392624401403e-05, "loss": 0.0274, "step": 4920 }, { "grad_norm": 0.391640305519104, "learning_rate": 9.996313700429801e-05, "loss": 0.0291, "step": 4930 }, { "grad_norm": 0.3909207880496979, "learning_rate": 9.996233922724836e-05, "loss": 0.0239, "step": 4940 }, { "grad_norm": 0.45053479075431824, "learning_rate": 9.996153291300141e-05, "loss": 0.0304, "step": 4950 }, { "grad_norm": 0.39472144842147827, "learning_rate": 9.996071806169494e-05, "loss": 0.0298, "step": 4960 }, { "grad_norm": 0.35102027654647827, "learning_rate": 9.995989467346817e-05, "loss": 0.0239, "step": 4970 }, { "grad_norm": 0.43529585003852844, "learning_rate": 9.995906274846183e-05, "loss": 0.0342, "step": 4980 }, { "grad_norm": 0.32274749875068665, "learning_rate": 9.995822228681803e-05, "loss": 0.0326, "step": 4990 }, { "grad_norm": 0.42742010951042175, "learning_rate": 9.99573732886804e-05, "loss": 0.0284, "step": 5000 }, { "grad_norm": 0.3681131899356842, "learning_rate": 9.995651575419402e-05, "loss": 0.0287, "step": 5010 }, { "grad_norm": 0.30882152915000916, "learning_rate": 9.995564968350541e-05, "loss": 0.0288, "step": 5020 }, { "grad_norm": 0.2617955505847931, "learning_rate": 9.995477507676256e-05, "loss": 0.0261, "step": 5030 }, { "grad_norm": 0.3584861159324646, "learning_rate": 9.995389193411493e-05, "loss": 0.0263, "step": 5040 }, { "grad_norm": 0.31754666566848755, "learning_rate": 9.995300025571339e-05, "loss": 0.0297, "step": 5050 }, { "grad_norm": 0.4789932072162628, "learning_rate": 9.995210004171034e-05, "loss": 0.0334, "step": 5060 }, { "grad_norm": 0.33284202218055725, "learning_rate": 9.995119129225956e-05, "loss": 0.0293, "step": 5070 }, { "grad_norm": 0.3922952115535736, "learning_rate": 9.995027400751637e-05, "loss": 0.0261, "step": 5080 }, { "grad_norm": 0.3923673629760742, "learning_rate": 9.994934818763751e-05, "loss": 0.028, "step": 5090 }, { "grad_norm": 0.39014914631843567, "learning_rate": 9.994841383278115e-05, "loss": 0.0289, "step": 5100 }, { "grad_norm": 0.4603292644023895, "learning_rate": 9.994747094310695e-05, "loss": 0.0288, "step": 5110 }, { "grad_norm": 0.3044225573539734, "learning_rate": 9.994651951877604e-05, "loss": 0.0248, "step": 5120 }, { "grad_norm": 0.34288331866264343, "learning_rate": 9.994555955995099e-05, "loss": 0.0262, "step": 5130 }, { "grad_norm": 0.3759622275829315, "learning_rate": 9.994459106679581e-05, "loss": 0.0275, "step": 5140 }, { "grad_norm": 0.35717976093292236, "learning_rate": 9.994361403947603e-05, "loss": 0.0286, "step": 5150 }, { "grad_norm": 0.31545737385749817, "learning_rate": 9.994262847815854e-05, "loss": 0.0282, "step": 5160 }, { "grad_norm": 0.35188308358192444, "learning_rate": 9.99416343830118e-05, "loss": 0.0268, "step": 5170 }, { "grad_norm": 0.3677300810813904, "learning_rate": 9.994063175420565e-05, "loss": 0.0248, "step": 5180 }, { "grad_norm": 0.31444409489631653, "learning_rate": 9.99396205919114e-05, "loss": 0.0281, "step": 5190 }, { "grad_norm": 0.3165115416049957, "learning_rate": 9.993860089630185e-05, "loss": 0.0266, "step": 5200 }, { "grad_norm": 0.42597025632858276, "learning_rate": 9.993757266755123e-05, "loss": 0.0333, "step": 5210 }, { "grad_norm": 0.337284117937088, "learning_rate": 9.993653590583522e-05, "loss": 0.0323, "step": 5220 }, { "grad_norm": 0.2979143261909485, "learning_rate": 9.993549061133102e-05, "loss": 0.0274, "step": 5230 }, { "grad_norm": 0.25128647685050964, "learning_rate": 9.993443678421719e-05, "loss": 0.0233, "step": 5240 }, { "grad_norm": 0.25309261679649353, "learning_rate": 9.993337442467384e-05, "loss": 0.0269, "step": 5250 }, { "grad_norm": 0.3741142749786377, "learning_rate": 9.993230353288248e-05, "loss": 0.027, "step": 5260 }, { "grad_norm": 0.35383710265159607, "learning_rate": 9.993122410902608e-05, "loss": 0.0277, "step": 5270 }, { "grad_norm": 0.30583280324935913, "learning_rate": 9.993013615328912e-05, "loss": 0.027, "step": 5280 }, { "grad_norm": 0.401326984167099, "learning_rate": 9.992903966585747e-05, "loss": 0.0276, "step": 5290 }, { "grad_norm": 0.4585466682910919, "learning_rate": 9.992793464691852e-05, "loss": 0.031, "step": 5300 }, { "grad_norm": 0.48512768745422363, "learning_rate": 9.992682109666105e-05, "loss": 0.034, "step": 5310 }, { "grad_norm": 0.37371712923049927, "learning_rate": 9.992569901527538e-05, "loss": 0.036, "step": 5320 }, { "grad_norm": 0.3006333112716675, "learning_rate": 9.99245684029532e-05, "loss": 0.0299, "step": 5330 }, { "grad_norm": 0.4140874147415161, "learning_rate": 9.992342925988774e-05, "loss": 0.0297, "step": 5340 }, { "grad_norm": 0.31902775168418884, "learning_rate": 9.992228158627361e-05, "loss": 0.0313, "step": 5350 }, { "grad_norm": 0.38593700528144836, "learning_rate": 9.992112538230693e-05, "loss": 0.024, "step": 5360 }, { "grad_norm": 0.3624003529548645, "learning_rate": 9.991996064818527e-05, "loss": 0.0252, "step": 5370 }, { "grad_norm": 0.22848744690418243, "learning_rate": 9.991878738410768e-05, "loss": 0.0248, "step": 5380 }, { "grad_norm": 0.3102940618991852, "learning_rate": 9.991760559027457e-05, "loss": 0.0249, "step": 5390 }, { "grad_norm": 0.2751339375972748, "learning_rate": 9.991641526688793e-05, "loss": 0.0232, "step": 5400 }, { "grad_norm": 0.2799088656902313, "learning_rate": 9.991521641415113e-05, "loss": 0.0252, "step": 5410 }, { "grad_norm": 0.27651193737983704, "learning_rate": 9.991400903226904e-05, "loss": 0.0251, "step": 5420 }, { "grad_norm": 0.31949442625045776, "learning_rate": 9.991279312144794e-05, "loss": 0.0269, "step": 5430 }, { "grad_norm": 0.32307636737823486, "learning_rate": 9.991156868189564e-05, "loss": 0.0254, "step": 5440 }, { "grad_norm": 0.32550451159477234, "learning_rate": 9.991033571382131e-05, "loss": 0.0276, "step": 5450 }, { "grad_norm": 0.29613396525382996, "learning_rate": 9.990909421743569e-05, "loss": 0.0262, "step": 5460 }, { "grad_norm": 0.4013030529022217, "learning_rate": 9.990784419295085e-05, "loss": 0.0262, "step": 5470 }, { "grad_norm": 0.4156981408596039, "learning_rate": 9.990658564058044e-05, "loss": 0.0289, "step": 5480 }, { "grad_norm": 0.2900552749633789, "learning_rate": 9.990531856053948e-05, "loss": 0.0278, "step": 5490 }, { "grad_norm": 0.3005337119102478, "learning_rate": 9.99040429530445e-05, "loss": 0.0231, "step": 5500 }, { "grad_norm": 0.31993240118026733, "learning_rate": 9.990275881831346e-05, "loss": 0.0285, "step": 5510 }, { "grad_norm": 0.36711108684539795, "learning_rate": 9.990146615656577e-05, "loss": 0.0289, "step": 5520 }, { "grad_norm": 0.35662129521369934, "learning_rate": 9.990016496802233e-05, "loss": 0.0265, "step": 5530 }, { "grad_norm": 0.37341928482055664, "learning_rate": 9.989885525290548e-05, "loss": 0.0258, "step": 5540 }, { "grad_norm": 0.2864595651626587, "learning_rate": 9.989753701143897e-05, "loss": 0.0258, "step": 5550 }, { "grad_norm": 0.36464595794677734, "learning_rate": 9.989621024384812e-05, "loss": 0.0266, "step": 5560 }, { "grad_norm": 0.35958221554756165, "learning_rate": 9.989487495035959e-05, "loss": 0.0264, "step": 5570 }, { "grad_norm": 0.4331085979938507, "learning_rate": 9.989353113120156e-05, "loss": 0.0221, "step": 5580 }, { "grad_norm": 0.3496827483177185, "learning_rate": 9.989217878660366e-05, "loss": 0.0252, "step": 5590 }, { "grad_norm": 0.29307499527931213, "learning_rate": 9.989081791679695e-05, "loss": 0.024, "step": 5600 }, { "grad_norm": 0.2947484254837036, "learning_rate": 9.988944852201397e-05, "loss": 0.0274, "step": 5610 }, { "grad_norm": 0.344237357378006, "learning_rate": 9.988807060248873e-05, "loss": 0.0252, "step": 5620 }, { "grad_norm": 0.3400457501411438, "learning_rate": 9.988668415845665e-05, "loss": 0.0239, "step": 5630 }, { "grad_norm": 0.32018736004829407, "learning_rate": 9.988528919015466e-05, "loss": 0.0234, "step": 5640 }, { "grad_norm": 0.3407381772994995, "learning_rate": 9.988388569782112e-05, "loss": 0.0325, "step": 5650 }, { "grad_norm": 0.2498631626367569, "learning_rate": 9.988247368169583e-05, "loss": 0.0335, "step": 5660 }, { "grad_norm": 0.3264145255088806, "learning_rate": 9.988105314202007e-05, "loss": 0.0266, "step": 5670 }, { "grad_norm": 0.4322798252105713, "learning_rate": 9.987962407903659e-05, "loss": 0.0274, "step": 5680 }, { "grad_norm": 0.31067368388175964, "learning_rate": 9.987818649298957e-05, "loss": 0.0252, "step": 5690 }, { "grad_norm": 0.3913861811161041, "learning_rate": 9.987674038412465e-05, "loss": 0.0273, "step": 5700 }, { "grad_norm": 0.3197999894618988, "learning_rate": 9.987528575268891e-05, "loss": 0.0273, "step": 5710 }, { "grad_norm": 0.26230892539024353, "learning_rate": 9.987382259893095e-05, "loss": 0.0273, "step": 5720 }, { "grad_norm": 0.25070348381996155, "learning_rate": 9.987235092310074e-05, "loss": 0.0271, "step": 5730 }, { "grad_norm": 0.33858785033226013, "learning_rate": 9.987087072544978e-05, "loss": 0.03, "step": 5740 }, { "grad_norm": 0.2819880247116089, "learning_rate": 9.9869382006231e-05, "loss": 0.029, "step": 5750 }, { "grad_norm": 0.31801971793174744, "learning_rate": 9.986788476569875e-05, "loss": 0.0241, "step": 5760 }, { "grad_norm": 0.29983675479888916, "learning_rate": 9.986637900410887e-05, "loss": 0.0275, "step": 5770 }, { "grad_norm": 0.32354503870010376, "learning_rate": 9.986486472171869e-05, "loss": 0.0258, "step": 5780 }, { "grad_norm": 0.31958305835723877, "learning_rate": 9.986334191878692e-05, "loss": 0.0285, "step": 5790 }, { "grad_norm": 0.22586646676063538, "learning_rate": 9.986181059557378e-05, "loss": 0.026, "step": 5800 }, { "grad_norm": 0.33091819286346436, "learning_rate": 9.986027075234094e-05, "loss": 0.024, "step": 5810 }, { "grad_norm": 0.3114239275455475, "learning_rate": 9.985872238935152e-05, "loss": 0.0264, "step": 5820 }, { "grad_norm": 0.35036522150039673, "learning_rate": 9.985716550687008e-05, "loss": 0.0265, "step": 5830 }, { "grad_norm": 0.4502059519290924, "learning_rate": 9.985560010516264e-05, "loss": 0.0273, "step": 5840 }, { "grad_norm": 0.350306898355484, "learning_rate": 9.985402618449668e-05, "loss": 0.0276, "step": 5850 }, { "grad_norm": 0.26350143551826477, "learning_rate": 9.985244374514118e-05, "loss": 0.0281, "step": 5860 }, { "grad_norm": 0.22907298803329468, "learning_rate": 9.985085278736651e-05, "loss": 0.0294, "step": 5870 }, { "grad_norm": 0.3642405569553375, "learning_rate": 9.984925331144452e-05, "loss": 0.028, "step": 5880 }, { "grad_norm": 0.2540161609649658, "learning_rate": 9.984764531764851e-05, "loss": 0.024, "step": 5890 }, { "grad_norm": 0.4379922151565552, "learning_rate": 9.984602880625326e-05, "loss": 0.0272, "step": 5900 }, { "grad_norm": 0.2899673879146576, "learning_rate": 9.9844403777535e-05, "loss": 0.0265, "step": 5910 }, { "grad_norm": 0.31457987427711487, "learning_rate": 9.984277023177135e-05, "loss": 0.024, "step": 5920 }, { "grad_norm": 0.29114067554473877, "learning_rate": 9.984112816924148e-05, "loss": 0.0223, "step": 5930 }, { "grad_norm": 0.3256438672542572, "learning_rate": 9.983947759022596e-05, "loss": 0.0259, "step": 5940 }, { "grad_norm": 0.2823372781276703, "learning_rate": 9.983781849500682e-05, "loss": 0.0249, "step": 5950 }, { "grad_norm": 0.3930516839027405, "learning_rate": 9.98361508838676e-05, "loss": 0.0256, "step": 5960 }, { "grad_norm": 0.259339839220047, "learning_rate": 9.98344747570932e-05, "loss": 0.0287, "step": 5970 }, { "grad_norm": 0.3057933449745178, "learning_rate": 9.983279011497004e-05, "loss": 0.0231, "step": 5980 }, { "grad_norm": 0.3228570818901062, "learning_rate": 9.983109695778596e-05, "loss": 0.0261, "step": 5990 }, { "grad_norm": 0.379792183637619, "learning_rate": 9.982939528583032e-05, "loss": 0.0289, "step": 6000 }, { "grad_norm": 0.26643720269203186, "learning_rate": 9.982768509939385e-05, "loss": 0.0235, "step": 6010 }, { "grad_norm": 0.2883794605731964, "learning_rate": 9.982596639876879e-05, "loss": 0.0246, "step": 6020 }, { "grad_norm": 0.29494962096214294, "learning_rate": 9.982423918424881e-05, "loss": 0.0227, "step": 6030 }, { "grad_norm": 0.3254585862159729, "learning_rate": 9.982250345612908e-05, "loss": 0.0271, "step": 6040 }, { "grad_norm": 0.27856260538101196, "learning_rate": 9.982075921470611e-05, "loss": 0.0268, "step": 6050 }, { "grad_norm": 0.29870742559432983, "learning_rate": 9.981900646027802e-05, "loss": 0.0224, "step": 6060 }, { "grad_norm": 0.30274173617362976, "learning_rate": 9.981724519314425e-05, "loss": 0.0244, "step": 6070 }, { "grad_norm": 0.40528756380081177, "learning_rate": 9.981547541360581e-05, "loss": 0.0397, "step": 6080 }, { "grad_norm": 0.25435206294059753, "learning_rate": 9.981369712196508e-05, "loss": 0.0227, "step": 6090 }, { "grad_norm": 0.3897985816001892, "learning_rate": 9.981191031852592e-05, "loss": 0.0309, "step": 6100 }, { "grad_norm": 0.28670549392700195, "learning_rate": 9.981011500359362e-05, "loss": 0.0277, "step": 6110 }, { "grad_norm": 0.375581830739975, "learning_rate": 9.9808311177475e-05, "loss": 0.0327, "step": 6120 }, { "grad_norm": 0.24414566159248352, "learning_rate": 9.980649884047826e-05, "loss": 0.0266, "step": 6130 }, { "grad_norm": 0.36874401569366455, "learning_rate": 9.980467799291307e-05, "loss": 0.025, "step": 6140 }, { "grad_norm": 0.3180181384086609, "learning_rate": 9.980284863509058e-05, "loss": 0.0311, "step": 6150 }, { "grad_norm": 0.3279063403606415, "learning_rate": 9.980101076732334e-05, "loss": 0.0252, "step": 6160 }, { "grad_norm": 0.34728679060935974, "learning_rate": 9.979916438992544e-05, "loss": 0.0279, "step": 6170 }, { "grad_norm": 0.3112865388393402, "learning_rate": 9.979730950321237e-05, "loss": 0.0251, "step": 6180 }, { "grad_norm": 0.36263659596443176, "learning_rate": 9.979544610750104e-05, "loss": 0.023, "step": 6190 }, { "grad_norm": 0.30460086464881897, "learning_rate": 9.97935742031099e-05, "loss": 0.0256, "step": 6200 }, { "grad_norm": 0.26921546459198, "learning_rate": 9.979169379035878e-05, "loss": 0.0261, "step": 6210 }, { "grad_norm": 0.3886951208114624, "learning_rate": 9.978980486956899e-05, "loss": 0.0272, "step": 6220 }, { "grad_norm": 0.3299066424369812, "learning_rate": 9.978790744106332e-05, "loss": 0.0262, "step": 6230 }, { "grad_norm": 0.4326258599758148, "learning_rate": 9.978600150516594e-05, "loss": 0.0255, "step": 6240 }, { "grad_norm": 0.27258017659187317, "learning_rate": 9.978408706220259e-05, "loss": 0.029, "step": 6250 }, { "grad_norm": 0.36450499296188354, "learning_rate": 9.978216411250032e-05, "loss": 0.0295, "step": 6260 }, { "grad_norm": 0.27262213826179504, "learning_rate": 9.978023265638778e-05, "loss": 0.022, "step": 6270 }, { "grad_norm": 0.34834396839141846, "learning_rate": 9.977829269419495e-05, "loss": 0.0243, "step": 6280 }, { "grad_norm": 0.24662905931472778, "learning_rate": 9.977634422625335e-05, "loss": 0.0226, "step": 6290 }, { "grad_norm": 0.35541146993637085, "learning_rate": 9.97743872528959e-05, "loss": 0.0288, "step": 6300 }, { "grad_norm": 0.3599545955657959, "learning_rate": 9.9772421774457e-05, "loss": 0.0267, "step": 6310 }, { "grad_norm": 0.4354398846626282, "learning_rate": 9.977044779127252e-05, "loss": 0.0273, "step": 6320 }, { "grad_norm": 0.26354163885116577, "learning_rate": 9.976846530367971e-05, "loss": 0.0293, "step": 6330 }, { "grad_norm": 0.36440908908843994, "learning_rate": 9.976647431201735e-05, "loss": 0.029, "step": 6340 }, { "grad_norm": 0.3888866603374481, "learning_rate": 9.976447481662568e-05, "loss": 0.0306, "step": 6350 }, { "grad_norm": 0.37514904141426086, "learning_rate": 9.976246681784629e-05, "loss": 0.0286, "step": 6360 }, { "grad_norm": 0.3882409632205963, "learning_rate": 9.976045031602234e-05, "loss": 0.027, "step": 6370 }, { "grad_norm": 0.4316557049751282, "learning_rate": 9.975842531149837e-05, "loss": 0.0277, "step": 6380 }, { "grad_norm": 0.3385840356349945, "learning_rate": 9.975639180462043e-05, "loss": 0.0284, "step": 6390 }, { "grad_norm": 0.3733552098274231, "learning_rate": 9.975434979573596e-05, "loss": 0.0263, "step": 6400 }, { "grad_norm": 0.23819752037525177, "learning_rate": 9.97522992851939e-05, "loss": 0.0302, "step": 6410 }, { "grad_norm": 0.3387473225593567, "learning_rate": 9.975024027334461e-05, "loss": 0.0291, "step": 6420 }, { "grad_norm": 0.35960039496421814, "learning_rate": 9.974817276053993e-05, "loss": 0.0256, "step": 6430 }, { "grad_norm": 0.31575262546539307, "learning_rate": 9.974609674713315e-05, "loss": 0.0257, "step": 6440 }, { "grad_norm": 0.3795393407344818, "learning_rate": 9.9744012233479e-05, "loss": 0.0252, "step": 6450 }, { "grad_norm": 0.37793704867362976, "learning_rate": 9.974191921993366e-05, "loss": 0.0248, "step": 6460 }, { "grad_norm": 0.26252955198287964, "learning_rate": 9.973981770685474e-05, "loss": 0.0231, "step": 6470 }, { "grad_norm": 0.370292067527771, "learning_rate": 9.97377076946014e-05, "loss": 0.0261, "step": 6480 }, { "grad_norm": 0.29160672426223755, "learning_rate": 9.973558918353412e-05, "loss": 0.0252, "step": 6490 }, { "grad_norm": 0.2608610689640045, "learning_rate": 9.973346217401494e-05, "loss": 0.0238, "step": 6500 }, { "grad_norm": 0.35520651936531067, "learning_rate": 9.973132666640726e-05, "loss": 0.0228, "step": 6510 }, { "grad_norm": 0.3565365970134735, "learning_rate": 9.972918266107602e-05, "loss": 0.023, "step": 6520 }, { "grad_norm": 0.2897922098636627, "learning_rate": 9.972703015838756e-05, "loss": 0.024, "step": 6530 }, { "grad_norm": 0.3715348541736603, "learning_rate": 9.97248691587097e-05, "loss": 0.0237, "step": 6540 }, { "grad_norm": 0.2511904835700989, "learning_rate": 9.972269966241166e-05, "loss": 0.0244, "step": 6550 }, { "grad_norm": 0.3517027795314789, "learning_rate": 9.972052166986417e-05, "loss": 0.0248, "step": 6560 }, { "grad_norm": 0.344151109457016, "learning_rate": 9.971833518143938e-05, "loss": 0.0282, "step": 6570 }, { "grad_norm": 0.35950830578804016, "learning_rate": 9.971614019751093e-05, "loss": 0.0234, "step": 6580 }, { "grad_norm": 0.3476487696170807, "learning_rate": 9.971393671845383e-05, "loss": 0.0248, "step": 6590 }, { "grad_norm": 0.21499888598918915, "learning_rate": 9.971172474464464e-05, "loss": 0.0285, "step": 6600 }, { "grad_norm": 0.2658209204673767, "learning_rate": 9.97095042764613e-05, "loss": 0.0271, "step": 6610 }, { "grad_norm": 0.28278571367263794, "learning_rate": 9.970727531428324e-05, "loss": 0.0196, "step": 6620 }, { "grad_norm": 0.30720406770706177, "learning_rate": 9.970503785849132e-05, "loss": 0.0294, "step": 6630 }, { "grad_norm": 0.24014626443386078, "learning_rate": 9.970279190946788e-05, "loss": 0.0265, "step": 6640 }, { "grad_norm": 0.2770478427410126, "learning_rate": 9.970053746759667e-05, "loss": 0.0241, "step": 6650 }, { "grad_norm": 0.2145395278930664, "learning_rate": 9.969827453326292e-05, "loss": 0.0258, "step": 6660 }, { "grad_norm": 0.39843660593032837, "learning_rate": 9.969600310685332e-05, "loss": 0.0284, "step": 6670 }, { "grad_norm": 0.3248130977153778, "learning_rate": 9.969372318875596e-05, "loss": 0.0266, "step": 6680 }, { "grad_norm": 0.3566984534263611, "learning_rate": 9.969143477936043e-05, "loss": 0.0254, "step": 6690 }, { "grad_norm": 0.2511415481567383, "learning_rate": 9.968913787905775e-05, "loss": 0.0258, "step": 6700 }, { "grad_norm": 0.30998170375823975, "learning_rate": 9.968683248824045e-05, "loss": 0.0252, "step": 6710 }, { "grad_norm": 0.24141177535057068, "learning_rate": 9.968451860730238e-05, "loss": 0.0234, "step": 6720 }, { "grad_norm": 0.38670268654823303, "learning_rate": 9.968219623663896e-05, "loss": 0.0243, "step": 6730 }, { "grad_norm": 0.2794952392578125, "learning_rate": 9.967986537664702e-05, "loss": 0.0252, "step": 6740 }, { "grad_norm": 0.28264540433883667, "learning_rate": 9.967752602772483e-05, "loss": 0.022, "step": 6750 }, { "grad_norm": 0.4374961853027344, "learning_rate": 9.967517819027212e-05, "loss": 0.0222, "step": 6760 }, { "grad_norm": 0.21650555729866028, "learning_rate": 9.967282186469009e-05, "loss": 0.0225, "step": 6770 }, { "grad_norm": 0.268446147441864, "learning_rate": 9.967045705138135e-05, "loss": 0.0228, "step": 6780 }, { "grad_norm": 0.2810943126678467, "learning_rate": 9.966808375074998e-05, "loss": 0.0246, "step": 6790 }, { "grad_norm": 0.23997174203395844, "learning_rate": 9.966570196320154e-05, "loss": 0.0251, "step": 6800 }, { "grad_norm": 0.32519206404685974, "learning_rate": 9.966331168914299e-05, "loss": 0.0237, "step": 6810 }, { "grad_norm": 0.2987738847732544, "learning_rate": 9.966091292898277e-05, "loss": 0.0281, "step": 6820 }, { "grad_norm": 0.2789584994316101, "learning_rate": 9.965850568313076e-05, "loss": 0.0229, "step": 6830 }, { "grad_norm": 0.3449133038520813, "learning_rate": 9.965608995199827e-05, "loss": 0.0251, "step": 6840 }, { "grad_norm": 0.27366310358047485, "learning_rate": 9.965366573599812e-05, "loss": 0.0262, "step": 6850 }, { "grad_norm": 0.37716758251190186, "learning_rate": 9.965123303554453e-05, "loss": 0.028, "step": 6860 }, { "grad_norm": 0.25658395886421204, "learning_rate": 9.964879185105317e-05, "loss": 0.0214, "step": 6870 }, { "grad_norm": 0.3422938287258148, "learning_rate": 9.964634218294119e-05, "loss": 0.0271, "step": 6880 }, { "grad_norm": 0.25678589940071106, "learning_rate": 9.964388403162714e-05, "loss": 0.02, "step": 6890 }, { "grad_norm": 0.22919145226478577, "learning_rate": 9.96414173975311e-05, "loss": 0.0207, "step": 6900 }, { "grad_norm": 0.3126279413700104, "learning_rate": 9.963894228107451e-05, "loss": 0.0205, "step": 6910 }, { "grad_norm": 0.2794678807258606, "learning_rate": 9.963645868268032e-05, "loss": 0.0215, "step": 6920 }, { "grad_norm": 0.27449947595596313, "learning_rate": 9.963396660277289e-05, "loss": 0.0225, "step": 6930 }, { "grad_norm": 0.22117270529270172, "learning_rate": 9.963146604177807e-05, "loss": 0.0211, "step": 6940 }, { "grad_norm": 0.2592596113681793, "learning_rate": 9.962895700012311e-05, "loss": 0.0221, "step": 6950 }, { "grad_norm": 0.30732080340385437, "learning_rate": 9.962643947823677e-05, "loss": 0.023, "step": 6960 }, { "grad_norm": 0.27195966243743896, "learning_rate": 9.962391347654921e-05, "loss": 0.0204, "step": 6970 }, { "grad_norm": 0.23686400055885315, "learning_rate": 9.962137899549204e-05, "loss": 0.0239, "step": 6980 }, { "grad_norm": 0.27474284172058105, "learning_rate": 9.961883603549835e-05, "loss": 0.0213, "step": 6990 }, { "grad_norm": 0.2044084370136261, "learning_rate": 9.961628459700267e-05, "loss": 0.0221, "step": 7000 }, { "grad_norm": 0.24072042107582092, "learning_rate": 9.961372468044095e-05, "loss": 0.0268, "step": 7010 }, { "grad_norm": 0.23477791249752045, "learning_rate": 9.961115628625062e-05, "loss": 0.0252, "step": 7020 }, { "grad_norm": 0.3573685586452484, "learning_rate": 9.960857941487056e-05, "loss": 0.0219, "step": 7030 }, { "grad_norm": 0.32161006331443787, "learning_rate": 9.960599406674106e-05, "loss": 0.0214, "step": 7040 }, { "grad_norm": 0.3077229857444763, "learning_rate": 9.960340024230393e-05, "loss": 0.0223, "step": 7050 }, { "grad_norm": 0.2585599720478058, "learning_rate": 9.960079794200232e-05, "loss": 0.024, "step": 7060 }, { "grad_norm": 0.3150414526462555, "learning_rate": 9.959818716628096e-05, "loss": 0.0219, "step": 7070 }, { "grad_norm": 0.1786259412765503, "learning_rate": 9.95955679155859e-05, "loss": 0.0231, "step": 7080 }, { "grad_norm": 0.2618967294692993, "learning_rate": 9.959294019036472e-05, "loss": 0.0245, "step": 7090 }, { "grad_norm": 0.2876232862472534, "learning_rate": 9.959030399106646e-05, "loss": 0.02, "step": 7100 }, { "grad_norm": 0.35348227620124817, "learning_rate": 9.958765931814153e-05, "loss": 0.0227, "step": 7110 }, { "grad_norm": 0.34540581703186035, "learning_rate": 9.958500617204184e-05, "loss": 0.0254, "step": 7120 }, { "grad_norm": 0.26851770281791687, "learning_rate": 9.958234455322075e-05, "loss": 0.0212, "step": 7130 }, { "grad_norm": 0.29515519738197327, "learning_rate": 9.957967446213308e-05, "loss": 0.0246, "step": 7140 }, { "grad_norm": 0.26822009682655334, "learning_rate": 9.957699589923501e-05, "loss": 0.0198, "step": 7150 }, { "grad_norm": 0.30752691626548767, "learning_rate": 9.957430886498431e-05, "loss": 0.0213, "step": 7160 }, { "grad_norm": 0.2837103605270386, "learning_rate": 9.957161335984008e-05, "loss": 0.0209, "step": 7170 }, { "grad_norm": 0.3537003993988037, "learning_rate": 9.956890938426291e-05, "loss": 0.0253, "step": 7180 }, { "grad_norm": 0.23121675848960876, "learning_rate": 9.956619693871482e-05, "loss": 0.0228, "step": 7190 }, { "grad_norm": 0.27134060859680176, "learning_rate": 9.956347602365934e-05, "loss": 0.0232, "step": 7200 }, { "grad_norm": 0.27788057923316956, "learning_rate": 9.956074663956135e-05, "loss": 0.0226, "step": 7210 }, { "grad_norm": 0.3972572386264801, "learning_rate": 9.955800878688726e-05, "loss": 0.0236, "step": 7220 }, { "grad_norm": 0.3795163333415985, "learning_rate": 9.955526246610489e-05, "loss": 0.0243, "step": 7230 }, { "grad_norm": 0.33989566564559937, "learning_rate": 9.955250767768349e-05, "loss": 0.021, "step": 7240 }, { "grad_norm": 0.3438964784145355, "learning_rate": 9.95497444220938e-05, "loss": 0.0219, "step": 7250 }, { "grad_norm": 0.2264646738767624, "learning_rate": 9.954697269980797e-05, "loss": 0.0208, "step": 7260 }, { "grad_norm": 0.34659022092819214, "learning_rate": 9.954419251129962e-05, "loss": 0.0221, "step": 7270 }, { "grad_norm": 0.2360236942768097, "learning_rate": 9.95414038570438e-05, "loss": 0.0186, "step": 7280 }, { "grad_norm": 0.2904353141784668, "learning_rate": 9.953860673751703e-05, "loss": 0.0236, "step": 7290 }, { "grad_norm": 0.22490639984607697, "learning_rate": 9.953580115319725e-05, "loss": 0.0231, "step": 7300 }, { "grad_norm": 0.31087008118629456, "learning_rate": 9.953298710456387e-05, "loss": 0.0219, "step": 7310 }, { "grad_norm": 0.2216058075428009, "learning_rate": 9.953016459209771e-05, "loss": 0.0208, "step": 7320 }, { "grad_norm": 0.3037148416042328, "learning_rate": 9.952733361628108e-05, "loss": 0.0242, "step": 7330 }, { "grad_norm": 0.2669990360736847, "learning_rate": 9.952449417759772e-05, "loss": 0.0244, "step": 7340 }, { "grad_norm": 0.3125574588775635, "learning_rate": 9.952164627653279e-05, "loss": 0.0275, "step": 7350 }, { "grad_norm": 0.25893890857696533, "learning_rate": 9.951878991357292e-05, "loss": 0.0225, "step": 7360 }, { "grad_norm": 0.3092728555202484, "learning_rate": 9.951592508920622e-05, "loss": 0.0236, "step": 7370 }, { "grad_norm": 0.3345997631549835, "learning_rate": 9.951305180392219e-05, "loss": 0.0214, "step": 7380 }, { "grad_norm": 0.2702217698097229, "learning_rate": 9.951017005821178e-05, "loss": 0.0245, "step": 7390 }, { "grad_norm": 0.29302236437797546, "learning_rate": 9.95072798525674e-05, "loss": 0.0227, "step": 7400 }, { "grad_norm": 0.2868140637874603, "learning_rate": 9.950438118748293e-05, "loss": 0.0265, "step": 7410 }, { "grad_norm": 0.22417591512203217, "learning_rate": 9.950147406345366e-05, "loss": 0.023, "step": 7420 }, { "grad_norm": 0.32782605290412903, "learning_rate": 9.949855848097635e-05, "loss": 0.0293, "step": 7430 }, { "grad_norm": 0.3313603103160858, "learning_rate": 9.949563444054916e-05, "loss": 0.0272, "step": 7440 }, { "grad_norm": 0.42528483271598816, "learning_rate": 9.949270194267178e-05, "loss": 0.0258, "step": 7450 }, { "grad_norm": 0.277904748916626, "learning_rate": 9.948976098784526e-05, "loss": 0.0274, "step": 7460 }, { "grad_norm": 0.3090466260910034, "learning_rate": 9.948681157657213e-05, "loss": 0.0288, "step": 7470 }, { "grad_norm": 0.25357040762901306, "learning_rate": 9.948385370935638e-05, "loss": 0.0233, "step": 7480 }, { "grad_norm": 0.33262163400650024, "learning_rate": 9.94808873867034e-05, "loss": 0.0254, "step": 7490 }, { "grad_norm": 0.27489006519317627, "learning_rate": 9.947791260912009e-05, "loss": 0.0263, "step": 7500 }, { "grad_norm": 0.2967894375324249, "learning_rate": 9.947492937711474e-05, "loss": 0.0221, "step": 7510 }, { "grad_norm": 0.34564009308815, "learning_rate": 9.947193769119707e-05, "loss": 0.0204, "step": 7520 }, { "grad_norm": 0.3163309395313263, "learning_rate": 9.946893755187834e-05, "loss": 0.023, "step": 7530 }, { "grad_norm": 0.2322370707988739, "learning_rate": 9.946592895967115e-05, "loss": 0.0229, "step": 7540 }, { "grad_norm": 0.2782721221446991, "learning_rate": 9.94629119150896e-05, "loss": 0.0221, "step": 7550 }, { "grad_norm": 0.3066134750843048, "learning_rate": 9.94598864186492e-05, "loss": 0.0242, "step": 7560 }, { "grad_norm": 0.2289065271615982, "learning_rate": 9.945685247086696e-05, "loss": 0.021, "step": 7570 }, { "grad_norm": 0.34459826350212097, "learning_rate": 9.945381007226129e-05, "loss": 0.0248, "step": 7580 }, { "grad_norm": 0.22146965563297272, "learning_rate": 9.945075922335203e-05, "loss": 0.0188, "step": 7590 }, { "grad_norm": 0.3041273057460785, "learning_rate": 9.944769992466049e-05, "loss": 0.0195, "step": 7600 }, { "grad_norm": 0.28410831093788147, "learning_rate": 9.944463217670945e-05, "loss": 0.0213, "step": 7610 }, { "grad_norm": 0.28770512342453003, "learning_rate": 9.944155598002307e-05, "loss": 0.0229, "step": 7620 }, { "grad_norm": 0.27829545736312866, "learning_rate": 9.943847133512701e-05, "loss": 0.0206, "step": 7630 }, { "grad_norm": 0.2785937786102295, "learning_rate": 9.943537824254834e-05, "loss": 0.0274, "step": 7640 }, { "grad_norm": 0.36524567008018494, "learning_rate": 9.943227670281559e-05, "loss": 0.0254, "step": 7650 }, { "grad_norm": 0.28669843077659607, "learning_rate": 9.942916671645873e-05, "loss": 0.0187, "step": 7660 }, { "grad_norm": 0.22254566848278046, "learning_rate": 9.942604828400916e-05, "loss": 0.0212, "step": 7670 }, { "grad_norm": 0.21527403593063354, "learning_rate": 9.942292140599975e-05, "loss": 0.0212, "step": 7680 }, { "grad_norm": 0.28070175647735596, "learning_rate": 9.94197860829648e-05, "loss": 0.0221, "step": 7690 }, { "grad_norm": 0.2861798405647278, "learning_rate": 9.941664231544004e-05, "loss": 0.0219, "step": 7700 }, { "grad_norm": 0.30972611904144287, "learning_rate": 9.941349010396264e-05, "loss": 0.0221, "step": 7710 }, { "grad_norm": 0.30068710446357727, "learning_rate": 9.941032944907125e-05, "loss": 0.0185, "step": 7720 }, { "grad_norm": 0.26021018624305725, "learning_rate": 9.940716035130596e-05, "loss": 0.0216, "step": 7730 }, { "grad_norm": 0.2565775215625763, "learning_rate": 9.940398281120821e-05, "loss": 0.0217, "step": 7740 }, { "grad_norm": 0.26501956582069397, "learning_rate": 9.940079682932102e-05, "loss": 0.0213, "step": 7750 }, { "grad_norm": 0.2285107523202896, "learning_rate": 9.939760240618877e-05, "loss": 0.0227, "step": 7760 }, { "grad_norm": 0.22557523846626282, "learning_rate": 9.939439954235729e-05, "loss": 0.0188, "step": 7770 }, { "grad_norm": 0.2671408951282501, "learning_rate": 9.939118823837387e-05, "loss": 0.0212, "step": 7780 }, { "grad_norm": 0.2570679783821106, "learning_rate": 9.938796849478725e-05, "loss": 0.0193, "step": 7790 }, { "grad_norm": 0.3333740234375, "learning_rate": 9.938474031214755e-05, "loss": 0.0223, "step": 7800 }, { "grad_norm": 0.2907090485095978, "learning_rate": 9.938150369100643e-05, "loss": 0.0242, "step": 7810 }, { "grad_norm": 0.23743444681167603, "learning_rate": 9.93782586319169e-05, "loss": 0.0188, "step": 7820 }, { "grad_norm": 0.2123517096042633, "learning_rate": 9.937500513543348e-05, "loss": 0.0193, "step": 7830 }, { "grad_norm": 0.3065265119075775, "learning_rate": 9.937174320211207e-05, "loss": 0.021, "step": 7840 }, { "grad_norm": 0.2762191891670227, "learning_rate": 9.936847283251009e-05, "loss": 0.0168, "step": 7850 }, { "grad_norm": 0.3357333242893219, "learning_rate": 9.936519402718632e-05, "loss": 0.0202, "step": 7860 }, { "grad_norm": 0.3089582026004791, "learning_rate": 9.936190678670102e-05, "loss": 0.0199, "step": 7870 }, { "grad_norm": 0.23728613555431366, "learning_rate": 9.935861111161593e-05, "loss": 0.0243, "step": 7880 }, { "grad_norm": 0.25413835048675537, "learning_rate": 9.935530700249416e-05, "loss": 0.0204, "step": 7890 }, { "grad_norm": 0.24897198379039764, "learning_rate": 9.935199445990028e-05, "loss": 0.0214, "step": 7900 }, { "grad_norm": 0.2816556692123413, "learning_rate": 9.934867348440033e-05, "loss": 0.0247, "step": 7910 }, { "grad_norm": 0.2887280583381653, "learning_rate": 9.934534407656176e-05, "loss": 0.0199, "step": 7920 }, { "grad_norm": 0.2286202758550644, "learning_rate": 9.93420062369535e-05, "loss": 0.021, "step": 7930 }, { "grad_norm": 0.2372996062040329, "learning_rate": 9.933865996614589e-05, "loss": 0.0238, "step": 7940 }, { "grad_norm": 0.2959100306034088, "learning_rate": 9.933530526471068e-05, "loss": 0.0174, "step": 7950 }, { "grad_norm": 0.3087533712387085, "learning_rate": 9.933194213322114e-05, "loss": 0.0202, "step": 7960 }, { "grad_norm": 0.27480682730674744, "learning_rate": 9.932857057225192e-05, "loss": 0.0215, "step": 7970 }, { "grad_norm": 0.24546213448047638, "learning_rate": 9.932519058237912e-05, "loss": 0.0217, "step": 7980 }, { "grad_norm": 0.21679815649986267, "learning_rate": 9.932180216418032e-05, "loss": 0.0201, "step": 7990 }, { "grad_norm": 0.28650423884391785, "learning_rate": 9.931840531823446e-05, "loss": 0.0191, "step": 8000 }, { "grad_norm": 0.27260828018188477, "learning_rate": 9.9315000045122e-05, "loss": 0.0217, "step": 8010 }, { "grad_norm": 0.3404346704483032, "learning_rate": 9.931158634542481e-05, "loss": 0.0255, "step": 8020 }, { "grad_norm": 0.24018193781375885, "learning_rate": 9.930816421972617e-05, "loss": 0.0195, "step": 8030 }, { "grad_norm": 0.27012455463409424, "learning_rate": 9.930473366861086e-05, "loss": 0.0231, "step": 8040 }, { "grad_norm": 0.27340877056121826, "learning_rate": 9.930129469266505e-05, "loss": 0.0201, "step": 8050 }, { "grad_norm": 0.2590246796607971, "learning_rate": 9.929784729247638e-05, "loss": 0.0199, "step": 8060 }, { "grad_norm": 0.25534921884536743, "learning_rate": 9.929439146863389e-05, "loss": 0.02, "step": 8070 }, { "grad_norm": 0.22269654273986816, "learning_rate": 9.92909272217281e-05, "loss": 0.0233, "step": 8080 }, { "grad_norm": 0.18803994357585907, "learning_rate": 9.928745455235097e-05, "loss": 0.0191, "step": 8090 }, { "grad_norm": 0.2787957191467285, "learning_rate": 9.928397346109588e-05, "loss": 0.0218, "step": 8100 }, { "grad_norm": 0.22910179197788239, "learning_rate": 9.928048394855762e-05, "loss": 0.0254, "step": 8110 }, { "grad_norm": 0.32484856247901917, "learning_rate": 9.92769860153325e-05, "loss": 0.0234, "step": 8120 }, { "grad_norm": 0.31516340374946594, "learning_rate": 9.927347966201819e-05, "loss": 0.0253, "step": 8130 }, { "grad_norm": 0.31549862027168274, "learning_rate": 9.926996488921383e-05, "loss": 0.0239, "step": 8140 }, { "grad_norm": 0.2296345829963684, "learning_rate": 9.926644169752001e-05, "loss": 0.0216, "step": 8150 }, { "grad_norm": 0.2681436240673065, "learning_rate": 9.926291008753875e-05, "loss": 0.0228, "step": 8160 }, { "grad_norm": 0.2856753468513489, "learning_rate": 9.92593700598735e-05, "loss": 0.0199, "step": 8170 }, { "grad_norm": 0.2967939078807831, "learning_rate": 9.925582161512915e-05, "loss": 0.0209, "step": 8180 }, { "grad_norm": 0.2454097867012024, "learning_rate": 9.925226475391205e-05, "loss": 0.0239, "step": 8190 }, { "grad_norm": 0.24952168762683868, "learning_rate": 9.924869947682993e-05, "loss": 0.0213, "step": 8200 }, { "grad_norm": 0.20657168328762054, "learning_rate": 9.924512578449204e-05, "loss": 0.0192, "step": 8210 }, { "grad_norm": 0.24005642533302307, "learning_rate": 9.924154367750901e-05, "loss": 0.0186, "step": 8220 }, { "grad_norm": 0.30981290340423584, "learning_rate": 9.923795315649293e-05, "loss": 0.0189, "step": 8230 }, { "grad_norm": 0.2986063063144684, "learning_rate": 9.92343542220573e-05, "loss": 0.0197, "step": 8240 }, { "grad_norm": 0.27910709381103516, "learning_rate": 9.92307468748171e-05, "loss": 0.0207, "step": 8250 }, { "grad_norm": 0.22119122743606567, "learning_rate": 9.922713111538873e-05, "loss": 0.0215, "step": 8260 }, { "grad_norm": 0.3021310269832611, "learning_rate": 9.922350694439003e-05, "loss": 0.0243, "step": 8270 }, { "grad_norm": 0.34141966700553894, "learning_rate": 9.921987436244024e-05, "loss": 0.0244, "step": 8280 }, { "grad_norm": 0.2035844326019287, "learning_rate": 9.921623337016008e-05, "loss": 0.0237, "step": 8290 }, { "grad_norm": 0.23248742520809174, "learning_rate": 9.921258396817172e-05, "loss": 0.0184, "step": 8300 }, { "grad_norm": 0.24480168521404266, "learning_rate": 9.920892615709874e-05, "loss": 0.023, "step": 8310 }, { "grad_norm": 0.2626880407333374, "learning_rate": 9.920525993756612e-05, "loss": 0.0217, "step": 8320 }, { "grad_norm": 0.2689662277698517, "learning_rate": 9.920158531020036e-05, "loss": 0.0202, "step": 8330 }, { "grad_norm": 0.24310344457626343, "learning_rate": 9.919790227562933e-05, "loss": 0.019, "step": 8340 }, { "grad_norm": 0.24847498536109924, "learning_rate": 9.919421083448237e-05, "loss": 0.0185, "step": 8350 }, { "grad_norm": 0.32883867621421814, "learning_rate": 9.919051098739022e-05, "loss": 0.02, "step": 8360 }, { "grad_norm": 0.2536257207393646, "learning_rate": 9.918680273498514e-05, "loss": 0.0195, "step": 8370 }, { "grad_norm": 0.3084844946861267, "learning_rate": 9.918308607790072e-05, "loss": 0.0216, "step": 8380 }, { "grad_norm": 0.2515258193016052, "learning_rate": 9.917936101677205e-05, "loss": 0.0201, "step": 8390 }, { "grad_norm": 0.2887676954269409, "learning_rate": 9.917562755223564e-05, "loss": 0.0197, "step": 8400 }, { "grad_norm": 0.2965984046459198, "learning_rate": 9.917188568492944e-05, "loss": 0.0198, "step": 8410 }, { "grad_norm": 0.26960834860801697, "learning_rate": 9.916813541549283e-05, "loss": 0.0217, "step": 8420 }, { "grad_norm": 0.21253548562526703, "learning_rate": 9.916437674456663e-05, "loss": 0.0171, "step": 8430 }, { "grad_norm": 0.22866769134998322, "learning_rate": 9.916060967279308e-05, "loss": 0.0229, "step": 8440 }, { "grad_norm": 0.22732798755168915, "learning_rate": 9.91568342008159e-05, "loss": 0.0211, "step": 8450 }, { "grad_norm": 0.2719951868057251, "learning_rate": 9.915305032928019e-05, "loss": 0.0215, "step": 8460 }, { "grad_norm": 0.2719557583332062, "learning_rate": 9.914925805883253e-05, "loss": 0.0245, "step": 8470 }, { "grad_norm": 0.23572009801864624, "learning_rate": 9.914545739012088e-05, "loss": 0.0206, "step": 8480 }, { "grad_norm": 0.2699815332889557, "learning_rate": 9.91416483237947e-05, "loss": 0.0192, "step": 8490 }, { "grad_norm": 0.3344748914241791, "learning_rate": 9.913783086050485e-05, "loss": 0.023, "step": 8500 }, { "grad_norm": 0.349850594997406, "learning_rate": 9.913400500090364e-05, "loss": 0.021, "step": 8510 }, { "grad_norm": 0.2529050409793854, "learning_rate": 9.913017074564479e-05, "loss": 0.0225, "step": 8520 }, { "grad_norm": 0.21583452820777893, "learning_rate": 9.912632809538348e-05, "loss": 0.0251, "step": 8530 }, { "grad_norm": 0.2747332751750946, "learning_rate": 9.912247705077629e-05, "loss": 0.0254, "step": 8540 }, { "grad_norm": 0.3079565167427063, "learning_rate": 9.911861761248127e-05, "loss": 0.0213, "step": 8550 }, { "grad_norm": 0.35290655493736267, "learning_rate": 9.91147497811579e-05, "loss": 0.0234, "step": 8560 }, { "grad_norm": 0.2906380593776703, "learning_rate": 9.911087355746709e-05, "loss": 0.0187, "step": 8570 }, { "grad_norm": 0.2500937879085541, "learning_rate": 9.910698894207117e-05, "loss": 0.0205, "step": 8580 }, { "grad_norm": 0.302474707365036, "learning_rate": 9.910309593563392e-05, "loss": 0.0195, "step": 8590 }, { "grad_norm": 0.1974867880344391, "learning_rate": 9.909919453882057e-05, "loss": 0.0172, "step": 8600 }, { "grad_norm": 0.1995481550693512, "learning_rate": 9.90952847522977e-05, "loss": 0.0188, "step": 8610 }, { "grad_norm": 0.25510191917419434, "learning_rate": 9.909136657673346e-05, "loss": 0.019, "step": 8620 }, { "grad_norm": 0.36278414726257324, "learning_rate": 9.908744001279731e-05, "loss": 0.0163, "step": 8630 }, { "grad_norm": 0.2502078711986542, "learning_rate": 9.90835050611602e-05, "loss": 0.0191, "step": 8640 }, { "grad_norm": 0.22118718922138214, "learning_rate": 9.90795617224945e-05, "loss": 0.0198, "step": 8650 }, { "grad_norm": 0.28529566526412964, "learning_rate": 9.907560999747405e-05, "loss": 0.0216, "step": 8660 }, { "grad_norm": 0.3020645081996918, "learning_rate": 9.907164988677408e-05, "loss": 0.0228, "step": 8670 }, { "grad_norm": 0.20996075868606567, "learning_rate": 9.906768139107124e-05, "loss": 0.0214, "step": 8680 }, { "grad_norm": 0.2751201391220093, "learning_rate": 9.906370451104367e-05, "loss": 0.0206, "step": 8690 }, { "grad_norm": 0.26365649700164795, "learning_rate": 9.905971924737088e-05, "loss": 0.0189, "step": 8700 }, { "grad_norm": 0.20908552408218384, "learning_rate": 9.905572560073387e-05, "loss": 0.0178, "step": 8710 }, { "grad_norm": 0.3124561607837677, "learning_rate": 9.905172357181501e-05, "loss": 0.0168, "step": 8720 }, { "grad_norm": 0.32375624775886536, "learning_rate": 9.904771316129817e-05, "loss": 0.022, "step": 8730 }, { "grad_norm": 0.3150860071182251, "learning_rate": 9.904369436986862e-05, "loss": 0.0204, "step": 8740 }, { "grad_norm": 0.2445487231016159, "learning_rate": 9.903966719821303e-05, "loss": 0.024, "step": 8750 }, { "grad_norm": 0.3076251447200775, "learning_rate": 9.903563164701956e-05, "loss": 0.0187, "step": 8760 }, { "grad_norm": 0.23613153398036957, "learning_rate": 9.903158771697778e-05, "loss": 0.0152, "step": 8770 }, { "grad_norm": 0.3381546437740326, "learning_rate": 9.902753540877867e-05, "loss": 0.0189, "step": 8780 }, { "grad_norm": 0.23882266879081726, "learning_rate": 9.902347472311466e-05, "loss": 0.0165, "step": 8790 }, { "grad_norm": 0.2385348677635193, "learning_rate": 9.901940566067962e-05, "loss": 0.0206, "step": 8800 }, { "grad_norm": 0.2439466416835785, "learning_rate": 9.901532822216883e-05, "loss": 0.0195, "step": 8810 }, { "grad_norm": 0.24287140369415283, "learning_rate": 9.901124240827904e-05, "loss": 0.0195, "step": 8820 }, { "grad_norm": 0.29148998856544495, "learning_rate": 9.900714821970835e-05, "loss": 0.0232, "step": 8830 }, { "grad_norm": 0.3299255073070526, "learning_rate": 9.900304565715641e-05, "loss": 0.0218, "step": 8840 }, { "grad_norm": 0.33771881461143494, "learning_rate": 9.899893472132419e-05, "loss": 0.0231, "step": 8850 }, { "grad_norm": 0.3040924370288849, "learning_rate": 9.899481541291415e-05, "loss": 0.0214, "step": 8860 }, { "grad_norm": 0.2298220843076706, "learning_rate": 9.899068773263016e-05, "loss": 0.0209, "step": 8870 }, { "grad_norm": 0.2254074066877365, "learning_rate": 9.898655168117754e-05, "loss": 0.0183, "step": 8880 }, { "grad_norm": 0.29933956265449524, "learning_rate": 9.898240725926302e-05, "loss": 0.0209, "step": 8890 }, { "grad_norm": 0.33035808801651, "learning_rate": 9.897825446759478e-05, "loss": 0.0171, "step": 8900 }, { "grad_norm": 0.23102939128875732, "learning_rate": 9.897409330688241e-05, "loss": 0.0193, "step": 8910 }, { "grad_norm": 0.30756890773773193, "learning_rate": 9.896992377783692e-05, "loss": 0.0175, "step": 8920 }, { "grad_norm": 0.28994250297546387, "learning_rate": 9.89657458811708e-05, "loss": 0.0179, "step": 8930 }, { "grad_norm": 0.25517037510871887, "learning_rate": 9.896155961759792e-05, "loss": 0.0163, "step": 8940 }, { "grad_norm": 0.2774934470653534, "learning_rate": 9.895736498783361e-05, "loss": 0.0201, "step": 8950 }, { "grad_norm": 0.25017687678337097, "learning_rate": 9.895316199259462e-05, "loss": 0.0225, "step": 8960 }, { "grad_norm": 0.2593897581100464, "learning_rate": 9.894895063259909e-05, "loss": 0.0185, "step": 8970 }, { "grad_norm": 0.22133105993270874, "learning_rate": 9.894473090856667e-05, "loss": 0.0206, "step": 8980 }, { "grad_norm": 0.2913318872451782, "learning_rate": 9.894050282121839e-05, "loss": 0.0191, "step": 8990 }, { "grad_norm": 0.21827608346939087, "learning_rate": 9.893626637127668e-05, "loss": 0.0231, "step": 9000 }, { "grad_norm": 0.281335711479187, "learning_rate": 9.893202155946546e-05, "loss": 0.0218, "step": 9010 }, { "grad_norm": 0.27791860699653625, "learning_rate": 9.892776838651006e-05, "loss": 0.0173, "step": 9020 }, { "grad_norm": 0.2529829740524292, "learning_rate": 9.892350685313722e-05, "loss": 0.0207, "step": 9030 }, { "grad_norm": 0.27629387378692627, "learning_rate": 9.891923696007513e-05, "loss": 0.0219, "step": 9040 }, { "grad_norm": 0.20805124938488007, "learning_rate": 9.891495870805336e-05, "loss": 0.0171, "step": 9050 }, { "grad_norm": 0.24703986942768097, "learning_rate": 9.891067209780298e-05, "loss": 0.0165, "step": 9060 }, { "grad_norm": 0.27597948908805847, "learning_rate": 9.890637713005646e-05, "loss": 0.0192, "step": 9070 }, { "grad_norm": 0.22957709431648254, "learning_rate": 9.890207380554767e-05, "loss": 0.0193, "step": 9080 }, { "grad_norm": 0.27875223755836487, "learning_rate": 9.889776212501196e-05, "loss": 0.0195, "step": 9090 }, { "grad_norm": 0.3139076232910156, "learning_rate": 9.889344208918605e-05, "loss": 0.0202, "step": 9100 }, { "grad_norm": 0.2542824149131775, "learning_rate": 9.888911369880812e-05, "loss": 0.0196, "step": 9110 }, { "grad_norm": 0.23710522055625916, "learning_rate": 9.888477695461777e-05, "loss": 0.0183, "step": 9120 }, { "grad_norm": 0.23061971366405487, "learning_rate": 9.888043185735607e-05, "loss": 0.0172, "step": 9130 }, { "grad_norm": 0.26238566637039185, "learning_rate": 9.887607840776542e-05, "loss": 0.0188, "step": 9140 }, { "grad_norm": 0.23451204597949982, "learning_rate": 9.887171660658975e-05, "loss": 0.0159, "step": 9150 }, { "grad_norm": 0.22244088351726532, "learning_rate": 9.886734645457435e-05, "loss": 0.0182, "step": 9160 }, { "grad_norm": 0.2853870689868927, "learning_rate": 9.886296795246597e-05, "loss": 0.0183, "step": 9170 }, { "grad_norm": 0.18138374388217926, "learning_rate": 9.885858110101276e-05, "loss": 0.016, "step": 9180 }, { "grad_norm": 0.2226170152425766, "learning_rate": 9.885418590096434e-05, "loss": 0.0173, "step": 9190 }, { "grad_norm": 0.24568381905555725, "learning_rate": 9.88497823530717e-05, "loss": 0.0195, "step": 9200 }, { "grad_norm": 0.2760298550128937, "learning_rate": 9.884537045808732e-05, "loss": 0.0201, "step": 9210 }, { "grad_norm": 0.2509033977985382, "learning_rate": 9.884095021676502e-05, "loss": 0.0207, "step": 9220 }, { "grad_norm": 0.22948306798934937, "learning_rate": 9.883652162986017e-05, "loss": 0.0175, "step": 9230 }, { "grad_norm": 0.277150422334671, "learning_rate": 9.883208469812943e-05, "loss": 0.0193, "step": 9240 }, { "grad_norm": 0.23250259459018707, "learning_rate": 9.882763942233098e-05, "loss": 0.0154, "step": 9250 }, { "grad_norm": 0.35213226079940796, "learning_rate": 9.882318580322441e-05, "loss": 0.0176, "step": 9260 }, { "grad_norm": 0.2961712181568146, "learning_rate": 9.881872384157067e-05, "loss": 0.019, "step": 9270 }, { "grad_norm": 0.3069742023944855, "learning_rate": 9.881425353813225e-05, "loss": 0.0218, "step": 9280 }, { "grad_norm": 0.211487278342247, "learning_rate": 9.880977489367296e-05, "loss": 0.0179, "step": 9290 }, { "grad_norm": 0.245164692401886, "learning_rate": 9.88052879089581e-05, "loss": 0.0183, "step": 9300 }, { "grad_norm": 0.2128800004720688, "learning_rate": 9.880079258475434e-05, "loss": 0.0173, "step": 9310 }, { "grad_norm": 0.17983685433864594, "learning_rate": 9.879628892182985e-05, "loss": 0.017, "step": 9320 }, { "grad_norm": 0.190949484705925, "learning_rate": 9.879177692095416e-05, "loss": 0.0169, "step": 9330 }, { "grad_norm": 0.2834857106208801, "learning_rate": 9.878725658289825e-05, "loss": 0.0167, "step": 9340 }, { "grad_norm": 0.2794448137283325, "learning_rate": 9.878272790843454e-05, "loss": 0.0198, "step": 9350 }, { "grad_norm": 0.32287853956222534, "learning_rate": 9.877819089833682e-05, "loss": 0.0168, "step": 9360 }, { "grad_norm": 0.2888623774051666, "learning_rate": 9.877364555338038e-05, "loss": 0.0151, "step": 9370 }, { "grad_norm": 0.2719094753265381, "learning_rate": 9.876909187434186e-05, "loss": 0.0184, "step": 9380 }, { "grad_norm": 0.19259122014045715, "learning_rate": 9.876452986199939e-05, "loss": 0.0147, "step": 9390 }, { "grad_norm": 0.24287700653076172, "learning_rate": 9.875995951713248e-05, "loss": 0.0168, "step": 9400 }, { "grad_norm": 0.22550752758979797, "learning_rate": 9.875538084052207e-05, "loss": 0.0143, "step": 9410 }, { "grad_norm": 0.2501773238182068, "learning_rate": 9.875079383295053e-05, "loss": 0.0187, "step": 9420 }, { "grad_norm": 0.246961772441864, "learning_rate": 9.874619849520167e-05, "loss": 0.0197, "step": 9430 }, { "grad_norm": 0.235249862074852, "learning_rate": 9.874159482806069e-05, "loss": 0.017, "step": 9440 }, { "grad_norm": 0.2319149672985077, "learning_rate": 9.873698283231426e-05, "loss": 0.0182, "step": 9450 }, { "grad_norm": 0.2937467694282532, "learning_rate": 9.87323625087504e-05, "loss": 0.0157, "step": 9460 }, { "grad_norm": 0.20445379614830017, "learning_rate": 9.872773385815863e-05, "loss": 0.0175, "step": 9470 }, { "grad_norm": 0.23815211653709412, "learning_rate": 9.872309688132986e-05, "loss": 0.0202, "step": 9480 }, { "grad_norm": 0.2300574779510498, "learning_rate": 9.871845157905639e-05, "loss": 0.0171, "step": 9490 }, { "grad_norm": 0.2861083149909973, "learning_rate": 9.871379795213201e-05, "loss": 0.0179, "step": 9500 }, { "grad_norm": 0.35299864411354065, "learning_rate": 9.87091360013519e-05, "loss": 0.0212, "step": 9510 }, { "grad_norm": 0.22333091497421265, "learning_rate": 9.870446572751262e-05, "loss": 0.0163, "step": 9520 }, { "grad_norm": 0.2101203203201294, "learning_rate": 9.869978713141224e-05, "loss": 0.0136, "step": 9530 }, { "grad_norm": 0.18730156123638153, "learning_rate": 9.869510021385016e-05, "loss": 0.0186, "step": 9540 }, { "grad_norm": 0.24485810101032257, "learning_rate": 9.869040497562727e-05, "loss": 0.0205, "step": 9550 }, { "grad_norm": 0.31712761521339417, "learning_rate": 9.868570141754587e-05, "loss": 0.0196, "step": 9560 }, { "grad_norm": 0.21031984686851501, "learning_rate": 9.868098954040965e-05, "loss": 0.0164, "step": 9570 }, { "grad_norm": 0.2796798348426819, "learning_rate": 9.867626934502374e-05, "loss": 0.022, "step": 9580 }, { "grad_norm": 0.3201524317264557, "learning_rate": 9.86715408321947e-05, "loss": 0.0242, "step": 9590 }, { "grad_norm": 0.42331281304359436, "learning_rate": 9.86668040027305e-05, "loss": 0.0196, "step": 9600 }, { "grad_norm": 0.2271026074886322, "learning_rate": 9.866205885744053e-05, "loss": 0.0219, "step": 9610 }, { "grad_norm": 0.24539417028427124, "learning_rate": 9.865730539713563e-05, "loss": 0.018, "step": 9620 }, { "grad_norm": 0.22698096930980682, "learning_rate": 9.8652543622628e-05, "loss": 0.016, "step": 9630 }, { "grad_norm": 0.2694137990474701, "learning_rate": 9.864777353473132e-05, "loss": 0.0201, "step": 9640 }, { "grad_norm": 0.18778268992900848, "learning_rate": 9.864299513426068e-05, "loss": 0.0195, "step": 9650 }, { "grad_norm": 0.21242208778858185, "learning_rate": 9.863820842203254e-05, "loss": 0.0201, "step": 9660 }, { "grad_norm": 0.24273771047592163, "learning_rate": 9.863341339886483e-05, "loss": 0.0165, "step": 9670 }, { "grad_norm": 0.18416234850883484, "learning_rate": 9.86286100655769e-05, "loss": 0.0145, "step": 9680 }, { "grad_norm": 0.23645086586475372, "learning_rate": 9.862379842298953e-05, "loss": 0.0215, "step": 9690 }, { "grad_norm": 0.21020041406154633, "learning_rate": 9.861897847192485e-05, "loss": 0.0209, "step": 9700 }, { "grad_norm": 0.2998260259628296, "learning_rate": 9.86141502132065e-05, "loss": 0.0172, "step": 9710 }, { "grad_norm": 0.20291729271411896, "learning_rate": 9.860931364765946e-05, "loss": 0.0194, "step": 9720 }, { "grad_norm": 0.2931438982486725, "learning_rate": 9.860446877611021e-05, "loss": 0.0156, "step": 9730 }, { "grad_norm": 0.20536458492279053, "learning_rate": 9.859961559938655e-05, "loss": 0.0172, "step": 9740 }, { "grad_norm": 0.21749648451805115, "learning_rate": 9.85947541183178e-05, "loss": 0.0197, "step": 9750 }, { "grad_norm": 0.23149776458740234, "learning_rate": 9.858988433373463e-05, "loss": 0.0193, "step": 9760 }, { "grad_norm": 0.2803376615047455, "learning_rate": 9.858500624646918e-05, "loss": 0.0187, "step": 9770 }, { "grad_norm": 0.31307947635650635, "learning_rate": 9.858011985735497e-05, "loss": 0.0165, "step": 9780 }, { "grad_norm": 0.2614188492298126, "learning_rate": 9.857522516722693e-05, "loss": 0.0176, "step": 9790 }, { "grad_norm": 0.2165190726518631, "learning_rate": 9.857032217692145e-05, "loss": 0.0201, "step": 9800 }, { "grad_norm": 0.2573375105857849, "learning_rate": 9.856541088727631e-05, "loss": 0.0258, "step": 9810 }, { "grad_norm": 0.26196473836898804, "learning_rate": 9.856049129913072e-05, "loss": 0.0203, "step": 9820 }, { "grad_norm": 0.2518160939216614, "learning_rate": 9.85555634133253e-05, "loss": 0.0214, "step": 9830 }, { "grad_norm": 0.25668537616729736, "learning_rate": 9.855062723070208e-05, "loss": 0.0203, "step": 9840 }, { "grad_norm": 0.25595083832740784, "learning_rate": 9.854568275210454e-05, "loss": 0.0198, "step": 9850 }, { "grad_norm": 0.2026129513978958, "learning_rate": 9.854072997837754e-05, "loss": 0.0163, "step": 9860 }, { "grad_norm": 0.1700882613658905, "learning_rate": 9.853576891036737e-05, "loss": 0.0141, "step": 9870 }, { "grad_norm": 0.2718278765678406, "learning_rate": 9.853079954892177e-05, "loss": 0.015, "step": 9880 }, { "grad_norm": 0.16607718169689178, "learning_rate": 9.852582189488983e-05, "loss": 0.0159, "step": 9890 }, { "grad_norm": 0.17840631306171417, "learning_rate": 9.852083594912212e-05, "loss": 0.0143, "step": 9900 }, { "grad_norm": 0.2276155799627304, "learning_rate": 9.851584171247058e-05, "loss": 0.0156, "step": 9910 }, { "grad_norm": 0.3217454254627228, "learning_rate": 9.851083918578863e-05, "loss": 0.015, "step": 9920 }, { "grad_norm": 0.27431929111480713, "learning_rate": 9.850582836993103e-05, "loss": 0.0166, "step": 9930 }, { "grad_norm": 0.24427984654903412, "learning_rate": 9.850080926575397e-05, "loss": 0.0129, "step": 9940 }, { "grad_norm": 0.2418101727962494, "learning_rate": 9.849578187411515e-05, "loss": 0.0181, "step": 9950 }, { "grad_norm": 0.25494489073753357, "learning_rate": 9.849074619587354e-05, "loss": 0.0158, "step": 9960 }, { "grad_norm": 0.19808529317378998, "learning_rate": 9.848570223188964e-05, "loss": 0.016, "step": 9970 }, { "grad_norm": 0.1782236397266388, "learning_rate": 9.848064998302531e-05, "loss": 0.0121, "step": 9980 }, { "grad_norm": 0.24685771763324738, "learning_rate": 9.847558945014386e-05, "loss": 0.0146, "step": 9990 }, { "grad_norm": 0.2394557148218155, "learning_rate": 9.847052063410996e-05, "loss": 0.0156, "step": 10000 }, { "grad_norm": 0.1754240095615387, "learning_rate": 9.846544353578977e-05, "loss": 0.0138, "step": 10010 }, { "grad_norm": 0.26807430386543274, "learning_rate": 9.846035815605081e-05, "loss": 0.0162, "step": 10020 }, { "grad_norm": 0.2516207695007324, "learning_rate": 9.845526449576204e-05, "loss": 0.0165, "step": 10030 }, { "grad_norm": 0.18897268176078796, "learning_rate": 9.845016255579383e-05, "loss": 0.0138, "step": 10040 }, { "grad_norm": 0.24694791436195374, "learning_rate": 9.844505233701794e-05, "loss": 0.0165, "step": 10050 }, { "grad_norm": 0.19386239349842072, "learning_rate": 9.843993384030757e-05, "loss": 0.0151, "step": 10060 }, { "grad_norm": 0.18242521584033966, "learning_rate": 9.843480706653737e-05, "loss": 0.015, "step": 10070 }, { "grad_norm": 0.2822193205356598, "learning_rate": 9.84296720165833e-05, "loss": 0.0233, "step": 10080 }, { "grad_norm": 0.3040410280227661, "learning_rate": 9.842452869132286e-05, "loss": 0.0167, "step": 10090 }, { "grad_norm": 0.25587883591651917, "learning_rate": 9.841937709163489e-05, "loss": 0.0157, "step": 10100 }, { "grad_norm": 0.15740805864334106, "learning_rate": 9.841421721839962e-05, "loss": 0.013, "step": 10110 }, { "grad_norm": 0.25446832180023193, "learning_rate": 9.840904907249879e-05, "loss": 0.0168, "step": 10120 }, { "grad_norm": 0.27269113063812256, "learning_rate": 9.840387265481545e-05, "loss": 0.0164, "step": 10130 }, { "grad_norm": 0.2735983431339264, "learning_rate": 9.839868796623411e-05, "loss": 0.0149, "step": 10140 }, { "grad_norm": 0.1822630912065506, "learning_rate": 9.839349500764072e-05, "loss": 0.015, "step": 10150 }, { "grad_norm": 0.22937966883182526, "learning_rate": 9.83882937799226e-05, "loss": 0.0164, "step": 10160 }, { "grad_norm": 0.24167777597904205, "learning_rate": 9.838308428396849e-05, "loss": 0.0186, "step": 10170 }, { "grad_norm": 0.2973092198371887, "learning_rate": 9.837786652066854e-05, "loss": 0.0179, "step": 10180 }, { "grad_norm": 0.2488224357366562, "learning_rate": 9.837264049091437e-05, "loss": 0.0173, "step": 10190 }, { "grad_norm": 0.22388020157814026, "learning_rate": 9.836740619559893e-05, "loss": 0.0164, "step": 10200 }, { "grad_norm": 0.2570743262767792, "learning_rate": 9.836216363561659e-05, "loss": 0.0213, "step": 10210 }, { "grad_norm": 0.2178541123867035, "learning_rate": 9.835691281186322e-05, "loss": 0.0207, "step": 10220 }, { "grad_norm": 0.21682696044445038, "learning_rate": 9.8351653725236e-05, "loss": 0.0166, "step": 10230 }, { "grad_norm": 0.19549618661403656, "learning_rate": 9.83463863766336e-05, "loss": 0.0135, "step": 10240 }, { "grad_norm": 0.22615855932235718, "learning_rate": 9.834111076695602e-05, "loss": 0.0163, "step": 10250 }, { "grad_norm": 0.2601812183856964, "learning_rate": 9.833582689710477e-05, "loss": 0.021, "step": 10260 }, { "grad_norm": 0.2073306143283844, "learning_rate": 9.833053476798268e-05, "loss": 0.0182, "step": 10270 }, { "grad_norm": 0.2721059024333954, "learning_rate": 9.832523438049404e-05, "loss": 0.02, "step": 10280 }, { "grad_norm": 0.21777226030826569, "learning_rate": 9.831992573554454e-05, "loss": 0.0143, "step": 10290 }, { "grad_norm": 0.20900002121925354, "learning_rate": 9.831460883404128e-05, "loss": 0.0169, "step": 10300 }, { "grad_norm": 0.21370741724967957, "learning_rate": 9.830928367689278e-05, "loss": 0.0155, "step": 10310 }, { "grad_norm": 0.26563024520874023, "learning_rate": 9.830395026500896e-05, "loss": 0.0155, "step": 10320 }, { "grad_norm": 0.18538890779018402, "learning_rate": 9.829860859930115e-05, "loss": 0.0174, "step": 10330 }, { "grad_norm": 0.16200962662696838, "learning_rate": 9.829325868068212e-05, "loss": 0.0136, "step": 10340 }, { "grad_norm": 0.23270055651664734, "learning_rate": 9.8287900510066e-05, "loss": 0.0196, "step": 10350 }, { "grad_norm": 0.2230253964662552, "learning_rate": 9.828253408836834e-05, "loss": 0.0185, "step": 10360 }, { "grad_norm": 0.25731655955314636, "learning_rate": 9.827715941650615e-05, "loss": 0.0159, "step": 10370 }, { "grad_norm": 0.2636203467845917, "learning_rate": 9.82717764953978e-05, "loss": 0.0136, "step": 10380 }, { "grad_norm": 0.22834895551204681, "learning_rate": 9.826638532596308e-05, "loss": 0.0152, "step": 10390 }, { "grad_norm": 0.20838871598243713, "learning_rate": 9.82609859091232e-05, "loss": 0.0141, "step": 10400 }, { "grad_norm": 0.18623574078083038, "learning_rate": 9.825557824580076e-05, "loss": 0.015, "step": 10410 }, { "grad_norm": 0.2753784954547882, "learning_rate": 9.82501623369198e-05, "loss": 0.0126, "step": 10420 }, { "grad_norm": 0.24720372259616852, "learning_rate": 9.824473818340574e-05, "loss": 0.0164, "step": 10430 }, { "grad_norm": 0.2455480843782425, "learning_rate": 9.823930578618541e-05, "loss": 0.0153, "step": 10440 }, { "grad_norm": 0.2537420690059662, "learning_rate": 9.823386514618709e-05, "loss": 0.0143, "step": 10450 }, { "grad_norm": 0.20705801248550415, "learning_rate": 9.82284162643404e-05, "loss": 0.0151, "step": 10460 }, { "grad_norm": 0.24314793944358826, "learning_rate": 9.822295914157642e-05, "loss": 0.0164, "step": 10470 }, { "grad_norm": 0.22014538943767548, "learning_rate": 9.821749377882763e-05, "loss": 0.0157, "step": 10480 }, { "grad_norm": 0.28572219610214233, "learning_rate": 9.821202017702791e-05, "loss": 0.0134, "step": 10490 }, { "grad_norm": 0.21523579955101013, "learning_rate": 9.820653833711253e-05, "loss": 0.0148, "step": 10500 }, { "grad_norm": 0.26703327894210815, "learning_rate": 9.820104826001822e-05, "loss": 0.0161, "step": 10510 }, { "grad_norm": 0.2505776882171631, "learning_rate": 9.819554994668305e-05, "loss": 0.0182, "step": 10520 }, { "grad_norm": 0.2924773693084717, "learning_rate": 9.819004339804654e-05, "loss": 0.0166, "step": 10530 }, { "grad_norm": 0.28838080167770386, "learning_rate": 9.818452861504961e-05, "loss": 0.0162, "step": 10540 }, { "grad_norm": 0.25664782524108887, "learning_rate": 9.81790055986346e-05, "loss": 0.0179, "step": 10550 }, { "grad_norm": 0.28512564301490784, "learning_rate": 9.817347434974523e-05, "loss": 0.0147, "step": 10560 }, { "grad_norm": 0.3013826012611389, "learning_rate": 9.816793486932664e-05, "loss": 0.0169, "step": 10570 }, { "grad_norm": 0.15560190379619598, "learning_rate": 9.816238715832538e-05, "loss": 0.0134, "step": 10580 }, { "grad_norm": 0.2610533535480499, "learning_rate": 9.815683121768939e-05, "loss": 0.0189, "step": 10590 }, { "grad_norm": 0.21622499823570251, "learning_rate": 9.815126704836804e-05, "loss": 0.0167, "step": 10600 }, { "grad_norm": 0.15747816860675812, "learning_rate": 9.81456946513121e-05, "loss": 0.0181, "step": 10610 }, { "grad_norm": 0.22088856995105743, "learning_rate": 9.814011402747373e-05, "loss": 0.02, "step": 10620 }, { "grad_norm": 0.21066085994243622, "learning_rate": 9.813452517780651e-05, "loss": 0.0135, "step": 10630 }, { "grad_norm": 0.24598297476768494, "learning_rate": 9.81289281032654e-05, "loss": 0.0178, "step": 10640 }, { "grad_norm": 0.2324071228504181, "learning_rate": 9.812332280480683e-05, "loss": 0.0157, "step": 10650 }, { "grad_norm": 0.20833313465118408, "learning_rate": 9.811770928338854e-05, "loss": 0.0174, "step": 10660 }, { "grad_norm": 0.2061978131532669, "learning_rate": 9.811208753996979e-05, "loss": 0.0139, "step": 10670 }, { "grad_norm": 0.23437908291816711, "learning_rate": 9.810645757551113e-05, "loss": 0.0133, "step": 10680 }, { "grad_norm": 0.23725295066833496, "learning_rate": 9.810081939097459e-05, "loss": 0.0151, "step": 10690 }, { "grad_norm": 0.24379436671733856, "learning_rate": 9.809517298732356e-05, "loss": 0.0156, "step": 10700 }, { "grad_norm": 0.23696082830429077, "learning_rate": 9.80895183655229e-05, "loss": 0.0142, "step": 10710 }, { "grad_norm": 0.24590438604354858, "learning_rate": 9.808385552653877e-05, "loss": 0.0181, "step": 10720 }, { "grad_norm": 0.2037208080291748, "learning_rate": 9.807818447133886e-05, "loss": 0.0184, "step": 10730 }, { "grad_norm": 0.24056102335453033, "learning_rate": 9.807250520089215e-05, "loss": 0.0142, "step": 10740 }, { "grad_norm": 0.2120945304632187, "learning_rate": 9.806681771616908e-05, "loss": 0.0137, "step": 10750 }, { "grad_norm": 0.16948482394218445, "learning_rate": 9.80611220181415e-05, "loss": 0.0148, "step": 10760 }, { "grad_norm": 0.2533979117870331, "learning_rate": 9.805541810778264e-05, "loss": 0.0164, "step": 10770 }, { "grad_norm": 0.195366770029068, "learning_rate": 9.804970598606716e-05, "loss": 0.0135, "step": 10780 }, { "grad_norm": 0.2806857228279114, "learning_rate": 9.804398565397106e-05, "loss": 0.0178, "step": 10790 }, { "grad_norm": 0.23171381652355194, "learning_rate": 9.803825711247183e-05, "loss": 0.0144, "step": 10800 }, { "grad_norm": 0.2068754881620407, "learning_rate": 9.803252036254831e-05, "loss": 0.0111, "step": 10810 }, { "grad_norm": 0.20503617823123932, "learning_rate": 9.802677540518076e-05, "loss": 0.0148, "step": 10820 }, { "grad_norm": 0.19369547069072723, "learning_rate": 9.802102224135081e-05, "loss": 0.0151, "step": 10830 }, { "grad_norm": 0.17833605408668518, "learning_rate": 9.801526087204155e-05, "loss": 0.0146, "step": 10840 }, { "grad_norm": 0.23266158998012543, "learning_rate": 9.800949129823743e-05, "loss": 0.0133, "step": 10850 }, { "grad_norm": 0.1919635385274887, "learning_rate": 9.80037135209243e-05, "loss": 0.0156, "step": 10860 }, { "grad_norm": 0.19318217039108276, "learning_rate": 9.799792754108946e-05, "loss": 0.0147, "step": 10870 }, { "grad_norm": 0.24286583065986633, "learning_rate": 9.799213335972152e-05, "loss": 0.0133, "step": 10880 }, { "grad_norm": 0.23550207912921906, "learning_rate": 9.798633097781058e-05, "loss": 0.0156, "step": 10890 }, { "grad_norm": 0.17499712109565735, "learning_rate": 9.79805203963481e-05, "loss": 0.0132, "step": 10900 }, { "grad_norm": 0.19854561984539032, "learning_rate": 9.797470161632697e-05, "loss": 0.0143, "step": 10910 }, { "grad_norm": 0.23960034549236298, "learning_rate": 9.796887463874145e-05, "loss": 0.0153, "step": 10920 }, { "grad_norm": 0.22385331988334656, "learning_rate": 9.796303946458718e-05, "loss": 0.0141, "step": 10930 }, { "grad_norm": 0.20773570239543915, "learning_rate": 9.795719609486127e-05, "loss": 0.0134, "step": 10940 }, { "grad_norm": 0.20105937123298645, "learning_rate": 9.795134453056219e-05, "loss": 0.0117, "step": 10950 }, { "grad_norm": 0.2722720503807068, "learning_rate": 9.794548477268979e-05, "loss": 0.0125, "step": 10960 }, { "grad_norm": 0.2009478360414505, "learning_rate": 9.793961682224537e-05, "loss": 0.0106, "step": 10970 }, { "grad_norm": 0.20811209082603455, "learning_rate": 9.793374068023156e-05, "loss": 0.013, "step": 10980 }, { "grad_norm": 0.23849129676818848, "learning_rate": 9.792785634765247e-05, "loss": 0.0168, "step": 10990 }, { "grad_norm": 0.2172967493534088, "learning_rate": 9.792196382551357e-05, "loss": 0.0158, "step": 11000 }, { "grad_norm": 0.1539393961429596, "learning_rate": 9.791606311482171e-05, "loss": 0.0129, "step": 11010 }, { "grad_norm": 0.1736888885498047, "learning_rate": 9.791015421658518e-05, "loss": 0.0137, "step": 11020 }, { "grad_norm": 0.22883132100105286, "learning_rate": 9.790423713181362e-05, "loss": 0.0147, "step": 11030 }, { "grad_norm": 0.18892978131771088, "learning_rate": 9.789831186151814e-05, "loss": 0.0129, "step": 11040 }, { "grad_norm": 0.22918939590454102, "learning_rate": 9.789237840671118e-05, "loss": 0.0143, "step": 11050 }, { "grad_norm": 0.24950075149536133, "learning_rate": 9.78864367684066e-05, "loss": 0.0141, "step": 11060 }, { "grad_norm": 0.17218132317066193, "learning_rate": 9.788048694761968e-05, "loss": 0.0132, "step": 11070 }, { "grad_norm": 0.21745890378952026, "learning_rate": 9.787452894536709e-05, "loss": 0.0167, "step": 11080 }, { "grad_norm": 0.19612179696559906, "learning_rate": 9.786856276266685e-05, "loss": 0.0137, "step": 11090 }, { "grad_norm": 0.23320898413658142, "learning_rate": 9.786258840053845e-05, "loss": 0.0159, "step": 11100 }, { "grad_norm": 0.1955249011516571, "learning_rate": 9.785660586000273e-05, "loss": 0.0135, "step": 11110 }, { "grad_norm": 0.26244667172431946, "learning_rate": 9.785061514208196e-05, "loss": 0.0146, "step": 11120 }, { "grad_norm": 0.19272613525390625, "learning_rate": 9.784461624779977e-05, "loss": 0.015, "step": 11130 }, { "grad_norm": 0.14533986151218414, "learning_rate": 9.783860917818123e-05, "loss": 0.017, "step": 11140 }, { "grad_norm": 0.2715647518634796, "learning_rate": 9.783259393425277e-05, "loss": 0.0164, "step": 11150 }, { "grad_norm": 0.1786218285560608, "learning_rate": 9.782657051704221e-05, "loss": 0.0122, "step": 11160 }, { "grad_norm": 0.2063017040491104, "learning_rate": 9.782053892757883e-05, "loss": 0.0119, "step": 11170 }, { "grad_norm": 0.1621958613395691, "learning_rate": 9.781449916689324e-05, "loss": 0.0168, "step": 11180 }, { "grad_norm": 0.2132951021194458, "learning_rate": 9.780845123601746e-05, "loss": 0.0144, "step": 11190 }, { "grad_norm": 0.14993464946746826, "learning_rate": 9.780239513598492e-05, "loss": 0.0195, "step": 11200 }, { "grad_norm": 0.23130080103874207, "learning_rate": 9.779633086783047e-05, "loss": 0.0208, "step": 11210 }, { "grad_norm": 0.3936920166015625, "learning_rate": 9.779025843259031e-05, "loss": 0.0177, "step": 11220 }, { "grad_norm": 0.2647641897201538, "learning_rate": 9.778417783130204e-05, "loss": 0.0191, "step": 11230 }, { "grad_norm": 0.2643890976905823, "learning_rate": 9.777808906500468e-05, "loss": 0.016, "step": 11240 }, { "grad_norm": 0.20167438685894012, "learning_rate": 9.777199213473862e-05, "loss": 0.0161, "step": 11250 }, { "grad_norm": 0.18461540341377258, "learning_rate": 9.77658870415457e-05, "loss": 0.0187, "step": 11260 }, { "grad_norm": 0.26040637493133545, "learning_rate": 9.775977378646906e-05, "loss": 0.0194, "step": 11270 }, { "grad_norm": 0.17105653882026672, "learning_rate": 9.775365237055331e-05, "loss": 0.0135, "step": 11280 }, { "grad_norm": 0.26176244020462036, "learning_rate": 9.774752279484445e-05, "loss": 0.0134, "step": 11290 }, { "grad_norm": 0.33665943145751953, "learning_rate": 9.774138506038984e-05, "loss": 0.0163, "step": 11300 }, { "grad_norm": 0.25891631841659546, "learning_rate": 9.773523916823826e-05, "loss": 0.0176, "step": 11310 }, { "grad_norm": 0.1912376433610916, "learning_rate": 9.772908511943986e-05, "loss": 0.0133, "step": 11320 }, { "grad_norm": 0.21398000419139862, "learning_rate": 9.77229229150462e-05, "loss": 0.0145, "step": 11330 }, { "grad_norm": 0.20456553995609283, "learning_rate": 9.771675255611024e-05, "loss": 0.0165, "step": 11340 }, { "grad_norm": 0.24459215998649597, "learning_rate": 9.771057404368632e-05, "loss": 0.0158, "step": 11350 }, { "grad_norm": 0.2462514489889145, "learning_rate": 9.770438737883018e-05, "loss": 0.0131, "step": 11360 }, { "grad_norm": 0.17075534164905548, "learning_rate": 9.769819256259898e-05, "loss": 0.0137, "step": 11370 }, { "grad_norm": 0.21827858686447144, "learning_rate": 9.769198959605119e-05, "loss": 0.0161, "step": 11380 }, { "grad_norm": 0.19395777583122253, "learning_rate": 9.768577848024678e-05, "loss": 0.0152, "step": 11390 }, { "grad_norm": 0.1649203598499298, "learning_rate": 9.767955921624702e-05, "loss": 0.0145, "step": 11400 }, { "grad_norm": 0.26081129908561707, "learning_rate": 9.767333180511465e-05, "loss": 0.0155, "step": 11410 }, { "grad_norm": 0.19940270483493805, "learning_rate": 9.766709624791373e-05, "loss": 0.0169, "step": 11420 }, { "grad_norm": 0.23676332831382751, "learning_rate": 9.766085254570975e-05, "loss": 0.0162, "step": 11430 }, { "grad_norm": 0.20112502574920654, "learning_rate": 9.76546006995696e-05, "loss": 0.0157, "step": 11440 }, { "grad_norm": 0.2697271704673767, "learning_rate": 9.764834071056155e-05, "loss": 0.0137, "step": 11450 }, { "grad_norm": 0.20241956412792206, "learning_rate": 9.764207257975526e-05, "loss": 0.0153, "step": 11460 }, { "grad_norm": 0.19349563121795654, "learning_rate": 9.763579630822179e-05, "loss": 0.0161, "step": 11470 }, { "grad_norm": 0.20074588060379028, "learning_rate": 9.762951189703356e-05, "loss": 0.0145, "step": 11480 }, { "grad_norm": 0.1609494537115097, "learning_rate": 9.762321934726442e-05, "loss": 0.0125, "step": 11490 }, { "grad_norm": 0.1819014847278595, "learning_rate": 9.761691865998959e-05, "loss": 0.0169, "step": 11500 }, { "grad_norm": 0.18950986862182617, "learning_rate": 9.76106098362857e-05, "loss": 0.0144, "step": 11510 }, { "grad_norm": 0.1943090558052063, "learning_rate": 9.760429287723072e-05, "loss": 0.0165, "step": 11520 }, { "grad_norm": 0.19565095007419586, "learning_rate": 9.759796778390406e-05, "loss": 0.0156, "step": 11530 }, { "grad_norm": 0.23411303758621216, "learning_rate": 9.759163455738653e-05, "loss": 0.016, "step": 11540 }, { "grad_norm": 0.18968619406223297, "learning_rate": 9.75852931987603e-05, "loss": 0.0114, "step": 11550 }, { "grad_norm": 0.17870382964611053, "learning_rate": 9.757894370910891e-05, "loss": 0.0178, "step": 11560 }, { "grad_norm": 0.2953875958919525, "learning_rate": 9.757258608951733e-05, "loss": 0.018, "step": 11570 }, { "grad_norm": 0.21542933583259583, "learning_rate": 9.75662203410719e-05, "loss": 0.0138, "step": 11580 }, { "grad_norm": 0.20973165333271027, "learning_rate": 9.755984646486034e-05, "loss": 0.0122, "step": 11590 }, { "grad_norm": 0.19787970185279846, "learning_rate": 9.75534644619718e-05, "loss": 0.0113, "step": 11600 }, { "grad_norm": 0.24178822338581085, "learning_rate": 9.754707433349676e-05, "loss": 0.0142, "step": 11610 }, { "grad_norm": 0.2883351743221283, "learning_rate": 9.754067608052715e-05, "loss": 0.0161, "step": 11620 }, { "grad_norm": 0.16223622858524323, "learning_rate": 9.753426970415622e-05, "loss": 0.0119, "step": 11630 }, { "grad_norm": 0.19482636451721191, "learning_rate": 9.752785520547868e-05, "loss": 0.0129, "step": 11640 }, { "grad_norm": 0.23974373936653137, "learning_rate": 9.752143258559056e-05, "loss": 0.0144, "step": 11650 }, { "grad_norm": 0.2303893119096756, "learning_rate": 9.751500184558933e-05, "loss": 0.014, "step": 11660 }, { "grad_norm": 0.17816472053527832, "learning_rate": 9.750856298657383e-05, "loss": 0.0115, "step": 11670 }, { "grad_norm": 0.17415903508663177, "learning_rate": 9.750211600964428e-05, "loss": 0.0119, "step": 11680 }, { "grad_norm": 0.2566029131412506, "learning_rate": 9.749566091590226e-05, "loss": 0.0154, "step": 11690 }, { "grad_norm": 0.21554125845432281, "learning_rate": 9.748919770645083e-05, "loss": 0.0171, "step": 11700 }, { "grad_norm": 0.2145404815673828, "learning_rate": 9.748272638239432e-05, "loss": 0.0143, "step": 11710 }, { "grad_norm": 0.2508895993232727, "learning_rate": 9.747624694483855e-05, "loss": 0.0153, "step": 11720 }, { "grad_norm": 0.18516705930233002, "learning_rate": 9.746975939489065e-05, "loss": 0.0118, "step": 11730 }, { "grad_norm": 0.11836741864681244, "learning_rate": 9.746326373365918e-05, "loss": 0.0121, "step": 11740 }, { "grad_norm": 0.2102271020412445, "learning_rate": 9.745675996225403e-05, "loss": 0.0135, "step": 11750 }, { "grad_norm": 0.19535014033317566, "learning_rate": 9.745024808178657e-05, "loss": 0.0136, "step": 11760 }, { "grad_norm": 0.17650283873081207, "learning_rate": 9.744372809336947e-05, "loss": 0.0126, "step": 11770 }, { "grad_norm": 0.18261343240737915, "learning_rate": 9.743719999811682e-05, "loss": 0.0174, "step": 11780 }, { "grad_norm": 0.2057546228170395, "learning_rate": 9.743066379714412e-05, "loss": 0.0123, "step": 11790 }, { "grad_norm": 0.19001702964305878, "learning_rate": 9.74241194915682e-05, "loss": 0.0181, "step": 11800 }, { "grad_norm": 0.17406384646892548, "learning_rate": 9.741756708250731e-05, "loss": 0.0129, "step": 11810 }, { "grad_norm": 0.3194144070148468, "learning_rate": 9.741100657108109e-05, "loss": 0.0166, "step": 11820 }, { "grad_norm": 0.236681267619133, "learning_rate": 9.740443795841054e-05, "loss": 0.0151, "step": 11830 }, { "grad_norm": 0.243289053440094, "learning_rate": 9.739786124561805e-05, "loss": 0.0127, "step": 11840 }, { "grad_norm": 0.17345502972602844, "learning_rate": 9.73912764338274e-05, "loss": 0.0152, "step": 11850 }, { "grad_norm": 0.1988266259431839, "learning_rate": 9.738468352416377e-05, "loss": 0.015, "step": 11860 }, { "grad_norm": 0.1950048953294754, "learning_rate": 9.737808251775369e-05, "loss": 0.0167, "step": 11870 }, { "grad_norm": 0.18098761141300201, "learning_rate": 9.737147341572512e-05, "loss": 0.0141, "step": 11880 }, { "grad_norm": 0.21458649635314941, "learning_rate": 9.736485621920735e-05, "loss": 0.0152, "step": 11890 }, { "grad_norm": 0.22993403673171997, "learning_rate": 9.735823092933108e-05, "loss": 0.0205, "step": 11900 }, { "grad_norm": 0.22885766625404358, "learning_rate": 9.735159754722838e-05, "loss": 0.0151, "step": 11910 }, { "grad_norm": 0.2161378413438797, "learning_rate": 9.734495607403275e-05, "loss": 0.0155, "step": 11920 }, { "grad_norm": 0.2589801847934723, "learning_rate": 9.733830651087901e-05, "loss": 0.0147, "step": 11930 }, { "grad_norm": 0.16658754646778107, "learning_rate": 9.733164885890338e-05, "loss": 0.0113, "step": 11940 }, { "grad_norm": 0.222255140542984, "learning_rate": 9.732498311924349e-05, "loss": 0.0163, "step": 11950 }, { "grad_norm": 0.19462761282920837, "learning_rate": 9.731830929303833e-05, "loss": 0.0153, "step": 11960 }, { "grad_norm": 0.25470465421676636, "learning_rate": 9.731162738142827e-05, "loss": 0.0145, "step": 11970 }, { "grad_norm": 0.2306894063949585, "learning_rate": 9.730493738555506e-05, "loss": 0.0177, "step": 11980 }, { "grad_norm": 0.24888823926448822, "learning_rate": 9.729823930656186e-05, "loss": 0.0136, "step": 11990 }, { "grad_norm": 0.19695280492305756, "learning_rate": 9.729153314559316e-05, "loss": 0.0151, "step": 12000 }, { "grad_norm": 0.1918860673904419, "learning_rate": 9.728481890379486e-05, "loss": 0.0178, "step": 12010 }, { "grad_norm": 0.25199922919273376, "learning_rate": 9.727809658231428e-05, "loss": 0.0162, "step": 12020 }, { "grad_norm": 0.30988651514053345, "learning_rate": 9.727136618230003e-05, "loss": 0.015, "step": 12030 }, { "grad_norm": 0.23671750724315643, "learning_rate": 9.726462770490219e-05, "loss": 0.0159, "step": 12040 }, { "grad_norm": 0.23443637788295746, "learning_rate": 9.725788115127214e-05, "loss": 0.0137, "step": 12050 }, { "grad_norm": 0.2114555835723877, "learning_rate": 9.725112652256274e-05, "loss": 0.0135, "step": 12060 }, { "grad_norm": 0.1787993460893631, "learning_rate": 9.724436381992812e-05, "loss": 0.0145, "step": 12070 }, { "grad_norm": 0.21207129955291748, "learning_rate": 9.723759304452387e-05, "loss": 0.0138, "step": 12080 }, { "grad_norm": 0.20612742006778717, "learning_rate": 9.72308141975069e-05, "loss": 0.0127, "step": 12090 }, { "grad_norm": 0.173976331949234, "learning_rate": 9.722402728003557e-05, "loss": 0.0132, "step": 12100 }, { "grad_norm": 0.2214282602071762, "learning_rate": 9.721723229326953e-05, "loss": 0.0142, "step": 12110 }, { "grad_norm": 0.20642225444316864, "learning_rate": 9.721042923836992e-05, "loss": 0.0159, "step": 12120 }, { "grad_norm": 0.21385575830936432, "learning_rate": 9.720361811649914e-05, "loss": 0.0136, "step": 12130 }, { "grad_norm": 0.19479818642139435, "learning_rate": 9.719679892882106e-05, "loss": 0.014, "step": 12140 }, { "grad_norm": 0.17757241427898407, "learning_rate": 9.718997167650085e-05, "loss": 0.0151, "step": 12150 }, { "grad_norm": 0.1810220330953598, "learning_rate": 9.718313636070515e-05, "loss": 0.0159, "step": 12160 }, { "grad_norm": 0.29622897505760193, "learning_rate": 9.717629298260192e-05, "loss": 0.0166, "step": 12170 }, { "grad_norm": 0.2663850486278534, "learning_rate": 9.716944154336047e-05, "loss": 0.0129, "step": 12180 }, { "grad_norm": 0.16547483205795288, "learning_rate": 9.716258204415157e-05, "loss": 0.0124, "step": 12190 }, { "grad_norm": 0.2074432522058487, "learning_rate": 9.715571448614728e-05, "loss": 0.0155, "step": 12200 }, { "grad_norm": 0.23037035763263702, "learning_rate": 9.71488388705211e-05, "loss": 0.0131, "step": 12210 }, { "grad_norm": 0.2116774469614029, "learning_rate": 9.714195519844788e-05, "loss": 0.0134, "step": 12220 }, { "grad_norm": 0.19751016795635223, "learning_rate": 9.713506347110386e-05, "loss": 0.0115, "step": 12230 }, { "grad_norm": 0.22175131738185883, "learning_rate": 9.712816368966662e-05, "loss": 0.0122, "step": 12240 }, { "grad_norm": 0.18327827751636505, "learning_rate": 9.712125585531517e-05, "loss": 0.0141, "step": 12250 }, { "grad_norm": 0.21461981534957886, "learning_rate": 9.711433996922988e-05, "loss": 0.0129, "step": 12260 }, { "grad_norm": 0.250983327627182, "learning_rate": 9.710741603259245e-05, "loss": 0.016, "step": 12270 }, { "grad_norm": 0.21501976251602173, "learning_rate": 9.710048404658603e-05, "loss": 0.0131, "step": 12280 }, { "grad_norm": 0.2281474769115448, "learning_rate": 9.709354401239508e-05, "loss": 0.0136, "step": 12290 }, { "grad_norm": 0.23707684874534607, "learning_rate": 9.708659593120546e-05, "loss": 0.014, "step": 12300 }, { "grad_norm": 0.23677802085876465, "learning_rate": 9.707963980420443e-05, "loss": 0.0151, "step": 12310 }, { "grad_norm": 0.21915605664253235, "learning_rate": 9.707267563258058e-05, "loss": 0.0117, "step": 12320 }, { "grad_norm": 0.19247928261756897, "learning_rate": 9.70657034175239e-05, "loss": 0.0135, "step": 12330 }, { "grad_norm": 0.20818576216697693, "learning_rate": 9.705872316022577e-05, "loss": 0.0159, "step": 12340 }, { "grad_norm": 0.20169655978679657, "learning_rate": 9.705173486187891e-05, "loss": 0.0124, "step": 12350 }, { "grad_norm": 0.2650585472583771, "learning_rate": 9.704473852367741e-05, "loss": 0.0181, "step": 12360 }, { "grad_norm": 0.21584799885749817, "learning_rate": 9.70377341468168e-05, "loss": 0.0152, "step": 12370 }, { "grad_norm": 0.23832382261753082, "learning_rate": 9.703072173249389e-05, "loss": 0.0161, "step": 12380 }, { "grad_norm": 0.15700982511043549, "learning_rate": 9.702370128190693e-05, "loss": 0.0142, "step": 12390 }, { "grad_norm": 0.20922964811325073, "learning_rate": 9.701667279625552e-05, "loss": 0.0162, "step": 12400 }, { "grad_norm": 0.21130944788455963, "learning_rate": 9.700963627674065e-05, "loss": 0.0148, "step": 12410 }, { "grad_norm": 0.19487608969211578, "learning_rate": 9.700259172456466e-05, "loss": 0.0136, "step": 12420 }, { "grad_norm": 0.21720550954341888, "learning_rate": 9.699553914093124e-05, "loss": 0.0118, "step": 12430 }, { "grad_norm": 0.17292998731136322, "learning_rate": 9.698847852704553e-05, "loss": 0.0133, "step": 12440 }, { "grad_norm": 0.19844205677509308, "learning_rate": 9.6981409884114e-05, "loss": 0.0202, "step": 12450 }, { "grad_norm": 0.238401398062706, "learning_rate": 9.697433321334443e-05, "loss": 0.015, "step": 12460 }, { "grad_norm": 0.22332410514354706, "learning_rate": 9.696724851594607e-05, "loss": 0.0179, "step": 12470 }, { "grad_norm": 0.20123136043548584, "learning_rate": 9.696015579312952e-05, "loss": 0.0135, "step": 12480 }, { "grad_norm": 0.19054199755191803, "learning_rate": 9.695305504610668e-05, "loss": 0.0135, "step": 12490 }, { "grad_norm": 0.2673513889312744, "learning_rate": 9.694594627609092e-05, "loss": 0.0129, "step": 12500 }, { "grad_norm": 0.21897579729557037, "learning_rate": 9.693882948429691e-05, "loss": 0.0134, "step": 12510 }, { "grad_norm": 0.20334528386592865, "learning_rate": 9.693170467194071e-05, "loss": 0.0135, "step": 12520 }, { "grad_norm": 0.1618494689464569, "learning_rate": 9.692457184023977e-05, "loss": 0.0149, "step": 12530 }, { "grad_norm": 0.29023900628089905, "learning_rate": 9.691743099041291e-05, "loss": 0.014, "step": 12540 }, { "grad_norm": 0.2053445428609848, "learning_rate": 9.691028212368027e-05, "loss": 0.014, "step": 12550 }, { "grad_norm": 0.22053773701190948, "learning_rate": 9.690312524126342e-05, "loss": 0.0121, "step": 12560 }, { "grad_norm": 0.18569473922252655, "learning_rate": 9.689596034438527e-05, "loss": 0.012, "step": 12570 }, { "grad_norm": 0.19107896089553833, "learning_rate": 9.688878743427012e-05, "loss": 0.0119, "step": 12580 }, { "grad_norm": 0.19678254425525665, "learning_rate": 9.688160651214359e-05, "loss": 0.0119, "step": 12590 }, { "grad_norm": 0.21327988803386688, "learning_rate": 9.687441757923273e-05, "loss": 0.013, "step": 12600 }, { "grad_norm": 0.21447890996932983, "learning_rate": 9.68672206367659e-05, "loss": 0.0123, "step": 12610 }, { "grad_norm": 0.2492232620716095, "learning_rate": 9.686001568597291e-05, "loss": 0.0155, "step": 12620 }, { "grad_norm": 0.22596243023872375, "learning_rate": 9.685280272808486e-05, "loss": 0.0131, "step": 12630 }, { "grad_norm": 0.24080075323581696, "learning_rate": 9.684558176433424e-05, "loss": 0.0133, "step": 12640 }, { "grad_norm": 0.20911791920661926, "learning_rate": 9.683835279595495e-05, "loss": 0.0123, "step": 12650 }, { "grad_norm": 0.1990547627210617, "learning_rate": 9.683111582418216e-05, "loss": 0.0113, "step": 12660 }, { "grad_norm": 0.22106999158859253, "learning_rate": 9.682387085025254e-05, "loss": 0.0113, "step": 12670 }, { "grad_norm": 0.2022857517004013, "learning_rate": 9.681661787540401e-05, "loss": 0.0143, "step": 12680 }, { "grad_norm": 0.16250751912593842, "learning_rate": 9.680935690087593e-05, "loss": 0.0146, "step": 12690 }, { "grad_norm": 0.19746215641498566, "learning_rate": 9.680208792790901e-05, "loss": 0.0118, "step": 12700 }, { "grad_norm": 0.23555974662303925, "learning_rate": 9.679481095774529e-05, "loss": 0.0137, "step": 12710 }, { "grad_norm": 0.19032442569732666, "learning_rate": 9.678752599162822e-05, "loss": 0.0132, "step": 12720 }, { "grad_norm": 0.190447598695755, "learning_rate": 9.678023303080259e-05, "loss": 0.0136, "step": 12730 }, { "grad_norm": 0.23806506395339966, "learning_rate": 9.677293207651459e-05, "loss": 0.0125, "step": 12740 }, { "grad_norm": 0.20570871233940125, "learning_rate": 9.676562313001173e-05, "loss": 0.0146, "step": 12750 }, { "grad_norm": 0.27830833196640015, "learning_rate": 9.675830619254293e-05, "loss": 0.0136, "step": 12760 }, { "grad_norm": 0.2635522782802582, "learning_rate": 9.675098126535843e-05, "loss": 0.0148, "step": 12770 }, { "grad_norm": 0.253230482339859, "learning_rate": 9.674364834970988e-05, "loss": 0.0166, "step": 12780 }, { "grad_norm": 0.18568511307239532, "learning_rate": 9.673630744685028e-05, "loss": 0.013, "step": 12790 }, { "grad_norm": 0.186501145362854, "learning_rate": 9.672895855803397e-05, "loss": 0.0138, "step": 12800 }, { "grad_norm": 0.2093854546546936, "learning_rate": 9.672160168451667e-05, "loss": 0.0115, "step": 12810 }, { "grad_norm": 0.20836998522281647, "learning_rate": 9.671423682755549e-05, "loss": 0.0128, "step": 12820 }, { "grad_norm": 0.18004031479358673, "learning_rate": 9.670686398840888e-05, "loss": 0.012, "step": 12830 }, { "grad_norm": 0.1701427698135376, "learning_rate": 9.669948316833664e-05, "loss": 0.0128, "step": 12840 }, { "grad_norm": 0.22954915463924408, "learning_rate": 9.669209436859997e-05, "loss": 0.0116, "step": 12850 }, { "grad_norm": 0.20675387978553772, "learning_rate": 9.66846975904614e-05, "loss": 0.0139, "step": 12860 }, { "grad_norm": 0.24447008967399597, "learning_rate": 9.667729283518483e-05, "loss": 0.0117, "step": 12870 }, { "grad_norm": 0.1778881549835205, "learning_rate": 9.666988010403557e-05, "loss": 0.0113, "step": 12880 }, { "grad_norm": 0.19125643372535706, "learning_rate": 9.66624593982802e-05, "loss": 0.0123, "step": 12890 }, { "grad_norm": 0.21332217752933502, "learning_rate": 9.665503071918675e-05, "loss": 0.0122, "step": 12900 }, { "grad_norm": 0.2274593561887741, "learning_rate": 9.664759406802456e-05, "loss": 0.0111, "step": 12910 }, { "grad_norm": 0.1975243091583252, "learning_rate": 9.664014944606437e-05, "loss": 0.0124, "step": 12920 }, { "grad_norm": 0.16666379570960999, "learning_rate": 9.663269685457822e-05, "loss": 0.0133, "step": 12930 }, { "grad_norm": 0.21614985167980194, "learning_rate": 9.662523629483962e-05, "loss": 0.0133, "step": 12940 }, { "grad_norm": 0.17069067060947418, "learning_rate": 9.661776776812333e-05, "loss": 0.0107, "step": 12950 }, { "grad_norm": 0.19676074385643005, "learning_rate": 9.661029127570553e-05, "loss": 0.0118, "step": 12960 }, { "grad_norm": 0.19468899071216583, "learning_rate": 9.660280681886373e-05, "loss": 0.0119, "step": 12970 }, { "grad_norm": 0.1738368421792984, "learning_rate": 9.659531439887685e-05, "loss": 0.0135, "step": 12980 }, { "grad_norm": 0.338390976190567, "learning_rate": 9.658781401702511e-05, "loss": 0.0191, "step": 12990 }, { "grad_norm": 0.14787018299102783, "learning_rate": 9.658030567459015e-05, "loss": 0.0118, "step": 13000 }, { "grad_norm": 0.2314724177122116, "learning_rate": 9.65727893728549e-05, "loss": 0.0136, "step": 13010 }, { "grad_norm": 0.2556465268135071, "learning_rate": 9.656526511310375e-05, "loss": 0.0125, "step": 13020 }, { "grad_norm": 0.21065476536750793, "learning_rate": 9.655773289662233e-05, "loss": 0.0123, "step": 13030 }, { "grad_norm": 0.1763332337141037, "learning_rate": 9.655019272469772e-05, "loss": 0.0171, "step": 13040 }, { "grad_norm": 0.23215681314468384, "learning_rate": 9.654264459861832e-05, "loss": 0.016, "step": 13050 }, { "grad_norm": 0.20545437932014465, "learning_rate": 9.653508851967391e-05, "loss": 0.0123, "step": 13060 }, { "grad_norm": 0.22269880771636963, "learning_rate": 9.65275244891556e-05, "loss": 0.0127, "step": 13070 }, { "grad_norm": 0.22275884449481964, "learning_rate": 9.651995250835591e-05, "loss": 0.011, "step": 13080 }, { "grad_norm": 0.20248469710350037, "learning_rate": 9.651237257856862e-05, "loss": 0.0123, "step": 13090 }, { "grad_norm": 0.1737489402294159, "learning_rate": 9.6504784701089e-05, "loss": 0.0116, "step": 13100 }, { "grad_norm": 0.18943087756633759, "learning_rate": 9.649718887721357e-05, "loss": 0.0139, "step": 13110 }, { "grad_norm": 0.136430561542511, "learning_rate": 9.648958510824028e-05, "loss": 0.0123, "step": 13120 }, { "grad_norm": 0.1758842021226883, "learning_rate": 9.648197339546837e-05, "loss": 0.0141, "step": 13130 }, { "grad_norm": 0.1496138870716095, "learning_rate": 9.647435374019851e-05, "loss": 0.014, "step": 13140 }, { "grad_norm": 0.25410985946655273, "learning_rate": 9.646672614373266e-05, "loss": 0.0161, "step": 13150 }, { "grad_norm": 0.2523193955421448, "learning_rate": 9.645909060737418e-05, "loss": 0.0131, "step": 13160 }, { "grad_norm": 0.1976139098405838, "learning_rate": 9.645144713242778e-05, "loss": 0.0157, "step": 13170 }, { "grad_norm": 0.2279968112707138, "learning_rate": 9.64437957201995e-05, "loss": 0.0136, "step": 13180 }, { "grad_norm": 0.2651370167732239, "learning_rate": 9.643613637199678e-05, "loss": 0.0144, "step": 13190 }, { "grad_norm": 0.20134291052818298, "learning_rate": 9.642846908912839e-05, "loss": 0.0147, "step": 13200 }, { "grad_norm": 0.17943178117275238, "learning_rate": 9.642079387290444e-05, "loss": 0.0129, "step": 13210 }, { "grad_norm": 0.15216077864170074, "learning_rate": 9.641311072463644e-05, "loss": 0.0145, "step": 13220 }, { "grad_norm": 0.20839959383010864, "learning_rate": 9.640541964563722e-05, "loss": 0.0151, "step": 13230 }, { "grad_norm": 0.22803819179534912, "learning_rate": 9.639772063722096e-05, "loss": 0.0161, "step": 13240 }, { "grad_norm": 0.16593527793884277, "learning_rate": 9.639001370070324e-05, "loss": 0.014, "step": 13250 }, { "grad_norm": 0.17700913548469543, "learning_rate": 9.638229883740095e-05, "loss": 0.0123, "step": 13260 }, { "grad_norm": 0.12104878574609756, "learning_rate": 9.637457604863233e-05, "loss": 0.0115, "step": 13270 }, { "grad_norm": 0.14419183135032654, "learning_rate": 9.636684533571703e-05, "loss": 0.0127, "step": 13280 }, { "grad_norm": 0.20738038420677185, "learning_rate": 9.635910669997599e-05, "loss": 0.0094, "step": 13290 }, { "grad_norm": 0.19834677875041962, "learning_rate": 9.635136014273154e-05, "loss": 0.0122, "step": 13300 }, { "grad_norm": 0.17657199501991272, "learning_rate": 9.634360566530735e-05, "loss": 0.0125, "step": 13310 }, { "grad_norm": 0.17982350289821625, "learning_rate": 9.633584326902845e-05, "loss": 0.0119, "step": 13320 }, { "grad_norm": 0.21946653723716736, "learning_rate": 9.632807295522124e-05, "loss": 0.0126, "step": 13330 }, { "grad_norm": 0.156845822930336, "learning_rate": 9.632029472521342e-05, "loss": 0.0114, "step": 13340 }, { "grad_norm": 0.15050974488258362, "learning_rate": 9.631250858033409e-05, "loss": 0.0111, "step": 13350 }, { "grad_norm": 0.1842520385980606, "learning_rate": 9.630471452191371e-05, "loss": 0.0122, "step": 13360 }, { "grad_norm": 0.1691090166568756, "learning_rate": 9.629691255128405e-05, "loss": 0.0133, "step": 13370 }, { "grad_norm": 0.15858113765716553, "learning_rate": 9.628910266977825e-05, "loss": 0.0108, "step": 13380 }, { "grad_norm": 0.1648901253938675, "learning_rate": 9.628128487873083e-05, "loss": 0.0106, "step": 13390 }, { "grad_norm": 0.20139192044734955, "learning_rate": 9.627345917947761e-05, "loss": 0.0122, "step": 13400 }, { "grad_norm": 0.1788090616464615, "learning_rate": 9.626562557335579e-05, "loss": 0.0109, "step": 13410 }, { "grad_norm": 0.16943249106407166, "learning_rate": 9.625778406170393e-05, "loss": 0.0103, "step": 13420 }, { "grad_norm": 0.20666798949241638, "learning_rate": 9.624993464586193e-05, "loss": 0.0137, "step": 13430 }, { "grad_norm": 0.3080921769142151, "learning_rate": 9.624207732717105e-05, "loss": 0.0166, "step": 13440 }, { "grad_norm": 0.26163098216056824, "learning_rate": 9.623421210697386e-05, "loss": 0.016, "step": 13450 }, { "grad_norm": 0.25470170378685, "learning_rate": 9.622633898661434e-05, "loss": 0.0163, "step": 13460 }, { "grad_norm": 0.19693048298358917, "learning_rate": 9.621845796743778e-05, "loss": 0.0171, "step": 13470 }, { "grad_norm": 0.2931872308254242, "learning_rate": 9.621056905079082e-05, "loss": 0.0139, "step": 13480 }, { "grad_norm": 0.2346125692129135, "learning_rate": 9.620267223802149e-05, "loss": 0.0132, "step": 13490 }, { "grad_norm": 0.1851264089345932, "learning_rate": 9.619476753047911e-05, "loss": 0.0128, "step": 13500 }, { "grad_norm": 0.17853733897209167, "learning_rate": 9.618685492951438e-05, "loss": 0.0141, "step": 13510 }, { "grad_norm": 0.24313093721866608, "learning_rate": 9.617893443647938e-05, "loss": 0.0115, "step": 13520 }, { "grad_norm": 0.23258541524410248, "learning_rate": 9.617100605272746e-05, "loss": 0.0135, "step": 13530 }, { "grad_norm": 0.16828927397727966, "learning_rate": 9.616306977961338e-05, "loss": 0.0114, "step": 13540 }, { "grad_norm": 0.2097148448228836, "learning_rate": 9.615512561849326e-05, "loss": 0.0134, "step": 13550 }, { "grad_norm": 0.17920081317424774, "learning_rate": 9.61471735707245e-05, "loss": 0.0191, "step": 13560 }, { "grad_norm": 0.20464016497135162, "learning_rate": 9.613921363766592e-05, "loss": 0.0142, "step": 13570 }, { "grad_norm": 0.18968866765499115, "learning_rate": 9.613124582067763e-05, "loss": 0.0114, "step": 13580 }, { "grad_norm": 0.21119016408920288, "learning_rate": 9.612327012112112e-05, "loss": 0.0128, "step": 13590 }, { "grad_norm": 0.23230792582035065, "learning_rate": 9.611528654035921e-05, "loss": 0.0112, "step": 13600 }, { "grad_norm": 0.2310476303100586, "learning_rate": 9.610729507975611e-05, "loss": 0.0126, "step": 13610 }, { "grad_norm": 0.31546148657798767, "learning_rate": 9.609929574067731e-05, "loss": 0.0122, "step": 13620 }, { "grad_norm": 0.1897694617509842, "learning_rate": 9.609128852448967e-05, "loss": 0.0126, "step": 13630 }, { "grad_norm": 0.18674330413341522, "learning_rate": 9.608327343256143e-05, "loss": 0.0129, "step": 13640 }, { "grad_norm": 0.2896610498428345, "learning_rate": 9.607525046626216e-05, "loss": 0.013, "step": 13650 }, { "grad_norm": 0.22099781036376953, "learning_rate": 9.606721962696272e-05, "loss": 0.0105, "step": 13660 }, { "grad_norm": 0.15367703139781952, "learning_rate": 9.60591809160354e-05, "loss": 0.0128, "step": 13670 }, { "grad_norm": 0.20447662472724915, "learning_rate": 9.605113433485378e-05, "loss": 0.0129, "step": 13680 }, { "grad_norm": 0.18628796935081482, "learning_rate": 9.604307988479279e-05, "loss": 0.0124, "step": 13690 }, { "grad_norm": 0.16758756339550018, "learning_rate": 9.603501756722876e-05, "loss": 0.0125, "step": 13700 }, { "grad_norm": 0.19130051136016846, "learning_rate": 9.602694738353927e-05, "loss": 0.0133, "step": 13710 }, { "grad_norm": 0.16210535168647766, "learning_rate": 9.601886933510331e-05, "loss": 0.0144, "step": 13720 }, { "grad_norm": 0.20045989751815796, "learning_rate": 9.60107834233012e-05, "loss": 0.0114, "step": 13730 }, { "grad_norm": 0.27753666043281555, "learning_rate": 9.60026896495146e-05, "loss": 0.0141, "step": 13740 }, { "grad_norm": 0.2200765609741211, "learning_rate": 9.599458801512652e-05, "loss": 0.0119, "step": 13750 }, { "grad_norm": 0.16610410809516907, "learning_rate": 9.598647852152129e-05, "loss": 0.0109, "step": 13760 }, { "grad_norm": 0.2590234875679016, "learning_rate": 9.597836117008462e-05, "loss": 0.013, "step": 13770 }, { "grad_norm": 0.21883350610733032, "learning_rate": 9.597023596220356e-05, "loss": 0.0146, "step": 13780 }, { "grad_norm": 0.20846083760261536, "learning_rate": 9.596210289926643e-05, "loss": 0.0121, "step": 13790 }, { "grad_norm": 0.19419755041599274, "learning_rate": 9.5953961982663e-05, "loss": 0.0125, "step": 13800 }, { "grad_norm": 0.20909379422664642, "learning_rate": 9.594581321378431e-05, "loss": 0.012, "step": 13810 }, { "grad_norm": 0.16283759474754333, "learning_rate": 9.593765659402276e-05, "loss": 0.0123, "step": 13820 }, { "grad_norm": 0.21666041016578674, "learning_rate": 9.59294921247721e-05, "loss": 0.0102, "step": 13830 }, { "grad_norm": 0.14986248314380646, "learning_rate": 9.59213198074274e-05, "loss": 0.0099, "step": 13840 }, { "grad_norm": 0.1601313203573227, "learning_rate": 9.59131396433851e-05, "loss": 0.0123, "step": 13850 }, { "grad_norm": 0.13878963887691498, "learning_rate": 9.590495163404297e-05, "loss": 0.0123, "step": 13860 }, { "grad_norm": 0.17252330482006073, "learning_rate": 9.589675578080009e-05, "loss": 0.0124, "step": 13870 }, { "grad_norm": 0.1607462614774704, "learning_rate": 9.588855208505694e-05, "loss": 0.012, "step": 13880 }, { "grad_norm": 0.15601523220539093, "learning_rate": 9.588034054821529e-05, "loss": 0.0111, "step": 13890 }, { "grad_norm": 0.12011834233999252, "learning_rate": 9.587212117167826e-05, "loss": 0.0117, "step": 13900 }, { "grad_norm": 0.15581297874450684, "learning_rate": 9.586389395685033e-05, "loss": 0.0089, "step": 13910 }, { "grad_norm": 0.18954381346702576, "learning_rate": 9.585565890513733e-05, "loss": 0.01, "step": 13920 }, { "grad_norm": 0.21545487642288208, "learning_rate": 9.584741601794636e-05, "loss": 0.0112, "step": 13930 }, { "grad_norm": 0.22622936964035034, "learning_rate": 9.58391652966859e-05, "loss": 0.0112, "step": 13940 }, { "grad_norm": 0.18680115044116974, "learning_rate": 9.583090674276583e-05, "loss": 0.0138, "step": 13950 }, { "grad_norm": 0.2538067400455475, "learning_rate": 9.582264035759726e-05, "loss": 0.0104, "step": 13960 }, { "grad_norm": 0.18242955207824707, "learning_rate": 9.58143661425927e-05, "loss": 0.0118, "step": 13970 }, { "grad_norm": 0.19359293580055237, "learning_rate": 9.580608409916601e-05, "loss": 0.0124, "step": 13980 }, { "grad_norm": 0.16159990429878235, "learning_rate": 9.579779422873233e-05, "loss": 0.0139, "step": 13990 }, { "grad_norm": 0.2343824952840805, "learning_rate": 9.578949653270819e-05, "loss": 0.0185, "step": 14000 }, { "grad_norm": 0.15912607312202454, "learning_rate": 9.578119101251144e-05, "loss": 0.0095, "step": 14010 }, { "grad_norm": 0.18060223758220673, "learning_rate": 9.577287766956127e-05, "loss": 0.0117, "step": 14020 }, { "grad_norm": 0.2119925320148468, "learning_rate": 9.57645565052782e-05, "loss": 0.0107, "step": 14030 }, { "grad_norm": 0.22522525489330292, "learning_rate": 9.575622752108407e-05, "loss": 0.0125, "step": 14040 }, { "grad_norm": 0.22323590517044067, "learning_rate": 9.57478907184021e-05, "loss": 0.0121, "step": 14050 }, { "grad_norm": 0.18764640390872955, "learning_rate": 9.573954609865681e-05, "loss": 0.0123, "step": 14060 }, { "grad_norm": 0.19646762311458588, "learning_rate": 9.573119366327408e-05, "loss": 0.0126, "step": 14070 }, { "grad_norm": 0.1957392543554306, "learning_rate": 9.57228334136811e-05, "loss": 0.0112, "step": 14080 }, { "grad_norm": 0.23398464918136597, "learning_rate": 9.571446535130641e-05, "loss": 0.013, "step": 14090 }, { "grad_norm": 0.14141827821731567, "learning_rate": 9.570608947757988e-05, "loss": 0.0107, "step": 14100 }, { "grad_norm": 0.14405810832977295, "learning_rate": 9.569770579393274e-05, "loss": 0.0109, "step": 14110 }, { "grad_norm": 0.14622075855731964, "learning_rate": 9.56893143017975e-05, "loss": 0.0109, "step": 14120 }, { "grad_norm": 0.18502038717269897, "learning_rate": 9.568091500260806e-05, "loss": 0.01, "step": 14130 }, { "grad_norm": 0.17815826833248138, "learning_rate": 9.567250789779961e-05, "loss": 0.0118, "step": 14140 }, { "grad_norm": 0.14917881786823273, "learning_rate": 9.566409298880872e-05, "loss": 0.0108, "step": 14150 }, { "grad_norm": 0.2091001570224762, "learning_rate": 9.565567027707326e-05, "loss": 0.0116, "step": 14160 }, { "grad_norm": 0.2158524990081787, "learning_rate": 9.56472397640324e-05, "loss": 0.0122, "step": 14170 }, { "grad_norm": 0.23961736261844635, "learning_rate": 9.563880145112675e-05, "loss": 0.016, "step": 14180 }, { "grad_norm": 0.29503676295280457, "learning_rate": 9.563035533979814e-05, "loss": 0.0127, "step": 14190 }, { "grad_norm": 0.16156092286109924, "learning_rate": 9.562190143148981e-05, "loss": 0.0144, "step": 14200 }, { "grad_norm": 0.18297231197357178, "learning_rate": 9.561343972764627e-05, "loss": 0.0099, "step": 14210 }, { "grad_norm": 0.17735528945922852, "learning_rate": 9.560497022971343e-05, "loss": 0.0117, "step": 14220 }, { "grad_norm": 0.1388481855392456, "learning_rate": 9.559649293913847e-05, "loss": 0.0116, "step": 14230 }, { "grad_norm": 0.20743213593959808, "learning_rate": 9.558800785736993e-05, "loss": 0.0131, "step": 14240 }, { "grad_norm": 0.2577696144580841, "learning_rate": 9.557951498585767e-05, "loss": 0.0138, "step": 14250 }, { "grad_norm": 0.17421551048755646, "learning_rate": 9.557101432605293e-05, "loss": 0.0117, "step": 14260 }, { "grad_norm": 0.2423105388879776, "learning_rate": 9.556250587940818e-05, "loss": 0.0116, "step": 14270 }, { "grad_norm": 0.2314424365758896, "learning_rate": 9.555398964737734e-05, "loss": 0.0124, "step": 14280 }, { "grad_norm": 0.18644170463085175, "learning_rate": 9.554546563141555e-05, "loss": 0.0124, "step": 14290 }, { "grad_norm": 0.18195770680904388, "learning_rate": 9.553693383297937e-05, "loss": 0.0156, "step": 14300 }, { "grad_norm": 0.22436760365962982, "learning_rate": 9.552839425352663e-05, "loss": 0.0119, "step": 14310 }, { "grad_norm": 0.23224669694900513, "learning_rate": 9.551984689451652e-05, "loss": 0.0129, "step": 14320 }, { "grad_norm": 0.2062879502773285, "learning_rate": 9.551129175740953e-05, "loss": 0.014, "step": 14330 }, { "grad_norm": 0.20815539360046387, "learning_rate": 9.550272884366754e-05, "loss": 0.0117, "step": 14340 }, { "grad_norm": 0.16853949427604675, "learning_rate": 9.549415815475369e-05, "loss": 0.0132, "step": 14350 }, { "grad_norm": 0.15748003125190735, "learning_rate": 9.548557969213247e-05, "loss": 0.0107, "step": 14360 }, { "grad_norm": 0.1884704977273941, "learning_rate": 9.547699345726972e-05, "loss": 0.0099, "step": 14370 }, { "grad_norm": 0.16366949677467346, "learning_rate": 9.546839945163257e-05, "loss": 0.0105, "step": 14380 }, { "grad_norm": 0.1604558527469635, "learning_rate": 9.545979767668953e-05, "loss": 0.0133, "step": 14390 }, { "grad_norm": 0.21014565229415894, "learning_rate": 9.54511881339104e-05, "loss": 0.0111, "step": 14400 }, { "grad_norm": 0.2039770781993866, "learning_rate": 9.54425708247663e-05, "loss": 0.014, "step": 14410 }, { "grad_norm": 0.159332275390625, "learning_rate": 9.543394575072972e-05, "loss": 0.0122, "step": 14420 }, { "grad_norm": 0.1820732206106186, "learning_rate": 9.542531291327441e-05, "loss": 0.0105, "step": 14430 }, { "grad_norm": 0.1962892860174179, "learning_rate": 9.541667231387552e-05, "loss": 0.011, "step": 14440 }, { "grad_norm": 0.18070976436138153, "learning_rate": 9.540802395400949e-05, "loss": 0.0106, "step": 14450 }, { "grad_norm": 0.21270209550857544, "learning_rate": 9.539936783515406e-05, "loss": 0.0113, "step": 14460 }, { "grad_norm": 0.22465237975120544, "learning_rate": 9.539070395878835e-05, "loss": 0.0135, "step": 14470 }, { "grad_norm": 0.2573768198490143, "learning_rate": 9.538203232639277e-05, "loss": 0.0112, "step": 14480 }, { "grad_norm": 0.17865630984306335, "learning_rate": 9.537335293944907e-05, "loss": 0.0126, "step": 14490 }, { "grad_norm": 0.18610382080078125, "learning_rate": 9.536466579944032e-05, "loss": 0.0108, "step": 14500 }, { "grad_norm": 0.18751148879528046, "learning_rate": 9.535597090785091e-05, "loss": 0.0123, "step": 14510 }, { "grad_norm": 0.1994241327047348, "learning_rate": 9.534726826616656e-05, "loss": 0.0105, "step": 14520 }, { "grad_norm": 0.1552632600069046, "learning_rate": 9.53385578758743e-05, "loss": 0.0114, "step": 14530 }, { "grad_norm": 0.2077675610780716, "learning_rate": 9.532983973846252e-05, "loss": 0.0137, "step": 14540 }, { "grad_norm": 0.1580284982919693, "learning_rate": 9.53211138554209e-05, "loss": 0.0107, "step": 14550 }, { "grad_norm": 0.1375611573457718, "learning_rate": 9.531238022824047e-05, "loss": 0.0102, "step": 14560 }, { "grad_norm": 0.15662042796611786, "learning_rate": 9.530363885841355e-05, "loss": 0.011, "step": 14570 }, { "grad_norm": 0.18453474342823029, "learning_rate": 9.52948897474338e-05, "loss": 0.0119, "step": 14580 }, { "grad_norm": 0.1739329695701599, "learning_rate": 9.528613289679622e-05, "loss": 0.0121, "step": 14590 }, { "grad_norm": 0.12928365170955658, "learning_rate": 9.52773683079971e-05, "loss": 0.0117, "step": 14600 }, { "grad_norm": 0.18512234091758728, "learning_rate": 9.526859598253407e-05, "loss": 0.0127, "step": 14610 }, { "grad_norm": 0.22287814319133759, "learning_rate": 9.525981592190609e-05, "loss": 0.0135, "step": 14620 }, { "grad_norm": 0.19097018241882324, "learning_rate": 9.525102812761342e-05, "loss": 0.0131, "step": 14630 }, { "grad_norm": 0.20662717521190643, "learning_rate": 9.524223260115768e-05, "loss": 0.0122, "step": 14640 }, { "grad_norm": 0.2029201090335846, "learning_rate": 9.523342934404175e-05, "loss": 0.0107, "step": 14650 }, { "grad_norm": 0.18307416141033173, "learning_rate": 9.522461835776989e-05, "loss": 0.0168, "step": 14660 }, { "grad_norm": 0.2225351780653, "learning_rate": 9.521579964384764e-05, "loss": 0.0114, "step": 14670 }, { "grad_norm": 0.18019673228263855, "learning_rate": 9.52069732037819e-05, "loss": 0.012, "step": 14680 }, { "grad_norm": 0.16745857894420624, "learning_rate": 9.519813903908083e-05, "loss": 0.0127, "step": 14690 }, { "grad_norm": 0.20044298470020294, "learning_rate": 9.5189297151254e-05, "loss": 0.0129, "step": 14700 }, { "grad_norm": 0.1328292191028595, "learning_rate": 9.518044754181218e-05, "loss": 0.0104, "step": 14710 }, { "grad_norm": 0.16237257421016693, "learning_rate": 9.51715902122676e-05, "loss": 0.0138, "step": 14720 }, { "grad_norm": 0.18252672255039215, "learning_rate": 9.516272516413368e-05, "loss": 0.013, "step": 14730 }, { "grad_norm": 0.22272147238254547, "learning_rate": 9.515385239892525e-05, "loss": 0.0135, "step": 14740 }, { "grad_norm": 0.17075961828231812, "learning_rate": 9.514497191815839e-05, "loss": 0.0121, "step": 14750 }, { "grad_norm": 0.20230898261070251, "learning_rate": 9.513608372335055e-05, "loss": 0.0107, "step": 14760 }, { "grad_norm": 0.15806947648525238, "learning_rate": 9.512718781602045e-05, "loss": 0.0137, "step": 14770 }, { "grad_norm": 0.1772410273551941, "learning_rate": 9.511828419768823e-05, "loss": 0.011, "step": 14780 }, { "grad_norm": 0.1746087521314621, "learning_rate": 9.510937286987521e-05, "loss": 0.0123, "step": 14790 }, { "grad_norm": 0.18134720623493195, "learning_rate": 9.510045383410408e-05, "loss": 0.0121, "step": 14800 }, { "grad_norm": 0.2010781168937683, "learning_rate": 9.509152709189892e-05, "loss": 0.0102, "step": 14810 }, { "grad_norm": 0.21682827174663544, "learning_rate": 9.508259264478504e-05, "loss": 0.013, "step": 14820 }, { "grad_norm": 0.19369572401046753, "learning_rate": 9.507365049428909e-05, "loss": 0.0122, "step": 14830 }, { "grad_norm": 0.16824190318584442, "learning_rate": 9.506470064193902e-05, "loss": 0.0126, "step": 14840 }, { "grad_norm": 0.2238263040781021, "learning_rate": 9.505574308926414e-05, "loss": 0.0115, "step": 14850 }, { "grad_norm": 0.15500077605247498, "learning_rate": 9.504677783779505e-05, "loss": 0.0087, "step": 14860 }, { "grad_norm": 0.18799801170825958, "learning_rate": 9.503780488906365e-05, "loss": 0.0096, "step": 14870 }, { "grad_norm": 0.18671441078186035, "learning_rate": 9.502882424460319e-05, "loss": 0.0112, "step": 14880 }, { "grad_norm": 0.23190374672412872, "learning_rate": 9.501983590594821e-05, "loss": 0.0109, "step": 14890 }, { "grad_norm": 0.19840143620967865, "learning_rate": 9.501083987463455e-05, "loss": 0.0115, "step": 14900 }, { "grad_norm": 0.1762484461069107, "learning_rate": 9.500183615219942e-05, "loss": 0.0105, "step": 14910 }, { "grad_norm": 0.1574656218290329, "learning_rate": 9.49928247401813e-05, "loss": 0.0129, "step": 14920 }, { "grad_norm": 0.23860879242420197, "learning_rate": 9.498380564011997e-05, "loss": 0.0127, "step": 14930 }, { "grad_norm": 0.23165160417556763, "learning_rate": 9.497477885355656e-05, "loss": 0.0129, "step": 14940 }, { "grad_norm": 0.1929062008857727, "learning_rate": 9.496574438203353e-05, "loss": 0.0101, "step": 14950 }, { "grad_norm": 0.1709241271018982, "learning_rate": 9.495670222709459e-05, "loss": 0.0099, "step": 14960 }, { "grad_norm": 0.16593599319458008, "learning_rate": 9.494765239028483e-05, "loss": 0.016, "step": 14970 }, { "grad_norm": 0.2253137081861496, "learning_rate": 9.493859487315057e-05, "loss": 0.0114, "step": 14980 }, { "grad_norm": 0.20457804203033447, "learning_rate": 9.492952967723953e-05, "loss": 0.0109, "step": 14990 }, { "grad_norm": 0.23775996267795563, "learning_rate": 9.492045680410068e-05, "loss": 0.0119, "step": 15000 }, { "grad_norm": 0.16126364469528198, "learning_rate": 9.491137625528436e-05, "loss": 0.0103, "step": 15010 }, { "grad_norm": 0.16344520449638367, "learning_rate": 9.490228803234215e-05, "loss": 0.0106, "step": 15020 }, { "grad_norm": 0.22316518425941467, "learning_rate": 9.489319213682701e-05, "loss": 0.0102, "step": 15030 }, { "grad_norm": 0.1914847195148468, "learning_rate": 9.488408857029316e-05, "loss": 0.0092, "step": 15040 }, { "grad_norm": 0.1722010225057602, "learning_rate": 9.487497733429616e-05, "loss": 0.0101, "step": 15050 }, { "grad_norm": 0.1832200288772583, "learning_rate": 9.486585843039286e-05, "loss": 0.0094, "step": 15060 }, { "grad_norm": 0.15141834318637848, "learning_rate": 9.485673186014143e-05, "loss": 0.0096, "step": 15070 }, { "grad_norm": 0.2053770273923874, "learning_rate": 9.484759762510137e-05, "loss": 0.0104, "step": 15080 }, { "grad_norm": 0.18039245903491974, "learning_rate": 9.483845572683346e-05, "loss": 0.0094, "step": 15090 }, { "grad_norm": 0.14187024533748627, "learning_rate": 9.48293061668998e-05, "loss": 0.0107, "step": 15100 }, { "grad_norm": 0.1933068037033081, "learning_rate": 9.48201489468638e-05, "loss": 0.0114, "step": 15110 }, { "grad_norm": 0.1453678160905838, "learning_rate": 9.481098406829016e-05, "loss": 0.0111, "step": 15120 }, { "grad_norm": 0.17023704946041107, "learning_rate": 9.480181153274495e-05, "loss": 0.0109, "step": 15130 }, { "grad_norm": 0.18779908120632172, "learning_rate": 9.479263134179548e-05, "loss": 0.0123, "step": 15140 }, { "grad_norm": 0.18925325572490692, "learning_rate": 9.478344349701039e-05, "loss": 0.0147, "step": 15150 }, { "grad_norm": 0.18424560129642487, "learning_rate": 9.477424799995964e-05, "loss": 0.0146, "step": 15160 }, { "grad_norm": 0.2117031365633011, "learning_rate": 9.476504485221448e-05, "loss": 0.0126, "step": 15170 }, { "grad_norm": 0.17977386713027954, "learning_rate": 9.475583405534748e-05, "loss": 0.0112, "step": 15180 }, { "grad_norm": 0.17783194780349731, "learning_rate": 9.474661561093251e-05, "loss": 0.0134, "step": 15190 }, { "grad_norm": 0.1314159631729126, "learning_rate": 9.473738952054478e-05, "loss": 0.0126, "step": 15200 }, { "grad_norm": 0.1352318823337555, "learning_rate": 9.472815578576073e-05, "loss": 0.0125, "step": 15210 }, { "grad_norm": 0.21507613360881805, "learning_rate": 9.471891440815817e-05, "loss": 0.0114, "step": 15220 }, { "grad_norm": 0.21785950660705566, "learning_rate": 9.470966538931621e-05, "loss": 0.0118, "step": 15230 }, { "grad_norm": 0.18424805998802185, "learning_rate": 9.470040873081525e-05, "loss": 0.0108, "step": 15240 }, { "grad_norm": 0.2594073712825775, "learning_rate": 9.469114443423698e-05, "loss": 0.0097, "step": 15250 }, { "grad_norm": 0.182132288813591, "learning_rate": 9.468187250116445e-05, "loss": 0.0115, "step": 15260 }, { "grad_norm": 0.14396469295024872, "learning_rate": 9.467259293318197e-05, "loss": 0.0104, "step": 15270 }, { "grad_norm": 0.18505769968032837, "learning_rate": 9.466330573187514e-05, "loss": 0.0123, "step": 15280 }, { "grad_norm": 0.15694692730903625, "learning_rate": 9.46540108988309e-05, "loss": 0.011, "step": 15290 }, { "grad_norm": 0.13998055458068848, "learning_rate": 9.46447084356375e-05, "loss": 0.01, "step": 15300 }, { "grad_norm": 0.1833702176809311, "learning_rate": 9.463539834388447e-05, "loss": 0.0096, "step": 15310 }, { "grad_norm": 0.15341836214065552, "learning_rate": 9.462608062516263e-05, "loss": 0.0112, "step": 15320 }, { "grad_norm": 0.14058543741703033, "learning_rate": 9.461675528106413e-05, "loss": 0.01, "step": 15330 }, { "grad_norm": 0.18226906657218933, "learning_rate": 9.460742231318244e-05, "loss": 0.0115, "step": 15340 }, { "grad_norm": 0.20096728205680847, "learning_rate": 9.459808172311229e-05, "loss": 0.0106, "step": 15350 }, { "grad_norm": 0.13457484543323517, "learning_rate": 9.458873351244972e-05, "loss": 0.0114, "step": 15360 }, { "grad_norm": 0.14988575875759125, "learning_rate": 9.457937768279211e-05, "loss": 0.0125, "step": 15370 }, { "grad_norm": 0.19069187343120575, "learning_rate": 9.45700142357381e-05, "loss": 0.013, "step": 15380 }, { "grad_norm": 0.15569889545440674, "learning_rate": 9.456064317288765e-05, "loss": 0.0105, "step": 15390 }, { "grad_norm": 0.21381890773773193, "learning_rate": 9.455126449584201e-05, "loss": 0.0127, "step": 15400 }, { "grad_norm": 0.1527254581451416, "learning_rate": 9.454187820620375e-05, "loss": 0.0116, "step": 15410 }, { "grad_norm": 0.20528289675712585, "learning_rate": 9.453248430557673e-05, "loss": 0.0116, "step": 15420 }, { "grad_norm": 0.1905265599489212, "learning_rate": 9.452308279556611e-05, "loss": 0.0124, "step": 15430 }, { "grad_norm": 0.2038540095090866, "learning_rate": 9.451367367777835e-05, "loss": 0.0123, "step": 15440 }, { "grad_norm": 0.13762138783931732, "learning_rate": 9.450425695382122e-05, "loss": 0.0102, "step": 15450 }, { "grad_norm": 0.15060704946517944, "learning_rate": 9.449483262530375e-05, "loss": 0.0114, "step": 15460 }, { "grad_norm": 0.20556603372097015, "learning_rate": 9.448540069383633e-05, "loss": 0.012, "step": 15470 }, { "grad_norm": 0.16441789269447327, "learning_rate": 9.447596116103061e-05, "loss": 0.0132, "step": 15480 }, { "grad_norm": 0.1712106466293335, "learning_rate": 9.446651402849955e-05, "loss": 0.0166, "step": 15490 }, { "grad_norm": 0.21122531592845917, "learning_rate": 9.44570592978574e-05, "loss": 0.0144, "step": 15500 }, { "grad_norm": 0.3201504349708557, "learning_rate": 9.444759697071972e-05, "loss": 0.0137, "step": 15510 }, { "grad_norm": 0.24950885772705078, "learning_rate": 9.443812704870336e-05, "loss": 0.0125, "step": 15520 }, { "grad_norm": 0.21007965505123138, "learning_rate": 9.442864953342649e-05, "loss": 0.0122, "step": 15530 }, { "grad_norm": 0.2072436809539795, "learning_rate": 9.441916442650852e-05, "loss": 0.0117, "step": 15540 }, { "grad_norm": 0.24840611219406128, "learning_rate": 9.440967172957023e-05, "loss": 0.0117, "step": 15550 }, { "grad_norm": 0.2008686065673828, "learning_rate": 9.440017144423364e-05, "loss": 0.0118, "step": 15560 }, { "grad_norm": 0.19014839828014374, "learning_rate": 9.439066357212209e-05, "loss": 0.0103, "step": 15570 }, { "grad_norm": 0.1838718056678772, "learning_rate": 9.438114811486022e-05, "loss": 0.0106, "step": 15580 }, { "grad_norm": 0.1163538247346878, "learning_rate": 9.4371625074074e-05, "loss": 0.012, "step": 15590 }, { "grad_norm": 0.15874513983726501, "learning_rate": 9.436209445139059e-05, "loss": 0.0118, "step": 15600 }, { "grad_norm": 0.17667248845100403, "learning_rate": 9.435255624843855e-05, "loss": 0.011, "step": 15610 }, { "grad_norm": 0.10872521251440048, "learning_rate": 9.43430104668477e-05, "loss": 0.0101, "step": 15620 }, { "grad_norm": 0.16394929587841034, "learning_rate": 9.433345710824914e-05, "loss": 0.0106, "step": 15630 }, { "grad_norm": 0.1317388266324997, "learning_rate": 9.432389617427529e-05, "loss": 0.0105, "step": 15640 }, { "grad_norm": 0.20525991916656494, "learning_rate": 9.431432766655984e-05, "loss": 0.0119, "step": 15650 }, { "grad_norm": 0.2577494978904724, "learning_rate": 9.430475158673778e-05, "loss": 0.0117, "step": 15660 }, { "grad_norm": 0.20722849667072296, "learning_rate": 9.429516793644542e-05, "loss": 0.0129, "step": 15670 }, { "grad_norm": 0.1979713886976242, "learning_rate": 9.428557671732034e-05, "loss": 0.0123, "step": 15680 }, { "grad_norm": 0.2142118662595749, "learning_rate": 9.42759779310014e-05, "loss": 0.0104, "step": 15690 }, { "grad_norm": 0.1876257061958313, "learning_rate": 9.426637157912879e-05, "loss": 0.0092, "step": 15700 }, { "grad_norm": 0.1360582560300827, "learning_rate": 9.425675766334397e-05, "loss": 0.0116, "step": 15710 }, { "grad_norm": 0.16547541320323944, "learning_rate": 9.424713618528968e-05, "loss": 0.0107, "step": 15720 }, { "grad_norm": 0.24643976986408234, "learning_rate": 9.423750714661e-05, "loss": 0.0132, "step": 15730 }, { "grad_norm": 0.18523229658603668, "learning_rate": 9.422787054895022e-05, "loss": 0.0113, "step": 15740 }, { "grad_norm": 0.200520858168602, "learning_rate": 9.4218226393957e-05, "loss": 0.0119, "step": 15750 }, { "grad_norm": 0.23182347416877747, "learning_rate": 9.420857468327828e-05, "loss": 0.0147, "step": 15760 }, { "grad_norm": 0.20129254460334778, "learning_rate": 9.419891541856323e-05, "loss": 0.0131, "step": 15770 }, { "grad_norm": 0.18380998075008392, "learning_rate": 9.41892486014624e-05, "loss": 0.0121, "step": 15780 }, { "grad_norm": 0.1847013235092163, "learning_rate": 9.417957423362756e-05, "loss": 0.0125, "step": 15790 }, { "grad_norm": 0.23999464511871338, "learning_rate": 9.416989231671178e-05, "loss": 0.0135, "step": 15800 }, { "grad_norm": 0.17285741865634918, "learning_rate": 9.416020285236946e-05, "loss": 0.0108, "step": 15810 }, { "grad_norm": 0.22651782631874084, "learning_rate": 9.415050584225626e-05, "loss": 0.0153, "step": 15820 }, { "grad_norm": 0.17620810866355896, "learning_rate": 9.414080128802914e-05, "loss": 0.0099, "step": 15830 }, { "grad_norm": 0.16937516629695892, "learning_rate": 9.413108919134632e-05, "loss": 0.0143, "step": 15840 }, { "grad_norm": 0.16407863795757294, "learning_rate": 9.412136955386734e-05, "loss": 0.0111, "step": 15850 }, { "grad_norm": 0.20121058821678162, "learning_rate": 9.411164237725303e-05, "loss": 0.0132, "step": 15860 }, { "grad_norm": 0.18804001808166504, "learning_rate": 9.41019076631655e-05, "loss": 0.0103, "step": 15870 }, { "grad_norm": 0.17337793111801147, "learning_rate": 9.409216541326815e-05, "loss": 0.0127, "step": 15880 }, { "grad_norm": 0.12972745299339294, "learning_rate": 9.408241562922564e-05, "loss": 0.0116, "step": 15890 }, { "grad_norm": 0.18696674704551697, "learning_rate": 9.407265831270395e-05, "loss": 0.013, "step": 15900 }, { "grad_norm": 0.19340850412845612, "learning_rate": 9.406289346537035e-05, "loss": 0.0138, "step": 15910 }, { "grad_norm": 0.19381558895111084, "learning_rate": 9.405312108889339e-05, "loss": 0.0151, "step": 15920 }, { "grad_norm": 0.24747201800346375, "learning_rate": 9.404334118494288e-05, "loss": 0.0122, "step": 15930 }, { "grad_norm": 0.23538626730442047, "learning_rate": 9.403355375518995e-05, "loss": 0.0126, "step": 15940 }, { "grad_norm": 0.24021022021770477, "learning_rate": 9.4023758801307e-05, "loss": 0.0151, "step": 15950 }, { "grad_norm": 0.2057894766330719, "learning_rate": 9.401395632496774e-05, "loss": 0.0107, "step": 15960 }, { "grad_norm": 0.159623920917511, "learning_rate": 9.400414632784711e-05, "loss": 0.012, "step": 15970 }, { "grad_norm": 0.19077570736408234, "learning_rate": 9.39943288116214e-05, "loss": 0.0111, "step": 15980 }, { "grad_norm": 0.15564627945423126, "learning_rate": 9.398450377796815e-05, "loss": 0.0116, "step": 15990 }, { "grad_norm": 0.20271463692188263, "learning_rate": 9.397467122856616e-05, "loss": 0.0133, "step": 16000 }, { "grad_norm": 0.22547896206378937, "learning_rate": 9.396483116509558e-05, "loss": 0.0128, "step": 16010 }, { "grad_norm": 0.1034667044878006, "learning_rate": 9.39549835892378e-05, "loss": 0.0128, "step": 16020 }, { "grad_norm": 0.15745271742343903, "learning_rate": 9.39451285026755e-05, "loss": 0.0103, "step": 16030 }, { "grad_norm": 0.135220006108284, "learning_rate": 9.393526590709262e-05, "loss": 0.0114, "step": 16040 }, { "grad_norm": 0.20619721710681915, "learning_rate": 9.392539580417444e-05, "loss": 0.0116, "step": 16050 }, { "grad_norm": 0.17908500134944916, "learning_rate": 9.391551819560747e-05, "loss": 0.0129, "step": 16060 }, { "grad_norm": 0.18188878893852234, "learning_rate": 9.390563308307955e-05, "loss": 0.0108, "step": 16070 }, { "grad_norm": 0.17066696286201477, "learning_rate": 9.389574046827974e-05, "loss": 0.0124, "step": 16080 }, { "grad_norm": 0.2163136601448059, "learning_rate": 9.388584035289845e-05, "loss": 0.0106, "step": 16090 }, { "grad_norm": 0.1876007616519928, "learning_rate": 9.387593273862732e-05, "loss": 0.0097, "step": 16100 }, { "grad_norm": 0.17763455212116241, "learning_rate": 9.386601762715929e-05, "loss": 0.0102, "step": 16110 }, { "grad_norm": 0.10857687145471573, "learning_rate": 9.38560950201886e-05, "loss": 0.0105, "step": 16120 }, { "grad_norm": 0.1252482682466507, "learning_rate": 9.384616491941071e-05, "loss": 0.012, "step": 16130 }, { "grad_norm": 0.18119314312934875, "learning_rate": 9.383622732652245e-05, "loss": 0.0117, "step": 16140 }, { "grad_norm": 0.21053089201450348, "learning_rate": 9.382628224322187e-05, "loss": 0.0111, "step": 16150 }, { "grad_norm": 0.159877747297287, "learning_rate": 9.381632967120829e-05, "loss": 0.0109, "step": 16160 }, { "grad_norm": 0.1887502521276474, "learning_rate": 9.380636961218235e-05, "loss": 0.0104, "step": 16170 }, { "grad_norm": 0.16638445854187012, "learning_rate": 9.379640206784597e-05, "loss": 0.011, "step": 16180 }, { "grad_norm": 0.20772601664066315, "learning_rate": 9.378642703990229e-05, "loss": 0.01, "step": 16190 }, { "grad_norm": 0.15460170805454254, "learning_rate": 9.37764445300558e-05, "loss": 0.0104, "step": 16200 }, { "grad_norm": 0.19159112870693207, "learning_rate": 9.376645454001222e-05, "loss": 0.0111, "step": 16210 }, { "grad_norm": 0.1661655157804489, "learning_rate": 9.375645707147858e-05, "loss": 0.011, "step": 16220 }, { "grad_norm": 0.15408770740032196, "learning_rate": 9.374645212616316e-05, "loss": 0.0113, "step": 16230 }, { "grad_norm": 0.19849105179309845, "learning_rate": 9.373643970577555e-05, "loss": 0.0115, "step": 16240 }, { "grad_norm": 0.10975421220064163, "learning_rate": 9.372641981202659e-05, "loss": 0.009, "step": 16250 }, { "grad_norm": 0.20956328511238098, "learning_rate": 9.37163924466284e-05, "loss": 0.0135, "step": 16260 }, { "grad_norm": 0.18191644549369812, "learning_rate": 9.370635761129438e-05, "loss": 0.0103, "step": 16270 }, { "grad_norm": 0.2027600258588791, "learning_rate": 9.36963153077392e-05, "loss": 0.014, "step": 16280 }, { "grad_norm": 0.166470468044281, "learning_rate": 9.368626553767888e-05, "loss": 0.0092, "step": 16290 }, { "grad_norm": 0.26142606139183044, "learning_rate": 9.367620830283057e-05, "loss": 0.0128, "step": 16300 }, { "grad_norm": 0.13567385077476501, "learning_rate": 9.366614360491281e-05, "loss": 0.0117, "step": 16310 }, { "grad_norm": 0.21719971299171448, "learning_rate": 9.365607144564539e-05, "loss": 0.0118, "step": 16320 }, { "grad_norm": 0.17160898447036743, "learning_rate": 9.364599182674934e-05, "loss": 0.0102, "step": 16330 }, { "grad_norm": 0.18181867897510529, "learning_rate": 9.3635904749947e-05, "loss": 0.0103, "step": 16340 }, { "grad_norm": 0.2025519460439682, "learning_rate": 9.362581021696202e-05, "loss": 0.0116, "step": 16350 }, { "grad_norm": 0.18127824366092682, "learning_rate": 9.361570822951921e-05, "loss": 0.0125, "step": 16360 }, { "grad_norm": 0.20383533835411072, "learning_rate": 9.360559878934476e-05, "loss": 0.011, "step": 16370 }, { "grad_norm": 0.2061832696199417, "learning_rate": 9.359548189816611e-05, "loss": 0.0098, "step": 16380 }, { "grad_norm": 0.21710814535617828, "learning_rate": 9.358535755771193e-05, "loss": 0.0111, "step": 16390 }, { "grad_norm": 0.17905116081237793, "learning_rate": 9.357522576971221e-05, "loss": 0.0143, "step": 16400 }, { "grad_norm": 0.20220008492469788, "learning_rate": 9.356508653589819e-05, "loss": 0.0121, "step": 16410 }, { "grad_norm": 0.16591329872608185, "learning_rate": 9.355493985800237e-05, "loss": 0.0129, "step": 16420 }, { "grad_norm": 0.1937415897846222, "learning_rate": 9.354478573775857e-05, "loss": 0.0113, "step": 16430 }, { "grad_norm": 0.2183019071817398, "learning_rate": 9.353462417690186e-05, "loss": 0.0104, "step": 16440 }, { "grad_norm": 0.15318861603736877, "learning_rate": 9.352445517716853e-05, "loss": 0.0124, "step": 16450 }, { "grad_norm": 0.17493972182273865, "learning_rate": 9.351427874029621e-05, "loss": 0.0125, "step": 16460 }, { "grad_norm": 0.1334065943956375, "learning_rate": 9.350409486802379e-05, "loss": 0.0104, "step": 16470 }, { "grad_norm": 0.1688057780265808, "learning_rate": 9.349390356209138e-05, "loss": 0.011, "step": 16480 }, { "grad_norm": 0.1413479745388031, "learning_rate": 9.348370482424042e-05, "loss": 0.0102, "step": 16490 }, { "grad_norm": 0.15052029490470886, "learning_rate": 9.347349865621357e-05, "loss": 0.0094, "step": 16500 }, { "grad_norm": 0.1371920257806778, "learning_rate": 9.346328505975481e-05, "loss": 0.0084, "step": 16510 }, { "grad_norm": 0.1462876945734024, "learning_rate": 9.345306403660936e-05, "loss": 0.0109, "step": 16520 }, { "grad_norm": 0.16693103313446045, "learning_rate": 9.344283558852371e-05, "loss": 0.0112, "step": 16530 }, { "grad_norm": 0.14364683628082275, "learning_rate": 9.343259971724563e-05, "loss": 0.0099, "step": 16540 }, { "grad_norm": 0.15616647899150848, "learning_rate": 9.342235642452413e-05, "loss": 0.0085, "step": 16550 }, { "grad_norm": 0.1889219731092453, "learning_rate": 9.341210571210954e-05, "loss": 0.0114, "step": 16560 }, { "grad_norm": 0.20391826331615448, "learning_rate": 9.340184758175338e-05, "loss": 0.0121, "step": 16570 }, { "grad_norm": 0.1739843636751175, "learning_rate": 9.339158203520854e-05, "loss": 0.0108, "step": 16580 }, { "grad_norm": 0.13114765286445618, "learning_rate": 9.338130907422908e-05, "loss": 0.0108, "step": 16590 }, { "grad_norm": 0.2175482213497162, "learning_rate": 9.337102870057037e-05, "loss": 0.0112, "step": 16600 }, { "grad_norm": 0.11543218791484833, "learning_rate": 9.336074091598907e-05, "loss": 0.0082, "step": 16610 }, { "grad_norm": 0.182450532913208, "learning_rate": 9.335044572224306e-05, "loss": 0.0093, "step": 16620 }, { "grad_norm": 0.13863219320774078, "learning_rate": 9.334014312109151e-05, "loss": 0.0124, "step": 16630 }, { "grad_norm": 0.16687512397766113, "learning_rate": 9.332983311429486e-05, "loss": 0.0097, "step": 16640 }, { "grad_norm": 0.1737656593322754, "learning_rate": 9.33195157036148e-05, "loss": 0.0138, "step": 16650 }, { "grad_norm": 0.1351543813943863, "learning_rate": 9.330919089081432e-05, "loss": 0.0103, "step": 16660 }, { "grad_norm": 0.1686643660068512, "learning_rate": 9.32988586776576e-05, "loss": 0.0118, "step": 16670 }, { "grad_norm": 0.13183560967445374, "learning_rate": 9.328851906591016e-05, "loss": 0.0088, "step": 16680 }, { "grad_norm": 0.15797922015190125, "learning_rate": 9.327817205733875e-05, "loss": 0.0094, "step": 16690 }, { "grad_norm": 0.1175687238574028, "learning_rate": 9.326781765371142e-05, "loss": 0.0085, "step": 16700 }, { "grad_norm": 0.1703043133020401, "learning_rate": 9.325745585679741e-05, "loss": 0.0107, "step": 16710 }, { "grad_norm": 0.14325842261314392, "learning_rate": 9.32470866683673e-05, "loss": 0.01, "step": 16720 }, { "grad_norm": 0.18299302458763123, "learning_rate": 9.323671009019288e-05, "loss": 0.0108, "step": 16730 }, { "grad_norm": 0.20465706288814545, "learning_rate": 9.322632612404725e-05, "loss": 0.0124, "step": 16740 }, { "grad_norm": 0.19222785532474518, "learning_rate": 9.321593477170471e-05, "loss": 0.0123, "step": 16750 }, { "grad_norm": 0.15877993404865265, "learning_rate": 9.320553603494088e-05, "loss": 0.0115, "step": 16760 }, { "grad_norm": 0.2412835955619812, "learning_rate": 9.319512991553261e-05, "loss": 0.0125, "step": 16770 }, { "grad_norm": 0.14797328412532806, "learning_rate": 9.318471641525803e-05, "loss": 0.0122, "step": 16780 }, { "grad_norm": 0.18106232583522797, "learning_rate": 9.317429553589652e-05, "loss": 0.0087, "step": 16790 }, { "grad_norm": 0.13116398453712463, "learning_rate": 9.316386727922873e-05, "loss": 0.0105, "step": 16800 }, { "grad_norm": 0.17075659334659576, "learning_rate": 9.315343164703656e-05, "loss": 0.0075, "step": 16810 }, { "grad_norm": 0.23030692338943481, "learning_rate": 9.314298864110316e-05, "loss": 0.0093, "step": 16820 }, { "grad_norm": 0.17746320366859436, "learning_rate": 9.313253826321295e-05, "loss": 0.0108, "step": 16830 }, { "grad_norm": 0.19537736475467682, "learning_rate": 9.312208051515165e-05, "loss": 0.0091, "step": 16840 }, { "grad_norm": 0.19754740595817566, "learning_rate": 9.311161539870618e-05, "loss": 0.0118, "step": 16850 }, { "grad_norm": 0.18998770415782928, "learning_rate": 9.310114291566474e-05, "loss": 0.0099, "step": 16860 }, { "grad_norm": 0.18413227796554565, "learning_rate": 9.309066306781679e-05, "loss": 0.0101, "step": 16870 }, { "grad_norm": 0.21695396304130554, "learning_rate": 9.308017585695306e-05, "loss": 0.0111, "step": 16880 }, { "grad_norm": 0.1763121336698532, "learning_rate": 9.306968128486552e-05, "loss": 0.0104, "step": 16890 }, { "grad_norm": 0.13290412724018097, "learning_rate": 9.30591793533474e-05, "loss": 0.011, "step": 16900 }, { "grad_norm": 0.23020347952842712, "learning_rate": 9.304867006419321e-05, "loss": 0.0097, "step": 16910 }, { "grad_norm": 0.1692233830690384, "learning_rate": 9.303815341919868e-05, "loss": 0.0131, "step": 16920 }, { "grad_norm": 0.1395568996667862, "learning_rate": 9.302762942016084e-05, "loss": 0.0088, "step": 16930 }, { "grad_norm": 0.138916477560997, "learning_rate": 9.301709806887792e-05, "loss": 0.0087, "step": 16940 }, { "grad_norm": 0.16115334630012512, "learning_rate": 9.300655936714948e-05, "loss": 0.0125, "step": 16950 }, { "grad_norm": 0.1524776667356491, "learning_rate": 9.299601331677627e-05, "loss": 0.0147, "step": 16960 }, { "grad_norm": 0.1953459531068802, "learning_rate": 9.298545991956033e-05, "loss": 0.012, "step": 16970 }, { "grad_norm": 0.15051716566085815, "learning_rate": 9.297489917730493e-05, "loss": 0.0107, "step": 16980 }, { "grad_norm": 0.19576884806156158, "learning_rate": 9.296433109181464e-05, "loss": 0.0099, "step": 16990 }, { "grad_norm": 0.23875348269939423, "learning_rate": 9.295375566489523e-05, "loss": 0.0112, "step": 17000 }, { "grad_norm": 0.17497260868549347, "learning_rate": 9.294317289835379e-05, "loss": 0.014, "step": 17010 }, { "grad_norm": 0.22423237562179565, "learning_rate": 9.293258279399859e-05, "loss": 0.012, "step": 17020 }, { "grad_norm": 0.21603506803512573, "learning_rate": 9.292198535363919e-05, "loss": 0.0112, "step": 17030 }, { "grad_norm": 0.23508486151695251, "learning_rate": 9.291138057908641e-05, "loss": 0.0118, "step": 17040 }, { "grad_norm": 0.17679181694984436, "learning_rate": 9.290076847215234e-05, "loss": 0.0133, "step": 17050 }, { "grad_norm": 0.20762789249420166, "learning_rate": 9.289014903465025e-05, "loss": 0.012, "step": 17060 }, { "grad_norm": 0.17075157165527344, "learning_rate": 9.287952226839475e-05, "loss": 0.0117, "step": 17070 }, { "grad_norm": 0.19514763355255127, "learning_rate": 9.286888817520164e-05, "loss": 0.0095, "step": 17080 }, { "grad_norm": 0.21586836874485016, "learning_rate": 9.285824675688803e-05, "loss": 0.0118, "step": 17090 }, { "grad_norm": 0.16450394690036774, "learning_rate": 9.28475980152722e-05, "loss": 0.0098, "step": 17100 }, { "grad_norm": 0.16561852395534515, "learning_rate": 9.283694195217379e-05, "loss": 0.0094, "step": 17110 }, { "grad_norm": 0.20682811737060547, "learning_rate": 9.282627856941356e-05, "loss": 0.0105, "step": 17120 }, { "grad_norm": 0.24419048428535461, "learning_rate": 9.281560786881363e-05, "loss": 0.01, "step": 17130 }, { "grad_norm": 0.20303574204444885, "learning_rate": 9.280492985219733e-05, "loss": 0.0131, "step": 17140 }, { "grad_norm": 0.15804888308048248, "learning_rate": 9.279424452138924e-05, "loss": 0.0114, "step": 17150 }, { "grad_norm": 0.172292560338974, "learning_rate": 9.278355187821517e-05, "loss": 0.0119, "step": 17160 }, { "grad_norm": 0.16081967949867249, "learning_rate": 9.277285192450224e-05, "loss": 0.01, "step": 17170 }, { "grad_norm": 0.15437063574790955, "learning_rate": 9.276214466207875e-05, "loss": 0.0095, "step": 17180 }, { "grad_norm": 0.20363979041576385, "learning_rate": 9.275143009277427e-05, "loss": 0.0126, "step": 17190 }, { "grad_norm": 0.16827349364757538, "learning_rate": 9.274070821841964e-05, "loss": 0.0086, "step": 17200 }, { "grad_norm": 0.1445135921239853, "learning_rate": 9.272997904084696e-05, "loss": 0.0077, "step": 17210 }, { "grad_norm": 0.15829049050807953, "learning_rate": 9.271924256188951e-05, "loss": 0.0104, "step": 17220 }, { "grad_norm": 0.17892737686634064, "learning_rate": 9.270849878338189e-05, "loss": 0.013, "step": 17230 }, { "grad_norm": 0.20640955865383148, "learning_rate": 9.269774770715991e-05, "loss": 0.0128, "step": 17240 }, { "grad_norm": 0.22245453298091888, "learning_rate": 9.268698933506061e-05, "loss": 0.0097, "step": 17250 }, { "grad_norm": 0.16872063279151917, "learning_rate": 9.267622366892235e-05, "loss": 0.0119, "step": 17260 }, { "grad_norm": 0.19297939538955688, "learning_rate": 9.266545071058465e-05, "loss": 0.0107, "step": 17270 }, { "grad_norm": 0.14495059847831726, "learning_rate": 9.265467046188833e-05, "loss": 0.0101, "step": 17280 }, { "grad_norm": 0.12450643628835678, "learning_rate": 9.264388292467543e-05, "loss": 0.0083, "step": 17290 }, { "grad_norm": 0.22900518774986267, "learning_rate": 9.263308810078926e-05, "loss": 0.0113, "step": 17300 }, { "grad_norm": 0.21908898651599884, "learning_rate": 9.262228599207434e-05, "loss": 0.0093, "step": 17310 }, { "grad_norm": 0.17558172345161438, "learning_rate": 9.261147660037647e-05, "loss": 0.0104, "step": 17320 }, { "grad_norm": 0.20010392367839813, "learning_rate": 9.26006599275427e-05, "loss": 0.0119, "step": 17330 }, { "grad_norm": 0.15722429752349854, "learning_rate": 9.258983597542124e-05, "loss": 0.01, "step": 17340 }, { "grad_norm": 0.21072454750537872, "learning_rate": 9.257900474586167e-05, "loss": 0.0122, "step": 17350 }, { "grad_norm": 0.22468362748622894, "learning_rate": 9.256816624071471e-05, "loss": 0.0105, "step": 17360 }, { "grad_norm": 0.16704890131950378, "learning_rate": 9.25573204618324e-05, "loss": 0.0093, "step": 17370 }, { "grad_norm": 0.16394901275634766, "learning_rate": 9.254646741106796e-05, "loss": 0.0106, "step": 17380 }, { "grad_norm": 0.1635238230228424, "learning_rate": 9.253560709027589e-05, "loss": 0.0099, "step": 17390 }, { "grad_norm": 0.17826445400714874, "learning_rate": 9.252473950131192e-05, "loss": 0.0119, "step": 17400 }, { "grad_norm": 0.14399191737174988, "learning_rate": 9.251386464603302e-05, "loss": 0.0111, "step": 17410 }, { "grad_norm": 0.1684136688709259, "learning_rate": 9.250298252629741e-05, "loss": 0.0098, "step": 17420 }, { "grad_norm": 0.14428351819515228, "learning_rate": 9.249209314396454e-05, "loss": 0.0097, "step": 17430 }, { "grad_norm": 0.14645890891551971, "learning_rate": 9.248119650089513e-05, "loss": 0.0111, "step": 17440 }, { "grad_norm": 0.19684679806232452, "learning_rate": 9.247029259895108e-05, "loss": 0.0119, "step": 17450 }, { "grad_norm": 0.14710451662540436, "learning_rate": 9.24593814399956e-05, "loss": 0.01, "step": 17460 }, { "grad_norm": 0.2020927369594574, "learning_rate": 9.244846302589309e-05, "loss": 0.0102, "step": 17470 }, { "grad_norm": 0.1904432475566864, "learning_rate": 9.243753735850923e-05, "loss": 0.0115, "step": 17480 }, { "grad_norm": 0.13694705069065094, "learning_rate": 9.24266044397109e-05, "loss": 0.0099, "step": 17490 }, { "grad_norm": 0.1293496936559677, "learning_rate": 9.241566427136624e-05, "loss": 0.0099, "step": 17500 }, { "grad_norm": 0.20948737859725952, "learning_rate": 9.240471685534463e-05, "loss": 0.011, "step": 17510 }, { "grad_norm": 0.20428875088691711, "learning_rate": 9.239376219351667e-05, "loss": 0.0085, "step": 17520 }, { "grad_norm": 0.15593887865543365, "learning_rate": 9.238280028775425e-05, "loss": 0.0102, "step": 17530 }, { "grad_norm": 0.20488639175891876, "learning_rate": 9.237183113993041e-05, "loss": 0.0099, "step": 17540 }, { "grad_norm": 0.2023521512746811, "learning_rate": 9.236085475191952e-05, "loss": 0.0081, "step": 17550 }, { "grad_norm": 0.13772903382778168, "learning_rate": 9.234987112559709e-05, "loss": 0.0094, "step": 17560 }, { "grad_norm": 0.18349184095859528, "learning_rate": 9.233888026283999e-05, "loss": 0.0098, "step": 17570 }, { "grad_norm": 0.17573170363903046, "learning_rate": 9.232788216552619e-05, "loss": 0.0104, "step": 17580 }, { "grad_norm": 0.10911629348993301, "learning_rate": 9.231687683553502e-05, "loss": 0.0104, "step": 17590 }, { "grad_norm": 0.140357106924057, "learning_rate": 9.230586427474698e-05, "loss": 0.0092, "step": 17600 }, { "grad_norm": 0.1548328399658203, "learning_rate": 9.229484448504379e-05, "loss": 0.0107, "step": 17610 }, { "grad_norm": 0.11343459039926529, "learning_rate": 9.228381746830843e-05, "loss": 0.009, "step": 17620 }, { "grad_norm": 0.1398957073688507, "learning_rate": 9.227278322642514e-05, "loss": 0.01, "step": 17630 }, { "grad_norm": 0.1547493040561676, "learning_rate": 9.226174176127937e-05, "loss": 0.0103, "step": 17640 }, { "grad_norm": 0.18698909878730774, "learning_rate": 9.22506930747578e-05, "loss": 0.0079, "step": 17650 }, { "grad_norm": 0.12965209782123566, "learning_rate": 9.223963716874831e-05, "loss": 0.0083, "step": 17660 }, { "grad_norm": 0.15863123536109924, "learning_rate": 9.222857404514012e-05, "loss": 0.0101, "step": 17670 }, { "grad_norm": 0.18137475848197937, "learning_rate": 9.221750370582355e-05, "loss": 0.013, "step": 17680 }, { "grad_norm": 0.1921229213476181, "learning_rate": 9.220642615269028e-05, "loss": 0.0106, "step": 17690 }, { "grad_norm": 0.16905413568019867, "learning_rate": 9.219534138763311e-05, "loss": 0.0113, "step": 17700 }, { "grad_norm": 0.14553889632225037, "learning_rate": 9.218424941254613e-05, "loss": 0.0092, "step": 17710 }, { "grad_norm": 0.14234048128128052, "learning_rate": 9.217315022932468e-05, "loss": 0.011, "step": 17720 }, { "grad_norm": 0.19855915009975433, "learning_rate": 9.216204383986528e-05, "loss": 0.0092, "step": 17730 }, { "grad_norm": 0.24433879554271698, "learning_rate": 9.215093024606574e-05, "loss": 0.0138, "step": 17740 }, { "grad_norm": 0.19682519137859344, "learning_rate": 9.213980944982506e-05, "loss": 0.01, "step": 17750 }, { "grad_norm": 0.20300355553627014, "learning_rate": 9.212868145304346e-05, "loss": 0.0142, "step": 17760 }, { "grad_norm": 0.20013585686683655, "learning_rate": 9.211754625762241e-05, "loss": 0.0114, "step": 17770 }, { "grad_norm": 0.2261241227388382, "learning_rate": 9.210640386546463e-05, "loss": 0.0113, "step": 17780 }, { "grad_norm": 0.14029461145401, "learning_rate": 9.209525427847405e-05, "loss": 0.0105, "step": 17790 }, { "grad_norm": 0.18380500376224518, "learning_rate": 9.208409749855583e-05, "loss": 0.0096, "step": 17800 }, { "grad_norm": 0.22642982006072998, "learning_rate": 9.207293352761633e-05, "loss": 0.0136, "step": 17810 }, { "grad_norm": 0.18682685494422913, "learning_rate": 9.206176236756319e-05, "loss": 0.0107, "step": 17820 }, { "grad_norm": 0.1430812031030655, "learning_rate": 9.205058402030525e-05, "loss": 0.009, "step": 17830 }, { "grad_norm": 0.17145058512687683, "learning_rate": 9.203939848775259e-05, "loss": 0.0106, "step": 17840 }, { "grad_norm": 0.1747303456068039, "learning_rate": 9.202820577181652e-05, "loss": 0.0103, "step": 17850 }, { "grad_norm": 0.20100173354148865, "learning_rate": 9.201700587440953e-05, "loss": 0.0108, "step": 17860 }, { "grad_norm": 0.32026898860931396, "learning_rate": 9.200579879744544e-05, "loss": 0.0142, "step": 17870 }, { "grad_norm": 0.2691306173801422, "learning_rate": 9.199458454283918e-05, "loss": 0.0108, "step": 17880 }, { "grad_norm": 0.20149311423301697, "learning_rate": 9.198336311250697e-05, "loss": 0.0128, "step": 17890 }, { "grad_norm": 0.15250153839588165, "learning_rate": 9.197213450836626e-05, "loss": 0.0107, "step": 17900 }, { "grad_norm": 0.1825007051229477, "learning_rate": 9.19608987323357e-05, "loss": 0.014, "step": 17910 }, { "grad_norm": 0.20010113716125488, "learning_rate": 9.194965578633517e-05, "loss": 0.0148, "step": 17920 }, { "grad_norm": 0.16261017322540283, "learning_rate": 9.193840567228582e-05, "loss": 0.0105, "step": 17930 }, { "grad_norm": 0.17308849096298218, "learning_rate": 9.192714839210994e-05, "loss": 0.0125, "step": 17940 }, { "grad_norm": 0.16624346375465393, "learning_rate": 9.19158839477311e-05, "loss": 0.0085, "step": 17950 }, { "grad_norm": 0.19828791916370392, "learning_rate": 9.190461234107411e-05, "loss": 0.0126, "step": 17960 }, { "grad_norm": 0.17262683808803558, "learning_rate": 9.189333357406496e-05, "loss": 0.0113, "step": 17970 }, { "grad_norm": 0.16263937950134277, "learning_rate": 9.188204764863089e-05, "loss": 0.0111, "step": 17980 }, { "grad_norm": 0.17353887856006622, "learning_rate": 9.187075456670033e-05, "loss": 0.0159, "step": 17990 }, { "grad_norm": 0.18612228333950043, "learning_rate": 9.1859454330203e-05, "loss": 0.0105, "step": 18000 }, { "grad_norm": 0.20704275369644165, "learning_rate": 9.18481469410698e-05, "loss": 0.0137, "step": 18010 }, { "grad_norm": 0.18507099151611328, "learning_rate": 9.183683240123281e-05, "loss": 0.0115, "step": 18020 }, { "grad_norm": 0.17819996178150177, "learning_rate": 9.182551071262541e-05, "loss": 0.0103, "step": 18030 }, { "grad_norm": 0.19553694128990173, "learning_rate": 9.181418187718218e-05, "loss": 0.0108, "step": 18040 }, { "grad_norm": 0.16865819692611694, "learning_rate": 9.180284589683888e-05, "loss": 0.0117, "step": 18050 }, { "grad_norm": 0.23855997622013092, "learning_rate": 9.17915027735325e-05, "loss": 0.0118, "step": 18060 }, { "grad_norm": 0.15100209414958954, "learning_rate": 9.178015250920133e-05, "loss": 0.0116, "step": 18070 }, { "grad_norm": 0.18104737997055054, "learning_rate": 9.176879510578477e-05, "loss": 0.0096, "step": 18080 }, { "grad_norm": 0.14761491119861603, "learning_rate": 9.17574305652235e-05, "loss": 0.0103, "step": 18090 }, { "grad_norm": 0.15721382200717926, "learning_rate": 9.174605888945942e-05, "loss": 0.0106, "step": 18100 }, { "grad_norm": 0.16068926453590393, "learning_rate": 9.173468008043564e-05, "loss": 0.0103, "step": 18110 }, { "grad_norm": 0.2419663965702057, "learning_rate": 9.172329414009648e-05, "loss": 0.0136, "step": 18120 }, { "grad_norm": 0.17502543330192566, "learning_rate": 9.171190107038747e-05, "loss": 0.0117, "step": 18130 }, { "grad_norm": 0.22029203176498413, "learning_rate": 9.170050087325541e-05, "loss": 0.0078, "step": 18140 }, { "grad_norm": 0.19393236935138702, "learning_rate": 9.168909355064824e-05, "loss": 0.0094, "step": 18150 }, { "grad_norm": 0.2260269671678543, "learning_rate": 9.167767910451519e-05, "loss": 0.0104, "step": 18160 }, { "grad_norm": 0.18246763944625854, "learning_rate": 9.166625753680669e-05, "loss": 0.0093, "step": 18170 }, { "grad_norm": 0.1860816925764084, "learning_rate": 9.165482884947431e-05, "loss": 0.0108, "step": 18180 }, { "grad_norm": 4.51571798324585, "learning_rate": 9.164339304447098e-05, "loss": 0.0173, "step": 18190 }, { "grad_norm": 0.19964918494224548, "learning_rate": 9.163195012375072e-05, "loss": 0.0139, "step": 18200 }, { "grad_norm": 0.18829962611198425, "learning_rate": 9.16205000892688e-05, "loss": 0.0118, "step": 18210 }, { "grad_norm": 0.14937153458595276, "learning_rate": 9.160904294298175e-05, "loss": 0.0116, "step": 18220 }, { "grad_norm": 0.23086127638816833, "learning_rate": 9.159757868684727e-05, "loss": 0.0127, "step": 18230 }, { "grad_norm": 0.1794380396604538, "learning_rate": 9.15861073228243e-05, "loss": 0.012, "step": 18240 }, { "grad_norm": 0.12811510264873505, "learning_rate": 9.157462885287296e-05, "loss": 0.0114, "step": 18250 }, { "grad_norm": 0.16685901582241058, "learning_rate": 9.156314327895461e-05, "loss": 0.0104, "step": 18260 }, { "grad_norm": 0.18784338235855103, "learning_rate": 9.155165060303185e-05, "loss": 0.0122, "step": 18270 }, { "grad_norm": 0.14777812361717224, "learning_rate": 9.154015082706841e-05, "loss": 0.0117, "step": 18280 }, { "grad_norm": 0.15585479140281677, "learning_rate": 9.152864395302936e-05, "loss": 0.0098, "step": 18290 }, { "grad_norm": 0.16360600292682648, "learning_rate": 9.151712998288085e-05, "loss": 0.008, "step": 18300 }, { "grad_norm": 0.17056754231452942, "learning_rate": 9.150560891859031e-05, "loss": 0.014, "step": 18310 }, { "grad_norm": 0.17636272311210632, "learning_rate": 9.14940807621264e-05, "loss": 0.0088, "step": 18320 }, { "grad_norm": 0.2215695083141327, "learning_rate": 9.148254551545894e-05, "loss": 0.0135, "step": 18330 }, { "grad_norm": 0.21183577179908752, "learning_rate": 9.147100318055901e-05, "loss": 0.0098, "step": 18340 }, { "grad_norm": 0.15353675186634064, "learning_rate": 9.145945375939888e-05, "loss": 0.0103, "step": 18350 }, { "grad_norm": 0.20014308393001556, "learning_rate": 9.144789725395203e-05, "loss": 0.0112, "step": 18360 }, { "grad_norm": 0.2103727161884308, "learning_rate": 9.14363336661931e-05, "loss": 0.0091, "step": 18370 }, { "grad_norm": 0.1965106874704361, "learning_rate": 9.142476299809806e-05, "loss": 0.0113, "step": 18380 }, { "grad_norm": 0.1845589578151703, "learning_rate": 9.1413185251644e-05, "loss": 0.0106, "step": 18390 }, { "grad_norm": 0.15897352993488312, "learning_rate": 9.140160042880923e-05, "loss": 0.0095, "step": 18400 }, { "grad_norm": 0.2092449963092804, "learning_rate": 9.139000853157327e-05, "loss": 0.01, "step": 18410 }, { "grad_norm": 0.19057436287403107, "learning_rate": 9.137840956191688e-05, "loss": 0.0094, "step": 18420 }, { "grad_norm": 0.15900982916355133, "learning_rate": 9.136680352182199e-05, "loss": 0.0115, "step": 18430 }, { "grad_norm": 0.20168660581111908, "learning_rate": 9.135519041327177e-05, "loss": 0.0152, "step": 18440 }, { "grad_norm": 0.2112208902835846, "learning_rate": 9.134357023825058e-05, "loss": 0.0117, "step": 18450 }, { "grad_norm": 0.18843357264995575, "learning_rate": 9.133194299874398e-05, "loss": 0.0098, "step": 18460 }, { "grad_norm": 0.17205122113227844, "learning_rate": 9.132030869673876e-05, "loss": 0.0098, "step": 18470 }, { "grad_norm": 0.15866626799106598, "learning_rate": 9.130866733422288e-05, "loss": 0.0089, "step": 18480 }, { "grad_norm": 0.16659054160118103, "learning_rate": 9.129701891318556e-05, "loss": 0.0131, "step": 18490 }, { "grad_norm": 0.14391355216503143, "learning_rate": 9.128536343561718e-05, "loss": 0.0104, "step": 18500 }, { "grad_norm": 0.1964254528284073, "learning_rate": 9.127370090350934e-05, "loss": 0.011, "step": 18510 }, { "grad_norm": 0.2152254730463028, "learning_rate": 9.126203131885487e-05, "loss": 0.0106, "step": 18520 }, { "grad_norm": 0.1562746912240982, "learning_rate": 9.125035468364775e-05, "loss": 0.0107, "step": 18530 }, { "grad_norm": 0.2044588327407837, "learning_rate": 9.123867099988322e-05, "loss": 0.0124, "step": 18540 }, { "grad_norm": 0.1897948980331421, "learning_rate": 9.122698026955769e-05, "loss": 0.012, "step": 18550 }, { "grad_norm": 0.13221685588359833, "learning_rate": 9.12152824946688e-05, "loss": 0.0094, "step": 18560 }, { "grad_norm": 0.11945097148418427, "learning_rate": 9.120357767721538e-05, "loss": 0.0083, "step": 18570 }, { "grad_norm": 0.12889905273914337, "learning_rate": 9.119186581919745e-05, "loss": 0.008, "step": 18580 }, { "grad_norm": 0.1638985574245453, "learning_rate": 9.118014692261624e-05, "loss": 0.0112, "step": 18590 }, { "grad_norm": 0.169545978307724, "learning_rate": 9.116842098947422e-05, "loss": 0.0089, "step": 18600 }, { "grad_norm": 0.21648181974887848, "learning_rate": 9.115668802177499e-05, "loss": 0.0112, "step": 18610 }, { "grad_norm": 0.2047313004732132, "learning_rate": 9.114494802152342e-05, "loss": 0.0118, "step": 18620 }, { "grad_norm": 0.17156647145748138, "learning_rate": 9.113320099072555e-05, "loss": 0.0112, "step": 18630 }, { "grad_norm": 0.11874315142631531, "learning_rate": 9.112144693138864e-05, "loss": 0.0093, "step": 18640 }, { "grad_norm": 0.1439257115125656, "learning_rate": 9.110968584552111e-05, "loss": 0.0084, "step": 18650 }, { "grad_norm": 0.10829796642065048, "learning_rate": 9.109791773513264e-05, "loss": 0.0086, "step": 18660 }, { "grad_norm": 0.1415509283542633, "learning_rate": 9.108614260223403e-05, "loss": 0.0088, "step": 18670 }, { "grad_norm": 0.18839243054389954, "learning_rate": 9.107436044883738e-05, "loss": 0.0102, "step": 18680 }, { "grad_norm": 0.21679893136024475, "learning_rate": 9.10625712769559e-05, "loss": 0.0099, "step": 18690 }, { "grad_norm": 0.1254817694425583, "learning_rate": 9.105077508860406e-05, "loss": 0.0094, "step": 18700 }, { "grad_norm": 0.19133183360099792, "learning_rate": 9.103897188579751e-05, "loss": 0.0082, "step": 18710 }, { "grad_norm": 0.14204716682434082, "learning_rate": 9.102716167055308e-05, "loss": 0.0084, "step": 18720 }, { "grad_norm": 0.18369914591312408, "learning_rate": 9.10153444448888e-05, "loss": 0.0102, "step": 18730 }, { "grad_norm": 0.19928686320781708, "learning_rate": 9.100352021082393e-05, "loss": 0.0098, "step": 18740 }, { "grad_norm": 0.14405891299247742, "learning_rate": 9.099168897037891e-05, "loss": 0.0093, "step": 18750 }, { "grad_norm": 0.16594533622264862, "learning_rate": 9.097985072557538e-05, "loss": 0.007, "step": 18760 }, { "grad_norm": 0.15557268261909485, "learning_rate": 9.096800547843615e-05, "loss": 0.0078, "step": 18770 }, { "grad_norm": 0.129851832985878, "learning_rate": 9.095615323098526e-05, "loss": 0.0094, "step": 18780 }, { "grad_norm": 0.13523262739181519, "learning_rate": 9.094429398524795e-05, "loss": 0.0077, "step": 18790 }, { "grad_norm": 0.1670275181531906, "learning_rate": 9.093242774325061e-05, "loss": 0.01, "step": 18800 }, { "grad_norm": 0.17933832108974457, "learning_rate": 9.092055450702088e-05, "loss": 0.0092, "step": 18810 }, { "grad_norm": 0.20414631068706512, "learning_rate": 9.090867427858756e-05, "loss": 0.0115, "step": 18820 }, { "grad_norm": 0.15938875079154968, "learning_rate": 9.089678705998066e-05, "loss": 0.009, "step": 18830 }, { "grad_norm": 0.2478603720664978, "learning_rate": 9.088489285323139e-05, "loss": 0.0098, "step": 18840 }, { "grad_norm": 0.23131291568279266, "learning_rate": 9.087299166037212e-05, "loss": 0.0095, "step": 18850 }, { "grad_norm": 0.16156005859375, "learning_rate": 9.086108348343647e-05, "loss": 0.0084, "step": 18860 }, { "grad_norm": 0.14017023146152496, "learning_rate": 9.08491683244592e-05, "loss": 0.0088, "step": 18870 }, { "grad_norm": 0.1756921410560608, "learning_rate": 9.08372461854763e-05, "loss": 0.0104, "step": 18880 }, { "grad_norm": 0.12739905714988708, "learning_rate": 9.082531706852492e-05, "loss": 0.013, "step": 18890 }, { "grad_norm": 0.15226775407791138, "learning_rate": 9.081338097564342e-05, "loss": 0.0088, "step": 18900 }, { "grad_norm": 0.16201721131801605, "learning_rate": 9.080143790887137e-05, "loss": 0.009, "step": 18910 }, { "grad_norm": 0.18769609928131104, "learning_rate": 9.07894878702495e-05, "loss": 0.0103, "step": 18920 }, { "grad_norm": 0.19615282118320465, "learning_rate": 9.077753086181974e-05, "loss": 0.0093, "step": 18930 }, { "grad_norm": 0.16989238560199738, "learning_rate": 9.076556688562524e-05, "loss": 0.0097, "step": 18940 }, { "grad_norm": 0.14171554148197174, "learning_rate": 9.075359594371029e-05, "loss": 0.0097, "step": 18950 }, { "grad_norm": 0.1644589900970459, "learning_rate": 9.07416180381204e-05, "loss": 0.0108, "step": 18960 }, { "grad_norm": 0.1827389895915985, "learning_rate": 9.072963317090228e-05, "loss": 0.0095, "step": 18970 }, { "grad_norm": 0.18689535558223724, "learning_rate": 9.071764134410382e-05, "loss": 0.0092, "step": 18980 }, { "grad_norm": 0.13951672613620758, "learning_rate": 9.070564255977407e-05, "loss": 0.0091, "step": 18990 }, { "grad_norm": 0.1972116380929947, "learning_rate": 9.06936368199633e-05, "loss": 0.0097, "step": 19000 }, { "grad_norm": 0.16435854136943817, "learning_rate": 9.0681624126723e-05, "loss": 0.0087, "step": 19010 }, { "grad_norm": 0.14630284905433655, "learning_rate": 9.066960448210576e-05, "loss": 0.0089, "step": 19020 }, { "grad_norm": 0.15102249383926392, "learning_rate": 9.065757788816543e-05, "loss": 0.0104, "step": 19030 }, { "grad_norm": 0.15458527207374573, "learning_rate": 9.064554434695705e-05, "loss": 0.0079, "step": 19040 }, { "grad_norm": 0.1968885362148285, "learning_rate": 9.063350386053677e-05, "loss": 0.0092, "step": 19050 }, { "grad_norm": 0.13306601345539093, "learning_rate": 9.062145643096202e-05, "loss": 0.0095, "step": 19060 }, { "grad_norm": 0.16544146835803986, "learning_rate": 9.060940206029136e-05, "loss": 0.0099, "step": 19070 }, { "grad_norm": 0.2010689675807953, "learning_rate": 9.059734075058457e-05, "loss": 0.0134, "step": 19080 }, { "grad_norm": 0.17839738726615906, "learning_rate": 9.058527250390257e-05, "loss": 0.0096, "step": 19090 }, { "grad_norm": 0.1642146110534668, "learning_rate": 9.057319732230752e-05, "loss": 0.0098, "step": 19100 }, { "grad_norm": 0.19990311563014984, "learning_rate": 9.056111520786273e-05, "loss": 0.0092, "step": 19110 }, { "grad_norm": 0.1328330934047699, "learning_rate": 9.054902616263268e-05, "loss": 0.0108, "step": 19120 }, { "grad_norm": 0.14270125329494476, "learning_rate": 9.05369301886831e-05, "loss": 0.0079, "step": 19130 }, { "grad_norm": 0.12138675898313522, "learning_rate": 9.052482728808083e-05, "loss": 0.0091, "step": 19140 }, { "grad_norm": 0.1273273080587387, "learning_rate": 9.051271746289391e-05, "loss": 0.0096, "step": 19150 }, { "grad_norm": 0.1251753866672516, "learning_rate": 9.050060071519162e-05, "loss": 0.008, "step": 19160 }, { "grad_norm": 0.2284957468509674, "learning_rate": 9.048847704704437e-05, "loss": 0.0092, "step": 19170 }, { "grad_norm": 0.156417578458786, "learning_rate": 9.047634646052376e-05, "loss": 0.0107, "step": 19180 }, { "grad_norm": 0.16911277174949646, "learning_rate": 9.046420895770256e-05, "loss": 0.0106, "step": 19190 }, { "grad_norm": 0.1837807297706604, "learning_rate": 9.045206454065473e-05, "loss": 0.0099, "step": 19200 }, { "grad_norm": 0.1905055195093155, "learning_rate": 9.043991321145546e-05, "loss": 0.0101, "step": 19210 }, { "grad_norm": 0.15957443416118622, "learning_rate": 9.042775497218105e-05, "loss": 0.0091, "step": 19220 }, { "grad_norm": 0.17404712736606598, "learning_rate": 9.041558982490901e-05, "loss": 0.0097, "step": 19230 }, { "grad_norm": 0.18278448283672333, "learning_rate": 9.040341777171805e-05, "loss": 0.0091, "step": 19240 }, { "grad_norm": 0.12604562938213348, "learning_rate": 9.039123881468802e-05, "loss": 0.0116, "step": 19250 }, { "grad_norm": 0.08818205446004868, "learning_rate": 9.037905295589998e-05, "loss": 0.008, "step": 19260 }, { "grad_norm": 0.14613915979862213, "learning_rate": 9.036686019743617e-05, "loss": 0.009, "step": 19270 }, { "grad_norm": 0.12129082530736923, "learning_rate": 9.035466054137997e-05, "loss": 0.01, "step": 19280 }, { "grad_norm": 0.14891387522220612, "learning_rate": 9.0342453989816e-05, "loss": 0.0113, "step": 19290 }, { "grad_norm": 0.14692844450473785, "learning_rate": 9.033024054483e-05, "loss": 0.008, "step": 19300 }, { "grad_norm": 0.1285972148180008, "learning_rate": 9.031802020850894e-05, "loss": 0.008, "step": 19310 }, { "grad_norm": 0.16465964913368225, "learning_rate": 9.030579298294092e-05, "loss": 0.0088, "step": 19320 }, { "grad_norm": 0.10367374122142792, "learning_rate": 9.029355887021524e-05, "loss": 0.0113, "step": 19330 }, { "grad_norm": 0.1811913251876831, "learning_rate": 9.028131787242238e-05, "loss": 0.0107, "step": 19340 }, { "grad_norm": 0.1120048388838768, "learning_rate": 9.026906999165399e-05, "loss": 0.0092, "step": 19350 }, { "grad_norm": 0.1600671112537384, "learning_rate": 9.025681523000291e-05, "loss": 0.009, "step": 19360 }, { "grad_norm": 0.12918366491794586, "learning_rate": 9.024455358956315e-05, "loss": 0.0077, "step": 19370 }, { "grad_norm": 0.1506340205669403, "learning_rate": 9.023228507242984e-05, "loss": 0.0097, "step": 19380 }, { "grad_norm": 0.1906869113445282, "learning_rate": 9.022000968069937e-05, "loss": 0.0106, "step": 19390 }, { "grad_norm": 0.16683560609817505, "learning_rate": 9.020772741646928e-05, "loss": 0.0095, "step": 19400 }, { "grad_norm": 0.13067030906677246, "learning_rate": 9.019543828183826e-05, "loss": 0.0116, "step": 19410 }, { "grad_norm": 0.1714831292629242, "learning_rate": 9.018314227890616e-05, "loss": 0.0081, "step": 19420 }, { "grad_norm": 0.1632365733385086, "learning_rate": 9.017083940977408e-05, "loss": 0.0091, "step": 19430 }, { "grad_norm": 0.16888487339019775, "learning_rate": 9.015852967654422e-05, "loss": 0.0106, "step": 19440 }, { "grad_norm": 0.2087714672088623, "learning_rate": 9.014621308131996e-05, "loss": 0.0094, "step": 19450 }, { "grad_norm": 0.19488246738910675, "learning_rate": 9.01338896262059e-05, "loss": 0.0093, "step": 19460 }, { "grad_norm": 0.17614944279193878, "learning_rate": 9.012155931330777e-05, "loss": 0.0111, "step": 19470 }, { "grad_norm": 0.16495341062545776, "learning_rate": 9.010922214473246e-05, "loss": 0.0118, "step": 19480 }, { "grad_norm": 0.17101287841796875, "learning_rate": 9.009687812258808e-05, "loss": 0.0086, "step": 19490 }, { "grad_norm": 0.137446790933609, "learning_rate": 9.00845272489839e-05, "loss": 0.0078, "step": 19500 }, { "grad_norm": 0.14649085700511932, "learning_rate": 9.007216952603031e-05, "loss": 0.0103, "step": 19510 }, { "grad_norm": 0.19474251568317413, "learning_rate": 9.005980495583894e-05, "loss": 0.0086, "step": 19520 }, { "grad_norm": 0.18051765859127045, "learning_rate": 9.004743354052252e-05, "loss": 0.0097, "step": 19530 }, { "grad_norm": 0.12117042392492294, "learning_rate": 9.003505528219503e-05, "loss": 0.0087, "step": 19540 }, { "grad_norm": 0.2229730188846588, "learning_rate": 9.002267018297154e-05, "loss": 0.0101, "step": 19550 }, { "grad_norm": 0.2403622716665268, "learning_rate": 9.001027824496834e-05, "loss": 0.009, "step": 19560 }, { "grad_norm": 0.14240741729736328, "learning_rate": 8.999787947030287e-05, "loss": 0.0104, "step": 19570 }, { "grad_norm": 0.1657925248146057, "learning_rate": 8.998547386109376e-05, "loss": 0.0072, "step": 19580 }, { "grad_norm": 0.13959449529647827, "learning_rate": 8.997306141946073e-05, "loss": 0.0102, "step": 19590 }, { "grad_norm": 0.20046782493591309, "learning_rate": 8.996064214752481e-05, "loss": 0.0098, "step": 19600 }, { "grad_norm": 0.16922780871391296, "learning_rate": 8.994821604740806e-05, "loss": 0.0093, "step": 19610 }, { "grad_norm": 0.15460842847824097, "learning_rate": 8.993578312123377e-05, "loss": 0.0096, "step": 19620 }, { "grad_norm": 0.12716709077358246, "learning_rate": 8.992334337112639e-05, "loss": 0.0096, "step": 19630 }, { "grad_norm": 0.137443408370018, "learning_rate": 8.991089679921154e-05, "loss": 0.0116, "step": 19640 }, { "grad_norm": 0.12162510305643082, "learning_rate": 8.989844340761599e-05, "loss": 0.0084, "step": 19650 }, { "grad_norm": 0.2272624373435974, "learning_rate": 8.988598319846768e-05, "loss": 0.0093, "step": 19660 }, { "grad_norm": 0.11573896557092667, "learning_rate": 8.987351617389574e-05, "loss": 0.0092, "step": 19670 }, { "grad_norm": 0.10991434752941132, "learning_rate": 8.98610423360304e-05, "loss": 0.0102, "step": 19680 }, { "grad_norm": 0.1840703934431076, "learning_rate": 8.984856168700317e-05, "loss": 0.0085, "step": 19690 }, { "grad_norm": 0.1732654720544815, "learning_rate": 8.983607422894658e-05, "loss": 0.0079, "step": 19700 }, { "grad_norm": 0.11920543015003204, "learning_rate": 8.982357996399442e-05, "loss": 0.0113, "step": 19710 }, { "grad_norm": 0.15270459651947021, "learning_rate": 8.981107889428164e-05, "loss": 0.0097, "step": 19720 }, { "grad_norm": 0.15397955477237701, "learning_rate": 8.979857102194428e-05, "loss": 0.0121, "step": 19730 }, { "grad_norm": 0.21793010830879211, "learning_rate": 8.978605634911968e-05, "loss": 0.0092, "step": 19740 }, { "grad_norm": 0.14166785776615143, "learning_rate": 8.977353487794616e-05, "loss": 0.0089, "step": 19750 }, { "grad_norm": 0.1617382913827896, "learning_rate": 8.976100661056334e-05, "loss": 0.0087, "step": 19760 }, { "grad_norm": 0.22711139917373657, "learning_rate": 8.974847154911197e-05, "loss": 0.0107, "step": 19770 }, { "grad_norm": 0.18404635787010193, "learning_rate": 8.973592969573393e-05, "loss": 0.0084, "step": 19780 }, { "grad_norm": 0.17584741115570068, "learning_rate": 8.972338105257228e-05, "loss": 0.0089, "step": 19790 }, { "grad_norm": 0.12070085853338242, "learning_rate": 8.971082562177125e-05, "loss": 0.0074, "step": 19800 }, { "grad_norm": 0.19725999236106873, "learning_rate": 8.96982634054762e-05, "loss": 0.0086, "step": 19810 }, { "grad_norm": 0.18131206929683685, "learning_rate": 8.96856944058337e-05, "loss": 0.0095, "step": 19820 }, { "grad_norm": 0.21993499994277954, "learning_rate": 8.967311862499144e-05, "loss": 0.0108, "step": 19830 }, { "grad_norm": 0.23408766090869904, "learning_rate": 8.966053606509825e-05, "loss": 0.0092, "step": 19840 }, { "grad_norm": 0.14820250868797302, "learning_rate": 8.964794672830417e-05, "loss": 0.009, "step": 19850 }, { "grad_norm": 0.21167834103107452, "learning_rate": 8.963535061676038e-05, "loss": 0.011, "step": 19860 }, { "grad_norm": 0.1500893235206604, "learning_rate": 8.962274773261918e-05, "loss": 0.011, "step": 19870 }, { "grad_norm": 0.16431879997253418, "learning_rate": 8.961013807803409e-05, "loss": 0.0115, "step": 19880 }, { "grad_norm": 0.138282909989357, "learning_rate": 8.959752165515973e-05, "loss": 0.0105, "step": 19890 }, { "grad_norm": 0.1592382788658142, "learning_rate": 8.958489846615193e-05, "loss": 0.0101, "step": 19900 }, { "grad_norm": 0.22258444130420685, "learning_rate": 8.957226851316762e-05, "loss": 0.012, "step": 19910 }, { "grad_norm": 0.18824760615825653, "learning_rate": 8.955963179836493e-05, "loss": 0.0092, "step": 19920 }, { "grad_norm": 0.18149307370185852, "learning_rate": 8.954698832390312e-05, "loss": 0.0092, "step": 19930 }, { "grad_norm": 0.1757144033908844, "learning_rate": 8.953433809194263e-05, "loss": 0.0093, "step": 19940 }, { "grad_norm": 0.17519481480121613, "learning_rate": 8.9521681104645e-05, "loss": 0.0089, "step": 19950 }, { "grad_norm": 0.1966841220855713, "learning_rate": 8.9509017364173e-05, "loss": 0.0102, "step": 19960 }, { "grad_norm": 0.17404960095882416, "learning_rate": 8.949634687269052e-05, "loss": 0.0105, "step": 19970 }, { "grad_norm": 0.17228345572948456, "learning_rate": 8.948366963236259e-05, "loss": 0.009, "step": 19980 }, { "grad_norm": 0.13497941195964813, "learning_rate": 8.947098564535538e-05, "loss": 0.0081, "step": 19990 }, { "grad_norm": 0.14388783276081085, "learning_rate": 8.945829491383627e-05, "loss": 0.0092, "step": 20000 }, { "grad_norm": 0.147835835814476, "learning_rate": 8.944559743997374e-05, "loss": 0.0118, "step": 20010 }, { "grad_norm": 0.15688154101371765, "learning_rate": 8.943289322593746e-05, "loss": 0.0092, "step": 20020 }, { "grad_norm": 0.18434692919254303, "learning_rate": 8.942018227389821e-05, "loss": 0.0093, "step": 20030 }, { "grad_norm": 0.1965884268283844, "learning_rate": 8.940746458602795e-05, "loss": 0.0089, "step": 20040 }, { "grad_norm": 0.15852010250091553, "learning_rate": 8.939474016449979e-05, "loss": 0.0099, "step": 20050 }, { "grad_norm": 0.1682555228471756, "learning_rate": 8.938200901148799e-05, "loss": 0.011, "step": 20060 }, { "grad_norm": 0.18402384221553802, "learning_rate": 8.936927112916795e-05, "loss": 0.0132, "step": 20070 }, { "grad_norm": 0.1843010038137436, "learning_rate": 8.935652651971622e-05, "loss": 0.0095, "step": 20080 }, { "grad_norm": 0.15877346694469452, "learning_rate": 8.934377518531052e-05, "loss": 0.0119, "step": 20090 }, { "grad_norm": 0.16498254239559174, "learning_rate": 8.933101712812967e-05, "loss": 0.0143, "step": 20100 }, { "grad_norm": 0.18379181623458862, "learning_rate": 8.931825235035374e-05, "loss": 0.0122, "step": 20110 }, { "grad_norm": 0.18471406400203705, "learning_rate": 8.930548085416382e-05, "loss": 0.0123, "step": 20120 }, { "grad_norm": 0.26300033926963806, "learning_rate": 8.92927026417422e-05, "loss": 0.0174, "step": 20130 }, { "grad_norm": 0.24894623458385468, "learning_rate": 8.92799177152724e-05, "loss": 0.013, "step": 20140 }, { "grad_norm": 0.15130683779716492, "learning_rate": 8.926712607693895e-05, "loss": 0.018, "step": 20150 }, { "grad_norm": 0.20298433303833008, "learning_rate": 8.925432772892762e-05, "loss": 0.0132, "step": 20160 }, { "grad_norm": 0.2462146282196045, "learning_rate": 8.924152267342529e-05, "loss": 0.0112, "step": 20170 }, { "grad_norm": 0.1266808807849884, "learning_rate": 8.922871091261998e-05, "loss": 0.0116, "step": 20180 }, { "grad_norm": 0.18371380865573883, "learning_rate": 8.92158924487009e-05, "loss": 0.013, "step": 20190 }, { "grad_norm": 0.1595817357301712, "learning_rate": 8.920306728385834e-05, "loss": 0.0095, "step": 20200 }, { "grad_norm": 0.17496085166931152, "learning_rate": 8.919023542028379e-05, "loss": 0.0111, "step": 20210 }, { "grad_norm": 0.18525680899620056, "learning_rate": 8.917739686016988e-05, "loss": 0.0125, "step": 20220 }, { "grad_norm": 0.15489336848258972, "learning_rate": 8.916455160571033e-05, "loss": 0.0083, "step": 20230 }, { "grad_norm": 0.21743056178092957, "learning_rate": 8.915169965910008e-05, "loss": 0.0091, "step": 20240 }, { "grad_norm": 0.16889688372612, "learning_rate": 8.913884102253514e-05, "loss": 0.0112, "step": 20250 }, { "grad_norm": 0.17794491350650787, "learning_rate": 8.912597569821273e-05, "loss": 0.0112, "step": 20260 }, { "grad_norm": 0.1834961622953415, "learning_rate": 8.911310368833118e-05, "loss": 0.0086, "step": 20270 }, { "grad_norm": 0.18842697143554688, "learning_rate": 8.910022499508994e-05, "loss": 0.0103, "step": 20280 }, { "grad_norm": 0.20461682975292206, "learning_rate": 8.908733962068965e-05, "loss": 0.0105, "step": 20290 }, { "grad_norm": 0.19234126806259155, "learning_rate": 8.907444756733207e-05, "loss": 0.0101, "step": 20300 }, { "grad_norm": 0.14743389189243317, "learning_rate": 8.906154883722006e-05, "loss": 0.0141, "step": 20310 }, { "grad_norm": 0.14421367645263672, "learning_rate": 8.904864343255773e-05, "loss": 0.0109, "step": 20320 }, { "grad_norm": 0.12880738079547882, "learning_rate": 8.90357313555502e-05, "loss": 0.0093, "step": 20330 }, { "grad_norm": 0.1548612266778946, "learning_rate": 8.90228126084038e-05, "loss": 0.0099, "step": 20340 }, { "grad_norm": 0.16964709758758545, "learning_rate": 8.900988719332601e-05, "loss": 0.0129, "step": 20350 }, { "grad_norm": 0.12761680781841278, "learning_rate": 8.899695511252542e-05, "loss": 0.0111, "step": 20360 }, { "grad_norm": 0.19747017323970795, "learning_rate": 8.898401636821176e-05, "loss": 0.01, "step": 20370 }, { "grad_norm": 0.18782611191272736, "learning_rate": 8.897107096259593e-05, "loss": 0.0107, "step": 20380 }, { "grad_norm": 0.14860716462135315, "learning_rate": 8.895811889788994e-05, "loss": 0.0133, "step": 20390 }, { "grad_norm": 0.14527547359466553, "learning_rate": 8.894516017630692e-05, "loss": 0.0104, "step": 20400 }, { "grad_norm": 0.1439020335674286, "learning_rate": 8.893219480006118e-05, "loss": 0.0095, "step": 20410 }, { "grad_norm": 0.18707561492919922, "learning_rate": 8.891922277136817e-05, "loss": 0.0061, "step": 20420 }, { "grad_norm": 0.17881737649440765, "learning_rate": 8.890624409244441e-05, "loss": 0.0073, "step": 20430 }, { "grad_norm": 0.1281491369009018, "learning_rate": 8.889325876550763e-05, "loss": 0.0088, "step": 20440 }, { "grad_norm": 0.18020236492156982, "learning_rate": 8.888026679277666e-05, "loss": 0.008, "step": 20450 }, { "grad_norm": 0.1765107959508896, "learning_rate": 8.886726817647147e-05, "loss": 0.0107, "step": 20460 }, { "grad_norm": 0.15949803590774536, "learning_rate": 8.885426291881319e-05, "loss": 0.0088, "step": 20470 }, { "grad_norm": 0.17603479325771332, "learning_rate": 8.884125102202401e-05, "loss": 0.0104, "step": 20480 }, { "grad_norm": 0.1663377732038498, "learning_rate": 8.882823248832736e-05, "loss": 0.0107, "step": 20490 }, { "grad_norm": 0.1421123892068863, "learning_rate": 8.881520731994772e-05, "loss": 0.0102, "step": 20500 }, { "grad_norm": 0.13009756803512573, "learning_rate": 8.880217551911077e-05, "loss": 0.0079, "step": 20510 }, { "grad_norm": 0.13086062669754028, "learning_rate": 8.878913708804323e-05, "loss": 0.0124, "step": 20520 }, { "grad_norm": 0.1627153754234314, "learning_rate": 8.877609202897308e-05, "loss": 0.0111, "step": 20530 }, { "grad_norm": 0.15220680832862854, "learning_rate": 8.876304034412933e-05, "loss": 0.0076, "step": 20540 }, { "grad_norm": 0.17454637587070465, "learning_rate": 8.874998203574214e-05, "loss": 0.0102, "step": 20550 }, { "grad_norm": 0.12742453813552856, "learning_rate": 8.873691710604284e-05, "loss": 0.0114, "step": 20560 }, { "grad_norm": 0.14872094988822937, "learning_rate": 8.872384555726387e-05, "loss": 0.0083, "step": 20570 }, { "grad_norm": 0.1544964611530304, "learning_rate": 8.871076739163878e-05, "loss": 0.0107, "step": 20580 }, { "grad_norm": 0.1322026550769806, "learning_rate": 8.86976826114023e-05, "loss": 0.0091, "step": 20590 }, { "grad_norm": 0.17459602653980255, "learning_rate": 8.868459121879023e-05, "loss": 0.0082, "step": 20600 }, { "grad_norm": 0.161843940615654, "learning_rate": 8.867149321603956e-05, "loss": 0.0088, "step": 20610 }, { "grad_norm": 0.14748162031173706, "learning_rate": 8.865838860538835e-05, "loss": 0.008, "step": 20620 }, { "grad_norm": 0.16001541912555695, "learning_rate": 8.864527738907585e-05, "loss": 0.0105, "step": 20630 }, { "grad_norm": 0.18648365139961243, "learning_rate": 8.863215956934239e-05, "loss": 0.0106, "step": 20640 }, { "grad_norm": 0.13830344378948212, "learning_rate": 8.861903514842947e-05, "loss": 0.0106, "step": 20650 }, { "grad_norm": 0.1513444185256958, "learning_rate": 8.860590412857966e-05, "loss": 0.012, "step": 20660 }, { "grad_norm": 0.16246888041496277, "learning_rate": 8.85927665120367e-05, "loss": 0.0101, "step": 20670 }, { "grad_norm": 0.21576759219169617, "learning_rate": 8.857962230104546e-05, "loss": 0.0091, "step": 20680 }, { "grad_norm": 0.14029008150100708, "learning_rate": 8.856647149785193e-05, "loss": 0.0091, "step": 20690 }, { "grad_norm": 0.12683404982089996, "learning_rate": 8.855331410470322e-05, "loss": 0.0101, "step": 20700 }, { "grad_norm": 0.11443334072828293, "learning_rate": 8.854015012384756e-05, "loss": 0.0099, "step": 20710 }, { "grad_norm": 0.1288835108280182, "learning_rate": 8.852697955753433e-05, "loss": 0.0099, "step": 20720 }, { "grad_norm": 0.18767325580120087, "learning_rate": 8.851380240801399e-05, "loss": 0.0119, "step": 20730 }, { "grad_norm": 0.20798027515411377, "learning_rate": 8.850061867753818e-05, "loss": 0.0132, "step": 20740 }, { "grad_norm": 0.20468351244926453, "learning_rate": 8.848742836835963e-05, "loss": 0.0101, "step": 20750 }, { "grad_norm": 0.1630464494228363, "learning_rate": 8.847423148273221e-05, "loss": 0.0101, "step": 20760 }, { "grad_norm": 0.18770332634449005, "learning_rate": 8.846102802291092e-05, "loss": 0.0114, "step": 20770 }, { "grad_norm": 0.16968050599098206, "learning_rate": 8.844781799115183e-05, "loss": 0.0088, "step": 20780 }, { "grad_norm": 0.15768319368362427, "learning_rate": 8.84346013897122e-05, "loss": 0.0094, "step": 20790 }, { "grad_norm": 0.1807251125574112, "learning_rate": 8.842137822085038e-05, "loss": 0.0084, "step": 20800 }, { "grad_norm": 0.16945362091064453, "learning_rate": 8.840814848682585e-05, "loss": 0.0078, "step": 20810 }, { "grad_norm": 0.1467355340719223, "learning_rate": 8.83949121898992e-05, "loss": 0.0094, "step": 20820 }, { "grad_norm": 0.1513415426015854, "learning_rate": 8.838166933233217e-05, "loss": 0.01, "step": 20830 }, { "grad_norm": 0.1440252810716629, "learning_rate": 8.83684199163876e-05, "loss": 0.0076, "step": 20840 }, { "grad_norm": 0.1731606125831604, "learning_rate": 8.835516394432943e-05, "loss": 0.0091, "step": 20850 }, { "grad_norm": 0.1797374039888382, "learning_rate": 8.834190141842276e-05, "loss": 0.0088, "step": 20860 }, { "grad_norm": 0.1358875036239624, "learning_rate": 8.83286323409338e-05, "loss": 0.0088, "step": 20870 }, { "grad_norm": 0.18755488097667694, "learning_rate": 8.831535671412986e-05, "loss": 0.0086, "step": 20880 }, { "grad_norm": 0.15445201098918915, "learning_rate": 8.830207454027938e-05, "loss": 0.0088, "step": 20890 }, { "grad_norm": 0.14534594118595123, "learning_rate": 8.828878582165192e-05, "loss": 0.0073, "step": 20900 }, { "grad_norm": 0.12666264176368713, "learning_rate": 8.827549056051818e-05, "loss": 0.01, "step": 20910 }, { "grad_norm": 0.12128260731697083, "learning_rate": 8.826218875914993e-05, "loss": 0.0092, "step": 20920 }, { "grad_norm": 0.1662335842847824, "learning_rate": 8.82488804198201e-05, "loss": 0.0084, "step": 20930 }, { "grad_norm": 0.15353526175022125, "learning_rate": 8.82355655448027e-05, "loss": 0.0111, "step": 20940 }, { "grad_norm": 0.13747651875019073, "learning_rate": 8.822224413637293e-05, "loss": 0.0071, "step": 20950 }, { "grad_norm": 0.17996343970298767, "learning_rate": 8.820891619680697e-05, "loss": 0.0083, "step": 20960 }, { "grad_norm": 0.09205595403909683, "learning_rate": 8.819558172838227e-05, "loss": 0.0083, "step": 20970 }, { "grad_norm": 0.1432253122329712, "learning_rate": 8.818224073337731e-05, "loss": 0.0089, "step": 20980 }, { "grad_norm": 0.1428586095571518, "learning_rate": 8.816889321407169e-05, "loss": 0.0073, "step": 20990 }, { "grad_norm": 0.1108134388923645, "learning_rate": 8.815553917274615e-05, "loss": 0.0094, "step": 21000 }, { "grad_norm": 0.18496842682361603, "learning_rate": 8.81421786116825e-05, "loss": 0.0075, "step": 21010 }, { "grad_norm": 0.17934785783290863, "learning_rate": 8.812881153316373e-05, "loss": 0.01, "step": 21020 }, { "grad_norm": 0.14300189912319183, "learning_rate": 8.81154379394739e-05, "loss": 0.0083, "step": 21030 }, { "grad_norm": 0.16332758963108063, "learning_rate": 8.810205783289818e-05, "loss": 0.0094, "step": 21040 }, { "grad_norm": 0.19431614875793457, "learning_rate": 8.808867121572286e-05, "loss": 0.0114, "step": 21050 }, { "grad_norm": 0.15155886113643646, "learning_rate": 8.807527809023537e-05, "loss": 0.0079, "step": 21060 }, { "grad_norm": 0.1678893268108368, "learning_rate": 8.80618784587242e-05, "loss": 0.0101, "step": 21070 }, { "grad_norm": 0.13568434119224548, "learning_rate": 8.804847232347902e-05, "loss": 0.0089, "step": 21080 }, { "grad_norm": 0.17379629611968994, "learning_rate": 8.803505968679054e-05, "loss": 0.0083, "step": 21090 }, { "grad_norm": 0.16590701043605804, "learning_rate": 8.802164055095061e-05, "loss": 0.008, "step": 21100 }, { "grad_norm": 0.18998488783836365, "learning_rate": 8.80082149182522e-05, "loss": 0.0119, "step": 21110 }, { "grad_norm": 0.15569783747196198, "learning_rate": 8.79947827909894e-05, "loss": 0.0083, "step": 21120 }, { "grad_norm": 0.1535004824399948, "learning_rate": 8.798134417145738e-05, "loss": 0.0108, "step": 21130 }, { "grad_norm": 0.171611025929451, "learning_rate": 8.796789906195243e-05, "loss": 0.0108, "step": 21140 }, { "grad_norm": 0.12317012250423431, "learning_rate": 8.795444746477195e-05, "loss": 0.0102, "step": 21150 }, { "grad_norm": 0.15823715925216675, "learning_rate": 8.794098938221446e-05, "loss": 0.0086, "step": 21160 }, { "grad_norm": 0.1402980536222458, "learning_rate": 8.792752481657957e-05, "loss": 0.0089, "step": 21170 }, { "grad_norm": 0.19027599692344666, "learning_rate": 8.791405377016802e-05, "loss": 0.0099, "step": 21180 }, { "grad_norm": 0.16966533660888672, "learning_rate": 8.790057624528163e-05, "loss": 0.0115, "step": 21190 }, { "grad_norm": 0.18563465774059296, "learning_rate": 8.788709224422333e-05, "loss": 0.0097, "step": 21200 }, { "grad_norm": 0.14873260259628296, "learning_rate": 8.787360176929717e-05, "loss": 0.0085, "step": 21210 }, { "grad_norm": 0.20226003229618073, "learning_rate": 8.786010482280834e-05, "loss": 0.0096, "step": 21220 }, { "grad_norm": 0.14826913177967072, "learning_rate": 8.784660140706306e-05, "loss": 0.0111, "step": 21230 }, { "grad_norm": 0.13569919764995575, "learning_rate": 8.783309152436872e-05, "loss": 0.0089, "step": 21240 }, { "grad_norm": 0.13431009650230408, "learning_rate": 8.781957517703375e-05, "loss": 0.0082, "step": 21250 }, { "grad_norm": 0.18029938638210297, "learning_rate": 8.780605236736776e-05, "loss": 0.0082, "step": 21260 }, { "grad_norm": 0.15958993136882782, "learning_rate": 8.779252309768142e-05, "loss": 0.0102, "step": 21270 }, { "grad_norm": 0.1650179624557495, "learning_rate": 8.777898737028652e-05, "loss": 0.0092, "step": 21280 }, { "grad_norm": 0.17632311582565308, "learning_rate": 8.776544518749591e-05, "loss": 0.0094, "step": 21290 }, { "grad_norm": 0.16970962285995483, "learning_rate": 8.775189655162364e-05, "loss": 0.0088, "step": 21300 }, { "grad_norm": 0.13105791807174683, "learning_rate": 8.773834146498474e-05, "loss": 0.0089, "step": 21310 }, { "grad_norm": 0.15332093834877014, "learning_rate": 8.772477992989545e-05, "loss": 0.0092, "step": 21320 }, { "grad_norm": 0.13461290299892426, "learning_rate": 8.771121194867304e-05, "loss": 0.0065, "step": 21330 }, { "grad_norm": 0.11316737532615662, "learning_rate": 8.769763752363589e-05, "loss": 0.0102, "step": 21340 }, { "grad_norm": 0.206864133477211, "learning_rate": 8.768405665710352e-05, "loss": 0.0103, "step": 21350 }, { "grad_norm": 0.20696663856506348, "learning_rate": 8.767046935139655e-05, "loss": 0.0085, "step": 21360 }, { "grad_norm": 0.12791360914707184, "learning_rate": 8.765687560883666e-05, "loss": 0.0091, "step": 21370 }, { "grad_norm": 0.1485678255558014, "learning_rate": 8.764327543174664e-05, "loss": 0.0074, "step": 21380 }, { "grad_norm": 0.14471104741096497, "learning_rate": 8.762966882245038e-05, "loss": 0.0086, "step": 21390 }, { "grad_norm": 0.1411740630865097, "learning_rate": 8.761605578327291e-05, "loss": 0.0087, "step": 21400 }, { "grad_norm": 0.1408502608537674, "learning_rate": 8.76024363165403e-05, "loss": 0.0086, "step": 21410 }, { "grad_norm": 0.14993083477020264, "learning_rate": 8.758881042457976e-05, "loss": 0.0091, "step": 21420 }, { "grad_norm": 0.149753600358963, "learning_rate": 8.757517810971957e-05, "loss": 0.0085, "step": 21430 }, { "grad_norm": 0.15406979620456696, "learning_rate": 8.756153937428913e-05, "loss": 0.0098, "step": 21440 }, { "grad_norm": 0.14690355956554413, "learning_rate": 8.754789422061889e-05, "loss": 0.0126, "step": 21450 }, { "grad_norm": 0.15162602066993713, "learning_rate": 8.753424265104052e-05, "loss": 0.0071, "step": 21460 }, { "grad_norm": 0.14627276360988617, "learning_rate": 8.752058466788659e-05, "loss": 0.0065, "step": 21470 }, { "grad_norm": 0.1447760909795761, "learning_rate": 8.750692027349097e-05, "loss": 0.0086, "step": 21480 }, { "grad_norm": 0.1507951021194458, "learning_rate": 8.749324947018847e-05, "loss": 0.0086, "step": 21490 }, { "grad_norm": 0.11452945321798325, "learning_rate": 8.747957226031507e-05, "loss": 0.0093, "step": 21500 }, { "grad_norm": 0.1186748594045639, "learning_rate": 8.746588864620787e-05, "loss": 0.0075, "step": 21510 }, { "grad_norm": 0.15099959075450897, "learning_rate": 8.745219863020498e-05, "loss": 0.0093, "step": 21520 }, { "grad_norm": 0.14981669187545776, "learning_rate": 8.743850221464564e-05, "loss": 0.0093, "step": 21530 }, { "grad_norm": 0.1558770090341568, "learning_rate": 8.742479940187026e-05, "loss": 0.0098, "step": 21540 }, { "grad_norm": 0.16649354994297028, "learning_rate": 8.74110901942202e-05, "loss": 0.0082, "step": 21550 }, { "grad_norm": 0.18098938465118408, "learning_rate": 8.739737459403803e-05, "loss": 0.0086, "step": 21560 }, { "grad_norm": 0.20455648005008698, "learning_rate": 8.738365260366737e-05, "loss": 0.007, "step": 21570 }, { "grad_norm": 0.13339188694953918, "learning_rate": 8.736992422545292e-05, "loss": 0.0099, "step": 21580 }, { "grad_norm": 0.15985487401485443, "learning_rate": 8.73561894617405e-05, "loss": 0.0098, "step": 21590 }, { "grad_norm": 0.16122588515281677, "learning_rate": 8.734244831487697e-05, "loss": 0.0104, "step": 21600 }, { "grad_norm": 0.16048534214496613, "learning_rate": 8.732870078721035e-05, "loss": 0.0087, "step": 21610 }, { "grad_norm": 0.14989882707595825, "learning_rate": 8.731494688108972e-05, "loss": 0.0097, "step": 21620 }, { "grad_norm": 0.12988166511058807, "learning_rate": 8.730118659886523e-05, "loss": 0.0097, "step": 21630 }, { "grad_norm": 0.14727984368801117, "learning_rate": 8.728741994288814e-05, "loss": 0.0099, "step": 21640 }, { "grad_norm": 0.17795129120349884, "learning_rate": 8.727364691551079e-05, "loss": 0.0076, "step": 21650 }, { "grad_norm": 0.1935635805130005, "learning_rate": 8.725986751908661e-05, "loss": 0.0107, "step": 21660 }, { "grad_norm": 0.18401823937892914, "learning_rate": 8.724608175597016e-05, "loss": 0.0094, "step": 21670 }, { "grad_norm": 0.22287335991859436, "learning_rate": 8.723228962851699e-05, "loss": 0.0119, "step": 21680 }, { "grad_norm": 0.14155560731887817, "learning_rate": 8.721849113908385e-05, "loss": 0.0099, "step": 21690 }, { "grad_norm": 0.19777485728263855, "learning_rate": 8.720468629002848e-05, "loss": 0.0107, "step": 21700 }, { "grad_norm": 0.1715584546327591, "learning_rate": 8.719087508370978e-05, "loss": 0.0128, "step": 21710 }, { "grad_norm": 0.16291949152946472, "learning_rate": 8.717705752248772e-05, "loss": 0.0145, "step": 21720 }, { "grad_norm": 0.1706172674894333, "learning_rate": 8.71632336087233e-05, "loss": 0.0128, "step": 21730 }, { "grad_norm": 0.13815855979919434, "learning_rate": 8.71494033447787e-05, "loss": 0.0098, "step": 21740 }, { "grad_norm": 0.14846907556056976, "learning_rate": 8.713556673301708e-05, "loss": 0.0105, "step": 21750 }, { "grad_norm": 0.12940293550491333, "learning_rate": 8.712172377580278e-05, "loss": 0.0094, "step": 21760 }, { "grad_norm": 0.15624397993087769, "learning_rate": 8.710787447550114e-05, "loss": 0.0114, "step": 21770 }, { "grad_norm": 0.23243139684200287, "learning_rate": 8.70940188344787e-05, "loss": 0.0086, "step": 21780 }, { "grad_norm": 0.18520568311214447, "learning_rate": 8.708015685510293e-05, "loss": 0.0086, "step": 21790 }, { "grad_norm": 0.1613849401473999, "learning_rate": 8.706628853974252e-05, "loss": 0.0086, "step": 21800 }, { "grad_norm": 0.11511119455099106, "learning_rate": 8.705241389076715e-05, "loss": 0.009, "step": 21810 }, { "grad_norm": 0.17820486426353455, "learning_rate": 8.703853291054764e-05, "loss": 0.0068, "step": 21820 }, { "grad_norm": 0.13014836609363556, "learning_rate": 8.702464560145587e-05, "loss": 0.0073, "step": 21830 }, { "grad_norm": 0.17509467899799347, "learning_rate": 8.701075196586476e-05, "loss": 0.0085, "step": 21840 }, { "grad_norm": 0.17867910861968994, "learning_rate": 8.699685200614842e-05, "loss": 0.0104, "step": 21850 }, { "grad_norm": 0.11835958063602448, "learning_rate": 8.698294572468193e-05, "loss": 0.0089, "step": 21860 }, { "grad_norm": 0.15372729301452637, "learning_rate": 8.696903312384148e-05, "loss": 0.0099, "step": 21870 }, { "grad_norm": 0.19566136598587036, "learning_rate": 8.695511420600439e-05, "loss": 0.0082, "step": 21880 }, { "grad_norm": 0.1848691999912262, "learning_rate": 8.694118897354901e-05, "loss": 0.0091, "step": 21890 }, { "grad_norm": 0.14756883680820465, "learning_rate": 8.692725742885478e-05, "loss": 0.01, "step": 21900 }, { "grad_norm": 0.11353421211242676, "learning_rate": 8.691331957430221e-05, "loss": 0.0083, "step": 21910 }, { "grad_norm": 0.18888171017169952, "learning_rate": 8.68993754122729e-05, "loss": 0.0084, "step": 21920 }, { "grad_norm": 0.14700759947299957, "learning_rate": 8.688542494514955e-05, "loss": 0.0086, "step": 21930 }, { "grad_norm": 0.13738450407981873, "learning_rate": 8.68714681753159e-05, "loss": 0.0088, "step": 21940 }, { "grad_norm": 0.1384442001581192, "learning_rate": 8.685750510515676e-05, "loss": 0.0071, "step": 21950 }, { "grad_norm": 0.13100382685661316, "learning_rate": 8.684353573705805e-05, "loss": 0.0085, "step": 21960 }, { "grad_norm": 0.15037290751934052, "learning_rate": 8.682956007340677e-05, "loss": 0.0097, "step": 21970 }, { "grad_norm": 0.17013344168663025, "learning_rate": 8.681557811659095e-05, "loss": 0.0087, "step": 21980 }, { "grad_norm": 0.1101604625582695, "learning_rate": 8.680158986899974e-05, "loss": 0.0083, "step": 21990 }, { "grad_norm": 0.2092888355255127, "learning_rate": 8.678759533302335e-05, "loss": 0.0069, "step": 22000 }, { "grad_norm": 0.12241429090499878, "learning_rate": 8.677359451105308e-05, "loss": 0.0077, "step": 22010 }, { "grad_norm": 0.146726593375206, "learning_rate": 8.675958740548123e-05, "loss": 0.0082, "step": 22020 }, { "grad_norm": 0.14201118052005768, "learning_rate": 8.674557401870129e-05, "loss": 0.0081, "step": 22030 }, { "grad_norm": 0.1677130162715912, "learning_rate": 8.673155435310775e-05, "loss": 0.0092, "step": 22040 }, { "grad_norm": 0.15196004509925842, "learning_rate": 8.671752841109617e-05, "loss": 0.0076, "step": 22050 }, { "grad_norm": 0.12279882282018661, "learning_rate": 8.670349619506321e-05, "loss": 0.0075, "step": 22060 }, { "grad_norm": 0.20865574479103088, "learning_rate": 8.66894577074066e-05, "loss": 0.0085, "step": 22070 }, { "grad_norm": 0.1505349576473236, "learning_rate": 8.667541295052513e-05, "loss": 0.0073, "step": 22080 }, { "grad_norm": 0.17009559273719788, "learning_rate": 8.666136192681865e-05, "loss": 0.0074, "step": 22090 }, { "grad_norm": 0.14084665477275848, "learning_rate": 8.664730463868811e-05, "loss": 0.011, "step": 22100 }, { "grad_norm": 0.11600727587938309, "learning_rate": 8.663324108853552e-05, "loss": 0.0076, "step": 22110 }, { "grad_norm": 0.18282461166381836, "learning_rate": 8.661917127876395e-05, "loss": 0.0087, "step": 22120 }, { "grad_norm": 0.19242940843105316, "learning_rate": 8.660509521177754e-05, "loss": 0.0122, "step": 22130 }, { "grad_norm": 0.08925726264715195, "learning_rate": 8.65910128899815e-05, "loss": 0.0097, "step": 22140 }, { "grad_norm": 0.13114477694034576, "learning_rate": 8.657692431578214e-05, "loss": 0.0094, "step": 22150 }, { "grad_norm": 0.1851496547460556, "learning_rate": 8.656282949158679e-05, "loss": 0.0089, "step": 22160 }, { "grad_norm": 0.16258478164672852, "learning_rate": 8.654872841980388e-05, "loss": 0.0096, "step": 22170 }, { "grad_norm": 0.20221491158008575, "learning_rate": 8.653462110284289e-05, "loss": 0.0097, "step": 22180 }, { "grad_norm": 0.17527253925800323, "learning_rate": 8.652050754311437e-05, "loss": 0.0097, "step": 22190 }, { "grad_norm": 0.18901871144771576, "learning_rate": 8.650638774302995e-05, "loss": 0.0092, "step": 22200 }, { "grad_norm": 0.12264750897884369, "learning_rate": 8.649226170500233e-05, "loss": 0.0093, "step": 22210 }, { "grad_norm": 0.22274623811244965, "learning_rate": 8.647812943144524e-05, "loss": 0.0106, "step": 22220 }, { "grad_norm": 0.16975559294223785, "learning_rate": 8.646399092477351e-05, "loss": 0.009, "step": 22230 }, { "grad_norm": 0.13443580269813538, "learning_rate": 8.644984618740301e-05, "loss": 0.0111, "step": 22240 }, { "grad_norm": 0.14483703672885895, "learning_rate": 8.643569522175073e-05, "loss": 0.014, "step": 22250 }, { "grad_norm": 0.18705080449581146, "learning_rate": 8.642153803023463e-05, "loss": 0.0118, "step": 22260 }, { "grad_norm": 0.12972380220890045, "learning_rate": 8.640737461527383e-05, "loss": 0.0087, "step": 22270 }, { "grad_norm": 0.18110236525535583, "learning_rate": 8.639320497928845e-05, "loss": 0.0094, "step": 22280 }, { "grad_norm": 0.14996415376663208, "learning_rate": 8.637902912469969e-05, "loss": 0.0094, "step": 22290 }, { "grad_norm": 0.13459724187850952, "learning_rate": 8.636484705392982e-05, "loss": 0.0075, "step": 22300 }, { "grad_norm": 0.16017264127731323, "learning_rate": 8.635065876940216e-05, "loss": 0.0084, "step": 22310 }, { "grad_norm": 0.11974802613258362, "learning_rate": 8.633646427354112e-05, "loss": 0.0074, "step": 22320 }, { "grad_norm": 0.15691709518432617, "learning_rate": 8.632226356877213e-05, "loss": 0.0099, "step": 22330 }, { "grad_norm": 0.15385952591896057, "learning_rate": 8.630805665752173e-05, "loss": 0.0089, "step": 22340 }, { "grad_norm": 0.226006880402565, "learning_rate": 8.629384354221748e-05, "loss": 0.0081, "step": 22350 }, { "grad_norm": 0.14815270900726318, "learning_rate": 8.627962422528797e-05, "loss": 0.0079, "step": 22360 }, { "grad_norm": 0.11892828345298767, "learning_rate": 8.626539870916296e-05, "loss": 0.0084, "step": 22370 }, { "grad_norm": 0.15354490280151367, "learning_rate": 8.625116699627317e-05, "loss": 0.0071, "step": 22380 }, { "grad_norm": 0.16464310884475708, "learning_rate": 8.623692908905041e-05, "loss": 0.0092, "step": 22390 }, { "grad_norm": 0.18106767535209656, "learning_rate": 8.622268498992755e-05, "loss": 0.0094, "step": 22400 }, { "grad_norm": 0.1306968331336975, "learning_rate": 8.620843470133851e-05, "loss": 0.0082, "step": 22410 }, { "grad_norm": 0.10282217711210251, "learning_rate": 8.619417822571829e-05, "loss": 0.0103, "step": 22420 }, { "grad_norm": 0.21157294511795044, "learning_rate": 8.617991556550292e-05, "loss": 0.01, "step": 22430 }, { "grad_norm": 0.17735153436660767, "learning_rate": 8.616564672312952e-05, "loss": 0.0097, "step": 22440 }, { "grad_norm": 0.17490364611148834, "learning_rate": 8.61513717010362e-05, "loss": 0.0083, "step": 22450 }, { "grad_norm": 0.16072827577590942, "learning_rate": 8.613709050166221e-05, "loss": 0.0077, "step": 22460 }, { "grad_norm": 0.13288040459156036, "learning_rate": 8.61228031274478e-05, "loss": 0.0084, "step": 22470 }, { "grad_norm": 0.1402299702167511, "learning_rate": 8.610850958083431e-05, "loss": 0.0073, "step": 22480 }, { "grad_norm": 0.11273656040430069, "learning_rate": 8.609420986426409e-05, "loss": 0.0069, "step": 22490 }, { "grad_norm": 0.1391109824180603, "learning_rate": 8.60799039801806e-05, "loss": 0.009, "step": 22500 }, { "grad_norm": 0.15640348196029663, "learning_rate": 8.606559193102828e-05, "loss": 0.0082, "step": 22510 }, { "grad_norm": 0.12015046179294586, "learning_rate": 8.605127371925273e-05, "loss": 0.0075, "step": 22520 }, { "grad_norm": 0.13741722702980042, "learning_rate": 8.603694934730047e-05, "loss": 0.0082, "step": 22530 }, { "grad_norm": 0.17806655168533325, "learning_rate": 8.602261881761919e-05, "loss": 0.0103, "step": 22540 }, { "grad_norm": 0.22282280027866364, "learning_rate": 8.600828213265759e-05, "loss": 0.008, "step": 22550 }, { "grad_norm": 0.14943021535873413, "learning_rate": 8.599393929486539e-05, "loss": 0.0087, "step": 22560 }, { "grad_norm": 0.14924006164073944, "learning_rate": 8.59795903066934e-05, "loss": 0.0079, "step": 22570 }, { "grad_norm": 0.08154485374689102, "learning_rate": 8.596523517059347e-05, "loss": 0.0065, "step": 22580 }, { "grad_norm": 0.1859111189842224, "learning_rate": 8.59508738890185e-05, "loss": 0.0077, "step": 22590 }, { "grad_norm": 0.14493921399116516, "learning_rate": 8.593650646442246e-05, "loss": 0.0084, "step": 22600 }, { "grad_norm": 0.16076160967350006, "learning_rate": 8.59221328992603e-05, "loss": 0.0075, "step": 22610 }, { "grad_norm": 0.15928223729133606, "learning_rate": 8.590775319598813e-05, "loss": 0.0073, "step": 22620 }, { "grad_norm": 0.1743868738412857, "learning_rate": 8.589336735706301e-05, "loss": 0.0083, "step": 22630 }, { "grad_norm": 0.16843697428703308, "learning_rate": 8.587897538494307e-05, "loss": 0.0085, "step": 22640 }, { "grad_norm": 0.16532890498638153, "learning_rate": 8.586457728208756e-05, "loss": 0.0108, "step": 22650 }, { "grad_norm": 0.1057385802268982, "learning_rate": 8.585017305095667e-05, "loss": 0.0091, "step": 22660 }, { "grad_norm": 0.167149618268013, "learning_rate": 8.583576269401173e-05, "loss": 0.0126, "step": 22670 }, { "grad_norm": 0.1496717631816864, "learning_rate": 8.582134621371504e-05, "loss": 0.0082, "step": 22680 }, { "grad_norm": 0.1435832679271698, "learning_rate": 8.580692361253e-05, "loss": 0.0091, "step": 22690 }, { "grad_norm": 0.17268945276737213, "learning_rate": 8.579249489292104e-05, "loss": 0.0084, "step": 22700 }, { "grad_norm": 0.15160487592220306, "learning_rate": 8.577806005735363e-05, "loss": 0.0104, "step": 22710 }, { "grad_norm": 0.15858876705169678, "learning_rate": 8.576361910829429e-05, "loss": 0.0085, "step": 22720 }, { "grad_norm": 0.14986282587051392, "learning_rate": 8.574917204821057e-05, "loss": 0.0082, "step": 22730 }, { "grad_norm": 0.2106049656867981, "learning_rate": 8.57347188795711e-05, "loss": 0.0097, "step": 22740 }, { "grad_norm": 0.15374639630317688, "learning_rate": 8.572025960484551e-05, "loss": 0.0092, "step": 22750 }, { "grad_norm": 0.16804300248622894, "learning_rate": 8.57057942265045e-05, "loss": 0.0082, "step": 22760 }, { "grad_norm": 0.17059926688671112, "learning_rate": 8.569132274701984e-05, "loss": 0.0093, "step": 22770 }, { "grad_norm": 0.14010784029960632, "learning_rate": 8.567684516886427e-05, "loss": 0.0097, "step": 22780 }, { "grad_norm": 0.13777612149715424, "learning_rate": 8.56623614945116e-05, "loss": 0.0083, "step": 22790 }, { "grad_norm": 0.15521343052387238, "learning_rate": 8.564787172643675e-05, "loss": 0.008, "step": 22800 }, { "grad_norm": 0.16364595293998718, "learning_rate": 8.563337586711559e-05, "loss": 0.0116, "step": 22810 }, { "grad_norm": 0.1444111466407776, "learning_rate": 8.561887391902506e-05, "loss": 0.0078, "step": 22820 }, { "grad_norm": 0.1589771807193756, "learning_rate": 8.560436588464316e-05, "loss": 0.0084, "step": 22830 }, { "grad_norm": 0.09806900471448898, "learning_rate": 8.55898517664489e-05, "loss": 0.0072, "step": 22840 }, { "grad_norm": 0.23046110570430756, "learning_rate": 8.557533156692236e-05, "loss": 0.0103, "step": 22850 }, { "grad_norm": 0.13260199129581451, "learning_rate": 8.556080528854467e-05, "loss": 0.0082, "step": 22860 }, { "grad_norm": 0.13908663392066956, "learning_rate": 8.554627293379791e-05, "loss": 0.0081, "step": 22870 }, { "grad_norm": 0.12739522755146027, "learning_rate": 8.553173450516531e-05, "loss": 0.0091, "step": 22880 }, { "grad_norm": 0.11863625794649124, "learning_rate": 8.551719000513108e-05, "loss": 0.0078, "step": 22890 }, { "grad_norm": 0.12059467285871506, "learning_rate": 8.550263943618049e-05, "loss": 0.0099, "step": 22900 }, { "grad_norm": 0.1288665533065796, "learning_rate": 8.54880828007998e-05, "loss": 0.0086, "step": 22910 }, { "grad_norm": 0.13914811611175537, "learning_rate": 8.547352010147637e-05, "loss": 0.0088, "step": 22920 }, { "grad_norm": 0.1484878659248352, "learning_rate": 8.545895134069855e-05, "loss": 0.008, "step": 22930 }, { "grad_norm": 0.13986296951770782, "learning_rate": 8.544437652095576e-05, "loss": 0.0091, "step": 22940 }, { "grad_norm": 0.15645194053649902, "learning_rate": 8.542979564473843e-05, "loss": 0.0085, "step": 22950 }, { "grad_norm": 0.15685541927814484, "learning_rate": 8.541520871453802e-05, "loss": 0.0092, "step": 22960 }, { "grad_norm": 0.20762299001216888, "learning_rate": 8.540061573284705e-05, "loss": 0.0089, "step": 22970 }, { "grad_norm": 0.132394477725029, "learning_rate": 8.538601670215906e-05, "loss": 0.0071, "step": 22980 }, { "grad_norm": 0.1739938110113144, "learning_rate": 8.537141162496864e-05, "loss": 0.0073, "step": 22990 }, { "grad_norm": 0.21212784945964813, "learning_rate": 8.535680050377137e-05, "loss": 0.0095, "step": 23000 }, { "grad_norm": 0.1512223333120346, "learning_rate": 8.534218334106391e-05, "loss": 0.0086, "step": 23010 }, { "grad_norm": 0.1430341601371765, "learning_rate": 8.532756013934393e-05, "loss": 0.0084, "step": 23020 }, { "grad_norm": 0.18717312812805176, "learning_rate": 8.531293090111012e-05, "loss": 0.007, "step": 23030 }, { "grad_norm": 0.12041860818862915, "learning_rate": 8.529829562886225e-05, "loss": 0.0087, "step": 23040 }, { "grad_norm": 0.16321322321891785, "learning_rate": 8.528365432510105e-05, "loss": 0.0083, "step": 23050 }, { "grad_norm": 0.15448802709579468, "learning_rate": 8.526900699232833e-05, "loss": 0.0086, "step": 23060 }, { "grad_norm": 0.16283129155635834, "learning_rate": 8.525435363304695e-05, "loss": 0.0085, "step": 23070 }, { "grad_norm": 0.1543436348438263, "learning_rate": 8.523969424976072e-05, "loss": 0.0097, "step": 23080 }, { "grad_norm": 0.19432125985622406, "learning_rate": 8.522502884497457e-05, "loss": 0.0088, "step": 23090 }, { "grad_norm": 0.1335378736257553, "learning_rate": 8.521035742119437e-05, "loss": 0.0086, "step": 23100 }, { "grad_norm": 0.1689736247062683, "learning_rate": 8.519567998092712e-05, "loss": 0.0081, "step": 23110 }, { "grad_norm": 0.1500156819820404, "learning_rate": 8.518099652668075e-05, "loss": 0.0065, "step": 23120 }, { "grad_norm": 0.1821824461221695, "learning_rate": 8.516630706096429e-05, "loss": 0.0071, "step": 23130 }, { "grad_norm": 0.18943239748477936, "learning_rate": 8.515161158628773e-05, "loss": 0.0074, "step": 23140 }, { "grad_norm": 0.12972944974899292, "learning_rate": 8.513691010516216e-05, "loss": 0.0092, "step": 23150 }, { "grad_norm": 0.10362493246793747, "learning_rate": 8.512220262009966e-05, "loss": 0.0083, "step": 23160 }, { "grad_norm": 0.13893628120422363, "learning_rate": 8.510748913361332e-05, "loss": 0.0077, "step": 23170 }, { "grad_norm": 0.13351117074489594, "learning_rate": 8.509276964821726e-05, "loss": 0.0079, "step": 23180 }, { "grad_norm": 0.17730984091758728, "learning_rate": 8.507804416642669e-05, "loss": 0.0112, "step": 23190 }, { "grad_norm": 0.14624427258968353, "learning_rate": 8.506331269075774e-05, "loss": 0.0121, "step": 23200 }, { "grad_norm": 0.14109055697917938, "learning_rate": 8.504857522372765e-05, "loss": 0.0074, "step": 23210 }, { "grad_norm": 0.13710932433605194, "learning_rate": 8.503383176785461e-05, "loss": 0.0071, "step": 23220 }, { "grad_norm": 0.16871728003025055, "learning_rate": 8.501908232565792e-05, "loss": 0.0101, "step": 23230 }, { "grad_norm": 0.1414748728275299, "learning_rate": 8.50043268996578e-05, "loss": 0.0091, "step": 23240 }, { "grad_norm": 0.12742440402507782, "learning_rate": 8.498956549237562e-05, "loss": 0.0068, "step": 23250 }, { "grad_norm": 0.11350473016500473, "learning_rate": 8.497479810633366e-05, "loss": 0.0069, "step": 23260 }, { "grad_norm": 0.12440665811300278, "learning_rate": 8.496002474405525e-05, "loss": 0.0078, "step": 23270 }, { "grad_norm": 0.1754416972398758, "learning_rate": 8.494524540806478e-05, "loss": 0.0095, "step": 23280 }, { "grad_norm": 0.1317754089832306, "learning_rate": 8.493046010088761e-05, "loss": 0.0086, "step": 23290 }, { "grad_norm": 0.13027413189411163, "learning_rate": 8.491566882505018e-05, "loss": 0.0094, "step": 23300 }, { "grad_norm": 0.16632133722305298, "learning_rate": 8.490087158307988e-05, "loss": 0.0091, "step": 23310 }, { "grad_norm": 0.16330435872077942, "learning_rate": 8.488606837750518e-05, "loss": 0.0097, "step": 23320 }, { "grad_norm": 0.13796716928482056, "learning_rate": 8.487125921085552e-05, "loss": 0.0081, "step": 23330 }, { "grad_norm": 0.15137845277786255, "learning_rate": 8.485644408566141e-05, "loss": 0.0067, "step": 23340 }, { "grad_norm": 0.18310536444187164, "learning_rate": 8.484162300445431e-05, "loss": 0.0083, "step": 23350 }, { "grad_norm": 0.15118390321731567, "learning_rate": 8.482679596976676e-05, "loss": 0.0121, "step": 23360 }, { "grad_norm": 0.1167507991194725, "learning_rate": 8.48119629841323e-05, "loss": 0.0099, "step": 23370 }, { "grad_norm": 0.15658825635910034, "learning_rate": 8.479712405008547e-05, "loss": 0.011, "step": 23380 }, { "grad_norm": 0.16631199419498444, "learning_rate": 8.478227917016184e-05, "loss": 0.0097, "step": 23390 }, { "grad_norm": 0.14414489269256592, "learning_rate": 8.476742834689801e-05, "loss": 0.0089, "step": 23400 }, { "grad_norm": 0.12373698502779007, "learning_rate": 8.475257158283157e-05, "loss": 0.0083, "step": 23410 }, { "grad_norm": 0.19017763435840607, "learning_rate": 8.473770888050112e-05, "loss": 0.0123, "step": 23420 }, { "grad_norm": 0.16472359001636505, "learning_rate": 8.47228402424463e-05, "loss": 0.0123, "step": 23430 }, { "grad_norm": 0.1476290225982666, "learning_rate": 8.470796567120775e-05, "loss": 0.009, "step": 23440 }, { "grad_norm": 0.1913626790046692, "learning_rate": 8.469308516932714e-05, "loss": 0.0104, "step": 23450 }, { "grad_norm": 0.13335461914539337, "learning_rate": 8.467819873934714e-05, "loss": 0.0098, "step": 23460 }, { "grad_norm": 0.17535535991191864, "learning_rate": 8.466330638381143e-05, "loss": 0.0096, "step": 23470 }, { "grad_norm": 0.13990291953086853, "learning_rate": 8.464840810526469e-05, "loss": 0.0082, "step": 23480 }, { "grad_norm": 0.11879418045282364, "learning_rate": 8.463350390625264e-05, "loss": 0.0062, "step": 23490 }, { "grad_norm": 0.18618546426296234, "learning_rate": 8.4618593789322e-05, "loss": 0.009, "step": 23500 }, { "grad_norm": 0.13717731833457947, "learning_rate": 8.46036777570205e-05, "loss": 0.0093, "step": 23510 }, { "grad_norm": 0.15293630957603455, "learning_rate": 8.458875581189688e-05, "loss": 0.0097, "step": 23520 }, { "grad_norm": 0.16488416492938995, "learning_rate": 8.457382795650092e-05, "loss": 0.0094, "step": 23530 }, { "grad_norm": 0.15533795952796936, "learning_rate": 8.455889419338335e-05, "loss": 0.0095, "step": 23540 }, { "grad_norm": 0.14597952365875244, "learning_rate": 8.454395452509593e-05, "loss": 0.0073, "step": 23550 }, { "grad_norm": 0.1507149338722229, "learning_rate": 8.452900895419146e-05, "loss": 0.0066, "step": 23560 }, { "grad_norm": 0.14820809662342072, "learning_rate": 8.451405748322376e-05, "loss": 0.0072, "step": 23570 }, { "grad_norm": 0.14186173677444458, "learning_rate": 8.449910011474759e-05, "loss": 0.011, "step": 23580 }, { "grad_norm": 0.15995678305625916, "learning_rate": 8.448413685131876e-05, "loss": 0.008, "step": 23590 }, { "grad_norm": 0.16827309131622314, "learning_rate": 8.446916769549407e-05, "loss": 0.0079, "step": 23600 }, { "grad_norm": 0.15678614377975464, "learning_rate": 8.445419264983136e-05, "loss": 0.0076, "step": 23610 }, { "grad_norm": 0.13983500003814697, "learning_rate": 8.443921171688947e-05, "loss": 0.007, "step": 23620 }, { "grad_norm": 0.1620873063802719, "learning_rate": 8.442422489922819e-05, "loss": 0.0076, "step": 23630 }, { "grad_norm": 0.18241889774799347, "learning_rate": 8.440923219940838e-05, "loss": 0.0074, "step": 23640 }, { "grad_norm": 0.15692946314811707, "learning_rate": 8.439423361999189e-05, "loss": 0.0074, "step": 23650 }, { "grad_norm": 0.18639500439167023, "learning_rate": 8.437922916354155e-05, "loss": 0.0077, "step": 23660 }, { "grad_norm": 0.21011175215244293, "learning_rate": 8.436421883262123e-05, "loss": 0.0094, "step": 23670 }, { "grad_norm": 0.17461271584033966, "learning_rate": 8.434920262979577e-05, "loss": 0.0074, "step": 23680 }, { "grad_norm": 0.1670936942100525, "learning_rate": 8.433418055763104e-05, "loss": 0.0079, "step": 23690 }, { "grad_norm": 0.1346222162246704, "learning_rate": 8.431915261869389e-05, "loss": 0.0079, "step": 23700 }, { "grad_norm": 0.17380619049072266, "learning_rate": 8.43041188155522e-05, "loss": 0.007, "step": 23710 }, { "grad_norm": 0.23704060912132263, "learning_rate": 8.428907915077481e-05, "loss": 0.0094, "step": 23720 }, { "grad_norm": 0.18594345450401306, "learning_rate": 8.42740336269316e-05, "loss": 0.0085, "step": 23730 }, { "grad_norm": 0.12806004285812378, "learning_rate": 8.425898224659345e-05, "loss": 0.0097, "step": 23740 }, { "grad_norm": 0.12307944148778915, "learning_rate": 8.42439250123322e-05, "loss": 0.0074, "step": 23750 }, { "grad_norm": 0.15474045276641846, "learning_rate": 8.422886192672076e-05, "loss": 0.0097, "step": 23760 }, { "grad_norm": 0.15516091883182526, "learning_rate": 8.421379299233297e-05, "loss": 0.0084, "step": 23770 }, { "grad_norm": 0.11265743523836136, "learning_rate": 8.419871821174371e-05, "loss": 0.0085, "step": 23780 }, { "grad_norm": 0.13571655750274658, "learning_rate": 8.418363758752884e-05, "loss": 0.0078, "step": 23790 }, { "grad_norm": 0.16037844121456146, "learning_rate": 8.416855112226523e-05, "loss": 0.0092, "step": 23800 }, { "grad_norm": 0.1660616248846054, "learning_rate": 8.415345881853075e-05, "loss": 0.0098, "step": 23810 }, { "grad_norm": 0.14531636238098145, "learning_rate": 8.413836067890426e-05, "loss": 0.0092, "step": 23820 }, { "grad_norm": 0.2688494622707367, "learning_rate": 8.41232567059656e-05, "loss": 0.0075, "step": 23830 }, { "grad_norm": 0.17400597035884857, "learning_rate": 8.410814690229565e-05, "loss": 0.01, "step": 23840 }, { "grad_norm": 0.14549218118190765, "learning_rate": 8.409303127047626e-05, "loss": 0.0108, "step": 23850 }, { "grad_norm": 0.1606181561946869, "learning_rate": 8.407790981309028e-05, "loss": 0.0091, "step": 23860 }, { "grad_norm": 0.17838574945926666, "learning_rate": 8.406278253272153e-05, "loss": 0.01, "step": 23870 }, { "grad_norm": 0.12436874955892563, "learning_rate": 8.404764943195487e-05, "loss": 0.0094, "step": 23880 }, { "grad_norm": 0.1352435201406479, "learning_rate": 8.403251051337613e-05, "loss": 0.0095, "step": 23890 }, { "grad_norm": 0.21535852551460266, "learning_rate": 8.401736577957214e-05, "loss": 0.0097, "step": 23900 }, { "grad_norm": 0.1911577433347702, "learning_rate": 8.40022152331307e-05, "loss": 0.0077, "step": 23910 }, { "grad_norm": 0.12699946761131287, "learning_rate": 8.398705887664064e-05, "loss": 0.0078, "step": 23920 }, { "grad_norm": 0.13794130086898804, "learning_rate": 8.397189671269177e-05, "loss": 0.0075, "step": 23930 }, { "grad_norm": 0.13688543438911438, "learning_rate": 8.395672874387488e-05, "loss": 0.0091, "step": 23940 }, { "grad_norm": 0.23540665209293365, "learning_rate": 8.394155497278177e-05, "loss": 0.0087, "step": 23950 }, { "grad_norm": 0.15092653036117554, "learning_rate": 8.392637540200523e-05, "loss": 0.0091, "step": 23960 }, { "grad_norm": 0.28820982575416565, "learning_rate": 8.391119003413902e-05, "loss": 0.0077, "step": 23970 }, { "grad_norm": 0.18640288710594177, "learning_rate": 8.38959988717779e-05, "loss": 0.0073, "step": 23980 }, { "grad_norm": 0.12529787421226501, "learning_rate": 8.388080191751764e-05, "loss": 0.0094, "step": 23990 }, { "grad_norm": 0.15340550243854523, "learning_rate": 8.386559917395496e-05, "loss": 0.0095, "step": 24000 }, { "grad_norm": 0.13838963210582733, "learning_rate": 8.385039064368761e-05, "loss": 0.0081, "step": 24010 }, { "grad_norm": 0.21984492242336273, "learning_rate": 8.383517632931431e-05, "loss": 0.0102, "step": 24020 }, { "grad_norm": 0.18203827738761902, "learning_rate": 8.381995623343477e-05, "loss": 0.0082, "step": 24030 }, { "grad_norm": 0.18765941262245178, "learning_rate": 8.380473035864968e-05, "loss": 0.0098, "step": 24040 }, { "grad_norm": 0.12374638020992279, "learning_rate": 8.378949870756076e-05, "loss": 0.0085, "step": 24050 }, { "grad_norm": 0.1279497742652893, "learning_rate": 8.377426128277063e-05, "loss": 0.0101, "step": 24060 }, { "grad_norm": 0.10530442744493484, "learning_rate": 8.375901808688298e-05, "loss": 0.0075, "step": 24070 }, { "grad_norm": 0.12596461176872253, "learning_rate": 8.374376912250246e-05, "loss": 0.0065, "step": 24080 }, { "grad_norm": 0.23836475610733032, "learning_rate": 8.372851439223468e-05, "loss": 0.0105, "step": 24090 }, { "grad_norm": 0.16613858938217163, "learning_rate": 8.371325389868627e-05, "loss": 0.009, "step": 24100 }, { "grad_norm": 0.15515220165252686, "learning_rate": 8.369798764446482e-05, "loss": 0.0087, "step": 24110 }, { "grad_norm": 0.17397856712341309, "learning_rate": 8.368271563217893e-05, "loss": 0.0089, "step": 24120 }, { "grad_norm": 0.15204942226409912, "learning_rate": 8.366743786443817e-05, "loss": 0.008, "step": 24130 }, { "grad_norm": 0.16364417970180511, "learning_rate": 8.365215434385309e-05, "loss": 0.0077, "step": 24140 }, { "grad_norm": 0.09290231764316559, "learning_rate": 8.36368650730352e-05, "loss": 0.0072, "step": 24150 }, { "grad_norm": 0.11042582243680954, "learning_rate": 8.362157005459705e-05, "loss": 0.0076, "step": 24160 }, { "grad_norm": 0.15467050671577454, "learning_rate": 8.360626929115213e-05, "loss": 0.0061, "step": 24170 }, { "grad_norm": 0.1737026423215866, "learning_rate": 8.359096278531492e-05, "loss": 0.0102, "step": 24180 }, { "grad_norm": 0.16571667790412903, "learning_rate": 8.357565053970088e-05, "loss": 0.0111, "step": 24190 }, { "grad_norm": 0.1550956517457962, "learning_rate": 8.356033255692647e-05, "loss": 0.0086, "step": 24200 }, { "grad_norm": 0.16491089761257172, "learning_rate": 8.354500883960911e-05, "loss": 0.0084, "step": 24210 }, { "grad_norm": 0.1552216112613678, "learning_rate": 8.352967939036717e-05, "loss": 0.0107, "step": 24220 }, { "grad_norm": 0.13795693218708038, "learning_rate": 8.35143442118201e-05, "loss": 0.0096, "step": 24230 }, { "grad_norm": 0.17104707658290863, "learning_rate": 8.349900330658819e-05, "loss": 0.009, "step": 24240 }, { "grad_norm": 0.1744030863046646, "learning_rate": 8.348365667729284e-05, "loss": 0.0085, "step": 24250 }, { "grad_norm": 0.14511331915855408, "learning_rate": 8.346830432655633e-05, "loss": 0.0073, "step": 24260 }, { "grad_norm": 0.1391277313232422, "learning_rate": 8.345294625700195e-05, "loss": 0.0067, "step": 24270 }, { "grad_norm": 0.13394324481487274, "learning_rate": 8.343758247125402e-05, "loss": 0.0066, "step": 24280 }, { "grad_norm": 0.16370795667171478, "learning_rate": 8.342221297193776e-05, "loss": 0.0097, "step": 24290 }, { "grad_norm": 0.10243801772594452, "learning_rate": 8.34068377616794e-05, "loss": 0.0092, "step": 24300 }, { "grad_norm": 0.15989556908607483, "learning_rate": 8.339145684310615e-05, "loss": 0.0089, "step": 24310 }, { "grad_norm": 0.13787512481212616, "learning_rate": 8.337607021884618e-05, "loss": 0.0075, "step": 24320 }, { "grad_norm": 0.11980132013559341, "learning_rate": 8.336067789152867e-05, "loss": 0.0074, "step": 24330 }, { "grad_norm": 0.16016213595867157, "learning_rate": 8.334527986378369e-05, "loss": 0.0079, "step": 24340 }, { "grad_norm": 0.1529388576745987, "learning_rate": 8.332987613824239e-05, "loss": 0.0079, "step": 24350 }, { "grad_norm": 0.1538882851600647, "learning_rate": 8.331446671753685e-05, "loss": 0.0088, "step": 24360 }, { "grad_norm": 0.15503114461898804, "learning_rate": 8.329905160430007e-05, "loss": 0.0096, "step": 24370 }, { "grad_norm": 0.22656786441802979, "learning_rate": 8.328363080116611e-05, "loss": 0.0083, "step": 24380 }, { "grad_norm": 0.13057094812393188, "learning_rate": 8.326820431076997e-05, "loss": 0.0082, "step": 24390 }, { "grad_norm": 0.14240135252475739, "learning_rate": 8.325277213574759e-05, "loss": 0.0075, "step": 24400 }, { "grad_norm": 0.143691286444664, "learning_rate": 8.32373342787359e-05, "loss": 0.0082, "step": 24410 }, { "grad_norm": 0.17073595523834229, "learning_rate": 8.322189074237285e-05, "loss": 0.0076, "step": 24420 }, { "grad_norm": 0.13808906078338623, "learning_rate": 8.32064415292973e-05, "loss": 0.0091, "step": 24430 }, { "grad_norm": 0.17309069633483887, "learning_rate": 8.319098664214907e-05, "loss": 0.0102, "step": 24440 }, { "grad_norm": 0.14596959948539734, "learning_rate": 8.3175526083569e-05, "loss": 0.0082, "step": 24450 }, { "grad_norm": 0.1214904710650444, "learning_rate": 8.316005985619889e-05, "loss": 0.007, "step": 24460 }, { "grad_norm": 0.15149767696857452, "learning_rate": 8.314458796268147e-05, "loss": 0.0074, "step": 24470 }, { "grad_norm": 0.15132389962673187, "learning_rate": 8.312911040566047e-05, "loss": 0.0082, "step": 24480 }, { "grad_norm": 0.2267143726348877, "learning_rate": 8.31136271877806e-05, "loss": 0.0103, "step": 24490 }, { "grad_norm": 0.1770864576101303, "learning_rate": 8.309813831168748e-05, "loss": 0.0103, "step": 24500 }, { "grad_norm": 0.12258036434650421, "learning_rate": 8.308264378002777e-05, "loss": 0.0091, "step": 24510 }, { "grad_norm": 0.1472238004207611, "learning_rate": 8.306714359544906e-05, "loss": 0.01, "step": 24520 }, { "grad_norm": 0.15215131640434265, "learning_rate": 8.30516377605999e-05, "loss": 0.009, "step": 24530 }, { "grad_norm": 0.1352398842573166, "learning_rate": 8.30361262781298e-05, "loss": 0.0108, "step": 24540 }, { "grad_norm": 0.1451585590839386, "learning_rate": 8.302060915068924e-05, "loss": 0.0105, "step": 24550 }, { "grad_norm": 0.2058936059474945, "learning_rate": 8.300508638092972e-05, "loss": 0.0095, "step": 24560 }, { "grad_norm": 0.1706135869026184, "learning_rate": 8.298955797150361e-05, "loss": 0.0107, "step": 24570 }, { "grad_norm": 0.1547766476869583, "learning_rate": 8.297402392506433e-05, "loss": 0.0078, "step": 24580 }, { "grad_norm": 0.133339986205101, "learning_rate": 8.295848424426617e-05, "loss": 0.0071, "step": 24590 }, { "grad_norm": 0.1682024449110031, "learning_rate": 8.29429389317645e-05, "loss": 0.01, "step": 24600 }, { "grad_norm": 0.17326992750167847, "learning_rate": 8.292738799021556e-05, "loss": 0.0088, "step": 24610 }, { "grad_norm": 0.14885488152503967, "learning_rate": 8.291183142227656e-05, "loss": 0.0096, "step": 24620 }, { "grad_norm": 0.12649750709533691, "learning_rate": 8.289626923060572e-05, "loss": 0.0082, "step": 24630 }, { "grad_norm": 0.16284127533435822, "learning_rate": 8.288070141786218e-05, "loss": 0.0075, "step": 24640 }, { "grad_norm": 0.1360795646905899, "learning_rate": 8.286512798670605e-05, "loss": 0.0071, "step": 24650 }, { "grad_norm": 0.1557023972272873, "learning_rate": 8.284954893979842e-05, "loss": 0.0092, "step": 24660 }, { "grad_norm": 0.13187538087368011, "learning_rate": 8.283396427980131e-05, "loss": 0.0086, "step": 24670 }, { "grad_norm": 0.1551636904478073, "learning_rate": 8.281837400937771e-05, "loss": 0.0098, "step": 24680 }, { "grad_norm": 0.14921322464942932, "learning_rate": 8.28027781311916e-05, "loss": 0.01, "step": 24690 }, { "grad_norm": 0.12917600572109222, "learning_rate": 8.278717664790785e-05, "loss": 0.0089, "step": 24700 }, { "grad_norm": 0.15602993965148926, "learning_rate": 8.277156956219234e-05, "loss": 0.007, "step": 24710 }, { "grad_norm": 0.1422022432088852, "learning_rate": 8.275595687671189e-05, "loss": 0.0098, "step": 24720 }, { "grad_norm": 0.1220056563615799, "learning_rate": 8.27403385941343e-05, "loss": 0.009, "step": 24730 }, { "grad_norm": 0.12367428839206696, "learning_rate": 8.272471471712828e-05, "loss": 0.008, "step": 24740 }, { "grad_norm": 0.13554024696350098, "learning_rate": 8.270908524836355e-05, "loss": 0.008, "step": 24750 }, { "grad_norm": 0.198228657245636, "learning_rate": 8.269345019051074e-05, "loss": 0.0112, "step": 24760 }, { "grad_norm": 0.10728701204061508, "learning_rate": 8.267780954624147e-05, "loss": 0.0075, "step": 24770 }, { "grad_norm": 0.13621795177459717, "learning_rate": 8.266216331822827e-05, "loss": 0.0088, "step": 24780 }, { "grad_norm": 0.16357938945293427, "learning_rate": 8.264651150914469e-05, "loss": 0.0096, "step": 24790 }, { "grad_norm": 0.1435835212469101, "learning_rate": 8.263085412166517e-05, "loss": 0.0078, "step": 24800 }, { "grad_norm": 0.17762449383735657, "learning_rate": 8.261519115846514e-05, "loss": 0.0105, "step": 24810 }, { "grad_norm": 0.16444063186645508, "learning_rate": 8.259952262222096e-05, "loss": 0.0074, "step": 24820 }, { "grad_norm": 0.1378450095653534, "learning_rate": 8.258384851560997e-05, "loss": 0.0074, "step": 24830 }, { "grad_norm": 0.12184792757034302, "learning_rate": 8.256816884131044e-05, "loss": 0.0065, "step": 24840 }, { "grad_norm": 0.1623823642730713, "learning_rate": 8.255248360200159e-05, "loss": 0.0079, "step": 24850 }, { "grad_norm": 0.08887766301631927, "learning_rate": 8.253679280036359e-05, "loss": 0.0055, "step": 24860 }, { "grad_norm": 0.18127526342868805, "learning_rate": 8.252109643907762e-05, "loss": 0.0072, "step": 24870 }, { "grad_norm": 0.1461808681488037, "learning_rate": 8.250539452082569e-05, "loss": 0.0073, "step": 24880 }, { "grad_norm": 0.14290866255760193, "learning_rate": 8.248968704829087e-05, "loss": 0.0061, "step": 24890 }, { "grad_norm": 0.15001080930233002, "learning_rate": 8.247397402415714e-05, "loss": 0.0071, "step": 24900 }, { "grad_norm": 0.16121523082256317, "learning_rate": 8.24582554511094e-05, "loss": 0.0069, "step": 24910 }, { "grad_norm": 0.11830412596464157, "learning_rate": 8.244253133183355e-05, "loss": 0.0109, "step": 24920 }, { "grad_norm": 0.17605248093605042, "learning_rate": 8.24268016690164e-05, "loss": 0.0083, "step": 24930 }, { "grad_norm": 0.13778233528137207, "learning_rate": 8.241106646534571e-05, "loss": 0.0075, "step": 24940 }, { "grad_norm": 0.11714702099561691, "learning_rate": 8.23953257235102e-05, "loss": 0.0073, "step": 24950 }, { "grad_norm": 0.15075574815273285, "learning_rate": 8.237957944619956e-05, "loss": 0.0084, "step": 24960 }, { "grad_norm": 0.1343645304441452, "learning_rate": 8.236382763610437e-05, "loss": 0.0067, "step": 24970 }, { "grad_norm": 0.17555934190750122, "learning_rate": 8.234807029591619e-05, "loss": 0.0104, "step": 24980 }, { "grad_norm": 0.13595199584960938, "learning_rate": 8.233230742832752e-05, "loss": 0.0086, "step": 24990 }, { "grad_norm": 0.15717925131320953, "learning_rate": 8.231653903603178e-05, "loss": 0.0073, "step": 25000 }, { "grad_norm": 0.1276298612356186, "learning_rate": 8.23007651217234e-05, "loss": 0.0069, "step": 25010 }, { "grad_norm": 0.14725914597511292, "learning_rate": 8.228498568809769e-05, "loss": 0.0078, "step": 25020 }, { "grad_norm": 0.13391730189323425, "learning_rate": 8.22692007378509e-05, "loss": 0.0088, "step": 25030 }, { "grad_norm": 0.1290981024503708, "learning_rate": 8.225341027368028e-05, "loss": 0.0067, "step": 25040 }, { "grad_norm": 0.12286317348480225, "learning_rate": 8.223761429828399e-05, "loss": 0.0083, "step": 25050 }, { "grad_norm": 0.11088910698890686, "learning_rate": 8.22218128143611e-05, "loss": 0.0079, "step": 25060 }, { "grad_norm": 0.15194855630397797, "learning_rate": 8.220600582461166e-05, "loss": 0.0101, "step": 25070 }, { "grad_norm": 0.11611106246709824, "learning_rate": 8.219019333173668e-05, "loss": 0.0072, "step": 25080 }, { "grad_norm": 0.15472717583179474, "learning_rate": 8.217437533843805e-05, "loss": 0.0083, "step": 25090 }, { "grad_norm": 0.13016290962696075, "learning_rate": 8.215855184741867e-05, "loss": 0.0095, "step": 25100 }, { "grad_norm": 0.12601931393146515, "learning_rate": 8.21427228613823e-05, "loss": 0.0078, "step": 25110 }, { "grad_norm": 0.11412260681390762, "learning_rate": 8.21268883830337e-05, "loss": 0.0091, "step": 25120 }, { "grad_norm": 0.11226717382669449, "learning_rate": 8.211104841507855e-05, "loss": 0.0083, "step": 25130 }, { "grad_norm": 0.21156975626945496, "learning_rate": 8.209520296022346e-05, "loss": 0.0091, "step": 25140 }, { "grad_norm": 0.16010960936546326, "learning_rate": 8.207935202117599e-05, "loss": 0.0091, "step": 25150 }, { "grad_norm": 0.1223597526550293, "learning_rate": 8.206349560064463e-05, "loss": 0.0093, "step": 25160 }, { "grad_norm": 0.1373172551393509, "learning_rate": 8.204763370133881e-05, "loss": 0.009, "step": 25170 }, { "grad_norm": 0.13352465629577637, "learning_rate": 8.203176632596892e-05, "loss": 0.0085, "step": 25180 }, { "grad_norm": 0.1445322483778, "learning_rate": 8.20158934772462e-05, "loss": 0.0082, "step": 25190 }, { "grad_norm": 0.12136314809322357, "learning_rate": 8.200001515788294e-05, "loss": 0.0062, "step": 25200 }, { "grad_norm": 0.1276751309633255, "learning_rate": 8.198413137059228e-05, "loss": 0.0069, "step": 25210 }, { "grad_norm": 0.16148194670677185, "learning_rate": 8.196824211808835e-05, "loss": 0.011, "step": 25220 }, { "grad_norm": 0.17919762432575226, "learning_rate": 8.195234740308617e-05, "loss": 0.0106, "step": 25230 }, { "grad_norm": 0.16907702386379242, "learning_rate": 8.193644722830171e-05, "loss": 0.0089, "step": 25240 }, { "grad_norm": 0.11739367991685867, "learning_rate": 8.19205415964519e-05, "loss": 0.0069, "step": 25250 }, { "grad_norm": 0.12316613644361496, "learning_rate": 8.190463051025456e-05, "loss": 0.0071, "step": 25260 }, { "grad_norm": 0.11387525498867035, "learning_rate": 8.188871397242843e-05, "loss": 0.0128, "step": 25270 }, { "grad_norm": 0.1927558332681656, "learning_rate": 8.187279198569326e-05, "loss": 0.0128, "step": 25280 }, { "grad_norm": 0.12786956131458282, "learning_rate": 8.185686455276966e-05, "loss": 0.01, "step": 25290 }, { "grad_norm": 0.10539710521697998, "learning_rate": 8.184093167637921e-05, "loss": 0.01, "step": 25300 }, { "grad_norm": 0.16535775363445282, "learning_rate": 8.182499335924437e-05, "loss": 0.0104, "step": 25310 }, { "grad_norm": 0.1421077847480774, "learning_rate": 8.18090496040886e-05, "loss": 0.0093, "step": 25320 }, { "grad_norm": 0.15622609853744507, "learning_rate": 8.179310041363621e-05, "loss": 0.0074, "step": 25330 }, { "grad_norm": 0.12270469218492508, "learning_rate": 8.17771457906125e-05, "loss": 0.0079, "step": 25340 }, { "grad_norm": 0.10320565104484558, "learning_rate": 8.176118573774371e-05, "loss": 0.0071, "step": 25350 }, { "grad_norm": 0.11709310859441757, "learning_rate": 8.174522025775692e-05, "loss": 0.0079, "step": 25360 }, { "grad_norm": 0.18644964694976807, "learning_rate": 8.172924935338022e-05, "loss": 0.0084, "step": 25370 }, { "grad_norm": 0.16542643308639526, "learning_rate": 8.171327302734262e-05, "loss": 0.0086, "step": 25380 }, { "grad_norm": 0.11702369898557663, "learning_rate": 8.169729128237401e-05, "loss": 0.0085, "step": 25390 }, { "grad_norm": 0.11475301533937454, "learning_rate": 8.168130412120525e-05, "loss": 0.0082, "step": 25400 }, { "grad_norm": 0.1502031832933426, "learning_rate": 8.16653115465681e-05, "loss": 0.011, "step": 25410 }, { "grad_norm": 0.12205925583839417, "learning_rate": 8.164931356119526e-05, "loss": 0.0098, "step": 25420 }, { "grad_norm": 0.1389015018939972, "learning_rate": 8.163331016782032e-05, "loss": 0.0079, "step": 25430 }, { "grad_norm": 0.11726157367229462, "learning_rate": 8.161730136917785e-05, "loss": 0.0089, "step": 25440 }, { "grad_norm": 0.1295177936553955, "learning_rate": 8.160128716800333e-05, "loss": 0.0062, "step": 25450 }, { "grad_norm": 0.18017259240150452, "learning_rate": 8.158526756703313e-05, "loss": 0.0094, "step": 25460 }, { "grad_norm": 0.11900224536657333, "learning_rate": 8.156924256900455e-05, "loss": 0.0101, "step": 25470 }, { "grad_norm": 0.160861536860466, "learning_rate": 8.155321217665584e-05, "loss": 0.0109, "step": 25480 }, { "grad_norm": 0.12305209040641785, "learning_rate": 8.153717639272614e-05, "loss": 0.0082, "step": 25490 }, { "grad_norm": 0.1947120577096939, "learning_rate": 8.152113521995555e-05, "loss": 0.0081, "step": 25500 }, { "grad_norm": 0.13120925426483154, "learning_rate": 8.150508866108505e-05, "loss": 0.0077, "step": 25510 }, { "grad_norm": 0.20151501893997192, "learning_rate": 8.148903671885657e-05, "loss": 0.0104, "step": 25520 }, { "grad_norm": 0.17026302218437195, "learning_rate": 8.147297939601292e-05, "loss": 0.009, "step": 25530 }, { "grad_norm": 0.2225688248872757, "learning_rate": 8.145691669529792e-05, "loss": 0.0099, "step": 25540 }, { "grad_norm": 0.161866694688797, "learning_rate": 8.144084861945618e-05, "loss": 0.0062, "step": 25550 }, { "grad_norm": 0.12391998618841171, "learning_rate": 8.142477517123333e-05, "loss": 0.0079, "step": 25560 }, { "grad_norm": 0.12286917865276337, "learning_rate": 8.140869635337586e-05, "loss": 0.0095, "step": 25570 }, { "grad_norm": 0.11980312317609787, "learning_rate": 8.139261216863123e-05, "loss": 0.0091, "step": 25580 }, { "grad_norm": 0.15748754143714905, "learning_rate": 8.137652261974776e-05, "loss": 0.0098, "step": 25590 }, { "grad_norm": 0.14778317511081696, "learning_rate": 8.136042770947472e-05, "loss": 0.0091, "step": 25600 }, { "grad_norm": 0.10684217512607574, "learning_rate": 8.134432744056228e-05, "loss": 0.0075, "step": 25610 }, { "grad_norm": 0.16511757671833038, "learning_rate": 8.132822181576158e-05, "loss": 0.0088, "step": 25620 }, { "grad_norm": 0.1482154130935669, "learning_rate": 8.131211083782459e-05, "loss": 0.008, "step": 25630 }, { "grad_norm": 0.21314089000225067, "learning_rate": 8.129599450950424e-05, "loss": 0.0078, "step": 25640 }, { "grad_norm": 0.13895824551582336, "learning_rate": 8.127987283355438e-05, "loss": 0.0075, "step": 25650 }, { "grad_norm": 0.1784747987985611, "learning_rate": 8.126374581272976e-05, "loss": 0.0082, "step": 25660 }, { "grad_norm": 0.16350367665290833, "learning_rate": 8.124761344978605e-05, "loss": 0.0115, "step": 25670 }, { "grad_norm": 0.145985946059227, "learning_rate": 8.12314757474798e-05, "loss": 0.01, "step": 25680 }, { "grad_norm": 0.13125447928905487, "learning_rate": 8.121533270856856e-05, "loss": 0.0063, "step": 25690 }, { "grad_norm": 0.18944494426250458, "learning_rate": 8.119918433581069e-05, "loss": 0.0088, "step": 25700 }, { "grad_norm": 0.13635630905628204, "learning_rate": 8.118303063196551e-05, "loss": 0.0095, "step": 25710 }, { "grad_norm": 0.08073864132165909, "learning_rate": 8.116687159979326e-05, "loss": 0.0065, "step": 25720 }, { "grad_norm": 0.13822415471076965, "learning_rate": 8.115070724205508e-05, "loss": 0.0078, "step": 25730 }, { "grad_norm": 0.13755376636981964, "learning_rate": 8.113453756151296e-05, "loss": 0.009, "step": 25740 }, { "grad_norm": 0.1428034007549286, "learning_rate": 8.111836256092995e-05, "loss": 0.008, "step": 25750 }, { "grad_norm": 0.16128848493099213, "learning_rate": 8.110218224306985e-05, "loss": 0.0085, "step": 25760 }, { "grad_norm": 0.14249487221240997, "learning_rate": 8.108599661069745e-05, "loss": 0.007, "step": 25770 }, { "grad_norm": 0.15794438123703003, "learning_rate": 8.106980566657845e-05, "loss": 0.0079, "step": 25780 }, { "grad_norm": 0.15273427963256836, "learning_rate": 8.10536094134794e-05, "loss": 0.0081, "step": 25790 }, { "grad_norm": 0.13611523807048798, "learning_rate": 8.103740785416783e-05, "loss": 0.0066, "step": 25800 }, { "grad_norm": 0.12480878084897995, "learning_rate": 8.102120099141212e-05, "loss": 0.0076, "step": 25810 }, { "grad_norm": 0.1343601793050766, "learning_rate": 8.100498882798163e-05, "loss": 0.0074, "step": 25820 }, { "grad_norm": 0.11820865422487259, "learning_rate": 8.09887713666465e-05, "loss": 0.0092, "step": 25830 }, { "grad_norm": 0.10615349560976028, "learning_rate": 8.09725486101779e-05, "loss": 0.008, "step": 25840 }, { "grad_norm": 0.09999980032444, "learning_rate": 8.095632056134784e-05, "loss": 0.0077, "step": 25850 }, { "grad_norm": 0.1750769019126892, "learning_rate": 8.094008722292925e-05, "loss": 0.0106, "step": 25860 }, { "grad_norm": 0.12956339120864868, "learning_rate": 8.092384859769598e-05, "loss": 0.0063, "step": 25870 }, { "grad_norm": 0.15834274888038635, "learning_rate": 8.090760468842275e-05, "loss": 0.0111, "step": 25880 }, { "grad_norm": 0.17200081050395966, "learning_rate": 8.089135549788521e-05, "loss": 0.0088, "step": 25890 }, { "grad_norm": 0.13935042917728424, "learning_rate": 8.087510102885987e-05, "loss": 0.0073, "step": 25900 }, { "grad_norm": 0.22116273641586304, "learning_rate": 8.085884128412422e-05, "loss": 0.0093, "step": 25910 }, { "grad_norm": 0.19687387347221375, "learning_rate": 8.084257626645659e-05, "loss": 0.0092, "step": 25920 }, { "grad_norm": 0.11191366612911224, "learning_rate": 8.08263059786362e-05, "loss": 0.0066, "step": 25930 }, { "grad_norm": 0.11428549140691757, "learning_rate": 8.081003042344325e-05, "loss": 0.0064, "step": 25940 }, { "grad_norm": 0.12460986524820328, "learning_rate": 8.079374960365872e-05, "loss": 0.0073, "step": 25950 }, { "grad_norm": 0.11985453218221664, "learning_rate": 8.077746352206463e-05, "loss": 0.0073, "step": 25960 }, { "grad_norm": 0.12319556623697281, "learning_rate": 8.076117218144377e-05, "loss": 0.0059, "step": 25970 }, { "grad_norm": 0.08589164167642593, "learning_rate": 8.074487558457991e-05, "loss": 0.0067, "step": 25980 }, { "grad_norm": 0.13777661323547363, "learning_rate": 8.072857373425768e-05, "loss": 0.0064, "step": 25990 }, { "grad_norm": 0.13290292024612427, "learning_rate": 8.071226663326264e-05, "loss": 0.0061, "step": 26000 }, { "grad_norm": 0.13098029792308807, "learning_rate": 8.069595428438121e-05, "loss": 0.0102, "step": 26010 }, { "grad_norm": 0.14743109047412872, "learning_rate": 8.067963669040072e-05, "loss": 0.0068, "step": 26020 }, { "grad_norm": 0.15216077864170074, "learning_rate": 8.066331385410942e-05, "loss": 0.0083, "step": 26030 }, { "grad_norm": 0.1442815214395523, "learning_rate": 8.064698577829641e-05, "loss": 0.0068, "step": 26040 }, { "grad_norm": 0.1347932666540146, "learning_rate": 8.063065246575175e-05, "loss": 0.0065, "step": 26050 }, { "grad_norm": 0.15916046500205994, "learning_rate": 8.061431391926631e-05, "loss": 0.0072, "step": 26060 }, { "grad_norm": 0.1660308837890625, "learning_rate": 8.059797014163195e-05, "loss": 0.008, "step": 26070 }, { "grad_norm": 0.14040562510490417, "learning_rate": 8.058162113564133e-05, "loss": 0.0083, "step": 26080 }, { "grad_norm": 0.14238840341567993, "learning_rate": 8.056526690408806e-05, "loss": 0.0086, "step": 26090 }, { "grad_norm": 0.13726045191287994, "learning_rate": 8.054890744976666e-05, "loss": 0.0083, "step": 26100 }, { "grad_norm": 0.16218577325344086, "learning_rate": 8.053254277547248e-05, "loss": 0.0073, "step": 26110 }, { "grad_norm": 0.12212906032800674, "learning_rate": 8.051617288400182e-05, "loss": 0.0071, "step": 26120 }, { "grad_norm": 0.16999363899230957, "learning_rate": 8.049979777815182e-05, "loss": 0.0079, "step": 26130 }, { "grad_norm": 0.14304973185062408, "learning_rate": 8.048341746072054e-05, "loss": 0.0077, "step": 26140 }, { "grad_norm": 0.12751543521881104, "learning_rate": 8.046703193450696e-05, "loss": 0.0091, "step": 26150 }, { "grad_norm": 0.12352657318115234, "learning_rate": 8.04506412023109e-05, "loss": 0.0067, "step": 26160 }, { "grad_norm": 0.1633034497499466, "learning_rate": 8.043424526693306e-05, "loss": 0.0099, "step": 26170 }, { "grad_norm": 0.13628274202346802, "learning_rate": 8.04178441311751e-05, "loss": 0.0069, "step": 26180 }, { "grad_norm": 0.12218693643808365, "learning_rate": 8.04014377978395e-05, "loss": 0.0068, "step": 26190 }, { "grad_norm": 0.14092695713043213, "learning_rate": 8.038502626972967e-05, "loss": 0.0084, "step": 26200 }, { "grad_norm": 0.12697041034698486, "learning_rate": 8.036860954964989e-05, "loss": 0.0079, "step": 26210 }, { "grad_norm": 0.1264396607875824, "learning_rate": 8.035218764040531e-05, "loss": 0.0079, "step": 26220 }, { "grad_norm": 0.17676284909248352, "learning_rate": 8.033576054480199e-05, "loss": 0.01, "step": 26230 }, { "grad_norm": 0.14463287591934204, "learning_rate": 8.031932826564688e-05, "loss": 0.0082, "step": 26240 }, { "grad_norm": 0.13225524127483368, "learning_rate": 8.030289080574782e-05, "loss": 0.0067, "step": 26250 }, { "grad_norm": 0.15877403318881989, "learning_rate": 8.028644816791349e-05, "loss": 0.0083, "step": 26260 }, { "grad_norm": 0.1314684897661209, "learning_rate": 8.027000035495351e-05, "loss": 0.0071, "step": 26270 }, { "grad_norm": 0.14602287113666534, "learning_rate": 8.025354736967836e-05, "loss": 0.006, "step": 26280 }, { "grad_norm": 0.0944100096821785, "learning_rate": 8.023708921489941e-05, "loss": 0.0082, "step": 26290 }, { "grad_norm": 0.1187836080789566, "learning_rate": 8.022062589342887e-05, "loss": 0.0082, "step": 26300 }, { "grad_norm": 0.16462911665439606, "learning_rate": 8.020415740807993e-05, "loss": 0.0065, "step": 26310 }, { "grad_norm": 0.1274576038122177, "learning_rate": 8.018768376166656e-05, "loss": 0.0082, "step": 26320 }, { "grad_norm": 0.15757863223552704, "learning_rate": 8.017120495700368e-05, "loss": 0.0092, "step": 26330 }, { "grad_norm": 0.1597461700439453, "learning_rate": 8.015472099690704e-05, "loss": 0.0084, "step": 26340 }, { "grad_norm": 0.139103502035141, "learning_rate": 8.013823188419332e-05, "loss": 0.0105, "step": 26350 }, { "grad_norm": 0.17005525529384613, "learning_rate": 8.012173762168006e-05, "loss": 0.0072, "step": 26360 }, { "grad_norm": 0.1564396768808365, "learning_rate": 8.010523821218567e-05, "loss": 0.007, "step": 26370 }, { "grad_norm": 0.15359917283058167, "learning_rate": 8.008873365852945e-05, "loss": 0.0085, "step": 26380 }, { "grad_norm": 0.14685699343681335, "learning_rate": 8.007222396353157e-05, "loss": 0.0101, "step": 26390 }, { "grad_norm": 0.12999030947685242, "learning_rate": 8.00557091300131e-05, "loss": 0.0075, "step": 26400 }, { "grad_norm": 0.15498977899551392, "learning_rate": 8.003918916079597e-05, "loss": 0.0082, "step": 26410 }, { "grad_norm": 0.11454295367002487, "learning_rate": 8.002266405870298e-05, "loss": 0.0077, "step": 26420 }, { "grad_norm": 0.140147402882576, "learning_rate": 8.000613382655782e-05, "loss": 0.0074, "step": 26430 }, { "grad_norm": 0.11601893603801727, "learning_rate": 7.998959846718505e-05, "loss": 0.0081, "step": 26440 }, { "grad_norm": 0.15308505296707153, "learning_rate": 7.997305798341012e-05, "loss": 0.0071, "step": 26450 }, { "grad_norm": 0.17755377292633057, "learning_rate": 7.995651237805937e-05, "loss": 0.008, "step": 26460 }, { "grad_norm": 0.17879386246204376, "learning_rate": 7.993996165395996e-05, "loss": 0.0087, "step": 26470 }, { "grad_norm": 0.1580052375793457, "learning_rate": 7.992340581393996e-05, "loss": 0.0079, "step": 26480 }, { "grad_norm": 0.18980222940444946, "learning_rate": 7.990684486082831e-05, "loss": 0.0085, "step": 26490 }, { "grad_norm": 0.131202831864357, "learning_rate": 7.989027879745482e-05, "loss": 0.0069, "step": 26500 }, { "grad_norm": 0.1336778700351715, "learning_rate": 7.98737076266502e-05, "loss": 0.0073, "step": 26510 }, { "grad_norm": 0.1487671434879303, "learning_rate": 7.985713135124598e-05, "loss": 0.0079, "step": 26520 }, { "grad_norm": 0.14393232762813568, "learning_rate": 7.98405499740746e-05, "loss": 0.0078, "step": 26530 }, { "grad_norm": 0.15245234966278076, "learning_rate": 7.98239634979694e-05, "loss": 0.01, "step": 26540 }, { "grad_norm": 0.1380636841058731, "learning_rate": 7.98073719257645e-05, "loss": 0.0087, "step": 26550 }, { "grad_norm": 0.11210367828607559, "learning_rate": 7.979077526029499e-05, "loss": 0.0081, "step": 26560 }, { "grad_norm": 0.13556592166423798, "learning_rate": 7.977417350439675e-05, "loss": 0.0098, "step": 26570 }, { "grad_norm": 0.16047652065753937, "learning_rate": 7.97575666609066e-05, "loss": 0.0095, "step": 26580 }, { "grad_norm": 0.1274237036705017, "learning_rate": 7.974095473266216e-05, "loss": 0.007, "step": 26590 }, { "grad_norm": 0.11965826898813248, "learning_rate": 7.972433772250198e-05, "loss": 0.0075, "step": 26600 }, { "grad_norm": 0.14397285878658295, "learning_rate": 7.970771563326544e-05, "loss": 0.0114, "step": 26610 }, { "grad_norm": 0.10142018646001816, "learning_rate": 7.96910884677928e-05, "loss": 0.0094, "step": 26620 }, { "grad_norm": 0.19911061227321625, "learning_rate": 7.967445622892523e-05, "loss": 0.0089, "step": 26630 }, { "grad_norm": 0.12097957730293274, "learning_rate": 7.965781891950465e-05, "loss": 0.0074, "step": 26640 }, { "grad_norm": 0.09512705355882645, "learning_rate": 7.964117654237397e-05, "loss": 0.0058, "step": 26650 }, { "grad_norm": 0.11228813230991364, "learning_rate": 7.962452910037692e-05, "loss": 0.0077, "step": 26660 }, { "grad_norm": 0.13747188448905945, "learning_rate": 7.96078765963581e-05, "loss": 0.0078, "step": 26670 }, { "grad_norm": 0.11917976289987564, "learning_rate": 7.95912190331629e-05, "loss": 0.0059, "step": 26680 }, { "grad_norm": 0.1204034686088562, "learning_rate": 7.957455641363772e-05, "loss": 0.0082, "step": 26690 }, { "grad_norm": 0.13071560859680176, "learning_rate": 7.955788874062968e-05, "loss": 0.0076, "step": 26700 }, { "grad_norm": 0.14575715363025665, "learning_rate": 7.95412160169869e-05, "loss": 0.0084, "step": 26710 }, { "grad_norm": 0.178354412317276, "learning_rate": 7.952453824555824e-05, "loss": 0.0071, "step": 26720 }, { "grad_norm": 0.14466696977615356, "learning_rate": 7.95078554291935e-05, "loss": 0.0069, "step": 26730 }, { "grad_norm": 0.13583283126354218, "learning_rate": 7.94911675707433e-05, "loss": 0.0056, "step": 26740 }, { "grad_norm": 0.12220821529626846, "learning_rate": 7.947447467305915e-05, "loss": 0.0072, "step": 26750 }, { "grad_norm": 0.10535510629415512, "learning_rate": 7.94577767389934e-05, "loss": 0.0073, "step": 26760 }, { "grad_norm": 0.1524219512939453, "learning_rate": 7.944107377139928e-05, "loss": 0.0067, "step": 26770 }, { "grad_norm": 0.1641198992729187, "learning_rate": 7.942436577313088e-05, "loss": 0.0089, "step": 26780 }, { "grad_norm": 0.11941149830818176, "learning_rate": 7.940765274704312e-05, "loss": 0.0059, "step": 26790 }, { "grad_norm": 0.12601564824581146, "learning_rate": 7.939093469599181e-05, "loss": 0.0069, "step": 26800 }, { "grad_norm": 0.14452984929084778, "learning_rate": 7.93742116228336e-05, "loss": 0.0073, "step": 26810 }, { "grad_norm": 0.1472085565328598, "learning_rate": 7.935748353042602e-05, "loss": 0.0079, "step": 26820 }, { "grad_norm": 0.17864902317523956, "learning_rate": 7.934075042162744e-05, "loss": 0.0091, "step": 26830 }, { "grad_norm": 0.18553312122821808, "learning_rate": 7.932401229929705e-05, "loss": 0.0086, "step": 26840 }, { "grad_norm": 0.12389305979013443, "learning_rate": 7.9307269166295e-05, "loss": 0.0078, "step": 26850 }, { "grad_norm": 0.1606261432170868, "learning_rate": 7.92905210254822e-05, "loss": 0.0069, "step": 26860 }, { "grad_norm": 0.15302066504955292, "learning_rate": 7.927376787972045e-05, "loss": 0.0097, "step": 26870 }, { "grad_norm": 0.18278980255126953, "learning_rate": 7.92570097318724e-05, "loss": 0.008, "step": 26880 }, { "grad_norm": 0.15024617314338684, "learning_rate": 7.924024658480158e-05, "loss": 0.0106, "step": 26890 }, { "grad_norm": 0.15371516346931458, "learning_rate": 7.922347844137233e-05, "loss": 0.0088, "step": 26900 }, { "grad_norm": 0.13591624796390533, "learning_rate": 7.92067053044499e-05, "loss": 0.007, "step": 26910 }, { "grad_norm": 0.14360937476158142, "learning_rate": 7.918992717690031e-05, "loss": 0.0076, "step": 26920 }, { "grad_norm": 0.14964547753334045, "learning_rate": 7.917314406159053e-05, "loss": 0.008, "step": 26930 }, { "grad_norm": 0.1119002029299736, "learning_rate": 7.915635596138832e-05, "loss": 0.0069, "step": 26940 }, { "grad_norm": 0.1677713692188263, "learning_rate": 7.913956287916228e-05, "loss": 0.0097, "step": 26950 }, { "grad_norm": 0.18623636662960052, "learning_rate": 7.912276481778193e-05, "loss": 0.0091, "step": 26960 }, { "grad_norm": 0.17331227660179138, "learning_rate": 7.910596178011759e-05, "loss": 0.0076, "step": 26970 }, { "grad_norm": 0.15020276606082916, "learning_rate": 7.908915376904043e-05, "loss": 0.0083, "step": 26980 }, { "grad_norm": 0.14698314666748047, "learning_rate": 7.907234078742247e-05, "loss": 0.0086, "step": 26990 }, { "grad_norm": 0.12887991964817047, "learning_rate": 7.90555228381366e-05, "loss": 0.007, "step": 27000 }, { "grad_norm": 0.22349095344543457, "learning_rate": 7.903869992405656e-05, "loss": 0.0085, "step": 27010 }, { "grad_norm": 0.1561284363269806, "learning_rate": 7.902187204805691e-05, "loss": 0.0085, "step": 27020 }, { "grad_norm": 0.1525171548128128, "learning_rate": 7.900503921301308e-05, "loss": 0.0071, "step": 27030 }, { "grad_norm": 0.12831000983715057, "learning_rate": 7.898820142180133e-05, "loss": 0.007, "step": 27040 }, { "grad_norm": 0.08536543697118759, "learning_rate": 7.897135867729879e-05, "loss": 0.0069, "step": 27050 }, { "grad_norm": 0.11665548384189606, "learning_rate": 7.89545109823834e-05, "loss": 0.0086, "step": 27060 }, { "grad_norm": 0.12160266190767288, "learning_rate": 7.8937658339934e-05, "loss": 0.0071, "step": 27070 }, { "grad_norm": 0.1392119973897934, "learning_rate": 7.892080075283026e-05, "loss": 0.0085, "step": 27080 }, { "grad_norm": 0.0960378423333168, "learning_rate": 7.890393822395263e-05, "loss": 0.006, "step": 27090 }, { "grad_norm": 0.14022231101989746, "learning_rate": 7.88870707561825e-05, "loss": 0.0067, "step": 27100 }, { "grad_norm": 0.14484000205993652, "learning_rate": 7.887019835240203e-05, "loss": 0.0062, "step": 27110 }, { "grad_norm": 0.11098676174879074, "learning_rate": 7.885332101549427e-05, "loss": 0.0072, "step": 27120 }, { "grad_norm": 0.14470969140529633, "learning_rate": 7.883643874834308e-05, "loss": 0.0073, "step": 27130 }, { "grad_norm": 0.1279807984828949, "learning_rate": 7.881955155383321e-05, "loss": 0.0066, "step": 27140 }, { "grad_norm": 0.14072473347187042, "learning_rate": 7.880265943485017e-05, "loss": 0.0076, "step": 27150 }, { "grad_norm": 0.12474218755960464, "learning_rate": 7.878576239428038e-05, "loss": 0.009, "step": 27160 }, { "grad_norm": 0.17234231531620026, "learning_rate": 7.87688604350111e-05, "loss": 0.0093, "step": 27170 }, { "grad_norm": 0.1181819960474968, "learning_rate": 7.875195355993042e-05, "loss": 0.0081, "step": 27180 }, { "grad_norm": 0.13745971024036407, "learning_rate": 7.873504177192724e-05, "loss": 0.0099, "step": 27190 }, { "grad_norm": 0.11643730849027634, "learning_rate": 7.87181250738913e-05, "loss": 0.0063, "step": 27200 }, { "grad_norm": 0.11329659074544907, "learning_rate": 7.870120346871324e-05, "loss": 0.0071, "step": 27210 }, { "grad_norm": 0.12453252822160721, "learning_rate": 7.86842769592845e-05, "loss": 0.0067, "step": 27220 }, { "grad_norm": 0.17725974321365356, "learning_rate": 7.866734554849732e-05, "loss": 0.007, "step": 27230 }, { "grad_norm": 0.12988559901714325, "learning_rate": 7.865040923924486e-05, "loss": 0.0066, "step": 27240 }, { "grad_norm": 0.1521323025226593, "learning_rate": 7.863346803442104e-05, "loss": 0.0079, "step": 27250 }, { "grad_norm": 0.14018762111663818, "learning_rate": 7.861652193692067e-05, "loss": 0.0076, "step": 27260 }, { "grad_norm": 0.11438073217868805, "learning_rate": 7.859957094963937e-05, "loss": 0.0065, "step": 27270 }, { "grad_norm": 0.09625651687383652, "learning_rate": 7.858261507547357e-05, "loss": 0.0063, "step": 27280 }, { "grad_norm": 0.1090465784072876, "learning_rate": 7.856565431732061e-05, "loss": 0.0072, "step": 27290 }, { "grad_norm": 0.16740891337394714, "learning_rate": 7.854868867807859e-05, "loss": 0.0084, "step": 27300 }, { "grad_norm": 0.10942424833774567, "learning_rate": 7.85317181606465e-05, "loss": 0.0099, "step": 27310 }, { "grad_norm": 0.20725414156913757, "learning_rate": 7.85147427679241e-05, "loss": 0.0077, "step": 27320 }, { "grad_norm": 0.13612179458141327, "learning_rate": 7.849776250281205e-05, "loss": 0.0068, "step": 27330 }, { "grad_norm": 0.10331705212593079, "learning_rate": 7.84807773682118e-05, "loss": 0.0065, "step": 27340 }, { "grad_norm": 0.11815029382705688, "learning_rate": 7.846378736702565e-05, "loss": 0.0063, "step": 27350 }, { "grad_norm": 0.0928235650062561, "learning_rate": 7.844679250215671e-05, "loss": 0.0061, "step": 27360 }, { "grad_norm": 0.13532410562038422, "learning_rate": 7.842979277650898e-05, "loss": 0.0076, "step": 27370 }, { "grad_norm": 0.1558816283941269, "learning_rate": 7.84127881929872e-05, "loss": 0.0069, "step": 27380 }, { "grad_norm": 0.1108589693903923, "learning_rate": 7.839577875449704e-05, "loss": 0.0063, "step": 27390 }, { "grad_norm": 0.15380986034870148, "learning_rate": 7.837876446394489e-05, "loss": 0.0059, "step": 27400 }, { "grad_norm": 0.13129396736621857, "learning_rate": 7.836174532423805e-05, "loss": 0.0062, "step": 27410 }, { "grad_norm": 0.11042232066392899, "learning_rate": 7.834472133828466e-05, "loss": 0.0073, "step": 27420 }, { "grad_norm": 0.1451578438282013, "learning_rate": 7.832769250899359e-05, "loss": 0.0085, "step": 27430 }, { "grad_norm": 0.1421964317560196, "learning_rate": 7.831065883927464e-05, "loss": 0.0073, "step": 27440 }, { "grad_norm": 0.11767441034317017, "learning_rate": 7.829362033203841e-05, "loss": 0.0087, "step": 27450 }, { "grad_norm": 0.1145339086651802, "learning_rate": 7.827657699019628e-05, "loss": 0.0062, "step": 27460 }, { "grad_norm": 0.11589214205741882, "learning_rate": 7.825952881666052e-05, "loss": 0.0089, "step": 27470 }, { "grad_norm": 0.11056577414274216, "learning_rate": 7.824247581434418e-05, "loss": 0.0081, "step": 27480 }, { "grad_norm": 0.11699116975069046, "learning_rate": 7.822541798616116e-05, "loss": 0.0061, "step": 27490 }, { "grad_norm": 0.14044412970542908, "learning_rate": 7.820835533502617e-05, "loss": 0.0071, "step": 27500 }, { "grad_norm": 0.2189895212650299, "learning_rate": 7.819128786385475e-05, "loss": 0.0068, "step": 27510 }, { "grad_norm": 0.18135730922222137, "learning_rate": 7.817421557556329e-05, "loss": 0.0071, "step": 27520 }, { "grad_norm": 0.1384219080209732, "learning_rate": 7.815713847306893e-05, "loss": 0.0061, "step": 27530 }, { "grad_norm": 0.13167288899421692, "learning_rate": 7.81400565592897e-05, "loss": 0.0079, "step": 27540 }, { "grad_norm": 0.10647698491811752, "learning_rate": 7.812296983714444e-05, "loss": 0.0072, "step": 27550 }, { "grad_norm": 0.10799508541822433, "learning_rate": 7.810587830955281e-05, "loss": 0.0054, "step": 27560 }, { "grad_norm": 0.11563495546579361, "learning_rate": 7.808878197943528e-05, "loss": 0.0079, "step": 27570 }, { "grad_norm": 0.16224618256092072, "learning_rate": 7.807168084971312e-05, "loss": 0.0076, "step": 27580 }, { "grad_norm": 0.16802223026752472, "learning_rate": 7.805457492330849e-05, "loss": 0.0085, "step": 27590 }, { "grad_norm": 0.10818915069103241, "learning_rate": 7.803746420314428e-05, "loss": 0.0071, "step": 27600 }, { "grad_norm": 0.1056530699133873, "learning_rate": 7.802034869214428e-05, "loss": 0.0085, "step": 27610 }, { "grad_norm": 0.1697094738483429, "learning_rate": 7.800322839323303e-05, "loss": 0.0104, "step": 27620 }, { "grad_norm": 0.1497020125389099, "learning_rate": 7.798610330933593e-05, "loss": 0.0081, "step": 27630 }, { "grad_norm": 0.13990487158298492, "learning_rate": 7.796897344337922e-05, "loss": 0.0069, "step": 27640 }, { "grad_norm": 0.14192765951156616, "learning_rate": 7.795183879828989e-05, "loss": 0.0081, "step": 27650 }, { "grad_norm": 0.1812434047460556, "learning_rate": 7.793469937699579e-05, "loss": 0.007, "step": 27660 }, { "grad_norm": 0.17372077703475952, "learning_rate": 7.791755518242558e-05, "loss": 0.0067, "step": 27670 }, { "grad_norm": 0.17334939539432526, "learning_rate": 7.790040621750876e-05, "loss": 0.0061, "step": 27680 }, { "grad_norm": 0.14690446853637695, "learning_rate": 7.788325248517558e-05, "loss": 0.0074, "step": 27690 }, { "grad_norm": 0.16990157961845398, "learning_rate": 7.786609398835715e-05, "loss": 0.0085, "step": 27700 }, { "grad_norm": 0.15515445172786713, "learning_rate": 7.784893072998541e-05, "loss": 0.0073, "step": 27710 }, { "grad_norm": 0.1358083039522171, "learning_rate": 7.783176271299306e-05, "loss": 0.0062, "step": 27720 }, { "grad_norm": 0.1441067010164261, "learning_rate": 7.781458994031368e-05, "loss": 0.0062, "step": 27730 }, { "grad_norm": 0.11652660369873047, "learning_rate": 7.779741241488161e-05, "loss": 0.007, "step": 27740 }, { "grad_norm": 0.16058731079101562, "learning_rate": 7.7780230139632e-05, "loss": 0.0076, "step": 27750 }, { "grad_norm": 0.14970663189888, "learning_rate": 7.776304311750087e-05, "loss": 0.0086, "step": 27760 }, { "grad_norm": 0.1711483895778656, "learning_rate": 7.7745851351425e-05, "loss": 0.0106, "step": 27770 }, { "grad_norm": 0.13376295566558838, "learning_rate": 7.772865484434197e-05, "loss": 0.0062, "step": 27780 }, { "grad_norm": 0.15013661980628967, "learning_rate": 7.77114535991902e-05, "loss": 0.0081, "step": 27790 }, { "grad_norm": 0.10601893067359924, "learning_rate": 7.769424761890893e-05, "loss": 0.0075, "step": 27800 }, { "grad_norm": 0.10187036544084549, "learning_rate": 7.767703690643817e-05, "loss": 0.0076, "step": 27810 }, { "grad_norm": 0.17082810401916504, "learning_rate": 7.76598214647188e-05, "loss": 0.0097, "step": 27820 }, { "grad_norm": 0.09483398497104645, "learning_rate": 7.764260129669241e-05, "loss": 0.0074, "step": 27830 }, { "grad_norm": 0.12208062410354614, "learning_rate": 7.76253764053015e-05, "loss": 0.0073, "step": 27840 }, { "grad_norm": 0.1324504315853119, "learning_rate": 7.760814679348932e-05, "loss": 0.0066, "step": 27850 }, { "grad_norm": 0.11421188712120056, "learning_rate": 7.759091246419992e-05, "loss": 0.0073, "step": 27860 }, { "grad_norm": 0.14668124914169312, "learning_rate": 7.75736734203782e-05, "loss": 0.0092, "step": 27870 }, { "grad_norm": 0.1350150853395462, "learning_rate": 7.755642966496985e-05, "loss": 0.0079, "step": 27880 }, { "grad_norm": 0.14107942581176758, "learning_rate": 7.753918120092132e-05, "loss": 0.0087, "step": 27890 }, { "grad_norm": 0.15983520448207855, "learning_rate": 7.752192803117993e-05, "loss": 0.0064, "step": 27900 }, { "grad_norm": 0.1410730928182602, "learning_rate": 7.750467015869377e-05, "loss": 0.009, "step": 27910 }, { "grad_norm": 0.1777774840593338, "learning_rate": 7.748740758641174e-05, "loss": 0.0065, "step": 27920 }, { "grad_norm": 0.11932098865509033, "learning_rate": 7.74701403172835e-05, "loss": 0.009, "step": 27930 }, { "grad_norm": 0.1350911259651184, "learning_rate": 7.745286835425962e-05, "loss": 0.0063, "step": 27940 }, { "grad_norm": 0.15263350307941437, "learning_rate": 7.743559170029138e-05, "loss": 0.0105, "step": 27950 }, { "grad_norm": 0.1853935569524765, "learning_rate": 7.741831035833087e-05, "loss": 0.0071, "step": 27960 }, { "grad_norm": 0.1507398933172226, "learning_rate": 7.740102433133102e-05, "loss": 0.0084, "step": 27970 }, { "grad_norm": 0.16300906240940094, "learning_rate": 7.738373362224553e-05, "loss": 0.0076, "step": 27980 }, { "grad_norm": 0.1462026685476303, "learning_rate": 7.73664382340289e-05, "loss": 0.0082, "step": 27990 }, { "grad_norm": 0.13785603642463684, "learning_rate": 7.734913816963647e-05, "loss": 0.0074, "step": 28000 }, { "grad_norm": 0.14127285778522491, "learning_rate": 7.73318334320243e-05, "loss": 0.0085, "step": 28010 }, { "grad_norm": 0.11712609976530075, "learning_rate": 7.731452402414934e-05, "loss": 0.0071, "step": 28020 }, { "grad_norm": 0.11071576178073883, "learning_rate": 7.729720994896928e-05, "loss": 0.0073, "step": 28030 }, { "grad_norm": 0.09419898688793182, "learning_rate": 7.727989120944262e-05, "loss": 0.0072, "step": 28040 }, { "grad_norm": 0.1718587875366211, "learning_rate": 7.726256780852865e-05, "loss": 0.0091, "step": 28050 }, { "grad_norm": 0.16333521902561188, "learning_rate": 7.724523974918749e-05, "loss": 0.0073, "step": 28060 }, { "grad_norm": 0.13963885605335236, "learning_rate": 7.722790703438002e-05, "loss": 0.0067, "step": 28070 }, { "grad_norm": 0.1395438015460968, "learning_rate": 7.72105696670679e-05, "loss": 0.0074, "step": 28080 }, { "grad_norm": 0.1315978616476059, "learning_rate": 7.719322765021364e-05, "loss": 0.0059, "step": 28090 }, { "grad_norm": 0.12654098868370056, "learning_rate": 7.717588098678051e-05, "loss": 0.0097, "step": 28100 }, { "grad_norm": 0.13411326706409454, "learning_rate": 7.715852967973258e-05, "loss": 0.0063, "step": 28110 }, { "grad_norm": 0.1079765260219574, "learning_rate": 7.714117373203474e-05, "loss": 0.0068, "step": 28120 }, { "grad_norm": 0.11811509728431702, "learning_rate": 7.712381314665259e-05, "loss": 0.0065, "step": 28130 }, { "grad_norm": 0.14151974022388458, "learning_rate": 7.710644792655261e-05, "loss": 0.0078, "step": 28140 }, { "grad_norm": 0.1371549665927887, "learning_rate": 7.708907807470207e-05, "loss": 0.0057, "step": 28150 }, { "grad_norm": 0.11175672709941864, "learning_rate": 7.707170359406896e-05, "loss": 0.0077, "step": 28160 }, { "grad_norm": 0.17937515676021576, "learning_rate": 7.705432448762213e-05, "loss": 0.0087, "step": 28170 }, { "grad_norm": 0.12658776342868805, "learning_rate": 7.703694075833117e-05, "loss": 0.0067, "step": 28180 }, { "grad_norm": 0.15023720264434814, "learning_rate": 7.70195524091665e-05, "loss": 0.0094, "step": 28190 }, { "grad_norm": 0.1418134570121765, "learning_rate": 7.70021594430993e-05, "loss": 0.0075, "step": 28200 }, { "grad_norm": 0.12327716499567032, "learning_rate": 7.698476186310157e-05, "loss": 0.0073, "step": 28210 }, { "grad_norm": 0.12356948107481003, "learning_rate": 7.696735967214608e-05, "loss": 0.0073, "step": 28220 }, { "grad_norm": 0.12814190983772278, "learning_rate": 7.694995287320636e-05, "loss": 0.0068, "step": 28230 }, { "grad_norm": 0.17106400430202484, "learning_rate": 7.693254146925679e-05, "loss": 0.0078, "step": 28240 }, { "grad_norm": 0.1776738315820694, "learning_rate": 7.691512546327251e-05, "loss": 0.0067, "step": 28250 }, { "grad_norm": 0.08866232633590698, "learning_rate": 7.689770485822939e-05, "loss": 0.0086, "step": 28260 }, { "grad_norm": 0.1527235358953476, "learning_rate": 7.688027965710416e-05, "loss": 0.0065, "step": 28270 }, { "grad_norm": 0.10651157051324844, "learning_rate": 7.686284986287433e-05, "loss": 0.0072, "step": 28280 }, { "grad_norm": 0.14396589994430542, "learning_rate": 7.684541547851817e-05, "loss": 0.0061, "step": 28290 }, { "grad_norm": 0.12821269035339355, "learning_rate": 7.68279765070147e-05, "loss": 0.0071, "step": 28300 }, { "grad_norm": 0.10129640251398087, "learning_rate": 7.68105329513438e-05, "loss": 0.0083, "step": 28310 }, { "grad_norm": 0.13292714953422546, "learning_rate": 7.67930848144861e-05, "loss": 0.0075, "step": 28320 }, { "grad_norm": 0.11997776478528976, "learning_rate": 7.6775632099423e-05, "loss": 0.0056, "step": 28330 }, { "grad_norm": 0.10218514502048492, "learning_rate": 7.675817480913667e-05, "loss": 0.0068, "step": 28340 }, { "grad_norm": 0.11801417171955109, "learning_rate": 7.674071294661011e-05, "loss": 0.0063, "step": 28350 }, { "grad_norm": 0.14216630160808563, "learning_rate": 7.672324651482707e-05, "loss": 0.0061, "step": 28360 }, { "grad_norm": 0.12854450941085815, "learning_rate": 7.670577551677209e-05, "loss": 0.0057, "step": 28370 }, { "grad_norm": 0.11951921880245209, "learning_rate": 7.668829995543047e-05, "loss": 0.0072, "step": 28380 }, { "grad_norm": 0.1808454543352127, "learning_rate": 7.667081983378832e-05, "loss": 0.0082, "step": 28390 }, { "grad_norm": 0.10870188474655151, "learning_rate": 7.66533351548325e-05, "loss": 0.0057, "step": 28400 }, { "grad_norm": 0.13373400270938873, "learning_rate": 7.663584592155069e-05, "loss": 0.0091, "step": 28410 }, { "grad_norm": 0.16264545917510986, "learning_rate": 7.661835213693129e-05, "loss": 0.007, "step": 28420 }, { "grad_norm": 0.1351575255393982, "learning_rate": 7.660085380396353e-05, "loss": 0.0073, "step": 28430 }, { "grad_norm": 0.1143224760890007, "learning_rate": 7.658335092563738e-05, "loss": 0.0065, "step": 28440 }, { "grad_norm": 0.12198404222726822, "learning_rate": 7.656584350494362e-05, "loss": 0.0067, "step": 28450 }, { "grad_norm": 0.187996044754982, "learning_rate": 7.654833154487378e-05, "loss": 0.0063, "step": 28460 }, { "grad_norm": 0.14753352105617523, "learning_rate": 7.653081504842017e-05, "loss": 0.0089, "step": 28470 }, { "grad_norm": 0.19266703724861145, "learning_rate": 7.65132940185759e-05, "loss": 0.0079, "step": 28480 }, { "grad_norm": 0.12550851702690125, "learning_rate": 7.649576845833481e-05, "loss": 0.0065, "step": 28490 }, { "grad_norm": 0.13949266076087952, "learning_rate": 7.647823837069156e-05, "loss": 0.0084, "step": 28500 }, { "grad_norm": 0.17621329426765442, "learning_rate": 7.646070375864156e-05, "loss": 0.0058, "step": 28510 }, { "grad_norm": 0.1347735971212387, "learning_rate": 7.644316462518099e-05, "loss": 0.0065, "step": 28520 }, { "grad_norm": 0.12145828455686569, "learning_rate": 7.642562097330679e-05, "loss": 0.0081, "step": 28530 }, { "grad_norm": 0.12779881060123444, "learning_rate": 7.640807280601671e-05, "loss": 0.0074, "step": 28540 }, { "grad_norm": 0.12850220501422882, "learning_rate": 7.639052012630927e-05, "loss": 0.0076, "step": 28550 }, { "grad_norm": 0.0960702896118164, "learning_rate": 7.63729629371837e-05, "loss": 0.0081, "step": 28560 }, { "grad_norm": 0.08495578169822693, "learning_rate": 7.635540124164009e-05, "loss": 0.0071, "step": 28570 }, { "grad_norm": 0.11168348789215088, "learning_rate": 7.633783504267922e-05, "loss": 0.0069, "step": 28580 }, { "grad_norm": 0.13618507981300354, "learning_rate": 7.632026434330269e-05, "loss": 0.0071, "step": 28590 }, { "grad_norm": 0.10781387239694595, "learning_rate": 7.630268914651282e-05, "loss": 0.0064, "step": 28600 }, { "grad_norm": 0.13330066204071045, "learning_rate": 7.628510945531278e-05, "loss": 0.007, "step": 28610 }, { "grad_norm": 0.12097222357988358, "learning_rate": 7.626752527270641e-05, "loss": 0.007, "step": 28620 }, { "grad_norm": 0.135736346244812, "learning_rate": 7.62499366016984e-05, "loss": 0.009, "step": 28630 }, { "grad_norm": 0.14202603697776794, "learning_rate": 7.623234344529416e-05, "loss": 0.0068, "step": 28640 }, { "grad_norm": 0.17221269011497498, "learning_rate": 7.62147458064999e-05, "loss": 0.0072, "step": 28650 }, { "grad_norm": 0.09825398772954941, "learning_rate": 7.619714368832254e-05, "loss": 0.0057, "step": 28660 }, { "grad_norm": 0.0892796665430069, "learning_rate": 7.61795370937698e-05, "loss": 0.0061, "step": 28670 }, { "grad_norm": 0.15165835618972778, "learning_rate": 7.61619260258502e-05, "loss": 0.0073, "step": 28680 }, { "grad_norm": 0.15053869783878326, "learning_rate": 7.614431048757298e-05, "loss": 0.0079, "step": 28690 }, { "grad_norm": 0.13946178555488586, "learning_rate": 7.612669048194814e-05, "loss": 0.0077, "step": 28700 }, { "grad_norm": 0.15681149065494537, "learning_rate": 7.610906601198646e-05, "loss": 0.0072, "step": 28710 }, { "grad_norm": 0.1339353770017624, "learning_rate": 7.60914370806995e-05, "loss": 0.0067, "step": 28720 }, { "grad_norm": 0.15553072094917297, "learning_rate": 7.607380369109953e-05, "loss": 0.0071, "step": 28730 }, { "grad_norm": 0.15855459868907928, "learning_rate": 7.605616584619961e-05, "loss": 0.0067, "step": 28740 }, { "grad_norm": 0.1396535038948059, "learning_rate": 7.603852354901362e-05, "loss": 0.008, "step": 28750 }, { "grad_norm": 0.12966077029705048, "learning_rate": 7.602087680255609e-05, "loss": 0.0069, "step": 28760 }, { "grad_norm": 0.146520733833313, "learning_rate": 7.600322560984238e-05, "loss": 0.0082, "step": 28770 }, { "grad_norm": 0.1281108260154724, "learning_rate": 7.598556997388863e-05, "loss": 0.0081, "step": 28780 }, { "grad_norm": 0.1535869687795639, "learning_rate": 7.596790989771166e-05, "loss": 0.0066, "step": 28790 }, { "grad_norm": 0.1169038712978363, "learning_rate": 7.595024538432914e-05, "loss": 0.0065, "step": 28800 }, { "grad_norm": 0.132165789604187, "learning_rate": 7.59325764367594e-05, "loss": 0.0066, "step": 28810 }, { "grad_norm": 0.10786860436201096, "learning_rate": 7.59149030580216e-05, "loss": 0.0084, "step": 28820 }, { "grad_norm": 0.10978053510189056, "learning_rate": 7.589722525113562e-05, "loss": 0.006, "step": 28830 }, { "grad_norm": 0.1463613063097, "learning_rate": 7.587954301912216e-05, "loss": 0.0072, "step": 28840 }, { "grad_norm": 0.13721773028373718, "learning_rate": 7.586185636500263e-05, "loss": 0.0064, "step": 28850 }, { "grad_norm": 0.0971483662724495, "learning_rate": 7.584416529179914e-05, "loss": 0.0063, "step": 28860 }, { "grad_norm": 0.11575228720903397, "learning_rate": 7.582646980253465e-05, "loss": 0.0063, "step": 28870 }, { "grad_norm": 0.10167846828699112, "learning_rate": 7.580876990023282e-05, "loss": 0.0062, "step": 28880 }, { "grad_norm": 0.09932701289653778, "learning_rate": 7.579106558791809e-05, "loss": 0.0055, "step": 28890 }, { "grad_norm": 0.0901600569486618, "learning_rate": 7.577335686861565e-05, "loss": 0.008, "step": 28900 }, { "grad_norm": 0.12250721454620361, "learning_rate": 7.575564374535141e-05, "loss": 0.0062, "step": 28910 }, { "grad_norm": 0.13999147713184357, "learning_rate": 7.573792622115207e-05, "loss": 0.0053, "step": 28920 }, { "grad_norm": 0.13823287189006805, "learning_rate": 7.572020429904507e-05, "loss": 0.008, "step": 28930 }, { "grad_norm": 0.135576069355011, "learning_rate": 7.570247798205861e-05, "loss": 0.0067, "step": 28940 }, { "grad_norm": 0.141694113612175, "learning_rate": 7.568474727322164e-05, "loss": 0.0072, "step": 28950 }, { "grad_norm": 0.13926462829113007, "learning_rate": 7.566701217556384e-05, "loss": 0.0085, "step": 28960 }, { "grad_norm": 0.13083449006080627, "learning_rate": 7.564927269211564e-05, "loss": 0.0078, "step": 28970 }, { "grad_norm": 0.12410129606723785, "learning_rate": 7.563152882590824e-05, "loss": 0.0077, "step": 28980 }, { "grad_norm": 0.1325904130935669, "learning_rate": 7.56137805799736e-05, "loss": 0.0074, "step": 28990 }, { "grad_norm": 0.1355864405632019, "learning_rate": 7.559602795734439e-05, "loss": 0.0079, "step": 29000 }, { "grad_norm": 0.07929284125566483, "learning_rate": 7.557827096105408e-05, "loss": 0.0073, "step": 29010 }, { "grad_norm": 0.07461497187614441, "learning_rate": 7.55605095941368e-05, "loss": 0.0075, "step": 29020 }, { "grad_norm": 0.13597078621387482, "learning_rate": 7.554274385962752e-05, "loss": 0.0063, "step": 29030 }, { "grad_norm": 0.1599554866552353, "learning_rate": 7.552497376056191e-05, "loss": 0.0093, "step": 29040 }, { "grad_norm": 0.0971389189362526, "learning_rate": 7.550719929997639e-05, "loss": 0.005, "step": 29050 }, { "grad_norm": 0.11639314144849777, "learning_rate": 7.548942048090813e-05, "loss": 0.007, "step": 29060 }, { "grad_norm": 0.14852796494960785, "learning_rate": 7.547163730639506e-05, "loss": 0.0088, "step": 29070 }, { "grad_norm": 0.08567163348197937, "learning_rate": 7.545384977947583e-05, "loss": 0.0069, "step": 29080 }, { "grad_norm": 0.13834863901138306, "learning_rate": 7.543605790318981e-05, "loss": 0.0076, "step": 29090 }, { "grad_norm": 0.17695015668869019, "learning_rate": 7.54182616805772e-05, "loss": 0.0077, "step": 29100 }, { "grad_norm": 0.13735145330429077, "learning_rate": 7.540046111467885e-05, "loss": 0.0064, "step": 29110 }, { "grad_norm": 0.15313728153705597, "learning_rate": 7.53826562085364e-05, "loss": 0.0077, "step": 29120 }, { "grad_norm": 0.10959731787443161, "learning_rate": 7.536484696519221e-05, "loss": 0.0074, "step": 29130 }, { "grad_norm": 0.10123907029628754, "learning_rate": 7.534703338768942e-05, "loss": 0.0084, "step": 29140 }, { "grad_norm": 0.17278282344341278, "learning_rate": 7.532921547907185e-05, "loss": 0.0079, "step": 29150 }, { "grad_norm": 0.16024890542030334, "learning_rate": 7.531139324238412e-05, "loss": 0.0081, "step": 29160 }, { "grad_norm": 0.15615427494049072, "learning_rate": 7.529356668067157e-05, "loss": 0.0073, "step": 29170 }, { "grad_norm": 0.1455867439508438, "learning_rate": 7.527573579698023e-05, "loss": 0.0069, "step": 29180 }, { "grad_norm": 0.0931829884648323, "learning_rate": 7.525790059435693e-05, "loss": 0.0063, "step": 29190 }, { "grad_norm": 0.10334453731775284, "learning_rate": 7.524006107584926e-05, "loss": 0.0055, "step": 29200 }, { "grad_norm": 0.10834261029958725, "learning_rate": 7.522221724450544e-05, "loss": 0.0057, "step": 29210 }, { "grad_norm": 0.1443396359682083, "learning_rate": 7.520436910337451e-05, "loss": 0.007, "step": 29220 }, { "grad_norm": 0.1928669810295105, "learning_rate": 7.518651665550627e-05, "loss": 0.0064, "step": 29230 }, { "grad_norm": 0.17017164826393127, "learning_rate": 7.516865990395117e-05, "loss": 0.0091, "step": 29240 }, { "grad_norm": 0.11957038193941116, "learning_rate": 7.515079885176047e-05, "loss": 0.0081, "step": 29250 }, { "grad_norm": 0.12275753170251846, "learning_rate": 7.513293350198612e-05, "loss": 0.0059, "step": 29260 }, { "grad_norm": 0.08397990465164185, "learning_rate": 7.511506385768081e-05, "loss": 0.0065, "step": 29270 }, { "grad_norm": 0.11484047025442123, "learning_rate": 7.509718992189801e-05, "loss": 0.0068, "step": 29280 }, { "grad_norm": 0.1445220708847046, "learning_rate": 7.507931169769182e-05, "loss": 0.0079, "step": 29290 }, { "grad_norm": 0.1197635754942894, "learning_rate": 7.506142918811722e-05, "loss": 0.009, "step": 29300 }, { "grad_norm": 0.15235157310962677, "learning_rate": 7.504354239622978e-05, "loss": 0.0063, "step": 29310 }, { "grad_norm": 0.13291731476783752, "learning_rate": 7.50256513250859e-05, "loss": 0.0074, "step": 29320 }, { "grad_norm": 0.13532754778862, "learning_rate": 7.500775597774265e-05, "loss": 0.0084, "step": 29330 }, { "grad_norm": 0.08697265386581421, "learning_rate": 7.498985635725788e-05, "loss": 0.0067, "step": 29340 }, { "grad_norm": 0.14196301996707916, "learning_rate": 7.497195246669012e-05, "loss": 0.0076, "step": 29350 }, { "grad_norm": 0.14909856021404266, "learning_rate": 7.495404430909868e-05, "loss": 0.008, "step": 29360 }, { "grad_norm": 0.14121761918067932, "learning_rate": 7.493613188754356e-05, "loss": 0.0077, "step": 29370 }, { "grad_norm": 0.12582723796367645, "learning_rate": 7.49182152050855e-05, "loss": 0.0068, "step": 29380 }, { "grad_norm": 0.12198494374752045, "learning_rate": 7.490029426478598e-05, "loss": 0.0065, "step": 29390 }, { "grad_norm": 0.16013400256633759, "learning_rate": 7.488236906970719e-05, "loss": 0.0073, "step": 29400 }, { "grad_norm": 0.1429757922887802, "learning_rate": 7.486443962291207e-05, "loss": 0.0056, "step": 29410 }, { "grad_norm": 0.13523362576961517, "learning_rate": 7.484650592746424e-05, "loss": 0.0056, "step": 29420 }, { "grad_norm": 0.10900384932756424, "learning_rate": 7.482856798642811e-05, "loss": 0.0075, "step": 29430 }, { "grad_norm": 0.12355779856443405, "learning_rate": 7.481062580286878e-05, "loss": 0.0067, "step": 29440 }, { "grad_norm": 0.10479866713285446, "learning_rate": 7.479267937985208e-05, "loss": 0.0067, "step": 29450 }, { "grad_norm": 0.11514369398355484, "learning_rate": 7.477472872044456e-05, "loss": 0.0065, "step": 29460 }, { "grad_norm": 0.11845243722200394, "learning_rate": 7.475677382771347e-05, "loss": 0.0054, "step": 29470 }, { "grad_norm": 0.1723024547100067, "learning_rate": 7.473881470472683e-05, "loss": 0.0079, "step": 29480 }, { "grad_norm": 0.10619061440229416, "learning_rate": 7.472085135455337e-05, "loss": 0.0061, "step": 29490 }, { "grad_norm": 0.17496991157531738, "learning_rate": 7.470288378026256e-05, "loss": 0.0078, "step": 29500 }, { "grad_norm": 0.09516683220863342, "learning_rate": 7.468491198492451e-05, "loss": 0.0081, "step": 29510 }, { "grad_norm": 0.17357979714870453, "learning_rate": 7.466693597161013e-05, "loss": 0.0075, "step": 29520 }, { "grad_norm": 0.1878298819065094, "learning_rate": 7.464895574339104e-05, "loss": 0.0061, "step": 29530 }, { "grad_norm": 0.14750629663467407, "learning_rate": 7.463097130333958e-05, "loss": 0.0094, "step": 29540 }, { "grad_norm": 0.10526272654533386, "learning_rate": 7.461298265452876e-05, "loss": 0.0064, "step": 29550 }, { "grad_norm": 0.1750611960887909, "learning_rate": 7.459498980003239e-05, "loss": 0.0083, "step": 29560 }, { "grad_norm": 0.14114190638065338, "learning_rate": 7.457699274292493e-05, "loss": 0.0067, "step": 29570 }, { "grad_norm": 0.1100429967045784, "learning_rate": 7.455899148628159e-05, "loss": 0.0064, "step": 29580 }, { "grad_norm": 0.11477026343345642, "learning_rate": 7.45409860331783e-05, "loss": 0.0067, "step": 29590 }, { "grad_norm": 0.21544605493545532, "learning_rate": 7.452297638669169e-05, "loss": 0.0093, "step": 29600 }, { "grad_norm": 0.19133901596069336, "learning_rate": 7.450496254989911e-05, "loss": 0.0093, "step": 29610 }, { "grad_norm": 0.1116044893860817, "learning_rate": 7.448694452587866e-05, "loss": 0.0076, "step": 29620 }, { "grad_norm": 0.1372603476047516, "learning_rate": 7.44689223177091e-05, "loss": 0.0057, "step": 29630 }, { "grad_norm": 0.14693982899188995, "learning_rate": 7.445089592846994e-05, "loss": 0.0062, "step": 29640 }, { "grad_norm": 0.07307519018650055, "learning_rate": 7.443286536124141e-05, "loss": 0.0059, "step": 29650 }, { "grad_norm": 0.10133342444896698, "learning_rate": 7.441483061910443e-05, "loss": 0.0061, "step": 29660 }, { "grad_norm": 0.11339282244443893, "learning_rate": 7.439679170514064e-05, "loss": 0.0058, "step": 29670 }, { "grad_norm": 0.14308054745197296, "learning_rate": 7.43787486224324e-05, "loss": 0.0056, "step": 29680 }, { "grad_norm": 0.12733028829097748, "learning_rate": 7.436070137406276e-05, "loss": 0.0062, "step": 29690 }, { "grad_norm": 0.13780421018600464, "learning_rate": 7.434264996311556e-05, "loss": 0.0069, "step": 29700 }, { "grad_norm": 0.13153068721294403, "learning_rate": 7.432459439267525e-05, "loss": 0.0071, "step": 29710 }, { "grad_norm": 0.1257350891828537, "learning_rate": 7.430653466582701e-05, "loss": 0.0062, "step": 29720 }, { "grad_norm": 0.10659483820199966, "learning_rate": 7.42884707856568e-05, "loss": 0.0087, "step": 29730 }, { "grad_norm": 0.11511796712875366, "learning_rate": 7.427040275525122e-05, "loss": 0.0054, "step": 29740 }, { "grad_norm": 0.11279997229576111, "learning_rate": 7.42523305776976e-05, "loss": 0.0072, "step": 29750 }, { "grad_norm": 0.10163195431232452, "learning_rate": 7.4234254256084e-05, "loss": 0.0066, "step": 29760 }, { "grad_norm": 0.16382940113544464, "learning_rate": 7.421617379349915e-05, "loss": 0.0084, "step": 29770 }, { "grad_norm": 0.11742410063743591, "learning_rate": 7.41980891930325e-05, "loss": 0.0065, "step": 29780 }, { "grad_norm": 0.10771370679140091, "learning_rate": 7.418000045777425e-05, "loss": 0.0071, "step": 29790 }, { "grad_norm": 0.17280972003936768, "learning_rate": 7.416190759081523e-05, "loss": 0.0096, "step": 29800 }, { "grad_norm": 0.11671552807092667, "learning_rate": 7.414381059524704e-05, "loss": 0.0055, "step": 29810 }, { "grad_norm": 0.09889433532953262, "learning_rate": 7.412570947416195e-05, "loss": 0.006, "step": 29820 }, { "grad_norm": 0.10408830642700195, "learning_rate": 7.410760423065295e-05, "loss": 0.0069, "step": 29830 }, { "grad_norm": 0.15460973978042603, "learning_rate": 7.408949486781372e-05, "loss": 0.007, "step": 29840 }, { "grad_norm": 0.18035301566123962, "learning_rate": 7.407138138873868e-05, "loss": 0.0084, "step": 29850 }, { "grad_norm": 0.13741302490234375, "learning_rate": 7.405326379652292e-05, "loss": 0.0084, "step": 29860 }, { "grad_norm": 0.2001935988664627, "learning_rate": 7.403514209426222e-05, "loss": 0.0074, "step": 29870 }, { "grad_norm": 0.12817658483982086, "learning_rate": 7.40170162850531e-05, "loss": 0.0085, "step": 29880 }, { "grad_norm": 0.15220984816551208, "learning_rate": 7.399888637199278e-05, "loss": 0.0102, "step": 29890 }, { "grad_norm": 0.14730770885944366, "learning_rate": 7.398075235817914e-05, "loss": 0.0065, "step": 29900 }, { "grad_norm": 0.15778638422489166, "learning_rate": 7.39626142467108e-05, "loss": 0.008, "step": 29910 }, { "grad_norm": 0.1616727113723755, "learning_rate": 7.394447204068706e-05, "loss": 0.007, "step": 29920 }, { "grad_norm": 0.16001898050308228, "learning_rate": 7.392632574320793e-05, "loss": 0.0091, "step": 29930 }, { "grad_norm": 0.1053328812122345, "learning_rate": 7.390817535737411e-05, "loss": 0.008, "step": 29940 }, { "grad_norm": 0.15874022245407104, "learning_rate": 7.389002088628703e-05, "loss": 0.0081, "step": 29950 }, { "grad_norm": 0.14371882379055023, "learning_rate": 7.387186233304877e-05, "loss": 0.0067, "step": 29960 }, { "grad_norm": 0.158756822347641, "learning_rate": 7.385369970076212e-05, "loss": 0.0062, "step": 29970 }, { "grad_norm": 0.11476915329694748, "learning_rate": 7.38355329925306e-05, "loss": 0.0053, "step": 29980 }, { "grad_norm": 0.1347552239894867, "learning_rate": 7.381736221145838e-05, "loss": 0.0062, "step": 29990 }, { "grad_norm": 0.09614349901676178, "learning_rate": 7.37991873606504e-05, "loss": 0.0068, "step": 30000 }, { "grad_norm": 0.1555899977684021, "learning_rate": 7.378100844321218e-05, "loss": 0.0061, "step": 30010 }, { "grad_norm": 0.1356513798236847, "learning_rate": 7.376282546225004e-05, "loss": 0.0092, "step": 30020 }, { "grad_norm": 0.09808346629142761, "learning_rate": 7.374463842087094e-05, "loss": 0.0073, "step": 30030 }, { "grad_norm": 0.09105196595191956, "learning_rate": 7.372644732218254e-05, "loss": 0.0056, "step": 30040 }, { "grad_norm": 0.12267874926328659, "learning_rate": 7.370825216929322e-05, "loss": 0.008, "step": 30050 }, { "grad_norm": 0.17066259682178497, "learning_rate": 7.369005296531205e-05, "loss": 0.0085, "step": 30060 }, { "grad_norm": 0.11245168745517731, "learning_rate": 7.367184971334873e-05, "loss": 0.0063, "step": 30070 }, { "grad_norm": 0.1304076611995697, "learning_rate": 7.365364241651371e-05, "loss": 0.0062, "step": 30080 }, { "grad_norm": 0.10783424973487854, "learning_rate": 7.363543107791815e-05, "loss": 0.0087, "step": 30090 }, { "grad_norm": 0.14573414623737335, "learning_rate": 7.361721570067384e-05, "loss": 0.0079, "step": 30100 }, { "grad_norm": 0.15246300399303436, "learning_rate": 7.359899628789331e-05, "loss": 0.0077, "step": 30110 }, { "grad_norm": 0.12424444407224655, "learning_rate": 7.358077284268974e-05, "loss": 0.0073, "step": 30120 }, { "grad_norm": 0.14157171547412872, "learning_rate": 7.356254536817702e-05, "loss": 0.0066, "step": 30130 }, { "grad_norm": 0.12023086100816727, "learning_rate": 7.354431386746973e-05, "loss": 0.0107, "step": 30140 }, { "grad_norm": 0.18836714327335358, "learning_rate": 7.352607834368316e-05, "loss": 0.0071, "step": 30150 }, { "grad_norm": 0.12917064130306244, "learning_rate": 7.350783879993324e-05, "loss": 0.0085, "step": 30160 }, { "grad_norm": 0.14641229808330536, "learning_rate": 7.348959523933658e-05, "loss": 0.0094, "step": 30170 }, { "grad_norm": 0.10532377660274506, "learning_rate": 7.347134766501057e-05, "loss": 0.0067, "step": 30180 }, { "grad_norm": 0.10171256214380264, "learning_rate": 7.345309608007315e-05, "loss": 0.0064, "step": 30190 }, { "grad_norm": 0.1261376440525055, "learning_rate": 7.343484048764308e-05, "loss": 0.0074, "step": 30200 }, { "grad_norm": 0.1347823441028595, "learning_rate": 7.341658089083972e-05, "loss": 0.0076, "step": 30210 }, { "grad_norm": 0.21467752754688263, "learning_rate": 7.339831729278313e-05, "loss": 0.0088, "step": 30220 }, { "grad_norm": 0.14448311924934387, "learning_rate": 7.338004969659404e-05, "loss": 0.0061, "step": 30230 }, { "grad_norm": 0.12601135671138763, "learning_rate": 7.336177810539391e-05, "loss": 0.0075, "step": 30240 }, { "grad_norm": 0.1830400824546814, "learning_rate": 7.334350252230485e-05, "loss": 0.0098, "step": 30250 }, { "grad_norm": 0.16374754905700684, "learning_rate": 7.332522295044965e-05, "loss": 0.0063, "step": 30260 }, { "grad_norm": 0.14562390744686127, "learning_rate": 7.33069393929518e-05, "loss": 0.0077, "step": 30270 }, { "grad_norm": 0.16640710830688477, "learning_rate": 7.328865185293545e-05, "loss": 0.0099, "step": 30280 }, { "grad_norm": 0.11976898461580276, "learning_rate": 7.327036033352546e-05, "loss": 0.0068, "step": 30290 }, { "grad_norm": 0.14318183064460754, "learning_rate": 7.325206483784733e-05, "loss": 0.0076, "step": 30300 }, { "grad_norm": 0.10241397470235825, "learning_rate": 7.323376536902724e-05, "loss": 0.0056, "step": 30310 }, { "grad_norm": 0.11885075271129608, "learning_rate": 7.321546193019213e-05, "loss": 0.0092, "step": 30320 }, { "grad_norm": 0.13016019761562347, "learning_rate": 7.31971545244695e-05, "loss": 0.0076, "step": 30330 }, { "grad_norm": 0.12933698296546936, "learning_rate": 7.31788431549876e-05, "loss": 0.0081, "step": 30340 }, { "grad_norm": 0.1528911590576172, "learning_rate": 7.316052782487534e-05, "loss": 0.0071, "step": 30350 }, { "grad_norm": 0.1719207912683487, "learning_rate": 7.314220853726234e-05, "loss": 0.008, "step": 30360 }, { "grad_norm": 0.11595571786165237, "learning_rate": 7.312388529527884e-05, "loss": 0.0081, "step": 30370 }, { "grad_norm": 0.1028403788805008, "learning_rate": 7.310555810205577e-05, "loss": 0.0062, "step": 30380 }, { "grad_norm": 0.12817135453224182, "learning_rate": 7.308722696072476e-05, "loss": 0.0069, "step": 30390 }, { "grad_norm": 0.10504058003425598, "learning_rate": 7.306889187441811e-05, "loss": 0.0058, "step": 30400 }, { "grad_norm": 0.09736049920320511, "learning_rate": 7.305055284626876e-05, "loss": 0.0063, "step": 30410 }, { "grad_norm": 0.14430348575115204, "learning_rate": 7.303220987941037e-05, "loss": 0.0072, "step": 30420 }, { "grad_norm": 0.10400990396738052, "learning_rate": 7.301386297697726e-05, "loss": 0.0042, "step": 30430 }, { "grad_norm": 0.12974299490451813, "learning_rate": 7.299551214210438e-05, "loss": 0.0051, "step": 30440 }, { "grad_norm": 0.1470595896244049, "learning_rate": 7.297715737792738e-05, "loss": 0.0059, "step": 30450 }, { "grad_norm": 0.1264185607433319, "learning_rate": 7.295879868758265e-05, "loss": 0.0061, "step": 30460 }, { "grad_norm": 0.15168413519859314, "learning_rate": 7.294043607420713e-05, "loss": 0.0062, "step": 30470 }, { "grad_norm": 0.11681348830461502, "learning_rate": 7.292206954093852e-05, "loss": 0.005, "step": 30480 }, { "grad_norm": 0.17322777211666107, "learning_rate": 7.290369909091515e-05, "loss": 0.0065, "step": 30490 }, { "grad_norm": 0.11681608110666275, "learning_rate": 7.2885324727276e-05, "loss": 0.0064, "step": 30500 }, { "grad_norm": 0.11262710392475128, "learning_rate": 7.286694645316076e-05, "loss": 0.0064, "step": 30510 }, { "grad_norm": 0.10140305757522583, "learning_rate": 7.284856427170982e-05, "loss": 0.0066, "step": 30520 }, { "grad_norm": 0.11601293087005615, "learning_rate": 7.283017818606414e-05, "loss": 0.0076, "step": 30530 }, { "grad_norm": 0.09992527961730957, "learning_rate": 7.28117881993654e-05, "loss": 0.0062, "step": 30540 }, { "grad_norm": 0.11907953768968582, "learning_rate": 7.279339431475598e-05, "loss": 0.0059, "step": 30550 }, { "grad_norm": 0.11732550710439682, "learning_rate": 7.277499653537887e-05, "loss": 0.0074, "step": 30560 }, { "grad_norm": 0.11317678540945053, "learning_rate": 7.275659486437776e-05, "loss": 0.0103, "step": 30570 }, { "grad_norm": 0.1223653107881546, "learning_rate": 7.273818930489695e-05, "loss": 0.0072, "step": 30580 }, { "grad_norm": 0.1237744390964508, "learning_rate": 7.271977986008151e-05, "loss": 0.0058, "step": 30590 }, { "grad_norm": 0.1170642301440239, "learning_rate": 7.270136653307705e-05, "loss": 0.0063, "step": 30600 }, { "grad_norm": 0.12309852987527847, "learning_rate": 7.268294932702994e-05, "loss": 0.0071, "step": 30610 }, { "grad_norm": 0.10829942673444748, "learning_rate": 7.266452824508719e-05, "loss": 0.0051, "step": 30620 }, { "grad_norm": 0.08929672092199326, "learning_rate": 7.264610329039643e-05, "loss": 0.0073, "step": 30630 }, { "grad_norm": 0.12147387117147446, "learning_rate": 7.262767446610599e-05, "loss": 0.0069, "step": 30640 }, { "grad_norm": 0.10848025977611542, "learning_rate": 7.260924177536485e-05, "loss": 0.006, "step": 30650 }, { "grad_norm": 0.14068840444087982, "learning_rate": 7.259080522132265e-05, "loss": 0.0078, "step": 30660 }, { "grad_norm": 0.09910403192043304, "learning_rate": 7.257236480712972e-05, "loss": 0.0061, "step": 30670 }, { "grad_norm": 0.1454652100801468, "learning_rate": 7.255392053593697e-05, "loss": 0.0062, "step": 30680 }, { "grad_norm": 0.11914924532175064, "learning_rate": 7.253547241089607e-05, "loss": 0.0076, "step": 30690 }, { "grad_norm": 0.1331644058227539, "learning_rate": 7.251702043515927e-05, "loss": 0.0061, "step": 30700 }, { "grad_norm": 0.1493767946958542, "learning_rate": 7.249856461187952e-05, "loss": 0.007, "step": 30710 }, { "grad_norm": 0.09980825334787369, "learning_rate": 7.248010494421042e-05, "loss": 0.0082, "step": 30720 }, { "grad_norm": 0.1775798350572586, "learning_rate": 7.246164143530622e-05, "loss": 0.0082, "step": 30730 }, { "grad_norm": 0.15466339886188507, "learning_rate": 7.244317408832181e-05, "loss": 0.0084, "step": 30740 }, { "grad_norm": 0.12360034883022308, "learning_rate": 7.242470290641279e-05, "loss": 0.0056, "step": 30750 }, { "grad_norm": 0.12497369199991226, "learning_rate": 7.240622789273536e-05, "loss": 0.0075, "step": 30760 }, { "grad_norm": 0.14407488703727722, "learning_rate": 7.238774905044638e-05, "loss": 0.0066, "step": 30770 }, { "grad_norm": 0.13044719398021698, "learning_rate": 7.236926638270341e-05, "loss": 0.0072, "step": 30780 }, { "grad_norm": 0.15271514654159546, "learning_rate": 7.23507798926646e-05, "loss": 0.0053, "step": 30790 }, { "grad_norm": 0.17244376242160797, "learning_rate": 7.23322895834888e-05, "loss": 0.0078, "step": 30800 }, { "grad_norm": 0.1209932342171669, "learning_rate": 7.231379545833552e-05, "loss": 0.0075, "step": 30810 }, { "grad_norm": 0.1570338010787964, "learning_rate": 7.229529752036487e-05, "loss": 0.0075, "step": 30820 }, { "grad_norm": 0.11220324784517288, "learning_rate": 7.227679577273765e-05, "loss": 0.0083, "step": 30830 }, { "grad_norm": 0.17767857015132904, "learning_rate": 7.225829021861529e-05, "loss": 0.0061, "step": 30840 }, { "grad_norm": 0.11394143849611282, "learning_rate": 7.223978086115992e-05, "loss": 0.0055, "step": 30850 }, { "grad_norm": 0.14104920625686646, "learning_rate": 7.222126770353425e-05, "loss": 0.0075, "step": 30860 }, { "grad_norm": 0.1342756152153015, "learning_rate": 7.22027507489017e-05, "loss": 0.0095, "step": 30870 }, { "grad_norm": 0.13003288209438324, "learning_rate": 7.218423000042627e-05, "loss": 0.0071, "step": 30880 }, { "grad_norm": 0.15155108273029327, "learning_rate": 7.216570546127268e-05, "loss": 0.0065, "step": 30890 }, { "grad_norm": 0.13680069148540497, "learning_rate": 7.214717713460626e-05, "loss": 0.0067, "step": 30900 }, { "grad_norm": 0.1628284752368927, "learning_rate": 7.2128645023593e-05, "loss": 0.0063, "step": 30910 }, { "grad_norm": 0.1403445452451706, "learning_rate": 7.211010913139951e-05, "loss": 0.0072, "step": 30920 }, { "grad_norm": 0.10389120131731033, "learning_rate": 7.209156946119308e-05, "loss": 0.0065, "step": 30930 }, { "grad_norm": 0.11602384597063065, "learning_rate": 7.207302601614166e-05, "loss": 0.0054, "step": 30940 }, { "grad_norm": 0.10973040014505386, "learning_rate": 7.205447879941378e-05, "loss": 0.0058, "step": 30950 }, { "grad_norm": 0.1345142126083374, "learning_rate": 7.203592781417866e-05, "loss": 0.007, "step": 30960 }, { "grad_norm": 0.13722531497478485, "learning_rate": 7.201737306360617e-05, "loss": 0.0063, "step": 30970 }, { "grad_norm": 0.07742427289485931, "learning_rate": 7.19988145508668e-05, "loss": 0.0063, "step": 30980 }, { "grad_norm": 0.14123636484146118, "learning_rate": 7.198025227913168e-05, "loss": 0.0089, "step": 30990 }, { "grad_norm": 0.14757521450519562, "learning_rate": 7.196168625157261e-05, "loss": 0.0056, "step": 31000 }, { "grad_norm": 0.10986930876970291, "learning_rate": 7.194311647136201e-05, "loss": 0.0072, "step": 31010 }, { "grad_norm": 0.12481585144996643, "learning_rate": 7.192454294167297e-05, "loss": 0.0071, "step": 31020 }, { "grad_norm": 0.08359262347221375, "learning_rate": 7.190596566567917e-05, "loss": 0.0055, "step": 31030 }, { "grad_norm": 0.1425025910139084, "learning_rate": 7.188738464655496e-05, "loss": 0.0083, "step": 31040 }, { "grad_norm": 0.12569282948970795, "learning_rate": 7.186879988747533e-05, "loss": 0.0057, "step": 31050 }, { "grad_norm": 0.12239884585142136, "learning_rate": 7.185021139161592e-05, "loss": 0.0055, "step": 31060 }, { "grad_norm": 0.09196406602859497, "learning_rate": 7.1831619162153e-05, "loss": 0.0064, "step": 31070 }, { "grad_norm": 0.13729557394981384, "learning_rate": 7.181302320226345e-05, "loss": 0.0069, "step": 31080 }, { "grad_norm": 0.11369138956069946, "learning_rate": 7.179442351512482e-05, "loss": 0.0057, "step": 31090 }, { "grad_norm": 0.15389439463615417, "learning_rate": 7.177582010391528e-05, "loss": 0.0094, "step": 31100 }, { "grad_norm": 0.13939452171325684, "learning_rate": 7.175721297181366e-05, "loss": 0.0061, "step": 31110 }, { "grad_norm": 0.1585662066936493, "learning_rate": 7.173860212199942e-05, "loss": 0.0077, "step": 31120 }, { "grad_norm": 0.11208760738372803, "learning_rate": 7.171998755765263e-05, "loss": 0.0057, "step": 31130 }, { "grad_norm": 0.1344427615404129, "learning_rate": 7.170136928195398e-05, "loss": 0.0068, "step": 31140 }, { "grad_norm": 0.1692243069410324, "learning_rate": 7.168274729808489e-05, "loss": 0.0064, "step": 31150 }, { "grad_norm": 0.12502823770046234, "learning_rate": 7.166412160922728e-05, "loss": 0.0079, "step": 31160 }, { "grad_norm": 0.13338792324066162, "learning_rate": 7.164549221856382e-05, "loss": 0.0071, "step": 31170 }, { "grad_norm": 0.13666480779647827, "learning_rate": 7.162685912927775e-05, "loss": 0.0069, "step": 31180 }, { "grad_norm": 0.1130950003862381, "learning_rate": 7.160822234455294e-05, "loss": 0.0065, "step": 31190 }, { "grad_norm": 0.16529004275798798, "learning_rate": 7.158958186757391e-05, "loss": 0.0061, "step": 31200 }, { "grad_norm": 0.14825168251991272, "learning_rate": 7.157093770152582e-05, "loss": 0.0076, "step": 31210 }, { "grad_norm": 0.15572699904441833, "learning_rate": 7.155228984959446e-05, "loss": 0.008, "step": 31220 }, { "grad_norm": 0.13581588864326477, "learning_rate": 7.153363831496621e-05, "loss": 0.0072, "step": 31230 }, { "grad_norm": 0.1408265382051468, "learning_rate": 7.151498310082811e-05, "loss": 0.0097, "step": 31240 }, { "grad_norm": 0.13464978337287903, "learning_rate": 7.149632421036784e-05, "loss": 0.0074, "step": 31250 }, { "grad_norm": 0.11289162933826447, "learning_rate": 7.147766164677369e-05, "loss": 0.0061, "step": 31260 }, { "grad_norm": 0.12314003705978394, "learning_rate": 7.145899541323459e-05, "loss": 0.0082, "step": 31270 }, { "grad_norm": 0.10766205191612244, "learning_rate": 7.144032551294007e-05, "loss": 0.009, "step": 31280 }, { "grad_norm": 0.14176416397094727, "learning_rate": 7.14216519490803e-05, "loss": 0.0081, "step": 31290 }, { "grad_norm": 0.16542577743530273, "learning_rate": 7.140297472484609e-05, "loss": 0.0092, "step": 31300 }, { "grad_norm": 0.11757177114486694, "learning_rate": 7.138429384342891e-05, "loss": 0.0054, "step": 31310 }, { "grad_norm": 0.13361772894859314, "learning_rate": 7.136560930802074e-05, "loss": 0.0063, "step": 31320 }, { "grad_norm": 0.09313588589429855, "learning_rate": 7.134692112181431e-05, "loss": 0.007, "step": 31330 }, { "grad_norm": 0.12163669615983963, "learning_rate": 7.13282292880029e-05, "loss": 0.0075, "step": 31340 }, { "grad_norm": 0.10218268632888794, "learning_rate": 7.130953380978043e-05, "loss": 0.0087, "step": 31350 }, { "grad_norm": 0.1308465301990509, "learning_rate": 7.129083469034144e-05, "loss": 0.0081, "step": 31360 }, { "grad_norm": 0.12505730986595154, "learning_rate": 7.127213193288112e-05, "loss": 0.0083, "step": 31370 }, { "grad_norm": 0.11012396961450577, "learning_rate": 7.125342554059522e-05, "loss": 0.0074, "step": 31380 }, { "grad_norm": 0.10603181272745132, "learning_rate": 7.12347155166802e-05, "loss": 0.0097, "step": 31390 }, { "grad_norm": 0.11219552904367447, "learning_rate": 7.121600186433306e-05, "loss": 0.007, "step": 31400 }, { "grad_norm": 0.17084912955760956, "learning_rate": 7.119728458675148e-05, "loss": 0.0073, "step": 31410 }, { "grad_norm": 0.12975731492042542, "learning_rate": 7.117856368713369e-05, "loss": 0.0089, "step": 31420 }, { "grad_norm": 0.13410170376300812, "learning_rate": 7.115983916867861e-05, "loss": 0.0069, "step": 31430 }, { "grad_norm": 0.14712177217006683, "learning_rate": 7.114111103458574e-05, "loss": 0.0066, "step": 31440 }, { "grad_norm": 0.11846580356359482, "learning_rate": 7.11223792880552e-05, "loss": 0.0067, "step": 31450 }, { "grad_norm": 0.13175082206726074, "learning_rate": 7.110364393228773e-05, "loss": 0.0081, "step": 31460 }, { "grad_norm": 0.09937341511249542, "learning_rate": 7.108490497048471e-05, "loss": 0.0065, "step": 31470 }, { "grad_norm": 0.12328564375638962, "learning_rate": 7.10661624058481e-05, "loss": 0.0068, "step": 31480 }, { "grad_norm": 0.1458781212568283, "learning_rate": 7.10474162415805e-05, "loss": 0.007, "step": 31490 }, { "grad_norm": 0.14410816133022308, "learning_rate": 7.102866648088511e-05, "loss": 0.0074, "step": 31500 }, { "grad_norm": 0.1316569596529007, "learning_rate": 7.100991312696576e-05, "loss": 0.0059, "step": 31510 }, { "grad_norm": 0.1439450979232788, "learning_rate": 7.099115618302686e-05, "loss": 0.0059, "step": 31520 }, { "grad_norm": 0.15857566893100739, "learning_rate": 7.097239565227349e-05, "loss": 0.0063, "step": 31530 }, { "grad_norm": 0.10871953517198563, "learning_rate": 7.09536315379113e-05, "loss": 0.0076, "step": 31540 }, { "grad_norm": 0.1386423408985138, "learning_rate": 7.093486384314656e-05, "loss": 0.0107, "step": 31550 }, { "grad_norm": 0.11415529996156693, "learning_rate": 7.091609257118616e-05, "loss": 0.0064, "step": 31560 }, { "grad_norm": 0.10508298873901367, "learning_rate": 7.08973177252376e-05, "loss": 0.005, "step": 31570 }, { "grad_norm": 0.10988558083772659, "learning_rate": 7.087853930850898e-05, "loss": 0.0055, "step": 31580 }, { "grad_norm": 0.15136708319187164, "learning_rate": 7.085975732420903e-05, "loss": 0.0064, "step": 31590 }, { "grad_norm": 0.11302974820137024, "learning_rate": 7.084097177554706e-05, "loss": 0.0076, "step": 31600 }, { "grad_norm": 0.12660805881023407, "learning_rate": 7.082218266573301e-05, "loss": 0.0057, "step": 31610 }, { "grad_norm": 0.19849136471748352, "learning_rate": 7.080338999797743e-05, "loss": 0.0096, "step": 31620 }, { "grad_norm": 0.11216739565134048, "learning_rate": 7.07845937754915e-05, "loss": 0.0052, "step": 31630 }, { "grad_norm": 0.15054433047771454, "learning_rate": 7.076579400148693e-05, "loss": 0.0073, "step": 31640 }, { "grad_norm": 0.15027685463428497, "learning_rate": 7.074699067917611e-05, "loss": 0.0089, "step": 31650 }, { "grad_norm": 0.15492865443229675, "learning_rate": 7.072818381177201e-05, "loss": 0.0074, "step": 31660 }, { "grad_norm": 0.11231452971696854, "learning_rate": 7.070937340248823e-05, "loss": 0.0077, "step": 31670 }, { "grad_norm": 0.103382408618927, "learning_rate": 7.069055945453893e-05, "loss": 0.0064, "step": 31680 }, { "grad_norm": 0.09719283878803253, "learning_rate": 7.067174197113892e-05, "loss": 0.0063, "step": 31690 }, { "grad_norm": 0.1414857655763626, "learning_rate": 7.065292095550355e-05, "loss": 0.007, "step": 31700 }, { "grad_norm": 0.14229634404182434, "learning_rate": 7.063409641084887e-05, "loss": 0.0052, "step": 31710 }, { "grad_norm": 0.13126401603221893, "learning_rate": 7.061526834039145e-05, "loss": 0.0059, "step": 31720 }, { "grad_norm": 0.1654602736234665, "learning_rate": 7.05964367473485e-05, "loss": 0.007, "step": 31730 }, { "grad_norm": 0.11859675496816635, "learning_rate": 7.057760163493783e-05, "loss": 0.0065, "step": 31740 }, { "grad_norm": 0.11623203009366989, "learning_rate": 7.055876300637783e-05, "loss": 0.0081, "step": 31750 }, { "grad_norm": 0.13670605421066284, "learning_rate": 7.053992086488753e-05, "loss": 0.0064, "step": 31760 }, { "grad_norm": 0.1433081030845642, "learning_rate": 7.052107521368651e-05, "loss": 0.0073, "step": 31770 }, { "grad_norm": 0.11080212146043777, "learning_rate": 7.0502226055995e-05, "loss": 0.0064, "step": 31780 }, { "grad_norm": 0.13930605351924896, "learning_rate": 7.048337339503379e-05, "loss": 0.0071, "step": 31790 }, { "grad_norm": 0.16834859549999237, "learning_rate": 7.046451723402427e-05, "loss": 0.0068, "step": 31800 }, { "grad_norm": 0.11644018441438675, "learning_rate": 7.044565757618848e-05, "loss": 0.0064, "step": 31810 }, { "grad_norm": 0.0908920168876648, "learning_rate": 7.042679442474899e-05, "loss": 0.0057, "step": 31820 }, { "grad_norm": 0.08247879892587662, "learning_rate": 7.040792778292902e-05, "loss": 0.0069, "step": 31830 }, { "grad_norm": 0.12683382630348206, "learning_rate": 7.038905765395234e-05, "loss": 0.0058, "step": 31840 }, { "grad_norm": 0.15725703537464142, "learning_rate": 7.037018404104334e-05, "loss": 0.007, "step": 31850 }, { "grad_norm": 0.1636931300163269, "learning_rate": 7.035130694742702e-05, "loss": 0.0063, "step": 31860 }, { "grad_norm": 0.15211121737957, "learning_rate": 7.033242637632897e-05, "loss": 0.0072, "step": 31870 }, { "grad_norm": 0.10432902723550797, "learning_rate": 7.031354233097534e-05, "loss": 0.0066, "step": 31880 }, { "grad_norm": 0.10950645059347153, "learning_rate": 7.029465481459289e-05, "loss": 0.0054, "step": 31890 }, { "grad_norm": 0.11318099498748779, "learning_rate": 7.027576383040898e-05, "loss": 0.0056, "step": 31900 }, { "grad_norm": 0.12332875281572342, "learning_rate": 7.025686938165159e-05, "loss": 0.0073, "step": 31910 }, { "grad_norm": 0.2010846883058548, "learning_rate": 7.023797147154924e-05, "loss": 0.0059, "step": 31920 }, { "grad_norm": 0.15108013153076172, "learning_rate": 7.021907010333111e-05, "loss": 0.0057, "step": 31930 }, { "grad_norm": 0.1468130350112915, "learning_rate": 7.020016528022685e-05, "loss": 0.0084, "step": 31940 }, { "grad_norm": 0.13617254793643951, "learning_rate": 7.018125700546683e-05, "loss": 0.0065, "step": 31950 }, { "grad_norm": 0.16009920835494995, "learning_rate": 7.016234528228196e-05, "loss": 0.0094, "step": 31960 }, { "grad_norm": 0.09845467656850815, "learning_rate": 7.014343011390372e-05, "loss": 0.0061, "step": 31970 }, { "grad_norm": 0.13166926801204681, "learning_rate": 7.01245115035642e-05, "loss": 0.0077, "step": 31980 }, { "grad_norm": 0.12193030118942261, "learning_rate": 7.010558945449606e-05, "loss": 0.0058, "step": 31990 }, { "grad_norm": 0.110274538397789, "learning_rate": 7.008666396993258e-05, "loss": 0.0068, "step": 32000 }, { "grad_norm": 0.14988300204277039, "learning_rate": 7.006773505310759e-05, "loss": 0.0081, "step": 32010 }, { "grad_norm": 0.12602531909942627, "learning_rate": 7.004880270725553e-05, "loss": 0.006, "step": 32020 }, { "grad_norm": 0.1074165552854538, "learning_rate": 7.002986693561144e-05, "loss": 0.0064, "step": 32030 }, { "grad_norm": 0.10382433235645294, "learning_rate": 7.001092774141089e-05, "loss": 0.005, "step": 32040 }, { "grad_norm": 0.1229141503572464, "learning_rate": 6.999198512789009e-05, "loss": 0.0052, "step": 32050 }, { "grad_norm": 0.12894493341445923, "learning_rate": 6.997303909828584e-05, "loss": 0.0075, "step": 32060 }, { "grad_norm": 0.1312023103237152, "learning_rate": 6.995408965583544e-05, "loss": 0.0049, "step": 32070 }, { "grad_norm": 0.13743674755096436, "learning_rate": 6.993513680377688e-05, "loss": 0.0076, "step": 32080 }, { "grad_norm": 0.11961134523153305, "learning_rate": 6.991618054534868e-05, "loss": 0.0073, "step": 32090 }, { "grad_norm": 0.12199226766824722, "learning_rate": 6.989722088378991e-05, "loss": 0.0059, "step": 32100 }, { "grad_norm": 0.11341144144535065, "learning_rate": 6.987825782234027e-05, "loss": 0.0054, "step": 32110 }, { "grad_norm": 0.12904159724712372, "learning_rate": 6.985929136424006e-05, "loss": 0.0062, "step": 32120 }, { "grad_norm": 0.13845859467983246, "learning_rate": 6.984032151273012e-05, "loss": 0.0061, "step": 32130 }, { "grad_norm": 0.08112654089927673, "learning_rate": 6.982134827105186e-05, "loss": 0.0072, "step": 32140 }, { "grad_norm": 0.11884342133998871, "learning_rate": 6.980237164244729e-05, "loss": 0.0081, "step": 32150 }, { "grad_norm": 0.09457279741764069, "learning_rate": 6.9783391630159e-05, "loss": 0.0059, "step": 32160 }, { "grad_norm": 0.1581314355134964, "learning_rate": 6.976440823743015e-05, "loss": 0.0065, "step": 32170 }, { "grad_norm": 0.21363703906536102, "learning_rate": 6.974542146750451e-05, "loss": 0.0071, "step": 32180 }, { "grad_norm": 0.09533154219388962, "learning_rate": 6.972643132362637e-05, "loss": 0.0055, "step": 32190 }, { "grad_norm": 0.13997261226177216, "learning_rate": 6.970743780904064e-05, "loss": 0.0061, "step": 32200 }, { "grad_norm": 0.1171097382903099, "learning_rate": 6.968844092699277e-05, "loss": 0.0101, "step": 32210 }, { "grad_norm": 0.1510801464319229, "learning_rate": 6.966944068072883e-05, "loss": 0.0075, "step": 32220 }, { "grad_norm": 0.09548208117485046, "learning_rate": 6.965043707349545e-05, "loss": 0.0067, "step": 32230 }, { "grad_norm": 0.170423224568367, "learning_rate": 6.963143010853982e-05, "loss": 0.0074, "step": 32240 }, { "grad_norm": 0.1530734896659851, "learning_rate": 6.961241978910971e-05, "loss": 0.0053, "step": 32250 }, { "grad_norm": 0.11722037941217422, "learning_rate": 6.959340611845344e-05, "loss": 0.008, "step": 32260 }, { "grad_norm": 0.09693169593811035, "learning_rate": 6.957438909981995e-05, "loss": 0.0065, "step": 32270 }, { "grad_norm": 0.14455614984035492, "learning_rate": 6.955536873645872e-05, "loss": 0.007, "step": 32280 }, { "grad_norm": 0.1483900547027588, "learning_rate": 6.953634503161982e-05, "loss": 0.0065, "step": 32290 }, { "grad_norm": 0.12781427800655365, "learning_rate": 6.951731798855387e-05, "loss": 0.006, "step": 32300 }, { "grad_norm": 0.14591097831726074, "learning_rate": 6.949828761051208e-05, "loss": 0.0057, "step": 32310 }, { "grad_norm": 0.10617135465145111, "learning_rate": 6.947925390074622e-05, "loss": 0.0044, "step": 32320 }, { "grad_norm": 0.08596891164779663, "learning_rate": 6.946021686250863e-05, "loss": 0.0065, "step": 32330 }, { "grad_norm": 0.10190331935882568, "learning_rate": 6.94411764990522e-05, "loss": 0.0059, "step": 32340 }, { "grad_norm": 0.08619502931833267, "learning_rate": 6.942213281363044e-05, "loss": 0.0055, "step": 32350 }, { "grad_norm": 0.12356659024953842, "learning_rate": 6.940308580949737e-05, "loss": 0.0082, "step": 32360 }, { "grad_norm": 0.16295650601387024, "learning_rate": 6.93840354899076e-05, "loss": 0.006, "step": 32370 }, { "grad_norm": 0.11136114597320557, "learning_rate": 6.936498185811633e-05, "loss": 0.0057, "step": 32380 }, { "grad_norm": 0.17765749990940094, "learning_rate": 6.93459249173793e-05, "loss": 0.0093, "step": 32390 }, { "grad_norm": 0.1740114688873291, "learning_rate": 6.932686467095279e-05, "loss": 0.0077, "step": 32400 }, { "grad_norm": 0.1714676171541214, "learning_rate": 6.930780112209373e-05, "loss": 0.0085, "step": 32410 }, { "grad_norm": 0.12473592162132263, "learning_rate": 6.92887342740595e-05, "loss": 0.007, "step": 32420 }, { "grad_norm": 0.15216630697250366, "learning_rate": 6.926966413010816e-05, "loss": 0.0059, "step": 32430 }, { "grad_norm": 0.11901868134737015, "learning_rate": 6.925059069349824e-05, "loss": 0.0062, "step": 32440 }, { "grad_norm": 0.14970643818378448, "learning_rate": 6.923151396748886e-05, "loss": 0.0064, "step": 32450 }, { "grad_norm": 0.11863505840301514, "learning_rate": 6.921243395533974e-05, "loss": 0.0068, "step": 32460 }, { "grad_norm": 0.12812843918800354, "learning_rate": 6.919335066031109e-05, "loss": 0.007, "step": 32470 }, { "grad_norm": 0.11829457432031631, "learning_rate": 6.917426408566379e-05, "loss": 0.007, "step": 32480 }, { "grad_norm": 0.11920687556266785, "learning_rate": 6.915517423465916e-05, "loss": 0.0059, "step": 32490 }, { "grad_norm": 0.14426469802856445, "learning_rate": 6.913608111055914e-05, "loss": 0.0051, "step": 32500 }, { "grad_norm": 0.11815864592790604, "learning_rate": 6.911698471662623e-05, "loss": 0.0056, "step": 32510 }, { "grad_norm": 0.1341223120689392, "learning_rate": 6.90978850561235e-05, "loss": 0.007, "step": 32520 }, { "grad_norm": 0.14955741167068481, "learning_rate": 6.907878213231454e-05, "loss": 0.0056, "step": 32530 }, { "grad_norm": 0.18823577463626862, "learning_rate": 6.90596759484635e-05, "loss": 0.0104, "step": 32540 }, { "grad_norm": 0.13600210845470428, "learning_rate": 6.904056650783514e-05, "loss": 0.006, "step": 32550 }, { "grad_norm": 0.10535318404436111, "learning_rate": 6.902145381369471e-05, "loss": 0.0055, "step": 32560 }, { "grad_norm": 0.11785245686769485, "learning_rate": 6.900233786930808e-05, "loss": 0.0065, "step": 32570 }, { "grad_norm": 0.13586099445819855, "learning_rate": 6.898321867794161e-05, "loss": 0.0073, "step": 32580 }, { "grad_norm": 0.11386747658252716, "learning_rate": 6.896409624286226e-05, "loss": 0.0079, "step": 32590 }, { "grad_norm": 0.12856806814670563, "learning_rate": 6.894497056733754e-05, "loss": 0.009, "step": 32600 }, { "grad_norm": 0.1567964106798172, "learning_rate": 6.89258416546355e-05, "loss": 0.0061, "step": 32610 }, { "grad_norm": 0.17350037395954132, "learning_rate": 6.890670950802474e-05, "loss": 0.0052, "step": 32620 }, { "grad_norm": 0.13114167749881744, "learning_rate": 6.88875741307744e-05, "loss": 0.0078, "step": 32630 }, { "grad_norm": 0.11144056171178818, "learning_rate": 6.886843552615425e-05, "loss": 0.0057, "step": 32640 }, { "grad_norm": 0.17046813666820526, "learning_rate": 6.884929369743451e-05, "loss": 0.0057, "step": 32650 }, { "grad_norm": 0.1581873595714569, "learning_rate": 6.8830148647886e-05, "loss": 0.0067, "step": 32660 }, { "grad_norm": 0.11566372215747833, "learning_rate": 6.88110003807801e-05, "loss": 0.0072, "step": 32670 }, { "grad_norm": 0.12055086344480515, "learning_rate": 6.87918488993887e-05, "loss": 0.0093, "step": 32680 }, { "grad_norm": 0.13020727038383484, "learning_rate": 6.877269420698431e-05, "loss": 0.0061, "step": 32690 }, { "grad_norm": 0.11866512149572372, "learning_rate": 6.875353630683989e-05, "loss": 0.0068, "step": 32700 }, { "grad_norm": 0.13174021244049072, "learning_rate": 6.873437520222905e-05, "loss": 0.005, "step": 32710 }, { "grad_norm": 0.14881274104118347, "learning_rate": 6.871521089642585e-05, "loss": 0.0053, "step": 32720 }, { "grad_norm": 0.11749391257762909, "learning_rate": 6.869604339270498e-05, "loss": 0.0088, "step": 32730 }, { "grad_norm": 0.09491220861673355, "learning_rate": 6.867687269434164e-05, "loss": 0.0048, "step": 32740 }, { "grad_norm": 0.12692517042160034, "learning_rate": 6.865769880461156e-05, "loss": 0.0069, "step": 32750 }, { "grad_norm": 0.1446305364370346, "learning_rate": 6.863852172679104e-05, "loss": 0.0066, "step": 32760 }, { "grad_norm": 0.13373303413391113, "learning_rate": 6.861934146415693e-05, "loss": 0.0072, "step": 32770 }, { "grad_norm": 0.13537456095218658, "learning_rate": 6.86001580199866e-05, "loss": 0.0081, "step": 32780 }, { "grad_norm": 0.1521194875240326, "learning_rate": 6.858097139755798e-05, "loss": 0.0066, "step": 32790 }, { "grad_norm": 0.13156142830848694, "learning_rate": 6.856178160014955e-05, "loss": 0.0066, "step": 32800 }, { "grad_norm": 0.12499303370714188, "learning_rate": 6.85425886310403e-05, "loss": 0.0062, "step": 32810 }, { "grad_norm": 0.12688249349594116, "learning_rate": 6.852339249350979e-05, "loss": 0.0052, "step": 32820 }, { "grad_norm": 0.1759263426065445, "learning_rate": 6.850419319083812e-05, "loss": 0.0079, "step": 32830 }, { "grad_norm": 0.13399426639080048, "learning_rate": 6.848499072630592e-05, "loss": 0.0058, "step": 32840 }, { "grad_norm": 0.12420302629470825, "learning_rate": 6.846578510319439e-05, "loss": 0.0053, "step": 32850 }, { "grad_norm": 0.11490388959646225, "learning_rate": 6.844657632478519e-05, "loss": 0.0057, "step": 32860 }, { "grad_norm": 0.10573747009038925, "learning_rate": 6.842736439436063e-05, "loss": 0.0051, "step": 32870 }, { "grad_norm": 0.09251821041107178, "learning_rate": 6.84081493152035e-05, "loss": 0.0075, "step": 32880 }, { "grad_norm": 0.09072784334421158, "learning_rate": 6.83889310905971e-05, "loss": 0.0059, "step": 32890 }, { "grad_norm": 0.12215492129325867, "learning_rate": 6.836970972382533e-05, "loss": 0.0062, "step": 32900 }, { "grad_norm": 0.08218535035848618, "learning_rate": 6.835048521817257e-05, "loss": 0.0047, "step": 32910 }, { "grad_norm": 0.11920158565044403, "learning_rate": 6.833125757692379e-05, "loss": 0.0069, "step": 32920 }, { "grad_norm": 0.12090899050235748, "learning_rate": 6.831202680336441e-05, "loss": 0.0054, "step": 32930 }, { "grad_norm": 0.11248431354761124, "learning_rate": 6.829279290078052e-05, "loss": 0.0083, "step": 32940 }, { "grad_norm": 0.11860623955726624, "learning_rate": 6.827355587245863e-05, "loss": 0.0064, "step": 32950 }, { "grad_norm": 0.12859268486499786, "learning_rate": 6.82543157216858e-05, "loss": 0.0054, "step": 32960 }, { "grad_norm": 0.13891290128231049, "learning_rate": 6.823507245174969e-05, "loss": 0.005, "step": 32970 }, { "grad_norm": 0.15355327725410461, "learning_rate": 6.821582606593841e-05, "loss": 0.0064, "step": 32980 }, { "grad_norm": 0.12577222287654877, "learning_rate": 6.81965765675407e-05, "loss": 0.0071, "step": 32990 }, { "grad_norm": 0.10903826355934143, "learning_rate": 6.81773239598457e-05, "loss": 0.0067, "step": 33000 }, { "grad_norm": 0.15878579020500183, "learning_rate": 6.815806824614319e-05, "loss": 0.0053, "step": 33010 }, { "grad_norm": 0.11110137403011322, "learning_rate": 6.813880942972343e-05, "loss": 0.0068, "step": 33020 }, { "grad_norm": 0.108706034719944, "learning_rate": 6.811954751387726e-05, "loss": 0.0091, "step": 33030 }, { "grad_norm": 0.11812929064035416, "learning_rate": 6.810028250189598e-05, "loss": 0.0073, "step": 33040 }, { "grad_norm": 0.13688361644744873, "learning_rate": 6.808101439707147e-05, "loss": 0.0081, "step": 33050 }, { "grad_norm": 0.17974264919757843, "learning_rate": 6.806174320269609e-05, "loss": 0.0075, "step": 33060 }, { "grad_norm": 0.13300219178199768, "learning_rate": 6.804246892206281e-05, "loss": 0.0065, "step": 33070 }, { "grad_norm": 0.14441828429698944, "learning_rate": 6.802319155846506e-05, "loss": 0.0076, "step": 33080 }, { "grad_norm": 0.13678082823753357, "learning_rate": 6.800391111519679e-05, "loss": 0.0084, "step": 33090 }, { "grad_norm": 0.11170594394207001, "learning_rate": 6.798462759555253e-05, "loss": 0.0069, "step": 33100 }, { "grad_norm": 0.11127248406410217, "learning_rate": 6.79653410028273e-05, "loss": 0.0066, "step": 33110 }, { "grad_norm": 0.1158316433429718, "learning_rate": 6.794605134031663e-05, "loss": 0.006, "step": 33120 }, { "grad_norm": 0.120576411485672, "learning_rate": 6.792675861131661e-05, "loss": 0.0067, "step": 33130 }, { "grad_norm": 0.22639340162277222, "learning_rate": 6.790746281912386e-05, "loss": 0.0101, "step": 33140 }, { "grad_norm": 0.13644206523895264, "learning_rate": 6.788816396703546e-05, "loss": 0.0075, "step": 33150 }, { "grad_norm": 0.18331341445446014, "learning_rate": 6.78688620583491e-05, "loss": 0.0072, "step": 33160 }, { "grad_norm": 0.19151894748210907, "learning_rate": 6.784955709636292e-05, "loss": 0.0084, "step": 33170 }, { "grad_norm": 0.10867583006620407, "learning_rate": 6.783024908437564e-05, "loss": 0.0076, "step": 33180 }, { "grad_norm": 0.16526398062705994, "learning_rate": 6.781093802568641e-05, "loss": 0.0068, "step": 33190 }, { "grad_norm": 0.10898017138242722, "learning_rate": 6.779162392359504e-05, "loss": 0.0064, "step": 33200 }, { "grad_norm": 0.11575289815664291, "learning_rate": 6.777230678140172e-05, "loss": 0.0086, "step": 33210 }, { "grad_norm": 0.10085420310497284, "learning_rate": 6.775298660240726e-05, "loss": 0.0065, "step": 33220 }, { "grad_norm": 0.12482839077711105, "learning_rate": 6.773366338991292e-05, "loss": 0.0068, "step": 33230 }, { "grad_norm": 0.1368260681629181, "learning_rate": 6.771433714722052e-05, "loss": 0.006, "step": 33240 }, { "grad_norm": 0.11660172045230865, "learning_rate": 6.769500787763239e-05, "loss": 0.0056, "step": 33250 }, { "grad_norm": 0.13596828281879425, "learning_rate": 6.76756755844514e-05, "loss": 0.0093, "step": 33260 }, { "grad_norm": 0.12142551690340042, "learning_rate": 6.765634027098087e-05, "loss": 0.007, "step": 33270 }, { "grad_norm": 0.13585205376148224, "learning_rate": 6.763700194052468e-05, "loss": 0.0074, "step": 33280 }, { "grad_norm": 0.1242847591638565, "learning_rate": 6.761766059638723e-05, "loss": 0.007, "step": 33290 }, { "grad_norm": 0.146986722946167, "learning_rate": 6.759831624187345e-05, "loss": 0.0062, "step": 33300 }, { "grad_norm": 0.16341665387153625, "learning_rate": 6.757896888028871e-05, "loss": 0.0071, "step": 33310 }, { "grad_norm": 0.15125823020935059, "learning_rate": 6.7559618514939e-05, "loss": 0.0058, "step": 33320 }, { "grad_norm": 0.1558421552181244, "learning_rate": 6.754026514913073e-05, "loss": 0.0101, "step": 33330 }, { "grad_norm": 0.12495585530996323, "learning_rate": 6.752090878617087e-05, "loss": 0.007, "step": 33340 }, { "grad_norm": 0.12977349758148193, "learning_rate": 6.75015494293669e-05, "loss": 0.005, "step": 33350 }, { "grad_norm": 0.10064344853162766, "learning_rate": 6.74821870820268e-05, "loss": 0.0064, "step": 33360 }, { "grad_norm": 0.10917828232049942, "learning_rate": 6.746282174745907e-05, "loss": 0.0059, "step": 33370 }, { "grad_norm": 0.11220680922269821, "learning_rate": 6.744345342897271e-05, "loss": 0.0068, "step": 33380 }, { "grad_norm": 0.1406562626361847, "learning_rate": 6.742408212987724e-05, "loss": 0.0063, "step": 33390 }, { "grad_norm": 0.13295094668865204, "learning_rate": 6.740470785348269e-05, "loss": 0.0059, "step": 33400 }, { "grad_norm": 0.1308528333902359, "learning_rate": 6.738533060309958e-05, "loss": 0.0073, "step": 33410 }, { "grad_norm": 0.11318212002515793, "learning_rate": 6.736595038203894e-05, "loss": 0.0062, "step": 33420 }, { "grad_norm": 0.10350700467824936, "learning_rate": 6.734656719361236e-05, "loss": 0.0051, "step": 33430 }, { "grad_norm": 0.10726059228181839, "learning_rate": 6.732718104113189e-05, "loss": 0.0086, "step": 33440 }, { "grad_norm": 0.11718510836362839, "learning_rate": 6.730779192791006e-05, "loss": 0.0061, "step": 33450 }, { "grad_norm": 0.11372233182191849, "learning_rate": 6.728839985725997e-05, "loss": 0.0063, "step": 33460 }, { "grad_norm": 0.12602092325687408, "learning_rate": 6.726900483249517e-05, "loss": 0.0069, "step": 33470 }, { "grad_norm": 0.14346399903297424, "learning_rate": 6.724960685692976e-05, "loss": 0.0074, "step": 33480 }, { "grad_norm": 0.10817793011665344, "learning_rate": 6.723020593387833e-05, "loss": 0.006, "step": 33490 }, { "grad_norm": 0.129950150847435, "learning_rate": 6.721080206665593e-05, "loss": 0.0072, "step": 33500 }, { "grad_norm": 0.10996613651514053, "learning_rate": 6.719139525857819e-05, "loss": 0.0083, "step": 33510 }, { "grad_norm": 0.11729319393634796, "learning_rate": 6.717198551296117e-05, "loss": 0.0051, "step": 33520 }, { "grad_norm": 0.10389848053455353, "learning_rate": 6.715257283312148e-05, "loss": 0.0081, "step": 33530 }, { "grad_norm": 0.11291000247001648, "learning_rate": 6.713315722237623e-05, "loss": 0.0049, "step": 33540 }, { "grad_norm": 0.10674845427274704, "learning_rate": 6.7113738684043e-05, "loss": 0.0059, "step": 33550 }, { "grad_norm": 0.07879087328910828, "learning_rate": 6.709431722143989e-05, "loss": 0.0058, "step": 33560 }, { "grad_norm": 0.16149263083934784, "learning_rate": 6.70748928378855e-05, "loss": 0.0055, "step": 33570 }, { "grad_norm": 0.1396009624004364, "learning_rate": 6.705546553669891e-05, "loss": 0.0069, "step": 33580 }, { "grad_norm": 0.1390090435743332, "learning_rate": 6.703603532119974e-05, "loss": 0.0071, "step": 33590 }, { "grad_norm": 0.1237192302942276, "learning_rate": 6.701660219470808e-05, "loss": 0.0053, "step": 33600 }, { "grad_norm": 0.13188335299491882, "learning_rate": 6.69971661605445e-05, "loss": 0.0078, "step": 33610 }, { "grad_norm": 0.14259974658489227, "learning_rate": 6.697772722203008e-05, "loss": 0.0065, "step": 33620 }, { "grad_norm": 0.15115977823734283, "learning_rate": 6.695828538248643e-05, "loss": 0.0068, "step": 33630 }, { "grad_norm": 0.08917150646448135, "learning_rate": 6.693884064523563e-05, "loss": 0.0064, "step": 33640 }, { "grad_norm": 0.1415840983390808, "learning_rate": 6.691939301360023e-05, "loss": 0.0073, "step": 33650 }, { "grad_norm": 0.11086742579936981, "learning_rate": 6.689994249090333e-05, "loss": 0.0058, "step": 33660 }, { "grad_norm": 0.08577506244182587, "learning_rate": 6.688048908046845e-05, "loss": 0.0058, "step": 33670 }, { "grad_norm": 0.10351734608411789, "learning_rate": 6.686103278561969e-05, "loss": 0.0055, "step": 33680 }, { "grad_norm": 0.16385816037654877, "learning_rate": 6.684157360968156e-05, "loss": 0.0082, "step": 33690 }, { "grad_norm": 0.09518611431121826, "learning_rate": 6.682211155597911e-05, "loss": 0.005, "step": 33700 }, { "grad_norm": 0.10847886651754379, "learning_rate": 6.680264662783789e-05, "loss": 0.0066, "step": 33710 }, { "grad_norm": 0.10823182761669159, "learning_rate": 6.678317882858391e-05, "loss": 0.0068, "step": 33720 }, { "grad_norm": 0.13298434019088745, "learning_rate": 6.67637081615437e-05, "loss": 0.0056, "step": 33730 }, { "grad_norm": 0.09178467094898224, "learning_rate": 6.674423463004427e-05, "loss": 0.0052, "step": 33740 }, { "grad_norm": 0.10983510315418243, "learning_rate": 6.672475823741308e-05, "loss": 0.0066, "step": 33750 }, { "grad_norm": 0.11143363267183304, "learning_rate": 6.670527898697811e-05, "loss": 0.0056, "step": 33760 }, { "grad_norm": 0.12376139312982559, "learning_rate": 6.668579688206788e-05, "loss": 0.0067, "step": 33770 }, { "grad_norm": 0.1093006357550621, "learning_rate": 6.666631192601131e-05, "loss": 0.0065, "step": 33780 }, { "grad_norm": 0.11481496691703796, "learning_rate": 6.664682412213785e-05, "loss": 0.0063, "step": 33790 }, { "grad_norm": 0.10276574641466141, "learning_rate": 6.662733347377745e-05, "loss": 0.0075, "step": 33800 }, { "grad_norm": 0.09370255470275879, "learning_rate": 6.660783998426051e-05, "loss": 0.0054, "step": 33810 }, { "grad_norm": 0.11675860732793808, "learning_rate": 6.658834365691794e-05, "loss": 0.0054, "step": 33820 }, { "grad_norm": 0.12654808163642883, "learning_rate": 6.656884449508115e-05, "loss": 0.0064, "step": 33830 }, { "grad_norm": 0.1380983591079712, "learning_rate": 6.654934250208198e-05, "loss": 0.0053, "step": 33840 }, { "grad_norm": 0.12630292773246765, "learning_rate": 6.65298376812528e-05, "loss": 0.0069, "step": 33850 }, { "grad_norm": 0.10598122328519821, "learning_rate": 6.651033003592646e-05, "loss": 0.0044, "step": 33860 }, { "grad_norm": 0.08971534669399261, "learning_rate": 6.649081956943626e-05, "loss": 0.0055, "step": 33870 }, { "grad_norm": 0.1164073571562767, "learning_rate": 6.647130628511604e-05, "loss": 0.0066, "step": 33880 }, { "grad_norm": 0.13454093039035797, "learning_rate": 6.645179018630005e-05, "loss": 0.0068, "step": 33890 }, { "grad_norm": 0.10740931332111359, "learning_rate": 6.643227127632309e-05, "loss": 0.007, "step": 33900 }, { "grad_norm": 0.10662277042865753, "learning_rate": 6.641274955852038e-05, "loss": 0.0053, "step": 33910 }, { "grad_norm": 0.10478927940130234, "learning_rate": 6.639322503622768e-05, "loss": 0.0063, "step": 33920 }, { "grad_norm": 0.13685904443264008, "learning_rate": 6.637369771278116e-05, "loss": 0.0093, "step": 33930 }, { "grad_norm": 0.10439807176589966, "learning_rate": 6.635416759151751e-05, "loss": 0.006, "step": 33940 }, { "grad_norm": 0.1175079271197319, "learning_rate": 6.633463467577394e-05, "loss": 0.0052, "step": 33950 }, { "grad_norm": 0.09768210351467133, "learning_rate": 6.631509896888803e-05, "loss": 0.007, "step": 33960 }, { "grad_norm": 0.13655820488929749, "learning_rate": 6.629556047419794e-05, "loss": 0.0066, "step": 33970 }, { "grad_norm": 0.16151030361652374, "learning_rate": 6.627601919504223e-05, "loss": 0.0072, "step": 33980 }, { "grad_norm": 0.15535348653793335, "learning_rate": 6.625647513476001e-05, "loss": 0.0065, "step": 33990 }, { "grad_norm": 0.12304475903511047, "learning_rate": 6.62369282966908e-05, "loss": 0.0071, "step": 34000 }, { "grad_norm": 0.1237955316901207, "learning_rate": 6.621737868417464e-05, "loss": 0.0054, "step": 34010 }, { "grad_norm": 0.1032385602593422, "learning_rate": 6.619782630055198e-05, "loss": 0.0065, "step": 34020 }, { "grad_norm": 0.14969183504581451, "learning_rate": 6.617827114916382e-05, "loss": 0.0084, "step": 34030 }, { "grad_norm": 0.11036194115877151, "learning_rate": 6.615871323335161e-05, "loss": 0.0055, "step": 34040 }, { "grad_norm": 0.11096203327178955, "learning_rate": 6.613915255645725e-05, "loss": 0.0057, "step": 34050 }, { "grad_norm": 0.14632773399353027, "learning_rate": 6.611958912182312e-05, "loss": 0.0091, "step": 34060 }, { "grad_norm": 0.14263559877872467, "learning_rate": 6.610002293279207e-05, "loss": 0.0059, "step": 34070 }, { "grad_norm": 0.10384848713874817, "learning_rate": 6.608045399270746e-05, "loss": 0.0056, "step": 34080 }, { "grad_norm": 0.1193741038441658, "learning_rate": 6.606088230491304e-05, "loss": 0.0071, "step": 34090 }, { "grad_norm": 0.09301339089870453, "learning_rate": 6.604130787275312e-05, "loss": 0.0054, "step": 34100 }, { "grad_norm": 0.13015806674957275, "learning_rate": 6.602173069957242e-05, "loss": 0.0081, "step": 34110 }, { "grad_norm": 0.11429945379495621, "learning_rate": 6.600215078871612e-05, "loss": 0.0074, "step": 34120 }, { "grad_norm": 0.09316085278987885, "learning_rate": 6.598256814352992e-05, "loss": 0.0072, "step": 34130 }, { "grad_norm": 0.09772580862045288, "learning_rate": 6.596298276735995e-05, "loss": 0.0051, "step": 34140 }, { "grad_norm": 0.1199788749217987, "learning_rate": 6.594339466355282e-05, "loss": 0.0067, "step": 34150 }, { "grad_norm": 0.12223421782255173, "learning_rate": 6.592380383545558e-05, "loss": 0.0054, "step": 34160 }, { "grad_norm": 0.16066913306713104, "learning_rate": 6.590421028641577e-05, "loss": 0.0055, "step": 34170 }, { "grad_norm": 0.09705045074224472, "learning_rate": 6.588461401978143e-05, "loss": 0.0068, "step": 34180 }, { "grad_norm": 0.13111701607704163, "learning_rate": 6.586501503890099e-05, "loss": 0.0054, "step": 34190 }, { "grad_norm": 0.13992270827293396, "learning_rate": 6.584541334712338e-05, "loss": 0.0054, "step": 34200 }, { "grad_norm": 0.09998250752687454, "learning_rate": 6.582580894779802e-05, "loss": 0.0049, "step": 34210 }, { "grad_norm": 0.116480253636837, "learning_rate": 6.580620184427473e-05, "loss": 0.0057, "step": 34220 }, { "grad_norm": 0.12757578492164612, "learning_rate": 6.578659203990385e-05, "loss": 0.0066, "step": 34230 }, { "grad_norm": 0.13432520627975464, "learning_rate": 6.576697953803615e-05, "loss": 0.0067, "step": 34240 }, { "grad_norm": 0.07968853414058685, "learning_rate": 6.57473643420229e-05, "loss": 0.006, "step": 34250 }, { "grad_norm": 0.16973111033439636, "learning_rate": 6.572774645521574e-05, "loss": 0.007, "step": 34260 }, { "grad_norm": 0.11407876014709473, "learning_rate": 6.570812588096688e-05, "loss": 0.0062, "step": 34270 }, { "grad_norm": 0.13512149453163147, "learning_rate": 6.568850262262893e-05, "loss": 0.0077, "step": 34280 }, { "grad_norm": 0.1493120640516281, "learning_rate": 6.566887668355497e-05, "loss": 0.0062, "step": 34290 }, { "grad_norm": 0.10753381997346878, "learning_rate": 6.564924806709851e-05, "loss": 0.0048, "step": 34300 }, { "grad_norm": 0.11404058337211609, "learning_rate": 6.562961677661359e-05, "loss": 0.005, "step": 34310 }, { "grad_norm": 0.10695597529411316, "learning_rate": 6.56099828154546e-05, "loss": 0.0068, "step": 34320 }, { "grad_norm": 0.10208834707736969, "learning_rate": 6.55903461869765e-05, "loss": 0.0059, "step": 34330 }, { "grad_norm": 0.09951476007699966, "learning_rate": 6.557070689453465e-05, "loss": 0.005, "step": 34340 }, { "grad_norm": 0.09122961014509201, "learning_rate": 6.555106494148482e-05, "loss": 0.0062, "step": 34350 }, { "grad_norm": 0.1256236582994461, "learning_rate": 6.553142033118333e-05, "loss": 0.0058, "step": 34360 }, { "grad_norm": 0.09821125864982605, "learning_rate": 6.551177306698688e-05, "loss": 0.0054, "step": 34370 }, { "grad_norm": 0.09407747536897659, "learning_rate": 6.549212315225267e-05, "loss": 0.0056, "step": 34380 }, { "grad_norm": 0.18870393931865692, "learning_rate": 6.547247059033833e-05, "loss": 0.0102, "step": 34390 }, { "grad_norm": 0.12375662475824356, "learning_rate": 6.545281538460193e-05, "loss": 0.0056, "step": 34400 }, { "grad_norm": 0.13735686242580414, "learning_rate": 6.543315753840202e-05, "loss": 0.0057, "step": 34410 }, { "grad_norm": 0.10078112781047821, "learning_rate": 6.541349705509758e-05, "loss": 0.005, "step": 34420 }, { "grad_norm": 0.11436805874109268, "learning_rate": 6.539383393804805e-05, "loss": 0.0057, "step": 34430 }, { "grad_norm": 0.11412984132766724, "learning_rate": 6.537416819061333e-05, "loss": 0.005, "step": 34440 }, { "grad_norm": 0.12147030979394913, "learning_rate": 6.535449981615375e-05, "loss": 0.0063, "step": 34450 }, { "grad_norm": 0.1403125822544098, "learning_rate": 6.53348288180301e-05, "loss": 0.0069, "step": 34460 }, { "grad_norm": 0.11832970380783081, "learning_rate": 6.531515519960361e-05, "loss": 0.0061, "step": 34470 }, { "grad_norm": 0.16409410536289215, "learning_rate": 6.529547896423597e-05, "loss": 0.0071, "step": 34480 }, { "grad_norm": 0.14836567640304565, "learning_rate": 6.52758001152893e-05, "loss": 0.0057, "step": 34490 }, { "grad_norm": 0.09532028436660767, "learning_rate": 6.525611865612618e-05, "loss": 0.006, "step": 34500 }, { "grad_norm": 0.15245583653450012, "learning_rate": 6.523643459010966e-05, "loss": 0.007, "step": 34510 }, { "grad_norm": 0.1259480118751526, "learning_rate": 6.521674792060317e-05, "loss": 0.0063, "step": 34520 }, { "grad_norm": 0.1112339049577713, "learning_rate": 6.519705865097063e-05, "loss": 0.0066, "step": 34530 }, { "grad_norm": 0.16811607778072357, "learning_rate": 6.517736678457641e-05, "loss": 0.0064, "step": 34540 }, { "grad_norm": 0.12169984728097916, "learning_rate": 6.515767232478534e-05, "loss": 0.0063, "step": 34550 }, { "grad_norm": 0.12114694714546204, "learning_rate": 6.51379752749626e-05, "loss": 0.0062, "step": 34560 }, { "grad_norm": 0.11963141709566116, "learning_rate": 6.511827563847393e-05, "loss": 0.0069, "step": 34570 }, { "grad_norm": 0.07806950062513351, "learning_rate": 6.509857341868542e-05, "loss": 0.0079, "step": 34580 }, { "grad_norm": 0.14455826580524445, "learning_rate": 6.507886861896367e-05, "loss": 0.0053, "step": 34590 }, { "grad_norm": 0.15752431750297546, "learning_rate": 6.505916124267567e-05, "loss": 0.005, "step": 34600 }, { "grad_norm": 0.1077558621764183, "learning_rate": 6.503945129318891e-05, "loss": 0.0073, "step": 34610 }, { "grad_norm": 0.17035645246505737, "learning_rate": 6.501973877387122e-05, "loss": 0.0062, "step": 34620 }, { "grad_norm": 0.08565754443407059, "learning_rate": 6.500002368809098e-05, "loss": 0.0073, "step": 34630 }, { "grad_norm": 0.11494510620832443, "learning_rate": 6.498030603921694e-05, "loss": 0.0072, "step": 34640 }, { "grad_norm": 0.13675296306610107, "learning_rate": 6.496058583061832e-05, "loss": 0.0086, "step": 34650 }, { "grad_norm": 0.10850148648023605, "learning_rate": 6.494086306566475e-05, "loss": 0.0055, "step": 34660 }, { "grad_norm": 0.1323881298303604, "learning_rate": 6.492113774772632e-05, "loss": 0.006, "step": 34670 }, { "grad_norm": 0.13428851962089539, "learning_rate": 6.490140988017354e-05, "loss": 0.0075, "step": 34680 }, { "grad_norm": 0.1339653879404068, "learning_rate": 6.488167946637736e-05, "loss": 0.0057, "step": 34690 }, { "grad_norm": 0.11053948104381561, "learning_rate": 6.486194650970915e-05, "loss": 0.0064, "step": 34700 }, { "grad_norm": 0.09813255816698074, "learning_rate": 6.48422110135408e-05, "loss": 0.0064, "step": 34710 }, { "grad_norm": 0.14949940145015717, "learning_rate": 6.482247298124451e-05, "loss": 0.0058, "step": 34720 }, { "grad_norm": 0.1392107754945755, "learning_rate": 6.480273241619297e-05, "loss": 0.0053, "step": 34730 }, { "grad_norm": 0.1257963627576828, "learning_rate": 6.478298932175933e-05, "loss": 0.0068, "step": 34740 }, { "grad_norm": 0.14106802642345428, "learning_rate": 6.476324370131712e-05, "loss": 0.0069, "step": 34750 }, { "grad_norm": 0.10813124477863312, "learning_rate": 6.474349555824036e-05, "loss": 0.0053, "step": 34760 }, { "grad_norm": 0.17305399477481842, "learning_rate": 6.472374489590342e-05, "loss": 0.0057, "step": 34770 }, { "grad_norm": 0.1120617464184761, "learning_rate": 6.470399171768118e-05, "loss": 0.0053, "step": 34780 }, { "grad_norm": 0.11179113388061523, "learning_rate": 6.468423602694891e-05, "loss": 0.005, "step": 34790 }, { "grad_norm": 0.10934678465127945, "learning_rate": 6.466447782708232e-05, "loss": 0.0049, "step": 34800 }, { "grad_norm": 0.13103188574314117, "learning_rate": 6.464471712145754e-05, "loss": 0.0066, "step": 34810 }, { "grad_norm": 0.13258223235607147, "learning_rate": 6.462495391345114e-05, "loss": 0.0052, "step": 34820 }, { "grad_norm": 0.0812569260597229, "learning_rate": 6.46051882064401e-05, "loss": 0.0065, "step": 34830 }, { "grad_norm": 0.12992048263549805, "learning_rate": 6.458542000380186e-05, "loss": 0.0056, "step": 34840 }, { "grad_norm": 0.10824316740036011, "learning_rate": 6.456564930891424e-05, "loss": 0.006, "step": 34850 }, { "grad_norm": 0.12186776846647263, "learning_rate": 6.454587612515555e-05, "loss": 0.0052, "step": 34860 }, { "grad_norm": 0.07737454771995544, "learning_rate": 6.452610045590444e-05, "loss": 0.0067, "step": 34870 }, { "grad_norm": 0.1105104312300682, "learning_rate": 6.450632230454005e-05, "loss": 0.0051, "step": 34880 }, { "grad_norm": 0.08108046650886536, "learning_rate": 6.448654167444195e-05, "loss": 0.0067, "step": 34890 }, { "grad_norm": 0.12702876329421997, "learning_rate": 6.446675856899005e-05, "loss": 0.0054, "step": 34900 }, { "grad_norm": 0.12663133442401886, "learning_rate": 6.444697299156481e-05, "loss": 0.0058, "step": 34910 }, { "grad_norm": 0.10102122277021408, "learning_rate": 6.442718494554701e-05, "loss": 0.0048, "step": 34920 }, { "grad_norm": 0.12947329878807068, "learning_rate": 6.440739443431787e-05, "loss": 0.0056, "step": 34930 }, { "grad_norm": 0.16340501606464386, "learning_rate": 6.438760146125906e-05, "loss": 0.007, "step": 34940 }, { "grad_norm": 0.1473550647497177, "learning_rate": 6.436780602975267e-05, "loss": 0.0056, "step": 34950 }, { "grad_norm": 0.12905694544315338, "learning_rate": 6.43480081431812e-05, "loss": 0.0064, "step": 34960 }, { "grad_norm": 0.11096642166376114, "learning_rate": 6.432820780492756e-05, "loss": 0.0058, "step": 34970 }, { "grad_norm": 0.11845264583826065, "learning_rate": 6.430840501837506e-05, "loss": 0.0084, "step": 34980 }, { "grad_norm": 0.08821011334657669, "learning_rate": 6.428859978690748e-05, "loss": 0.0059, "step": 34990 }, { "grad_norm": 0.1267869919538498, "learning_rate": 6.426879211390901e-05, "loss": 0.005, "step": 35000 }, { "grad_norm": 0.1660020649433136, "learning_rate": 6.424898200276422e-05, "loss": 0.0109, "step": 35010 }, { "grad_norm": 0.10386480391025543, "learning_rate": 6.42291694568581e-05, "loss": 0.0055, "step": 35020 }, { "grad_norm": 0.12391127645969391, "learning_rate": 6.42093544795761e-05, "loss": 0.0076, "step": 35030 }, { "grad_norm": 0.10213647782802582, "learning_rate": 6.418953707430403e-05, "loss": 0.0064, "step": 35040 }, { "grad_norm": 0.0971711203455925, "learning_rate": 6.416971724442819e-05, "loss": 0.0051, "step": 35050 }, { "grad_norm": 0.07595653086900711, "learning_rate": 6.414989499333519e-05, "loss": 0.0054, "step": 35060 }, { "grad_norm": 0.15580816566944122, "learning_rate": 6.413007032441214e-05, "loss": 0.0066, "step": 35070 }, { "grad_norm": 0.1545766144990921, "learning_rate": 6.411024324104653e-05, "loss": 0.0089, "step": 35080 }, { "grad_norm": 0.13207249343395233, "learning_rate": 6.409041374662628e-05, "loss": 0.0069, "step": 35090 }, { "grad_norm": 0.15572728216648102, "learning_rate": 6.407058184453967e-05, "loss": 0.0069, "step": 35100 }, { "grad_norm": 0.1250016838312149, "learning_rate": 6.405074753817548e-05, "loss": 0.0093, "step": 35110 }, { "grad_norm": 0.11972321569919586, "learning_rate": 6.40309108309228e-05, "loss": 0.0067, "step": 35120 }, { "grad_norm": 0.12343825399875641, "learning_rate": 6.401107172617122e-05, "loss": 0.0058, "step": 35130 }, { "grad_norm": 0.11851410567760468, "learning_rate": 6.399123022731068e-05, "loss": 0.0052, "step": 35140 }, { "grad_norm": 0.12153975665569305, "learning_rate": 6.397138633773157e-05, "loss": 0.0052, "step": 35150 }, { "grad_norm": 0.15140078961849213, "learning_rate": 6.395154006082463e-05, "loss": 0.006, "step": 35160 }, { "grad_norm": 0.10285845398902893, "learning_rate": 6.393169139998109e-05, "loss": 0.0052, "step": 35170 }, { "grad_norm": 0.14452405273914337, "learning_rate": 6.39118403585925e-05, "loss": 0.0073, "step": 35180 }, { "grad_norm": 0.15213768184185028, "learning_rate": 6.38919869400509e-05, "loss": 0.0074, "step": 35190 }, { "grad_norm": 0.10985288769006729, "learning_rate": 6.387213114774865e-05, "loss": 0.0084, "step": 35200 }, { "grad_norm": 0.10567064583301544, "learning_rate": 6.385227298507863e-05, "loss": 0.0054, "step": 35210 }, { "grad_norm": 0.17493367195129395, "learning_rate": 6.3832412455434e-05, "loss": 0.0061, "step": 35220 }, { "grad_norm": 0.15622742474079132, "learning_rate": 6.381254956220841e-05, "loss": 0.0064, "step": 35230 }, { "grad_norm": 0.08735550194978714, "learning_rate": 6.379268430879586e-05, "loss": 0.0065, "step": 35240 }, { "grad_norm": 0.1337163746356964, "learning_rate": 6.37728166985908e-05, "loss": 0.0066, "step": 35250 }, { "grad_norm": 0.10017233341932297, "learning_rate": 6.375294673498804e-05, "loss": 0.0078, "step": 35260 }, { "grad_norm": 0.14839619398117065, "learning_rate": 6.373307442138284e-05, "loss": 0.0058, "step": 35270 }, { "grad_norm": 0.1301998645067215, "learning_rate": 6.371319976117081e-05, "loss": 0.0052, "step": 35280 }, { "grad_norm": 0.13631854951381683, "learning_rate": 6.3693322757748e-05, "loss": 0.0067, "step": 35290 }, { "grad_norm": 0.10779570043087006, "learning_rate": 6.367344341451086e-05, "loss": 0.007, "step": 35300 }, { "grad_norm": 0.1776195764541626, "learning_rate": 6.36535617348562e-05, "loss": 0.0063, "step": 35310 }, { "grad_norm": 0.1319110095500946, "learning_rate": 6.363367772218128e-05, "loss": 0.0076, "step": 35320 }, { "grad_norm": 0.13091683387756348, "learning_rate": 6.36137913798837e-05, "loss": 0.0072, "step": 35330 }, { "grad_norm": 0.12504181265830994, "learning_rate": 6.359390271136151e-05, "loss": 0.0074, "step": 35340 }, { "grad_norm": 0.1188458725810051, "learning_rate": 6.357401172001314e-05, "loss": 0.007, "step": 35350 }, { "grad_norm": 0.09023315459489822, "learning_rate": 6.355411840923742e-05, "loss": 0.0054, "step": 35360 }, { "grad_norm": 0.09253997355699539, "learning_rate": 6.353422278243358e-05, "loss": 0.009, "step": 35370 }, { "grad_norm": 0.08824611455202103, "learning_rate": 6.351432484300121e-05, "loss": 0.0047, "step": 35380 }, { "grad_norm": 0.09007848054170609, "learning_rate": 6.349442459434036e-05, "loss": 0.0068, "step": 35390 }, { "grad_norm": 0.14328952133655548, "learning_rate": 6.34745220398514e-05, "loss": 0.0067, "step": 35400 }, { "grad_norm": 0.11789999157190323, "learning_rate": 6.345461718293518e-05, "loss": 0.0058, "step": 35410 }, { "grad_norm": 0.16320645809173584, "learning_rate": 6.343471002699286e-05, "loss": 0.0078, "step": 35420 }, { "grad_norm": 0.10768771916627884, "learning_rate": 6.341480057542602e-05, "loss": 0.007, "step": 35430 }, { "grad_norm": 0.15354683995246887, "learning_rate": 6.339488883163667e-05, "loss": 0.0099, "step": 35440 }, { "grad_norm": 0.11034418642520905, "learning_rate": 6.337497479902716e-05, "loss": 0.0061, "step": 35450 }, { "grad_norm": 0.11796967685222626, "learning_rate": 6.335505848100027e-05, "loss": 0.0072, "step": 35460 }, { "grad_norm": 0.1364738494157791, "learning_rate": 6.333513988095915e-05, "loss": 0.0066, "step": 35470 }, { "grad_norm": 0.09521741420030594, "learning_rate": 6.331521900230735e-05, "loss": 0.0054, "step": 35480 }, { "grad_norm": 0.13342131674289703, "learning_rate": 6.329529584844878e-05, "loss": 0.004, "step": 35490 }, { "grad_norm": 0.10246925801038742, "learning_rate": 6.327537042278777e-05, "loss": 0.0067, "step": 35500 }, { "grad_norm": 0.16949057579040527, "learning_rate": 6.325544272872905e-05, "loss": 0.0081, "step": 35510 }, { "grad_norm": 0.15122704207897186, "learning_rate": 6.323551276967771e-05, "loss": 0.0069, "step": 35520 }, { "grad_norm": 0.1228032112121582, "learning_rate": 6.321558054903922e-05, "loss": 0.0066, "step": 35530 }, { "grad_norm": 0.13939110934734344, "learning_rate": 6.319564607021947e-05, "loss": 0.0061, "step": 35540 }, { "grad_norm": 0.1162976399064064, "learning_rate": 6.31757093366247e-05, "loss": 0.0054, "step": 35550 }, { "grad_norm": 0.08468236774206161, "learning_rate": 6.315577035166154e-05, "loss": 0.0061, "step": 35560 }, { "grad_norm": 0.10311800986528397, "learning_rate": 6.313582911873708e-05, "loss": 0.0064, "step": 35570 }, { "grad_norm": 0.12045856565237045, "learning_rate": 6.311588564125865e-05, "loss": 0.0057, "step": 35580 }, { "grad_norm": 0.15259452164173126, "learning_rate": 6.30959399226341e-05, "loss": 0.0068, "step": 35590 }, { "grad_norm": 0.13691993057727814, "learning_rate": 6.30759919662716e-05, "loss": 0.0059, "step": 35600 }, { "grad_norm": 0.09924398362636566, "learning_rate": 6.30560417755797e-05, "loss": 0.006, "step": 35610 }, { "grad_norm": 0.1492663472890854, "learning_rate": 6.303608935396735e-05, "loss": 0.008, "step": 35620 }, { "grad_norm": 0.11876239627599716, "learning_rate": 6.301613470484386e-05, "loss": 0.0078, "step": 35630 }, { "grad_norm": 0.15779463946819305, "learning_rate": 6.299617783161893e-05, "loss": 0.0089, "step": 35640 }, { "grad_norm": 0.1504649817943573, "learning_rate": 6.297621873770266e-05, "loss": 0.0056, "step": 35650 }, { "grad_norm": 0.06800302863121033, "learning_rate": 6.29562574265055e-05, "loss": 0.0059, "step": 35660 }, { "grad_norm": 0.0961829274892807, "learning_rate": 6.293629390143834e-05, "loss": 0.006, "step": 35670 }, { "grad_norm": 0.09540294110774994, "learning_rate": 6.291632816591232e-05, "loss": 0.0061, "step": 35680 }, { "grad_norm": 0.10379024595022202, "learning_rate": 6.28963602233391e-05, "loss": 0.0087, "step": 35690 }, { "grad_norm": 0.13485145568847656, "learning_rate": 6.287639007713062e-05, "loss": 0.0064, "step": 35700 }, { "grad_norm": 0.13456694781780243, "learning_rate": 6.285641773069926e-05, "loss": 0.0057, "step": 35710 }, { "grad_norm": 0.14666564762592316, "learning_rate": 6.283644318745773e-05, "loss": 0.0088, "step": 35720 }, { "grad_norm": 0.11863841116428375, "learning_rate": 6.281646645081912e-05, "loss": 0.0079, "step": 35730 }, { "grad_norm": 0.1443621814250946, "learning_rate": 6.279648752419693e-05, "loss": 0.0058, "step": 35740 }, { "grad_norm": 0.14277106523513794, "learning_rate": 6.2776506411005e-05, "loss": 0.006, "step": 35750 }, { "grad_norm": 0.1008681133389473, "learning_rate": 6.275652311465758e-05, "loss": 0.0062, "step": 35760 }, { "grad_norm": 0.09114456921815872, "learning_rate": 6.273653763856926e-05, "loss": 0.007, "step": 35770 }, { "grad_norm": 0.12466621398925781, "learning_rate": 6.271654998615501e-05, "loss": 0.0041, "step": 35780 }, { "grad_norm": 0.11460698395967484, "learning_rate": 6.269656016083013e-05, "loss": 0.0049, "step": 35790 }, { "grad_norm": 0.1325109452009201, "learning_rate": 6.267656816601038e-05, "loss": 0.007, "step": 35800 }, { "grad_norm": 0.1503581553697586, "learning_rate": 6.265657400511185e-05, "loss": 0.0077, "step": 35810 }, { "grad_norm": 0.13095256686210632, "learning_rate": 6.263657768155098e-05, "loss": 0.0053, "step": 35820 }, { "grad_norm": 0.11675000935792923, "learning_rate": 6.261657919874457e-05, "loss": 0.0053, "step": 35830 }, { "grad_norm": 0.11554872989654541, "learning_rate": 6.259657856010986e-05, "loss": 0.0059, "step": 35840 }, { "grad_norm": 0.13927839696407318, "learning_rate": 6.257657576906439e-05, "loss": 0.0066, "step": 35850 }, { "grad_norm": 0.09460251033306122, "learning_rate": 6.255657082902609e-05, "loss": 0.0043, "step": 35860 }, { "grad_norm": 0.11488205194473267, "learning_rate": 6.253656374341325e-05, "loss": 0.0084, "step": 35870 }, { "grad_norm": 0.12481539696455002, "learning_rate": 6.251655451564457e-05, "loss": 0.0047, "step": 35880 }, { "grad_norm": 0.12993162870407104, "learning_rate": 6.249654314913902e-05, "loss": 0.0047, "step": 35890 }, { "grad_norm": 0.10140015184879303, "learning_rate": 6.247652964731604e-05, "loss": 0.0056, "step": 35900 }, { "grad_norm": 0.21459105610847473, "learning_rate": 6.245651401359537e-05, "loss": 0.0077, "step": 35910 }, { "grad_norm": 0.1252630352973938, "learning_rate": 6.243649625139715e-05, "loss": 0.0051, "step": 35920 }, { "grad_norm": 0.0925574079155922, "learning_rate": 6.241647636414185e-05, "loss": 0.0055, "step": 35930 }, { "grad_norm": 0.09622730314731598, "learning_rate": 6.239645435525034e-05, "loss": 0.0054, "step": 35940 }, { "grad_norm": 0.09776458144187927, "learning_rate": 6.237643022814381e-05, "loss": 0.0056, "step": 35950 }, { "grad_norm": 0.0943792387843132, "learning_rate": 6.235640398624386e-05, "loss": 0.0047, "step": 35960 }, { "grad_norm": 0.13087914884090424, "learning_rate": 6.233637563297243e-05, "loss": 0.0061, "step": 35970 }, { "grad_norm": 0.11050648987293243, "learning_rate": 6.23163451717518e-05, "loss": 0.0057, "step": 35980 }, { "grad_norm": 0.1355978548526764, "learning_rate": 6.229631260600463e-05, "loss": 0.0074, "step": 35990 }, { "grad_norm": 0.11031665652990341, "learning_rate": 6.227627793915392e-05, "loss": 0.0071, "step": 36000 }, { "grad_norm": 0.11102855205535889, "learning_rate": 6.225624117462309e-05, "loss": 0.0058, "step": 36010 }, { "grad_norm": 0.1320943385362625, "learning_rate": 6.223620231583586e-05, "loss": 0.0053, "step": 36020 }, { "grad_norm": 0.11569681763648987, "learning_rate": 6.221616136621629e-05, "loss": 0.0075, "step": 36030 }, { "grad_norm": 0.10273642092943192, "learning_rate": 6.219611832918887e-05, "loss": 0.0066, "step": 36040 }, { "grad_norm": 0.08548104017972946, "learning_rate": 6.217607320817838e-05, "loss": 0.0063, "step": 36050 }, { "grad_norm": 0.08233034610748291, "learning_rate": 6.215602600661001e-05, "loss": 0.0071, "step": 36060 }, { "grad_norm": 0.09884423762559891, "learning_rate": 6.213597672790925e-05, "loss": 0.0058, "step": 36070 }, { "grad_norm": 0.13573987782001495, "learning_rate": 6.2115925375502e-05, "loss": 0.0062, "step": 36080 }, { "grad_norm": 0.14526589214801788, "learning_rate": 6.209587195281447e-05, "loss": 0.0071, "step": 36090 }, { "grad_norm": 0.14804315567016602, "learning_rate": 6.207581646327324e-05, "loss": 0.0077, "step": 36100 }, { "grad_norm": 0.12752169370651245, "learning_rate": 6.205575891030526e-05, "loss": 0.0064, "step": 36110 }, { "grad_norm": 0.13639366626739502, "learning_rate": 6.203569929733781e-05, "loss": 0.0058, "step": 36120 }, { "grad_norm": 0.11975555866956711, "learning_rate": 6.201563762779852e-05, "loss": 0.0056, "step": 36130 }, { "grad_norm": 0.10389046370983124, "learning_rate": 6.199557390511538e-05, "loss": 0.0047, "step": 36140 }, { "grad_norm": 0.1596573293209076, "learning_rate": 6.197550813271675e-05, "loss": 0.0057, "step": 36150 }, { "grad_norm": 0.08039844781160355, "learning_rate": 6.195544031403131e-05, "loss": 0.0049, "step": 36160 }, { "grad_norm": 0.08295224606990814, "learning_rate": 6.19353704524881e-05, "loss": 0.0069, "step": 36170 }, { "grad_norm": 0.08182092756032944, "learning_rate": 6.191529855151652e-05, "loss": 0.0051, "step": 36180 }, { "grad_norm": 0.11505705118179321, "learning_rate": 6.189522461454629e-05, "loss": 0.0052, "step": 36190 }, { "grad_norm": 0.09799502044916153, "learning_rate": 6.187514864500752e-05, "loss": 0.0048, "step": 36200 }, { "grad_norm": 0.11098231375217438, "learning_rate": 6.185507064633062e-05, "loss": 0.0048, "step": 36210 }, { "grad_norm": 0.09429214149713516, "learning_rate": 6.18349906219464e-05, "loss": 0.0045, "step": 36220 }, { "grad_norm": 0.11359428614377975, "learning_rate": 6.181490857528596e-05, "loss": 0.0068, "step": 36230 }, { "grad_norm": 0.11565151810646057, "learning_rate": 6.179482450978077e-05, "loss": 0.006, "step": 36240 }, { "grad_norm": 0.12690052390098572, "learning_rate": 6.177473842886269e-05, "loss": 0.0053, "step": 36250 }, { "grad_norm": 0.185203418135643, "learning_rate": 6.175465033596382e-05, "loss": 0.0059, "step": 36260 }, { "grad_norm": 0.13214221596717834, "learning_rate": 6.173456023451671e-05, "loss": 0.0062, "step": 36270 }, { "grad_norm": 0.13992877304553986, "learning_rate": 6.171446812795422e-05, "loss": 0.0065, "step": 36280 }, { "grad_norm": 0.14104034006595612, "learning_rate": 6.169437401970949e-05, "loss": 0.0055, "step": 36290 }, { "grad_norm": 0.13645298779010773, "learning_rate": 6.16742779132161e-05, "loss": 0.006, "step": 36300 }, { "grad_norm": 0.1307234764099121, "learning_rate": 6.165417981190789e-05, "loss": 0.0051, "step": 36310 }, { "grad_norm": 0.1025160476565361, "learning_rate": 6.16340797192191e-05, "loss": 0.0054, "step": 36320 }, { "grad_norm": 0.15037775039672852, "learning_rate": 6.161397763858427e-05, "loss": 0.0063, "step": 36330 }, { "grad_norm": 0.12047364562749863, "learning_rate": 6.159387357343834e-05, "loss": 0.0063, "step": 36340 }, { "grad_norm": 0.14395512640476227, "learning_rate": 6.157376752721648e-05, "loss": 0.0054, "step": 36350 }, { "grad_norm": 0.1270807981491089, "learning_rate": 6.155365950335428e-05, "loss": 0.0046, "step": 36360 }, { "grad_norm": 0.12180770933628082, "learning_rate": 6.153354950528768e-05, "loss": 0.0055, "step": 36370 }, { "grad_norm": 0.1539231687784195, "learning_rate": 6.151343753645293e-05, "loss": 0.0067, "step": 36380 }, { "grad_norm": 0.0902710035443306, "learning_rate": 6.149332360028657e-05, "loss": 0.0073, "step": 36390 }, { "grad_norm": 0.10998381674289703, "learning_rate": 6.147320770022555e-05, "loss": 0.0049, "step": 36400 }, { "grad_norm": 0.13508538901805878, "learning_rate": 6.145308983970715e-05, "loss": 0.0061, "step": 36410 }, { "grad_norm": 0.14042963087558746, "learning_rate": 6.143297002216892e-05, "loss": 0.0054, "step": 36420 }, { "grad_norm": 0.14046764373779297, "learning_rate": 6.141284825104882e-05, "loss": 0.0067, "step": 36430 }, { "grad_norm": 0.12572690844535828, "learning_rate": 6.13927245297851e-05, "loss": 0.0056, "step": 36440 }, { "grad_norm": 0.1738177239894867, "learning_rate": 6.137259886181633e-05, "loss": 0.0056, "step": 36450 }, { "grad_norm": 0.09534664452075958, "learning_rate": 6.135247125058145e-05, "loss": 0.0049, "step": 36460 }, { "grad_norm": 0.12125861644744873, "learning_rate": 6.133234169951974e-05, "loss": 0.0054, "step": 36470 }, { "grad_norm": 0.08202433586120605, "learning_rate": 6.131221021207078e-05, "loss": 0.0049, "step": 36480 }, { "grad_norm": 0.11549786478281021, "learning_rate": 6.129207679167448e-05, "loss": 0.005, "step": 36490 }, { "grad_norm": 0.08914320170879364, "learning_rate": 6.127194144177109e-05, "loss": 0.0055, "step": 36500 }, { "grad_norm": 0.11365353316068649, "learning_rate": 6.125180416580118e-05, "loss": 0.0047, "step": 36510 }, { "grad_norm": 0.09843189269304276, "learning_rate": 6.123166496720571e-05, "loss": 0.0041, "step": 36520 }, { "grad_norm": 0.09494759887456894, "learning_rate": 6.121152384942588e-05, "loss": 0.0049, "step": 36530 }, { "grad_norm": 0.11894529312849045, "learning_rate": 6.119138081590324e-05, "loss": 0.0065, "step": 36540 }, { "grad_norm": 0.122207410633564, "learning_rate": 6.117123587007971e-05, "loss": 0.0054, "step": 36550 }, { "grad_norm": 0.09370506554841995, "learning_rate": 6.11510890153975e-05, "loss": 0.0052, "step": 36560 }, { "grad_norm": 0.08819311857223511, "learning_rate": 6.113094025529916e-05, "loss": 0.0049, "step": 36570 }, { "grad_norm": 0.11446943134069443, "learning_rate": 6.111078959322757e-05, "loss": 0.0076, "step": 36580 }, { "grad_norm": 0.1330704540014267, "learning_rate": 6.109063703262592e-05, "loss": 0.0074, "step": 36590 }, { "grad_norm": 0.09788267314434052, "learning_rate": 6.107048257693772e-05, "loss": 0.0058, "step": 36600 }, { "grad_norm": 0.11517348140478134, "learning_rate": 6.105032622960683e-05, "loss": 0.007, "step": 36610 }, { "grad_norm": 0.07609164714813232, "learning_rate": 6.103016799407743e-05, "loss": 0.0057, "step": 36620 }, { "grad_norm": 0.10528570413589478, "learning_rate": 6.1010007873793984e-05, "loss": 0.0048, "step": 36630 }, { "grad_norm": 0.1329493224620819, "learning_rate": 6.098984587220131e-05, "loss": 0.005, "step": 36640 }, { "grad_norm": 0.12674130499362946, "learning_rate": 6.096968199274456e-05, "loss": 0.0069, "step": 36650 }, { "grad_norm": 0.1126563772559166, "learning_rate": 6.0949516238869166e-05, "loss": 0.0043, "step": 36660 }, { "grad_norm": 0.12763026356697083, "learning_rate": 6.092934861402092e-05, "loss": 0.0053, "step": 36670 }, { "grad_norm": 0.08941725641489029, "learning_rate": 6.0909179121645924e-05, "loss": 0.0061, "step": 36680 }, { "grad_norm": 0.10794370621442795, "learning_rate": 6.0889007765190576e-05, "loss": 0.007, "step": 36690 }, { "grad_norm": 0.09326696395874023, "learning_rate": 6.086883454810162e-05, "loss": 0.0058, "step": 36700 }, { "grad_norm": 0.09636040031909943, "learning_rate": 6.0848659473826084e-05, "loss": 0.0047, "step": 36710 }, { "grad_norm": 0.12441830337047577, "learning_rate": 6.082848254581138e-05, "loss": 0.006, "step": 36720 }, { "grad_norm": 0.09494519978761673, "learning_rate": 6.080830376750517e-05, "loss": 0.0053, "step": 36730 }, { "grad_norm": 0.08423857390880585, "learning_rate": 6.0788123142355445e-05, "loss": 0.0053, "step": 36740 }, { "grad_norm": 0.0944245308637619, "learning_rate": 6.076794067381052e-05, "loss": 0.0077, "step": 36750 }, { "grad_norm": 0.11584695428609848, "learning_rate": 6.074775636531905e-05, "loss": 0.008, "step": 36760 }, { "grad_norm": 0.09357872605323792, "learning_rate": 6.072757022032997e-05, "loss": 0.0044, "step": 36770 }, { "grad_norm": 0.10044287145137787, "learning_rate": 6.070738224229253e-05, "loss": 0.0056, "step": 36780 }, { "grad_norm": 0.09351596236228943, "learning_rate": 6.0687192434656314e-05, "loss": 0.0059, "step": 36790 }, { "grad_norm": 0.1131017878651619, "learning_rate": 6.066700080087121e-05, "loss": 0.0068, "step": 36800 }, { "grad_norm": 0.11809520423412323, "learning_rate": 6.0646807344387424e-05, "loss": 0.0047, "step": 36810 }, { "grad_norm": 0.1328526884317398, "learning_rate": 6.062661206865543e-05, "loss": 0.0068, "step": 36820 }, { "grad_norm": 0.13059304654598236, "learning_rate": 6.06064149771261e-05, "loss": 0.0081, "step": 36830 }, { "grad_norm": 0.15483056008815765, "learning_rate": 6.058621607325051e-05, "loss": 0.0059, "step": 36840 }, { "grad_norm": 0.12291513383388519, "learning_rate": 6.056601536048014e-05, "loss": 0.0065, "step": 36850 }, { "grad_norm": 0.13097184896469116, "learning_rate": 6.0545812842266725e-05, "loss": 0.0058, "step": 36860 }, { "grad_norm": 0.132433220744133, "learning_rate": 6.052560852206232e-05, "loss": 0.0056, "step": 36870 }, { "grad_norm": 0.1147381141781807, "learning_rate": 6.05054024033193e-05, "loss": 0.0065, "step": 36880 }, { "grad_norm": 0.11721406131982803, "learning_rate": 6.048519448949032e-05, "loss": 0.0054, "step": 36890 }, { "grad_norm": 0.14817707240581512, "learning_rate": 6.046498478402839e-05, "loss": 0.0056, "step": 36900 }, { "grad_norm": 0.13700927793979645, "learning_rate": 6.044477329038677e-05, "loss": 0.0092, "step": 36910 }, { "grad_norm": 0.10428560525178909, "learning_rate": 6.042456001201906e-05, "loss": 0.0049, "step": 36920 }, { "grad_norm": 0.09890741109848022, "learning_rate": 6.040434495237917e-05, "loss": 0.005, "step": 36930 }, { "grad_norm": 0.1546543538570404, "learning_rate": 6.0384128114921256e-05, "loss": 0.0065, "step": 36940 }, { "grad_norm": 0.12800949811935425, "learning_rate": 6.036390950309987e-05, "loss": 0.008, "step": 36950 }, { "grad_norm": 0.13619518280029297, "learning_rate": 6.0343689120369805e-05, "loss": 0.0082, "step": 36960 }, { "grad_norm": 0.1130853220820427, "learning_rate": 6.032346697018616e-05, "loss": 0.0062, "step": 36970 }, { "grad_norm": 0.08616665750741959, "learning_rate": 6.0303243056004375e-05, "loss": 0.0053, "step": 36980 }, { "grad_norm": 0.13712169229984283, "learning_rate": 6.0283017381280136e-05, "loss": 0.0049, "step": 36990 }, { "grad_norm": 0.09937847405672073, "learning_rate": 6.026278994946948e-05, "loss": 0.0071, "step": 37000 }, { "grad_norm": 0.14556512236595154, "learning_rate": 6.02425607640287e-05, "loss": 0.0058, "step": 37010 }, { "grad_norm": 0.10103484988212585, "learning_rate": 6.022232982841441e-05, "loss": 0.0052, "step": 37020 }, { "grad_norm": 0.10373767465353012, "learning_rate": 6.020209714608355e-05, "loss": 0.0055, "step": 37030 }, { "grad_norm": 0.09819290041923523, "learning_rate": 6.018186272049331e-05, "loss": 0.0079, "step": 37040 }, { "grad_norm": 0.11789806932210922, "learning_rate": 6.01616265551012e-05, "loss": 0.0064, "step": 37050 }, { "grad_norm": 0.10544770210981369, "learning_rate": 6.014138865336503e-05, "loss": 0.0073, "step": 37060 }, { "grad_norm": 0.09191467612981796, "learning_rate": 6.0121149018742905e-05, "loss": 0.0054, "step": 37070 }, { "grad_norm": 0.1285027414560318, "learning_rate": 6.010090765469325e-05, "loss": 0.0051, "step": 37080 }, { "grad_norm": 0.1402404010295868, "learning_rate": 6.008066456467473e-05, "loss": 0.0078, "step": 37090 }, { "grad_norm": 0.17498087882995605, "learning_rate": 6.0060419752146335e-05, "loss": 0.007, "step": 37100 }, { "grad_norm": 0.14063416421413422, "learning_rate": 6.0040173220567353e-05, "loss": 0.0059, "step": 37110 }, { "grad_norm": 0.12015987187623978, "learning_rate": 6.001992497339737e-05, "loss": 0.0046, "step": 37120 }, { "grad_norm": 0.10202109068632126, "learning_rate": 5.999967501409626e-05, "loss": 0.0049, "step": 37130 }, { "grad_norm": 0.08738668262958527, "learning_rate": 5.997942334612418e-05, "loss": 0.0042, "step": 37140 }, { "grad_norm": 0.09508261829614639, "learning_rate": 5.995916997294158e-05, "loss": 0.0059, "step": 37150 }, { "grad_norm": 0.10206039249897003, "learning_rate": 5.9938914898009214e-05, "loss": 0.0048, "step": 37160 }, { "grad_norm": 0.07848095893859863, "learning_rate": 5.991865812478813e-05, "loss": 0.0061, "step": 37170 }, { "grad_norm": 0.08442611992359161, "learning_rate": 5.989839965673964e-05, "loss": 0.0064, "step": 37180 }, { "grad_norm": 0.11539589613676071, "learning_rate": 5.987813949732539e-05, "loss": 0.0082, "step": 37190 }, { "grad_norm": 0.10676910728216171, "learning_rate": 5.9857877650007255e-05, "loss": 0.0058, "step": 37200 }, { "grad_norm": 0.07825785130262375, "learning_rate": 5.983761411824744e-05, "loss": 0.0082, "step": 37210 }, { "grad_norm": 0.1149170994758606, "learning_rate": 5.981734890550844e-05, "loss": 0.0061, "step": 37220 }, { "grad_norm": 0.09783732146024704, "learning_rate": 5.979708201525301e-05, "loss": 0.0069, "step": 37230 }, { "grad_norm": 0.10756868869066238, "learning_rate": 5.977681345094422e-05, "loss": 0.0048, "step": 37240 }, { "grad_norm": 0.11172902584075928, "learning_rate": 5.97565432160454e-05, "loss": 0.0059, "step": 37250 }, { "grad_norm": 0.15417815744876862, "learning_rate": 5.9736271314020186e-05, "loss": 0.0059, "step": 37260 }, { "grad_norm": 0.22834841907024384, "learning_rate": 5.971599774833251e-05, "loss": 0.0053, "step": 37270 }, { "grad_norm": 0.16207613050937653, "learning_rate": 5.9695722522446525e-05, "loss": 0.0069, "step": 37280 }, { "grad_norm": 0.1058015376329422, "learning_rate": 5.9675445639826765e-05, "loss": 0.0041, "step": 37290 }, { "grad_norm": 0.14352931082248688, "learning_rate": 5.965516710393796e-05, "loss": 0.0068, "step": 37300 }, { "grad_norm": 0.11767376959323883, "learning_rate": 5.963488691824516e-05, "loss": 0.0049, "step": 37310 }, { "grad_norm": 0.13268347084522247, "learning_rate": 5.96146050862137e-05, "loss": 0.0093, "step": 37320 }, { "grad_norm": 0.14430157840251923, "learning_rate": 5.959432161130919e-05, "loss": 0.0074, "step": 37330 }, { "grad_norm": 0.11629001796245575, "learning_rate": 5.9574036496997545e-05, "loss": 0.0067, "step": 37340 }, { "grad_norm": 0.09321969002485275, "learning_rate": 5.955374974674488e-05, "loss": 0.0065, "step": 37350 }, { "grad_norm": 0.09336128830909729, "learning_rate": 5.9533461364017696e-05, "loss": 0.0062, "step": 37360 }, { "grad_norm": 0.09833911061286926, "learning_rate": 5.9513171352282716e-05, "loss": 0.0047, "step": 37370 }, { "grad_norm": 0.13827945291996002, "learning_rate": 5.949287971500692e-05, "loss": 0.0052, "step": 37380 }, { "grad_norm": 0.08813643455505371, "learning_rate": 5.947258645565762e-05, "loss": 0.0064, "step": 37390 }, { "grad_norm": 0.1286672055721283, "learning_rate": 5.945229157770237e-05, "loss": 0.0056, "step": 37400 }, { "grad_norm": 0.10104095935821533, "learning_rate": 5.9431995084609006e-05, "loss": 0.0066, "step": 37410 }, { "grad_norm": 0.11151748895645142, "learning_rate": 5.941169697984564e-05, "loss": 0.0056, "step": 37420 }, { "grad_norm": 0.0985148623585701, "learning_rate": 5.9391397266880675e-05, "loss": 0.0046, "step": 37430 }, { "grad_norm": 0.11666695773601532, "learning_rate": 5.937109594918279e-05, "loss": 0.0075, "step": 37440 }, { "grad_norm": 0.13157783448696136, "learning_rate": 5.9350793030220884e-05, "loss": 0.0056, "step": 37450 }, { "grad_norm": 0.11777343600988388, "learning_rate": 5.933048851346421e-05, "loss": 0.0069, "step": 37460 }, { "grad_norm": 0.13353288173675537, "learning_rate": 5.931018240238224e-05, "loss": 0.0068, "step": 37470 }, { "grad_norm": 0.19897234439849854, "learning_rate": 5.928987470044471e-05, "loss": 0.0052, "step": 37480 }, { "grad_norm": 0.13081854581832886, "learning_rate": 5.9269565411121695e-05, "loss": 0.0063, "step": 37490 }, { "grad_norm": 0.1511060893535614, "learning_rate": 5.924925453788347e-05, "loss": 0.0061, "step": 37500 }, { "grad_norm": 0.11055292934179306, "learning_rate": 5.92289420842006e-05, "loss": 0.0059, "step": 37510 }, { "grad_norm": 0.09573431313037872, "learning_rate": 5.9208628053543945e-05, "loss": 0.0071, "step": 37520 }, { "grad_norm": 0.141632080078125, "learning_rate": 5.918831244938462e-05, "loss": 0.0066, "step": 37530 }, { "grad_norm": 0.11411077529191971, "learning_rate": 5.9167995275194e-05, "loss": 0.0054, "step": 37540 }, { "grad_norm": 0.11256667226552963, "learning_rate": 5.914767653444373e-05, "loss": 0.0056, "step": 37550 }, { "grad_norm": 0.08401435613632202, "learning_rate": 5.912735623060572e-05, "loss": 0.0055, "step": 37560 }, { "grad_norm": 0.09673341363668442, "learning_rate": 5.910703436715217e-05, "loss": 0.0074, "step": 37570 }, { "grad_norm": 0.11576376110315323, "learning_rate": 5.908671094755552e-05, "loss": 0.0053, "step": 37580 }, { "grad_norm": 0.17839424312114716, "learning_rate": 5.906638597528851e-05, "loss": 0.0052, "step": 37590 }, { "grad_norm": 0.07917571067810059, "learning_rate": 5.9046059453824076e-05, "loss": 0.0037, "step": 37600 }, { "grad_norm": 0.11280021071434021, "learning_rate": 5.9025731386635505e-05, "loss": 0.0051, "step": 37610 }, { "grad_norm": 0.0958729237318039, "learning_rate": 5.900540177719629e-05, "loss": 0.0047, "step": 37620 }, { "grad_norm": 0.10196629911661148, "learning_rate": 5.898507062898021e-05, "loss": 0.0059, "step": 37630 }, { "grad_norm": 0.10788202285766602, "learning_rate": 5.8964737945461316e-05, "loss": 0.0052, "step": 37640 }, { "grad_norm": 0.10579739511013031, "learning_rate": 5.8944403730113885e-05, "loss": 0.0068, "step": 37650 }, { "grad_norm": 0.08152730762958527, "learning_rate": 5.892406798641248e-05, "loss": 0.0076, "step": 37660 }, { "grad_norm": 0.1293915957212448, "learning_rate": 5.890373071783193e-05, "loss": 0.0097, "step": 37670 }, { "grad_norm": 0.13431206345558167, "learning_rate": 5.888339192784732e-05, "loss": 0.008, "step": 37680 }, { "grad_norm": 0.1345987617969513, "learning_rate": 5.8863051619934003e-05, "loss": 0.0076, "step": 37690 }, { "grad_norm": 0.10608742386102676, "learning_rate": 5.8842709797567554e-05, "loss": 0.0057, "step": 37700 }, { "grad_norm": 0.11818996071815491, "learning_rate": 5.8822366464223855e-05, "loss": 0.0067, "step": 37710 }, { "grad_norm": 0.09794426709413528, "learning_rate": 5.880202162337901e-05, "loss": 0.0055, "step": 37720 }, { "grad_norm": 0.08837119489908218, "learning_rate": 5.8781675278509405e-05, "loss": 0.0046, "step": 37730 }, { "grad_norm": 0.11799014359712601, "learning_rate": 5.8761327433091696e-05, "loss": 0.0062, "step": 37740 }, { "grad_norm": 0.09312669187784195, "learning_rate": 5.874097809060275e-05, "loss": 0.0053, "step": 37750 }, { "grad_norm": 0.10779466480016708, "learning_rate": 5.87206272545197e-05, "loss": 0.0053, "step": 37760 }, { "grad_norm": 0.10351727902889252, "learning_rate": 5.8700274928319955e-05, "loss": 0.0047, "step": 37770 }, { "grad_norm": 0.13241328299045563, "learning_rate": 5.867992111548118e-05, "loss": 0.0068, "step": 37780 }, { "grad_norm": 0.11523287743330002, "learning_rate": 5.865956581948131e-05, "loss": 0.0046, "step": 37790 }, { "grad_norm": 0.09947049617767334, "learning_rate": 5.863920904379845e-05, "loss": 0.0049, "step": 37800 }, { "grad_norm": 0.07354239374399185, "learning_rate": 5.8618850791911064e-05, "loss": 0.0062, "step": 37810 }, { "grad_norm": 0.10853035748004913, "learning_rate": 5.859849106729779e-05, "loss": 0.0054, "step": 37820 }, { "grad_norm": 0.14359880983829498, "learning_rate": 5.857812987343758e-05, "loss": 0.004, "step": 37830 }, { "grad_norm": 0.1170760989189148, "learning_rate": 5.855776721380957e-05, "loss": 0.0053, "step": 37840 }, { "grad_norm": 0.09993131458759308, "learning_rate": 5.8537403091893217e-05, "loss": 0.0061, "step": 37850 }, { "grad_norm": 0.07813096046447754, "learning_rate": 5.851703751116816e-05, "loss": 0.0051, "step": 37860 }, { "grad_norm": 0.12377522140741348, "learning_rate": 5.8496670475114336e-05, "loss": 0.0045, "step": 37870 }, { "grad_norm": 0.08557409048080444, "learning_rate": 5.84763019872119e-05, "loss": 0.0052, "step": 37880 }, { "grad_norm": 0.09173883497714996, "learning_rate": 5.845593205094131e-05, "loss": 0.0053, "step": 37890 }, { "grad_norm": 0.10267133265733719, "learning_rate": 5.843556066978318e-05, "loss": 0.0042, "step": 37900 }, { "grad_norm": 0.09891466796398163, "learning_rate": 5.8415187847218455e-05, "loss": 0.0046, "step": 37910 }, { "grad_norm": 0.12294520437717438, "learning_rate": 5.839481358672827e-05, "loss": 0.0058, "step": 37920 }, { "grad_norm": 0.12706148624420166, "learning_rate": 5.837443789179407e-05, "loss": 0.006, "step": 37930 }, { "grad_norm": 0.10284029692411423, "learning_rate": 5.8354060765897445e-05, "loss": 0.0052, "step": 37940 }, { "grad_norm": 0.08741800487041473, "learning_rate": 5.8333682212520334e-05, "loss": 0.0051, "step": 37950 }, { "grad_norm": 0.09257597476243973, "learning_rate": 5.831330223514486e-05, "loss": 0.0054, "step": 37960 }, { "grad_norm": 0.12388670444488525, "learning_rate": 5.8292920837253396e-05, "loss": 0.0059, "step": 37970 }, { "grad_norm": 0.14768187701702118, "learning_rate": 5.827253802232857e-05, "loss": 0.006, "step": 37980 }, { "grad_norm": 0.10476501286029816, "learning_rate": 5.825215379385327e-05, "loss": 0.0046, "step": 37990 }, { "grad_norm": 0.12711690366268158, "learning_rate": 5.823176815531057e-05, "loss": 0.0081, "step": 38000 }, { "grad_norm": 0.10867907106876373, "learning_rate": 5.8211381110183826e-05, "loss": 0.0055, "step": 38010 }, { "grad_norm": 0.1340719759464264, "learning_rate": 5.8190992661956645e-05, "loss": 0.0053, "step": 38020 }, { "grad_norm": 0.1507115215063095, "learning_rate": 5.817060281411284e-05, "loss": 0.004, "step": 38030 }, { "grad_norm": 0.0898575633764267, "learning_rate": 5.815021157013647e-05, "loss": 0.0047, "step": 38040 }, { "grad_norm": 0.11913049966096878, "learning_rate": 5.8129818933511856e-05, "loss": 0.0055, "step": 38050 }, { "grad_norm": 0.08895684033632278, "learning_rate": 5.8109424907723544e-05, "loss": 0.0045, "step": 38060 }, { "grad_norm": 0.12408197671175003, "learning_rate": 5.80890294962563e-05, "loss": 0.0041, "step": 38070 }, { "grad_norm": 0.13686779141426086, "learning_rate": 5.806863270259515e-05, "loss": 0.0049, "step": 38080 }, { "grad_norm": 0.09511645883321762, "learning_rate": 5.804823453022536e-05, "loss": 0.0072, "step": 38090 }, { "grad_norm": 0.10019544512033463, "learning_rate": 5.80278349826324e-05, "loss": 0.0072, "step": 38100 }, { "grad_norm": 0.09202519804239273, "learning_rate": 5.8007434063302014e-05, "loss": 0.005, "step": 38110 }, { "grad_norm": 0.2542141079902649, "learning_rate": 5.7987031775720136e-05, "loss": 0.0062, "step": 38120 }, { "grad_norm": 0.12892033159732819, "learning_rate": 5.7966628123372976e-05, "loss": 0.0078, "step": 38130 }, { "grad_norm": 0.11183182895183563, "learning_rate": 5.7946223109746956e-05, "loss": 0.0042, "step": 38140 }, { "grad_norm": 0.12645414471626282, "learning_rate": 5.7925816738328754e-05, "loss": 0.0046, "step": 38150 }, { "grad_norm": 0.09001846611499786, "learning_rate": 5.790540901260521e-05, "loss": 0.0056, "step": 38160 }, { "grad_norm": 0.11238036304712296, "learning_rate": 5.788499993606351e-05, "loss": 0.0071, "step": 38170 }, { "grad_norm": 0.10627032816410065, "learning_rate": 5.786458951219096e-05, "loss": 0.0046, "step": 38180 }, { "grad_norm": 0.08802805840969086, "learning_rate": 5.784417774447517e-05, "loss": 0.006, "step": 38190 }, { "grad_norm": 0.10352794080972672, "learning_rate": 5.782376463640393e-05, "loss": 0.0052, "step": 38200 }, { "grad_norm": 0.10387017577886581, "learning_rate": 5.780335019146531e-05, "loss": 0.0055, "step": 38210 }, { "grad_norm": 0.13458703458309174, "learning_rate": 5.778293441314755e-05, "loss": 0.0052, "step": 38220 }, { "grad_norm": 0.09981469810009003, "learning_rate": 5.776251730493917e-05, "loss": 0.0067, "step": 38230 }, { "grad_norm": 0.1063244417309761, "learning_rate": 5.774209887032887e-05, "loss": 0.0052, "step": 38240 }, { "grad_norm": 0.1519206315279007, "learning_rate": 5.772167911280565e-05, "loss": 0.0068, "step": 38250 }, { "grad_norm": 0.1672995686531067, "learning_rate": 5.770125803585864e-05, "loss": 0.0059, "step": 38260 }, { "grad_norm": 0.1288476586341858, "learning_rate": 5.768083564297726e-05, "loss": 0.0054, "step": 38270 }, { "grad_norm": 0.1382034718990326, "learning_rate": 5.766041193765114e-05, "loss": 0.0053, "step": 38280 }, { "grad_norm": 0.13764113187789917, "learning_rate": 5.763998692337015e-05, "loss": 0.0058, "step": 38290 }, { "grad_norm": 0.09192713350057602, "learning_rate": 5.761956060362433e-05, "loss": 0.0056, "step": 38300 }, { "grad_norm": 0.11527059227228165, "learning_rate": 5.7599132981904e-05, "loss": 0.0058, "step": 38310 }, { "grad_norm": 0.09744983166456223, "learning_rate": 5.75787040616997e-05, "loss": 0.0045, "step": 38320 }, { "grad_norm": 0.12177260965108871, "learning_rate": 5.755827384650212e-05, "loss": 0.0069, "step": 38330 }, { "grad_norm": 0.09166449308395386, "learning_rate": 5.753784233980228e-05, "loss": 0.0086, "step": 38340 }, { "grad_norm": 0.13360092043876648, "learning_rate": 5.751740954509135e-05, "loss": 0.0063, "step": 38350 }, { "grad_norm": 0.11717817187309265, "learning_rate": 5.7496975465860715e-05, "loss": 0.0057, "step": 38360 }, { "grad_norm": 0.16083309054374695, "learning_rate": 5.747654010560202e-05, "loss": 0.0063, "step": 38370 }, { "grad_norm": 0.1895354837179184, "learning_rate": 5.7456103467807097e-05, "loss": 0.0057, "step": 38380 }, { "grad_norm": 0.13271693885326385, "learning_rate": 5.7435665555968046e-05, "loss": 0.0051, "step": 38390 }, { "grad_norm": 0.11983364820480347, "learning_rate": 5.74152263735771e-05, "loss": 0.0048, "step": 38400 }, { "grad_norm": 0.1480664312839508, "learning_rate": 5.739478592412677e-05, "loss": 0.0066, "step": 38410 }, { "grad_norm": 0.11797454953193665, "learning_rate": 5.7374344211109766e-05, "loss": 0.0069, "step": 38420 }, { "grad_norm": 0.12498890608549118, "learning_rate": 5.735390123801904e-05, "loss": 0.0057, "step": 38430 }, { "grad_norm": 0.1048809364438057, "learning_rate": 5.7333457008347704e-05, "loss": 0.0052, "step": 38440 }, { "grad_norm": 0.10139677673578262, "learning_rate": 5.7313011525589156e-05, "loss": 0.0056, "step": 38450 }, { "grad_norm": 0.09651999175548553, "learning_rate": 5.729256479323694e-05, "loss": 0.0068, "step": 38460 }, { "grad_norm": 0.15195995569229126, "learning_rate": 5.727211681478485e-05, "loss": 0.0065, "step": 38470 }, { "grad_norm": 0.09437990933656693, "learning_rate": 5.7251667593726886e-05, "loss": 0.0054, "step": 38480 }, { "grad_norm": 0.1386851668357849, "learning_rate": 5.723121713355728e-05, "loss": 0.0062, "step": 38490 }, { "grad_norm": 0.1015835776925087, "learning_rate": 5.721076543777044e-05, "loss": 0.0074, "step": 38500 }, { "grad_norm": 0.1148638054728508, "learning_rate": 5.7190312509860986e-05, "loss": 0.0072, "step": 38510 }, { "grad_norm": 0.16180577874183655, "learning_rate": 5.716985835332379e-05, "loss": 0.0054, "step": 38520 }, { "grad_norm": 0.1396602839231491, "learning_rate": 5.714940297165389e-05, "loss": 0.0052, "step": 38530 }, { "grad_norm": 0.1076541319489479, "learning_rate": 5.712894636834656e-05, "loss": 0.0067, "step": 38540 }, { "grad_norm": 0.1148613765835762, "learning_rate": 5.7108488546897276e-05, "loss": 0.0082, "step": 38550 }, { "grad_norm": 0.13857197761535645, "learning_rate": 5.708802951080172e-05, "loss": 0.0051, "step": 38560 }, { "grad_norm": 0.09946268796920776, "learning_rate": 5.706756926355576e-05, "loss": 0.0059, "step": 38570 }, { "grad_norm": 0.12204092741012573, "learning_rate": 5.704710780865554e-05, "loss": 0.0061, "step": 38580 }, { "grad_norm": 0.10066212713718414, "learning_rate": 5.7026645149597325e-05, "loss": 0.0066, "step": 38590 }, { "grad_norm": 0.09048330783843994, "learning_rate": 5.700618128987764e-05, "loss": 0.0058, "step": 38600 }, { "grad_norm": 0.12205114960670471, "learning_rate": 5.698571623299317e-05, "loss": 0.0059, "step": 38610 }, { "grad_norm": 0.10123937577009201, "learning_rate": 5.696524998244086e-05, "loss": 0.0081, "step": 38620 }, { "grad_norm": 0.11809827387332916, "learning_rate": 5.6944782541717836e-05, "loss": 0.0053, "step": 38630 }, { "grad_norm": 0.11032017320394516, "learning_rate": 5.69243139143214e-05, "loss": 0.0057, "step": 38640 }, { "grad_norm": 0.10561713576316833, "learning_rate": 5.6903844103749125e-05, "loss": 0.0062, "step": 38650 }, { "grad_norm": 0.18175004422664642, "learning_rate": 5.688337311349869e-05, "loss": 0.0057, "step": 38660 }, { "grad_norm": 0.13463765382766724, "learning_rate": 5.6862900947068074e-05, "loss": 0.0054, "step": 38670 }, { "grad_norm": 0.12163037806749344, "learning_rate": 5.6842427607955374e-05, "loss": 0.0059, "step": 38680 }, { "grad_norm": 0.11572253704071045, "learning_rate": 5.682195309965893e-05, "loss": 0.0059, "step": 38690 }, { "grad_norm": 0.11143351346254349, "learning_rate": 5.6801477425677294e-05, "loss": 0.0049, "step": 38700 }, { "grad_norm": 0.1294214278459549, "learning_rate": 5.678100058950917e-05, "loss": 0.0058, "step": 38710 }, { "grad_norm": 0.13959702849388123, "learning_rate": 5.676052259465352e-05, "loss": 0.0064, "step": 38720 }, { "grad_norm": 0.17885053157806396, "learning_rate": 5.674004344460945e-05, "loss": 0.0071, "step": 38730 }, { "grad_norm": 0.1264467090368271, "learning_rate": 5.6719563142876295e-05, "loss": 0.0062, "step": 38740 }, { "grad_norm": 0.12918847799301147, "learning_rate": 5.669908169295359e-05, "loss": 0.0068, "step": 38750 }, { "grad_norm": 0.1132354587316513, "learning_rate": 5.667859909834105e-05, "loss": 0.0047, "step": 38760 }, { "grad_norm": 0.12717746198177338, "learning_rate": 5.6658115362538565e-05, "loss": 0.0067, "step": 38770 }, { "grad_norm": 0.09492313861846924, "learning_rate": 5.663763048904628e-05, "loss": 0.0055, "step": 38780 }, { "grad_norm": 0.11349866539239883, "learning_rate": 5.661714448136447e-05, "loss": 0.005, "step": 38790 }, { "grad_norm": 0.09188491851091385, "learning_rate": 5.659665734299366e-05, "loss": 0.0057, "step": 38800 }, { "grad_norm": 0.10303585976362228, "learning_rate": 5.6576169077434516e-05, "loss": 0.0056, "step": 38810 }, { "grad_norm": 0.09737834334373474, "learning_rate": 5.6555679688187944e-05, "loss": 0.0065, "step": 38820 }, { "grad_norm": 0.12427139282226562, "learning_rate": 5.6535189178755e-05, "loss": 0.0063, "step": 38830 }, { "grad_norm": 0.08713073283433914, "learning_rate": 5.651469755263698e-05, "loss": 0.006, "step": 38840 }, { "grad_norm": 0.08866796642541885, "learning_rate": 5.6494204813335316e-05, "loss": 0.0052, "step": 38850 }, { "grad_norm": 0.08499416708946228, "learning_rate": 5.647371096435168e-05, "loss": 0.0078, "step": 38860 }, { "grad_norm": 0.14066217839717865, "learning_rate": 5.645321600918788e-05, "loss": 0.0056, "step": 38870 }, { "grad_norm": 0.15445423126220703, "learning_rate": 5.643271995134597e-05, "loss": 0.0058, "step": 38880 }, { "grad_norm": 0.09278527647256851, "learning_rate": 5.641222279432814e-05, "loss": 0.0055, "step": 38890 }, { "grad_norm": 0.12994198501110077, "learning_rate": 5.6391724541636834e-05, "loss": 0.0067, "step": 38900 }, { "grad_norm": 0.12376018613576889, "learning_rate": 5.6371225196774605e-05, "loss": 0.0058, "step": 38910 }, { "grad_norm": 0.10877134650945663, "learning_rate": 5.635072476324423e-05, "loss": 0.0049, "step": 38920 }, { "grad_norm": 0.10218673944473267, "learning_rate": 5.63302232445487e-05, "loss": 0.0043, "step": 38930 }, { "grad_norm": 0.08936423063278198, "learning_rate": 5.6309720644191144e-05, "loss": 0.0051, "step": 38940 }, { "grad_norm": 0.13233348727226257, "learning_rate": 5.628921696567491e-05, "loss": 0.0072, "step": 38950 }, { "grad_norm": 0.1160845085978508, "learning_rate": 5.62687122125035e-05, "loss": 0.006, "step": 38960 }, { "grad_norm": 0.11260562390089035, "learning_rate": 5.624820638818062e-05, "loss": 0.0047, "step": 38970 }, { "grad_norm": 0.13690243661403656, "learning_rate": 5.6227699496210164e-05, "loss": 0.0053, "step": 38980 }, { "grad_norm": 0.0943179726600647, "learning_rate": 5.6207191540096195e-05, "loss": 0.0057, "step": 38990 }, { "grad_norm": 0.123174749314785, "learning_rate": 5.618668252334296e-05, "loss": 0.005, "step": 39000 }, { "grad_norm": 0.21085584163665771, "learning_rate": 5.616617244945488e-05, "loss": 0.0072, "step": 39010 }, { "grad_norm": 0.1048666462302208, "learning_rate": 5.614566132193656e-05, "loss": 0.005, "step": 39020 }, { "grad_norm": 0.13922494649887085, "learning_rate": 5.612514914429282e-05, "loss": 0.0073, "step": 39030 }, { "grad_norm": 0.11118263006210327, "learning_rate": 5.610463592002863e-05, "loss": 0.0077, "step": 39040 }, { "grad_norm": 0.08605330437421799, "learning_rate": 5.608412165264909e-05, "loss": 0.0044, "step": 39050 }, { "grad_norm": 0.07485409826040268, "learning_rate": 5.606360634565959e-05, "loss": 0.0066, "step": 39060 }, { "grad_norm": 0.1276780515909195, "learning_rate": 5.604309000256559e-05, "loss": 0.0066, "step": 39070 }, { "grad_norm": 0.10332977026700974, "learning_rate": 5.6022572626872785e-05, "loss": 0.0052, "step": 39080 }, { "grad_norm": 0.15742932260036469, "learning_rate": 5.600205422208704e-05, "loss": 0.0062, "step": 39090 }, { "grad_norm": 0.07148842513561249, "learning_rate": 5.59815347917144e-05, "loss": 0.0069, "step": 39100 }, { "grad_norm": 0.16288639605045319, "learning_rate": 5.596101433926103e-05, "loss": 0.005, "step": 39110 }, { "grad_norm": 0.1426219940185547, "learning_rate": 5.5940492868233364e-05, "loss": 0.0083, "step": 39120 }, { "grad_norm": 0.1274726241827011, "learning_rate": 5.591997038213793e-05, "loss": 0.007, "step": 39130 }, { "grad_norm": 0.16095511615276337, "learning_rate": 5.5899446884481475e-05, "loss": 0.0066, "step": 39140 }, { "grad_norm": 0.0802743211388588, "learning_rate": 5.5878922378770906e-05, "loss": 0.0041, "step": 39150 }, { "grad_norm": 0.14316882193088531, "learning_rate": 5.5858396868513285e-05, "loss": 0.005, "step": 39160 }, { "grad_norm": 0.07357762008905411, "learning_rate": 5.583787035721586e-05, "loss": 0.0041, "step": 39170 }, { "grad_norm": 0.10014133155345917, "learning_rate": 5.581734284838606e-05, "loss": 0.0043, "step": 39180 }, { "grad_norm": 0.09278934448957443, "learning_rate": 5.579681434553147e-05, "loss": 0.0052, "step": 39190 }, { "grad_norm": 0.14410734176635742, "learning_rate": 5.5776284852159854e-05, "loss": 0.0049, "step": 39200 }, { "grad_norm": 0.12421439588069916, "learning_rate": 5.575575437177913e-05, "loss": 0.0056, "step": 39210 }, { "grad_norm": 0.11589460074901581, "learning_rate": 5.573522290789742e-05, "loss": 0.0059, "step": 39220 }, { "grad_norm": 0.09705767035484314, "learning_rate": 5.571469046402298e-05, "loss": 0.0051, "step": 39230 }, { "grad_norm": 0.16920694708824158, "learning_rate": 5.5694157043664205e-05, "loss": 0.0087, "step": 39240 }, { "grad_norm": 0.0783579871058464, "learning_rate": 5.567362265032975e-05, "loss": 0.0037, "step": 39250 }, { "grad_norm": 0.10943212360143661, "learning_rate": 5.565308728752836e-05, "loss": 0.0041, "step": 39260 }, { "grad_norm": 0.13761495053768158, "learning_rate": 5.5632550958768956e-05, "loss": 0.0055, "step": 39270 }, { "grad_norm": 0.10427641868591309, "learning_rate": 5.5612013667560636e-05, "loss": 0.0067, "step": 39280 }, { "grad_norm": 0.08609066158533096, "learning_rate": 5.5591475417412676e-05, "loss": 0.0078, "step": 39290 }, { "grad_norm": 0.1004895269870758, "learning_rate": 5.557093621183451e-05, "loss": 0.0048, "step": 39300 }, { "grad_norm": 0.1013583391904831, "learning_rate": 5.55503960543357e-05, "loss": 0.0061, "step": 39310 }, { "grad_norm": 0.1307770013809204, "learning_rate": 5.552985494842601e-05, "loss": 0.0054, "step": 39320 }, { "grad_norm": 0.0828033834695816, "learning_rate": 5.550931289761534e-05, "loss": 0.0065, "step": 39330 }, { "grad_norm": 0.09330964833498001, "learning_rate": 5.548876990541378e-05, "loss": 0.0045, "step": 39340 }, { "grad_norm": 0.14984364807605743, "learning_rate": 5.5468225975331564e-05, "loss": 0.0065, "step": 39350 }, { "grad_norm": 0.15237267315387726, "learning_rate": 5.544768111087909e-05, "loss": 0.0061, "step": 39360 }, { "grad_norm": 0.13363225758075714, "learning_rate": 5.54271353155669e-05, "loss": 0.0045, "step": 39370 }, { "grad_norm": 0.10859384387731552, "learning_rate": 5.5406588592905715e-05, "loss": 0.0048, "step": 39380 }, { "grad_norm": 0.08465974777936935, "learning_rate": 5.5386040946406416e-05, "loss": 0.0053, "step": 39390 }, { "grad_norm": 0.10948667675256729, "learning_rate": 5.536549237958004e-05, "loss": 0.0043, "step": 39400 }, { "grad_norm": 0.11408185958862305, "learning_rate": 5.5344942895937744e-05, "loss": 0.005, "step": 39410 }, { "grad_norm": 0.08422670513391495, "learning_rate": 5.5324392498990904e-05, "loss": 0.0058, "step": 39420 }, { "grad_norm": 0.1441541314125061, "learning_rate": 5.5303841192251e-05, "loss": 0.0054, "step": 39430 }, { "grad_norm": 0.09813977777957916, "learning_rate": 5.52832889792297e-05, "loss": 0.0061, "step": 39440 }, { "grad_norm": 0.11167572438716888, "learning_rate": 5.526273586343881e-05, "loss": 0.0058, "step": 39450 }, { "grad_norm": 0.1359296441078186, "learning_rate": 5.5242181848390306e-05, "loss": 0.0054, "step": 39460 }, { "grad_norm": 0.1200910210609436, "learning_rate": 5.5221626937596285e-05, "loss": 0.0046, "step": 39470 }, { "grad_norm": 0.10959915816783905, "learning_rate": 5.520107113456903e-05, "loss": 0.0047, "step": 39480 }, { "grad_norm": 0.13259956240653992, "learning_rate": 5.5180514442820974e-05, "loss": 0.0046, "step": 39490 }, { "grad_norm": 0.10698435455560684, "learning_rate": 5.515995686586469e-05, "loss": 0.0046, "step": 39500 }, { "grad_norm": 0.19127912819385529, "learning_rate": 5.5139398407212916e-05, "loss": 0.0089, "step": 39510 }, { "grad_norm": 0.13578838109970093, "learning_rate": 5.511883907037849e-05, "loss": 0.0067, "step": 39520 }, { "grad_norm": 0.07168131321668625, "learning_rate": 5.509827885887449e-05, "loss": 0.0043, "step": 39530 }, { "grad_norm": 0.10643479228019714, "learning_rate": 5.507771777621406e-05, "loss": 0.0057, "step": 39540 }, { "grad_norm": 0.11851809173822403, "learning_rate": 5.505715582591052e-05, "loss": 0.0053, "step": 39550 }, { "grad_norm": 0.09071749448776245, "learning_rate": 5.50365930114774e-05, "loss": 0.0046, "step": 39560 }, { "grad_norm": 0.11636558920145035, "learning_rate": 5.5016029336428255e-05, "loss": 0.0052, "step": 39570 }, { "grad_norm": 0.0980425626039505, "learning_rate": 5.49954648042769e-05, "loss": 0.0063, "step": 39580 }, { "grad_norm": 0.11863236874341965, "learning_rate": 5.4974899418537226e-05, "loss": 0.0068, "step": 39590 }, { "grad_norm": 0.12690339982509613, "learning_rate": 5.4954333182723316e-05, "loss": 0.0047, "step": 39600 }, { "grad_norm": 0.10176634043455124, "learning_rate": 5.493376610034937e-05, "loss": 0.0061, "step": 39610 }, { "grad_norm": 0.13746316730976105, "learning_rate": 5.4913198174929735e-05, "loss": 0.0044, "step": 39620 }, { "grad_norm": 0.13738271594047546, "learning_rate": 5.489262940997889e-05, "loss": 0.0054, "step": 39630 }, { "grad_norm": 0.15045219659805298, "learning_rate": 5.487205980901151e-05, "loss": 0.0048, "step": 39640 }, { "grad_norm": 0.09968002885580063, "learning_rate": 5.485148937554234e-05, "loss": 0.0037, "step": 39650 }, { "grad_norm": 0.1199556440114975, "learning_rate": 5.483091811308635e-05, "loss": 0.0053, "step": 39660 }, { "grad_norm": 0.11283005774021149, "learning_rate": 5.4810346025158564e-05, "loss": 0.0052, "step": 39670 }, { "grad_norm": 0.11575522273778915, "learning_rate": 5.478977311527421e-05, "loss": 0.0058, "step": 39680 }, { "grad_norm": 0.10742393136024475, "learning_rate": 5.476919938694863e-05, "loss": 0.0057, "step": 39690 }, { "grad_norm": 0.10278838872909546, "learning_rate": 5.474862484369733e-05, "loss": 0.0061, "step": 39700 }, { "grad_norm": 0.11688017100095749, "learning_rate": 5.472804948903589e-05, "loss": 0.0068, "step": 39710 }, { "grad_norm": 0.07686378061771393, "learning_rate": 5.470747332648013e-05, "loss": 0.0048, "step": 39720 }, { "grad_norm": 0.11253542453050613, "learning_rate": 5.468689635954591e-05, "loss": 0.0057, "step": 39730 }, { "grad_norm": 0.13933274149894714, "learning_rate": 5.46663185917493e-05, "loss": 0.0077, "step": 39740 }, { "grad_norm": 0.14415597915649414, "learning_rate": 5.464574002660645e-05, "loss": 0.0049, "step": 39750 }, { "grad_norm": 0.15201959013938904, "learning_rate": 5.4625160667633724e-05, "loss": 0.0063, "step": 39760 }, { "grad_norm": 0.09477008879184723, "learning_rate": 5.4604580518347505e-05, "loss": 0.005, "step": 39770 }, { "grad_norm": 0.09794244170188904, "learning_rate": 5.458399958226442e-05, "loss": 0.0047, "step": 39780 }, { "grad_norm": 0.09299091249704361, "learning_rate": 5.456341786290119e-05, "loss": 0.0064, "step": 39790 }, { "grad_norm": 0.09825598448514938, "learning_rate": 5.454283536377465e-05, "loss": 0.0065, "step": 39800 }, { "grad_norm": 0.12121113389730453, "learning_rate": 5.452225208840179e-05, "loss": 0.0056, "step": 39810 }, { "grad_norm": 0.12579426169395447, "learning_rate": 5.450166804029975e-05, "loss": 0.0062, "step": 39820 }, { "grad_norm": 0.09286539256572723, "learning_rate": 5.448108322298574e-05, "loss": 0.0049, "step": 39830 }, { "grad_norm": 0.11105462163686752, "learning_rate": 5.446049763997717e-05, "loss": 0.0062, "step": 39840 }, { "grad_norm": 0.10635166615247726, "learning_rate": 5.4439911294791546e-05, "loss": 0.0066, "step": 39850 }, { "grad_norm": 0.11695592850446701, "learning_rate": 5.441932419094652e-05, "loss": 0.0055, "step": 39860 }, { "grad_norm": 0.12554852664470673, "learning_rate": 5.439873633195985e-05, "loss": 0.0047, "step": 39870 }, { "grad_norm": 0.08204241096973419, "learning_rate": 5.437814772134947e-05, "loss": 0.0056, "step": 39880 }, { "grad_norm": 0.13193204998970032, "learning_rate": 5.4357558362633366e-05, "loss": 0.0069, "step": 39890 }, { "grad_norm": 0.10106369853019714, "learning_rate": 5.4336968259329726e-05, "loss": 0.0065, "step": 39900 }, { "grad_norm": 0.09931599348783493, "learning_rate": 5.431637741495681e-05, "loss": 0.0054, "step": 39910 }, { "grad_norm": 0.07174727320671082, "learning_rate": 5.429578583303307e-05, "loss": 0.0064, "step": 39920 }, { "grad_norm": 0.08575862646102905, "learning_rate": 5.427519351707701e-05, "loss": 0.0062, "step": 39930 }, { "grad_norm": 0.08801109343767166, "learning_rate": 5.42546004706073e-05, "loss": 0.0047, "step": 39940 }, { "grad_norm": 0.07979391515254974, "learning_rate": 5.4234006697142735e-05, "loss": 0.0042, "step": 39950 }, { "grad_norm": 0.11049610376358032, "learning_rate": 5.421341220020224e-05, "loss": 0.0056, "step": 39960 }, { "grad_norm": 0.07135337591171265, "learning_rate": 5.419281698330482e-05, "loss": 0.0042, "step": 39970 }, { "grad_norm": 0.08262263238430023, "learning_rate": 5.4172221049969665e-05, "loss": 0.0055, "step": 39980 }, { "grad_norm": 0.11894860863685608, "learning_rate": 5.415162440371604e-05, "loss": 0.005, "step": 39990 }, { "grad_norm": 0.12138494849205017, "learning_rate": 5.413102704806334e-05, "loss": 0.0048, "step": 40000 } ], "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": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 48, "trial_name": null, "trial_params": null }