{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.0123779773712158, "learning_rate": 1.8e-07, "loss": 1.5772, "step": 10 }, { "grad_norm": 1.2592813968658447, "learning_rate": 3.8e-07, "loss": 1.5971, "step": 20 }, { "grad_norm": 1.0761749744415283, "learning_rate": 5.8e-07, "loss": 1.5405, "step": 30 }, { "grad_norm": 1.417466163635254, "learning_rate": 7.8e-07, "loss": 1.5672, "step": 40 }, { "grad_norm": 1.114725112915039, "learning_rate": 9.8e-07, "loss": 1.5523, "step": 50 }, { "grad_norm": 1.4429272413253784, "learning_rate": 1.18e-06, "loss": 1.5279, "step": 60 }, { "grad_norm": 1.815200924873352, "learning_rate": 1.3800000000000001e-06, "loss": 1.4263, "step": 70 }, { "grad_norm": 1.7818489074707031, "learning_rate": 1.5800000000000003e-06, "loss": 1.3722, "step": 80 }, { "grad_norm": 1.2510874271392822, "learning_rate": 1.7800000000000001e-06, "loss": 1.3051, "step": 90 }, { "grad_norm": 1.4966543912887573, "learning_rate": 1.98e-06, "loss": 1.3201, "step": 100 }, { "grad_norm": 1.9242680072784424, "learning_rate": 2.1800000000000003e-06, "loss": 1.2924, "step": 110 }, { "grad_norm": 2.277714967727661, "learning_rate": 2.38e-06, "loss": 1.2726, "step": 120 }, { "grad_norm": 1.4786337614059448, "learning_rate": 2.5800000000000003e-06, "loss": 1.2504, "step": 130 }, { "grad_norm": 2.0217652320861816, "learning_rate": 2.78e-06, "loss": 1.217, "step": 140 }, { "grad_norm": 1.3985671997070312, "learning_rate": 2.9800000000000003e-06, "loss": 1.1506, "step": 150 }, { "grad_norm": 2.0294013023376465, "learning_rate": 3.1800000000000005e-06, "loss": 1.1792, "step": 160 }, { "grad_norm": 1.9859572649002075, "learning_rate": 3.38e-06, "loss": 1.1204, "step": 170 }, { "grad_norm": 1.7574998140335083, "learning_rate": 3.58e-06, "loss": 1.1042, "step": 180 }, { "grad_norm": 1.7298672199249268, "learning_rate": 3.7800000000000002e-06, "loss": 1.0953, "step": 190 }, { "grad_norm": 2.2437376976013184, "learning_rate": 3.98e-06, "loss": 1.1279, "step": 200 }, { "grad_norm": 1.8126857280731201, "learning_rate": 4.18e-06, "loss": 1.1202, "step": 210 }, { "grad_norm": 2.03757905960083, "learning_rate": 4.38e-06, "loss": 1.1198, "step": 220 }, { "grad_norm": 1.6197428703308105, "learning_rate": 4.58e-06, "loss": 1.114, "step": 230 }, { "grad_norm": 1.523368000984192, "learning_rate": 4.780000000000001e-06, "loss": 1.0877, "step": 240 }, { "grad_norm": 1.3007570505142212, "learning_rate": 4.98e-06, "loss": 1.0815, "step": 250 }, { "grad_norm": 1.6922616958618164, "learning_rate": 5.18e-06, "loss": 1.0629, "step": 260 }, { "grad_norm": 1.5264747142791748, "learning_rate": 5.38e-06, "loss": 1.0753, "step": 270 }, { "grad_norm": 1.6138008832931519, "learning_rate": 5.580000000000001e-06, "loss": 1.0705, "step": 280 }, { "grad_norm": 1.4698479175567627, "learning_rate": 5.78e-06, "loss": 1.1044, "step": 290 }, { "grad_norm": 1.5180444717407227, "learning_rate": 5.98e-06, "loss": 1.0708, "step": 300 }, { "grad_norm": 1.6491835117340088, "learning_rate": 6.18e-06, "loss": 1.109, "step": 310 }, { "grad_norm": 1.6250134706497192, "learning_rate": 6.38e-06, "loss": 1.0901, "step": 320 }, { "grad_norm": 1.4406590461730957, "learning_rate": 6.58e-06, "loss": 1.0679, "step": 330 }, { "grad_norm": 1.8877729177474976, "learning_rate": 6.78e-06, "loss": 1.0905, "step": 340 }, { "grad_norm": 1.6018575429916382, "learning_rate": 6.98e-06, "loss": 1.0781, "step": 350 }, { "grad_norm": 1.3599878549575806, "learning_rate": 7.180000000000001e-06, "loss": 1.0815, "step": 360 }, { "grad_norm": 1.7996174097061157, "learning_rate": 7.3800000000000005e-06, "loss": 1.0838, "step": 370 }, { "grad_norm": 1.579229712486267, "learning_rate": 7.580000000000001e-06, "loss": 1.0951, "step": 380 }, { "grad_norm": 1.5960150957107544, "learning_rate": 7.78e-06, "loss": 1.0811, "step": 390 }, { "grad_norm": 1.5551671981811523, "learning_rate": 7.98e-06, "loss": 1.0705, "step": 400 }, { "grad_norm": 1.4811607599258423, "learning_rate": 8.18e-06, "loss": 1.0841, "step": 410 }, { "grad_norm": 1.5013585090637207, "learning_rate": 8.380000000000001e-06, "loss": 1.069, "step": 420 }, { "grad_norm": 1.5804296731948853, "learning_rate": 8.580000000000001e-06, "loss": 1.0496, "step": 430 }, { "grad_norm": 2.3538012504577637, "learning_rate": 8.78e-06, "loss": 1.0419, "step": 440 }, { "grad_norm": 1.5198365449905396, "learning_rate": 8.98e-06, "loss": 1.0108, "step": 450 }, { "grad_norm": 1.5919795036315918, "learning_rate": 9.180000000000002e-06, "loss": 1.0049, "step": 460 }, { "grad_norm": 1.9657613039016724, "learning_rate": 9.38e-06, "loss": 1.0033, "step": 470 }, { "grad_norm": 3.16713285446167, "learning_rate": 9.58e-06, "loss": 0.9805, "step": 480 }, { "grad_norm": 2.532460927963257, "learning_rate": 9.78e-06, "loss": 0.9656, "step": 490 }, { "grad_norm": 2.2535314559936523, "learning_rate": 9.980000000000001e-06, "loss": 0.9233, "step": 500 }, { "grad_norm": 2.711146354675293, "learning_rate": 1.018e-05, "loss": 0.8972, "step": 510 }, { "grad_norm": 3.0120809078216553, "learning_rate": 1.038e-05, "loss": 0.8459, "step": 520 }, { "grad_norm": 2.84019136428833, "learning_rate": 1.058e-05, "loss": 0.838, "step": 530 }, { "grad_norm": 2.4315738677978516, "learning_rate": 1.0780000000000002e-05, "loss": 0.797, "step": 540 }, { "grad_norm": 2.76203989982605, "learning_rate": 1.098e-05, "loss": 0.7967, "step": 550 }, { "grad_norm": 2.6901087760925293, "learning_rate": 1.118e-05, "loss": 0.7772, "step": 560 }, { "grad_norm": 2.5516774654388428, "learning_rate": 1.1380000000000001e-05, "loss": 0.7481, "step": 570 }, { "grad_norm": 2.3370094299316406, "learning_rate": 1.1580000000000001e-05, "loss": 0.7019, "step": 580 }, { "grad_norm": 2.8282368183135986, "learning_rate": 1.178e-05, "loss": 0.6977, "step": 590 }, { "grad_norm": 3.0147855281829834, "learning_rate": 1.198e-05, "loss": 0.6565, "step": 600 }, { "grad_norm": 3.1915533542633057, "learning_rate": 1.2180000000000002e-05, "loss": 0.6563, "step": 610 }, { "grad_norm": 2.3931844234466553, "learning_rate": 1.238e-05, "loss": 0.6191, "step": 620 }, { "grad_norm": 2.611110210418701, "learning_rate": 1.258e-05, "loss": 0.6154, "step": 630 }, { "grad_norm": 2.852053642272949, "learning_rate": 1.278e-05, "loss": 0.5596, "step": 640 }, { "grad_norm": 3.309039831161499, "learning_rate": 1.2980000000000001e-05, "loss": 0.5194, "step": 650 }, { "grad_norm": 3.4503345489501953, "learning_rate": 1.3180000000000001e-05, "loss": 0.5324, "step": 660 }, { "grad_norm": 2.9398770332336426, "learning_rate": 1.338e-05, "loss": 0.4768, "step": 670 }, { "grad_norm": 3.146461248397827, "learning_rate": 1.358e-05, "loss": 0.4653, "step": 680 }, { "grad_norm": 3.0712332725524902, "learning_rate": 1.3780000000000002e-05, "loss": 0.4838, "step": 690 }, { "grad_norm": 3.5572497844696045, "learning_rate": 1.3980000000000002e-05, "loss": 0.4759, "step": 700 }, { "grad_norm": 3.879347324371338, "learning_rate": 1.4180000000000001e-05, "loss": 0.3955, "step": 710 }, { "grad_norm": 3.6180357933044434, "learning_rate": 1.4380000000000001e-05, "loss": 0.4312, "step": 720 }, { "grad_norm": 3.2686767578125, "learning_rate": 1.4580000000000003e-05, "loss": 0.38, "step": 730 }, { "grad_norm": 2.565136432647705, "learning_rate": 1.4779999999999999e-05, "loss": 0.3431, "step": 740 }, { "grad_norm": 3.2176859378814697, "learning_rate": 1.4979999999999999e-05, "loss": 0.3775, "step": 750 }, { "grad_norm": 3.1082537174224854, "learning_rate": 1.518e-05, "loss": 0.3471, "step": 760 }, { "grad_norm": 2.531458854675293, "learning_rate": 1.538e-05, "loss": 0.3523, "step": 770 }, { "grad_norm": 3.067186117172241, "learning_rate": 1.558e-05, "loss": 0.3758, "step": 780 }, { "grad_norm": 2.748722553253174, "learning_rate": 1.578e-05, "loss": 0.2995, "step": 790 }, { "grad_norm": 2.715473175048828, "learning_rate": 1.598e-05, "loss": 0.2724, "step": 800 }, { "grad_norm": 3.0197741985321045, "learning_rate": 1.618e-05, "loss": 0.2714, "step": 810 }, { "grad_norm": 2.977109909057617, "learning_rate": 1.6380000000000002e-05, "loss": 0.2556, "step": 820 }, { "grad_norm": 3.122105836868286, "learning_rate": 1.658e-05, "loss": 0.2792, "step": 830 }, { "grad_norm": 3.4769370555877686, "learning_rate": 1.6780000000000002e-05, "loss": 0.2717, "step": 840 }, { "grad_norm": 2.785459041595459, "learning_rate": 1.698e-05, "loss": 0.2832, "step": 850 }, { "grad_norm": 3.0209527015686035, "learning_rate": 1.718e-05, "loss": 0.241, "step": 860 }, { "grad_norm": 2.4730193614959717, "learning_rate": 1.7380000000000003e-05, "loss": 0.2803, "step": 870 }, { "grad_norm": 4.172440528869629, "learning_rate": 1.758e-05, "loss": 0.2592, "step": 880 }, { "grad_norm": 2.5514471530914307, "learning_rate": 1.7780000000000003e-05, "loss": 0.257, "step": 890 }, { "grad_norm": 3.6078784465789795, "learning_rate": 1.798e-05, "loss": 0.25, "step": 900 }, { "grad_norm": 3.213683843612671, "learning_rate": 1.818e-05, "loss": 0.2213, "step": 910 }, { "grad_norm": 2.3577466011047363, "learning_rate": 1.838e-05, "loss": 0.1991, "step": 920 }, { "grad_norm": 2.766584873199463, "learning_rate": 1.858e-05, "loss": 0.2242, "step": 930 }, { "grad_norm": 4.6130051612854, "learning_rate": 1.878e-05, "loss": 0.2418, "step": 940 }, { "grad_norm": 2.324368715286255, "learning_rate": 1.898e-05, "loss": 0.2059, "step": 950 }, { "grad_norm": 2.9643149375915527, "learning_rate": 1.918e-05, "loss": 0.2359, "step": 960 }, { "grad_norm": 3.999295949935913, "learning_rate": 1.938e-05, "loss": 0.2427, "step": 970 }, { "grad_norm": 3.05621600151062, "learning_rate": 1.9580000000000002e-05, "loss": 0.2179, "step": 980 }, { "grad_norm": 3.1065828800201416, "learning_rate": 1.978e-05, "loss": 0.1892, "step": 990 }, { "grad_norm": 3.2218329906463623, "learning_rate": 1.9980000000000002e-05, "loss": 0.2019, "step": 1000 }, { "grad_norm": 2.1365933418273926, "learning_rate": 2.0180000000000003e-05, "loss": 0.2251, "step": 1010 }, { "grad_norm": 2.4452767372131348, "learning_rate": 2.038e-05, "loss": 0.2315, "step": 1020 }, { "grad_norm": 3.948934316635132, "learning_rate": 2.0580000000000003e-05, "loss": 0.1937, "step": 1030 }, { "grad_norm": 2.3805477619171143, "learning_rate": 2.078e-05, "loss": 0.2247, "step": 1040 }, { "grad_norm": 2.7627761363983154, "learning_rate": 2.098e-05, "loss": 0.203, "step": 1050 }, { "grad_norm": 2.405463695526123, "learning_rate": 2.118e-05, "loss": 0.2185, "step": 1060 }, { "grad_norm": 2.0984795093536377, "learning_rate": 2.138e-05, "loss": 0.2154, "step": 1070 }, { "grad_norm": 2.5384204387664795, "learning_rate": 2.158e-05, "loss": 0.2277, "step": 1080 }, { "grad_norm": 1.6165398359298706, "learning_rate": 2.178e-05, "loss": 0.1908, "step": 1090 }, { "grad_norm": 2.891535758972168, "learning_rate": 2.198e-05, "loss": 0.1563, "step": 1100 }, { "grad_norm": 3.131582260131836, "learning_rate": 2.218e-05, "loss": 0.244, "step": 1110 }, { "grad_norm": 1.789388656616211, "learning_rate": 2.2380000000000003e-05, "loss": 0.1613, "step": 1120 }, { "grad_norm": 2.271132469177246, "learning_rate": 2.258e-05, "loss": 0.2129, "step": 1130 }, { "grad_norm": 2.8645498752593994, "learning_rate": 2.2780000000000002e-05, "loss": 0.2507, "step": 1140 }, { "grad_norm": 2.4185400009155273, "learning_rate": 2.298e-05, "loss": 0.2203, "step": 1150 }, { "grad_norm": 1.621439814567566, "learning_rate": 2.318e-05, "loss": 0.193, "step": 1160 }, { "grad_norm": 2.1171038150787354, "learning_rate": 2.3380000000000003e-05, "loss": 0.1838, "step": 1170 }, { "grad_norm": 2.629486083984375, "learning_rate": 2.358e-05, "loss": 0.1706, "step": 1180 }, { "grad_norm": 2.822793483734131, "learning_rate": 2.3780000000000003e-05, "loss": 0.1945, "step": 1190 }, { "grad_norm": 4.603678226470947, "learning_rate": 2.398e-05, "loss": 0.2002, "step": 1200 }, { "grad_norm": 4.413877964019775, "learning_rate": 2.418e-05, "loss": 0.2108, "step": 1210 }, { "grad_norm": 2.159128189086914, "learning_rate": 2.438e-05, "loss": 0.1725, "step": 1220 }, { "grad_norm": 3.321786403656006, "learning_rate": 2.4580000000000002e-05, "loss": 0.1947, "step": 1230 }, { "grad_norm": 2.582263469696045, "learning_rate": 2.478e-05, "loss": 0.1838, "step": 1240 }, { "grad_norm": 2.466743230819702, "learning_rate": 2.498e-05, "loss": 0.182, "step": 1250 }, { "grad_norm": 3.4856057167053223, "learning_rate": 2.5180000000000003e-05, "loss": 0.1937, "step": 1260 }, { "grad_norm": 2.0983784198760986, "learning_rate": 2.5380000000000004e-05, "loss": 0.1618, "step": 1270 }, { "grad_norm": 1.5973272323608398, "learning_rate": 2.5580000000000002e-05, "loss": 0.1859, "step": 1280 }, { "grad_norm": 1.8653725385665894, "learning_rate": 2.5779999999999997e-05, "loss": 0.1867, "step": 1290 }, { "grad_norm": 1.7918272018432617, "learning_rate": 2.598e-05, "loss": 0.165, "step": 1300 }, { "grad_norm": 1.9755750894546509, "learning_rate": 2.618e-05, "loss": 0.1894, "step": 1310 }, { "grad_norm": 2.7814016342163086, "learning_rate": 2.6379999999999998e-05, "loss": 0.1716, "step": 1320 }, { "grad_norm": 1.8250502347946167, "learning_rate": 2.658e-05, "loss": 0.1475, "step": 1330 }, { "grad_norm": 1.6864312887191772, "learning_rate": 2.678e-05, "loss": 0.1558, "step": 1340 }, { "grad_norm": 1.9889020919799805, "learning_rate": 2.698e-05, "loss": 0.1865, "step": 1350 }, { "grad_norm": 2.8556220531463623, "learning_rate": 2.718e-05, "loss": 0.1701, "step": 1360 }, { "grad_norm": 2.6687557697296143, "learning_rate": 2.738e-05, "loss": 0.16, "step": 1370 }, { "grad_norm": 2.0207419395446777, "learning_rate": 2.758e-05, "loss": 0.1626, "step": 1380 }, { "grad_norm": 2.7340898513793945, "learning_rate": 2.778e-05, "loss": 0.1744, "step": 1390 }, { "grad_norm": 2.182796001434326, "learning_rate": 2.798e-05, "loss": 0.154, "step": 1400 }, { "grad_norm": 1.7736445665359497, "learning_rate": 2.818e-05, "loss": 0.1697, "step": 1410 }, { "grad_norm": 1.706506609916687, "learning_rate": 2.8380000000000003e-05, "loss": 0.1823, "step": 1420 }, { "grad_norm": 1.7401882410049438, "learning_rate": 2.858e-05, "loss": 0.1459, "step": 1430 }, { "grad_norm": 1.5290840864181519, "learning_rate": 2.8780000000000002e-05, "loss": 0.1628, "step": 1440 }, { "grad_norm": 1.8572039604187012, "learning_rate": 2.898e-05, "loss": 0.182, "step": 1450 }, { "grad_norm": 1.9522302150726318, "learning_rate": 2.9180000000000002e-05, "loss": 0.192, "step": 1460 }, { "grad_norm": 2.1385412216186523, "learning_rate": 2.9380000000000003e-05, "loss": 0.169, "step": 1470 }, { "grad_norm": 2.047760486602783, "learning_rate": 2.958e-05, "loss": 0.15, "step": 1480 }, { "grad_norm": 2.0927658081054688, "learning_rate": 2.9780000000000003e-05, "loss": 0.1646, "step": 1490 }, { "grad_norm": 2.162412166595459, "learning_rate": 2.998e-05, "loss": 0.1739, "step": 1500 }, { "grad_norm": 1.798634648323059, "learning_rate": 3.0180000000000002e-05, "loss": 0.1671, "step": 1510 }, { "grad_norm": 2.3737175464630127, "learning_rate": 3.0380000000000004e-05, "loss": 0.1849, "step": 1520 }, { "grad_norm": 1.4577254056930542, "learning_rate": 3.058e-05, "loss": 0.1696, "step": 1530 }, { "grad_norm": 1.7961758375167847, "learning_rate": 3.078e-05, "loss": 0.1598, "step": 1540 }, { "grad_norm": 1.6929172277450562, "learning_rate": 3.0980000000000005e-05, "loss": 0.1368, "step": 1550 }, { "grad_norm": 2.9677109718322754, "learning_rate": 3.118e-05, "loss": 0.1814, "step": 1560 }, { "grad_norm": 2.068610668182373, "learning_rate": 3.138e-05, "loss": 0.157, "step": 1570 }, { "grad_norm": 1.6375513076782227, "learning_rate": 3.1580000000000006e-05, "loss": 0.1512, "step": 1580 }, { "grad_norm": 1.9513837099075317, "learning_rate": 3.1780000000000004e-05, "loss": 0.1859, "step": 1590 }, { "grad_norm": 2.020920515060425, "learning_rate": 3.198e-05, "loss": 0.1609, "step": 1600 }, { "grad_norm": 2.077425956726074, "learning_rate": 3.218e-05, "loss": 0.1526, "step": 1610 }, { "grad_norm": 1.822500228881836, "learning_rate": 3.238e-05, "loss": 0.1603, "step": 1620 }, { "grad_norm": 2.236448049545288, "learning_rate": 3.2579999999999996e-05, "loss": 0.1996, "step": 1630 }, { "grad_norm": 1.5125468969345093, "learning_rate": 3.278e-05, "loss": 0.1649, "step": 1640 }, { "grad_norm": 1.9673361778259277, "learning_rate": 3.298e-05, "loss": 0.1644, "step": 1650 }, { "grad_norm": 1.8034127950668335, "learning_rate": 3.318e-05, "loss": 0.1613, "step": 1660 }, { "grad_norm": 2.0889620780944824, "learning_rate": 3.338e-05, "loss": 0.1354, "step": 1670 }, { "grad_norm": 1.8533321619033813, "learning_rate": 3.358e-05, "loss": 0.1479, "step": 1680 }, { "grad_norm": 1.8870692253112793, "learning_rate": 3.378e-05, "loss": 0.1543, "step": 1690 }, { "grad_norm": 1.6594520807266235, "learning_rate": 3.398e-05, "loss": 0.1657, "step": 1700 }, { "grad_norm": 2.2707393169403076, "learning_rate": 3.418e-05, "loss": 0.183, "step": 1710 }, { "grad_norm": 1.627942681312561, "learning_rate": 3.438e-05, "loss": 0.1714, "step": 1720 }, { "grad_norm": 1.5584015846252441, "learning_rate": 3.4580000000000004e-05, "loss": 0.1577, "step": 1730 }, { "grad_norm": 2.2574708461761475, "learning_rate": 3.478e-05, "loss": 0.1439, "step": 1740 }, { "grad_norm": 1.817442536354065, "learning_rate": 3.498e-05, "loss": 0.1595, "step": 1750 }, { "grad_norm": 2.080709457397461, "learning_rate": 3.518e-05, "loss": 0.1462, "step": 1760 }, { "grad_norm": 1.5563063621520996, "learning_rate": 3.5380000000000003e-05, "loss": 0.148, "step": 1770 }, { "grad_norm": 1.2271009683609009, "learning_rate": 3.558e-05, "loss": 0.1574, "step": 1780 }, { "grad_norm": 1.6118823289871216, "learning_rate": 3.578e-05, "loss": 0.142, "step": 1790 }, { "grad_norm": 1.5805829763412476, "learning_rate": 3.5980000000000004e-05, "loss": 0.1491, "step": 1800 }, { "grad_norm": 1.5975985527038574, "learning_rate": 3.618e-05, "loss": 0.1468, "step": 1810 }, { "grad_norm": 2.0941834449768066, "learning_rate": 3.638e-05, "loss": 0.1532, "step": 1820 }, { "grad_norm": 2.408301591873169, "learning_rate": 3.6580000000000006e-05, "loss": 0.1696, "step": 1830 }, { "grad_norm": 1.7141588926315308, "learning_rate": 3.6780000000000004e-05, "loss": 0.1705, "step": 1840 }, { "grad_norm": 1.285997986793518, "learning_rate": 3.698e-05, "loss": 0.1903, "step": 1850 }, { "grad_norm": 1.589271068572998, "learning_rate": 3.7180000000000007e-05, "loss": 0.1782, "step": 1860 }, { "grad_norm": 1.1627285480499268, "learning_rate": 3.7380000000000005e-05, "loss": 0.1324, "step": 1870 }, { "grad_norm": 1.364999771118164, "learning_rate": 3.758e-05, "loss": 0.1238, "step": 1880 }, { "grad_norm": 1.7939364910125732, "learning_rate": 3.778000000000001e-05, "loss": 0.1577, "step": 1890 }, { "grad_norm": 1.236767292022705, "learning_rate": 3.7980000000000006e-05, "loss": 0.1398, "step": 1900 }, { "grad_norm": 2.488252878189087, "learning_rate": 3.818e-05, "loss": 0.1785, "step": 1910 }, { "grad_norm": 1.2973952293395996, "learning_rate": 3.838e-05, "loss": 0.1428, "step": 1920 }, { "grad_norm": 1.535369634628296, "learning_rate": 3.858e-05, "loss": 0.1554, "step": 1930 }, { "grad_norm": 2.0359420776367188, "learning_rate": 3.878e-05, "loss": 0.155, "step": 1940 }, { "grad_norm": 1.2050608396530151, "learning_rate": 3.898e-05, "loss": 0.1334, "step": 1950 }, { "grad_norm": 1.8280251026153564, "learning_rate": 3.918e-05, "loss": 0.1821, "step": 1960 }, { "grad_norm": 1.402255654335022, "learning_rate": 3.938e-05, "loss": 0.1394, "step": 1970 }, { "grad_norm": 1.546366810798645, "learning_rate": 3.958e-05, "loss": 0.1774, "step": 1980 }, { "grad_norm": 2.004380941390991, "learning_rate": 3.978e-05, "loss": 0.1357, "step": 1990 }, { "grad_norm": 1.256800651550293, "learning_rate": 3.998e-05, "loss": 0.1508, "step": 2000 }, { "grad_norm": 1.4216458797454834, "learning_rate": 4.018e-05, "loss": 0.1829, "step": 2010 }, { "grad_norm": 1.5985032320022583, "learning_rate": 4.038e-05, "loss": 0.1469, "step": 2020 }, { "grad_norm": 1.5800328254699707, "learning_rate": 4.058e-05, "loss": 0.141, "step": 2030 }, { "grad_norm": 1.5726516246795654, "learning_rate": 4.078e-05, "loss": 0.1619, "step": 2040 }, { "grad_norm": 1.5090371370315552, "learning_rate": 4.0980000000000004e-05, "loss": 0.1595, "step": 2050 }, { "grad_norm": 1.9011012315750122, "learning_rate": 4.118e-05, "loss": 0.1431, "step": 2060 }, { "grad_norm": 1.0359629392623901, "learning_rate": 4.138e-05, "loss": 0.1295, "step": 2070 }, { "grad_norm": 1.0391626358032227, "learning_rate": 4.1580000000000005e-05, "loss": 0.1521, "step": 2080 }, { "grad_norm": 1.568859338760376, "learning_rate": 4.178e-05, "loss": 0.1456, "step": 2090 }, { "grad_norm": 1.414298176765442, "learning_rate": 4.198e-05, "loss": 0.1582, "step": 2100 }, { "grad_norm": 1.3606810569763184, "learning_rate": 4.2180000000000006e-05, "loss": 0.1723, "step": 2110 }, { "grad_norm": 1.2081323862075806, "learning_rate": 4.2380000000000004e-05, "loss": 0.137, "step": 2120 }, { "grad_norm": 1.3424915075302124, "learning_rate": 4.258e-05, "loss": 0.1494, "step": 2130 }, { "grad_norm": 1.9334129095077515, "learning_rate": 4.278e-05, "loss": 0.1155, "step": 2140 }, { "grad_norm": 1.7535039186477661, "learning_rate": 4.2980000000000005e-05, "loss": 0.1648, "step": 2150 }, { "grad_norm": 1.0176442861557007, "learning_rate": 4.318e-05, "loss": 0.1412, "step": 2160 }, { "grad_norm": 0.8946642279624939, "learning_rate": 4.338e-05, "loss": 0.1516, "step": 2170 }, { "grad_norm": 1.1042178869247437, "learning_rate": 4.3580000000000006e-05, "loss": 0.1672, "step": 2180 }, { "grad_norm": 1.2726041078567505, "learning_rate": 4.3780000000000004e-05, "loss": 0.1409, "step": 2190 }, { "grad_norm": 1.3028396368026733, "learning_rate": 4.398e-05, "loss": 0.1484, "step": 2200 }, { "grad_norm": 1.410158395767212, "learning_rate": 4.418000000000001e-05, "loss": 0.1458, "step": 2210 }, { "grad_norm": 1.7590463161468506, "learning_rate": 4.438e-05, "loss": 0.1403, "step": 2220 }, { "grad_norm": 1.0751593112945557, "learning_rate": 4.458e-05, "loss": 0.1399, "step": 2230 }, { "grad_norm": 1.107905387878418, "learning_rate": 4.478e-05, "loss": 0.1484, "step": 2240 }, { "grad_norm": 1.0702955722808838, "learning_rate": 4.498e-05, "loss": 0.1587, "step": 2250 }, { "grad_norm": 1.6148322820663452, "learning_rate": 4.518e-05, "loss": 0.1237, "step": 2260 }, { "grad_norm": 1.096171259880066, "learning_rate": 4.538e-05, "loss": 0.1461, "step": 2270 }, { "grad_norm": 1.2530033588409424, "learning_rate": 4.558e-05, "loss": 0.1386, "step": 2280 }, { "grad_norm": 1.558445930480957, "learning_rate": 4.578e-05, "loss": 0.1868, "step": 2290 }, { "grad_norm": 1.2754096984863281, "learning_rate": 4.5980000000000004e-05, "loss": 0.136, "step": 2300 }, { "grad_norm": 1.3250893354415894, "learning_rate": 4.618e-05, "loss": 0.1379, "step": 2310 }, { "grad_norm": 1.0821837186813354, "learning_rate": 4.638e-05, "loss": 0.1293, "step": 2320 }, { "grad_norm": 1.1718374490737915, "learning_rate": 4.6580000000000005e-05, "loss": 0.1562, "step": 2330 }, { "grad_norm": 0.9599266648292542, "learning_rate": 4.678e-05, "loss": 0.1417, "step": 2340 }, { "grad_norm": 0.8694728016853333, "learning_rate": 4.698e-05, "loss": 0.1406, "step": 2350 }, { "grad_norm": 0.9845587611198425, "learning_rate": 4.718e-05, "loss": 0.1475, "step": 2360 }, { "grad_norm": 1.0917835235595703, "learning_rate": 4.7380000000000004e-05, "loss": 0.1439, "step": 2370 }, { "grad_norm": 1.3107490539550781, "learning_rate": 4.758e-05, "loss": 0.1636, "step": 2380 }, { "grad_norm": 1.0397471189498901, "learning_rate": 4.778e-05, "loss": 0.1571, "step": 2390 }, { "grad_norm": 0.9217429757118225, "learning_rate": 4.7980000000000005e-05, "loss": 0.1395, "step": 2400 }, { "grad_norm": 0.9929723739624023, "learning_rate": 4.818e-05, "loss": 0.1503, "step": 2410 }, { "grad_norm": 1.2613180875778198, "learning_rate": 4.838e-05, "loss": 0.1951, "step": 2420 }, { "grad_norm": 1.2167319059371948, "learning_rate": 4.8580000000000006e-05, "loss": 0.1487, "step": 2430 }, { "grad_norm": 1.6300249099731445, "learning_rate": 4.8780000000000004e-05, "loss": 0.1537, "step": 2440 }, { "grad_norm": 1.1018568277359009, "learning_rate": 4.898e-05, "loss": 0.1264, "step": 2450 }, { "grad_norm": 0.9911161661148071, "learning_rate": 4.918000000000001e-05, "loss": 0.1318, "step": 2460 }, { "grad_norm": 2.6974103450775146, "learning_rate": 4.9380000000000005e-05, "loss": 0.1278, "step": 2470 }, { "grad_norm": 1.407912015914917, "learning_rate": 4.958e-05, "loss": 0.1572, "step": 2480 }, { "grad_norm": 1.2365713119506836, "learning_rate": 4.978e-05, "loss": 0.1372, "step": 2490 }, { "grad_norm": 1.1842284202575684, "learning_rate": 4.9980000000000006e-05, "loss": 0.1308, "step": 2500 }, { "grad_norm": 0.8795945644378662, "learning_rate": 5.0180000000000004e-05, "loss": 0.1191, "step": 2510 }, { "grad_norm": 1.3221439123153687, "learning_rate": 5.038e-05, "loss": 0.1413, "step": 2520 }, { "grad_norm": 1.0067603588104248, "learning_rate": 5.058000000000001e-05, "loss": 0.1221, "step": 2530 }, { "grad_norm": 1.088184118270874, "learning_rate": 5.0780000000000005e-05, "loss": 0.1472, "step": 2540 }, { "grad_norm": 1.5008631944656372, "learning_rate": 5.098e-05, "loss": 0.1582, "step": 2550 }, { "grad_norm": 0.9963534474372864, "learning_rate": 5.118000000000001e-05, "loss": 0.1473, "step": 2560 }, { "grad_norm": 1.1055828332901, "learning_rate": 5.1380000000000006e-05, "loss": 0.146, "step": 2570 }, { "grad_norm": 1.0491725206375122, "learning_rate": 5.1580000000000004e-05, "loss": 0.1525, "step": 2580 }, { "grad_norm": 0.8924846649169922, "learning_rate": 5.178000000000001e-05, "loss": 0.1308, "step": 2590 }, { "grad_norm": 1.0470463037490845, "learning_rate": 5.198000000000001e-05, "loss": 0.1359, "step": 2600 }, { "grad_norm": 1.3156784772872925, "learning_rate": 5.2180000000000005e-05, "loss": 0.1464, "step": 2610 }, { "grad_norm": 1.1278554201126099, "learning_rate": 5.238000000000001e-05, "loss": 0.1188, "step": 2620 }, { "grad_norm": 1.3091026544570923, "learning_rate": 5.258000000000001e-05, "loss": 0.132, "step": 2630 }, { "grad_norm": 1.0020482540130615, "learning_rate": 5.2780000000000006e-05, "loss": 0.1382, "step": 2640 }, { "grad_norm": 0.8735929727554321, "learning_rate": 5.2980000000000004e-05, "loss": 0.1356, "step": 2650 }, { "grad_norm": 0.9417508244514465, "learning_rate": 5.318000000000001e-05, "loss": 0.1408, "step": 2660 }, { "grad_norm": 1.0128512382507324, "learning_rate": 5.338000000000001e-05, "loss": 0.1486, "step": 2670 }, { "grad_norm": 1.3693221807479858, "learning_rate": 5.3580000000000005e-05, "loss": 0.1628, "step": 2680 }, { "grad_norm": 0.9086475372314453, "learning_rate": 5.378e-05, "loss": 0.1301, "step": 2690 }, { "grad_norm": 1.0671173334121704, "learning_rate": 5.3979999999999995e-05, "loss": 0.1215, "step": 2700 }, { "grad_norm": 1.0182924270629883, "learning_rate": 5.418e-05, "loss": 0.1382, "step": 2710 }, { "grad_norm": 1.3047832250595093, "learning_rate": 5.438e-05, "loss": 0.157, "step": 2720 }, { "grad_norm": 0.9403024315834045, "learning_rate": 5.4579999999999996e-05, "loss": 0.1188, "step": 2730 }, { "grad_norm": 0.8718661069869995, "learning_rate": 5.478e-05, "loss": 0.12, "step": 2740 }, { "grad_norm": 1.1081429719924927, "learning_rate": 5.498e-05, "loss": 0.1507, "step": 2750 }, { "grad_norm": 1.4521064758300781, "learning_rate": 5.518e-05, "loss": 0.1523, "step": 2760 }, { "grad_norm": 1.5146392583847046, "learning_rate": 5.538e-05, "loss": 0.1504, "step": 2770 }, { "grad_norm": 1.0381076335906982, "learning_rate": 5.558e-05, "loss": 0.1605, "step": 2780 }, { "grad_norm": 0.7862001657485962, "learning_rate": 5.578e-05, "loss": 0.1519, "step": 2790 }, { "grad_norm": 1.001607894897461, "learning_rate": 5.5979999999999996e-05, "loss": 0.1284, "step": 2800 }, { "grad_norm": 0.9604479670524597, "learning_rate": 5.618e-05, "loss": 0.1607, "step": 2810 }, { "grad_norm": 1.3250788450241089, "learning_rate": 5.638e-05, "loss": 0.1342, "step": 2820 }, { "grad_norm": 0.9351118206977844, "learning_rate": 5.658e-05, "loss": 0.1201, "step": 2830 }, { "grad_norm": 1.137865424156189, "learning_rate": 5.678e-05, "loss": 0.1545, "step": 2840 }, { "grad_norm": 0.9193785786628723, "learning_rate": 5.698e-05, "loss": 0.1482, "step": 2850 }, { "grad_norm": 1.0393178462982178, "learning_rate": 5.718e-05, "loss": 0.1347, "step": 2860 }, { "grad_norm": 1.109897494316101, "learning_rate": 5.738e-05, "loss": 0.1382, "step": 2870 }, { "grad_norm": 1.0689271688461304, "learning_rate": 5.758e-05, "loss": 0.1229, "step": 2880 }, { "grad_norm": 0.7826744914054871, "learning_rate": 5.778e-05, "loss": 0.1321, "step": 2890 }, { "grad_norm": 1.0121920108795166, "learning_rate": 5.7980000000000004e-05, "loss": 0.145, "step": 2900 }, { "grad_norm": 1.1530460119247437, "learning_rate": 5.818e-05, "loss": 0.1192, "step": 2910 }, { "grad_norm": 0.836128830909729, "learning_rate": 5.838e-05, "loss": 0.138, "step": 2920 }, { "grad_norm": 1.160669207572937, "learning_rate": 5.858e-05, "loss": 0.1153, "step": 2930 }, { "grad_norm": 1.115050196647644, "learning_rate": 5.878e-05, "loss": 0.1244, "step": 2940 }, { "grad_norm": 1.0579824447631836, "learning_rate": 5.898e-05, "loss": 0.1487, "step": 2950 }, { "grad_norm": 0.8120213150978088, "learning_rate": 5.918e-05, "loss": 0.1509, "step": 2960 }, { "grad_norm": 0.8935179710388184, "learning_rate": 5.9380000000000004e-05, "loss": 0.1117, "step": 2970 }, { "grad_norm": 1.2384790182113647, "learning_rate": 5.958e-05, "loss": 0.1532, "step": 2980 }, { "grad_norm": 1.1415910720825195, "learning_rate": 5.978e-05, "loss": 0.1419, "step": 2990 }, { "grad_norm": 1.0822911262512207, "learning_rate": 5.9980000000000005e-05, "loss": 0.1432, "step": 3000 }, { "grad_norm": 1.0970814228057861, "learning_rate": 6.018e-05, "loss": 0.113, "step": 3010 }, { "grad_norm": 1.3073798418045044, "learning_rate": 6.038e-05, "loss": 0.1277, "step": 3020 }, { "grad_norm": 5.491157054901123, "learning_rate": 6.0580000000000006e-05, "loss": 0.1224, "step": 3030 }, { "grad_norm": 0.7912977933883667, "learning_rate": 6.0780000000000004e-05, "loss": 0.1272, "step": 3040 }, { "grad_norm": 0.9220079779624939, "learning_rate": 6.098e-05, "loss": 0.1461, "step": 3050 }, { "grad_norm": 0.9473837614059448, "learning_rate": 6.118000000000001e-05, "loss": 0.1431, "step": 3060 }, { "grad_norm": 0.919746458530426, "learning_rate": 6.138e-05, "loss": 0.1092, "step": 3070 }, { "grad_norm": 1.1982078552246094, "learning_rate": 6.158e-05, "loss": 0.133, "step": 3080 }, { "grad_norm": 1.079372525215149, "learning_rate": 6.178000000000001e-05, "loss": 0.1449, "step": 3090 }, { "grad_norm": 1.2312853336334229, "learning_rate": 6.198e-05, "loss": 0.1298, "step": 3100 }, { "grad_norm": 1.0408254861831665, "learning_rate": 6.218e-05, "loss": 0.1121, "step": 3110 }, { "grad_norm": 0.9950939416885376, "learning_rate": 6.238000000000001e-05, "loss": 0.1073, "step": 3120 }, { "grad_norm": 1.0491548776626587, "learning_rate": 6.258e-05, "loss": 0.1305, "step": 3130 }, { "grad_norm": 1.2887829542160034, "learning_rate": 6.278e-05, "loss": 0.1271, "step": 3140 }, { "grad_norm": 2.077991008758545, "learning_rate": 6.298000000000001e-05, "loss": 0.1367, "step": 3150 }, { "grad_norm": 0.8200010061264038, "learning_rate": 6.318e-05, "loss": 0.1276, "step": 3160 }, { "grad_norm": 0.8443949222564697, "learning_rate": 6.338e-05, "loss": 0.1154, "step": 3170 }, { "grad_norm": 1.0253164768218994, "learning_rate": 6.358000000000001e-05, "loss": 0.1331, "step": 3180 }, { "grad_norm": 1.013732671737671, "learning_rate": 6.378e-05, "loss": 0.1362, "step": 3190 }, { "grad_norm": 1.0341250896453857, "learning_rate": 6.398000000000001e-05, "loss": 0.1185, "step": 3200 }, { "grad_norm": 0.7432531714439392, "learning_rate": 6.418000000000001e-05, "loss": 0.1126, "step": 3210 }, { "grad_norm": 1.1711965799331665, "learning_rate": 6.438e-05, "loss": 0.1455, "step": 3220 }, { "grad_norm": 1.0196319818496704, "learning_rate": 6.458000000000001e-05, "loss": 0.1402, "step": 3230 }, { "grad_norm": 0.7232682704925537, "learning_rate": 6.478000000000001e-05, "loss": 0.1059, "step": 3240 }, { "grad_norm": 0.8626006245613098, "learning_rate": 6.498e-05, "loss": 0.1415, "step": 3250 }, { "grad_norm": 0.9152103066444397, "learning_rate": 6.518000000000001e-05, "loss": 0.1235, "step": 3260 }, { "grad_norm": 0.9466662406921387, "learning_rate": 6.538000000000001e-05, "loss": 0.1262, "step": 3270 }, { "grad_norm": 0.8707005977630615, "learning_rate": 6.558e-05, "loss": 0.1117, "step": 3280 }, { "grad_norm": 0.8935055732727051, "learning_rate": 6.578000000000001e-05, "loss": 0.1247, "step": 3290 }, { "grad_norm": 0.9482165575027466, "learning_rate": 6.598e-05, "loss": 0.1047, "step": 3300 }, { "grad_norm": 1.049156665802002, "learning_rate": 6.618e-05, "loss": 0.1233, "step": 3310 }, { "grad_norm": 0.9205184578895569, "learning_rate": 6.638e-05, "loss": 0.1314, "step": 3320 }, { "grad_norm": 0.9299852252006531, "learning_rate": 6.658e-05, "loss": 0.1104, "step": 3330 }, { "grad_norm": 0.8577566742897034, "learning_rate": 6.678e-05, "loss": 0.1117, "step": 3340 }, { "grad_norm": 0.7743574380874634, "learning_rate": 6.698e-05, "loss": 0.1212, "step": 3350 }, { "grad_norm": 1.162043571472168, "learning_rate": 6.718e-05, "loss": 0.1243, "step": 3360 }, { "grad_norm": 1.1843407154083252, "learning_rate": 6.738e-05, "loss": 0.1265, "step": 3370 }, { "grad_norm": 0.9706444144248962, "learning_rate": 6.758e-05, "loss": 0.1801, "step": 3380 }, { "grad_norm": 1.3471264839172363, "learning_rate": 6.778e-05, "loss": 0.1293, "step": 3390 }, { "grad_norm": 0.8715254068374634, "learning_rate": 6.798e-05, "loss": 0.1256, "step": 3400 }, { "grad_norm": 0.9000633955001831, "learning_rate": 6.818e-05, "loss": 0.1189, "step": 3410 }, { "grad_norm": 1.330948829650879, "learning_rate": 6.838e-05, "loss": 0.1327, "step": 3420 }, { "grad_norm": 0.8700183629989624, "learning_rate": 6.858e-05, "loss": 0.1177, "step": 3430 }, { "grad_norm": 1.0181759595870972, "learning_rate": 6.878e-05, "loss": 0.1547, "step": 3440 }, { "grad_norm": 0.8934319019317627, "learning_rate": 6.898e-05, "loss": 0.1234, "step": 3450 }, { "grad_norm": 0.626194417476654, "learning_rate": 6.918e-05, "loss": 0.1065, "step": 3460 }, { "grad_norm": 0.8738490343093872, "learning_rate": 6.938e-05, "loss": 0.1449, "step": 3470 }, { "grad_norm": 0.7483816742897034, "learning_rate": 6.958e-05, "loss": 0.1249, "step": 3480 }, { "grad_norm": 0.9700235724449158, "learning_rate": 6.978e-05, "loss": 0.1247, "step": 3490 }, { "grad_norm": 1.0292222499847412, "learning_rate": 6.998e-05, "loss": 0.1245, "step": 3500 }, { "grad_norm": 0.6823210120201111, "learning_rate": 7.018e-05, "loss": 0.1119, "step": 3510 }, { "grad_norm": 0.9872069954872131, "learning_rate": 7.038e-05, "loss": 0.1256, "step": 3520 }, { "grad_norm": 1.5486799478530884, "learning_rate": 7.058e-05, "loss": 0.1294, "step": 3530 }, { "grad_norm": 0.8458973169326782, "learning_rate": 7.078e-05, "loss": 0.1101, "step": 3540 }, { "grad_norm": 0.8220863938331604, "learning_rate": 7.098e-05, "loss": 0.1166, "step": 3550 }, { "grad_norm": 0.8676125407218933, "learning_rate": 7.118e-05, "loss": 0.0982, "step": 3560 }, { "grad_norm": 0.9623662829399109, "learning_rate": 7.138e-05, "loss": 0.1482, "step": 3570 }, { "grad_norm": 0.7676408886909485, "learning_rate": 7.158e-05, "loss": 0.1391, "step": 3580 }, { "grad_norm": 0.9886109232902527, "learning_rate": 7.178000000000001e-05, "loss": 0.13, "step": 3590 }, { "grad_norm": 0.8842405676841736, "learning_rate": 7.198e-05, "loss": 0.122, "step": 3600 }, { "grad_norm": 0.7992447018623352, "learning_rate": 7.218e-05, "loss": 0.1278, "step": 3610 }, { "grad_norm": 0.9829670190811157, "learning_rate": 7.238000000000001e-05, "loss": 0.1181, "step": 3620 }, { "grad_norm": 0.9435042142868042, "learning_rate": 7.258e-05, "loss": 0.1132, "step": 3630 }, { "grad_norm": 0.8246570229530334, "learning_rate": 7.278e-05, "loss": 0.1342, "step": 3640 }, { "grad_norm": 0.982610285282135, "learning_rate": 7.298000000000001e-05, "loss": 0.1506, "step": 3650 }, { "grad_norm": 0.8966707587242126, "learning_rate": 7.318e-05, "loss": 0.1268, "step": 3660 }, { "grad_norm": 0.9189737439155579, "learning_rate": 7.338e-05, "loss": 0.149, "step": 3670 }, { "grad_norm": 0.8977679014205933, "learning_rate": 7.358000000000001e-05, "loss": 0.1328, "step": 3680 }, { "grad_norm": 0.7961472868919373, "learning_rate": 7.378e-05, "loss": 0.122, "step": 3690 }, { "grad_norm": 0.8363505005836487, "learning_rate": 7.398e-05, "loss": 0.1064, "step": 3700 }, { "grad_norm": 0.8843754529953003, "learning_rate": 7.418000000000001e-05, "loss": 0.1338, "step": 3710 }, { "grad_norm": 0.793368399143219, "learning_rate": 7.438e-05, "loss": 0.1185, "step": 3720 }, { "grad_norm": 0.8231315612792969, "learning_rate": 7.458000000000001e-05, "loss": 0.1075, "step": 3730 }, { "grad_norm": 0.9755131602287292, "learning_rate": 7.478e-05, "loss": 0.1163, "step": 3740 }, { "grad_norm": 1.0981413125991821, "learning_rate": 7.498e-05, "loss": 0.1379, "step": 3750 }, { "grad_norm": 1.0049636363983154, "learning_rate": 7.518000000000001e-05, "loss": 0.1081, "step": 3760 }, { "grad_norm": 0.8955798745155334, "learning_rate": 7.538e-05, "loss": 0.128, "step": 3770 }, { "grad_norm": 1.1221083402633667, "learning_rate": 7.558e-05, "loss": 0.1213, "step": 3780 }, { "grad_norm": 1.0752737522125244, "learning_rate": 7.578000000000001e-05, "loss": 0.1264, "step": 3790 }, { "grad_norm": 0.7539939880371094, "learning_rate": 7.598e-05, "loss": 0.1649, "step": 3800 }, { "grad_norm": 1.1379523277282715, "learning_rate": 7.618e-05, "loss": 0.1268, "step": 3810 }, { "grad_norm": 0.8095458149909973, "learning_rate": 7.638000000000001e-05, "loss": 0.1204, "step": 3820 }, { "grad_norm": 0.9824821352958679, "learning_rate": 7.658e-05, "loss": 0.1341, "step": 3830 }, { "grad_norm": 0.961646318435669, "learning_rate": 7.678000000000001e-05, "loss": 0.1178, "step": 3840 }, { "grad_norm": 1.101017951965332, "learning_rate": 7.698000000000001e-05, "loss": 0.147, "step": 3850 }, { "grad_norm": 1.0121127367019653, "learning_rate": 7.718e-05, "loss": 0.1199, "step": 3860 }, { "grad_norm": 0.7488695979118347, "learning_rate": 7.738000000000001e-05, "loss": 0.1215, "step": 3870 }, { "grad_norm": 0.7515673041343689, "learning_rate": 7.758000000000001e-05, "loss": 0.1194, "step": 3880 }, { "grad_norm": 1.11250901222229, "learning_rate": 7.778e-05, "loss": 0.1165, "step": 3890 }, { "grad_norm": 0.976610004901886, "learning_rate": 7.798000000000001e-05, "loss": 0.1158, "step": 3900 }, { "grad_norm": 1.1686677932739258, "learning_rate": 7.818000000000001e-05, "loss": 0.127, "step": 3910 }, { "grad_norm": 1.0152363777160645, "learning_rate": 7.838e-05, "loss": 0.1309, "step": 3920 }, { "grad_norm": 0.7506522536277771, "learning_rate": 7.858000000000001e-05, "loss": 0.1061, "step": 3930 }, { "grad_norm": 0.8613129258155823, "learning_rate": 7.878e-05, "loss": 0.133, "step": 3940 }, { "grad_norm": 1.6940351724624634, "learning_rate": 7.897999999999999e-05, "loss": 0.116, "step": 3950 }, { "grad_norm": 0.8246526718139648, "learning_rate": 7.918e-05, "loss": 0.1383, "step": 3960 }, { "grad_norm": 0.9086644053459167, "learning_rate": 7.938e-05, "loss": 0.1115, "step": 3970 }, { "grad_norm": 0.7878873944282532, "learning_rate": 7.958e-05, "loss": 0.1285, "step": 3980 }, { "grad_norm": 1.0113192796707153, "learning_rate": 7.978e-05, "loss": 0.139, "step": 3990 }, { "grad_norm": 1.1238508224487305, "learning_rate": 7.998e-05, "loss": 0.1141, "step": 4000 }, { "grad_norm": 1.0596835613250732, "learning_rate": 8.018e-05, "loss": 0.1255, "step": 4010 }, { "grad_norm": 0.8605086207389832, "learning_rate": 8.038e-05, "loss": 0.1247, "step": 4020 }, { "grad_norm": 1.318610668182373, "learning_rate": 8.058e-05, "loss": 0.137, "step": 4030 }, { "grad_norm": 0.7198920249938965, "learning_rate": 8.078e-05, "loss": 0.1082, "step": 4040 }, { "grad_norm": 0.6317945718765259, "learning_rate": 8.098e-05, "loss": 0.1217, "step": 4050 }, { "grad_norm": 0.8681421875953674, "learning_rate": 8.118e-05, "loss": 0.1189, "step": 4060 }, { "grad_norm": 1.0694999694824219, "learning_rate": 8.138e-05, "loss": 0.1183, "step": 4070 }, { "grad_norm": 0.6798802614212036, "learning_rate": 8.158e-05, "loss": 0.1146, "step": 4080 }, { "grad_norm": 0.7432214021682739, "learning_rate": 8.178e-05, "loss": 0.1225, "step": 4090 }, { "grad_norm": 0.8001123070716858, "learning_rate": 8.198e-05, "loss": 0.1261, "step": 4100 }, { "grad_norm": 0.835138738155365, "learning_rate": 8.218e-05, "loss": 0.1276, "step": 4110 }, { "grad_norm": 0.8194931149482727, "learning_rate": 8.238000000000001e-05, "loss": 0.1356, "step": 4120 }, { "grad_norm": 0.6871868968009949, "learning_rate": 8.258e-05, "loss": 0.1196, "step": 4130 }, { "grad_norm": 0.8054322004318237, "learning_rate": 8.278e-05, "loss": 0.1236, "step": 4140 }, { "grad_norm": 0.7791943550109863, "learning_rate": 8.298000000000001e-05, "loss": 0.135, "step": 4150 }, { "grad_norm": 0.9443441033363342, "learning_rate": 8.318e-05, "loss": 0.1065, "step": 4160 }, { "grad_norm": 0.9478850960731506, "learning_rate": 8.338e-05, "loss": 0.1193, "step": 4170 }, { "grad_norm": 0.9623135328292847, "learning_rate": 8.358e-05, "loss": 0.1146, "step": 4180 }, { "grad_norm": 1.2297431230545044, "learning_rate": 8.378e-05, "loss": 0.1211, "step": 4190 }, { "grad_norm": 0.6216996312141418, "learning_rate": 8.398e-05, "loss": 0.1164, "step": 4200 }, { "grad_norm": 0.9036362171173096, "learning_rate": 8.418e-05, "loss": 0.1183, "step": 4210 }, { "grad_norm": 1.0739314556121826, "learning_rate": 8.438e-05, "loss": 0.1342, "step": 4220 }, { "grad_norm": 0.8552709221839905, "learning_rate": 8.458e-05, "loss": 0.141, "step": 4230 }, { "grad_norm": 0.6629366874694824, "learning_rate": 8.478e-05, "loss": 0.1325, "step": 4240 }, { "grad_norm": 0.9772053956985474, "learning_rate": 8.498e-05, "loss": 0.131, "step": 4250 }, { "grad_norm": 0.7080503106117249, "learning_rate": 8.518000000000001e-05, "loss": 0.1251, "step": 4260 }, { "grad_norm": 0.6630628108978271, "learning_rate": 8.538e-05, "loss": 0.1251, "step": 4270 }, { "grad_norm": 0.7783237099647522, "learning_rate": 8.558e-05, "loss": 0.1019, "step": 4280 }, { "grad_norm": 0.8315276503562927, "learning_rate": 8.578000000000001e-05, "loss": 0.1162, "step": 4290 }, { "grad_norm": 0.6526811122894287, "learning_rate": 8.598e-05, "loss": 0.1089, "step": 4300 }, { "grad_norm": 0.8474971652030945, "learning_rate": 8.618e-05, "loss": 0.1243, "step": 4310 }, { "grad_norm": 1.5416561365127563, "learning_rate": 8.638000000000001e-05, "loss": 0.128, "step": 4320 }, { "grad_norm": 0.8189437985420227, "learning_rate": 8.658e-05, "loss": 0.12, "step": 4330 }, { "grad_norm": 1.0306397676467896, "learning_rate": 8.678e-05, "loss": 0.0937, "step": 4340 }, { "grad_norm": 0.8891005516052246, "learning_rate": 8.698000000000001e-05, "loss": 0.1066, "step": 4350 }, { "grad_norm": 0.7563629150390625, "learning_rate": 8.718e-05, "loss": 0.1137, "step": 4360 }, { "grad_norm": 0.6978031992912292, "learning_rate": 8.738000000000001e-05, "loss": 0.1482, "step": 4370 }, { "grad_norm": 0.769284725189209, "learning_rate": 8.758000000000001e-05, "loss": 0.101, "step": 4380 }, { "grad_norm": 0.6408653855323792, "learning_rate": 8.778e-05, "loss": 0.0951, "step": 4390 }, { "grad_norm": 0.9697969555854797, "learning_rate": 8.798000000000001e-05, "loss": 0.1353, "step": 4400 }, { "grad_norm": 0.8653203845024109, "learning_rate": 8.818000000000001e-05, "loss": 0.12, "step": 4410 }, { "grad_norm": 0.7591384053230286, "learning_rate": 8.838e-05, "loss": 0.1246, "step": 4420 }, { "grad_norm": 0.9175883531570435, "learning_rate": 8.858000000000001e-05, "loss": 0.1083, "step": 4430 }, { "grad_norm": 0.5308096408843994, "learning_rate": 8.878000000000001e-05, "loss": 0.1142, "step": 4440 }, { "grad_norm": 0.6460272073745728, "learning_rate": 8.898e-05, "loss": 0.1384, "step": 4450 }, { "grad_norm": 0.5898608565330505, "learning_rate": 8.918000000000001e-05, "loss": 0.1189, "step": 4460 }, { "grad_norm": 0.7239465713500977, "learning_rate": 8.938e-05, "loss": 0.1016, "step": 4470 }, { "grad_norm": 0.7201537489891052, "learning_rate": 8.958e-05, "loss": 0.1428, "step": 4480 }, { "grad_norm": 0.6845909953117371, "learning_rate": 8.978000000000001e-05, "loss": 0.1078, "step": 4490 }, { "grad_norm": 0.8923078179359436, "learning_rate": 8.998e-05, "loss": 0.1196, "step": 4500 }, { "grad_norm": 0.7770247459411621, "learning_rate": 9.018000000000001e-05, "loss": 0.1304, "step": 4510 }, { "grad_norm": 0.7313761115074158, "learning_rate": 9.038000000000001e-05, "loss": 0.1186, "step": 4520 }, { "grad_norm": 0.6737341284751892, "learning_rate": 9.058e-05, "loss": 0.104, "step": 4530 }, { "grad_norm": 0.9458636045455933, "learning_rate": 9.078000000000001e-05, "loss": 0.1324, "step": 4540 }, { "grad_norm": 0.6922023892402649, "learning_rate": 9.098000000000001e-05, "loss": 0.1361, "step": 4550 }, { "grad_norm": 0.8497938513755798, "learning_rate": 9.118e-05, "loss": 0.1109, "step": 4560 }, { "grad_norm": 0.8230882287025452, "learning_rate": 9.138e-05, "loss": 0.1277, "step": 4570 }, { "grad_norm": 1.3211842775344849, "learning_rate": 9.158e-05, "loss": 0.1362, "step": 4580 }, { "grad_norm": 1.0629652738571167, "learning_rate": 9.178e-05, "loss": 0.1217, "step": 4590 }, { "grad_norm": 0.7923469543457031, "learning_rate": 9.198e-05, "loss": 0.125, "step": 4600 }, { "grad_norm": 0.7292318940162659, "learning_rate": 9.218e-05, "loss": 0.1267, "step": 4610 }, { "grad_norm": 0.9909964203834534, "learning_rate": 9.238e-05, "loss": 0.1244, "step": 4620 }, { "grad_norm": 0.9953761100769043, "learning_rate": 9.258e-05, "loss": 0.1287, "step": 4630 }, { "grad_norm": 0.7584828734397888, "learning_rate": 9.278e-05, "loss": 0.1262, "step": 4640 }, { "grad_norm": 0.8964294195175171, "learning_rate": 9.298e-05, "loss": 0.11, "step": 4650 }, { "grad_norm": 0.7902924418449402, "learning_rate": 9.318e-05, "loss": 0.1114, "step": 4660 }, { "grad_norm": 0.9389833211898804, "learning_rate": 9.338e-05, "loss": 0.1291, "step": 4670 }, { "grad_norm": 0.9911646246910095, "learning_rate": 9.358e-05, "loss": 0.1227, "step": 4680 }, { "grad_norm": 0.7277824282646179, "learning_rate": 9.378e-05, "loss": 0.132, "step": 4690 }, { "grad_norm": 0.6697057485580444, "learning_rate": 9.398e-05, "loss": 0.117, "step": 4700 }, { "grad_norm": 0.7450379133224487, "learning_rate": 9.418e-05, "loss": 0.1274, "step": 4710 }, { "grad_norm": 0.5721610188484192, "learning_rate": 9.438e-05, "loss": 0.1066, "step": 4720 }, { "grad_norm": 0.6637871861457825, "learning_rate": 9.458e-05, "loss": 0.1363, "step": 4730 }, { "grad_norm": 1.1047896146774292, "learning_rate": 9.478e-05, "loss": 0.1335, "step": 4740 }, { "grad_norm": 0.6347559690475464, "learning_rate": 9.498e-05, "loss": 0.1045, "step": 4750 }, { "grad_norm": 0.5785782337188721, "learning_rate": 9.518000000000001e-05, "loss": 0.1193, "step": 4760 }, { "grad_norm": 0.7533589005470276, "learning_rate": 9.538e-05, "loss": 0.1157, "step": 4770 }, { "grad_norm": 0.8706145882606506, "learning_rate": 9.558e-05, "loss": 0.1405, "step": 4780 }, { "grad_norm": 0.8841857314109802, "learning_rate": 9.578000000000001e-05, "loss": 0.1123, "step": 4790 }, { "grad_norm": 0.8008820414543152, "learning_rate": 9.598e-05, "loss": 0.1224, "step": 4800 }, { "grad_norm": 0.8914652466773987, "learning_rate": 9.618e-05, "loss": 0.1176, "step": 4810 }, { "grad_norm": 0.7804062366485596, "learning_rate": 9.638000000000001e-05, "loss": 0.1306, "step": 4820 }, { "grad_norm": 0.9108445048332214, "learning_rate": 9.658e-05, "loss": 0.1316, "step": 4830 }, { "grad_norm": 0.8574646711349487, "learning_rate": 9.678e-05, "loss": 0.1234, "step": 4840 }, { "grad_norm": 0.688258945941925, "learning_rate": 9.698000000000001e-05, "loss": 0.1133, "step": 4850 }, { "grad_norm": 0.90996915102005, "learning_rate": 9.718e-05, "loss": 0.1194, "step": 4860 }, { "grad_norm": 1.153320550918579, "learning_rate": 9.738e-05, "loss": 0.1176, "step": 4870 }, { "grad_norm": 0.8261082172393799, "learning_rate": 9.758000000000001e-05, "loss": 0.1204, "step": 4880 }, { "grad_norm": 1.0097440481185913, "learning_rate": 9.778e-05, "loss": 0.1415, "step": 4890 }, { "grad_norm": 1.0156983137130737, "learning_rate": 9.798000000000001e-05, "loss": 0.1311, "step": 4900 }, { "grad_norm": 0.6081116795539856, "learning_rate": 9.818000000000001e-05, "loss": 0.1088, "step": 4910 }, { "grad_norm": 0.9029247164726257, "learning_rate": 9.838e-05, "loss": 0.1224, "step": 4920 }, { "grad_norm": 0.7199517488479614, "learning_rate": 9.858000000000001e-05, "loss": 0.1294, "step": 4930 }, { "grad_norm": 0.7281032204627991, "learning_rate": 9.878e-05, "loss": 0.1203, "step": 4940 }, { "grad_norm": 0.6472499370574951, "learning_rate": 9.898e-05, "loss": 0.1211, "step": 4950 }, { "grad_norm": 0.5994446873664856, "learning_rate": 9.918000000000001e-05, "loss": 0.1217, "step": 4960 }, { "grad_norm": 0.547116219997406, "learning_rate": 9.938e-05, "loss": 0.1318, "step": 4970 }, { "grad_norm": 0.6716684103012085, "learning_rate": 9.958e-05, "loss": 0.1299, "step": 4980 }, { "grad_norm": 0.7187948822975159, "learning_rate": 9.978000000000001e-05, "loss": 0.1143, "step": 4990 }, { "grad_norm": 0.7584694027900696, "learning_rate": 9.998e-05, "loss": 0.1276, "step": 5000 }, { "grad_norm": 0.8097333312034607, "learning_rate": 9.999999778549045e-05, "loss": 0.122, "step": 5010 }, { "grad_norm": 0.7407689094543457, "learning_rate": 9.999999013039593e-05, "loss": 0.1088, "step": 5020 }, { "grad_norm": 0.7704765796661377, "learning_rate": 9.999997700737766e-05, "loss": 0.1183, "step": 5030 }, { "grad_norm": 0.6142639517784119, "learning_rate": 9.999995841643709e-05, "loss": 0.1127, "step": 5040 }, { "grad_norm": 0.7784140706062317, "learning_rate": 9.999993435757623e-05, "loss": 0.1202, "step": 5050 }, { "grad_norm": 0.7912461161613464, "learning_rate": 9.999990483079773e-05, "loss": 0.1017, "step": 5060 }, { "grad_norm": 0.7326473593711853, "learning_rate": 9.999986983610481e-05, "loss": 0.1331, "step": 5070 }, { "grad_norm": 0.9136129021644592, "learning_rate": 9.99998293735013e-05, "loss": 0.131, "step": 5080 }, { "grad_norm": 1.3194608688354492, "learning_rate": 9.999978344299161e-05, "loss": 0.1283, "step": 5090 }, { "grad_norm": 0.8343186378479004, "learning_rate": 9.99997320445808e-05, "loss": 0.1233, "step": 5100 }, { "grad_norm": 0.7859112024307251, "learning_rate": 9.999967517827444e-05, "loss": 0.1226, "step": 5110 }, { "grad_norm": 0.9310383200645447, "learning_rate": 9.999961284407879e-05, "loss": 0.1307, "step": 5120 }, { "grad_norm": 0.606707751750946, "learning_rate": 9.999954504200067e-05, "loss": 0.112, "step": 5130 }, { "grad_norm": 0.7887547016143799, "learning_rate": 9.999947177204744e-05, "loss": 0.1184, "step": 5140 }, { "grad_norm": 0.6138298511505127, "learning_rate": 9.999939303422718e-05, "loss": 0.0948, "step": 5150 }, { "grad_norm": 0.650134265422821, "learning_rate": 9.999930882854847e-05, "loss": 0.1247, "step": 5160 }, { "grad_norm": 0.8729849457740784, "learning_rate": 9.999921915502051e-05, "loss": 0.1149, "step": 5170 }, { "grad_norm": 0.8307914137840271, "learning_rate": 9.99991240136531e-05, "loss": 0.1189, "step": 5180 }, { "grad_norm": 0.5874277353286743, "learning_rate": 9.999902340445668e-05, "loss": 0.1427, "step": 5190 }, { "grad_norm": 0.6438550353050232, "learning_rate": 9.999891732744224e-05, "loss": 0.1112, "step": 5200 }, { "grad_norm": 0.872652530670166, "learning_rate": 9.999880578262135e-05, "loss": 0.1262, "step": 5210 }, { "grad_norm": 0.7612065076828003, "learning_rate": 9.999868877000624e-05, "loss": 0.1102, "step": 5220 }, { "grad_norm": 0.7282289862632751, "learning_rate": 9.99985662896097e-05, "loss": 0.1188, "step": 5230 }, { "grad_norm": 0.6984679102897644, "learning_rate": 9.999843834144513e-05, "loss": 0.1058, "step": 5240 }, { "grad_norm": 0.7927218079566956, "learning_rate": 9.99983049255265e-05, "loss": 0.1311, "step": 5250 }, { "grad_norm": 0.7279465794563293, "learning_rate": 9.999816604186843e-05, "loss": 0.1139, "step": 5260 }, { "grad_norm": 0.7168864011764526, "learning_rate": 9.999802169048609e-05, "loss": 0.1151, "step": 5270 }, { "grad_norm": 0.6782646179199219, "learning_rate": 9.999787187139527e-05, "loss": 0.1222, "step": 5280 }, { "grad_norm": 2.3269991874694824, "learning_rate": 9.999771658461234e-05, "loss": 0.1185, "step": 5290 }, { "grad_norm": 1.2924493551254272, "learning_rate": 9.999755583015431e-05, "loss": 0.1199, "step": 5300 }, { "grad_norm": 1.1270288228988647, "learning_rate": 9.999738960803874e-05, "loss": 0.1122, "step": 5310 }, { "grad_norm": 0.77693772315979, "learning_rate": 9.99972179182838e-05, "loss": 0.1109, "step": 5320 }, { "grad_norm": 0.845192551612854, "learning_rate": 9.99970407609083e-05, "loss": 0.1422, "step": 5330 }, { "grad_norm": 0.786543607711792, "learning_rate": 9.999685813593159e-05, "loss": 0.1129, "step": 5340 }, { "grad_norm": 0.6172980666160583, "learning_rate": 9.999667004337362e-05, "loss": 0.0994, "step": 5350 }, { "grad_norm": 0.6958621144294739, "learning_rate": 9.9996476483255e-05, "loss": 0.1328, "step": 5360 }, { "grad_norm": 0.6689847707748413, "learning_rate": 9.999627745559688e-05, "loss": 0.0986, "step": 5370 }, { "grad_norm": 1.052056074142456, "learning_rate": 9.999607296042101e-05, "loss": 0.0998, "step": 5380 }, { "grad_norm": 0.9679405093193054, "learning_rate": 9.99958629977498e-05, "loss": 0.1163, "step": 5390 }, { "grad_norm": 0.8768700361251831, "learning_rate": 9.999564756760615e-05, "loss": 0.0982, "step": 5400 }, { "grad_norm": 0.5092124342918396, "learning_rate": 9.999542667001366e-05, "loss": 0.0875, "step": 5410 }, { "grad_norm": 0.5524652600288391, "learning_rate": 9.999520030499647e-05, "loss": 0.1091, "step": 5420 }, { "grad_norm": 0.9938234090805054, "learning_rate": 9.999496847257936e-05, "loss": 0.0969, "step": 5430 }, { "grad_norm": 0.6909302473068237, "learning_rate": 9.999473117278764e-05, "loss": 0.1196, "step": 5440 }, { "grad_norm": 0.7078397274017334, "learning_rate": 9.999448840564731e-05, "loss": 0.0947, "step": 5450 }, { "grad_norm": 0.6329946517944336, "learning_rate": 9.999424017118488e-05, "loss": 0.092, "step": 5460 }, { "grad_norm": 0.5741941332817078, "learning_rate": 9.999398646942751e-05, "loss": 0.0891, "step": 5470 }, { "grad_norm": 0.7554945349693298, "learning_rate": 9.999372730040296e-05, "loss": 0.0913, "step": 5480 }, { "grad_norm": 0.6719297766685486, "learning_rate": 9.999346266413953e-05, "loss": 0.1086, "step": 5490 }, { "grad_norm": 0.6465921401977539, "learning_rate": 9.99931925606662e-05, "loss": 0.1185, "step": 5500 }, { "grad_norm": 1.0734106302261353, "learning_rate": 9.99929169900125e-05, "loss": 0.1115, "step": 5510 }, { "grad_norm": 0.9209766387939453, "learning_rate": 9.999263595220855e-05, "loss": 0.1023, "step": 5520 }, { "grad_norm": 0.9032283425331116, "learning_rate": 9.99923494472851e-05, "loss": 0.1033, "step": 5530 }, { "grad_norm": 0.8004978895187378, "learning_rate": 9.999205747527348e-05, "loss": 0.0982, "step": 5540 }, { "grad_norm": 0.7424494624137878, "learning_rate": 9.999176003620561e-05, "loss": 0.1186, "step": 5550 }, { "grad_norm": 1.0033990144729614, "learning_rate": 9.999145713011405e-05, "loss": 0.1191, "step": 5560 }, { "grad_norm": 0.8311755061149597, "learning_rate": 9.999114875703186e-05, "loss": 0.1149, "step": 5570 }, { "grad_norm": 0.8066667318344116, "learning_rate": 9.999083491699281e-05, "loss": 0.1279, "step": 5580 }, { "grad_norm": 0.7865925431251526, "learning_rate": 9.999051561003123e-05, "loss": 0.1374, "step": 5590 }, { "grad_norm": 0.6896798610687256, "learning_rate": 9.999019083618202e-05, "loss": 0.1387, "step": 5600 }, { "grad_norm": 0.6828899383544922, "learning_rate": 9.99898605954807e-05, "loss": 0.1414, "step": 5610 }, { "grad_norm": 1.2459744215011597, "learning_rate": 9.998952488796338e-05, "loss": 0.1225, "step": 5620 }, { "grad_norm": 0.6159306764602661, "learning_rate": 9.998918371366676e-05, "loss": 0.121, "step": 5630 }, { "grad_norm": 0.771047055721283, "learning_rate": 9.99888370726282e-05, "loss": 0.1121, "step": 5640 }, { "grad_norm": 1.2727919816970825, "learning_rate": 9.998848496488556e-05, "loss": 0.1527, "step": 5650 }, { "grad_norm": 0.646817147731781, "learning_rate": 9.998812739047736e-05, "loss": 0.1082, "step": 5660 }, { "grad_norm": 0.7428675293922424, "learning_rate": 9.99877643494427e-05, "loss": 0.1135, "step": 5670 }, { "grad_norm": 1.9454262256622314, "learning_rate": 9.998739584182128e-05, "loss": 0.1269, "step": 5680 }, { "grad_norm": 0.8500282168388367, "learning_rate": 9.998702186765342e-05, "loss": 0.1115, "step": 5690 }, { "grad_norm": 0.7048953175544739, "learning_rate": 9.998664242698e-05, "loss": 0.1113, "step": 5700 }, { "grad_norm": 0.8148061037063599, "learning_rate": 9.998625751984251e-05, "loss": 0.1233, "step": 5710 }, { "grad_norm": 0.6520211100578308, "learning_rate": 9.998586714628307e-05, "loss": 0.0945, "step": 5720 }, { "grad_norm": 0.6272139549255371, "learning_rate": 9.998547130634432e-05, "loss": 0.1222, "step": 5730 }, { "grad_norm": 0.6528464555740356, "learning_rate": 9.99850700000696e-05, "loss": 0.1227, "step": 5740 }, { "grad_norm": 0.9150359034538269, "learning_rate": 9.998466322750278e-05, "loss": 0.1118, "step": 5750 }, { "grad_norm": 0.5753370523452759, "learning_rate": 9.998425098868834e-05, "loss": 0.1104, "step": 5760 }, { "grad_norm": 0.9958245754241943, "learning_rate": 9.998383328367136e-05, "loss": 0.1392, "step": 5770 }, { "grad_norm": 0.7484912872314453, "learning_rate": 9.99834101124975e-05, "loss": 0.1331, "step": 5780 }, { "grad_norm": 0.8178614974021912, "learning_rate": 9.998298147521309e-05, "loss": 0.1073, "step": 5790 }, { "grad_norm": 1.0829527378082275, "learning_rate": 9.998254737186496e-05, "loss": 0.1422, "step": 5800 }, { "grad_norm": 0.6730647087097168, "learning_rate": 9.99821078025006e-05, "loss": 0.1173, "step": 5810 }, { "grad_norm": 0.8277758955955505, "learning_rate": 9.998166276716807e-05, "loss": 0.1111, "step": 5820 }, { "grad_norm": 0.6888526082038879, "learning_rate": 9.998121226591606e-05, "loss": 0.1306, "step": 5830 }, { "grad_norm": 0.6991539597511292, "learning_rate": 9.998075629879382e-05, "loss": 0.118, "step": 5840 }, { "grad_norm": 0.8838464021682739, "learning_rate": 9.99802948658512e-05, "loss": 0.1127, "step": 5850 }, { "grad_norm": 0.7014954090118408, "learning_rate": 9.99798279671387e-05, "loss": 0.1312, "step": 5860 }, { "grad_norm": 0.6007761359214783, "learning_rate": 9.997935560270734e-05, "loss": 0.1162, "step": 5870 }, { "grad_norm": 0.6544389724731445, "learning_rate": 9.997887777260879e-05, "loss": 0.1249, "step": 5880 }, { "grad_norm": 0.6624070405960083, "learning_rate": 9.997839447689532e-05, "loss": 0.1357, "step": 5890 }, { "grad_norm": 0.7427312135696411, "learning_rate": 9.997790571561978e-05, "loss": 0.1304, "step": 5900 }, { "grad_norm": 0.7825894951820374, "learning_rate": 9.99774114888356e-05, "loss": 0.1505, "step": 5910 }, { "grad_norm": 0.7645909190177917, "learning_rate": 9.997691179659684e-05, "loss": 0.1261, "step": 5920 }, { "grad_norm": 0.6106961965560913, "learning_rate": 9.997640663895815e-05, "loss": 0.1034, "step": 5930 }, { "grad_norm": 1.120179295539856, "learning_rate": 9.997589601597477e-05, "loss": 0.1255, "step": 5940 }, { "grad_norm": 0.8279895186424255, "learning_rate": 9.997537992770252e-05, "loss": 0.1106, "step": 5950 }, { "grad_norm": 0.638211727142334, "learning_rate": 9.997485837419788e-05, "loss": 0.1118, "step": 5960 }, { "grad_norm": 0.7381488680839539, "learning_rate": 9.997433135551786e-05, "loss": 0.1262, "step": 5970 }, { "grad_norm": 0.8406988978385925, "learning_rate": 9.997379887172009e-05, "loss": 0.1027, "step": 5980 }, { "grad_norm": 0.7410876750946045, "learning_rate": 9.997326092286281e-05, "loss": 0.1245, "step": 5990 }, { "grad_norm": 0.9072580933570862, "learning_rate": 9.997271750900486e-05, "loss": 0.147, "step": 6000 }, { "grad_norm": 0.7866866588592529, "learning_rate": 9.997216863020565e-05, "loss": 0.1335, "step": 6010 }, { "grad_norm": 0.6848472952842712, "learning_rate": 9.99716142865252e-05, "loss": 0.1295, "step": 6020 }, { "grad_norm": 1.2715457677841187, "learning_rate": 9.997105447802415e-05, "loss": 0.1289, "step": 6030 }, { "grad_norm": 0.9471465945243835, "learning_rate": 9.997048920476373e-05, "loss": 0.1128, "step": 6040 }, { "grad_norm": 0.8487175703048706, "learning_rate": 9.996991846680572e-05, "loss": 0.1271, "step": 6050 }, { "grad_norm": 0.608380138874054, "learning_rate": 9.996934226421257e-05, "loss": 0.1137, "step": 6060 }, { "grad_norm": 0.6203828454017639, "learning_rate": 9.996876059704726e-05, "loss": 0.1158, "step": 6070 }, { "grad_norm": 0.8525974154472351, "learning_rate": 9.996817346537343e-05, "loss": 0.1212, "step": 6080 }, { "grad_norm": 0.708416223526001, "learning_rate": 9.996758086925526e-05, "loss": 0.1181, "step": 6090 }, { "grad_norm": 0.6676358580589294, "learning_rate": 9.996698280875759e-05, "loss": 0.1282, "step": 6100 }, { "grad_norm": 0.6486712694168091, "learning_rate": 9.99663792839458e-05, "loss": 0.108, "step": 6110 }, { "grad_norm": 0.8409522175788879, "learning_rate": 9.99657702948859e-05, "loss": 0.1201, "step": 6120 }, { "grad_norm": 0.6411033272743225, "learning_rate": 9.996515584164448e-05, "loss": 0.0985, "step": 6130 }, { "grad_norm": 0.9455837607383728, "learning_rate": 9.996453592428873e-05, "loss": 0.135, "step": 6140 }, { "grad_norm": 0.5661848783493042, "learning_rate": 9.996391054288646e-05, "loss": 0.1428, "step": 6150 }, { "grad_norm": 0.7145506143569946, "learning_rate": 9.996327969750605e-05, "loss": 0.1331, "step": 6160 }, { "grad_norm": 0.5469978451728821, "learning_rate": 9.996264338821649e-05, "loss": 0.115, "step": 6170 }, { "grad_norm": 0.7116473317146301, "learning_rate": 9.996200161508735e-05, "loss": 0.1034, "step": 6180 }, { "grad_norm": 0.8504451513290405, "learning_rate": 9.996135437818885e-05, "loss": 0.1242, "step": 6190 }, { "grad_norm": 0.6361265778541565, "learning_rate": 9.996070167759175e-05, "loss": 0.1002, "step": 6200 }, { "grad_norm": 0.7034767270088196, "learning_rate": 9.996004351336743e-05, "loss": 0.1159, "step": 6210 }, { "grad_norm": 0.8632635474205017, "learning_rate": 9.995937988558785e-05, "loss": 0.0972, "step": 6220 }, { "grad_norm": 0.7642154693603516, "learning_rate": 9.995871079432561e-05, "loss": 0.1222, "step": 6230 }, { "grad_norm": 0.7689366340637207, "learning_rate": 9.995803623965389e-05, "loss": 0.1207, "step": 6240 }, { "grad_norm": 0.9495342373847961, "learning_rate": 9.995735622164641e-05, "loss": 0.1215, "step": 6250 }, { "grad_norm": 0.7694157958030701, "learning_rate": 9.995667074037758e-05, "loss": 0.0776, "step": 6260 }, { "grad_norm": 1.0087436437606812, "learning_rate": 9.995597979592232e-05, "loss": 0.1097, "step": 6270 }, { "grad_norm": 0.7731265425682068, "learning_rate": 9.995528338835625e-05, "loss": 0.1119, "step": 6280 }, { "grad_norm": 0.6373992562294006, "learning_rate": 9.995458151775547e-05, "loss": 0.0973, "step": 6290 }, { "grad_norm": 0.4782901406288147, "learning_rate": 9.995387418419677e-05, "loss": 0.106, "step": 6300 }, { "grad_norm": 0.6121023297309875, "learning_rate": 9.99531613877575e-05, "loss": 0.0936, "step": 6310 }, { "grad_norm": 0.7521550059318542, "learning_rate": 9.995244312851559e-05, "loss": 0.1049, "step": 6320 }, { "grad_norm": 1.111263394355774, "learning_rate": 9.995171940654961e-05, "loss": 0.1175, "step": 6330 }, { "grad_norm": 0.8199871778488159, "learning_rate": 9.995099022193871e-05, "loss": 0.1169, "step": 6340 }, { "grad_norm": 0.6349664926528931, "learning_rate": 9.995025557476261e-05, "loss": 0.1017, "step": 6350 }, { "grad_norm": 0.792345404624939, "learning_rate": 9.994951546510165e-05, "loss": 0.1032, "step": 6360 }, { "grad_norm": 0.5385950207710266, "learning_rate": 9.994876989303679e-05, "loss": 0.0988, "step": 6370 }, { "grad_norm": 0.7575928568840027, "learning_rate": 9.994801885864955e-05, "loss": 0.1017, "step": 6380 }, { "grad_norm": 1.2382690906524658, "learning_rate": 9.994726236202205e-05, "loss": 0.1182, "step": 6390 }, { "grad_norm": 0.6063915491104126, "learning_rate": 9.994650040323704e-05, "loss": 0.1416, "step": 6400 }, { "grad_norm": 0.6278910040855408, "learning_rate": 9.994573298237784e-05, "loss": 0.0869, "step": 6410 }, { "grad_norm": 0.5774757266044617, "learning_rate": 9.994496009952837e-05, "loss": 0.1045, "step": 6420 }, { "grad_norm": 0.7098646759986877, "learning_rate": 9.994418175477316e-05, "loss": 0.1151, "step": 6430 }, { "grad_norm": 0.79350745677948, "learning_rate": 9.994339794819733e-05, "loss": 0.099, "step": 6440 }, { "grad_norm": 0.6286329627037048, "learning_rate": 9.994260867988658e-05, "loss": 0.1076, "step": 6450 }, { "grad_norm": 0.573400616645813, "learning_rate": 9.994181394992723e-05, "loss": 0.1114, "step": 6460 }, { "grad_norm": 0.6061807870864868, "learning_rate": 9.994101375840618e-05, "loss": 0.1034, "step": 6470 }, { "grad_norm": 0.7640385627746582, "learning_rate": 9.994020810541098e-05, "loss": 0.0883, "step": 6480 }, { "grad_norm": 0.7788577079772949, "learning_rate": 9.99393969910297e-05, "loss": 0.1265, "step": 6490 }, { "grad_norm": 1.3423335552215576, "learning_rate": 9.993858041535104e-05, "loss": 0.0931, "step": 6500 }, { "grad_norm": 0.596864640712738, "learning_rate": 9.99377583784643e-05, "loss": 0.091, "step": 6510 }, { "grad_norm": 0.9552411437034607, "learning_rate": 9.993693088045939e-05, "loss": 0.1063, "step": 6520 }, { "grad_norm": 0.7189499735832214, "learning_rate": 9.99360979214268e-05, "loss": 0.1013, "step": 6530 }, { "grad_norm": 0.4693257808685303, "learning_rate": 9.99352595014576e-05, "loss": 0.1013, "step": 6540 }, { "grad_norm": 1.1687116622924805, "learning_rate": 9.993441562064354e-05, "loss": 0.1062, "step": 6550 }, { "grad_norm": 0.741500198841095, "learning_rate": 9.993356627907685e-05, "loss": 0.0979, "step": 6560 }, { "grad_norm": 0.5457954406738281, "learning_rate": 9.99327114768504e-05, "loss": 0.1025, "step": 6570 }, { "grad_norm": 0.6071977019309998, "learning_rate": 9.99318512140577e-05, "loss": 0.0829, "step": 6580 }, { "grad_norm": 0.7474352121353149, "learning_rate": 9.993098549079284e-05, "loss": 0.1161, "step": 6590 }, { "grad_norm": 0.701296865940094, "learning_rate": 9.993011430715047e-05, "loss": 0.1045, "step": 6600 }, { "grad_norm": 0.8130127787590027, "learning_rate": 9.992923766322586e-05, "loss": 0.0902, "step": 6610 }, { "grad_norm": 0.7053923010826111, "learning_rate": 9.99283555591149e-05, "loss": 0.0867, "step": 6620 }, { "grad_norm": 0.6304624080657959, "learning_rate": 9.992746799491404e-05, "loss": 0.1029, "step": 6630 }, { "grad_norm": 0.6376467347145081, "learning_rate": 9.992657497072033e-05, "loss": 0.1022, "step": 6640 }, { "grad_norm": 1.1694518327713013, "learning_rate": 9.992567648663147e-05, "loss": 0.1343, "step": 6650 }, { "grad_norm": 0.5721333622932434, "learning_rate": 9.992477254274568e-05, "loss": 0.103, "step": 6660 }, { "grad_norm": 0.7451930046081543, "learning_rate": 9.992386313916183e-05, "loss": 0.0971, "step": 6670 }, { "grad_norm": 0.6524651050567627, "learning_rate": 9.992294827597934e-05, "loss": 0.1245, "step": 6680 }, { "grad_norm": 0.6796344518661499, "learning_rate": 9.992202795329831e-05, "loss": 0.1071, "step": 6690 }, { "grad_norm": 0.8958333730697632, "learning_rate": 9.992110217121936e-05, "loss": 0.0933, "step": 6700 }, { "grad_norm": 0.8517448902130127, "learning_rate": 9.992017092984372e-05, "loss": 0.126, "step": 6710 }, { "grad_norm": 0.7035553455352783, "learning_rate": 9.991923422927326e-05, "loss": 0.1163, "step": 6720 }, { "grad_norm": 0.6544493436813354, "learning_rate": 9.991829206961037e-05, "loss": 0.119, "step": 6730 }, { "grad_norm": 0.5995206236839294, "learning_rate": 9.991734445095813e-05, "loss": 0.1381, "step": 6740 }, { "grad_norm": 0.8549278378486633, "learning_rate": 9.991639137342015e-05, "loss": 0.0983, "step": 6750 }, { "grad_norm": 0.8013144731521606, "learning_rate": 9.991543283710064e-05, "loss": 0.0978, "step": 6760 }, { "grad_norm": 0.7477031350135803, "learning_rate": 9.991446884210445e-05, "loss": 0.1093, "step": 6770 }, { "grad_norm": 0.8919029235839844, "learning_rate": 9.9913499388537e-05, "loss": 0.1408, "step": 6780 }, { "grad_norm": 1.1830756664276123, "learning_rate": 9.99125244765043e-05, "loss": 0.094, "step": 6790 }, { "grad_norm": 0.7424704432487488, "learning_rate": 9.991154410611296e-05, "loss": 0.1452, "step": 6800 }, { "grad_norm": 0.6624036431312561, "learning_rate": 9.99105582774702e-05, "loss": 0.1029, "step": 6810 }, { "grad_norm": 0.9582640528678894, "learning_rate": 9.990956699068384e-05, "loss": 0.1318, "step": 6820 }, { "grad_norm": 0.6177154183387756, "learning_rate": 9.990857024586224e-05, "loss": 0.0944, "step": 6830 }, { "grad_norm": 0.7286909222602844, "learning_rate": 9.990756804311446e-05, "loss": 0.1117, "step": 6840 }, { "grad_norm": 0.6533923149108887, "learning_rate": 9.990656038255006e-05, "loss": 0.1085, "step": 6850 }, { "grad_norm": 0.6459093689918518, "learning_rate": 9.990554726427926e-05, "loss": 0.0924, "step": 6860 }, { "grad_norm": 0.7543391585350037, "learning_rate": 9.990452868841284e-05, "loss": 0.1303, "step": 6870 }, { "grad_norm": 0.614220380783081, "learning_rate": 9.99035046550622e-05, "loss": 0.1186, "step": 6880 }, { "grad_norm": 0.49243733286857605, "learning_rate": 9.99024751643393e-05, "loss": 0.1026, "step": 6890 }, { "grad_norm": 0.7387738227844238, "learning_rate": 9.990144021635677e-05, "loss": 0.1096, "step": 6900 }, { "grad_norm": 0.6556506156921387, "learning_rate": 9.990039981122775e-05, "loss": 0.085, "step": 6910 }, { "grad_norm": 1.0792850255966187, "learning_rate": 9.989935394906602e-05, "loss": 0.1068, "step": 6920 }, { "grad_norm": 0.8073018789291382, "learning_rate": 9.989830262998598e-05, "loss": 0.1178, "step": 6930 }, { "grad_norm": 0.7906090617179871, "learning_rate": 9.989724585410259e-05, "loss": 0.1023, "step": 6940 }, { "grad_norm": 1.300902247428894, "learning_rate": 9.989618362153139e-05, "loss": 0.1185, "step": 6950 }, { "grad_norm": 0.5854198932647705, "learning_rate": 9.989511593238859e-05, "loss": 0.111, "step": 6960 }, { "grad_norm": 0.7179349660873413, "learning_rate": 9.98940427867909e-05, "loss": 0.1267, "step": 6970 }, { "grad_norm": 0.5958865880966187, "learning_rate": 9.989296418485573e-05, "loss": 0.11, "step": 6980 }, { "grad_norm": 0.587315559387207, "learning_rate": 9.989188012670101e-05, "loss": 0.0871, "step": 6990 }, { "grad_norm": 0.599157452583313, "learning_rate": 9.989079061244528e-05, "loss": 0.0978, "step": 7000 }, { "grad_norm": 0.7566514611244202, "learning_rate": 9.988969564220769e-05, "loss": 0.1205, "step": 7010 }, { "grad_norm": 0.6836740374565125, "learning_rate": 9.988859521610801e-05, "loss": 0.0957, "step": 7020 }, { "grad_norm": 0.6540045738220215, "learning_rate": 9.988748933426656e-05, "loss": 0.1369, "step": 7030 }, { "grad_norm": 0.6203954815864563, "learning_rate": 9.988637799680428e-05, "loss": 0.1, "step": 7040 }, { "grad_norm": 0.5743858218193054, "learning_rate": 9.98852612038427e-05, "loss": 0.0868, "step": 7050 }, { "grad_norm": 1.2902864217758179, "learning_rate": 9.988413895550397e-05, "loss": 0.0921, "step": 7060 }, { "grad_norm": 0.6918584704399109, "learning_rate": 9.98830112519108e-05, "loss": 0.1122, "step": 7070 }, { "grad_norm": 0.6307616829872131, "learning_rate": 9.98818780931865e-05, "loss": 0.1315, "step": 7080 }, { "grad_norm": 0.7040257453918457, "learning_rate": 9.988073947945502e-05, "loss": 0.0987, "step": 7090 }, { "grad_norm": 0.561911940574646, "learning_rate": 9.987959541084087e-05, "loss": 0.0914, "step": 7100 }, { "grad_norm": 0.6228032112121582, "learning_rate": 9.987844588746915e-05, "loss": 0.11, "step": 7110 }, { "grad_norm": 0.8638731837272644, "learning_rate": 9.987729090946558e-05, "loss": 0.0941, "step": 7120 }, { "grad_norm": 0.5588858723640442, "learning_rate": 9.987613047695647e-05, "loss": 0.0864, "step": 7130 }, { "grad_norm": 0.7254958748817444, "learning_rate": 9.987496459006871e-05, "loss": 0.115, "step": 7140 }, { "grad_norm": 1.2023282051086426, "learning_rate": 9.987379324892982e-05, "loss": 0.1398, "step": 7150 }, { "grad_norm": 1.1282340288162231, "learning_rate": 9.987261645366788e-05, "loss": 0.1378, "step": 7160 }, { "grad_norm": 0.8147374987602234, "learning_rate": 9.987143420441158e-05, "loss": 0.1361, "step": 7170 }, { "grad_norm": 0.8654886484146118, "learning_rate": 9.987024650129022e-05, "loss": 0.1177, "step": 7180 }, { "grad_norm": 0.3837297260761261, "learning_rate": 9.986905334443368e-05, "loss": 0.1081, "step": 7190 }, { "grad_norm": 0.6652655005455017, "learning_rate": 9.986785473397245e-05, "loss": 0.1088, "step": 7200 }, { "grad_norm": 0.7752277255058289, "learning_rate": 9.98666506700376e-05, "loss": 0.1152, "step": 7210 }, { "grad_norm": 0.7567798495292664, "learning_rate": 9.986544115276081e-05, "loss": 0.114, "step": 7220 }, { "grad_norm": 0.7572298049926758, "learning_rate": 9.986422618227433e-05, "loss": 0.1092, "step": 7230 }, { "grad_norm": 1.2933357954025269, "learning_rate": 9.986300575871106e-05, "loss": 0.1263, "step": 7240 }, { "grad_norm": 0.7324004173278809, "learning_rate": 9.986177988220444e-05, "loss": 0.1081, "step": 7250 }, { "grad_norm": 0.8174881935119629, "learning_rate": 9.986054855288856e-05, "loss": 0.099, "step": 7260 }, { "grad_norm": 0.6456823348999023, "learning_rate": 9.985931177089802e-05, "loss": 0.1219, "step": 7270 }, { "grad_norm": 0.7910546064376831, "learning_rate": 9.985806953636814e-05, "loss": 0.1235, "step": 7280 }, { "grad_norm": 0.8470240235328674, "learning_rate": 9.985682184943471e-05, "loss": 0.1126, "step": 7290 }, { "grad_norm": 0.6002684235572815, "learning_rate": 9.98555687102342e-05, "loss": 0.1049, "step": 7300 }, { "grad_norm": 0.6892115473747253, "learning_rate": 9.985431011890367e-05, "loss": 0.1157, "step": 7310 }, { "grad_norm": 0.8660969734191895, "learning_rate": 9.985304607558075e-05, "loss": 0.0992, "step": 7320 }, { "grad_norm": 0.523968517780304, "learning_rate": 9.985177658040364e-05, "loss": 0.1048, "step": 7330 }, { "grad_norm": 0.6408233046531677, "learning_rate": 9.985050163351119e-05, "loss": 0.1084, "step": 7340 }, { "grad_norm": 0.9961560368537903, "learning_rate": 9.984922123504286e-05, "loss": 0.0909, "step": 7350 }, { "grad_norm": 0.9361294507980347, "learning_rate": 9.984793538513862e-05, "loss": 0.101, "step": 7360 }, { "grad_norm": 0.6315744519233704, "learning_rate": 9.984664408393912e-05, "loss": 0.1064, "step": 7370 }, { "grad_norm": 0.5830110311508179, "learning_rate": 9.984534733158556e-05, "loss": 0.1102, "step": 7380 }, { "grad_norm": 0.8164078593254089, "learning_rate": 9.984404512821977e-05, "loss": 0.1146, "step": 7390 }, { "grad_norm": 0.5781572461128235, "learning_rate": 9.984273747398411e-05, "loss": 0.1227, "step": 7400 }, { "grad_norm": 0.7136433720588684, "learning_rate": 9.984142436902165e-05, "loss": 0.1005, "step": 7410 }, { "grad_norm": 0.5706701874732971, "learning_rate": 9.984010581347596e-05, "loss": 0.1048, "step": 7420 }, { "grad_norm": 0.5926202535629272, "learning_rate": 9.983878180749121e-05, "loss": 0.1036, "step": 7430 }, { "grad_norm": 0.5660136938095093, "learning_rate": 9.983745235121222e-05, "loss": 0.0921, "step": 7440 }, { "grad_norm": 0.7398651838302612, "learning_rate": 9.983611744478438e-05, "loss": 0.0998, "step": 7450 }, { "grad_norm": 0.8279762864112854, "learning_rate": 9.983477708835365e-05, "loss": 0.0916, "step": 7460 }, { "grad_norm": 0.6729668974876404, "learning_rate": 9.983343128206664e-05, "loss": 0.1159, "step": 7470 }, { "grad_norm": 0.7657408714294434, "learning_rate": 9.983208002607049e-05, "loss": 0.1123, "step": 7480 }, { "grad_norm": 0.7253252863883972, "learning_rate": 9.9830723320513e-05, "loss": 0.1086, "step": 7490 }, { "grad_norm": 0.8242231011390686, "learning_rate": 9.982936116554254e-05, "loss": 0.1, "step": 7500 }, { "grad_norm": 0.5171955823898315, "learning_rate": 9.982799356130803e-05, "loss": 0.0891, "step": 7510 }, { "grad_norm": 0.6877511143684387, "learning_rate": 9.982662050795908e-05, "loss": 0.0924, "step": 7520 }, { "grad_norm": 0.8286205530166626, "learning_rate": 9.982524200564583e-05, "loss": 0.0958, "step": 7530 }, { "grad_norm": 0.5926327705383301, "learning_rate": 9.982385805451901e-05, "loss": 0.099, "step": 7540 }, { "grad_norm": 0.618775486946106, "learning_rate": 9.982246865472998e-05, "loss": 0.0869, "step": 7550 }, { "grad_norm": 0.5822294354438782, "learning_rate": 9.982107380643069e-05, "loss": 0.0952, "step": 7560 }, { "grad_norm": 0.7124553322792053, "learning_rate": 9.981967350977368e-05, "loss": 0.133, "step": 7570 }, { "grad_norm": 0.5625077486038208, "learning_rate": 9.981826776491208e-05, "loss": 0.0934, "step": 7580 }, { "grad_norm": 0.7454460263252258, "learning_rate": 9.98168565719996e-05, "loss": 0.1245, "step": 7590 }, { "grad_norm": 0.5401722192764282, "learning_rate": 9.98154399311906e-05, "loss": 0.099, "step": 7600 }, { "grad_norm": 0.7878639101982117, "learning_rate": 9.981401784263997e-05, "loss": 0.1092, "step": 7610 }, { "grad_norm": 0.6826326847076416, "learning_rate": 9.981259030650326e-05, "loss": 0.1064, "step": 7620 }, { "grad_norm": 0.5889582633972168, "learning_rate": 9.981115732293655e-05, "loss": 0.097, "step": 7630 }, { "grad_norm": 0.5846942067146301, "learning_rate": 9.980971889209659e-05, "loss": 0.1063, "step": 7640 }, { "grad_norm": 0.5088595151901245, "learning_rate": 9.980827501414064e-05, "loss": 0.0887, "step": 7650 }, { "grad_norm": 0.7028477787971497, "learning_rate": 9.980682568922663e-05, "loss": 0.0935, "step": 7660 }, { "grad_norm": 0.7146445512771606, "learning_rate": 9.980537091751304e-05, "loss": 0.0899, "step": 7670 }, { "grad_norm": 0.6292643547058105, "learning_rate": 9.980391069915897e-05, "loss": 0.0923, "step": 7680 }, { "grad_norm": 0.7098272442817688, "learning_rate": 9.98024450343241e-05, "loss": 0.1018, "step": 7690 }, { "grad_norm": 0.6994541883468628, "learning_rate": 9.980097392316872e-05, "loss": 0.0929, "step": 7700 }, { "grad_norm": 0.6640215516090393, "learning_rate": 9.97994973658537e-05, "loss": 0.1106, "step": 7710 }, { "grad_norm": 0.6034817695617676, "learning_rate": 9.979801536254054e-05, "loss": 0.0969, "step": 7720 }, { "grad_norm": 0.4768582880496979, "learning_rate": 9.979652791339127e-05, "loss": 0.1032, "step": 7730 }, { "grad_norm": 0.6246818900108337, "learning_rate": 9.97950350185686e-05, "loss": 0.0907, "step": 7740 }, { "grad_norm": 0.41733643412590027, "learning_rate": 9.979353667823574e-05, "loss": 0.0998, "step": 7750 }, { "grad_norm": 0.6471352577209473, "learning_rate": 9.979203289255658e-05, "loss": 0.1107, "step": 7760 }, { "grad_norm": 0.6448581218719482, "learning_rate": 9.979052366169557e-05, "loss": 0.1095, "step": 7770 }, { "grad_norm": 0.5251234769821167, "learning_rate": 9.978900898581775e-05, "loss": 0.1036, "step": 7780 }, { "grad_norm": 0.5599926710128784, "learning_rate": 9.978748886508875e-05, "loss": 0.0858, "step": 7790 }, { "grad_norm": 0.7059524059295654, "learning_rate": 9.978596329967484e-05, "loss": 0.1041, "step": 7800 }, { "grad_norm": 0.6871547102928162, "learning_rate": 9.978443228974284e-05, "loss": 0.1212, "step": 7810 }, { "grad_norm": 0.6181379556655884, "learning_rate": 9.978289583546015e-05, "loss": 0.1041, "step": 7820 }, { "grad_norm": 0.7409477829933167, "learning_rate": 9.978135393699484e-05, "loss": 0.0928, "step": 7830 }, { "grad_norm": 0.6198902726173401, "learning_rate": 9.977980659451548e-05, "loss": 0.1065, "step": 7840 }, { "grad_norm": 0.6263271570205688, "learning_rate": 9.977825380819135e-05, "loss": 0.0909, "step": 7850 }, { "grad_norm": 0.5699934363365173, "learning_rate": 9.97766955781922e-05, "loss": 0.1026, "step": 7860 }, { "grad_norm": 0.9844276309013367, "learning_rate": 9.977513190468848e-05, "loss": 0.1051, "step": 7870 }, { "grad_norm": 0.6637774109840393, "learning_rate": 9.977356278785116e-05, "loss": 0.1078, "step": 7880 }, { "grad_norm": 0.5410853028297424, "learning_rate": 9.977198822785184e-05, "loss": 0.0991, "step": 7890 }, { "grad_norm": 0.6363478899002075, "learning_rate": 9.977040822486273e-05, "loss": 0.1053, "step": 7900 }, { "grad_norm": 0.527121901512146, "learning_rate": 9.97688227790566e-05, "loss": 0.1095, "step": 7910 }, { "grad_norm": 0.5377683639526367, "learning_rate": 9.976723189060684e-05, "loss": 0.1174, "step": 7920 }, { "grad_norm": 0.5638037919998169, "learning_rate": 9.976563555968742e-05, "loss": 0.0774, "step": 7930 }, { "grad_norm": 0.7966604232788086, "learning_rate": 9.976403378647292e-05, "loss": 0.1091, "step": 7940 }, { "grad_norm": 0.9077863693237305, "learning_rate": 9.97624265711385e-05, "loss": 0.109, "step": 7950 }, { "grad_norm": 0.967947781085968, "learning_rate": 9.976081391385993e-05, "loss": 0.1171, "step": 7960 }, { "grad_norm": 0.5890535116195679, "learning_rate": 9.975919581481356e-05, "loss": 0.1058, "step": 7970 }, { "grad_norm": 0.9056243896484375, "learning_rate": 9.975757227417634e-05, "loss": 0.0941, "step": 7980 }, { "grad_norm": 0.7753492593765259, "learning_rate": 9.975594329212586e-05, "loss": 0.1064, "step": 7990 }, { "grad_norm": 0.8312535881996155, "learning_rate": 9.97543088688402e-05, "loss": 0.1065, "step": 8000 }, { "grad_norm": 0.9977442026138306, "learning_rate": 9.975266900449814e-05, "loss": 0.0999, "step": 8010 }, { "grad_norm": 0.6400183439254761, "learning_rate": 9.975102369927898e-05, "loss": 0.0997, "step": 8020 }, { "grad_norm": 0.7184542417526245, "learning_rate": 9.974937295336269e-05, "loss": 0.0998, "step": 8030 }, { "grad_norm": 0.8201456069946289, "learning_rate": 9.974771676692975e-05, "loss": 0.1023, "step": 8040 }, { "grad_norm": 0.5776672959327698, "learning_rate": 9.974605514016131e-05, "loss": 0.1053, "step": 8050 }, { "grad_norm": 1.0033966302871704, "learning_rate": 9.974438807323907e-05, "loss": 0.0885, "step": 8060 }, { "grad_norm": 0.6392910480499268, "learning_rate": 9.974271556634535e-05, "loss": 0.0925, "step": 8070 }, { "grad_norm": 1.073622703552246, "learning_rate": 9.974103761966302e-05, "loss": 0.118, "step": 8080 }, { "grad_norm": 0.9556176662445068, "learning_rate": 9.973935423337563e-05, "loss": 0.1139, "step": 8090 }, { "grad_norm": 0.7895596623420715, "learning_rate": 9.973766540766722e-05, "loss": 0.1152, "step": 8100 }, { "grad_norm": 0.6960121393203735, "learning_rate": 9.97359711427225e-05, "loss": 0.112, "step": 8110 }, { "grad_norm": 0.554836094379425, "learning_rate": 9.973427143872677e-05, "loss": 0.12, "step": 8120 }, { "grad_norm": 1.156132698059082, "learning_rate": 9.973256629586589e-05, "loss": 0.1105, "step": 8130 }, { "grad_norm": 0.7066123485565186, "learning_rate": 9.973085571432632e-05, "loss": 0.1307, "step": 8140 }, { "grad_norm": 0.9223375916481018, "learning_rate": 9.972913969429513e-05, "loss": 0.1182, "step": 8150 }, { "grad_norm": 0.7267870903015137, "learning_rate": 9.972741823596e-05, "loss": 0.1085, "step": 8160 }, { "grad_norm": 0.5747233033180237, "learning_rate": 9.972569133950917e-05, "loss": 0.1167, "step": 8170 }, { "grad_norm": 0.8608458638191223, "learning_rate": 9.972395900513151e-05, "loss": 0.0936, "step": 8180 }, { "grad_norm": 0.8685992956161499, "learning_rate": 9.972222123301645e-05, "loss": 0.1016, "step": 8190 }, { "grad_norm": 0.7149431109428406, "learning_rate": 9.972047802335403e-05, "loss": 0.0979, "step": 8200 }, { "grad_norm": 0.5511358380317688, "learning_rate": 9.971872937633488e-05, "loss": 0.0885, "step": 8210 }, { "grad_norm": 0.6487357020378113, "learning_rate": 9.971697529215024e-05, "loss": 0.1021, "step": 8220 }, { "grad_norm": 0.8163998126983643, "learning_rate": 9.971521577099192e-05, "loss": 0.0984, "step": 8230 }, { "grad_norm": 0.6670188307762146, "learning_rate": 9.971345081305236e-05, "loss": 0.0994, "step": 8240 }, { "grad_norm": 0.8917937874794006, "learning_rate": 9.971168041852456e-05, "loss": 0.1033, "step": 8250 }, { "grad_norm": 0.478788286447525, "learning_rate": 9.970990458760215e-05, "loss": 0.0843, "step": 8260 }, { "grad_norm": 0.4705009162425995, "learning_rate": 9.970812332047929e-05, "loss": 0.0961, "step": 8270 }, { "grad_norm": 0.9321383237838745, "learning_rate": 9.97063366173508e-05, "loss": 0.1093, "step": 8280 }, { "grad_norm": 0.5960950255393982, "learning_rate": 9.970454447841207e-05, "loss": 0.0929, "step": 8290 }, { "grad_norm": 0.5823997259140015, "learning_rate": 9.970274690385909e-05, "loss": 0.0992, "step": 8300 }, { "grad_norm": 0.46725746989250183, "learning_rate": 9.970094389388844e-05, "loss": 0.0854, "step": 8310 }, { "grad_norm": 0.5755862593650818, "learning_rate": 9.969913544869728e-05, "loss": 0.084, "step": 8320 }, { "grad_norm": 0.5948111414909363, "learning_rate": 9.96973215684834e-05, "loss": 0.0949, "step": 8330 }, { "grad_norm": 0.5133479237556458, "learning_rate": 9.969550225344513e-05, "loss": 0.0896, "step": 8340 }, { "grad_norm": 0.6101265549659729, "learning_rate": 9.969367750378147e-05, "loss": 0.1068, "step": 8350 }, { "grad_norm": 0.7074673175811768, "learning_rate": 9.969184731969194e-05, "loss": 0.0879, "step": 8360 }, { "grad_norm": 0.6916816234588623, "learning_rate": 9.96900117013767e-05, "loss": 0.1024, "step": 8370 }, { "grad_norm": 0.638656735420227, "learning_rate": 9.96881706490365e-05, "loss": 0.0979, "step": 8380 }, { "grad_norm": 0.551520586013794, "learning_rate": 9.968632416287265e-05, "loss": 0.0822, "step": 8390 }, { "grad_norm": 0.5007894039154053, "learning_rate": 9.96844722430871e-05, "loss": 0.0892, "step": 8400 }, { "grad_norm": 0.6441226005554199, "learning_rate": 9.968261488988235e-05, "loss": 0.1023, "step": 8410 }, { "grad_norm": 0.7050287127494812, "learning_rate": 9.968075210346155e-05, "loss": 0.0913, "step": 8420 }, { "grad_norm": 0.6543415188789368, "learning_rate": 9.967888388402839e-05, "loss": 0.0946, "step": 8430 }, { "grad_norm": 0.5279092788696289, "learning_rate": 9.967701023178717e-05, "loss": 0.0885, "step": 8440 }, { "grad_norm": 0.7630616426467896, "learning_rate": 9.967513114694282e-05, "loss": 0.0919, "step": 8450 }, { "grad_norm": 0.5811266899108887, "learning_rate": 9.967324662970079e-05, "loss": 0.0894, "step": 8460 }, { "grad_norm": 0.8450945615768433, "learning_rate": 9.96713566802672e-05, "loss": 0.1003, "step": 8470 }, { "grad_norm": 0.6716780662536621, "learning_rate": 9.966946129884873e-05, "loss": 0.0921, "step": 8480 }, { "grad_norm": 0.7099688649177551, "learning_rate": 9.966756048565265e-05, "loss": 0.1025, "step": 8490 }, { "grad_norm": 0.744357705116272, "learning_rate": 9.966565424088681e-05, "loss": 0.1042, "step": 8500 }, { "grad_norm": 0.948112964630127, "learning_rate": 9.96637425647597e-05, "loss": 0.0932, "step": 8510 }, { "grad_norm": 0.611077070236206, "learning_rate": 9.966182545748038e-05, "loss": 0.0908, "step": 8520 }, { "grad_norm": 0.8932161927223206, "learning_rate": 9.96599029192585e-05, "loss": 0.1099, "step": 8530 }, { "grad_norm": 0.6509459614753723, "learning_rate": 9.965797495030428e-05, "loss": 0.1099, "step": 8540 }, { "grad_norm": 0.9268403053283691, "learning_rate": 9.96560415508286e-05, "loss": 0.1031, "step": 8550 }, { "grad_norm": 0.6787495017051697, "learning_rate": 9.965410272104286e-05, "loss": 0.0979, "step": 8560 }, { "grad_norm": 0.6672255396842957, "learning_rate": 9.96521584611591e-05, "loss": 0.1309, "step": 8570 }, { "grad_norm": 1.15591561794281, "learning_rate": 9.965020877138994e-05, "loss": 0.1274, "step": 8580 }, { "grad_norm": 0.8600249290466309, "learning_rate": 9.964825365194861e-05, "loss": 0.1035, "step": 8590 }, { "grad_norm": 0.6393437385559082, "learning_rate": 9.96462931030489e-05, "loss": 0.0928, "step": 8600 }, { "grad_norm": 0.8349064588546753, "learning_rate": 9.96443271249052e-05, "loss": 0.0948, "step": 8610 }, { "grad_norm": 0.43431711196899414, "learning_rate": 9.964235571773255e-05, "loss": 0.1208, "step": 8620 }, { "grad_norm": 0.9981090426445007, "learning_rate": 9.96403788817465e-05, "loss": 0.1098, "step": 8630 }, { "grad_norm": 0.6598801016807556, "learning_rate": 9.963839661716325e-05, "loss": 0.1043, "step": 8640 }, { "grad_norm": 0.5233941674232483, "learning_rate": 9.963640892419958e-05, "loss": 0.0994, "step": 8650 }, { "grad_norm": 0.9144399762153625, "learning_rate": 9.963441580307286e-05, "loss": 0.0938, "step": 8660 }, { "grad_norm": 0.5333747863769531, "learning_rate": 9.963241725400104e-05, "loss": 0.0916, "step": 8670 }, { "grad_norm": 0.673471987247467, "learning_rate": 9.963041327720271e-05, "loss": 0.0991, "step": 8680 }, { "grad_norm": 0.629513680934906, "learning_rate": 9.962840387289697e-05, "loss": 0.107, "step": 8690 }, { "grad_norm": 0.9806175231933594, "learning_rate": 9.962638904130363e-05, "loss": 0.1033, "step": 8700 }, { "grad_norm": 0.6740489602088928, "learning_rate": 9.962436878264298e-05, "loss": 0.1121, "step": 8710 }, { "grad_norm": 0.7624882459640503, "learning_rate": 9.962234309713598e-05, "loss": 0.1179, "step": 8720 }, { "grad_norm": 0.6788387298583984, "learning_rate": 9.962031198500414e-05, "loss": 0.1051, "step": 8730 }, { "grad_norm": 0.4840719997882843, "learning_rate": 9.961827544646958e-05, "loss": 0.0866, "step": 8740 }, { "grad_norm": 0.6003747582435608, "learning_rate": 9.961623348175501e-05, "loss": 0.1043, "step": 8750 }, { "grad_norm": 0.5761746168136597, "learning_rate": 9.961418609108377e-05, "loss": 0.0876, "step": 8760 }, { "grad_norm": 0.5460626482963562, "learning_rate": 9.961213327467971e-05, "loss": 0.1032, "step": 8770 }, { "grad_norm": 0.6096686124801636, "learning_rate": 9.961007503276736e-05, "loss": 0.0924, "step": 8780 }, { "grad_norm": 0.6094332933425903, "learning_rate": 9.960801136557179e-05, "loss": 0.0807, "step": 8790 }, { "grad_norm": 0.5727656483650208, "learning_rate": 9.960594227331866e-05, "loss": 0.0811, "step": 8800 }, { "grad_norm": 0.38649576902389526, "learning_rate": 9.960386775623429e-05, "loss": 0.0767, "step": 8810 }, { "grad_norm": 0.5891106128692627, "learning_rate": 9.96017878145455e-05, "loss": 0.0929, "step": 8820 }, { "grad_norm": 0.7247765064239502, "learning_rate": 9.959970244847977e-05, "loss": 0.0964, "step": 8830 }, { "grad_norm": 0.67998206615448, "learning_rate": 9.959761165826518e-05, "loss": 0.099, "step": 8840 }, { "grad_norm": 0.7607550621032715, "learning_rate": 9.959551544413033e-05, "loss": 0.094, "step": 8850 }, { "grad_norm": 0.6950315237045288, "learning_rate": 9.959341380630448e-05, "loss": 0.0917, "step": 8860 }, { "grad_norm": 0.831373929977417, "learning_rate": 9.959130674501746e-05, "loss": 0.1003, "step": 8870 }, { "grad_norm": 0.7025822997093201, "learning_rate": 9.958919426049968e-05, "loss": 0.106, "step": 8880 }, { "grad_norm": 0.6558560729026794, "learning_rate": 9.958707635298219e-05, "loss": 0.0818, "step": 8890 }, { "grad_norm": 0.3935665190219879, "learning_rate": 9.958495302269657e-05, "loss": 0.0923, "step": 8900 }, { "grad_norm": 0.5417048335075378, "learning_rate": 9.958282426987503e-05, "loss": 0.0982, "step": 8910 }, { "grad_norm": 0.7931092977523804, "learning_rate": 9.95806900947504e-05, "loss": 0.1047, "step": 8920 }, { "grad_norm": 0.6426995992660522, "learning_rate": 9.957855049755604e-05, "loss": 0.0873, "step": 8930 }, { "grad_norm": 0.6017234921455383, "learning_rate": 9.957640547852593e-05, "loss": 0.1041, "step": 8940 }, { "grad_norm": 0.5937118530273438, "learning_rate": 9.957425503789466e-05, "loss": 0.0947, "step": 8950 }, { "grad_norm": 0.6697285771369934, "learning_rate": 9.957209917589738e-05, "loss": 0.0897, "step": 8960 }, { "grad_norm": 0.5932798385620117, "learning_rate": 9.956993789276987e-05, "loss": 0.0884, "step": 8970 }, { "grad_norm": 0.6564937829971313, "learning_rate": 9.956777118874847e-05, "loss": 0.0986, "step": 8980 }, { "grad_norm": 0.8389794826507568, "learning_rate": 9.956559906407016e-05, "loss": 0.123, "step": 8990 }, { "grad_norm": 0.6147750616073608, "learning_rate": 9.956342151897245e-05, "loss": 0.1037, "step": 9000 }, { "grad_norm": 0.6153735518455505, "learning_rate": 9.956123855369346e-05, "loss": 0.0885, "step": 9010 }, { "grad_norm": 0.5581565499305725, "learning_rate": 9.955905016847196e-05, "loss": 0.088, "step": 9020 }, { "grad_norm": 0.5609540939331055, "learning_rate": 9.955685636354723e-05, "loss": 0.1149, "step": 9030 }, { "grad_norm": 0.7585353851318359, "learning_rate": 9.95546571391592e-05, "loss": 0.1054, "step": 9040 }, { "grad_norm": 0.7393069863319397, "learning_rate": 9.955245249554837e-05, "loss": 0.0879, "step": 9050 }, { "grad_norm": 0.6135903596878052, "learning_rate": 9.955024243295582e-05, "loss": 0.1234, "step": 9060 }, { "grad_norm": 0.5489068627357483, "learning_rate": 9.954802695162328e-05, "loss": 0.0939, "step": 9070 }, { "grad_norm": 0.628690779209137, "learning_rate": 9.954580605179302e-05, "loss": 0.094, "step": 9080 }, { "grad_norm": 0.8445379137992859, "learning_rate": 9.954357973370788e-05, "loss": 0.1113, "step": 9090 }, { "grad_norm": 0.7510657906532288, "learning_rate": 9.954134799761135e-05, "loss": 0.1316, "step": 9100 }, { "grad_norm": 0.6101550459861755, "learning_rate": 9.953911084374748e-05, "loss": 0.1052, "step": 9110 }, { "grad_norm": 0.8560345768928528, "learning_rate": 9.953686827236093e-05, "loss": 0.115, "step": 9120 }, { "grad_norm": 0.7025312781333923, "learning_rate": 9.953462028369695e-05, "loss": 0.1191, "step": 9130 }, { "grad_norm": 0.5925542116165161, "learning_rate": 9.953236687800136e-05, "loss": 0.1077, "step": 9140 }, { "grad_norm": 0.6827165484428406, "learning_rate": 9.95301080555206e-05, "loss": 0.1082, "step": 9150 }, { "grad_norm": 0.4603462219238281, "learning_rate": 9.952784381650171e-05, "loss": 0.0816, "step": 9160 }, { "grad_norm": 0.6318110227584839, "learning_rate": 9.952557416119226e-05, "loss": 0.1058, "step": 9170 }, { "grad_norm": 0.59056156873703, "learning_rate": 9.95232990898405e-05, "loss": 0.0875, "step": 9180 }, { "grad_norm": 0.5820726752281189, "learning_rate": 9.95210186026952e-05, "loss": 0.0784, "step": 9190 }, { "grad_norm": 0.48817384243011475, "learning_rate": 9.951873270000576e-05, "loss": 0.0926, "step": 9200 }, { "grad_norm": 0.653788149356842, "learning_rate": 9.951644138202216e-05, "loss": 0.1108, "step": 9210 }, { "grad_norm": 0.7093116044998169, "learning_rate": 9.951414464899498e-05, "loss": 0.0888, "step": 9220 }, { "grad_norm": 0.5402491688728333, "learning_rate": 9.951184250117538e-05, "loss": 0.0852, "step": 9230 }, { "grad_norm": 0.6951597929000854, "learning_rate": 9.950953493881513e-05, "loss": 0.0931, "step": 9240 }, { "grad_norm": 0.6522396206855774, "learning_rate": 9.950722196216658e-05, "loss": 0.0809, "step": 9250 }, { "grad_norm": 0.5228630900382996, "learning_rate": 9.950490357148265e-05, "loss": 0.0715, "step": 9260 }, { "grad_norm": 0.518257200717926, "learning_rate": 9.950257976701692e-05, "loss": 0.1054, "step": 9270 }, { "grad_norm": 0.6635202169418335, "learning_rate": 9.950025054902348e-05, "loss": 0.0952, "step": 9280 }, { "grad_norm": 0.8413674831390381, "learning_rate": 9.949791591775706e-05, "loss": 0.0876, "step": 9290 }, { "grad_norm": 0.4485960304737091, "learning_rate": 9.949557587347298e-05, "loss": 0.0969, "step": 9300 }, { "grad_norm": 0.7077145576477051, "learning_rate": 9.949323041642713e-05, "loss": 0.0896, "step": 9310 }, { "grad_norm": 0.7149465680122375, "learning_rate": 9.949087954687602e-05, "loss": 0.112, "step": 9320 }, { "grad_norm": 0.6154718995094299, "learning_rate": 9.948852326507672e-05, "loss": 0.0947, "step": 9330 }, { "grad_norm": 0.5053884983062744, "learning_rate": 9.948616157128694e-05, "loss": 0.0868, "step": 9340 }, { "grad_norm": 0.6794482469558716, "learning_rate": 9.948379446576493e-05, "loss": 0.0906, "step": 9350 }, { "grad_norm": 0.5819694995880127, "learning_rate": 9.948142194876952e-05, "loss": 0.0837, "step": 9360 }, { "grad_norm": 0.4772723913192749, "learning_rate": 9.947904402056024e-05, "loss": 0.0915, "step": 9370 }, { "grad_norm": 0.681365430355072, "learning_rate": 9.947666068139708e-05, "loss": 0.094, "step": 9380 }, { "grad_norm": 0.7606827020645142, "learning_rate": 9.947427193154071e-05, "loss": 0.0911, "step": 9390 }, { "grad_norm": 0.6238269805908203, "learning_rate": 9.947187777125233e-05, "loss": 0.1198, "step": 9400 }, { "grad_norm": 0.5875886678695679, "learning_rate": 9.946947820079377e-05, "loss": 0.1015, "step": 9410 }, { "grad_norm": 1.0176740884780884, "learning_rate": 9.946707322042747e-05, "loss": 0.1014, "step": 9420 }, { "grad_norm": 0.6588602662086487, "learning_rate": 9.94646628304164e-05, "loss": 0.0969, "step": 9430 }, { "grad_norm": 0.6370160579681396, "learning_rate": 9.946224703102418e-05, "loss": 0.0935, "step": 9440 }, { "grad_norm": 0.666295051574707, "learning_rate": 9.945982582251498e-05, "loss": 0.0842, "step": 9450 }, { "grad_norm": 0.6589059233665466, "learning_rate": 9.94573992051536e-05, "loss": 0.0853, "step": 9460 }, { "grad_norm": 0.3947792649269104, "learning_rate": 9.94549671792054e-05, "loss": 0.1045, "step": 9470 }, { "grad_norm": 0.8736461997032166, "learning_rate": 9.945252974493635e-05, "loss": 0.0895, "step": 9480 }, { "grad_norm": 0.43033891916275024, "learning_rate": 9.9450086902613e-05, "loss": 0.0887, "step": 9490 }, { "grad_norm": 0.6110168099403381, "learning_rate": 9.944763865250248e-05, "loss": 0.0833, "step": 9500 }, { "grad_norm": 0.5140467286109924, "learning_rate": 9.944518499487254e-05, "loss": 0.071, "step": 9510 }, { "grad_norm": 0.5241640210151672, "learning_rate": 9.944272592999151e-05, "loss": 0.0914, "step": 9520 }, { "grad_norm": 0.47680336236953735, "learning_rate": 9.94402614581283e-05, "loss": 0.0792, "step": 9530 }, { "grad_norm": 0.7131528854370117, "learning_rate": 9.943779157955244e-05, "loss": 0.0876, "step": 9540 }, { "grad_norm": 0.5098943114280701, "learning_rate": 9.943531629453403e-05, "loss": 0.0799, "step": 9550 }, { "grad_norm": 0.6029288172721863, "learning_rate": 9.943283560334375e-05, "loss": 0.0896, "step": 9560 }, { "grad_norm": 0.5550011396408081, "learning_rate": 9.943034950625288e-05, "loss": 0.0928, "step": 9570 }, { "grad_norm": 0.48939818143844604, "learning_rate": 9.942785800353332e-05, "loss": 0.0802, "step": 9580 }, { "grad_norm": 0.5595859885215759, "learning_rate": 9.942536109545751e-05, "loss": 0.0824, "step": 9590 }, { "grad_norm": 0.8067134618759155, "learning_rate": 9.942285878229853e-05, "loss": 0.0834, "step": 9600 }, { "grad_norm": 0.8584873676300049, "learning_rate": 9.942035106433001e-05, "loss": 0.0899, "step": 9610 }, { "grad_norm": 0.9439913630485535, "learning_rate": 9.94178379418262e-05, "loss": 0.1035, "step": 9620 }, { "grad_norm": 0.9014241695404053, "learning_rate": 9.941531941506194e-05, "loss": 0.0825, "step": 9630 }, { "grad_norm": 0.6445736885070801, "learning_rate": 9.941279548431263e-05, "loss": 0.0886, "step": 9640 }, { "grad_norm": 0.5154881477355957, "learning_rate": 9.941026614985431e-05, "loss": 0.1102, "step": 9650 }, { "grad_norm": 0.6517970561981201, "learning_rate": 9.940773141196357e-05, "loss": 0.0888, "step": 9660 }, { "grad_norm": 0.5800061821937561, "learning_rate": 9.94051912709176e-05, "loss": 0.094, "step": 9670 }, { "grad_norm": 0.572985053062439, "learning_rate": 9.940264572699421e-05, "loss": 0.081, "step": 9680 }, { "grad_norm": 0.6314157247543335, "learning_rate": 9.940009478047174e-05, "loss": 0.0908, "step": 9690 }, { "grad_norm": 0.5311369299888611, "learning_rate": 9.939753843162918e-05, "loss": 0.1036, "step": 9700 }, { "grad_norm": 0.6840944290161133, "learning_rate": 9.939497668074609e-05, "loss": 0.0873, "step": 9710 }, { "grad_norm": 0.6523741483688354, "learning_rate": 9.93924095281026e-05, "loss": 0.0938, "step": 9720 }, { "grad_norm": 0.5941541790962219, "learning_rate": 9.938983697397948e-05, "loss": 0.0956, "step": 9730 }, { "grad_norm": 0.5164909958839417, "learning_rate": 9.938725901865805e-05, "loss": 0.0911, "step": 9740 }, { "grad_norm": 0.5474421977996826, "learning_rate": 9.93846756624202e-05, "loss": 0.0701, "step": 9750 }, { "grad_norm": 0.8233975768089294, "learning_rate": 9.938208690554849e-05, "loss": 0.1053, "step": 9760 }, { "grad_norm": 0.6455035209655762, "learning_rate": 9.9379492748326e-05, "loss": 0.0981, "step": 9770 }, { "grad_norm": 0.8309957981109619, "learning_rate": 9.937689319103641e-05, "loss": 0.1027, "step": 9780 }, { "grad_norm": 0.8173476457595825, "learning_rate": 9.937428823396404e-05, "loss": 0.1004, "step": 9790 }, { "grad_norm": 0.6616349220275879, "learning_rate": 9.937167787739372e-05, "loss": 0.088, "step": 9800 }, { "grad_norm": 0.5801679491996765, "learning_rate": 9.936906212161095e-05, "loss": 0.0891, "step": 9810 }, { "grad_norm": 0.5916625261306763, "learning_rate": 9.936644096690176e-05, "loss": 0.0974, "step": 9820 }, { "grad_norm": 0.6781185865402222, "learning_rate": 9.936381441355282e-05, "loss": 0.0893, "step": 9830 }, { "grad_norm": 0.6079961061477661, "learning_rate": 9.936118246185136e-05, "loss": 0.1102, "step": 9840 }, { "grad_norm": 1.03074312210083, "learning_rate": 9.935854511208518e-05, "loss": 0.1078, "step": 9850 }, { "grad_norm": 0.6191421151161194, "learning_rate": 9.935590236454272e-05, "loss": 0.0924, "step": 9860 }, { "grad_norm": 0.4758634567260742, "learning_rate": 9.935325421951298e-05, "loss": 0.074, "step": 9870 }, { "grad_norm": 0.5899062752723694, "learning_rate": 9.935060067728557e-05, "loss": 0.0857, "step": 9880 }, { "grad_norm": 0.48882678151130676, "learning_rate": 9.934794173815067e-05, "loss": 0.0717, "step": 9890 }, { "grad_norm": 0.7235378623008728, "learning_rate": 9.934527740239906e-05, "loss": 0.0921, "step": 9900 }, { "grad_norm": 0.5958234667778015, "learning_rate": 9.934260767032209e-05, "loss": 0.0938, "step": 9910 }, { "grad_norm": 0.5855137705802917, "learning_rate": 9.933993254221172e-05, "loss": 0.0878, "step": 9920 }, { "grad_norm": 1.0007387399673462, "learning_rate": 9.933725201836053e-05, "loss": 0.0809, "step": 9930 }, { "grad_norm": 1.1400527954101562, "learning_rate": 9.933456609906162e-05, "loss": 0.0904, "step": 9940 }, { "grad_norm": 0.5046648383140564, "learning_rate": 9.933187478460875e-05, "loss": 0.0922, "step": 9950 }, { "grad_norm": 0.5775107145309448, "learning_rate": 9.93291780752962e-05, "loss": 0.098, "step": 9960 }, { "grad_norm": 0.5537809729576111, "learning_rate": 9.932647597141893e-05, "loss": 0.0922, "step": 9970 }, { "grad_norm": 0.6988746523857117, "learning_rate": 9.932376847327239e-05, "loss": 0.0967, "step": 9980 }, { "grad_norm": 0.6465769410133362, "learning_rate": 9.932105558115268e-05, "loss": 0.0966, "step": 9990 }, { "grad_norm": 0.5747760534286499, "learning_rate": 9.931833729535651e-05, "loss": 0.0962, "step": 10000 }, { "grad_norm": 0.7711770534515381, "learning_rate": 9.931561361618111e-05, "loss": 0.1029, "step": 10010 }, { "grad_norm": 0.5785977244377136, "learning_rate": 9.931288454392435e-05, "loss": 0.1006, "step": 10020 }, { "grad_norm": 0.6897149085998535, "learning_rate": 9.931015007888467e-05, "loss": 0.0961, "step": 10030 }, { "grad_norm": 0.7360920906066895, "learning_rate": 9.930741022136112e-05, "loss": 0.0978, "step": 10040 }, { "grad_norm": 0.5918619632720947, "learning_rate": 9.930466497165333e-05, "loss": 0.0902, "step": 10050 }, { "grad_norm": 0.6284366250038147, "learning_rate": 9.93019143300615e-05, "loss": 0.0861, "step": 10060 }, { "grad_norm": 0.5609793663024902, "learning_rate": 9.929915829688644e-05, "loss": 0.0809, "step": 10070 }, { "grad_norm": 0.41863253712654114, "learning_rate": 9.929639687242955e-05, "loss": 0.0734, "step": 10080 }, { "grad_norm": 0.5486583709716797, "learning_rate": 9.929363005699281e-05, "loss": 0.0888, "step": 10090 }, { "grad_norm": 0.5346747040748596, "learning_rate": 9.92908578508788e-05, "loss": 0.0859, "step": 10100 }, { "grad_norm": 0.5968537926673889, "learning_rate": 9.928808025439069e-05, "loss": 0.1041, "step": 10110 }, { "grad_norm": 0.6496557593345642, "learning_rate": 9.928529726783223e-05, "loss": 0.0818, "step": 10120 }, { "grad_norm": 0.7321464419364929, "learning_rate": 9.928250889150774e-05, "loss": 0.0915, "step": 10130 }, { "grad_norm": 0.6541480422019958, "learning_rate": 9.92797151257222e-05, "loss": 0.0888, "step": 10140 }, { "grad_norm": 0.5819696187973022, "learning_rate": 9.927691597078108e-05, "loss": 0.0963, "step": 10150 }, { "grad_norm": 0.6760955452919006, "learning_rate": 9.927411142699053e-05, "loss": 0.0923, "step": 10160 }, { "grad_norm": 0.5408718585968018, "learning_rate": 9.927130149465725e-05, "loss": 0.0912, "step": 10170 }, { "grad_norm": 0.7529031038284302, "learning_rate": 9.92684861740885e-05, "loss": 0.1135, "step": 10180 }, { "grad_norm": 0.6885493397712708, "learning_rate": 9.926566546559217e-05, "loss": 0.1042, "step": 10190 }, { "grad_norm": 0.631066083908081, "learning_rate": 9.926283936947673e-05, "loss": 0.0897, "step": 10200 }, { "grad_norm": 0.5149457454681396, "learning_rate": 9.926000788605126e-05, "loss": 0.0911, "step": 10210 }, { "grad_norm": 0.5825347900390625, "learning_rate": 9.92571710156254e-05, "loss": 0.0814, "step": 10220 }, { "grad_norm": 0.5460373163223267, "learning_rate": 9.925432875850936e-05, "loss": 0.0921, "step": 10230 }, { "grad_norm": 0.8511856198310852, "learning_rate": 9.925148111501396e-05, "loss": 0.0898, "step": 10240 }, { "grad_norm": 0.7740609049797058, "learning_rate": 9.924862808545066e-05, "loss": 0.0883, "step": 10250 }, { "grad_norm": 0.6458532214164734, "learning_rate": 9.924576967013141e-05, "loss": 0.0705, "step": 10260 }, { "grad_norm": 0.6041314601898193, "learning_rate": 9.924290586936887e-05, "loss": 0.0795, "step": 10270 }, { "grad_norm": 0.6276288032531738, "learning_rate": 9.924003668347614e-05, "loss": 0.102, "step": 10280 }, { "grad_norm": 0.6768761277198792, "learning_rate": 9.923716211276704e-05, "loss": 0.1004, "step": 10290 }, { "grad_norm": 0.6299011707305908, "learning_rate": 9.923428215755594e-05, "loss": 0.0791, "step": 10300 }, { "grad_norm": 0.7273299694061279, "learning_rate": 9.923139681815775e-05, "loss": 0.0833, "step": 10310 }, { "grad_norm": 0.8557558059692383, "learning_rate": 9.922850609488801e-05, "loss": 0.0843, "step": 10320 }, { "grad_norm": 0.6510015726089478, "learning_rate": 9.922560998806287e-05, "loss": 0.0841, "step": 10330 }, { "grad_norm": 0.5022733211517334, "learning_rate": 9.922270849799905e-05, "loss": 0.1015, "step": 10340 }, { "grad_norm": 0.6957213878631592, "learning_rate": 9.92198016250138e-05, "loss": 0.0932, "step": 10350 }, { "grad_norm": 0.6874309182167053, "learning_rate": 9.921688936942506e-05, "loss": 0.0909, "step": 10360 }, { "grad_norm": 0.6830790638923645, "learning_rate": 9.921397173155129e-05, "loss": 0.0809, "step": 10370 }, { "grad_norm": 0.7161532044410706, "learning_rate": 9.921104871171157e-05, "loss": 0.0772, "step": 10380 }, { "grad_norm": 0.46115514636039734, "learning_rate": 9.920812031022554e-05, "loss": 0.0857, "step": 10390 }, { "grad_norm": 0.8013975620269775, "learning_rate": 9.920518652741348e-05, "loss": 0.0847, "step": 10400 }, { "grad_norm": 0.6667234897613525, "learning_rate": 9.920224736359618e-05, "loss": 0.1021, "step": 10410 }, { "grad_norm": 0.9894588589668274, "learning_rate": 9.91993028190951e-05, "loss": 0.1061, "step": 10420 }, { "grad_norm": 0.6485292911529541, "learning_rate": 9.919635289423222e-05, "loss": 0.0992, "step": 10430 }, { "grad_norm": 0.5197357535362244, "learning_rate": 9.919339758933015e-05, "loss": 0.0795, "step": 10440 }, { "grad_norm": 0.4720841944217682, "learning_rate": 9.919043690471209e-05, "loss": 0.0958, "step": 10450 }, { "grad_norm": 0.5697386264801025, "learning_rate": 9.91874708407018e-05, "loss": 0.09, "step": 10460 }, { "grad_norm": 0.5151691436767578, "learning_rate": 9.918449939762367e-05, "loss": 0.0718, "step": 10470 }, { "grad_norm": 0.39393502473831177, "learning_rate": 9.91815225758026e-05, "loss": 0.0889, "step": 10480 }, { "grad_norm": 0.48335176706314087, "learning_rate": 9.917854037556419e-05, "loss": 0.0787, "step": 10490 }, { "grad_norm": 0.7056258916854858, "learning_rate": 9.917555279723454e-05, "loss": 0.0827, "step": 10500 }, { "grad_norm": 0.7350947856903076, "learning_rate": 9.917255984114036e-05, "loss": 0.1039, "step": 10510 }, { "grad_norm": 0.53476482629776, "learning_rate": 9.916956150760896e-05, "loss": 0.0754, "step": 10520 }, { "grad_norm": 0.5096080899238586, "learning_rate": 9.916655779696826e-05, "loss": 0.0905, "step": 10530 }, { "grad_norm": 0.5544942617416382, "learning_rate": 9.916354870954671e-05, "loss": 0.0827, "step": 10540 }, { "grad_norm": 0.7090858221054077, "learning_rate": 9.91605342456734e-05, "loss": 0.095, "step": 10550 }, { "grad_norm": 0.6855466961860657, "learning_rate": 9.915751440567795e-05, "loss": 0.0887, "step": 10560 }, { "grad_norm": 0.569249153137207, "learning_rate": 9.915448918989066e-05, "loss": 0.0888, "step": 10570 }, { "grad_norm": 0.8053779006004333, "learning_rate": 9.915145859864232e-05, "loss": 0.0924, "step": 10580 }, { "grad_norm": 0.8127243518829346, "learning_rate": 9.914842263226437e-05, "loss": 0.1047, "step": 10590 }, { "grad_norm": 0.41371217370033264, "learning_rate": 9.914538129108882e-05, "loss": 0.0983, "step": 10600 }, { "grad_norm": 0.6080191135406494, "learning_rate": 9.914233457544825e-05, "loss": 0.1171, "step": 10610 }, { "grad_norm": 0.6057019233703613, "learning_rate": 9.913928248567586e-05, "loss": 0.1196, "step": 10620 }, { "grad_norm": 0.6278406381607056, "learning_rate": 9.913622502210542e-05, "loss": 0.0762, "step": 10630 }, { "grad_norm": 0.6529204249382019, "learning_rate": 9.913316218507128e-05, "loss": 0.1034, "step": 10640 }, { "grad_norm": 0.487213134765625, "learning_rate": 9.91300939749084e-05, "loss": 0.0845, "step": 10650 }, { "grad_norm": 0.5826277732849121, "learning_rate": 9.91270203919523e-05, "loss": 0.1112, "step": 10660 }, { "grad_norm": 1.1855205297470093, "learning_rate": 9.912394143653912e-05, "loss": 0.1, "step": 10670 }, { "grad_norm": 0.812404990196228, "learning_rate": 9.912085710900555e-05, "loss": 0.0914, "step": 10680 }, { "grad_norm": 0.4701129198074341, "learning_rate": 9.911776740968892e-05, "loss": 0.088, "step": 10690 }, { "grad_norm": 0.5675243139266968, "learning_rate": 9.911467233892709e-05, "loss": 0.1137, "step": 10700 }, { "grad_norm": 0.8508665561676025, "learning_rate": 9.911157189705853e-05, "loss": 0.1108, "step": 10710 }, { "grad_norm": 0.5275500416755676, "learning_rate": 9.910846608442229e-05, "loss": 0.0802, "step": 10720 }, { "grad_norm": 0.6483001112937927, "learning_rate": 9.910535490135805e-05, "loss": 0.0791, "step": 10730 }, { "grad_norm": 0.7111926674842834, "learning_rate": 9.910223834820603e-05, "loss": 0.0866, "step": 10740 }, { "grad_norm": 0.6972798705101013, "learning_rate": 9.909911642530703e-05, "loss": 0.1051, "step": 10750 }, { "grad_norm": 0.6809383034706116, "learning_rate": 9.909598913300249e-05, "loss": 0.1103, "step": 10760 }, { "grad_norm": 0.6984972357749939, "learning_rate": 9.909285647163438e-05, "loss": 0.1044, "step": 10770 }, { "grad_norm": 0.7053111791610718, "learning_rate": 9.908971844154531e-05, "loss": 0.0977, "step": 10780 }, { "grad_norm": 0.6673281788825989, "learning_rate": 9.908657504307843e-05, "loss": 0.0886, "step": 10790 }, { "grad_norm": 0.5296409130096436, "learning_rate": 9.908342627657751e-05, "loss": 0.0949, "step": 10800 }, { "grad_norm": 0.5381871461868286, "learning_rate": 9.908027214238689e-05, "loss": 0.0837, "step": 10810 }, { "grad_norm": 0.6000400185585022, "learning_rate": 9.90771126408515e-05, "loss": 0.0934, "step": 10820 }, { "grad_norm": 0.46067312359809875, "learning_rate": 9.907394777231685e-05, "loss": 0.0899, "step": 10830 }, { "grad_norm": 0.5863708257675171, "learning_rate": 9.907077753712905e-05, "loss": 0.1066, "step": 10840 }, { "grad_norm": 0.6757685542106628, "learning_rate": 9.906760193563482e-05, "loss": 0.0931, "step": 10850 }, { "grad_norm": 0.41250208020210266, "learning_rate": 9.906442096818139e-05, "loss": 0.0737, "step": 10860 }, { "grad_norm": 0.7215993404388428, "learning_rate": 9.906123463511665e-05, "loss": 0.1031, "step": 10870 }, { "grad_norm": 0.5049706697463989, "learning_rate": 9.905804293678907e-05, "loss": 0.0854, "step": 10880 }, { "grad_norm": 0.7197475433349609, "learning_rate": 9.905484587354766e-05, "loss": 0.1014, "step": 10890 }, { "grad_norm": 0.6106626987457275, "learning_rate": 9.905164344574205e-05, "loss": 0.0868, "step": 10900 }, { "grad_norm": 0.6415876746177673, "learning_rate": 9.904843565372248e-05, "loss": 0.074, "step": 10910 }, { "grad_norm": 0.5193730592727661, "learning_rate": 9.904522249783972e-05, "loss": 0.0858, "step": 10920 }, { "grad_norm": 0.6268952488899231, "learning_rate": 9.904200397844517e-05, "loss": 0.1057, "step": 10930 }, { "grad_norm": 0.6219902038574219, "learning_rate": 9.903878009589078e-05, "loss": 0.0911, "step": 10940 }, { "grad_norm": 0.7596555352210999, "learning_rate": 9.903555085052915e-05, "loss": 0.0862, "step": 10950 }, { "grad_norm": 0.6565003991127014, "learning_rate": 9.903231624271338e-05, "loss": 0.0858, "step": 10960 }, { "grad_norm": 0.5735461115837097, "learning_rate": 9.902907627279724e-05, "loss": 0.0814, "step": 10970 }, { "grad_norm": 0.5842372179031372, "learning_rate": 9.902583094113504e-05, "loss": 0.0864, "step": 10980 }, { "grad_norm": 0.6899083852767944, "learning_rate": 9.902258024808168e-05, "loss": 0.0912, "step": 10990 }, { "grad_norm": 0.538342297077179, "learning_rate": 9.901932419399264e-05, "loss": 0.084, "step": 11000 }, { "grad_norm": 0.7451401352882385, "learning_rate": 9.9016062779224e-05, "loss": 0.1042, "step": 11010 }, { "grad_norm": 0.48144468665122986, "learning_rate": 9.901279600413242e-05, "loss": 0.0887, "step": 11020 }, { "grad_norm": 0.8057117462158203, "learning_rate": 9.900952386907518e-05, "loss": 0.0941, "step": 11030 }, { "grad_norm": 0.7034852504730225, "learning_rate": 9.90062463744101e-05, "loss": 0.0861, "step": 11040 }, { "grad_norm": 0.7778409719467163, "learning_rate": 9.900296352049558e-05, "loss": 0.0748, "step": 11050 }, { "grad_norm": 0.4221575856208801, "learning_rate": 9.899967530769065e-05, "loss": 0.1031, "step": 11060 }, { "grad_norm": 0.589333713054657, "learning_rate": 9.899638173635489e-05, "loss": 0.0784, "step": 11070 }, { "grad_norm": 0.5208377242088318, "learning_rate": 9.899308280684849e-05, "loss": 0.0875, "step": 11080 }, { "grad_norm": 0.5199598670005798, "learning_rate": 9.898977851953222e-05, "loss": 0.0896, "step": 11090 }, { "grad_norm": 0.5817688703536987, "learning_rate": 9.898646887476741e-05, "loss": 0.0927, "step": 11100 }, { "grad_norm": 0.564644455909729, "learning_rate": 9.898315387291603e-05, "loss": 0.0941, "step": 11110 }, { "grad_norm": 0.47875678539276123, "learning_rate": 9.89798335143406e-05, "loss": 0.0785, "step": 11120 }, { "grad_norm": 0.664882242679596, "learning_rate": 9.897650779940419e-05, "loss": 0.0905, "step": 11130 }, { "grad_norm": 0.7931351065635681, "learning_rate": 9.897317672847054e-05, "loss": 0.077, "step": 11140 }, { "grad_norm": 0.768649697303772, "learning_rate": 9.89698403019039e-05, "loss": 0.0888, "step": 11150 }, { "grad_norm": 0.6234953999519348, "learning_rate": 9.896649852006917e-05, "loss": 0.0818, "step": 11160 }, { "grad_norm": 0.7101189494132996, "learning_rate": 9.896315138333177e-05, "loss": 0.0797, "step": 11170 }, { "grad_norm": 0.5877615213394165, "learning_rate": 9.895979889205774e-05, "loss": 0.0789, "step": 11180 }, { "grad_norm": 0.5037212371826172, "learning_rate": 9.895644104661372e-05, "loss": 0.0783, "step": 11190 }, { "grad_norm": 0.6302365064620972, "learning_rate": 9.895307784736691e-05, "loss": 0.0868, "step": 11200 }, { "grad_norm": 0.6463252902030945, "learning_rate": 9.894970929468512e-05, "loss": 0.0862, "step": 11210 }, { "grad_norm": 0.5823260545730591, "learning_rate": 9.89463353889367e-05, "loss": 0.0887, "step": 11220 }, { "grad_norm": 0.5497224926948547, "learning_rate": 9.894295613049065e-05, "loss": 0.1121, "step": 11230 }, { "grad_norm": 0.5901570320129395, "learning_rate": 9.893957151971649e-05, "loss": 0.0898, "step": 11240 }, { "grad_norm": 0.6721956729888916, "learning_rate": 9.893618155698436e-05, "loss": 0.0814, "step": 11250 }, { "grad_norm": 0.6438552737236023, "learning_rate": 9.8932786242665e-05, "loss": 0.081, "step": 11260 }, { "grad_norm": 0.47347232699394226, "learning_rate": 9.89293855771297e-05, "loss": 0.1087, "step": 11270 }, { "grad_norm": 0.636427104473114, "learning_rate": 9.892597956075036e-05, "loss": 0.0949, "step": 11280 }, { "grad_norm": 0.5267882347106934, "learning_rate": 9.892256819389947e-05, "loss": 0.0854, "step": 11290 }, { "grad_norm": 0.48036742210388184, "learning_rate": 9.891915147695006e-05, "loss": 0.087, "step": 11300 }, { "grad_norm": 0.6850181818008423, "learning_rate": 9.891572941027577e-05, "loss": 0.1048, "step": 11310 }, { "grad_norm": 0.6502511501312256, "learning_rate": 9.89123019942509e-05, "loss": 0.0866, "step": 11320 }, { "grad_norm": 0.585538387298584, "learning_rate": 9.89088692292502e-05, "loss": 0.0806, "step": 11330 }, { "grad_norm": 0.6350550651550293, "learning_rate": 9.89054311156491e-05, "loss": 0.0846, "step": 11340 }, { "grad_norm": 0.7823606729507446, "learning_rate": 9.890198765382357e-05, "loss": 0.1159, "step": 11350 }, { "grad_norm": 0.6971085667610168, "learning_rate": 9.889853884415021e-05, "loss": 0.0834, "step": 11360 }, { "grad_norm": 1.120776891708374, "learning_rate": 9.889508468700614e-05, "loss": 0.0723, "step": 11370 }, { "grad_norm": 0.47671714425086975, "learning_rate": 9.889162518276915e-05, "loss": 0.0783, "step": 11380 }, { "grad_norm": 0.4176432192325592, "learning_rate": 9.888816033181752e-05, "loss": 0.0913, "step": 11390 }, { "grad_norm": 1.0489407777786255, "learning_rate": 9.888469013453018e-05, "loss": 0.1104, "step": 11400 }, { "grad_norm": 1.0785038471221924, "learning_rate": 9.888121459128663e-05, "loss": 0.0953, "step": 11410 }, { "grad_norm": 0.4876071512699127, "learning_rate": 9.887773370246693e-05, "loss": 0.1062, "step": 11420 }, { "grad_norm": 0.6605433225631714, "learning_rate": 9.887424746845177e-05, "loss": 0.0936, "step": 11430 }, { "grad_norm": 1.0163161754608154, "learning_rate": 9.887075588962239e-05, "loss": 0.0928, "step": 11440 }, { "grad_norm": 0.6314312815666199, "learning_rate": 9.88672589663606e-05, "loss": 0.1041, "step": 11450 }, { "grad_norm": 0.49001774191856384, "learning_rate": 9.886375669904886e-05, "loss": 0.1013, "step": 11460 }, { "grad_norm": 0.6546870470046997, "learning_rate": 9.886024908807014e-05, "loss": 0.0864, "step": 11470 }, { "grad_norm": 1.0844155550003052, "learning_rate": 9.885673613380806e-05, "loss": 0.0841, "step": 11480 }, { "grad_norm": 0.570473313331604, "learning_rate": 9.885321783664676e-05, "loss": 0.0798, "step": 11490 }, { "grad_norm": 0.6116921901702881, "learning_rate": 9.884969419697101e-05, "loss": 0.0828, "step": 11500 }, { "grad_norm": 0.6094731092453003, "learning_rate": 9.884616521516614e-05, "loss": 0.0979, "step": 11510 }, { "grad_norm": 0.5682856440544128, "learning_rate": 9.88426308916181e-05, "loss": 0.0905, "step": 11520 }, { "grad_norm": 0.7442985773086548, "learning_rate": 9.883909122671335e-05, "loss": 0.0865, "step": 11530 }, { "grad_norm": 0.575162410736084, "learning_rate": 9.883554622083904e-05, "loss": 0.096, "step": 11540 }, { "grad_norm": 0.9375356435775757, "learning_rate": 9.88319958743828e-05, "loss": 0.1012, "step": 11550 }, { "grad_norm": 0.7483826279640198, "learning_rate": 9.882844018773291e-05, "loss": 0.0846, "step": 11560 }, { "grad_norm": 0.6586488485336304, "learning_rate": 9.882487916127823e-05, "loss": 0.0973, "step": 11570 }, { "grad_norm": 0.5701059699058533, "learning_rate": 9.882131279540815e-05, "loss": 0.1154, "step": 11580 }, { "grad_norm": 0.783669650554657, "learning_rate": 9.881774109051271e-05, "loss": 0.1005, "step": 11590 }, { "grad_norm": 0.6591687202453613, "learning_rate": 9.881416404698252e-05, "loss": 0.0951, "step": 11600 }, { "grad_norm": 0.49394452571868896, "learning_rate": 9.881058166520873e-05, "loss": 0.0927, "step": 11610 }, { "grad_norm": 0.7389425039291382, "learning_rate": 9.880699394558311e-05, "loss": 0.1074, "step": 11620 }, { "grad_norm": 0.5632372498512268, "learning_rate": 9.880340088849801e-05, "loss": 0.1058, "step": 11630 }, { "grad_norm": 0.9832507371902466, "learning_rate": 9.879980249434637e-05, "loss": 0.0858, "step": 11640 }, { "grad_norm": 0.6272990107536316, "learning_rate": 9.879619876352168e-05, "loss": 0.0942, "step": 11650 }, { "grad_norm": 0.5288619995117188, "learning_rate": 9.879258969641809e-05, "loss": 0.0899, "step": 11660 }, { "grad_norm": 0.865157961845398, "learning_rate": 9.878897529343023e-05, "loss": 0.0958, "step": 11670 }, { "grad_norm": 0.7136399745941162, "learning_rate": 9.878535555495338e-05, "loss": 0.0949, "step": 11680 }, { "grad_norm": 0.7557562589645386, "learning_rate": 9.87817304813834e-05, "loss": 0.1066, "step": 11690 }, { "grad_norm": 0.5844379663467407, "learning_rate": 9.877810007311671e-05, "loss": 0.0951, "step": 11700 }, { "grad_norm": 0.5322182178497314, "learning_rate": 9.877446433055035e-05, "loss": 0.0944, "step": 11710 }, { "grad_norm": 0.7415554523468018, "learning_rate": 9.877082325408191e-05, "loss": 0.0931, "step": 11720 }, { "grad_norm": 0.989256739616394, "learning_rate": 9.876717684410954e-05, "loss": 0.0813, "step": 11730 }, { "grad_norm": 1.0029146671295166, "learning_rate": 9.876352510103204e-05, "loss": 0.0821, "step": 11740 }, { "grad_norm": 0.6734974384307861, "learning_rate": 9.875986802524875e-05, "loss": 0.0803, "step": 11750 }, { "grad_norm": 0.5477232933044434, "learning_rate": 9.87562056171596e-05, "loss": 0.0801, "step": 11760 }, { "grad_norm": 0.4523962438106537, "learning_rate": 9.875253787716511e-05, "loss": 0.1036, "step": 11770 }, { "grad_norm": 0.6823940277099609, "learning_rate": 9.874886480566637e-05, "loss": 0.0923, "step": 11780 }, { "grad_norm": 0.5803208351135254, "learning_rate": 9.874518640306507e-05, "loss": 0.0834, "step": 11790 }, { "grad_norm": 0.5781999230384827, "learning_rate": 9.874150266976347e-05, "loss": 0.1015, "step": 11800 }, { "grad_norm": 0.6428901553153992, "learning_rate": 9.873781360616443e-05, "loss": 0.0983, "step": 11810 }, { "grad_norm": 0.7610725164413452, "learning_rate": 9.873411921267137e-05, "loss": 0.0848, "step": 11820 }, { "grad_norm": 0.735373854637146, "learning_rate": 9.873041948968829e-05, "loss": 0.093, "step": 11830 }, { "grad_norm": 0.5662447214126587, "learning_rate": 9.872671443761981e-05, "loss": 0.0877, "step": 11840 }, { "grad_norm": 0.7061771750450134, "learning_rate": 9.872300405687109e-05, "loss": 0.0766, "step": 11850 }, { "grad_norm": 0.5117701888084412, "learning_rate": 9.871928834784792e-05, "loss": 0.0893, "step": 11860 }, { "grad_norm": 0.7672353386878967, "learning_rate": 9.871556731095661e-05, "loss": 0.098, "step": 11870 }, { "grad_norm": 0.7833588719367981, "learning_rate": 9.871184094660411e-05, "loss": 0.1036, "step": 11880 }, { "grad_norm": 0.590908408164978, "learning_rate": 9.870810925519791e-05, "loss": 0.0929, "step": 11890 }, { "grad_norm": 0.591867208480835, "learning_rate": 9.870437223714612e-05, "loss": 0.0873, "step": 11900 }, { "grad_norm": 0.46225371956825256, "learning_rate": 9.87006298928574e-05, "loss": 0.0869, "step": 11910 }, { "grad_norm": 0.46159201860427856, "learning_rate": 9.869688222274103e-05, "loss": 0.0721, "step": 11920 }, { "grad_norm": 0.47698280215263367, "learning_rate": 9.869312922720681e-05, "loss": 0.0869, "step": 11930 }, { "grad_norm": 0.4497362971305847, "learning_rate": 9.868937090666521e-05, "loss": 0.0823, "step": 11940 }, { "grad_norm": 0.3717261552810669, "learning_rate": 9.86856072615272e-05, "loss": 0.0681, "step": 11950 }, { "grad_norm": 0.7544646263122559, "learning_rate": 9.868183829220438e-05, "loss": 0.0764, "step": 11960 }, { "grad_norm": 0.842075526714325, "learning_rate": 9.867806399910893e-05, "loss": 0.0773, "step": 11970 }, { "grad_norm": 0.6852490305900574, "learning_rate": 9.867428438265356e-05, "loss": 0.0856, "step": 11980 }, { "grad_norm": 0.6609559059143066, "learning_rate": 9.867049944325165e-05, "loss": 0.1007, "step": 11990 }, { "grad_norm": 0.6038932800292969, "learning_rate": 9.86667091813171e-05, "loss": 0.0993, "step": 12000 }, { "grad_norm": 0.6033244132995605, "learning_rate": 9.866291359726438e-05, "loss": 0.091, "step": 12010 }, { "grad_norm": 0.8369919657707214, "learning_rate": 9.865911269150861e-05, "loss": 0.095, "step": 12020 }, { "grad_norm": 0.6335798501968384, "learning_rate": 9.865530646446544e-05, "loss": 0.0828, "step": 12030 }, { "grad_norm": 0.6164636611938477, "learning_rate": 9.86514949165511e-05, "loss": 0.0939, "step": 12040 }, { "grad_norm": 0.589577853679657, "learning_rate": 9.864767804818243e-05, "loss": 0.0983, "step": 12050 }, { "grad_norm": 0.6622703671455383, "learning_rate": 9.86438558597768e-05, "loss": 0.1, "step": 12060 }, { "grad_norm": 0.4205518960952759, "learning_rate": 9.864002835175225e-05, "loss": 0.0984, "step": 12070 }, { "grad_norm": 0.5587154030799866, "learning_rate": 9.863619552452734e-05, "loss": 0.0747, "step": 12080 }, { "grad_norm": 0.541785717010498, "learning_rate": 9.863235737852119e-05, "loss": 0.0598, "step": 12090 }, { "grad_norm": 0.8794878125190735, "learning_rate": 9.862851391415356e-05, "loss": 0.0824, "step": 12100 }, { "grad_norm": 0.6132757067680359, "learning_rate": 9.862466513184477e-05, "loss": 0.0879, "step": 12110 }, { "grad_norm": 0.5535143613815308, "learning_rate": 9.86208110320157e-05, "loss": 0.0743, "step": 12120 }, { "grad_norm": 0.7657831907272339, "learning_rate": 9.861695161508784e-05, "loss": 0.079, "step": 12130 }, { "grad_norm": 0.6155270934104919, "learning_rate": 9.861308688148324e-05, "loss": 0.082, "step": 12140 }, { "grad_norm": 0.6244989037513733, "learning_rate": 9.860921683162455e-05, "loss": 0.0846, "step": 12150 }, { "grad_norm": 0.6988179683685303, "learning_rate": 9.860534146593499e-05, "loss": 0.102, "step": 12160 }, { "grad_norm": 0.5941486358642578, "learning_rate": 9.860146078483836e-05, "loss": 0.0939, "step": 12170 }, { "grad_norm": 0.6906565427780151, "learning_rate": 9.859757478875905e-05, "loss": 0.089, "step": 12180 }, { "grad_norm": 0.584325909614563, "learning_rate": 9.859368347812204e-05, "loss": 0.0801, "step": 12190 }, { "grad_norm": 0.6763774752616882, "learning_rate": 9.858978685335285e-05, "loss": 0.0754, "step": 12200 }, { "grad_norm": 0.6870662569999695, "learning_rate": 9.858588491487763e-05, "loss": 0.0939, "step": 12210 }, { "grad_norm": 0.4981749653816223, "learning_rate": 9.858197766312308e-05, "loss": 0.0733, "step": 12220 }, { "grad_norm": 0.5769076347351074, "learning_rate": 9.857806509851649e-05, "loss": 0.0883, "step": 12230 }, { "grad_norm": 0.4677540361881256, "learning_rate": 9.857414722148574e-05, "loss": 0.089, "step": 12240 }, { "grad_norm": 0.6019818186759949, "learning_rate": 9.857022403245928e-05, "loss": 0.0949, "step": 12250 }, { "grad_norm": 0.5918934941291809, "learning_rate": 9.856629553186615e-05, "loss": 0.0928, "step": 12260 }, { "grad_norm": 0.5479993224143982, "learning_rate": 9.856236172013595e-05, "loss": 0.0834, "step": 12270 }, { "grad_norm": 0.6020552515983582, "learning_rate": 9.85584225976989e-05, "loss": 0.0891, "step": 12280 }, { "grad_norm": 1.1915258169174194, "learning_rate": 9.855447816498575e-05, "loss": 0.1121, "step": 12290 }, { "grad_norm": 0.8108825087547302, "learning_rate": 9.855052842242787e-05, "loss": 0.0813, "step": 12300 }, { "grad_norm": 0.60265052318573, "learning_rate": 9.85465733704572e-05, "loss": 0.0905, "step": 12310 }, { "grad_norm": 0.8330880403518677, "learning_rate": 9.854261300950624e-05, "loss": 0.0976, "step": 12320 }, { "grad_norm": 0.5754004120826721, "learning_rate": 9.853864734000813e-05, "loss": 0.0853, "step": 12330 }, { "grad_norm": 0.5606362223625183, "learning_rate": 9.85346763623965e-05, "loss": 0.1015, "step": 12340 }, { "grad_norm": 0.6360895037651062, "learning_rate": 9.853070007710564e-05, "loss": 0.0812, "step": 12350 }, { "grad_norm": 0.640274167060852, "learning_rate": 9.85267184845704e-05, "loss": 0.0808, "step": 12360 }, { "grad_norm": 0.737802267074585, "learning_rate": 9.852273158522616e-05, "loss": 0.1096, "step": 12370 }, { "grad_norm": 0.6802354454994202, "learning_rate": 9.851873937950896e-05, "loss": 0.0837, "step": 12380 }, { "grad_norm": 0.6598166227340698, "learning_rate": 9.851474186785537e-05, "loss": 0.0947, "step": 12390 }, { "grad_norm": 0.7289209365844727, "learning_rate": 9.851073905070254e-05, "loss": 0.095, "step": 12400 }, { "grad_norm": 0.9690006971359253, "learning_rate": 9.850673092848824e-05, "loss": 0.1143, "step": 12410 }, { "grad_norm": 0.6668169498443604, "learning_rate": 9.850271750165077e-05, "loss": 0.0718, "step": 12420 }, { "grad_norm": 0.9728941917419434, "learning_rate": 9.849869877062902e-05, "loss": 0.1, "step": 12430 }, { "grad_norm": 0.6364112496376038, "learning_rate": 9.849467473586252e-05, "loss": 0.0954, "step": 12440 }, { "grad_norm": 0.667014479637146, "learning_rate": 9.849064539779127e-05, "loss": 0.0863, "step": 12450 }, { "grad_norm": 0.7143647074699402, "learning_rate": 9.848661075685594e-05, "loss": 0.0896, "step": 12460 }, { "grad_norm": 0.6506759524345398, "learning_rate": 9.848257081349778e-05, "loss": 0.0987, "step": 12470 }, { "grad_norm": 0.5242937803268433, "learning_rate": 9.847852556815856e-05, "loss": 0.0848, "step": 12480 }, { "grad_norm": 0.6101565361022949, "learning_rate": 9.847447502128067e-05, "loss": 0.0951, "step": 12490 }, { "grad_norm": 0.5846071243286133, "learning_rate": 9.847041917330708e-05, "loss": 0.0934, "step": 12500 }, { "grad_norm": 0.5360637307167053, "learning_rate": 9.846635802468132e-05, "loss": 0.0911, "step": 12510 }, { "grad_norm": 0.5762774348258972, "learning_rate": 9.84622915758475e-05, "loss": 0.0821, "step": 12520 }, { "grad_norm": 0.5330851078033447, "learning_rate": 9.845821982725034e-05, "loss": 0.077, "step": 12530 }, { "grad_norm": 0.560939610004425, "learning_rate": 9.845414277933514e-05, "loss": 0.0836, "step": 12540 }, { "grad_norm": 0.6193766593933105, "learning_rate": 9.845006043254771e-05, "loss": 0.0935, "step": 12550 }, { "grad_norm": 0.7083591818809509, "learning_rate": 9.844597278733451e-05, "loss": 0.0901, "step": 12560 }, { "grad_norm": 1.3162645101547241, "learning_rate": 9.844187984414259e-05, "loss": 0.0784, "step": 12570 }, { "grad_norm": 0.5834406018257141, "learning_rate": 9.84377816034195e-05, "loss": 0.0794, "step": 12580 }, { "grad_norm": 0.6130528450012207, "learning_rate": 9.843367806561345e-05, "loss": 0.0835, "step": 12590 }, { "grad_norm": 0.5808354616165161, "learning_rate": 9.842956923117317e-05, "loss": 0.0738, "step": 12600 }, { "grad_norm": 0.4870161712169647, "learning_rate": 9.842545510054802e-05, "loss": 0.0621, "step": 12610 }, { "grad_norm": 0.6653697490692139, "learning_rate": 9.842133567418792e-05, "loss": 0.0753, "step": 12620 }, { "grad_norm": 0.472342848777771, "learning_rate": 9.841721095254333e-05, "loss": 0.0669, "step": 12630 }, { "grad_norm": 0.6931574940681458, "learning_rate": 9.841308093606537e-05, "loss": 0.0803, "step": 12640 }, { "grad_norm": 0.5263307094573975, "learning_rate": 9.840894562520565e-05, "loss": 0.0775, "step": 12650 }, { "grad_norm": 0.9952122569084167, "learning_rate": 9.840480502041642e-05, "loss": 0.1031, "step": 12660 }, { "grad_norm": 0.5485883355140686, "learning_rate": 9.840065912215049e-05, "loss": 0.08, "step": 12670 }, { "grad_norm": 0.6218370795249939, "learning_rate": 9.839650793086124e-05, "loss": 0.1034, "step": 12680 }, { "grad_norm": 0.40278366208076477, "learning_rate": 9.839235144700265e-05, "loss": 0.0784, "step": 12690 }, { "grad_norm": 0.4900897443294525, "learning_rate": 9.838818967102926e-05, "loss": 0.078, "step": 12700 }, { "grad_norm": 0.5322786569595337, "learning_rate": 9.83840226033962e-05, "loss": 0.0799, "step": 12710 }, { "grad_norm": 0.5833542943000793, "learning_rate": 9.837985024455918e-05, "loss": 0.0719, "step": 12720 }, { "grad_norm": 0.5322002172470093, "learning_rate": 9.837567259497447e-05, "loss": 0.0768, "step": 12730 }, { "grad_norm": 0.5234118103981018, "learning_rate": 9.837148965509894e-05, "loss": 0.0753, "step": 12740 }, { "grad_norm": 1.0674549341201782, "learning_rate": 9.836730142539001e-05, "loss": 0.075, "step": 12750 }, { "grad_norm": 0.6931951642036438, "learning_rate": 9.836310790630574e-05, "loss": 0.1056, "step": 12760 }, { "grad_norm": 0.8436711430549622, "learning_rate": 9.83589090983047e-05, "loss": 0.0921, "step": 12770 }, { "grad_norm": 0.7304542660713196, "learning_rate": 9.835470500184605e-05, "loss": 0.086, "step": 12780 }, { "grad_norm": 0.6635867953300476, "learning_rate": 9.835049561738957e-05, "loss": 0.084, "step": 12790 }, { "grad_norm": 0.7168521285057068, "learning_rate": 9.834628094539558e-05, "loss": 0.0806, "step": 12800 }, { "grad_norm": 0.8213154077529907, "learning_rate": 9.834206098632499e-05, "loss": 0.0961, "step": 12810 }, { "grad_norm": 0.5746742486953735, "learning_rate": 9.833783574063931e-05, "loss": 0.1047, "step": 12820 }, { "grad_norm": 0.7375064492225647, "learning_rate": 9.833360520880058e-05, "loss": 0.1038, "step": 12830 }, { "grad_norm": 0.5347227454185486, "learning_rate": 9.832936939127144e-05, "loss": 0.0817, "step": 12840 }, { "grad_norm": 0.7746209502220154, "learning_rate": 9.832512828851515e-05, "loss": 0.1122, "step": 12850 }, { "grad_norm": 0.5697305798530579, "learning_rate": 9.832088190099546e-05, "loss": 0.0842, "step": 12860 }, { "grad_norm": 0.636292040348053, "learning_rate": 9.831663022917679e-05, "loss": 0.0892, "step": 12870 }, { "grad_norm": 0.6060138940811157, "learning_rate": 9.831237327352407e-05, "loss": 0.0793, "step": 12880 }, { "grad_norm": 0.6782150864601135, "learning_rate": 9.830811103450286e-05, "loss": 0.0856, "step": 12890 }, { "grad_norm": 0.5504676699638367, "learning_rate": 9.830384351257924e-05, "loss": 0.0868, "step": 12900 }, { "grad_norm": 0.7911955118179321, "learning_rate": 9.829957070821993e-05, "loss": 0.0872, "step": 12910 }, { "grad_norm": 0.5913336277008057, "learning_rate": 9.829529262189218e-05, "loss": 0.0672, "step": 12920 }, { "grad_norm": 0.5269050002098083, "learning_rate": 9.829100925406385e-05, "loss": 0.0776, "step": 12930 }, { "grad_norm": 0.6718311309814453, "learning_rate": 9.828672060520333e-05, "loss": 0.0833, "step": 12940 }, { "grad_norm": 1.045335292816162, "learning_rate": 9.828242667577966e-05, "loss": 0.0686, "step": 12950 }, { "grad_norm": 0.7453465461730957, "learning_rate": 9.82781274662624e-05, "loss": 0.0854, "step": 12960 }, { "grad_norm": 0.6559438109397888, "learning_rate": 9.82738229771217e-05, "loss": 0.085, "step": 12970 }, { "grad_norm": 0.47651731967926025, "learning_rate": 9.826951320882829e-05, "loss": 0.0783, "step": 12980 }, { "grad_norm": 0.9461248517036438, "learning_rate": 9.826519816185351e-05, "loss": 0.0768, "step": 12990 }, { "grad_norm": 0.6621875166893005, "learning_rate": 9.826087783666921e-05, "loss": 0.0844, "step": 13000 }, { "grad_norm": 0.5792906880378723, "learning_rate": 9.825655223374787e-05, "loss": 0.0985, "step": 13010 }, { "grad_norm": 0.8468583226203918, "learning_rate": 9.825222135356253e-05, "loss": 0.103, "step": 13020 }, { "grad_norm": 0.611594021320343, "learning_rate": 9.82478851965868e-05, "loss": 0.0988, "step": 13030 }, { "grad_norm": 0.5092558860778809, "learning_rate": 9.82435437632949e-05, "loss": 0.0689, "step": 13040 }, { "grad_norm": 0.7316983342170715, "learning_rate": 9.823919705416158e-05, "loss": 0.087, "step": 13050 }, { "grad_norm": 0.8789944052696228, "learning_rate": 9.82348450696622e-05, "loss": 0.0906, "step": 13060 }, { "grad_norm": 0.6997138261795044, "learning_rate": 9.823048781027268e-05, "loss": 0.1039, "step": 13070 }, { "grad_norm": 0.6180400252342224, "learning_rate": 9.822612527646953e-05, "loss": 0.0903, "step": 13080 }, { "grad_norm": 0.613204300403595, "learning_rate": 9.822175746872984e-05, "loss": 0.0996, "step": 13090 }, { "grad_norm": 0.7401745319366455, "learning_rate": 9.821738438753123e-05, "loss": 0.0856, "step": 13100 }, { "grad_norm": 0.6860005855560303, "learning_rate": 9.821300603335196e-05, "loss": 0.0853, "step": 13110 }, { "grad_norm": 0.5117696523666382, "learning_rate": 9.820862240667085e-05, "loss": 0.0977, "step": 13120 }, { "grad_norm": 0.6708999276161194, "learning_rate": 9.820423350796726e-05, "loss": 0.0807, "step": 13130 }, { "grad_norm": 0.948148787021637, "learning_rate": 9.819983933772118e-05, "loss": 0.0893, "step": 13140 }, { "grad_norm": 0.7195096611976624, "learning_rate": 9.819543989641314e-05, "loss": 0.0841, "step": 13150 }, { "grad_norm": 0.4697469174861908, "learning_rate": 9.819103518452423e-05, "loss": 0.0815, "step": 13160 }, { "grad_norm": 0.6724469065666199, "learning_rate": 9.818662520253618e-05, "loss": 0.096, "step": 13170 }, { "grad_norm": 0.485237717628479, "learning_rate": 9.818220995093126e-05, "loss": 0.0794, "step": 13180 }, { "grad_norm": 0.7158761620521545, "learning_rate": 9.817778943019228e-05, "loss": 0.0812, "step": 13190 }, { "grad_norm": 0.9045717120170593, "learning_rate": 9.81733636408027e-05, "loss": 0.0918, "step": 13200 }, { "grad_norm": 0.6968690752983093, "learning_rate": 9.816893258324649e-05, "loss": 0.0672, "step": 13210 }, { "grad_norm": 0.6688721776008606, "learning_rate": 9.816449625800823e-05, "loss": 0.0701, "step": 13220 }, { "grad_norm": 0.5121040344238281, "learning_rate": 9.816005466557308e-05, "loss": 0.0828, "step": 13230 }, { "grad_norm": 0.6223666071891785, "learning_rate": 9.815560780642674e-05, "loss": 0.0995, "step": 13240 }, { "grad_norm": 0.5140635371208191, "learning_rate": 9.815115568105555e-05, "loss": 0.075, "step": 13250 }, { "grad_norm": 0.5156570672988892, "learning_rate": 9.814669828994638e-05, "loss": 0.0882, "step": 13260 }, { "grad_norm": 0.5102458000183105, "learning_rate": 9.814223563358665e-05, "loss": 0.0805, "step": 13270 }, { "grad_norm": 0.582868218421936, "learning_rate": 9.813776771246443e-05, "loss": 0.0781, "step": 13280 }, { "grad_norm": 0.4768914580345154, "learning_rate": 9.813329452706829e-05, "loss": 0.0793, "step": 13290 }, { "grad_norm": 0.5658367276191711, "learning_rate": 9.812881607788744e-05, "loss": 0.0814, "step": 13300 }, { "grad_norm": 0.3524906635284424, "learning_rate": 9.812433236541163e-05, "loss": 0.0801, "step": 13310 }, { "grad_norm": 0.7667546272277832, "learning_rate": 9.811984339013116e-05, "loss": 0.088, "step": 13320 }, { "grad_norm": 0.48765718936920166, "learning_rate": 9.811534915253698e-05, "loss": 0.0727, "step": 13330 }, { "grad_norm": 0.5358437299728394, "learning_rate": 9.811084965312056e-05, "loss": 0.0885, "step": 13340 }, { "grad_norm": 0.7595519423484802, "learning_rate": 9.810634489237396e-05, "loss": 0.0806, "step": 13350 }, { "grad_norm": 0.7877787947654724, "learning_rate": 9.81018348707898e-05, "loss": 0.1007, "step": 13360 }, { "grad_norm": 0.671251654624939, "learning_rate": 9.809731958886131e-05, "loss": 0.0815, "step": 13370 }, { "grad_norm": 0.7652460932731628, "learning_rate": 9.809279904708224e-05, "loss": 0.1014, "step": 13380 }, { "grad_norm": 0.5511196851730347, "learning_rate": 9.808827324594699e-05, "loss": 0.0919, "step": 13390 }, { "grad_norm": 0.4690355956554413, "learning_rate": 9.808374218595046e-05, "loss": 0.0982, "step": 13400 }, { "grad_norm": 0.698087751865387, "learning_rate": 9.80792058675882e-05, "loss": 0.0853, "step": 13410 }, { "grad_norm": 0.5304948091506958, "learning_rate": 9.807466429135627e-05, "loss": 0.0821, "step": 13420 }, { "grad_norm": 0.49970388412475586, "learning_rate": 9.807011745775132e-05, "loss": 0.0862, "step": 13430 }, { "grad_norm": 0.49381953477859497, "learning_rate": 9.806556536727061e-05, "loss": 0.0802, "step": 13440 }, { "grad_norm": 0.5194369554519653, "learning_rate": 9.806100802041193e-05, "loss": 0.1035, "step": 13450 }, { "grad_norm": 0.45977750420570374, "learning_rate": 9.805644541767368e-05, "loss": 0.0931, "step": 13460 }, { "grad_norm": 0.448278546333313, "learning_rate": 9.805187755955478e-05, "loss": 0.0935, "step": 13470 }, { "grad_norm": 0.355703741312027, "learning_rate": 9.804730444655483e-05, "loss": 0.0777, "step": 13480 }, { "grad_norm": 0.5918381810188293, "learning_rate": 9.804272607917388e-05, "loss": 0.0902, "step": 13490 }, { "grad_norm": 0.6067783832550049, "learning_rate": 9.803814245791265e-05, "loss": 0.0742, "step": 13500 }, { "grad_norm": 0.5939932465553284, "learning_rate": 9.803355358327239e-05, "loss": 0.0886, "step": 13510 }, { "grad_norm": 0.44758141040802, "learning_rate": 9.802895945575492e-05, "loss": 0.0807, "step": 13520 }, { "grad_norm": 0.596287190914154, "learning_rate": 9.802436007586266e-05, "loss": 0.0976, "step": 13530 }, { "grad_norm": 0.5169705152511597, "learning_rate": 9.801975544409858e-05, "loss": 0.068, "step": 13540 }, { "grad_norm": 0.71782386302948, "learning_rate": 9.801514556096625e-05, "loss": 0.0789, "step": 13550 }, { "grad_norm": 0.5426823496818542, "learning_rate": 9.801053042696977e-05, "loss": 0.0836, "step": 13560 }, { "grad_norm": 0.4492107033729553, "learning_rate": 9.800591004261388e-05, "loss": 0.0725, "step": 13570 }, { "grad_norm": 0.604655921459198, "learning_rate": 9.800128440840385e-05, "loss": 0.0796, "step": 13580 }, { "grad_norm": 0.599908173084259, "learning_rate": 9.799665352484552e-05, "loss": 0.1026, "step": 13590 }, { "grad_norm": 0.6784942150115967, "learning_rate": 9.799201739244532e-05, "loss": 0.111, "step": 13600 }, { "grad_norm": 0.5797923803329468, "learning_rate": 9.798737601171025e-05, "loss": 0.082, "step": 13610 }, { "grad_norm": 0.6086942553520203, "learning_rate": 9.79827293831479e-05, "loss": 0.099, "step": 13620 }, { "grad_norm": 0.4960779547691345, "learning_rate": 9.797807750726638e-05, "loss": 0.0693, "step": 13630 }, { "grad_norm": 0.49975162744522095, "learning_rate": 9.797342038457446e-05, "loss": 0.0761, "step": 13640 }, { "grad_norm": 0.6119379997253418, "learning_rate": 9.796875801558141e-05, "loss": 0.0842, "step": 13650 }, { "grad_norm": 0.6582351326942444, "learning_rate": 9.79640904007971e-05, "loss": 0.0973, "step": 13660 }, { "grad_norm": 0.7012725472450256, "learning_rate": 9.795941754073199e-05, "loss": 0.079, "step": 13670 }, { "grad_norm": 0.5100688934326172, "learning_rate": 9.795473943589705e-05, "loss": 0.0742, "step": 13680 }, { "grad_norm": 0.8134443163871765, "learning_rate": 9.795005608680394e-05, "loss": 0.0819, "step": 13690 }, { "grad_norm": 0.8062583804130554, "learning_rate": 9.794536749396477e-05, "loss": 0.1185, "step": 13700 }, { "grad_norm": 0.6305544972419739, "learning_rate": 9.79406736578923e-05, "loss": 0.1016, "step": 13710 }, { "grad_norm": 0.5936004519462585, "learning_rate": 9.793597457909984e-05, "loss": 0.0945, "step": 13720 }, { "grad_norm": 0.4877772927284241, "learning_rate": 9.793127025810127e-05, "loss": 0.0807, "step": 13730 }, { "grad_norm": 0.5549484491348267, "learning_rate": 9.792656069541104e-05, "loss": 0.0828, "step": 13740 }, { "grad_norm": 0.7842921614646912, "learning_rate": 9.79218458915442e-05, "loss": 0.0857, "step": 13750 }, { "grad_norm": 0.5358951687812805, "learning_rate": 9.791712584701634e-05, "loss": 0.0881, "step": 13760 }, { "grad_norm": 0.43399038910865784, "learning_rate": 9.791240056234364e-05, "loss": 0.0778, "step": 13770 }, { "grad_norm": 0.5131624341011047, "learning_rate": 9.790767003804283e-05, "loss": 0.0773, "step": 13780 }, { "grad_norm": 0.5057307481765747, "learning_rate": 9.790293427463126e-05, "loss": 0.0731, "step": 13790 }, { "grad_norm": 0.5771039724349976, "learning_rate": 9.789819327262684e-05, "loss": 0.0763, "step": 13800 }, { "grad_norm": 0.624729573726654, "learning_rate": 9.7893447032548e-05, "loss": 0.0943, "step": 13810 }, { "grad_norm": 0.8537721037864685, "learning_rate": 9.78886955549138e-05, "loss": 0.0931, "step": 13820 }, { "grad_norm": 0.5207360982894897, "learning_rate": 9.788393884024387e-05, "loss": 0.0957, "step": 13830 }, { "grad_norm": 0.5922811031341553, "learning_rate": 9.787917688905836e-05, "loss": 0.0917, "step": 13840 }, { "grad_norm": 0.5597848892211914, "learning_rate": 9.787440970187807e-05, "loss": 0.069, "step": 13850 }, { "grad_norm": 0.8003806471824646, "learning_rate": 9.786963727922429e-05, "loss": 0.0861, "step": 13860 }, { "grad_norm": 0.7630233764648438, "learning_rate": 9.786485962161897e-05, "loss": 0.082, "step": 13870 }, { "grad_norm": 0.9925993084907532, "learning_rate": 9.786007672958455e-05, "loss": 0.0864, "step": 13880 }, { "grad_norm": 0.7174705862998962, "learning_rate": 9.78552886036441e-05, "loss": 0.0887, "step": 13890 }, { "grad_norm": 0.721694827079773, "learning_rate": 9.785049524432124e-05, "loss": 0.0831, "step": 13900 }, { "grad_norm": 0.7085034251213074, "learning_rate": 9.784569665214016e-05, "loss": 0.0956, "step": 13910 }, { "grad_norm": 0.6397092342376709, "learning_rate": 9.784089282762563e-05, "loss": 0.0856, "step": 13920 }, { "grad_norm": 0.5527456402778625, "learning_rate": 9.7836083771303e-05, "loss": 0.0915, "step": 13930 }, { "grad_norm": 0.6820744872093201, "learning_rate": 9.783126948369817e-05, "loss": 0.0928, "step": 13940 }, { "grad_norm": 0.6270164847373962, "learning_rate": 9.78264499653376e-05, "loss": 0.0891, "step": 13950 }, { "grad_norm": 0.5187535881996155, "learning_rate": 9.782162521674838e-05, "loss": 0.0934, "step": 13960 }, { "grad_norm": 0.5177817344665527, "learning_rate": 9.781679523845812e-05, "loss": 0.1032, "step": 13970 }, { "grad_norm": 0.8058435916900635, "learning_rate": 9.781196003099502e-05, "loss": 0.0932, "step": 13980 }, { "grad_norm": 0.4640423357486725, "learning_rate": 9.780711959488786e-05, "loss": 0.0688, "step": 13990 }, { "grad_norm": 0.4269220530986786, "learning_rate": 9.780227393066599e-05, "loss": 0.0714, "step": 14000 }, { "grad_norm": 0.5072428584098816, "learning_rate": 9.77974230388593e-05, "loss": 0.0882, "step": 14010 }, { "grad_norm": 0.5702025890350342, "learning_rate": 9.779256691999829e-05, "loss": 0.1007, "step": 14020 }, { "grad_norm": 0.6140459179878235, "learning_rate": 9.778770557461403e-05, "loss": 0.1007, "step": 14030 }, { "grad_norm": 0.6525800824165344, "learning_rate": 9.778283900323812e-05, "loss": 0.0756, "step": 14040 }, { "grad_norm": 0.6245748996734619, "learning_rate": 9.777796720640277e-05, "loss": 0.0848, "step": 14050 }, { "grad_norm": 0.5173306465148926, "learning_rate": 9.777309018464078e-05, "loss": 0.0672, "step": 14060 }, { "grad_norm": 0.5295171141624451, "learning_rate": 9.776820793848547e-05, "loss": 0.0791, "step": 14070 }, { "grad_norm": 0.6045582890510559, "learning_rate": 9.776332046847075e-05, "loss": 0.0703, "step": 14080 }, { "grad_norm": 0.7282732725143433, "learning_rate": 9.775842777513111e-05, "loss": 0.0804, "step": 14090 }, { "grad_norm": 0.5420275926589966, "learning_rate": 9.775352985900163e-05, "loss": 0.0915, "step": 14100 }, { "grad_norm": 0.7024986743927002, "learning_rate": 9.774862672061791e-05, "loss": 0.085, "step": 14110 }, { "grad_norm": 0.45623767375946045, "learning_rate": 9.774371836051616e-05, "loss": 0.0863, "step": 14120 }, { "grad_norm": 0.8860058784484863, "learning_rate": 9.773880477923315e-05, "loss": 0.0862, "step": 14130 }, { "grad_norm": 0.8947244882583618, "learning_rate": 9.773388597730623e-05, "loss": 0.0878, "step": 14140 }, { "grad_norm": 1.1519330739974976, "learning_rate": 9.77289619552733e-05, "loss": 0.0947, "step": 14150 }, { "grad_norm": 0.7419194579124451, "learning_rate": 9.772403271367285e-05, "loss": 0.0954, "step": 14160 }, { "grad_norm": 0.478170782327652, "learning_rate": 9.771909825304396e-05, "loss": 0.0844, "step": 14170 }, { "grad_norm": 0.4884116053581238, "learning_rate": 9.771415857392619e-05, "loss": 0.0999, "step": 14180 }, { "grad_norm": 1.5261956453323364, "learning_rate": 9.770921367685978e-05, "loss": 0.1049, "step": 14190 }, { "grad_norm": 0.7102028727531433, "learning_rate": 9.770426356238551e-05, "loss": 0.0896, "step": 14200 }, { "grad_norm": 0.6639786958694458, "learning_rate": 9.769930823104469e-05, "loss": 0.0962, "step": 14210 }, { "grad_norm": 0.5203274488449097, "learning_rate": 9.769434768337926e-05, "loss": 0.0905, "step": 14220 }, { "grad_norm": 0.5992172360420227, "learning_rate": 9.768938191993164e-05, "loss": 0.0856, "step": 14230 }, { "grad_norm": 0.6054806709289551, "learning_rate": 9.768441094124494e-05, "loss": 0.0829, "step": 14240 }, { "grad_norm": 0.5927041172981262, "learning_rate": 9.767943474786275e-05, "loss": 0.0974, "step": 14250 }, { "grad_norm": 0.6235475540161133, "learning_rate": 9.767445334032923e-05, "loss": 0.0628, "step": 14260 }, { "grad_norm": 0.9256267547607422, "learning_rate": 9.766946671918919e-05, "loss": 0.0962, "step": 14270 }, { "grad_norm": 0.6740541458129883, "learning_rate": 9.766447488498796e-05, "loss": 0.0711, "step": 14280 }, { "grad_norm": 0.5747337937355042, "learning_rate": 9.765947783827139e-05, "loss": 0.0799, "step": 14290 }, { "grad_norm": 0.7617997527122498, "learning_rate": 9.765447557958599e-05, "loss": 0.0756, "step": 14300 }, { "grad_norm": 0.5719621181488037, "learning_rate": 9.764946810947879e-05, "loss": 0.076, "step": 14310 }, { "grad_norm": 0.42914748191833496, "learning_rate": 9.764445542849738e-05, "loss": 0.0916, "step": 14320 }, { "grad_norm": 0.6283689141273499, "learning_rate": 9.763943753718998e-05, "loss": 0.0851, "step": 14330 }, { "grad_norm": 0.5818926095962524, "learning_rate": 9.76344144361053e-05, "loss": 0.0777, "step": 14340 }, { "grad_norm": 0.6926578283309937, "learning_rate": 9.762938612579269e-05, "loss": 0.0932, "step": 14350 }, { "grad_norm": 0.6217115521430969, "learning_rate": 9.762435260680202e-05, "loss": 0.0808, "step": 14360 }, { "grad_norm": 0.8428274393081665, "learning_rate": 9.761931387968373e-05, "loss": 0.0884, "step": 14370 }, { "grad_norm": 0.44191214442253113, "learning_rate": 9.76142699449889e-05, "loss": 0.0741, "step": 14380 }, { "grad_norm": 0.47866061329841614, "learning_rate": 9.760922080326908e-05, "loss": 0.0922, "step": 14390 }, { "grad_norm": 0.5946398973464966, "learning_rate": 9.760416645507644e-05, "loss": 0.0796, "step": 14400 }, { "grad_norm": 0.5273325443267822, "learning_rate": 9.759910690096375e-05, "loss": 0.0827, "step": 14410 }, { "grad_norm": 0.6479106545448303, "learning_rate": 9.759404214148429e-05, "loss": 0.0894, "step": 14420 }, { "grad_norm": 0.5395489931106567, "learning_rate": 9.758897217719191e-05, "loss": 0.0844, "step": 14430 }, { "grad_norm": 0.5594136714935303, "learning_rate": 9.758389700864113e-05, "loss": 0.0765, "step": 14440 }, { "grad_norm": 0.9559556841850281, "learning_rate": 9.757881663638688e-05, "loss": 0.0951, "step": 14450 }, { "grad_norm": 0.5803672075271606, "learning_rate": 9.757373106098478e-05, "loss": 0.0854, "step": 14460 }, { "grad_norm": 0.9653894305229187, "learning_rate": 9.756864028299097e-05, "loss": 0.0919, "step": 14470 }, { "grad_norm": 0.6580811738967896, "learning_rate": 9.75635443029622e-05, "loss": 0.0921, "step": 14480 }, { "grad_norm": 0.6135696172714233, "learning_rate": 9.755844312145572e-05, "loss": 0.0787, "step": 14490 }, { "grad_norm": 0.5326455235481262, "learning_rate": 9.755333673902941e-05, "loss": 0.0998, "step": 14500 }, { "grad_norm": 0.6260656118392944, "learning_rate": 9.75482251562417e-05, "loss": 0.078, "step": 14510 }, { "grad_norm": 0.5673891305923462, "learning_rate": 9.754310837365155e-05, "loss": 0.0766, "step": 14520 }, { "grad_norm": 0.5099547505378723, "learning_rate": 9.753798639181856e-05, "loss": 0.1003, "step": 14530 }, { "grad_norm": 0.6554941534996033, "learning_rate": 9.753285921130286e-05, "loss": 0.0798, "step": 14540 }, { "grad_norm": 0.587726354598999, "learning_rate": 9.752772683266512e-05, "loss": 0.0788, "step": 14550 }, { "grad_norm": 0.8592720031738281, "learning_rate": 9.752258925646665e-05, "loss": 0.088, "step": 14560 }, { "grad_norm": 0.8383951783180237, "learning_rate": 9.751744648326926e-05, "loss": 0.073, "step": 14570 }, { "grad_norm": 0.6329952478408813, "learning_rate": 9.751229851363536e-05, "loss": 0.0829, "step": 14580 }, { "grad_norm": 0.5833433270454407, "learning_rate": 9.750714534812793e-05, "loss": 0.1137, "step": 14590 }, { "grad_norm": 0.6974145174026489, "learning_rate": 9.750198698731053e-05, "loss": 0.0937, "step": 14600 }, { "grad_norm": 2.092395305633545, "learning_rate": 9.749682343174722e-05, "loss": 0.0807, "step": 14610 }, { "grad_norm": 0.4423055052757263, "learning_rate": 9.749165468200272e-05, "loss": 0.0756, "step": 14620 }, { "grad_norm": 0.49080541729927063, "learning_rate": 9.748648073864229e-05, "loss": 0.0846, "step": 14630 }, { "grad_norm": 0.5380358695983887, "learning_rate": 9.748130160223168e-05, "loss": 0.0707, "step": 14640 }, { "grad_norm": 2.6828255653381348, "learning_rate": 9.747611727333734e-05, "loss": 0.0998, "step": 14650 }, { "grad_norm": 0.5975819230079651, "learning_rate": 9.74709277525262e-05, "loss": 0.0871, "step": 14660 }, { "grad_norm": 0.4825593829154968, "learning_rate": 9.746573304036576e-05, "loss": 0.0893, "step": 14670 }, { "grad_norm": 0.7324846386909485, "learning_rate": 9.746053313742412e-05, "loss": 0.0677, "step": 14680 }, { "grad_norm": 0.6200469136238098, "learning_rate": 9.745532804426994e-05, "loss": 0.0879, "step": 14690 }, { "grad_norm": 0.6012254357337952, "learning_rate": 9.745011776147242e-05, "loss": 0.0769, "step": 14700 }, { "grad_norm": 0.793343186378479, "learning_rate": 9.744490228960138e-05, "loss": 0.0894, "step": 14710 }, { "grad_norm": 0.6612326502799988, "learning_rate": 9.743968162922713e-05, "loss": 0.0735, "step": 14720 }, { "grad_norm": 0.6182848215103149, "learning_rate": 9.743445578092064e-05, "loss": 0.0927, "step": 14730 }, { "grad_norm": 0.41753271222114563, "learning_rate": 9.742922474525338e-05, "loss": 0.0937, "step": 14740 }, { "grad_norm": 0.7963870763778687, "learning_rate": 9.742398852279741e-05, "loss": 0.0713, "step": 14750 }, { "grad_norm": 0.5651687383651733, "learning_rate": 9.741874711412535e-05, "loss": 0.0964, "step": 14760 }, { "grad_norm": 0.7157779335975647, "learning_rate": 9.741350051981042e-05, "loss": 0.0771, "step": 14770 }, { "grad_norm": 0.7745670080184937, "learning_rate": 9.740824874042633e-05, "loss": 0.0701, "step": 14780 }, { "grad_norm": 0.8933091163635254, "learning_rate": 9.740299177654746e-05, "loss": 0.0723, "step": 14790 }, { "grad_norm": 0.7278118133544922, "learning_rate": 9.739772962874867e-05, "loss": 0.0859, "step": 14800 }, { "grad_norm": 0.4291171431541443, "learning_rate": 9.739246229760541e-05, "loss": 0.0698, "step": 14810 }, { "grad_norm": 0.5806509256362915, "learning_rate": 9.738718978369376e-05, "loss": 0.0842, "step": 14820 }, { "grad_norm": 0.491334468126297, "learning_rate": 9.738191208759025e-05, "loss": 0.0867, "step": 14830 }, { "grad_norm": 0.7584007978439331, "learning_rate": 9.73766292098721e-05, "loss": 0.1028, "step": 14840 }, { "grad_norm": 0.6311973333358765, "learning_rate": 9.737134115111699e-05, "loss": 0.1055, "step": 14850 }, { "grad_norm": 0.6109609007835388, "learning_rate": 9.736604791190323e-05, "loss": 0.0875, "step": 14860 }, { "grad_norm": 0.4657033383846283, "learning_rate": 9.73607494928097e-05, "loss": 0.0696, "step": 14870 }, { "grad_norm": 0.4808611571788788, "learning_rate": 9.735544589441581e-05, "loss": 0.0822, "step": 14880 }, { "grad_norm": 0.5751407742500305, "learning_rate": 9.735013711730154e-05, "loss": 0.0683, "step": 14890 }, { "grad_norm": 0.4113808572292328, "learning_rate": 9.734482316204747e-05, "loss": 0.0676, "step": 14900 }, { "grad_norm": 0.7678288817405701, "learning_rate": 9.733950402923473e-05, "loss": 0.0754, "step": 14910 }, { "grad_norm": 0.4112948477268219, "learning_rate": 9.7334179719445e-05, "loss": 0.0667, "step": 14920 }, { "grad_norm": 0.6317963600158691, "learning_rate": 9.732885023326053e-05, "loss": 0.0821, "step": 14930 }, { "grad_norm": 0.5007804036140442, "learning_rate": 9.732351557126418e-05, "loss": 0.1086, "step": 14940 }, { "grad_norm": 0.6909951567649841, "learning_rate": 9.731817573403929e-05, "loss": 0.0869, "step": 14950 }, { "grad_norm": 0.8755314946174622, "learning_rate": 9.731283072216985e-05, "loss": 0.0812, "step": 14960 }, { "grad_norm": 0.5689918994903564, "learning_rate": 9.730748053624039e-05, "loss": 0.0686, "step": 14970 }, { "grad_norm": 0.7928932309150696, "learning_rate": 9.730212517683598e-05, "loss": 0.0975, "step": 14980 }, { "grad_norm": 0.5167352557182312, "learning_rate": 9.729676464454228e-05, "loss": 0.079, "step": 14990 }, { "grad_norm": 0.4183594882488251, "learning_rate": 9.72913989399455e-05, "loss": 0.0762, "step": 15000 }, { "grad_norm": 0.4991823434829712, "learning_rate": 9.728602806363242e-05, "loss": 0.0709, "step": 15010 }, { "grad_norm": 0.5366695523262024, "learning_rate": 9.728065201619043e-05, "loss": 0.0808, "step": 15020 }, { "grad_norm": 0.5741288661956787, "learning_rate": 9.727527079820742e-05, "loss": 0.0842, "step": 15030 }, { "grad_norm": 0.6003832221031189, "learning_rate": 9.726988441027186e-05, "loss": 0.101, "step": 15040 }, { "grad_norm": 0.548185408115387, "learning_rate": 9.726449285297281e-05, "loss": 0.0959, "step": 15050 }, { "grad_norm": 0.6540589928627014, "learning_rate": 9.72590961268999e-05, "loss": 0.1005, "step": 15060 }, { "grad_norm": 0.8414568901062012, "learning_rate": 9.725369423264328e-05, "loss": 0.0818, "step": 15070 }, { "grad_norm": 0.6006522178649902, "learning_rate": 9.72482871707937e-05, "loss": 0.0762, "step": 15080 }, { "grad_norm": 0.6051322221755981, "learning_rate": 9.724287494194247e-05, "loss": 0.0891, "step": 15090 }, { "grad_norm": 0.48553386330604553, "learning_rate": 9.723745754668147e-05, "loss": 0.0826, "step": 15100 }, { "grad_norm": 0.49134984612464905, "learning_rate": 9.723203498560313e-05, "loss": 0.0928, "step": 15110 }, { "grad_norm": 0.4504914879798889, "learning_rate": 9.722660725930046e-05, "loss": 0.0684, "step": 15120 }, { "grad_norm": 0.7915201187133789, "learning_rate": 9.722117436836702e-05, "loss": 0.0767, "step": 15130 }, { "grad_norm": 0.6557609438896179, "learning_rate": 9.721573631339696e-05, "loss": 0.095, "step": 15140 }, { "grad_norm": 0.8674782514572144, "learning_rate": 9.721029309498494e-05, "loss": 0.1131, "step": 15150 }, { "grad_norm": 0.7394811511039734, "learning_rate": 9.720484471372627e-05, "loss": 0.0952, "step": 15160 }, { "grad_norm": 0.5537139773368835, "learning_rate": 9.719939117021673e-05, "loss": 0.0909, "step": 15170 }, { "grad_norm": 0.7428058385848999, "learning_rate": 9.719393246505275e-05, "loss": 0.098, "step": 15180 }, { "grad_norm": 0.670072078704834, "learning_rate": 9.718846859883128e-05, "loss": 0.0865, "step": 15190 }, { "grad_norm": 0.455806165933609, "learning_rate": 9.718299957214982e-05, "loss": 0.0954, "step": 15200 }, { "grad_norm": 0.5648370385169983, "learning_rate": 9.717752538560646e-05, "loss": 0.0867, "step": 15210 }, { "grad_norm": 0.6606319546699524, "learning_rate": 9.717204603979986e-05, "loss": 0.0894, "step": 15220 }, { "grad_norm": 0.5582156777381897, "learning_rate": 9.716656153532922e-05, "loss": 0.0767, "step": 15230 }, { "grad_norm": 0.5604267716407776, "learning_rate": 9.716107187279434e-05, "loss": 0.0857, "step": 15240 }, { "grad_norm": 0.6794465184211731, "learning_rate": 9.715557705279555e-05, "loss": 0.1017, "step": 15250 }, { "grad_norm": 0.6594938039779663, "learning_rate": 9.715007707593372e-05, "loss": 0.1055, "step": 15260 }, { "grad_norm": 0.6019705533981323, "learning_rate": 9.714457194281036e-05, "loss": 0.079, "step": 15270 }, { "grad_norm": 0.6402590274810791, "learning_rate": 9.713906165402751e-05, "loss": 0.0859, "step": 15280 }, { "grad_norm": 0.8008766770362854, "learning_rate": 9.713354621018774e-05, "loss": 0.1004, "step": 15290 }, { "grad_norm": 0.41313090920448303, "learning_rate": 9.712802561189422e-05, "loss": 0.071, "step": 15300 }, { "grad_norm": 0.4983275532722473, "learning_rate": 9.712249985975069e-05, "loss": 0.0865, "step": 15310 }, { "grad_norm": 0.46958544850349426, "learning_rate": 9.71169689543614e-05, "loss": 0.0691, "step": 15320 }, { "grad_norm": 0.81980961561203, "learning_rate": 9.711143289633123e-05, "loss": 0.0916, "step": 15330 }, { "grad_norm": 0.533220648765564, "learning_rate": 9.710589168626561e-05, "loss": 0.0851, "step": 15340 }, { "grad_norm": 1.2870486974716187, "learning_rate": 9.710034532477048e-05, "loss": 0.0894, "step": 15350 }, { "grad_norm": 0.44277042150497437, "learning_rate": 9.709479381245239e-05, "loss": 0.0699, "step": 15360 }, { "grad_norm": 0.5884377360343933, "learning_rate": 9.708923714991847e-05, "loss": 0.0888, "step": 15370 }, { "grad_norm": 0.6152501702308655, "learning_rate": 9.708367533777638e-05, "loss": 0.0805, "step": 15380 }, { "grad_norm": 0.6188784837722778, "learning_rate": 9.707810837663431e-05, "loss": 0.0807, "step": 15390 }, { "grad_norm": 0.6857043504714966, "learning_rate": 9.707253626710113e-05, "loss": 0.0928, "step": 15400 }, { "grad_norm": 0.6958150267601013, "learning_rate": 9.706695900978613e-05, "loss": 0.0736, "step": 15410 }, { "grad_norm": 0.6556610465049744, "learning_rate": 9.706137660529926e-05, "loss": 0.081, "step": 15420 }, { "grad_norm": 0.6967801451683044, "learning_rate": 9.705578905425101e-05, "loss": 0.0953, "step": 15430 }, { "grad_norm": 0.6920786499977112, "learning_rate": 9.705019635725241e-05, "loss": 0.0831, "step": 15440 }, { "grad_norm": 0.503474235534668, "learning_rate": 9.704459851491508e-05, "loss": 0.0893, "step": 15450 }, { "grad_norm": 0.6115251779556274, "learning_rate": 9.703899552785118e-05, "loss": 0.0715, "step": 15460 }, { "grad_norm": 0.8802318572998047, "learning_rate": 9.703338739667346e-05, "loss": 0.0718, "step": 15470 }, { "grad_norm": 0.5823684334754944, "learning_rate": 9.70277741219952e-05, "loss": 0.0847, "step": 15480 }, { "grad_norm": 0.7613281607627869, "learning_rate": 9.702215570443027e-05, "loss": 0.0819, "step": 15490 }, { "grad_norm": 0.5690234899520874, "learning_rate": 9.701653214459309e-05, "loss": 0.074, "step": 15500 }, { "grad_norm": 1.2997939586639404, "learning_rate": 9.701090344309865e-05, "loss": 0.0876, "step": 15510 }, { "grad_norm": 0.8795703053474426, "learning_rate": 9.700526960056247e-05, "loss": 0.0731, "step": 15520 }, { "grad_norm": 0.6050471663475037, "learning_rate": 9.699963061760068e-05, "loss": 0.0669, "step": 15530 }, { "grad_norm": 0.5155009627342224, "learning_rate": 9.699398649482997e-05, "loss": 0.0703, "step": 15540 }, { "grad_norm": 0.5621236562728882, "learning_rate": 9.698833723286753e-05, "loss": 0.0627, "step": 15550 }, { "grad_norm": 0.47868433594703674, "learning_rate": 9.698268283233118e-05, "loss": 0.0807, "step": 15560 }, { "grad_norm": 0.7378695011138916, "learning_rate": 9.697702329383929e-05, "loss": 0.0785, "step": 15570 }, { "grad_norm": 0.39869236946105957, "learning_rate": 9.697135861801074e-05, "loss": 0.0807, "step": 15580 }, { "grad_norm": 0.5239235162734985, "learning_rate": 9.696568880546505e-05, "loss": 0.0777, "step": 15590 }, { "grad_norm": 0.5851592421531677, "learning_rate": 9.696001385682223e-05, "loss": 0.0758, "step": 15600 }, { "grad_norm": 0.41289716958999634, "learning_rate": 9.695433377270291e-05, "loss": 0.0801, "step": 15610 }, { "grad_norm": 0.4365018606185913, "learning_rate": 9.694864855372824e-05, "loss": 0.0692, "step": 15620 }, { "grad_norm": 0.713319718837738, "learning_rate": 9.694295820051995e-05, "loss": 0.0737, "step": 15630 }, { "grad_norm": 0.728102445602417, "learning_rate": 9.693726271370032e-05, "loss": 0.0742, "step": 15640 }, { "grad_norm": 0.6354525685310364, "learning_rate": 9.693156209389221e-05, "loss": 0.0809, "step": 15650 }, { "grad_norm": 0.5751070976257324, "learning_rate": 9.692585634171905e-05, "loss": 0.0836, "step": 15660 }, { "grad_norm": 0.46671661734580994, "learning_rate": 9.692014545780476e-05, "loss": 0.0703, "step": 15670 }, { "grad_norm": 0.5405368208885193, "learning_rate": 9.691442944277393e-05, "loss": 0.0701, "step": 15680 }, { "grad_norm": 0.5011789202690125, "learning_rate": 9.690870829725162e-05, "loss": 0.0656, "step": 15690 }, { "grad_norm": 0.62565016746521, "learning_rate": 9.69029820218635e-05, "loss": 0.0815, "step": 15700 }, { "grad_norm": 0.609950840473175, "learning_rate": 9.689725061723579e-05, "loss": 0.0896, "step": 15710 }, { "grad_norm": 0.5160030722618103, "learning_rate": 9.689151408399527e-05, "loss": 0.0821, "step": 15720 }, { "grad_norm": 0.6367046236991882, "learning_rate": 9.688577242276924e-05, "loss": 0.0954, "step": 15730 }, { "grad_norm": 0.5906247496604919, "learning_rate": 9.688002563418566e-05, "loss": 0.0992, "step": 15740 }, { "grad_norm": 0.5819428563117981, "learning_rate": 9.687427371887293e-05, "loss": 0.0789, "step": 15750 }, { "grad_norm": 0.672824501991272, "learning_rate": 9.686851667746012e-05, "loss": 0.0792, "step": 15760 }, { "grad_norm": 0.5743110179901123, "learning_rate": 9.686275451057677e-05, "loss": 0.0764, "step": 15770 }, { "grad_norm": 0.8266882300376892, "learning_rate": 9.685698721885308e-05, "loss": 0.0873, "step": 15780 }, { "grad_norm": 0.5601295232772827, "learning_rate": 9.68512148029197e-05, "loss": 0.0804, "step": 15790 }, { "grad_norm": 0.6543629169464111, "learning_rate": 9.684543726340791e-05, "loss": 0.0835, "step": 15800 }, { "grad_norm": 0.6755616664886475, "learning_rate": 9.683965460094952e-05, "loss": 0.0672, "step": 15810 }, { "grad_norm": 0.6341456770896912, "learning_rate": 9.683386681617694e-05, "loss": 0.0757, "step": 15820 }, { "grad_norm": 0.6120376586914062, "learning_rate": 9.68280739097231e-05, "loss": 0.0748, "step": 15830 }, { "grad_norm": 0.6154187917709351, "learning_rate": 9.682227588222148e-05, "loss": 0.0679, "step": 15840 }, { "grad_norm": 1.529565453529358, "learning_rate": 9.681647273430618e-05, "loss": 0.0648, "step": 15850 }, { "grad_norm": 0.7987861037254333, "learning_rate": 9.681066446661182e-05, "loss": 0.0763, "step": 15860 }, { "grad_norm": 0.7486475110054016, "learning_rate": 9.680485107977357e-05, "loss": 0.0757, "step": 15870 }, { "grad_norm": 0.7514899373054504, "learning_rate": 9.679903257442716e-05, "loss": 0.0839, "step": 15880 }, { "grad_norm": 0.7499273419380188, "learning_rate": 9.679320895120891e-05, "loss": 0.0803, "step": 15890 }, { "grad_norm": 0.38587504625320435, "learning_rate": 9.67873802107557e-05, "loss": 0.0825, "step": 15900 }, { "grad_norm": 0.49702689051628113, "learning_rate": 9.67815463537049e-05, "loss": 0.079, "step": 15910 }, { "grad_norm": 0.6568658947944641, "learning_rate": 9.677570738069457e-05, "loss": 0.0766, "step": 15920 }, { "grad_norm": 0.6565911769866943, "learning_rate": 9.676986329236318e-05, "loss": 0.1035, "step": 15930 }, { "grad_norm": 0.665070116519928, "learning_rate": 9.676401408934987e-05, "loss": 0.0844, "step": 15940 }, { "grad_norm": 0.575432300567627, "learning_rate": 9.675815977229428e-05, "loss": 0.0873, "step": 15950 }, { "grad_norm": 0.7156361937522888, "learning_rate": 9.675230034183664e-05, "loss": 0.0728, "step": 15960 }, { "grad_norm": 0.5840632915496826, "learning_rate": 9.674643579861773e-05, "loss": 0.1012, "step": 15970 }, { "grad_norm": 0.925099790096283, "learning_rate": 9.674056614327886e-05, "loss": 0.0754, "step": 15980 }, { "grad_norm": 0.5301706194877625, "learning_rate": 9.673469137646198e-05, "loss": 0.0877, "step": 15990 }, { "grad_norm": 0.6031053066253662, "learning_rate": 9.67288114988095e-05, "loss": 0.0812, "step": 16000 }, { "grad_norm": 0.672027587890625, "learning_rate": 9.672292651096447e-05, "loss": 0.0774, "step": 16010 }, { "grad_norm": 0.6742087602615356, "learning_rate": 9.671703641357042e-05, "loss": 0.0707, "step": 16020 }, { "grad_norm": 0.5104454159736633, "learning_rate": 9.67111412072715e-05, "loss": 0.0916, "step": 16030 }, { "grad_norm": 0.5594448447227478, "learning_rate": 9.670524089271242e-05, "loss": 0.0803, "step": 16040 }, { "grad_norm": 0.5557490587234497, "learning_rate": 9.669933547053842e-05, "loss": 0.0696, "step": 16050 }, { "grad_norm": 0.5685398578643799, "learning_rate": 9.669342494139531e-05, "loss": 0.0826, "step": 16060 }, { "grad_norm": 0.6030092835426331, "learning_rate": 9.668750930592943e-05, "loss": 0.0724, "step": 16070 }, { "grad_norm": 0.9423360824584961, "learning_rate": 9.668158856478775e-05, "loss": 0.07, "step": 16080 }, { "grad_norm": 0.772117555141449, "learning_rate": 9.66756627186177e-05, "loss": 0.0885, "step": 16090 }, { "grad_norm": 0.6777312159538269, "learning_rate": 9.666973176806737e-05, "loss": 0.08, "step": 16100 }, { "grad_norm": 0.5652921795845032, "learning_rate": 9.666379571378534e-05, "loss": 0.0848, "step": 16110 }, { "grad_norm": 0.6728422045707703, "learning_rate": 9.665785455642076e-05, "loss": 0.0985, "step": 16120 }, { "grad_norm": 0.7408113479614258, "learning_rate": 9.665190829662337e-05, "loss": 0.0969, "step": 16130 }, { "grad_norm": 0.5340995192527771, "learning_rate": 9.664595693504342e-05, "loss": 0.0669, "step": 16140 }, { "grad_norm": 0.6132147312164307, "learning_rate": 9.664000047233175e-05, "loss": 0.1054, "step": 16150 }, { "grad_norm": 0.7286325097084045, "learning_rate": 9.663403890913976e-05, "loss": 0.0765, "step": 16160 }, { "grad_norm": 0.7701099514961243, "learning_rate": 9.662807224611938e-05, "loss": 0.0856, "step": 16170 }, { "grad_norm": 0.9004387855529785, "learning_rate": 9.662210048392311e-05, "loss": 0.0818, "step": 16180 }, { "grad_norm": 0.7134156823158264, "learning_rate": 9.661612362320405e-05, "loss": 0.087, "step": 16190 }, { "grad_norm": 0.6784546375274658, "learning_rate": 9.661014166461579e-05, "loss": 0.0901, "step": 16200 }, { "grad_norm": 0.5349096059799194, "learning_rate": 9.66041546088125e-05, "loss": 0.085, "step": 16210 }, { "grad_norm": 0.4440859854221344, "learning_rate": 9.659816245644895e-05, "loss": 0.0695, "step": 16220 }, { "grad_norm": 0.6888728737831116, "learning_rate": 9.65921652081804e-05, "loss": 0.0728, "step": 16230 }, { "grad_norm": 0.6593837738037109, "learning_rate": 9.658616286466271e-05, "loss": 0.0789, "step": 16240 }, { "grad_norm": 0.46986278891563416, "learning_rate": 9.65801554265523e-05, "loss": 0.0866, "step": 16250 }, { "grad_norm": 0.8587180376052856, "learning_rate": 9.657414289450612e-05, "loss": 0.0796, "step": 16260 }, { "grad_norm": 0.836544930934906, "learning_rate": 9.656812526918171e-05, "loss": 0.0853, "step": 16270 }, { "grad_norm": 0.6602417230606079, "learning_rate": 9.656210255123712e-05, "loss": 0.081, "step": 16280 }, { "grad_norm": 0.6758567690849304, "learning_rate": 9.6556074741331e-05, "loss": 0.0732, "step": 16290 }, { "grad_norm": 0.40078842639923096, "learning_rate": 9.655004184012256e-05, "loss": 0.0671, "step": 16300 }, { "grad_norm": 0.5803987383842468, "learning_rate": 9.654400384827152e-05, "loss": 0.081, "step": 16310 }, { "grad_norm": 0.5979173183441162, "learning_rate": 9.653796076643818e-05, "loss": 0.0672, "step": 16320 }, { "grad_norm": 1.0288535356521606, "learning_rate": 9.653191259528344e-05, "loss": 0.0727, "step": 16330 }, { "grad_norm": 0.6033034324645996, "learning_rate": 9.65258593354687e-05, "loss": 0.0811, "step": 16340 }, { "grad_norm": 0.5335120558738708, "learning_rate": 9.651980098765591e-05, "loss": 0.0726, "step": 16350 }, { "grad_norm": 0.5661507844924927, "learning_rate": 9.651373755250765e-05, "loss": 0.0994, "step": 16360 }, { "grad_norm": 0.713793933391571, "learning_rate": 9.650766903068697e-05, "loss": 0.0854, "step": 16370 }, { "grad_norm": 0.545923113822937, "learning_rate": 9.650159542285753e-05, "loss": 0.0706, "step": 16380 }, { "grad_norm": 0.7278067469596863, "learning_rate": 9.649551672968353e-05, "loss": 0.099, "step": 16390 }, { "grad_norm": 0.6401772499084473, "learning_rate": 9.648943295182973e-05, "loss": 0.087, "step": 16400 }, { "grad_norm": 0.6276611089706421, "learning_rate": 9.648334408996144e-05, "loss": 0.0853, "step": 16410 }, { "grad_norm": 0.4299986958503723, "learning_rate": 9.647725014474452e-05, "loss": 0.0723, "step": 16420 }, { "grad_norm": 0.44166961312294006, "learning_rate": 9.64711511168454e-05, "loss": 0.0762, "step": 16430 }, { "grad_norm": 0.613264799118042, "learning_rate": 9.646504700693108e-05, "loss": 0.0845, "step": 16440 }, { "grad_norm": 0.5555150508880615, "learning_rate": 9.645893781566907e-05, "loss": 0.0769, "step": 16450 }, { "grad_norm": 0.5744794011116028, "learning_rate": 9.645282354372744e-05, "loss": 0.0972, "step": 16460 }, { "grad_norm": 1.1830590963363647, "learning_rate": 9.644670419177491e-05, "loss": 0.0751, "step": 16470 }, { "grad_norm": 0.6759348511695862, "learning_rate": 9.644057976048062e-05, "loss": 0.0807, "step": 16480 }, { "grad_norm": 0.7217559218406677, "learning_rate": 9.643445025051435e-05, "loss": 0.0841, "step": 16490 }, { "grad_norm": 1.3232226371765137, "learning_rate": 9.642831566254641e-05, "loss": 0.0827, "step": 16500 }, { "grad_norm": 0.80866539478302, "learning_rate": 9.642217599724769e-05, "loss": 0.0776, "step": 16510 }, { "grad_norm": 0.5279644727706909, "learning_rate": 9.64160312552896e-05, "loss": 0.0838, "step": 16520 }, { "grad_norm": 0.40475186705589294, "learning_rate": 9.64098814373441e-05, "loss": 0.0839, "step": 16530 }, { "grad_norm": 0.6497613191604614, "learning_rate": 9.640372654408374e-05, "loss": 0.0837, "step": 16540 }, { "grad_norm": 0.5339824557304382, "learning_rate": 9.639756657618162e-05, "loss": 0.0881, "step": 16550 }, { "grad_norm": 0.6662784814834595, "learning_rate": 9.639140153431138e-05, "loss": 0.0798, "step": 16560 }, { "grad_norm": 1.4058856964111328, "learning_rate": 9.638523141914721e-05, "loss": 0.0893, "step": 16570 }, { "grad_norm": 0.6905684471130371, "learning_rate": 9.637905623136388e-05, "loss": 0.0906, "step": 16580 }, { "grad_norm": 0.844718873500824, "learning_rate": 9.637287597163669e-05, "loss": 0.0835, "step": 16590 }, { "grad_norm": 0.5598976612091064, "learning_rate": 9.63666906406415e-05, "loss": 0.0679, "step": 16600 }, { "grad_norm": 0.5718483924865723, "learning_rate": 9.636050023905473e-05, "loss": 0.076, "step": 16610 }, { "grad_norm": 0.5675366520881653, "learning_rate": 9.635430476755336e-05, "loss": 0.0688, "step": 16620 }, { "grad_norm": 0.5351285934448242, "learning_rate": 9.63481042268149e-05, "loss": 0.0755, "step": 16630 }, { "grad_norm": 0.5602129697799683, "learning_rate": 9.634189861751745e-05, "loss": 0.0794, "step": 16640 }, { "grad_norm": 0.3901274800300598, "learning_rate": 9.633568794033967e-05, "loss": 0.0796, "step": 16650 }, { "grad_norm": 0.5271885395050049, "learning_rate": 9.63294721959607e-05, "loss": 0.0772, "step": 16660 }, { "grad_norm": 0.506773829460144, "learning_rate": 9.63232513850603e-05, "loss": 0.1009, "step": 16670 }, { "grad_norm": 0.7503607869148254, "learning_rate": 9.631702550831878e-05, "loss": 0.0903, "step": 16680 }, { "grad_norm": 0.49465957283973694, "learning_rate": 9.631079456641698e-05, "loss": 0.0707, "step": 16690 }, { "grad_norm": 0.6168495416641235, "learning_rate": 9.630455856003632e-05, "loss": 0.0734, "step": 16700 }, { "grad_norm": 0.616600513458252, "learning_rate": 9.629831748985876e-05, "loss": 0.0849, "step": 16710 }, { "grad_norm": 0.7015532851219177, "learning_rate": 9.629207135656679e-05, "loss": 0.0841, "step": 16720 }, { "grad_norm": 0.5745207667350769, "learning_rate": 9.628582016084353e-05, "loss": 0.0756, "step": 16730 }, { "grad_norm": 0.5892488360404968, "learning_rate": 9.627956390337254e-05, "loss": 0.0713, "step": 16740 }, { "grad_norm": 0.5546716451644897, "learning_rate": 9.627330258483802e-05, "loss": 0.0794, "step": 16750 }, { "grad_norm": 1.1549252271652222, "learning_rate": 9.62670362059247e-05, "loss": 0.0658, "step": 16760 }, { "grad_norm": 0.5518957376480103, "learning_rate": 9.626076476731786e-05, "loss": 0.07, "step": 16770 }, { "grad_norm": 0.526567816734314, "learning_rate": 9.625448826970336e-05, "loss": 0.0875, "step": 16780 }, { "grad_norm": 0.9242820739746094, "learning_rate": 9.624820671376755e-05, "loss": 0.0801, "step": 16790 }, { "grad_norm": 0.6019503474235535, "learning_rate": 9.62419201001974e-05, "loss": 0.0753, "step": 16800 }, { "grad_norm": 0.5873798131942749, "learning_rate": 9.623562842968037e-05, "loss": 0.0703, "step": 16810 }, { "grad_norm": 0.6576668620109558, "learning_rate": 9.622933170290454e-05, "loss": 0.0699, "step": 16820 }, { "grad_norm": 0.49259376525878906, "learning_rate": 9.622302992055849e-05, "loss": 0.0811, "step": 16830 }, { "grad_norm": 0.5320648550987244, "learning_rate": 9.62167230833314e-05, "loss": 0.0657, "step": 16840 }, { "grad_norm": 0.663390040397644, "learning_rate": 9.621041119191295e-05, "loss": 0.0745, "step": 16850 }, { "grad_norm": 0.7805367112159729, "learning_rate": 9.620409424699342e-05, "loss": 0.0783, "step": 16860 }, { "grad_norm": 0.5912261009216309, "learning_rate": 9.619777224926359e-05, "loss": 0.0625, "step": 16870 }, { "grad_norm": 0.46970921754837036, "learning_rate": 9.619144519941485e-05, "loss": 0.0766, "step": 16880 }, { "grad_norm": 0.5316357016563416, "learning_rate": 9.618511309813912e-05, "loss": 0.0862, "step": 16890 }, { "grad_norm": 0.7488037943840027, "learning_rate": 9.617877594612886e-05, "loss": 0.0736, "step": 16900 }, { "grad_norm": 0.500272810459137, "learning_rate": 9.617243374407707e-05, "loss": 0.069, "step": 16910 }, { "grad_norm": 0.6179171204566956, "learning_rate": 9.616608649267736e-05, "loss": 0.0798, "step": 16920 }, { "grad_norm": 0.6084524989128113, "learning_rate": 9.615973419262385e-05, "loss": 0.0686, "step": 16930 }, { "grad_norm": 1.373538613319397, "learning_rate": 9.615337684461119e-05, "loss": 0.0695, "step": 16940 }, { "grad_norm": 0.5825751423835754, "learning_rate": 9.614701444933465e-05, "loss": 0.0829, "step": 16950 }, { "grad_norm": 0.5534944534301758, "learning_rate": 9.614064700748997e-05, "loss": 0.0819, "step": 16960 }, { "grad_norm": 0.7599756717681885, "learning_rate": 9.613427451977352e-05, "loss": 0.0672, "step": 16970 }, { "grad_norm": 0.8191048502922058, "learning_rate": 9.612789698688216e-05, "loss": 0.0832, "step": 16980 }, { "grad_norm": 0.5740129351615906, "learning_rate": 9.612151440951334e-05, "loss": 0.0792, "step": 16990 }, { "grad_norm": 0.7543590664863586, "learning_rate": 9.611512678836506e-05, "loss": 0.0769, "step": 17000 }, { "grad_norm": 0.5294600129127502, "learning_rate": 9.610873412413584e-05, "loss": 0.0835, "step": 17010 }, { "grad_norm": 1.1332221031188965, "learning_rate": 9.610233641752476e-05, "loss": 0.0973, "step": 17020 }, { "grad_norm": 0.7135267853736877, "learning_rate": 9.609593366923151e-05, "loss": 0.0739, "step": 17030 }, { "grad_norm": 0.6658530235290527, "learning_rate": 9.608952587995625e-05, "loss": 0.078, "step": 17040 }, { "grad_norm": 0.5668891668319702, "learning_rate": 9.608311305039972e-05, "loss": 0.0852, "step": 17050 }, { "grad_norm": 0.8059354424476624, "learning_rate": 9.607669518126326e-05, "loss": 0.0889, "step": 17060 }, { "grad_norm": 0.539526104927063, "learning_rate": 9.607027227324866e-05, "loss": 0.073, "step": 17070 }, { "grad_norm": 0.5834903717041016, "learning_rate": 9.606384432705837e-05, "loss": 0.0851, "step": 17080 }, { "grad_norm": 0.6552129983901978, "learning_rate": 9.60574113433953e-05, "loss": 0.0863, "step": 17090 }, { "grad_norm": 0.5327224731445312, "learning_rate": 9.6050973322963e-05, "loss": 0.0804, "step": 17100 }, { "grad_norm": 0.5889337062835693, "learning_rate": 9.604453026646547e-05, "loss": 0.0947, "step": 17110 }, { "grad_norm": 0.5215093493461609, "learning_rate": 9.603808217460735e-05, "loss": 0.08, "step": 17120 }, { "grad_norm": 0.6934828162193298, "learning_rate": 9.603162904809377e-05, "loss": 0.0801, "step": 17130 }, { "grad_norm": 0.598479688167572, "learning_rate": 9.602517088763045e-05, "loss": 0.0648, "step": 17140 }, { "grad_norm": 0.43802815675735474, "learning_rate": 9.601870769392365e-05, "loss": 0.0943, "step": 17150 }, { "grad_norm": 0.48330172896385193, "learning_rate": 9.601223946768017e-05, "loss": 0.0661, "step": 17160 }, { "grad_norm": 0.3958703577518463, "learning_rate": 9.600576620960734e-05, "loss": 0.0696, "step": 17170 }, { "grad_norm": 1.030131459236145, "learning_rate": 9.599928792041308e-05, "loss": 0.0744, "step": 17180 }, { "grad_norm": 0.8363309502601624, "learning_rate": 9.599280460080587e-05, "loss": 0.0856, "step": 17190 }, { "grad_norm": 0.429643452167511, "learning_rate": 9.59863162514947e-05, "loss": 0.0861, "step": 17200 }, { "grad_norm": 0.5978304147720337, "learning_rate": 9.597982287318911e-05, "loss": 0.0715, "step": 17210 }, { "grad_norm": 0.4941202998161316, "learning_rate": 9.597332446659923e-05, "loss": 0.0791, "step": 17220 }, { "grad_norm": 0.5508356690406799, "learning_rate": 9.59668210324357e-05, "loss": 0.0752, "step": 17230 }, { "grad_norm": 0.46487775444984436, "learning_rate": 9.596031257140974e-05, "loss": 0.069, "step": 17240 }, { "grad_norm": 1.35657799243927, "learning_rate": 9.59537990842331e-05, "loss": 0.0969, "step": 17250 }, { "grad_norm": 0.4542594254016876, "learning_rate": 9.594728057161806e-05, "loss": 0.0658, "step": 17260 }, { "grad_norm": 0.6187313795089722, "learning_rate": 9.594075703427752e-05, "loss": 0.0803, "step": 17270 }, { "grad_norm": 0.5091332793235779, "learning_rate": 9.593422847292486e-05, "loss": 0.065, "step": 17280 }, { "grad_norm": 0.630781352519989, "learning_rate": 9.592769488827402e-05, "loss": 0.0817, "step": 17290 }, { "grad_norm": 0.5884510278701782, "learning_rate": 9.592115628103952e-05, "loss": 0.074, "step": 17300 }, { "grad_norm": 0.38708674907684326, "learning_rate": 9.591461265193643e-05, "loss": 0.0649, "step": 17310 }, { "grad_norm": 0.9849005937576294, "learning_rate": 9.590806400168032e-05, "loss": 0.0672, "step": 17320 }, { "grad_norm": 0.4543546140193939, "learning_rate": 9.590151033098735e-05, "loss": 0.0635, "step": 17330 }, { "grad_norm": 0.5382592678070068, "learning_rate": 9.589495164057423e-05, "loss": 0.0717, "step": 17340 }, { "grad_norm": 0.4346521496772766, "learning_rate": 9.58883879311582e-05, "loss": 0.079, "step": 17350 }, { "grad_norm": 0.5597408413887024, "learning_rate": 9.588181920345705e-05, "loss": 0.0751, "step": 17360 }, { "grad_norm": 0.4123472571372986, "learning_rate": 9.587524545818913e-05, "loss": 0.0792, "step": 17370 }, { "grad_norm": 0.9935568571090698, "learning_rate": 9.586866669607335e-05, "loss": 0.078, "step": 17380 }, { "grad_norm": 1.083804965019226, "learning_rate": 9.586208291782915e-05, "loss": 0.0787, "step": 17390 }, { "grad_norm": 0.6952612400054932, "learning_rate": 9.58554941241765e-05, "loss": 0.0912, "step": 17400 }, { "grad_norm": 0.586972177028656, "learning_rate": 9.584890031583596e-05, "loss": 0.08, "step": 17410 }, { "grad_norm": 0.5718162059783936, "learning_rate": 9.584230149352861e-05, "loss": 0.0739, "step": 17420 }, { "grad_norm": 0.7744045853614807, "learning_rate": 9.58356976579761e-05, "loss": 0.0777, "step": 17430 }, { "grad_norm": 0.5892927050590515, "learning_rate": 9.58290888099006e-05, "loss": 0.0825, "step": 17440 }, { "grad_norm": 0.4446142017841339, "learning_rate": 9.582247495002486e-05, "loss": 0.084, "step": 17450 }, { "grad_norm": 0.8184815049171448, "learning_rate": 9.581585607907214e-05, "loss": 0.0756, "step": 17460 }, { "grad_norm": 0.5819581747055054, "learning_rate": 9.580923219776628e-05, "loss": 0.0717, "step": 17470 }, { "grad_norm": 0.5446973443031311, "learning_rate": 9.580260330683167e-05, "loss": 0.0698, "step": 17480 }, { "grad_norm": 0.4782049357891083, "learning_rate": 9.579596940699322e-05, "loss": 0.0716, "step": 17490 }, { "grad_norm": 0.5758213996887207, "learning_rate": 9.578933049897643e-05, "loss": 0.0701, "step": 17500 }, { "grad_norm": 0.7211636900901794, "learning_rate": 9.578268658350728e-05, "loss": 0.0721, "step": 17510 }, { "grad_norm": 0.635639488697052, "learning_rate": 9.577603766131235e-05, "loss": 0.0775, "step": 17520 }, { "grad_norm": 1.2600281238555908, "learning_rate": 9.576938373311878e-05, "loss": 0.0807, "step": 17530 }, { "grad_norm": 1.028373122215271, "learning_rate": 9.576272479965421e-05, "loss": 0.0759, "step": 17540 }, { "grad_norm": 0.7128778696060181, "learning_rate": 9.575606086164687e-05, "loss": 0.09, "step": 17550 }, { "grad_norm": 0.5866314172744751, "learning_rate": 9.57493919198255e-05, "loss": 0.0958, "step": 17560 }, { "grad_norm": 0.5995969772338867, "learning_rate": 9.57427179749194e-05, "loss": 0.0939, "step": 17570 }, { "grad_norm": 0.5367289781570435, "learning_rate": 9.573603902765846e-05, "loss": 0.0677, "step": 17580 }, { "grad_norm": 0.4474431872367859, "learning_rate": 9.572935507877304e-05, "loss": 0.0804, "step": 17590 }, { "grad_norm": 0.6570374369621277, "learning_rate": 9.57226661289941e-05, "loss": 0.0904, "step": 17600 }, { "grad_norm": 0.6290913224220276, "learning_rate": 9.571597217905315e-05, "loss": 0.0855, "step": 17610 }, { "grad_norm": 0.8188625574111938, "learning_rate": 9.57092732296822e-05, "loss": 0.0753, "step": 17620 }, { "grad_norm": 0.46327680349349976, "learning_rate": 9.570256928161385e-05, "loss": 0.067, "step": 17630 }, { "grad_norm": 0.42995673418045044, "learning_rate": 9.569586033558126e-05, "loss": 0.0831, "step": 17640 }, { "grad_norm": 0.5399401783943176, "learning_rate": 9.568914639231807e-05, "loss": 0.0742, "step": 17650 }, { "grad_norm": 0.751863956451416, "learning_rate": 9.568242745255852e-05, "loss": 0.0744, "step": 17660 }, { "grad_norm": 0.5107289552688599, "learning_rate": 9.567570351703739e-05, "loss": 0.0829, "step": 17670 }, { "grad_norm": 0.6563920378684998, "learning_rate": 9.566897458649001e-05, "loss": 0.0786, "step": 17680 }, { "grad_norm": 0.58626788854599, "learning_rate": 9.566224066165221e-05, "loss": 0.0783, "step": 17690 }, { "grad_norm": 0.5834264159202576, "learning_rate": 9.565550174326043e-05, "loss": 0.0857, "step": 17700 }, { "grad_norm": 0.6038605570793152, "learning_rate": 9.564875783205162e-05, "loss": 0.0738, "step": 17710 }, { "grad_norm": 0.7554025650024414, "learning_rate": 9.564200892876328e-05, "loss": 0.0781, "step": 17720 }, { "grad_norm": 0.623759925365448, "learning_rate": 9.563525503413348e-05, "loss": 0.0739, "step": 17730 }, { "grad_norm": 0.5201154947280884, "learning_rate": 9.562849614890079e-05, "loss": 0.0792, "step": 17740 }, { "grad_norm": 0.36476224660873413, "learning_rate": 9.562173227380436e-05, "loss": 0.0879, "step": 17750 }, { "grad_norm": 0.6029587984085083, "learning_rate": 9.561496340958389e-05, "loss": 0.0743, "step": 17760 }, { "grad_norm": 0.6324467062950134, "learning_rate": 9.560818955697959e-05, "loss": 0.0772, "step": 17770 }, { "grad_norm": 0.6056416630744934, "learning_rate": 9.560141071673228e-05, "loss": 0.0748, "step": 17780 }, { "grad_norm": 0.7047393918037415, "learning_rate": 9.559462688958323e-05, "loss": 0.0549, "step": 17790 }, { "grad_norm": 0.474890798330307, "learning_rate": 9.558783807627434e-05, "loss": 0.0596, "step": 17800 }, { "grad_norm": 0.5954223275184631, "learning_rate": 9.558104427754801e-05, "loss": 0.0973, "step": 17810 }, { "grad_norm": 0.6211732029914856, "learning_rate": 9.557424549414722e-05, "loss": 0.0843, "step": 17820 }, { "grad_norm": 0.4860486090183258, "learning_rate": 9.556744172681546e-05, "loss": 0.076, "step": 17830 }, { "grad_norm": 1.0149219036102295, "learning_rate": 9.556063297629677e-05, "loss": 0.0771, "step": 17840 }, { "grad_norm": 0.5844089984893799, "learning_rate": 9.555381924333578e-05, "loss": 0.0662, "step": 17850 }, { "grad_norm": 0.5287757515907288, "learning_rate": 9.554700052867758e-05, "loss": 0.0721, "step": 17860 }, { "grad_norm": 0.6228827834129333, "learning_rate": 9.554017683306789e-05, "loss": 0.0755, "step": 17870 }, { "grad_norm": 0.4923996031284332, "learning_rate": 9.553334815725294e-05, "loss": 0.0732, "step": 17880 }, { "grad_norm": 0.4956567883491516, "learning_rate": 9.552651450197949e-05, "loss": 0.076, "step": 17890 }, { "grad_norm": 0.5773857235908508, "learning_rate": 9.551967586799486e-05, "loss": 0.0712, "step": 17900 }, { "grad_norm": 0.6286770105361938, "learning_rate": 9.551283225604692e-05, "loss": 0.0855, "step": 17910 }, { "grad_norm": 0.6082272529602051, "learning_rate": 9.550598366688406e-05, "loss": 0.0901, "step": 17920 }, { "grad_norm": 0.6068474054336548, "learning_rate": 9.549913010125526e-05, "loss": 0.079, "step": 17930 }, { "grad_norm": 0.40832266211509705, "learning_rate": 9.549227155990999e-05, "loss": 0.0797, "step": 17940 }, { "grad_norm": 0.6123285889625549, "learning_rate": 9.548540804359828e-05, "loss": 0.079, "step": 17950 }, { "grad_norm": 0.6539498567581177, "learning_rate": 9.547853955307077e-05, "loss": 0.0846, "step": 17960 }, { "grad_norm": 0.498290479183197, "learning_rate": 9.547166608907853e-05, "loss": 0.0711, "step": 17970 }, { "grad_norm": 0.5343847870826721, "learning_rate": 9.546478765237326e-05, "loss": 0.0934, "step": 17980 }, { "grad_norm": 0.6256905794143677, "learning_rate": 9.545790424370715e-05, "loss": 0.0753, "step": 17990 }, { "grad_norm": 0.3872273564338684, "learning_rate": 9.5451015863833e-05, "loss": 0.0841, "step": 18000 }, { "grad_norm": 0.642597496509552, "learning_rate": 9.544412251350408e-05, "loss": 0.0776, "step": 18010 }, { "grad_norm": 0.6333182454109192, "learning_rate": 9.543722419347422e-05, "loss": 0.0717, "step": 18020 }, { "grad_norm": 0.5806918144226074, "learning_rate": 9.543032090449788e-05, "loss": 0.0869, "step": 18030 }, { "grad_norm": 0.6192120909690857, "learning_rate": 9.542341264732992e-05, "loss": 0.0816, "step": 18040 }, { "grad_norm": 0.5251660943031311, "learning_rate": 9.541649942272585e-05, "loss": 0.0815, "step": 18050 }, { "grad_norm": 0.6176384687423706, "learning_rate": 9.54095812314417e-05, "loss": 0.0669, "step": 18060 }, { "grad_norm": 0.6223798394203186, "learning_rate": 9.540265807423401e-05, "loss": 0.0728, "step": 18070 }, { "grad_norm": 0.4997309744358063, "learning_rate": 9.53957299518599e-05, "loss": 0.0692, "step": 18080 }, { "grad_norm": 0.5136397480964661, "learning_rate": 9.5388796865077e-05, "loss": 0.083, "step": 18090 }, { "grad_norm": 0.627738893032074, "learning_rate": 9.538185881464353e-05, "loss": 0.0822, "step": 18100 }, { "grad_norm": 0.5355615615844727, "learning_rate": 9.537491580131821e-05, "loss": 0.0663, "step": 18110 }, { "grad_norm": 0.44114649295806885, "learning_rate": 9.53679678258603e-05, "loss": 0.0704, "step": 18120 }, { "grad_norm": 0.9678211212158203, "learning_rate": 9.536101488902966e-05, "loss": 0.0713, "step": 18130 }, { "grad_norm": 0.552668571472168, "learning_rate": 9.535405699158663e-05, "loss": 0.0741, "step": 18140 }, { "grad_norm": 0.9326058030128479, "learning_rate": 9.53470941342921e-05, "loss": 0.0664, "step": 18150 }, { "grad_norm": 0.7126494646072388, "learning_rate": 9.534012631790756e-05, "loss": 0.0718, "step": 18160 }, { "grad_norm": 1.1306008100509644, "learning_rate": 9.533315354319494e-05, "loss": 0.0923, "step": 18170 }, { "grad_norm": 0.5603679418563843, "learning_rate": 9.532617581091682e-05, "loss": 0.0818, "step": 18180 }, { "grad_norm": 0.737546980381012, "learning_rate": 9.531919312183629e-05, "loss": 0.0788, "step": 18190 }, { "grad_norm": 0.8051674365997314, "learning_rate": 9.531220547671688e-05, "loss": 0.0692, "step": 18200 }, { "grad_norm": 0.522976815700531, "learning_rate": 9.530521287632285e-05, "loss": 0.0796, "step": 18210 }, { "grad_norm": 0.8106547594070435, "learning_rate": 9.529821532141884e-05, "loss": 0.0762, "step": 18220 }, { "grad_norm": 0.44546273350715637, "learning_rate": 9.52912128127701e-05, "loss": 0.0659, "step": 18230 }, { "grad_norm": 0.5577492117881775, "learning_rate": 9.528420535114244e-05, "loss": 0.0771, "step": 18240 }, { "grad_norm": 0.723899781703949, "learning_rate": 9.527719293730215e-05, "loss": 0.0891, "step": 18250 }, { "grad_norm": 0.4808765649795532, "learning_rate": 9.527017557201611e-05, "loss": 0.0827, "step": 18260 }, { "grad_norm": 0.6366971731185913, "learning_rate": 9.526315325605176e-05, "loss": 0.0945, "step": 18270 }, { "grad_norm": 0.4808683693408966, "learning_rate": 9.525612599017699e-05, "loss": 0.0662, "step": 18280 }, { "grad_norm": 0.6552632451057434, "learning_rate": 9.524909377516033e-05, "loss": 0.0815, "step": 18290 }, { "grad_norm": 0.6019235849380493, "learning_rate": 9.524205661177081e-05, "loss": 0.0711, "step": 18300 }, { "grad_norm": 0.4884364902973175, "learning_rate": 9.523501450077801e-05, "loss": 0.0771, "step": 18310 }, { "grad_norm": 0.5973943471908569, "learning_rate": 9.522796744295202e-05, "loss": 0.0825, "step": 18320 }, { "grad_norm": 0.7029218673706055, "learning_rate": 9.522091543906352e-05, "loss": 0.0798, "step": 18330 }, { "grad_norm": 0.8556959629058838, "learning_rate": 9.521385848988369e-05, "loss": 0.0808, "step": 18340 }, { "grad_norm": 0.4892061948776245, "learning_rate": 9.520679659618428e-05, "loss": 0.0648, "step": 18350 }, { "grad_norm": 0.6667073965072632, "learning_rate": 9.519972975873754e-05, "loss": 0.076, "step": 18360 }, { "grad_norm": 0.47897958755493164, "learning_rate": 9.519265797831633e-05, "loss": 0.0664, "step": 18370 }, { "grad_norm": 0.6823270916938782, "learning_rate": 9.518558125569399e-05, "loss": 0.0724, "step": 18380 }, { "grad_norm": 0.606510579586029, "learning_rate": 9.517849959164442e-05, "loss": 0.0764, "step": 18390 }, { "grad_norm": 1.0507557392120361, "learning_rate": 9.517141298694205e-05, "loss": 0.0924, "step": 18400 }, { "grad_norm": 0.5947656631469727, "learning_rate": 9.516432144236188e-05, "loss": 0.0773, "step": 18410 }, { "grad_norm": 0.5952451825141907, "learning_rate": 9.515722495867941e-05, "loss": 0.0823, "step": 18420 }, { "grad_norm": 0.7478041052818298, "learning_rate": 9.515012353667072e-05, "loss": 0.0781, "step": 18430 }, { "grad_norm": 0.6044105887413025, "learning_rate": 9.51430171771124e-05, "loss": 0.0619, "step": 18440 }, { "grad_norm": 0.668965756893158, "learning_rate": 9.513590588078159e-05, "loss": 0.0715, "step": 18450 }, { "grad_norm": 0.6745938062667847, "learning_rate": 9.512878964845597e-05, "loss": 0.0771, "step": 18460 }, { "grad_norm": 1.939295768737793, "learning_rate": 9.512166848091377e-05, "loss": 0.0922, "step": 18470 }, { "grad_norm": 0.5764667987823486, "learning_rate": 9.511454237893376e-05, "loss": 0.093, "step": 18480 }, { "grad_norm": 0.7394214868545532, "learning_rate": 9.51074113432952e-05, "loss": 0.0689, "step": 18490 }, { "grad_norm": 0.5980542898178101, "learning_rate": 9.510027537477797e-05, "loss": 0.0967, "step": 18500 }, { "grad_norm": 0.6557061076164246, "learning_rate": 9.509313447416242e-05, "loss": 0.0794, "step": 18510 }, { "grad_norm": 0.5150856971740723, "learning_rate": 9.508598864222949e-05, "loss": 0.0775, "step": 18520 }, { "grad_norm": 0.6888570785522461, "learning_rate": 9.507883787976062e-05, "loss": 0.0699, "step": 18530 }, { "grad_norm": 0.4500201642513275, "learning_rate": 9.507168218753781e-05, "loss": 0.069, "step": 18540 }, { "grad_norm": 0.6249091625213623, "learning_rate": 9.506452156634362e-05, "loss": 0.0834, "step": 18550 }, { "grad_norm": 0.513998806476593, "learning_rate": 9.505735601696109e-05, "loss": 0.088, "step": 18560 }, { "grad_norm": 0.6408042311668396, "learning_rate": 9.505018554017385e-05, "loss": 0.0691, "step": 18570 }, { "grad_norm": 0.8076256513595581, "learning_rate": 9.504301013676604e-05, "loss": 0.0694, "step": 18580 }, { "grad_norm": 0.5584895014762878, "learning_rate": 9.503582980752238e-05, "loss": 0.0731, "step": 18590 }, { "grad_norm": 0.8377382159233093, "learning_rate": 9.502864455322809e-05, "loss": 0.0863, "step": 18600 }, { "grad_norm": 0.5545018911361694, "learning_rate": 9.502145437466891e-05, "loss": 0.0818, "step": 18610 }, { "grad_norm": 0.4840364158153534, "learning_rate": 9.501425927263116e-05, "loss": 0.0888, "step": 18620 }, { "grad_norm": 0.7052289843559265, "learning_rate": 9.500705924790172e-05, "loss": 0.1045, "step": 18630 }, { "grad_norm": 0.9905157685279846, "learning_rate": 9.499985430126794e-05, "loss": 0.083, "step": 18640 }, { "grad_norm": 0.5827069282531738, "learning_rate": 9.499264443351775e-05, "loss": 0.0771, "step": 18650 }, { "grad_norm": 0.7637944221496582, "learning_rate": 9.498542964543961e-05, "loss": 0.0758, "step": 18660 }, { "grad_norm": 0.5172498822212219, "learning_rate": 9.497820993782252e-05, "loss": 0.0827, "step": 18670 }, { "grad_norm": 0.6335750222206116, "learning_rate": 9.497098531145601e-05, "loss": 0.0746, "step": 18680 }, { "grad_norm": 0.9582322835922241, "learning_rate": 9.496375576713017e-05, "loss": 0.0835, "step": 18690 }, { "grad_norm": 0.6134570240974426, "learning_rate": 9.49565213056356e-05, "loss": 0.0756, "step": 18700 }, { "grad_norm": 0.4532028138637543, "learning_rate": 9.494928192776342e-05, "loss": 0.0743, "step": 18710 }, { "grad_norm": 0.9581607580184937, "learning_rate": 9.494203763430538e-05, "loss": 0.0805, "step": 18720 }, { "grad_norm": 0.4794679880142212, "learning_rate": 9.493478842605366e-05, "loss": 0.1024, "step": 18730 }, { "grad_norm": 0.7964400053024292, "learning_rate": 9.492753430380105e-05, "loss": 0.0772, "step": 18740 }, { "grad_norm": 0.7833442687988281, "learning_rate": 9.492027526834083e-05, "loss": 0.0663, "step": 18750 }, { "grad_norm": 0.8817667365074158, "learning_rate": 9.491301132046684e-05, "loss": 0.0826, "step": 18760 }, { "grad_norm": 0.9258068799972534, "learning_rate": 9.490574246097345e-05, "loss": 0.0968, "step": 18770 }, { "grad_norm": 0.6755976676940918, "learning_rate": 9.48984686906556e-05, "loss": 0.0823, "step": 18780 }, { "grad_norm": 0.630975604057312, "learning_rate": 9.489119001030871e-05, "loss": 0.0698, "step": 18790 }, { "grad_norm": 0.5130486488342285, "learning_rate": 9.488390642072878e-05, "loss": 0.0876, "step": 18800 }, { "grad_norm": 0.49444714188575745, "learning_rate": 9.48766179227123e-05, "loss": 0.0847, "step": 18810 }, { "grad_norm": 1.151257038116455, "learning_rate": 9.486932451705636e-05, "loss": 0.1, "step": 18820 }, { "grad_norm": 0.4779968857765198, "learning_rate": 9.486202620455857e-05, "loss": 0.0779, "step": 18830 }, { "grad_norm": 0.5047234296798706, "learning_rate": 9.485472298601704e-05, "loss": 0.0744, "step": 18840 }, { "grad_norm": 0.4566652178764343, "learning_rate": 9.484741486223043e-05, "loss": 0.085, "step": 18850 }, { "grad_norm": 0.5799393653869629, "learning_rate": 9.484010183399797e-05, "loss": 0.0706, "step": 18860 }, { "grad_norm": 0.5448815226554871, "learning_rate": 9.483278390211938e-05, "loss": 0.074, "step": 18870 }, { "grad_norm": 0.5255581736564636, "learning_rate": 9.482546106739496e-05, "loss": 0.0603, "step": 18880 }, { "grad_norm": 0.5526302456855774, "learning_rate": 9.48181333306255e-05, "loss": 0.0685, "step": 18890 }, { "grad_norm": 0.5176563262939453, "learning_rate": 9.481080069261237e-05, "loss": 0.0816, "step": 18900 }, { "grad_norm": 0.5866497755050659, "learning_rate": 9.480346315415745e-05, "loss": 0.0804, "step": 18910 }, { "grad_norm": 0.9260342717170715, "learning_rate": 9.479612071606314e-05, "loss": 0.0631, "step": 18920 }, { "grad_norm": 0.8348374366760254, "learning_rate": 9.478877337913244e-05, "loss": 0.0684, "step": 18930 }, { "grad_norm": 0.5882585048675537, "learning_rate": 9.478142114416881e-05, "loss": 0.0673, "step": 18940 }, { "grad_norm": 0.6994442939758301, "learning_rate": 9.47740640119763e-05, "loss": 0.0717, "step": 18950 }, { "grad_norm": 0.5690302848815918, "learning_rate": 9.476670198335947e-05, "loss": 0.0765, "step": 18960 }, { "grad_norm": 0.6554953455924988, "learning_rate": 9.47593350591234e-05, "loss": 0.077, "step": 18970 }, { "grad_norm": 0.44983890652656555, "learning_rate": 9.475196324007376e-05, "loss": 0.0803, "step": 18980 }, { "grad_norm": 0.5375405550003052, "learning_rate": 9.474458652701669e-05, "loss": 0.0697, "step": 18990 }, { "grad_norm": 0.6007333397865295, "learning_rate": 9.473720492075892e-05, "loss": 0.0826, "step": 19000 }, { "grad_norm": 0.9073498845100403, "learning_rate": 9.472981842210768e-05, "loss": 0.0751, "step": 19010 }, { "grad_norm": 0.9211255311965942, "learning_rate": 9.472242703187074e-05, "loss": 0.0701, "step": 19020 }, { "grad_norm": 0.5710247755050659, "learning_rate": 9.471503075085643e-05, "loss": 0.0675, "step": 19030 }, { "grad_norm": 1.0458873510360718, "learning_rate": 9.470762957987359e-05, "loss": 0.0886, "step": 19040 }, { "grad_norm": 0.5265045166015625, "learning_rate": 9.470022351973158e-05, "loss": 0.0872, "step": 19050 }, { "grad_norm": 0.7128584384918213, "learning_rate": 9.469281257124034e-05, "loss": 0.0814, "step": 19060 }, { "grad_norm": 0.7840448617935181, "learning_rate": 9.46853967352103e-05, "loss": 0.0749, "step": 19070 }, { "grad_norm": 0.7321930527687073, "learning_rate": 9.467797601245246e-05, "loss": 0.0717, "step": 19080 }, { "grad_norm": 0.7129189968109131, "learning_rate": 9.467055040377834e-05, "loss": 0.0809, "step": 19090 }, { "grad_norm": 0.5246298909187317, "learning_rate": 9.466311990999999e-05, "loss": 0.0848, "step": 19100 }, { "grad_norm": 0.8664183020591736, "learning_rate": 9.465568453193e-05, "loss": 0.0801, "step": 19110 }, { "grad_norm": 0.5265381932258606, "learning_rate": 9.464824427038148e-05, "loss": 0.0675, "step": 19120 }, { "grad_norm": 0.5574256181716919, "learning_rate": 9.46407991261681e-05, "loss": 0.0888, "step": 19130 }, { "grad_norm": 0.6403401494026184, "learning_rate": 9.463334910010404e-05, "loss": 0.06, "step": 19140 }, { "grad_norm": 0.4418877959251404, "learning_rate": 9.462589419300403e-05, "loss": 0.0607, "step": 19150 }, { "grad_norm": 0.893259584903717, "learning_rate": 9.461843440568333e-05, "loss": 0.0863, "step": 19160 }, { "grad_norm": 0.46030357480049133, "learning_rate": 9.461096973895773e-05, "loss": 0.0791, "step": 19170 }, { "grad_norm": 0.4621909558773041, "learning_rate": 9.460350019364355e-05, "loss": 0.0685, "step": 19180 }, { "grad_norm": 0.6803387403488159, "learning_rate": 9.459602577055764e-05, "loss": 0.0989, "step": 19190 }, { "grad_norm": 0.5090517997741699, "learning_rate": 9.45885464705174e-05, "loss": 0.072, "step": 19200 }, { "grad_norm": 0.7509467601776123, "learning_rate": 9.458106229434076e-05, "loss": 0.0861, "step": 19210 }, { "grad_norm": 0.7561288475990295, "learning_rate": 9.457357324284617e-05, "loss": 0.0809, "step": 19220 }, { "grad_norm": 0.788061797618866, "learning_rate": 9.456607931685262e-05, "loss": 0.0759, "step": 19230 }, { "grad_norm": 0.5670823454856873, "learning_rate": 9.455858051717965e-05, "loss": 0.0746, "step": 19240 }, { "grad_norm": 0.5277376174926758, "learning_rate": 9.45510768446473e-05, "loss": 0.0651, "step": 19250 }, { "grad_norm": 1.0165036916732788, "learning_rate": 9.454356830007618e-05, "loss": 0.0623, "step": 19260 }, { "grad_norm": 1.0234378576278687, "learning_rate": 9.45360548842874e-05, "loss": 0.0632, "step": 19270 }, { "grad_norm": 0.47642025351524353, "learning_rate": 9.452853659810261e-05, "loss": 0.0651, "step": 19280 }, { "grad_norm": 0.630540668964386, "learning_rate": 9.452101344234401e-05, "loss": 0.0883, "step": 19290 }, { "grad_norm": 0.39319512248039246, "learning_rate": 9.451348541783431e-05, "loss": 0.0709, "step": 19300 }, { "grad_norm": 1.3396128416061401, "learning_rate": 9.450595252539678e-05, "loss": 0.0809, "step": 19310 }, { "grad_norm": 0.8677427768707275, "learning_rate": 9.449841476585518e-05, "loss": 0.0873, "step": 19320 }, { "grad_norm": 0.586819052696228, "learning_rate": 9.449087214003384e-05, "loss": 0.0869, "step": 19330 }, { "grad_norm": 0.8873380422592163, "learning_rate": 9.448332464875765e-05, "loss": 0.0925, "step": 19340 }, { "grad_norm": 0.4885033965110779, "learning_rate": 9.447577229285192e-05, "loss": 0.0956, "step": 19350 }, { "grad_norm": 0.5538535118103027, "learning_rate": 9.446821507314261e-05, "loss": 0.0688, "step": 19360 }, { "grad_norm": 1.000794768333435, "learning_rate": 9.446065299045617e-05, "loss": 0.0786, "step": 19370 }, { "grad_norm": 0.5307129621505737, "learning_rate": 9.445308604561955e-05, "loss": 0.0802, "step": 19380 }, { "grad_norm": 0.9378606677055359, "learning_rate": 9.444551423946028e-05, "loss": 0.0761, "step": 19390 }, { "grad_norm": 0.7933962345123291, "learning_rate": 9.443793757280638e-05, "loss": 0.0819, "step": 19400 }, { "grad_norm": 0.9976608157157898, "learning_rate": 9.443035604648646e-05, "loss": 0.0763, "step": 19410 }, { "grad_norm": 0.6403400897979736, "learning_rate": 9.44227696613296e-05, "loss": 0.071, "step": 19420 }, { "grad_norm": 0.5678724646568298, "learning_rate": 9.441517841816542e-05, "loss": 0.0906, "step": 19430 }, { "grad_norm": 0.7104395627975464, "learning_rate": 9.440758231782413e-05, "loss": 0.0675, "step": 19440 }, { "grad_norm": 0.5445203185081482, "learning_rate": 9.439998136113639e-05, "loss": 0.0824, "step": 19450 }, { "grad_norm": 0.493480384349823, "learning_rate": 9.439237554893344e-05, "loss": 0.0744, "step": 19460 }, { "grad_norm": 0.5811070203781128, "learning_rate": 9.438476488204705e-05, "loss": 0.0853, "step": 19470 }, { "grad_norm": 0.5097109079360962, "learning_rate": 9.43771493613095e-05, "loss": 0.0752, "step": 19480 }, { "grad_norm": 0.5314359068870544, "learning_rate": 9.436952898755362e-05, "loss": 0.08, "step": 19490 }, { "grad_norm": 0.7489134669303894, "learning_rate": 9.436190376161276e-05, "loss": 0.0631, "step": 19500 }, { "grad_norm": 1.1208659410476685, "learning_rate": 9.43542736843208e-05, "loss": 0.1018, "step": 19510 }, { "grad_norm": 0.7497509121894836, "learning_rate": 9.434663875651216e-05, "loss": 0.0639, "step": 19520 }, { "grad_norm": 0.7274243235588074, "learning_rate": 9.433899897902177e-05, "loss": 0.0783, "step": 19530 }, { "grad_norm": 0.6130602955818176, "learning_rate": 9.433135435268511e-05, "loss": 0.0626, "step": 19540 }, { "grad_norm": 0.3887886106967926, "learning_rate": 9.432370487833819e-05, "loss": 0.0754, "step": 19550 }, { "grad_norm": 0.36641183495521545, "learning_rate": 9.431605055681756e-05, "loss": 0.0699, "step": 19560 }, { "grad_norm": 0.607620894908905, "learning_rate": 9.430839138896026e-05, "loss": 0.0786, "step": 19570 }, { "grad_norm": 0.698514997959137, "learning_rate": 9.43007273756039e-05, "loss": 0.0837, "step": 19580 }, { "grad_norm": 0.5568724274635315, "learning_rate": 9.429305851758658e-05, "loss": 0.0797, "step": 19590 }, { "grad_norm": 0.5418863296508789, "learning_rate": 9.428538481574699e-05, "loss": 0.0718, "step": 19600 }, { "grad_norm": 0.5279957056045532, "learning_rate": 9.42777062709243e-05, "loss": 0.0776, "step": 19610 }, { "grad_norm": 0.4871143698692322, "learning_rate": 9.427002288395821e-05, "loss": 0.0649, "step": 19620 }, { "grad_norm": 0.6319656372070312, "learning_rate": 9.426233465568898e-05, "loss": 0.0697, "step": 19630 }, { "grad_norm": 0.49785321950912476, "learning_rate": 9.42546415869574e-05, "loss": 0.069, "step": 19640 }, { "grad_norm": 0.4969632029533386, "learning_rate": 9.424694367860473e-05, "loss": 0.0541, "step": 19650 }, { "grad_norm": 0.615479588508606, "learning_rate": 9.423924093147284e-05, "loss": 0.0576, "step": 19660 }, { "grad_norm": 0.5274770855903625, "learning_rate": 9.423153334640407e-05, "loss": 0.0753, "step": 19670 }, { "grad_norm": 0.4428713023662567, "learning_rate": 9.42238209242413e-05, "loss": 0.058, "step": 19680 }, { "grad_norm": 0.38339728116989136, "learning_rate": 9.421610366582798e-05, "loss": 0.0812, "step": 19690 }, { "grad_norm": 0.6719378232955933, "learning_rate": 9.420838157200803e-05, "loss": 0.0714, "step": 19700 }, { "grad_norm": 0.5764545202255249, "learning_rate": 9.420065464362594e-05, "loss": 0.0885, "step": 19710 }, { "grad_norm": 0.7598796486854553, "learning_rate": 9.419292288152673e-05, "loss": 0.0739, "step": 19720 }, { "grad_norm": 0.45409464836120605, "learning_rate": 9.418518628655588e-05, "loss": 0.0712, "step": 19730 }, { "grad_norm": 0.44987693428993225, "learning_rate": 9.417744485955951e-05, "loss": 0.0626, "step": 19740 }, { "grad_norm": 0.5203339457511902, "learning_rate": 9.41696986013842e-05, "loss": 0.0704, "step": 19750 }, { "grad_norm": 0.6839748024940491, "learning_rate": 9.416194751287705e-05, "loss": 0.1031, "step": 19760 }, { "grad_norm": 0.6661019921302795, "learning_rate": 9.415419159488572e-05, "loss": 0.0625, "step": 19770 }, { "grad_norm": 0.7630220055580139, "learning_rate": 9.414643084825837e-05, "loss": 0.0799, "step": 19780 }, { "grad_norm": 0.5350911617279053, "learning_rate": 9.413866527384372e-05, "loss": 0.0731, "step": 19790 }, { "grad_norm": 0.8733565807342529, "learning_rate": 9.4130894872491e-05, "loss": 0.0773, "step": 19800 }, { "grad_norm": 0.9347449541091919, "learning_rate": 9.412311964504998e-05, "loss": 0.0782, "step": 19810 }, { "grad_norm": 0.5372282266616821, "learning_rate": 9.411533959237091e-05, "loss": 0.0821, "step": 19820 }, { "grad_norm": 0.36681973934173584, "learning_rate": 9.410755471530464e-05, "loss": 0.0558, "step": 19830 }, { "grad_norm": 0.7583183646202087, "learning_rate": 9.40997650147025e-05, "loss": 0.0811, "step": 19840 }, { "grad_norm": 0.6455779075622559, "learning_rate": 9.409197049141637e-05, "loss": 0.085, "step": 19850 }, { "grad_norm": 0.488014280796051, "learning_rate": 9.408417114629863e-05, "loss": 0.0772, "step": 19860 }, { "grad_norm": 0.4469667375087738, "learning_rate": 9.40763669802022e-05, "loss": 0.0645, "step": 19870 }, { "grad_norm": 0.6899768114089966, "learning_rate": 9.406855799398056e-05, "loss": 0.0867, "step": 19880 }, { "grad_norm": 0.6070411205291748, "learning_rate": 9.406074418848767e-05, "loss": 0.0779, "step": 19890 }, { "grad_norm": 0.7084612846374512, "learning_rate": 9.405292556457805e-05, "loss": 0.093, "step": 19900 }, { "grad_norm": 0.5223615169525146, "learning_rate": 9.404510212310671e-05, "loss": 0.0832, "step": 19910 }, { "grad_norm": 0.4414767026901245, "learning_rate": 9.403727386492924e-05, "loss": 0.0731, "step": 19920 }, { "grad_norm": 0.43180006742477417, "learning_rate": 9.40294407909017e-05, "loss": 0.07, "step": 19930 }, { "grad_norm": 0.8093221783638, "learning_rate": 9.40216029018807e-05, "loss": 0.0824, "step": 19940 }, { "grad_norm": 0.6834551692008972, "learning_rate": 9.401376019872338e-05, "loss": 0.0753, "step": 19950 }, { "grad_norm": 0.5785003304481506, "learning_rate": 9.400591268228746e-05, "loss": 0.0793, "step": 19960 }, { "grad_norm": 0.7068715691566467, "learning_rate": 9.399806035343106e-05, "loss": 0.0767, "step": 19970 }, { "grad_norm": 0.4608284831047058, "learning_rate": 9.399020321301294e-05, "loss": 0.0715, "step": 19980 }, { "grad_norm": 0.4653612971305847, "learning_rate": 9.398234126189234e-05, "loss": 0.087, "step": 19990 }, { "grad_norm": 0.5154697299003601, "learning_rate": 9.397447450092902e-05, "loss": 0.0728, "step": 20000 }, { "grad_norm": 0.40154650807380676, "learning_rate": 9.39666029309833e-05, "loss": 0.0637, "step": 20010 }, { "grad_norm": 0.5873201489448547, "learning_rate": 9.395872655291596e-05, "loss": 0.0723, "step": 20020 }, { "grad_norm": 0.5686072111129761, "learning_rate": 9.395084536758838e-05, "loss": 0.0751, "step": 20030 }, { "grad_norm": 0.7568693161010742, "learning_rate": 9.394295937586243e-05, "loss": 0.0988, "step": 20040 }, { "grad_norm": 0.5316621661186218, "learning_rate": 9.393506857860052e-05, "loss": 0.076, "step": 20050 }, { "grad_norm": 0.5676901936531067, "learning_rate": 9.392717297666555e-05, "loss": 0.0628, "step": 20060 }, { "grad_norm": 0.7122811675071716, "learning_rate": 9.391927257092101e-05, "loss": 0.0821, "step": 20070 }, { "grad_norm": 0.6762874722480774, "learning_rate": 9.391136736223085e-05, "loss": 0.0801, "step": 20080 }, { "grad_norm": 0.4632270038127899, "learning_rate": 9.390345735145956e-05, "loss": 0.0793, "step": 20090 }, { "grad_norm": 0.8494360446929932, "learning_rate": 9.389554253947219e-05, "loss": 0.0797, "step": 20100 }, { "grad_norm": 0.46178826689720154, "learning_rate": 9.388762292713428e-05, "loss": 0.0859, "step": 20110 }, { "grad_norm": 0.7213112115859985, "learning_rate": 9.38796985153119e-05, "loss": 0.0813, "step": 20120 }, { "grad_norm": 0.5752737522125244, "learning_rate": 9.387176930487169e-05, "loss": 0.0834, "step": 20130 }, { "grad_norm": 1.1513139009475708, "learning_rate": 9.386383529668072e-05, "loss": 0.0712, "step": 20140 }, { "grad_norm": 0.9386141896247864, "learning_rate": 9.385589649160669e-05, "loss": 0.0738, "step": 20150 }, { "grad_norm": 0.48107048869132996, "learning_rate": 9.384795289051775e-05, "loss": 0.081, "step": 20160 }, { "grad_norm": 0.6776064038276672, "learning_rate": 9.384000449428261e-05, "loss": 0.078, "step": 20170 }, { "grad_norm": 0.7836536765098572, "learning_rate": 9.383205130377048e-05, "loss": 0.0771, "step": 20180 }, { "grad_norm": 0.42180249094963074, "learning_rate": 9.382409331985114e-05, "loss": 0.0685, "step": 20190 }, { "grad_norm": 1.004732370376587, "learning_rate": 9.381613054339482e-05, "loss": 0.0701, "step": 20200 }, { "grad_norm": 0.5367337465286255, "learning_rate": 9.380816297527235e-05, "loss": 0.0775, "step": 20210 }, { "grad_norm": 0.6334348320960999, "learning_rate": 9.380019061635506e-05, "loss": 0.0723, "step": 20220 }, { "grad_norm": 0.4549106955528259, "learning_rate": 9.379221346751474e-05, "loss": 0.0875, "step": 20230 }, { "grad_norm": 0.5815432071685791, "learning_rate": 9.378423152962382e-05, "loss": 0.0702, "step": 20240 }, { "grad_norm": 0.5931077003479004, "learning_rate": 9.377624480355517e-05, "loss": 0.066, "step": 20250 }, { "grad_norm": 0.6404350399971008, "learning_rate": 9.376825329018219e-05, "loss": 0.0717, "step": 20260 }, { "grad_norm": 0.7834556102752686, "learning_rate": 9.376025699037884e-05, "loss": 0.0788, "step": 20270 }, { "grad_norm": 0.7045333385467529, "learning_rate": 9.37522559050196e-05, "loss": 0.0751, "step": 20280 }, { "grad_norm": 0.5094547271728516, "learning_rate": 9.37442500349794e-05, "loss": 0.0748, "step": 20290 }, { "grad_norm": 0.6013284921646118, "learning_rate": 9.373623938113381e-05, "loss": 0.0706, "step": 20300 }, { "grad_norm": 0.5240951776504517, "learning_rate": 9.372822394435883e-05, "loss": 0.0816, "step": 20310 }, { "grad_norm": 0.41674524545669556, "learning_rate": 9.372020372553102e-05, "loss": 0.0708, "step": 20320 }, { "grad_norm": 0.5175225138664246, "learning_rate": 9.371217872552746e-05, "loss": 0.0773, "step": 20330 }, { "grad_norm": 0.49478620290756226, "learning_rate": 9.370414894522576e-05, "loss": 0.0716, "step": 20340 }, { "grad_norm": 0.6959102153778076, "learning_rate": 9.369611438550406e-05, "loss": 0.0839, "step": 20350 }, { "grad_norm": 0.9456933736801147, "learning_rate": 9.368807504724095e-05, "loss": 0.0834, "step": 20360 }, { "grad_norm": 0.8598072528839111, "learning_rate": 9.368003093131565e-05, "loss": 0.0856, "step": 20370 }, { "grad_norm": 0.5516594052314758, "learning_rate": 9.367198203860785e-05, "loss": 0.091, "step": 20380 }, { "grad_norm": 0.47388869524002075, "learning_rate": 9.366392836999774e-05, "loss": 0.0626, "step": 20390 }, { "grad_norm": 0.6635566353797913, "learning_rate": 9.365586992636607e-05, "loss": 0.0759, "step": 20400 }, { "grad_norm": 0.5797755122184753, "learning_rate": 9.364780670859412e-05, "loss": 0.0674, "step": 20410 }, { "grad_norm": 0.4793945848941803, "learning_rate": 9.363973871756364e-05, "loss": 0.069, "step": 20420 }, { "grad_norm": 0.7349984645843506, "learning_rate": 9.363166595415696e-05, "loss": 0.0764, "step": 20430 }, { "grad_norm": 0.6822915077209473, "learning_rate": 9.362358841925686e-05, "loss": 0.0733, "step": 20440 }, { "grad_norm": 0.5246841907501221, "learning_rate": 9.361550611374674e-05, "loss": 0.0781, "step": 20450 }, { "grad_norm": 0.46759071946144104, "learning_rate": 9.360741903851043e-05, "loss": 0.0989, "step": 20460 }, { "grad_norm": 0.7825506925582886, "learning_rate": 9.359932719443236e-05, "loss": 0.0993, "step": 20470 }, { "grad_norm": 0.47387075424194336, "learning_rate": 9.35912305823974e-05, "loss": 0.0711, "step": 20480 }, { "grad_norm": 0.6542593240737915, "learning_rate": 9.358312920329101e-05, "loss": 0.0829, "step": 20490 }, { "grad_norm": 0.5741099119186401, "learning_rate": 9.357502305799914e-05, "loss": 0.0663, "step": 20500 }, { "grad_norm": 0.5515969395637512, "learning_rate": 9.356691214740824e-05, "loss": 0.0629, "step": 20510 }, { "grad_norm": 0.6034601926803589, "learning_rate": 9.355879647240535e-05, "loss": 0.0729, "step": 20520 }, { "grad_norm": 0.5431663393974304, "learning_rate": 9.355067603387798e-05, "loss": 0.0788, "step": 20530 }, { "grad_norm": 0.5755314826965332, "learning_rate": 9.354255083271412e-05, "loss": 0.0856, "step": 20540 }, { "grad_norm": 0.6474774479866028, "learning_rate": 9.353442086980239e-05, "loss": 0.0752, "step": 20550 }, { "grad_norm": 0.4264608919620514, "learning_rate": 9.352628614603185e-05, "loss": 0.076, "step": 20560 }, { "grad_norm": 0.7262622714042664, "learning_rate": 9.351814666229209e-05, "loss": 0.0778, "step": 20570 }, { "grad_norm": 0.8270207047462463, "learning_rate": 9.351000241947324e-05, "loss": 0.0714, "step": 20580 }, { "grad_norm": 0.6069445013999939, "learning_rate": 9.350185341846594e-05, "loss": 0.074, "step": 20590 }, { "grad_norm": 0.3831814229488373, "learning_rate": 9.349369966016134e-05, "loss": 0.0725, "step": 20600 }, { "grad_norm": 0.7241611480712891, "learning_rate": 9.348554114545117e-05, "loss": 0.0989, "step": 20610 }, { "grad_norm": 0.5332961082458496, "learning_rate": 9.347737787522758e-05, "loss": 0.081, "step": 20620 }, { "grad_norm": 0.5007699131965637, "learning_rate": 9.346920985038332e-05, "loss": 0.09, "step": 20630 }, { "grad_norm": 0.6888684034347534, "learning_rate": 9.346103707181162e-05, "loss": 0.0829, "step": 20640 }, { "grad_norm": 0.43374744057655334, "learning_rate": 9.345285954040626e-05, "loss": 0.0688, "step": 20650 }, { "grad_norm": 0.5367761254310608, "learning_rate": 9.34446772570615e-05, "loss": 0.0769, "step": 20660 }, { "grad_norm": 0.571151852607727, "learning_rate": 9.343649022267214e-05, "loss": 0.0635, "step": 20670 }, { "grad_norm": 1.6600213050842285, "learning_rate": 9.342829843813353e-05, "loss": 0.0809, "step": 20680 }, { "grad_norm": 0.6786293387413025, "learning_rate": 9.342010190434149e-05, "loss": 0.0845, "step": 20690 }, { "grad_norm": 0.9520493149757385, "learning_rate": 9.34119006221924e-05, "loss": 0.0903, "step": 20700 }, { "grad_norm": 0.8003093004226685, "learning_rate": 9.340369459258313e-05, "loss": 0.0788, "step": 20710 }, { "grad_norm": 0.7839937210083008, "learning_rate": 9.339548381641106e-05, "loss": 0.0865, "step": 20720 }, { "grad_norm": 0.6424801349639893, "learning_rate": 9.338726829457413e-05, "loss": 0.0857, "step": 20730 }, { "grad_norm": 0.6975129246711731, "learning_rate": 9.337904802797078e-05, "loss": 0.0693, "step": 20740 }, { "grad_norm": 0.6456841826438904, "learning_rate": 9.337082301749993e-05, "loss": 0.0741, "step": 20750 }, { "grad_norm": 0.5647163391113281, "learning_rate": 9.336259326406109e-05, "loss": 0.0757, "step": 20760 }, { "grad_norm": 0.6775135397911072, "learning_rate": 9.335435876855427e-05, "loss": 0.0856, "step": 20770 }, { "grad_norm": 0.6119949221611023, "learning_rate": 9.334611953187994e-05, "loss": 0.0758, "step": 20780 }, { "grad_norm": 0.564015805721283, "learning_rate": 9.333787555493914e-05, "loss": 0.0685, "step": 20790 }, { "grad_norm": 0.9211437106132507, "learning_rate": 9.332962683863345e-05, "loss": 0.0892, "step": 20800 }, { "grad_norm": 0.4495052397251129, "learning_rate": 9.332137338386489e-05, "loss": 0.0731, "step": 20810 }, { "grad_norm": 0.4953277111053467, "learning_rate": 9.33131151915361e-05, "loss": 0.0907, "step": 20820 }, { "grad_norm": 0.6432408690452576, "learning_rate": 9.330485226255012e-05, "loss": 0.0723, "step": 20830 }, { "grad_norm": 0.4377516210079193, "learning_rate": 9.329658459781061e-05, "loss": 0.0708, "step": 20840 }, { "grad_norm": 0.37772423028945923, "learning_rate": 9.328831219822172e-05, "loss": 0.0765, "step": 20850 }, { "grad_norm": 0.5259264707565308, "learning_rate": 9.328003506468808e-05, "loss": 0.0859, "step": 20860 }, { "grad_norm": 0.6071046590805054, "learning_rate": 9.327175319811488e-05, "loss": 0.0726, "step": 20870 }, { "grad_norm": 0.5833677053451538, "learning_rate": 9.326346659940781e-05, "loss": 0.0714, "step": 20880 }, { "grad_norm": 0.4038889706134796, "learning_rate": 9.325517526947308e-05, "loss": 0.0719, "step": 20890 }, { "grad_norm": 0.5404285788536072, "learning_rate": 9.32468792092174e-05, "loss": 0.0793, "step": 20900 }, { "grad_norm": 0.48517394065856934, "learning_rate": 9.323857841954803e-05, "loss": 0.0739, "step": 20910 }, { "grad_norm": 0.819131076335907, "learning_rate": 9.323027290137276e-05, "loss": 0.0741, "step": 20920 }, { "grad_norm": 0.5148729681968689, "learning_rate": 9.322196265559981e-05, "loss": 0.0785, "step": 20930 }, { "grad_norm": 0.5254265069961548, "learning_rate": 9.321364768313803e-05, "loss": 0.0777, "step": 20940 }, { "grad_norm": 0.7080069184303284, "learning_rate": 9.32053279848967e-05, "loss": 0.0795, "step": 20950 }, { "grad_norm": 0.7114814519882202, "learning_rate": 9.319700356178567e-05, "loss": 0.0772, "step": 20960 }, { "grad_norm": 0.6030357480049133, "learning_rate": 9.318867441471527e-05, "loss": 0.0691, "step": 20970 }, { "grad_norm": 0.5845833420753479, "learning_rate": 9.318034054459637e-05, "loss": 0.0575, "step": 20980 }, { "grad_norm": 0.4869217872619629, "learning_rate": 9.317200195234034e-05, "loss": 0.0833, "step": 20990 }, { "grad_norm": 0.49960431456565857, "learning_rate": 9.316365863885909e-05, "loss": 0.0858, "step": 21000 }, { "grad_norm": 0.7861506938934326, "learning_rate": 9.315531060506502e-05, "loss": 0.0894, "step": 21010 }, { "grad_norm": 0.6237455010414124, "learning_rate": 9.314695785187108e-05, "loss": 0.0787, "step": 21020 }, { "grad_norm": 0.573911726474762, "learning_rate": 9.313860038019069e-05, "loss": 0.0729, "step": 21030 }, { "grad_norm": 1.5265581607818604, "learning_rate": 9.313023819093782e-05, "loss": 0.0967, "step": 21040 }, { "grad_norm": 0.6644562482833862, "learning_rate": 9.312187128502695e-05, "loss": 0.0826, "step": 21050 }, { "grad_norm": 0.9454654455184937, "learning_rate": 9.311349966337307e-05, "loss": 0.0921, "step": 21060 }, { "grad_norm": 0.8267561793327332, "learning_rate": 9.310512332689169e-05, "loss": 0.0831, "step": 21070 }, { "grad_norm": 0.7067095637321472, "learning_rate": 9.309674227649883e-05, "loss": 0.0743, "step": 21080 }, { "grad_norm": 0.8025397658348083, "learning_rate": 9.308835651311103e-05, "loss": 0.073, "step": 21090 }, { "grad_norm": 0.6005744338035583, "learning_rate": 9.307996603764533e-05, "loss": 0.0799, "step": 21100 }, { "grad_norm": 0.4947572648525238, "learning_rate": 9.307157085101932e-05, "loss": 0.0729, "step": 21110 }, { "grad_norm": 1.1122636795043945, "learning_rate": 9.306317095415109e-05, "loss": 0.0731, "step": 21120 }, { "grad_norm": 0.411735475063324, "learning_rate": 9.305476634795922e-05, "loss": 0.0678, "step": 21130 }, { "grad_norm": 0.48267924785614014, "learning_rate": 9.304635703336284e-05, "loss": 0.0705, "step": 21140 }, { "grad_norm": 0.4980216324329376, "learning_rate": 9.303794301128157e-05, "loss": 0.0671, "step": 21150 }, { "grad_norm": 0.5308330655097961, "learning_rate": 9.302952428263555e-05, "loss": 0.0749, "step": 21160 }, { "grad_norm": 0.8720249533653259, "learning_rate": 9.302110084834545e-05, "loss": 0.0699, "step": 21170 }, { "grad_norm": 0.4713704288005829, "learning_rate": 9.301267270933245e-05, "loss": 0.0555, "step": 21180 }, { "grad_norm": 0.6575721502304077, "learning_rate": 9.300423986651823e-05, "loss": 0.0726, "step": 21190 }, { "grad_norm": 0.567232072353363, "learning_rate": 9.299580232082501e-05, "loss": 0.0732, "step": 21200 }, { "grad_norm": 0.75435870885849, "learning_rate": 9.298736007317547e-05, "loss": 0.0639, "step": 21210 }, { "grad_norm": 0.5894647836685181, "learning_rate": 9.297891312449288e-05, "loss": 0.0731, "step": 21220 }, { "grad_norm": 0.5258097648620605, "learning_rate": 9.297046147570094e-05, "loss": 0.0776, "step": 21230 }, { "grad_norm": 1.0500924587249756, "learning_rate": 9.296200512772396e-05, "loss": 0.0729, "step": 21240 }, { "grad_norm": 0.7626551389694214, "learning_rate": 9.295354408148668e-05, "loss": 0.0712, "step": 21250 }, { "grad_norm": 0.6126708984375, "learning_rate": 9.294507833791441e-05, "loss": 0.0709, "step": 21260 }, { "grad_norm": 1.262044072151184, "learning_rate": 9.293660789793295e-05, "loss": 0.0799, "step": 21270 }, { "grad_norm": 0.44715139269828796, "learning_rate": 9.292813276246858e-05, "loss": 0.0707, "step": 21280 }, { "grad_norm": 0.5887997150421143, "learning_rate": 9.291965293244816e-05, "loss": 0.0694, "step": 21290 }, { "grad_norm": 0.5684338212013245, "learning_rate": 9.291116840879904e-05, "loss": 0.0637, "step": 21300 }, { "grad_norm": 0.7084646224975586, "learning_rate": 9.290267919244904e-05, "loss": 0.0746, "step": 21310 }, { "grad_norm": 0.9628027081489563, "learning_rate": 9.289418528432655e-05, "loss": 0.0776, "step": 21320 }, { "grad_norm": 0.9650506377220154, "learning_rate": 9.288568668536045e-05, "loss": 0.0992, "step": 21330 }, { "grad_norm": 0.5841540098190308, "learning_rate": 9.287718339648013e-05, "loss": 0.0724, "step": 21340 }, { "grad_norm": 1.0323702096939087, "learning_rate": 9.28686754186155e-05, "loss": 0.0798, "step": 21350 }, { "grad_norm": 0.7812164425849915, "learning_rate": 9.286016275269698e-05, "loss": 0.0969, "step": 21360 }, { "grad_norm": 0.4603539705276489, "learning_rate": 9.285164539965551e-05, "loss": 0.0725, "step": 21370 }, { "grad_norm": 0.4688640236854553, "learning_rate": 9.284312336042251e-05, "loss": 0.0677, "step": 21380 }, { "grad_norm": 0.6576806306838989, "learning_rate": 9.283459663592996e-05, "loss": 0.0673, "step": 21390 }, { "grad_norm": 0.5775365829467773, "learning_rate": 9.282606522711033e-05, "loss": 0.0694, "step": 21400 }, { "grad_norm": 0.6584802865982056, "learning_rate": 9.281752913489657e-05, "loss": 0.0874, "step": 21410 }, { "grad_norm": 0.48039430379867554, "learning_rate": 9.280898836022222e-05, "loss": 0.068, "step": 21420 }, { "grad_norm": 0.5904822945594788, "learning_rate": 9.280044290402126e-05, "loss": 0.0652, "step": 21430 }, { "grad_norm": 0.5305498242378235, "learning_rate": 9.279189276722821e-05, "loss": 0.065, "step": 21440 }, { "grad_norm": 0.5979146361351013, "learning_rate": 9.278333795077812e-05, "loss": 0.0722, "step": 21450 }, { "grad_norm": 0.5303770303726196, "learning_rate": 9.27747784556065e-05, "loss": 0.0591, "step": 21460 }, { "grad_norm": 0.7317502498626709, "learning_rate": 9.276621428264942e-05, "loss": 0.0832, "step": 21470 }, { "grad_norm": 0.6018704175949097, "learning_rate": 9.275764543284345e-05, "loss": 0.0723, "step": 21480 }, { "grad_norm": 0.5525532960891724, "learning_rate": 9.274907190712566e-05, "loss": 0.0713, "step": 21490 }, { "grad_norm": 1.2577598094940186, "learning_rate": 9.274049370643363e-05, "loss": 0.086, "step": 21500 }, { "grad_norm": 1.3472070693969727, "learning_rate": 9.273191083170547e-05, "loss": 0.0798, "step": 21510 }, { "grad_norm": 0.5549918413162231, "learning_rate": 9.27233232838798e-05, "loss": 0.0701, "step": 21520 }, { "grad_norm": 0.4363701641559601, "learning_rate": 9.27147310638957e-05, "loss": 0.0722, "step": 21530 }, { "grad_norm": 0.3797568380832672, "learning_rate": 9.270613417269286e-05, "loss": 0.0799, "step": 21540 }, { "grad_norm": 0.5676025152206421, "learning_rate": 9.269753261121138e-05, "loss": 0.074, "step": 21550 }, { "grad_norm": 0.40456128120422363, "learning_rate": 9.268892638039194e-05, "loss": 0.0708, "step": 21560 }, { "grad_norm": 0.48637211322784424, "learning_rate": 9.268031548117569e-05, "loss": 0.0786, "step": 21570 }, { "grad_norm": 0.4202674329280853, "learning_rate": 9.26716999145043e-05, "loss": 0.0615, "step": 21580 }, { "grad_norm": 0.661077082157135, "learning_rate": 9.266307968131998e-05, "loss": 0.0682, "step": 21590 }, { "grad_norm": 0.6673147678375244, "learning_rate": 9.26544547825654e-05, "loss": 0.0629, "step": 21600 }, { "grad_norm": 0.4699094593524933, "learning_rate": 9.264582521918376e-05, "loss": 0.0694, "step": 21610 }, { "grad_norm": 0.656116783618927, "learning_rate": 9.263719099211881e-05, "loss": 0.062, "step": 21620 }, { "grad_norm": 0.708877444267273, "learning_rate": 9.262855210231476e-05, "loss": 0.0802, "step": 21630 }, { "grad_norm": 0.49893173575401306, "learning_rate": 9.261990855071633e-05, "loss": 0.0764, "step": 21640 }, { "grad_norm": 0.6864891052246094, "learning_rate": 9.261126033826878e-05, "loss": 0.0651, "step": 21650 }, { "grad_norm": 0.44954952597618103, "learning_rate": 9.260260746591786e-05, "loss": 0.0657, "step": 21660 }, { "grad_norm": 0.4358295500278473, "learning_rate": 9.259394993460985e-05, "loss": 0.0695, "step": 21670 }, { "grad_norm": 0.6423358917236328, "learning_rate": 9.258528774529151e-05, "loss": 0.0665, "step": 21680 }, { "grad_norm": 0.5047716498374939, "learning_rate": 9.257662089891013e-05, "loss": 0.0634, "step": 21690 }, { "grad_norm": 0.5606237053871155, "learning_rate": 9.25679493964135e-05, "loss": 0.0715, "step": 21700 }, { "grad_norm": 0.46408769488334656, "learning_rate": 9.255927323874994e-05, "loss": 0.0569, "step": 21710 }, { "grad_norm": 0.5816148519515991, "learning_rate": 9.255059242686822e-05, "loss": 0.086, "step": 21720 }, { "grad_norm": 0.4532894194126129, "learning_rate": 9.254190696171769e-05, "loss": 0.0725, "step": 21730 }, { "grad_norm": 0.6839203238487244, "learning_rate": 9.25332168442482e-05, "loss": 0.0949, "step": 21740 }, { "grad_norm": 0.5487110614776611, "learning_rate": 9.252452207541004e-05, "loss": 0.0658, "step": 21750 }, { "grad_norm": 0.8942174911499023, "learning_rate": 9.251582265615409e-05, "loss": 0.093, "step": 21760 }, { "grad_norm": 0.6807253360748291, "learning_rate": 9.250711858743169e-05, "loss": 0.0842, "step": 21770 }, { "grad_norm": 0.528158962726593, "learning_rate": 9.24984098701947e-05, "loss": 0.0895, "step": 21780 }, { "grad_norm": 0.49781352281570435, "learning_rate": 9.248969650539552e-05, "loss": 0.084, "step": 21790 }, { "grad_norm": 0.6196095943450928, "learning_rate": 9.2480978493987e-05, "loss": 0.0824, "step": 21800 }, { "grad_norm": 0.5167220234870911, "learning_rate": 9.247225583692256e-05, "loss": 0.0726, "step": 21810 }, { "grad_norm": 0.730483889579773, "learning_rate": 9.246352853515607e-05, "loss": 0.0737, "step": 21820 }, { "grad_norm": 0.6186240315437317, "learning_rate": 9.245479658964194e-05, "loss": 0.0682, "step": 21830 }, { "grad_norm": 0.7158065438270569, "learning_rate": 9.244606000133507e-05, "loss": 0.0743, "step": 21840 }, { "grad_norm": 0.38624751567840576, "learning_rate": 9.24373187711909e-05, "loss": 0.0643, "step": 21850 }, { "grad_norm": 0.7153670787811279, "learning_rate": 9.242857290016537e-05, "loss": 0.0709, "step": 21860 }, { "grad_norm": 0.5051316022872925, "learning_rate": 9.241982238921488e-05, "loss": 0.0797, "step": 21870 }, { "grad_norm": 0.5633518695831299, "learning_rate": 9.24110672392964e-05, "loss": 0.0571, "step": 21880 }, { "grad_norm": 0.39877137541770935, "learning_rate": 9.240230745136737e-05, "loss": 0.0683, "step": 21890 }, { "grad_norm": 0.6040505766868591, "learning_rate": 9.239354302638575e-05, "loss": 0.0634, "step": 21900 }, { "grad_norm": 0.6027412414550781, "learning_rate": 9.238477396531e-05, "loss": 0.0764, "step": 21910 }, { "grad_norm": 0.5603957176208496, "learning_rate": 9.23760002690991e-05, "loss": 0.0643, "step": 21920 }, { "grad_norm": 0.4485113024711609, "learning_rate": 9.236722193871252e-05, "loss": 0.0711, "step": 21930 }, { "grad_norm": 0.43217670917510986, "learning_rate": 9.235843897511023e-05, "loss": 0.0755, "step": 21940 }, { "grad_norm": 1.28190016746521, "learning_rate": 9.234965137925276e-05, "loss": 0.0647, "step": 21950 }, { "grad_norm": 0.535374104976654, "learning_rate": 9.234085915210108e-05, "loss": 0.0894, "step": 21960 }, { "grad_norm": 0.7242379784584045, "learning_rate": 9.23320622946167e-05, "loss": 0.0772, "step": 21970 }, { "grad_norm": 0.5189918279647827, "learning_rate": 9.232326080776163e-05, "loss": 0.0634, "step": 21980 }, { "grad_norm": 1.0418996810913086, "learning_rate": 9.23144546924984e-05, "loss": 0.0649, "step": 21990 }, { "grad_norm": 0.7784345746040344, "learning_rate": 9.230564394979e-05, "loss": 0.0671, "step": 22000 }, { "grad_norm": 0.7005460858345032, "learning_rate": 9.22968285806e-05, "loss": 0.0601, "step": 22010 }, { "grad_norm": 0.6678425669670105, "learning_rate": 9.228800858589242e-05, "loss": 0.0757, "step": 22020 }, { "grad_norm": 0.4533831477165222, "learning_rate": 9.227918396663179e-05, "loss": 0.0761, "step": 22030 }, { "grad_norm": 1.1087011098861694, "learning_rate": 9.227035472378319e-05, "loss": 0.0764, "step": 22040 }, { "grad_norm": 0.48972287774086, "learning_rate": 9.226152085831213e-05, "loss": 0.0729, "step": 22050 }, { "grad_norm": 0.6629826426506042, "learning_rate": 9.22526823711847e-05, "loss": 0.0659, "step": 22060 }, { "grad_norm": 0.4575481712818146, "learning_rate": 9.224383926336745e-05, "loss": 0.0682, "step": 22070 }, { "grad_norm": 0.4238409101963043, "learning_rate": 9.223499153582744e-05, "loss": 0.0528, "step": 22080 }, { "grad_norm": 0.5909479260444641, "learning_rate": 9.222613918953226e-05, "loss": 0.0778, "step": 22090 }, { "grad_norm": 0.6417248845100403, "learning_rate": 9.221728222544999e-05, "loss": 0.0748, "step": 22100 }, { "grad_norm": 0.5707734227180481, "learning_rate": 9.22084206445492e-05, "loss": 0.0791, "step": 22110 }, { "grad_norm": 0.5781667232513428, "learning_rate": 9.2199554447799e-05, "loss": 0.0884, "step": 22120 }, { "grad_norm": 0.4197624921798706, "learning_rate": 9.219068363616897e-05, "loss": 0.076, "step": 22130 }, { "grad_norm": 0.6413480639457703, "learning_rate": 9.218180821062919e-05, "loss": 0.0774, "step": 22140 }, { "grad_norm": 0.7748538255691528, "learning_rate": 9.21729281721503e-05, "loss": 0.0845, "step": 22150 }, { "grad_norm": 1.1405123472213745, "learning_rate": 9.216404352170339e-05, "loss": 0.0851, "step": 22160 }, { "grad_norm": 0.5881456136703491, "learning_rate": 9.215515426026007e-05, "loss": 0.0728, "step": 22170 }, { "grad_norm": 0.6399917602539062, "learning_rate": 9.214626038879246e-05, "loss": 0.068, "step": 22180 }, { "grad_norm": 0.7334755063056946, "learning_rate": 9.21373619082732e-05, "loss": 0.0757, "step": 22190 }, { "grad_norm": 0.8365433812141418, "learning_rate": 9.212845881967535e-05, "loss": 0.0729, "step": 22200 }, { "grad_norm": 0.7609314322471619, "learning_rate": 9.211955112397262e-05, "loss": 0.0697, "step": 22210 }, { "grad_norm": 0.43525034189224243, "learning_rate": 9.211063882213909e-05, "loss": 0.0763, "step": 22220 }, { "grad_norm": 0.6000063419342041, "learning_rate": 9.210172191514942e-05, "loss": 0.0683, "step": 22230 }, { "grad_norm": 0.5336659550666809, "learning_rate": 9.209280040397874e-05, "loss": 0.0744, "step": 22240 }, { "grad_norm": 0.5591168403625488, "learning_rate": 9.208387428960268e-05, "loss": 0.0615, "step": 22250 }, { "grad_norm": 0.6226292252540588, "learning_rate": 9.20749435729974e-05, "loss": 0.0744, "step": 22260 }, { "grad_norm": 0.6163890957832336, "learning_rate": 9.206600825513957e-05, "loss": 0.0831, "step": 22270 }, { "grad_norm": 0.49550315737724304, "learning_rate": 9.20570683370063e-05, "loss": 0.0763, "step": 22280 }, { "grad_norm": 0.6621290445327759, "learning_rate": 9.204812381957528e-05, "loss": 0.0677, "step": 22290 }, { "grad_norm": 0.6253287196159363, "learning_rate": 9.203917470382465e-05, "loss": 0.0667, "step": 22300 }, { "grad_norm": 0.6747132539749146, "learning_rate": 9.203022099073309e-05, "loss": 0.0707, "step": 22310 }, { "grad_norm": 0.6686506271362305, "learning_rate": 9.202126268127976e-05, "loss": 0.0754, "step": 22320 }, { "grad_norm": 0.4495733082294464, "learning_rate": 9.20122997764443e-05, "loss": 0.0757, "step": 22330 }, { "grad_norm": 0.9563446044921875, "learning_rate": 9.200333227720692e-05, "loss": 0.0824, "step": 22340 }, { "grad_norm": 1.4633294343948364, "learning_rate": 9.199436018454826e-05, "loss": 0.0728, "step": 22350 }, { "grad_norm": 0.4866946339607239, "learning_rate": 9.198538349944952e-05, "loss": 0.0702, "step": 22360 }, { "grad_norm": 0.47071215510368347, "learning_rate": 9.197640222289234e-05, "loss": 0.0846, "step": 22370 }, { "grad_norm": 0.5977785587310791, "learning_rate": 9.196741635585895e-05, "loss": 0.0725, "step": 22380 }, { "grad_norm": 0.47980955243110657, "learning_rate": 9.195842589933199e-05, "loss": 0.062, "step": 22390 }, { "grad_norm": 0.4977458417415619, "learning_rate": 9.194943085429466e-05, "loss": 0.0751, "step": 22400 }, { "grad_norm": 0.4165741801261902, "learning_rate": 9.194043122173065e-05, "loss": 0.0909, "step": 22410 }, { "grad_norm": 0.5949761867523193, "learning_rate": 9.193142700262413e-05, "loss": 0.066, "step": 22420 }, { "grad_norm": 0.471554696559906, "learning_rate": 9.192241819795979e-05, "loss": 0.0577, "step": 22430 }, { "grad_norm": 0.4056625962257385, "learning_rate": 9.191340480872284e-05, "loss": 0.0686, "step": 22440 }, { "grad_norm": 0.5622199177742004, "learning_rate": 9.190438683589895e-05, "loss": 0.0837, "step": 22450 }, { "grad_norm": 0.4797905683517456, "learning_rate": 9.189536428047432e-05, "loss": 0.07, "step": 22460 }, { "grad_norm": 0.4022717773914337, "learning_rate": 9.188633714343564e-05, "loss": 0.0647, "step": 22470 }, { "grad_norm": 0.7214434742927551, "learning_rate": 9.18773054257701e-05, "loss": 0.0847, "step": 22480 }, { "grad_norm": 0.6687266230583191, "learning_rate": 9.18682691284654e-05, "loss": 0.065, "step": 22490 }, { "grad_norm": 0.5080021023750305, "learning_rate": 9.185922825250974e-05, "loss": 0.0706, "step": 22500 }, { "grad_norm": 0.586633563041687, "learning_rate": 9.185018279889181e-05, "loss": 0.0561, "step": 22510 }, { "grad_norm": 0.7056379318237305, "learning_rate": 9.184113276860082e-05, "loss": 0.0644, "step": 22520 }, { "grad_norm": 0.5560154914855957, "learning_rate": 9.183207816262645e-05, "loss": 0.0743, "step": 22530 }, { "grad_norm": 0.5981187224388123, "learning_rate": 9.182301898195891e-05, "loss": 0.0709, "step": 22540 }, { "grad_norm": 0.5016118884086609, "learning_rate": 9.181395522758889e-05, "loss": 0.0782, "step": 22550 }, { "grad_norm": 0.6767295598983765, "learning_rate": 9.180488690050759e-05, "loss": 0.0899, "step": 22560 }, { "grad_norm": 0.598806619644165, "learning_rate": 9.179581400170671e-05, "loss": 0.069, "step": 22570 }, { "grad_norm": 0.7289533019065857, "learning_rate": 9.178673653217845e-05, "loss": 0.0615, "step": 22580 }, { "grad_norm": 0.5900623202323914, "learning_rate": 9.177765449291551e-05, "loss": 0.0638, "step": 22590 }, { "grad_norm": 0.9886161088943481, "learning_rate": 9.176856788491109e-05, "loss": 0.0877, "step": 22600 }, { "grad_norm": 0.5644418597221375, "learning_rate": 9.175947670915887e-05, "loss": 0.0703, "step": 22610 }, { "grad_norm": 0.7629654407501221, "learning_rate": 9.175038096665309e-05, "loss": 0.0724, "step": 22620 }, { "grad_norm": 0.5175313353538513, "learning_rate": 9.17412806583884e-05, "loss": 0.0553, "step": 22630 }, { "grad_norm": 0.7743253707885742, "learning_rate": 9.173217578536002e-05, "loss": 0.0894, "step": 22640 }, { "grad_norm": 0.5745189785957336, "learning_rate": 9.172306634856362e-05, "loss": 0.0751, "step": 22650 }, { "grad_norm": 0.5335894227027893, "learning_rate": 9.171395234899545e-05, "loss": 0.0822, "step": 22660 }, { "grad_norm": 0.5322475433349609, "learning_rate": 9.170483378765214e-05, "loss": 0.0738, "step": 22670 }, { "grad_norm": 0.5148568153381348, "learning_rate": 9.169571066553091e-05, "loss": 0.0706, "step": 22680 }, { "grad_norm": 0.4971812665462494, "learning_rate": 9.168658298362946e-05, "loss": 0.0734, "step": 22690 }, { "grad_norm": 0.6007739305496216, "learning_rate": 9.167745074294598e-05, "loss": 0.0776, "step": 22700 }, { "grad_norm": 0.4717002511024475, "learning_rate": 9.166831394447913e-05, "loss": 0.0738, "step": 22710 }, { "grad_norm": 0.710172176361084, "learning_rate": 9.165917258922812e-05, "loss": 0.065, "step": 22720 }, { "grad_norm": 0.6936858296394348, "learning_rate": 9.165002667819262e-05, "loss": 0.0769, "step": 22730 }, { "grad_norm": 0.462791383266449, "learning_rate": 9.164087621237282e-05, "loss": 0.0667, "step": 22740 }, { "grad_norm": 0.5864302515983582, "learning_rate": 9.163172119276942e-05, "loss": 0.0855, "step": 22750 }, { "grad_norm": 0.5157551169395447, "learning_rate": 9.162256162038358e-05, "loss": 0.0873, "step": 22760 }, { "grad_norm": 0.7438257932662964, "learning_rate": 9.161339749621698e-05, "loss": 0.0702, "step": 22770 }, { "grad_norm": 0.4166215658187866, "learning_rate": 9.160422882127177e-05, "loss": 0.0692, "step": 22780 }, { "grad_norm": 0.41809165477752686, "learning_rate": 9.159505559655069e-05, "loss": 0.0692, "step": 22790 }, { "grad_norm": 0.8451606631278992, "learning_rate": 9.158587782305684e-05, "loss": 0.0759, "step": 22800 }, { "grad_norm": 0.5289822220802307, "learning_rate": 9.157669550179391e-05, "loss": 0.0735, "step": 22810 }, { "grad_norm": 0.36830979585647583, "learning_rate": 9.156750863376609e-05, "loss": 0.0764, "step": 22820 }, { "grad_norm": 0.567619264125824, "learning_rate": 9.155831721997801e-05, "loss": 0.0784, "step": 22830 }, { "grad_norm": 0.7037906646728516, "learning_rate": 9.154912126143484e-05, "loss": 0.0781, "step": 22840 }, { "grad_norm": 0.6488338708877563, "learning_rate": 9.153992075914224e-05, "loss": 0.0887, "step": 22850 }, { "grad_norm": 0.6063115000724792, "learning_rate": 9.153071571410635e-05, "loss": 0.0656, "step": 22860 }, { "grad_norm": 0.5199649930000305, "learning_rate": 9.152150612733384e-05, "loss": 0.0665, "step": 22870 }, { "grad_norm": 0.5298653244972229, "learning_rate": 9.151229199983184e-05, "loss": 0.0705, "step": 22880 }, { "grad_norm": 0.6225855946540833, "learning_rate": 9.150307333260802e-05, "loss": 0.0785, "step": 22890 }, { "grad_norm": 0.5865479111671448, "learning_rate": 9.149385012667048e-05, "loss": 0.0917, "step": 22900 }, { "grad_norm": 0.7478147149085999, "learning_rate": 9.148462238302788e-05, "loss": 0.0902, "step": 22910 }, { "grad_norm": 0.8061582446098328, "learning_rate": 9.147539010268936e-05, "loss": 0.0915, "step": 22920 }, { "grad_norm": 0.808742344379425, "learning_rate": 9.14661532866645e-05, "loss": 0.0876, "step": 22930 }, { "grad_norm": 0.7968221306800842, "learning_rate": 9.145691193596348e-05, "loss": 0.0683, "step": 22940 }, { "grad_norm": 0.6150432229042053, "learning_rate": 9.144766605159691e-05, "loss": 0.1018, "step": 22950 }, { "grad_norm": 0.9805545806884766, "learning_rate": 9.14384156345759e-05, "loss": 0.0848, "step": 22960 }, { "grad_norm": 0.7391517162322998, "learning_rate": 9.142916068591204e-05, "loss": 0.0695, "step": 22970 }, { "grad_norm": 0.6922769546508789, "learning_rate": 9.141990120661746e-05, "loss": 0.0747, "step": 22980 }, { "grad_norm": 0.5901910662651062, "learning_rate": 9.141063719770475e-05, "loss": 0.0777, "step": 22990 }, { "grad_norm": 0.47312185168266296, "learning_rate": 9.140136866018704e-05, "loss": 0.0715, "step": 23000 }, { "grad_norm": 0.5442067980766296, "learning_rate": 9.139209559507788e-05, "loss": 0.0681, "step": 23010 }, { "grad_norm": 0.5504363179206848, "learning_rate": 9.13828180033914e-05, "loss": 0.0676, "step": 23020 }, { "grad_norm": 0.6436241269111633, "learning_rate": 9.137353588614212e-05, "loss": 0.0756, "step": 23030 }, { "grad_norm": 0.4065932631492615, "learning_rate": 9.136424924434519e-05, "loss": 0.0798, "step": 23040 }, { "grad_norm": 0.6381221413612366, "learning_rate": 9.135495807901615e-05, "loss": 0.081, "step": 23050 }, { "grad_norm": 0.43104133009910583, "learning_rate": 9.134566239117108e-05, "loss": 0.0719, "step": 23060 }, { "grad_norm": 0.8120679259300232, "learning_rate": 9.13363621818265e-05, "loss": 0.0584, "step": 23070 }, { "grad_norm": 0.8851833939552307, "learning_rate": 9.132705745199953e-05, "loss": 0.0823, "step": 23080 }, { "grad_norm": 0.3797968924045563, "learning_rate": 9.131774820270768e-05, "loss": 0.0652, "step": 23090 }, { "grad_norm": 0.6577905416488647, "learning_rate": 9.130843443496901e-05, "loss": 0.0723, "step": 23100 }, { "grad_norm": 0.6188557744026184, "learning_rate": 9.129911614980206e-05, "loss": 0.0713, "step": 23110 }, { "grad_norm": 0.6282963752746582, "learning_rate": 9.128979334822584e-05, "loss": 0.0597, "step": 23120 }, { "grad_norm": 0.633120059967041, "learning_rate": 9.128046603125992e-05, "loss": 0.0632, "step": 23130 }, { "grad_norm": 0.4437216818332672, "learning_rate": 9.12711341999243e-05, "loss": 0.074, "step": 23140 }, { "grad_norm": 0.5596481561660767, "learning_rate": 9.12617978552395e-05, "loss": 0.084, "step": 23150 }, { "grad_norm": 0.3773511052131653, "learning_rate": 9.12524569982265e-05, "loss": 0.0622, "step": 23160 }, { "grad_norm": 0.6703824400901794, "learning_rate": 9.124311162990684e-05, "loss": 0.0685, "step": 23170 }, { "grad_norm": 0.5909264087677002, "learning_rate": 9.12337617513025e-05, "loss": 0.0707, "step": 23180 }, { "grad_norm": 0.5789366364479065, "learning_rate": 9.122440736343596e-05, "loss": 0.0656, "step": 23190 }, { "grad_norm": 0.3535482883453369, "learning_rate": 9.12150484673302e-05, "loss": 0.0714, "step": 23200 }, { "grad_norm": 0.4958033561706543, "learning_rate": 9.120568506400873e-05, "loss": 0.07, "step": 23210 }, { "grad_norm": 0.495263934135437, "learning_rate": 9.119631715449548e-05, "loss": 0.0673, "step": 23220 }, { "grad_norm": 0.5313053727149963, "learning_rate": 9.118694473981493e-05, "loss": 0.0695, "step": 23230 }, { "grad_norm": 0.7208168506622314, "learning_rate": 9.117756782099203e-05, "loss": 0.0843, "step": 23240 }, { "grad_norm": 0.7067644596099854, "learning_rate": 9.11681863990522e-05, "loss": 0.0738, "step": 23250 }, { "grad_norm": 0.9873151183128357, "learning_rate": 9.115880047502142e-05, "loss": 0.0735, "step": 23260 }, { "grad_norm": 0.6790745854377747, "learning_rate": 9.114941004992609e-05, "loss": 0.0883, "step": 23270 }, { "grad_norm": 0.5192537903785706, "learning_rate": 9.114001512479317e-05, "loss": 0.0728, "step": 23280 }, { "grad_norm": 1.0324995517730713, "learning_rate": 9.113061570065003e-05, "loss": 0.0854, "step": 23290 }, { "grad_norm": 0.5493882894515991, "learning_rate": 9.112121177852459e-05, "loss": 0.0684, "step": 23300 }, { "grad_norm": 0.5074159502983093, "learning_rate": 9.111180335944527e-05, "loss": 0.0775, "step": 23310 }, { "grad_norm": 0.6067661643028259, "learning_rate": 9.110239044444093e-05, "loss": 0.0712, "step": 23320 }, { "grad_norm": 0.6258233785629272, "learning_rate": 9.109297303454099e-05, "loss": 0.0786, "step": 23330 }, { "grad_norm": 0.5481348633766174, "learning_rate": 9.108355113077526e-05, "loss": 0.0718, "step": 23340 }, { "grad_norm": 0.6897430419921875, "learning_rate": 9.107412473417419e-05, "loss": 0.0826, "step": 23350 }, { "grad_norm": 0.6348762512207031, "learning_rate": 9.106469384576858e-05, "loss": 0.0709, "step": 23360 }, { "grad_norm": 1.1623107194900513, "learning_rate": 9.105525846658978e-05, "loss": 0.0873, "step": 23370 }, { "grad_norm": 0.6496245265007019, "learning_rate": 9.104581859766965e-05, "loss": 0.0885, "step": 23380 }, { "grad_norm": 0.6211602091789246, "learning_rate": 9.10363742400405e-05, "loss": 0.0804, "step": 23390 }, { "grad_norm": 0.8270382285118103, "learning_rate": 9.102692539473518e-05, "loss": 0.0795, "step": 23400 }, { "grad_norm": 0.7332601547241211, "learning_rate": 9.101747206278697e-05, "loss": 0.0831, "step": 23410 }, { "grad_norm": 0.5240169167518616, "learning_rate": 9.100801424522968e-05, "loss": 0.0705, "step": 23420 }, { "grad_norm": 0.45555126667022705, "learning_rate": 9.099855194309762e-05, "loss": 0.0795, "step": 23430 }, { "grad_norm": 0.5029414296150208, "learning_rate": 9.098908515742554e-05, "loss": 0.0682, "step": 23440 }, { "grad_norm": 0.5374996066093445, "learning_rate": 9.097961388924873e-05, "loss": 0.0795, "step": 23450 }, { "grad_norm": 0.644845187664032, "learning_rate": 9.097013813960298e-05, "loss": 0.0764, "step": 23460 }, { "grad_norm": 1.105367660522461, "learning_rate": 9.09606579095245e-05, "loss": 0.0792, "step": 23470 }, { "grad_norm": 0.5513213872909546, "learning_rate": 9.095117320005008e-05, "loss": 0.0691, "step": 23480 }, { "grad_norm": 1.423638105392456, "learning_rate": 9.094168401221691e-05, "loss": 0.0737, "step": 23490 }, { "grad_norm": 1.104515790939331, "learning_rate": 9.093219034706273e-05, "loss": 0.0939, "step": 23500 }, { "grad_norm": 1.0144938230514526, "learning_rate": 9.092269220562577e-05, "loss": 0.0848, "step": 23510 }, { "grad_norm": 0.5064188241958618, "learning_rate": 9.09131895889447e-05, "loss": 0.0985, "step": 23520 }, { "grad_norm": 0.36493849754333496, "learning_rate": 9.090368249805873e-05, "loss": 0.0628, "step": 23530 }, { "grad_norm": 0.5004419684410095, "learning_rate": 9.089417093400754e-05, "loss": 0.0748, "step": 23540 }, { "grad_norm": 0.4850045144557953, "learning_rate": 9.088465489783131e-05, "loss": 0.0577, "step": 23550 }, { "grad_norm": 0.581127941608429, "learning_rate": 9.087513439057068e-05, "loss": 0.0747, "step": 23560 }, { "grad_norm": 0.5474236011505127, "learning_rate": 9.08656094132668e-05, "loss": 0.0737, "step": 23570 }, { "grad_norm": 0.45168864727020264, "learning_rate": 9.085607996696134e-05, "loss": 0.0701, "step": 23580 }, { "grad_norm": 0.5918739438056946, "learning_rate": 9.084654605269639e-05, "loss": 0.0858, "step": 23590 }, { "grad_norm": 0.45732375979423523, "learning_rate": 9.083700767151457e-05, "loss": 0.0693, "step": 23600 }, { "grad_norm": 0.43579286336898804, "learning_rate": 9.082746482445898e-05, "loss": 0.0621, "step": 23610 }, { "grad_norm": 0.7500902414321899, "learning_rate": 9.081791751257325e-05, "loss": 0.0629, "step": 23620 }, { "grad_norm": 0.5985294580459595, "learning_rate": 9.080836573690142e-05, "loss": 0.0635, "step": 23630 }, { "grad_norm": 0.5816349387168884, "learning_rate": 9.079880949848805e-05, "loss": 0.0785, "step": 23640 }, { "grad_norm": 0.5317317843437195, "learning_rate": 9.078924879837822e-05, "loss": 0.0693, "step": 23650 }, { "grad_norm": 0.6130500435829163, "learning_rate": 9.077968363761747e-05, "loss": 0.0766, "step": 23660 }, { "grad_norm": 1.0549744367599487, "learning_rate": 9.077011401725182e-05, "loss": 0.0597, "step": 23670 }, { "grad_norm": 0.5317274928092957, "learning_rate": 9.07605399383278e-05, "loss": 0.0718, "step": 23680 }, { "grad_norm": 0.545748770236969, "learning_rate": 9.075096140189243e-05, "loss": 0.0738, "step": 23690 }, { "grad_norm": 0.5574052929878235, "learning_rate": 9.074137840899318e-05, "loss": 0.0626, "step": 23700 }, { "grad_norm": 2.4508957862854004, "learning_rate": 9.073179096067804e-05, "loss": 0.0905, "step": 23710 }, { "grad_norm": 0.5995832681655884, "learning_rate": 9.072219905799549e-05, "loss": 0.0792, "step": 23720 }, { "grad_norm": 0.5672137141227722, "learning_rate": 9.071260270199447e-05, "loss": 0.0819, "step": 23730 }, { "grad_norm": 0.5203406810760498, "learning_rate": 9.070300189372441e-05, "loss": 0.0675, "step": 23740 }, { "grad_norm": 0.7676020264625549, "learning_rate": 9.069339663423528e-05, "loss": 0.0695, "step": 23750 }, { "grad_norm": 0.9196402430534363, "learning_rate": 9.068378692457747e-05, "loss": 0.0817, "step": 23760 }, { "grad_norm": 0.46115365624427795, "learning_rate": 9.067417276580189e-05, "loss": 0.0687, "step": 23770 }, { "grad_norm": 0.3948163092136383, "learning_rate": 9.066455415895993e-05, "loss": 0.0627, "step": 23780 }, { "grad_norm": 0.46574240922927856, "learning_rate": 9.065493110510346e-05, "loss": 0.0789, "step": 23790 }, { "grad_norm": 0.5666951537132263, "learning_rate": 9.064530360528484e-05, "loss": 0.0681, "step": 23800 }, { "grad_norm": 0.5036430358886719, "learning_rate": 9.063567166055695e-05, "loss": 0.0631, "step": 23810 }, { "grad_norm": 0.581517219543457, "learning_rate": 9.062603527197308e-05, "loss": 0.0653, "step": 23820 }, { "grad_norm": 0.5158708691596985, "learning_rate": 9.06163944405871e-05, "loss": 0.0629, "step": 23830 }, { "grad_norm": 0.8786114454269409, "learning_rate": 9.060674916745327e-05, "loss": 0.0737, "step": 23840 }, { "grad_norm": 0.419704794883728, "learning_rate": 9.05970994536264e-05, "loss": 0.0636, "step": 23850 }, { "grad_norm": 0.5648536682128906, "learning_rate": 9.05874453001618e-05, "loss": 0.0633, "step": 23860 }, { "grad_norm": 1.0997377634048462, "learning_rate": 9.057778670811517e-05, "loss": 0.0744, "step": 23870 }, { "grad_norm": 0.6611408591270447, "learning_rate": 9.056812367854281e-05, "loss": 0.0712, "step": 23880 }, { "grad_norm": 0.46212804317474365, "learning_rate": 9.055845621250143e-05, "loss": 0.0654, "step": 23890 }, { "grad_norm": 0.4894605875015259, "learning_rate": 9.054878431104825e-05, "loss": 0.0779, "step": 23900 }, { "grad_norm": 0.605091392993927, "learning_rate": 9.0539107975241e-05, "loss": 0.0724, "step": 23910 }, { "grad_norm": 0.7037986516952515, "learning_rate": 9.052942720613784e-05, "loss": 0.07, "step": 23920 }, { "grad_norm": 0.512283444404602, "learning_rate": 9.051974200479745e-05, "loss": 0.0581, "step": 23930 }, { "grad_norm": 0.6182085871696472, "learning_rate": 9.051005237227901e-05, "loss": 0.0531, "step": 23940 }, { "grad_norm": 0.5502643585205078, "learning_rate": 9.050035830964215e-05, "loss": 0.0571, "step": 23950 }, { "grad_norm": 0.6613551378250122, "learning_rate": 9.049065981794698e-05, "loss": 0.0847, "step": 23960 }, { "grad_norm": 0.4305879771709442, "learning_rate": 9.048095689825414e-05, "loss": 0.0674, "step": 23970 }, { "grad_norm": 0.5271011590957642, "learning_rate": 9.047124955162472e-05, "loss": 0.061, "step": 23980 }, { "grad_norm": 0.7692633867263794, "learning_rate": 9.046153777912028e-05, "loss": 0.0733, "step": 23990 }, { "grad_norm": 0.722133994102478, "learning_rate": 9.045182158180292e-05, "loss": 0.0625, "step": 24000 }, { "grad_norm": 0.4976543188095093, "learning_rate": 9.044210096073516e-05, "loss": 0.0585, "step": 24010 }, { "grad_norm": 0.4957445561885834, "learning_rate": 9.043237591698004e-05, "loss": 0.0627, "step": 24020 }, { "grad_norm": 0.423623263835907, "learning_rate": 9.04226464516011e-05, "loss": 0.0571, "step": 24030 }, { "grad_norm": 0.6277263164520264, "learning_rate": 9.041291256566229e-05, "loss": 0.0777, "step": 24040 }, { "grad_norm": 0.5922040343284607, "learning_rate": 9.040317426022814e-05, "loss": 0.0622, "step": 24050 }, { "grad_norm": 0.6711040735244751, "learning_rate": 9.03934315363636e-05, "loss": 0.0778, "step": 24060 }, { "grad_norm": 0.5226536989212036, "learning_rate": 9.038368439513409e-05, "loss": 0.0566, "step": 24070 }, { "grad_norm": 0.6007493734359741, "learning_rate": 9.03739328376056e-05, "loss": 0.0768, "step": 24080 }, { "grad_norm": 0.6695069670677185, "learning_rate": 9.036417686484451e-05, "loss": 0.0644, "step": 24090 }, { "grad_norm": 0.460510790348053, "learning_rate": 9.035441647791773e-05, "loss": 0.0577, "step": 24100 }, { "grad_norm": 0.49411433935165405, "learning_rate": 9.034465167789263e-05, "loss": 0.0712, "step": 24110 }, { "grad_norm": 0.5686188340187073, "learning_rate": 9.033488246583706e-05, "loss": 0.0658, "step": 24120 }, { "grad_norm": 0.465436726808548, "learning_rate": 9.032510884281941e-05, "loss": 0.0743, "step": 24130 }, { "grad_norm": 0.4572250545024872, "learning_rate": 9.031533080990848e-05, "loss": 0.064, "step": 24140 }, { "grad_norm": 2.6119892597198486, "learning_rate": 9.030554836817358e-05, "loss": 0.0712, "step": 24150 }, { "grad_norm": 0.45649319887161255, "learning_rate": 9.029576151868451e-05, "loss": 0.0657, "step": 24160 }, { "grad_norm": 0.3808409571647644, "learning_rate": 9.028597026251155e-05, "loss": 0.0592, "step": 24170 }, { "grad_norm": 0.6978375911712646, "learning_rate": 9.027617460072547e-05, "loss": 0.072, "step": 24180 }, { "grad_norm": 0.5735536813735962, "learning_rate": 9.026637453439745e-05, "loss": 0.0757, "step": 24190 }, { "grad_norm": 0.5134896039962769, "learning_rate": 9.025657006459927e-05, "loss": 0.0584, "step": 24200 }, { "grad_norm": 0.8409344553947449, "learning_rate": 9.024676119240311e-05, "loss": 0.0597, "step": 24210 }, { "grad_norm": 0.44419413805007935, "learning_rate": 9.023694791888166e-05, "loss": 0.0692, "step": 24220 }, { "grad_norm": 0.5164256691932678, "learning_rate": 9.022713024510808e-05, "loss": 0.0675, "step": 24230 }, { "grad_norm": 0.697429358959198, "learning_rate": 9.021730817215601e-05, "loss": 0.0673, "step": 24240 }, { "grad_norm": 0.6662896871566772, "learning_rate": 9.02074817010996e-05, "loss": 0.0802, "step": 24250 }, { "grad_norm": 0.5485411286354065, "learning_rate": 9.019765083301342e-05, "loss": 0.0635, "step": 24260 }, { "grad_norm": 0.4691145718097687, "learning_rate": 9.01878155689726e-05, "loss": 0.0784, "step": 24270 }, { "grad_norm": 0.5419954061508179, "learning_rate": 9.017797591005268e-05, "loss": 0.0601, "step": 24280 }, { "grad_norm": 0.4672199487686157, "learning_rate": 9.016813185732972e-05, "loss": 0.0791, "step": 24290 }, { "grad_norm": 0.5731874108314514, "learning_rate": 9.015828341188027e-05, "loss": 0.0789, "step": 24300 }, { "grad_norm": 0.7229285836219788, "learning_rate": 9.01484305747813e-05, "loss": 0.0702, "step": 24310 }, { "grad_norm": 0.7349870800971985, "learning_rate": 9.013857334711033e-05, "loss": 0.0695, "step": 24320 }, { "grad_norm": 0.5970000624656677, "learning_rate": 9.012871172994534e-05, "loss": 0.0608, "step": 24330 }, { "grad_norm": 0.6332318186759949, "learning_rate": 9.011884572436476e-05, "loss": 0.0746, "step": 24340 }, { "grad_norm": 0.6942461133003235, "learning_rate": 9.010897533144754e-05, "loss": 0.0696, "step": 24350 }, { "grad_norm": 0.698792576789856, "learning_rate": 9.009910055227306e-05, "loss": 0.0663, "step": 24360 }, { "grad_norm": 0.6114485263824463, "learning_rate": 9.008922138792124e-05, "loss": 0.0622, "step": 24370 }, { "grad_norm": 0.45317280292510986, "learning_rate": 9.007933783947244e-05, "loss": 0.0644, "step": 24380 }, { "grad_norm": 0.651989221572876, "learning_rate": 9.006944990800752e-05, "loss": 0.058, "step": 24390 }, { "grad_norm": 1.093383550643921, "learning_rate": 9.005955759460779e-05, "loss": 0.0753, "step": 24400 }, { "grad_norm": 0.6758864521980286, "learning_rate": 9.004966090035508e-05, "loss": 0.0679, "step": 24410 }, { "grad_norm": 0.314484566450119, "learning_rate": 9.003975982633166e-05, "loss": 0.0637, "step": 24420 }, { "grad_norm": 0.5121755599975586, "learning_rate": 9.00298543736203e-05, "loss": 0.0667, "step": 24430 }, { "grad_norm": 0.8029468655586243, "learning_rate": 9.001994454330427e-05, "loss": 0.0621, "step": 24440 }, { "grad_norm": 0.5489897131919861, "learning_rate": 9.001003033646727e-05, "loss": 0.0698, "step": 24450 }, { "grad_norm": 0.6016038656234741, "learning_rate": 9.00001117541935e-05, "loss": 0.0657, "step": 24460 }, { "grad_norm": 0.5445589423179626, "learning_rate": 8.999018879756764e-05, "loss": 0.0724, "step": 24470 }, { "grad_norm": 0.6388660073280334, "learning_rate": 8.998026146767487e-05, "loss": 0.0612, "step": 24480 }, { "grad_norm": 0.8286705613136292, "learning_rate": 8.99703297656008e-05, "loss": 0.076, "step": 24490 }, { "grad_norm": 0.5032833218574524, "learning_rate": 8.996039369243156e-05, "loss": 0.0719, "step": 24500 }, { "grad_norm": 0.779630184173584, "learning_rate": 8.995045324925378e-05, "loss": 0.0584, "step": 24510 }, { "grad_norm": 0.5298855900764465, "learning_rate": 8.994050843715448e-05, "loss": 0.071, "step": 24520 }, { "grad_norm": 0.6800863742828369, "learning_rate": 8.993055925722121e-05, "loss": 0.0657, "step": 24530 }, { "grad_norm": 0.5007458329200745, "learning_rate": 8.992060571054202e-05, "loss": 0.0758, "step": 24540 }, { "grad_norm": 0.5497590899467468, "learning_rate": 8.991064779820542e-05, "loss": 0.0748, "step": 24550 }, { "grad_norm": 0.4176234006881714, "learning_rate": 8.990068552130036e-05, "loss": 0.0726, "step": 24560 }, { "grad_norm": 0.6225422024726868, "learning_rate": 8.989071888091634e-05, "loss": 0.0683, "step": 24570 }, { "grad_norm": 0.527533233165741, "learning_rate": 8.988074787814329e-05, "loss": 0.0605, "step": 24580 }, { "grad_norm": 0.7465860247612, "learning_rate": 8.987077251407158e-05, "loss": 0.069, "step": 24590 }, { "grad_norm": 0.40151575207710266, "learning_rate": 8.986079278979216e-05, "loss": 0.066, "step": 24600 }, { "grad_norm": 0.5938043594360352, "learning_rate": 8.985080870639635e-05, "loss": 0.0781, "step": 24610 }, { "grad_norm": 0.4887818396091461, "learning_rate": 8.984082026497603e-05, "loss": 0.0569, "step": 24620 }, { "grad_norm": 0.6190551519393921, "learning_rate": 8.98308274666235e-05, "loss": 0.0778, "step": 24630 }, { "grad_norm": 0.45200976729393005, "learning_rate": 8.982083031243155e-05, "loss": 0.0646, "step": 24640 }, { "grad_norm": 0.6943804621696472, "learning_rate": 8.98108288034935e-05, "loss": 0.064, "step": 24650 }, { "grad_norm": 0.543048083782196, "learning_rate": 8.980082294090305e-05, "loss": 0.0685, "step": 24660 }, { "grad_norm": 0.6098179817199707, "learning_rate": 8.979081272575443e-05, "loss": 0.0792, "step": 24670 }, { "grad_norm": 0.8349655270576477, "learning_rate": 8.978079815914236e-05, "loss": 0.0772, "step": 24680 }, { "grad_norm": 0.7772045731544495, "learning_rate": 8.977077924216202e-05, "loss": 0.0791, "step": 24690 }, { "grad_norm": 0.9242666959762573, "learning_rate": 8.976075597590905e-05, "loss": 0.0769, "step": 24700 }, { "grad_norm": 0.5412113666534424, "learning_rate": 8.975072836147958e-05, "loss": 0.0631, "step": 24710 }, { "grad_norm": 0.7152572274208069, "learning_rate": 8.974069639997025e-05, "loss": 0.067, "step": 24720 }, { "grad_norm": 0.6172972321510315, "learning_rate": 8.973066009247808e-05, "loss": 0.0786, "step": 24730 }, { "grad_norm": 0.5582907199859619, "learning_rate": 8.972061944010066e-05, "loss": 0.0799, "step": 24740 }, { "grad_norm": 0.5956586599349976, "learning_rate": 8.971057444393603e-05, "loss": 0.074, "step": 24750 }, { "grad_norm": 0.45348823070526123, "learning_rate": 8.970052510508268e-05, "loss": 0.0714, "step": 24760 }, { "grad_norm": 0.5566409826278687, "learning_rate": 8.969047142463959e-05, "loss": 0.0746, "step": 24770 }, { "grad_norm": 0.611997127532959, "learning_rate": 8.968041340370621e-05, "loss": 0.0666, "step": 24780 }, { "grad_norm": 0.4434497654438019, "learning_rate": 8.96703510433825e-05, "loss": 0.0706, "step": 24790 }, { "grad_norm": 0.9101011157035828, "learning_rate": 8.966028434476883e-05, "loss": 0.0652, "step": 24800 }, { "grad_norm": 0.5029658079147339, "learning_rate": 8.96502133089661e-05, "loss": 0.0605, "step": 24810 }, { "grad_norm": 0.7312551736831665, "learning_rate": 8.964013793707564e-05, "loss": 0.0718, "step": 24820 }, { "grad_norm": 0.6117269396781921, "learning_rate": 8.963005823019932e-05, "loss": 0.0764, "step": 24830 }, { "grad_norm": 0.5315427780151367, "learning_rate": 8.961997418943939e-05, "loss": 0.0612, "step": 24840 }, { "grad_norm": 0.5407394170761108, "learning_rate": 8.960988581589865e-05, "loss": 0.0728, "step": 24850 }, { "grad_norm": 0.7049873471260071, "learning_rate": 8.959979311068037e-05, "loss": 0.0602, "step": 24860 }, { "grad_norm": 0.5013410449028015, "learning_rate": 8.958969607488823e-05, "loss": 0.0848, "step": 24870 }, { "grad_norm": 1.3550091981887817, "learning_rate": 8.957959470962647e-05, "loss": 0.0587, "step": 24880 }, { "grad_norm": 0.4855756163597107, "learning_rate": 8.956948901599971e-05, "loss": 0.066, "step": 24890 }, { "grad_norm": 0.4805621802806854, "learning_rate": 8.955937899511315e-05, "loss": 0.0697, "step": 24900 }, { "grad_norm": 0.6143919825553894, "learning_rate": 8.954926464807238e-05, "loss": 0.0676, "step": 24910 }, { "grad_norm": 0.3808324635028839, "learning_rate": 8.953914597598347e-05, "loss": 0.0658, "step": 24920 }, { "grad_norm": 0.5890918374061584, "learning_rate": 8.952902297995303e-05, "loss": 0.0578, "step": 24930 }, { "grad_norm": 0.46791714429855347, "learning_rate": 8.951889566108804e-05, "loss": 0.0718, "step": 24940 }, { "grad_norm": 0.7666253447532654, "learning_rate": 8.950876402049606e-05, "loss": 0.0724, "step": 24950 }, { "grad_norm": 0.7704698443412781, "learning_rate": 8.949862805928504e-05, "loss": 0.058, "step": 24960 }, { "grad_norm": 0.4658684730529785, "learning_rate": 8.948848777856343e-05, "loss": 0.0745, "step": 24970 }, { "grad_norm": 1.6131972074508667, "learning_rate": 8.947834317944017e-05, "loss": 0.0723, "step": 24980 }, { "grad_norm": 0.5215590000152588, "learning_rate": 8.946819426302466e-05, "loss": 0.0756, "step": 24990 }, { "grad_norm": 0.5662876963615417, "learning_rate": 8.945804103042676e-05, "loss": 0.0582, "step": 25000 }, { "grad_norm": 0.8987593650817871, "learning_rate": 8.944788348275681e-05, "loss": 0.0615, "step": 25010 }, { "grad_norm": 0.6642352342605591, "learning_rate": 8.943772162112565e-05, "loss": 0.0759, "step": 25020 }, { "grad_norm": 0.5936944484710693, "learning_rate": 8.942755544664454e-05, "loss": 0.0661, "step": 25030 }, { "grad_norm": 0.6854842901229858, "learning_rate": 8.941738496042525e-05, "loss": 0.0739, "step": 25040 }, { "grad_norm": 0.4564007520675659, "learning_rate": 8.940721016357999e-05, "loss": 0.0753, "step": 25050 }, { "grad_norm": 0.5846689343452454, "learning_rate": 8.939703105722148e-05, "loss": 0.0793, "step": 25060 }, { "grad_norm": 0.5546371936798096, "learning_rate": 8.93868476424629e-05, "loss": 0.0639, "step": 25070 }, { "grad_norm": 0.758438229560852, "learning_rate": 8.937665992041786e-05, "loss": 0.0723, "step": 25080 }, { "grad_norm": 1.0552650690078735, "learning_rate": 8.93664678922005e-05, "loss": 0.0748, "step": 25090 }, { "grad_norm": 0.8728475570678711, "learning_rate": 8.93562715589254e-05, "loss": 0.0671, "step": 25100 }, { "grad_norm": 0.457790732383728, "learning_rate": 8.934607092170762e-05, "loss": 0.067, "step": 25110 }, { "grad_norm": 0.5213972926139832, "learning_rate": 8.933586598166266e-05, "loss": 0.073, "step": 25120 }, { "grad_norm": 0.5621369481086731, "learning_rate": 8.932565673990655e-05, "loss": 0.0873, "step": 25130 }, { "grad_norm": 0.5174733400344849, "learning_rate": 8.931544319755574e-05, "loss": 0.0663, "step": 25140 }, { "grad_norm": 0.41729429364204407, "learning_rate": 8.930522535572718e-05, "loss": 0.0656, "step": 25150 }, { "grad_norm": 0.3914867639541626, "learning_rate": 8.929500321553826e-05, "loss": 0.0628, "step": 25160 }, { "grad_norm": 0.5237804651260376, "learning_rate": 8.928477677810686e-05, "loss": 0.0729, "step": 25170 }, { "grad_norm": 0.460306853055954, "learning_rate": 8.927454604455137e-05, "loss": 0.0719, "step": 25180 }, { "grad_norm": 0.6047822833061218, "learning_rate": 8.926431101599053e-05, "loss": 0.066, "step": 25190 }, { "grad_norm": 0.6208035945892334, "learning_rate": 8.925407169354369e-05, "loss": 0.0639, "step": 25200 }, { "grad_norm": 0.4665940999984741, "learning_rate": 8.92438280783306e-05, "loss": 0.0834, "step": 25210 }, { "grad_norm": 0.637924313545227, "learning_rate": 8.923358017147146e-05, "loss": 0.0886, "step": 25220 }, { "grad_norm": 0.4978000223636627, "learning_rate": 8.922332797408697e-05, "loss": 0.0787, "step": 25230 }, { "grad_norm": 0.40654680132865906, "learning_rate": 8.921307148729831e-05, "loss": 0.0572, "step": 25240 }, { "grad_norm": 0.4747518301010132, "learning_rate": 8.920281071222712e-05, "loss": 0.0634, "step": 25250 }, { "grad_norm": 1.4377286434173584, "learning_rate": 8.919254564999548e-05, "loss": 0.0544, "step": 25260 }, { "grad_norm": 0.39978957176208496, "learning_rate": 8.918227630172598e-05, "loss": 0.0599, "step": 25270 }, { "grad_norm": 0.604842483997345, "learning_rate": 8.917200266854165e-05, "loss": 0.0669, "step": 25280 }, { "grad_norm": 0.44223153591156006, "learning_rate": 8.9161724751566e-05, "loss": 0.0629, "step": 25290 }, { "grad_norm": 0.42662203311920166, "learning_rate": 8.915144255192302e-05, "loss": 0.0712, "step": 25300 }, { "grad_norm": 0.9364541172981262, "learning_rate": 8.914115607073714e-05, "loss": 0.0833, "step": 25310 }, { "grad_norm": 0.9895086884498596, "learning_rate": 8.913086530913327e-05, "loss": 0.0649, "step": 25320 }, { "grad_norm": 0.372599720954895, "learning_rate": 8.912057026823681e-05, "loss": 0.0722, "step": 25330 }, { "grad_norm": 0.5117264986038208, "learning_rate": 8.91102709491736e-05, "loss": 0.0738, "step": 25340 }, { "grad_norm": 0.5101721882820129, "learning_rate": 8.909996735306996e-05, "loss": 0.0657, "step": 25350 }, { "grad_norm": 0.6271215677261353, "learning_rate": 8.908965948105268e-05, "loss": 0.0673, "step": 25360 }, { "grad_norm": 0.6410520672798157, "learning_rate": 8.907934733424901e-05, "loss": 0.0651, "step": 25370 }, { "grad_norm": 0.41420307755470276, "learning_rate": 8.906903091378666e-05, "loss": 0.0667, "step": 25380 }, { "grad_norm": 0.4421137869358063, "learning_rate": 8.905871022079384e-05, "loss": 0.0686, "step": 25390 }, { "grad_norm": 0.47082051634788513, "learning_rate": 8.90483852563992e-05, "loss": 0.0745, "step": 25400 }, { "grad_norm": 0.5502517819404602, "learning_rate": 8.903805602173185e-05, "loss": 0.0664, "step": 25410 }, { "grad_norm": 0.4537426829338074, "learning_rate": 8.902772251792137e-05, "loss": 0.0592, "step": 25420 }, { "grad_norm": 0.5530076026916504, "learning_rate": 8.901738474609786e-05, "loss": 0.0719, "step": 25430 }, { "grad_norm": 0.4996613562107086, "learning_rate": 8.900704270739179e-05, "loss": 0.0683, "step": 25440 }, { "grad_norm": 0.4755780100822449, "learning_rate": 8.89966964029342e-05, "loss": 0.073, "step": 25450 }, { "grad_norm": 0.3899479806423187, "learning_rate": 8.898634583385652e-05, "loss": 0.047, "step": 25460 }, { "grad_norm": 0.5533406138420105, "learning_rate": 8.897599100129065e-05, "loss": 0.0796, "step": 25470 }, { "grad_norm": 0.46623310446739197, "learning_rate": 8.896563190636903e-05, "loss": 0.0593, "step": 25480 }, { "grad_norm": 0.36206570267677307, "learning_rate": 8.895526855022448e-05, "loss": 0.0672, "step": 25490 }, { "grad_norm": 0.42892155051231384, "learning_rate": 8.894490093399033e-05, "loss": 0.0722, "step": 25500 }, { "grad_norm": 0.35974064469337463, "learning_rate": 8.893452905880035e-05, "loss": 0.0699, "step": 25510 }, { "grad_norm": 0.49980881810188293, "learning_rate": 8.892415292578883e-05, "loss": 0.0686, "step": 25520 }, { "grad_norm": 0.5105143189430237, "learning_rate": 8.891377253609046e-05, "loss": 0.0633, "step": 25530 }, { "grad_norm": 0.5278148055076599, "learning_rate": 8.890338789084043e-05, "loss": 0.0642, "step": 25540 }, { "grad_norm": 0.7507418990135193, "learning_rate": 8.88929989911744e-05, "loss": 0.0632, "step": 25550 }, { "grad_norm": 0.6787919402122498, "learning_rate": 8.888260583822847e-05, "loss": 0.0716, "step": 25560 }, { "grad_norm": 0.542240560054779, "learning_rate": 8.887220843313921e-05, "loss": 0.067, "step": 25570 }, { "grad_norm": 0.6381548047065735, "learning_rate": 8.88618067770437e-05, "loss": 0.069, "step": 25580 }, { "grad_norm": 0.9170910716056824, "learning_rate": 8.885140087107942e-05, "loss": 0.0789, "step": 25590 }, { "grad_norm": 0.4486556053161621, "learning_rate": 8.884099071638436e-05, "loss": 0.0618, "step": 25600 }, { "grad_norm": 0.6018874049186707, "learning_rate": 8.883057631409695e-05, "loss": 0.0902, "step": 25610 }, { "grad_norm": 0.6468281149864197, "learning_rate": 8.882015766535608e-05, "loss": 0.0709, "step": 25620 }, { "grad_norm": 0.6377519965171814, "learning_rate": 8.880973477130115e-05, "loss": 0.0725, "step": 25630 }, { "grad_norm": 0.6092180609703064, "learning_rate": 8.879930763307197e-05, "loss": 0.0596, "step": 25640 }, { "grad_norm": 0.5847712159156799, "learning_rate": 8.878887625180884e-05, "loss": 0.0766, "step": 25650 }, { "grad_norm": 0.5324114561080933, "learning_rate": 8.877844062865253e-05, "loss": 0.0584, "step": 25660 }, { "grad_norm": 0.5401694774627686, "learning_rate": 8.876800076474424e-05, "loss": 0.0626, "step": 25670 }, { "grad_norm": 0.44359973073005676, "learning_rate": 8.875755666122568e-05, "loss": 0.055, "step": 25680 }, { "grad_norm": 0.388741672039032, "learning_rate": 8.8747108319239e-05, "loss": 0.0567, "step": 25690 }, { "grad_norm": 0.5134117007255554, "learning_rate": 8.87366557399268e-05, "loss": 0.0585, "step": 25700 }, { "grad_norm": 0.611312985420227, "learning_rate": 8.872619892443217e-05, "loss": 0.0727, "step": 25710 }, { "grad_norm": 0.47849541902542114, "learning_rate": 8.871573787389865e-05, "loss": 0.0553, "step": 25720 }, { "grad_norm": 0.6327455639839172, "learning_rate": 8.870527258947024e-05, "loss": 0.0697, "step": 25730 }, { "grad_norm": 0.47683221101760864, "learning_rate": 8.869480307229143e-05, "loss": 0.0495, "step": 25740 }, { "grad_norm": 0.6958595514297485, "learning_rate": 8.868432932350712e-05, "loss": 0.0624, "step": 25750 }, { "grad_norm": 0.5456107258796692, "learning_rate": 8.867385134426272e-05, "loss": 0.0713, "step": 25760 }, { "grad_norm": 0.5020712614059448, "learning_rate": 8.866336913570407e-05, "loss": 0.0601, "step": 25770 }, { "grad_norm": 0.3760659992694855, "learning_rate": 8.865288269897751e-05, "loss": 0.0548, "step": 25780 }, { "grad_norm": 0.6388717889785767, "learning_rate": 8.864239203522981e-05, "loss": 0.0594, "step": 25790 }, { "grad_norm": 0.4319595992565155, "learning_rate": 8.863189714560822e-05, "loss": 0.0656, "step": 25800 }, { "grad_norm": 0.4408852159976959, "learning_rate": 8.862139803126043e-05, "loss": 0.0575, "step": 25810 }, { "grad_norm": 0.6461766362190247, "learning_rate": 8.861089469333463e-05, "loss": 0.0662, "step": 25820 }, { "grad_norm": 0.5416937470436096, "learning_rate": 8.860038713297944e-05, "loss": 0.0525, "step": 25830 }, { "grad_norm": 0.5785108804702759, "learning_rate": 8.858987535134394e-05, "loss": 0.0668, "step": 25840 }, { "grad_norm": 0.5446697473526001, "learning_rate": 8.857935934957769e-05, "loss": 0.0589, "step": 25850 }, { "grad_norm": 0.5465359091758728, "learning_rate": 8.856883912883071e-05, "loss": 0.0572, "step": 25860 }, { "grad_norm": 0.43676647543907166, "learning_rate": 8.855831469025346e-05, "loss": 0.0547, "step": 25870 }, { "grad_norm": 0.5023271441459656, "learning_rate": 8.854778603499689e-05, "loss": 0.0823, "step": 25880 }, { "grad_norm": 0.9399457573890686, "learning_rate": 8.85372531642124e-05, "loss": 0.0625, "step": 25890 }, { "grad_norm": 0.4619523584842682, "learning_rate": 8.852671607905185e-05, "loss": 0.0615, "step": 25900 }, { "grad_norm": 0.6331958770751953, "learning_rate": 8.851617478066754e-05, "loss": 0.0741, "step": 25910 }, { "grad_norm": 0.5399779081344604, "learning_rate": 8.850562927021227e-05, "loss": 0.0683, "step": 25920 }, { "grad_norm": 0.5717593431472778, "learning_rate": 8.849507954883928e-05, "loss": 0.0642, "step": 25930 }, { "grad_norm": 0.600711464881897, "learning_rate": 8.848452561770226e-05, "loss": 0.0646, "step": 25940 }, { "grad_norm": 0.750355064868927, "learning_rate": 8.847396747795538e-05, "loss": 0.0668, "step": 25950 }, { "grad_norm": 0.6541047692298889, "learning_rate": 8.846340513075327e-05, "loss": 0.068, "step": 25960 }, { "grad_norm": 0.5098267793655396, "learning_rate": 8.845283857725099e-05, "loss": 0.0769, "step": 25970 }, { "grad_norm": 0.565790593624115, "learning_rate": 8.844226781860409e-05, "loss": 0.0685, "step": 25980 }, { "grad_norm": 0.9189888834953308, "learning_rate": 8.84316928559686e-05, "loss": 0.0845, "step": 25990 }, { "grad_norm": 0.8858381509780884, "learning_rate": 8.842111369050094e-05, "loss": 0.085, "step": 26000 }, { "grad_norm": 0.515421450138092, "learning_rate": 8.841053032335808e-05, "loss": 0.0717, "step": 26010 }, { "grad_norm": 0.5023024678230286, "learning_rate": 8.839994275569735e-05, "loss": 0.0735, "step": 26020 }, { "grad_norm": 0.6162002086639404, "learning_rate": 8.838935098867662e-05, "loss": 0.0688, "step": 26030 }, { "grad_norm": 0.6111297607421875, "learning_rate": 8.837875502345418e-05, "loss": 0.079, "step": 26040 }, { "grad_norm": 0.523127555847168, "learning_rate": 8.83681548611888e-05, "loss": 0.0721, "step": 26050 }, { "grad_norm": 0.6623544096946716, "learning_rate": 8.835755050303969e-05, "loss": 0.0625, "step": 26060 }, { "grad_norm": 0.8751317858695984, "learning_rate": 8.834694195016653e-05, "loss": 0.0723, "step": 26070 }, { "grad_norm": 0.7277916669845581, "learning_rate": 8.833632920372942e-05, "loss": 0.0866, "step": 26080 }, { "grad_norm": 0.6237527132034302, "learning_rate": 8.832571226488903e-05, "loss": 0.0704, "step": 26090 }, { "grad_norm": 0.601581871509552, "learning_rate": 8.831509113480634e-05, "loss": 0.0657, "step": 26100 }, { "grad_norm": 0.7650189399719238, "learning_rate": 8.83044658146429e-05, "loss": 0.0813, "step": 26110 }, { "grad_norm": 0.4631790220737457, "learning_rate": 8.829383630556067e-05, "loss": 0.0572, "step": 26120 }, { "grad_norm": 0.7780558466911316, "learning_rate": 8.828320260872207e-05, "loss": 0.0718, "step": 26130 }, { "grad_norm": 0.5369163155555725, "learning_rate": 8.827256472529e-05, "loss": 0.0655, "step": 26140 }, { "grad_norm": 0.5557616353034973, "learning_rate": 8.826192265642778e-05, "loss": 0.0687, "step": 26150 }, { "grad_norm": 0.4562006890773773, "learning_rate": 8.825127640329923e-05, "loss": 0.0675, "step": 26160 }, { "grad_norm": 1.2229535579681396, "learning_rate": 8.824062596706861e-05, "loss": 0.0744, "step": 26170 }, { "grad_norm": 0.5006884932518005, "learning_rate": 8.822997134890062e-05, "loss": 0.0769, "step": 26180 }, { "grad_norm": 0.760069727897644, "learning_rate": 8.821931254996044e-05, "loss": 0.0646, "step": 26190 }, { "grad_norm": 0.4525846540927887, "learning_rate": 8.82086495714137e-05, "loss": 0.0746, "step": 26200 }, { "grad_norm": 0.6977397203445435, "learning_rate": 8.81979824144265e-05, "loss": 0.0791, "step": 26210 }, { "grad_norm": 0.49494150280952454, "learning_rate": 8.818731108016536e-05, "loss": 0.0682, "step": 26220 }, { "grad_norm": 0.5927958488464355, "learning_rate": 8.81766355697973e-05, "loss": 0.076, "step": 26230 }, { "grad_norm": 0.7106823921203613, "learning_rate": 8.816595588448977e-05, "loss": 0.0783, "step": 26240 }, { "grad_norm": 0.6694095134735107, "learning_rate": 8.81552720254107e-05, "loss": 0.073, "step": 26250 }, { "grad_norm": 0.5495017766952515, "learning_rate": 8.814458399372842e-05, "loss": 0.0868, "step": 26260 }, { "grad_norm": 0.7097170948982239, "learning_rate": 8.813389179061181e-05, "loss": 0.0695, "step": 26270 }, { "grad_norm": 0.7064859867095947, "learning_rate": 8.812319541723012e-05, "loss": 0.0863, "step": 26280 }, { "grad_norm": 0.7897859215736389, "learning_rate": 8.811249487475309e-05, "loss": 0.0862, "step": 26290 }, { "grad_norm": 0.5338631272315979, "learning_rate": 8.810179016435092e-05, "loss": 0.0782, "step": 26300 }, { "grad_norm": 0.6450602412223816, "learning_rate": 8.809108128719428e-05, "loss": 0.09, "step": 26310 }, { "grad_norm": 0.6575629711151123, "learning_rate": 8.808036824445424e-05, "loss": 0.0705, "step": 26320 }, { "grad_norm": 0.48225489258766174, "learning_rate": 8.806965103730238e-05, "loss": 0.077, "step": 26330 }, { "grad_norm": 0.606593906879425, "learning_rate": 8.805892966691074e-05, "loss": 0.066, "step": 26340 }, { "grad_norm": 0.5624226927757263, "learning_rate": 8.804820413445175e-05, "loss": 0.065, "step": 26350 }, { "grad_norm": 0.39571887254714966, "learning_rate": 8.803747444109837e-05, "loss": 0.0809, "step": 26360 }, { "grad_norm": 0.5555494427680969, "learning_rate": 8.802674058802399e-05, "loss": 0.0551, "step": 26370 }, { "grad_norm": 0.6282844543457031, "learning_rate": 8.801600257640241e-05, "loss": 0.0756, "step": 26380 }, { "grad_norm": 0.5708553194999695, "learning_rate": 8.800526040740795e-05, "loss": 0.0874, "step": 26390 }, { "grad_norm": 0.6517091393470764, "learning_rate": 8.799451408221535e-05, "loss": 0.0687, "step": 26400 }, { "grad_norm": 0.6836376786231995, "learning_rate": 8.798376360199982e-05, "loss": 0.0728, "step": 26410 }, { "grad_norm": 0.4254152774810791, "learning_rate": 8.797300896793701e-05, "loss": 0.0624, "step": 26420 }, { "grad_norm": 0.7134705185890198, "learning_rate": 8.796225018120302e-05, "loss": 0.0721, "step": 26430 }, { "grad_norm": 0.5198557376861572, "learning_rate": 8.795148724297444e-05, "loss": 0.0657, "step": 26440 }, { "grad_norm": 0.9311607480049133, "learning_rate": 8.794072015442825e-05, "loss": 0.0746, "step": 26450 }, { "grad_norm": 0.7182212471961975, "learning_rate": 8.792994891674198e-05, "loss": 0.0683, "step": 26460 }, { "grad_norm": 0.5101274251937866, "learning_rate": 8.79191735310935e-05, "loss": 0.0688, "step": 26470 }, { "grad_norm": 0.4178113341331482, "learning_rate": 8.790839399866122e-05, "loss": 0.0583, "step": 26480 }, { "grad_norm": 0.7476213574409485, "learning_rate": 8.789761032062397e-05, "loss": 0.0625, "step": 26490 }, { "grad_norm": 0.45838814973831177, "learning_rate": 8.788682249816103e-05, "loss": 0.0675, "step": 26500 }, { "grad_norm": 0.640048086643219, "learning_rate": 8.787603053245215e-05, "loss": 0.0898, "step": 26510 }, { "grad_norm": 0.5269314646720886, "learning_rate": 8.78652344246775e-05, "loss": 0.072, "step": 26520 }, { "grad_norm": 0.5073434710502625, "learning_rate": 8.785443417601776e-05, "loss": 0.0669, "step": 26530 }, { "grad_norm": 0.4695936143398285, "learning_rate": 8.784362978765401e-05, "loss": 0.0551, "step": 26540 }, { "grad_norm": 0.5236226320266724, "learning_rate": 8.783282126076779e-05, "loss": 0.0632, "step": 26550 }, { "grad_norm": 0.5389395356178284, "learning_rate": 8.782200859654112e-05, "loss": 0.0653, "step": 26560 }, { "grad_norm": 0.5536391139030457, "learning_rate": 8.781119179615646e-05, "loss": 0.0636, "step": 26570 }, { "grad_norm": 0.6311397552490234, "learning_rate": 8.780037086079674e-05, "loss": 0.0691, "step": 26580 }, { "grad_norm": 0.49087703227996826, "learning_rate": 8.778954579164527e-05, "loss": 0.0676, "step": 26590 }, { "grad_norm": 0.5869646072387695, "learning_rate": 8.777871658988588e-05, "loss": 0.0755, "step": 26600 }, { "grad_norm": 0.583307147026062, "learning_rate": 8.776788325670285e-05, "loss": 0.0658, "step": 26610 }, { "grad_norm": 0.4879336357116699, "learning_rate": 8.775704579328089e-05, "loss": 0.0689, "step": 26620 }, { "grad_norm": 0.5510200262069702, "learning_rate": 8.774620420080517e-05, "loss": 0.063, "step": 26630 }, { "grad_norm": 0.4697266221046448, "learning_rate": 8.773535848046131e-05, "loss": 0.0535, "step": 26640 }, { "grad_norm": 0.562708854675293, "learning_rate": 8.772450863343538e-05, "loss": 0.0696, "step": 26650 }, { "grad_norm": 0.7537121176719666, "learning_rate": 8.77136546609139e-05, "loss": 0.0783, "step": 26660 }, { "grad_norm": 0.8901656866073608, "learning_rate": 8.770279656408385e-05, "loss": 0.0577, "step": 26670 }, { "grad_norm": 0.5889557003974915, "learning_rate": 8.769193434413265e-05, "loss": 0.0628, "step": 26680 }, { "grad_norm": 0.6261551380157471, "learning_rate": 8.76810680022482e-05, "loss": 0.0746, "step": 26690 }, { "grad_norm": 0.51272052526474, "learning_rate": 8.767019753961878e-05, "loss": 0.0775, "step": 26700 }, { "grad_norm": 0.7813248634338379, "learning_rate": 8.765932295743321e-05, "loss": 0.0655, "step": 26710 }, { "grad_norm": 0.4213522970676422, "learning_rate": 8.764844425688068e-05, "loss": 0.0736, "step": 26720 }, { "grad_norm": 0.517858624458313, "learning_rate": 8.763756143915092e-05, "loss": 0.0615, "step": 26730 }, { "grad_norm": 0.6090403199195862, "learning_rate": 8.7626674505434e-05, "loss": 0.0771, "step": 26740 }, { "grad_norm": 0.4919608235359192, "learning_rate": 8.761578345692053e-05, "loss": 0.0827, "step": 26750 }, { "grad_norm": 0.5778172016143799, "learning_rate": 8.760488829480156e-05, "loss": 0.0693, "step": 26760 }, { "grad_norm": 0.5274073481559753, "learning_rate": 8.759398902026854e-05, "loss": 0.059, "step": 26770 }, { "grad_norm": 0.44478094577789307, "learning_rate": 8.758308563451339e-05, "loss": 0.0756, "step": 26780 }, { "grad_norm": 0.6838617920875549, "learning_rate": 8.75721781387285e-05, "loss": 0.0709, "step": 26790 }, { "grad_norm": 0.43151992559432983, "learning_rate": 8.75612665341067e-05, "loss": 0.0625, "step": 26800 }, { "grad_norm": 0.6017817258834839, "learning_rate": 8.755035082184126e-05, "loss": 0.0644, "step": 26810 }, { "grad_norm": 0.4250005781650543, "learning_rate": 8.753943100312592e-05, "loss": 0.0763, "step": 26820 }, { "grad_norm": 0.5267670154571533, "learning_rate": 8.752850707915484e-05, "loss": 0.0549, "step": 26830 }, { "grad_norm": 0.6640056371688843, "learning_rate": 8.751757905112264e-05, "loss": 0.0847, "step": 26840 }, { "grad_norm": 0.6545857787132263, "learning_rate": 8.75066469202244e-05, "loss": 0.0752, "step": 26850 }, { "grad_norm": 0.46465864777565, "learning_rate": 8.749571068765567e-05, "loss": 0.0591, "step": 26860 }, { "grad_norm": 0.4215736389160156, "learning_rate": 8.748477035461238e-05, "loss": 0.0609, "step": 26870 }, { "grad_norm": 0.48093289136886597, "learning_rate": 8.747382592229095e-05, "loss": 0.0672, "step": 26880 }, { "grad_norm": 0.5331687927246094, "learning_rate": 8.746287739188828e-05, "loss": 0.0808, "step": 26890 }, { "grad_norm": 0.5533828139305115, "learning_rate": 8.745192476460165e-05, "loss": 0.0597, "step": 26900 }, { "grad_norm": 0.5923279523849487, "learning_rate": 8.744096804162882e-05, "loss": 0.0765, "step": 26910 }, { "grad_norm": 0.9099106192588806, "learning_rate": 8.743000722416804e-05, "loss": 0.071, "step": 26920 }, { "grad_norm": 0.5571293234825134, "learning_rate": 8.741904231341793e-05, "loss": 0.0667, "step": 26930 }, { "grad_norm": 0.77666836977005, "learning_rate": 8.740807331057762e-05, "loss": 0.0791, "step": 26940 }, { "grad_norm": 0.601661741733551, "learning_rate": 8.739710021684667e-05, "loss": 0.0714, "step": 26950 }, { "grad_norm": 0.5246697068214417, "learning_rate": 8.738612303342503e-05, "loss": 0.0674, "step": 26960 }, { "grad_norm": 0.6122221946716309, "learning_rate": 8.73751417615132e-05, "loss": 0.0588, "step": 26970 }, { "grad_norm": 0.3428884744644165, "learning_rate": 8.736415640231208e-05, "loss": 0.0517, "step": 26980 }, { "grad_norm": 0.44167184829711914, "learning_rate": 8.735316695702297e-05, "loss": 0.0562, "step": 26990 }, { "grad_norm": 0.4104129672050476, "learning_rate": 8.734217342684769e-05, "loss": 0.0566, "step": 27000 }, { "grad_norm": 0.3708201050758362, "learning_rate": 8.733117581298847e-05, "loss": 0.0566, "step": 27010 }, { "grad_norm": 0.5330785512924194, "learning_rate": 8.732017411664796e-05, "loss": 0.0643, "step": 27020 }, { "grad_norm": 1.1658421754837036, "learning_rate": 8.730916833902936e-05, "loss": 0.0599, "step": 27030 }, { "grad_norm": 0.3786047101020813, "learning_rate": 8.729815848133618e-05, "loss": 0.0784, "step": 27040 }, { "grad_norm": 0.44661879539489746, "learning_rate": 8.728714454477247e-05, "loss": 0.0551, "step": 27050 }, { "grad_norm": 0.44362854957580566, "learning_rate": 8.727612653054269e-05, "loss": 0.0631, "step": 27060 }, { "grad_norm": 0.44777148962020874, "learning_rate": 8.726510443985176e-05, "loss": 0.0551, "step": 27070 }, { "grad_norm": 0.5718213319778442, "learning_rate": 8.725407827390503e-05, "loss": 0.0668, "step": 27080 }, { "grad_norm": 0.5852091312408447, "learning_rate": 8.724304803390833e-05, "loss": 0.0648, "step": 27090 }, { "grad_norm": 0.3914780914783478, "learning_rate": 8.723201372106788e-05, "loss": 0.0536, "step": 27100 }, { "grad_norm": 0.7082642912864685, "learning_rate": 8.722097533659038e-05, "loss": 0.068, "step": 27110 }, { "grad_norm": 0.5388979911804199, "learning_rate": 8.720993288168299e-05, "loss": 0.0653, "step": 27120 }, { "grad_norm": 0.6334106922149658, "learning_rate": 8.719888635755327e-05, "loss": 0.0589, "step": 27130 }, { "grad_norm": 0.453548789024353, "learning_rate": 8.718783576540928e-05, "loss": 0.066, "step": 27140 }, { "grad_norm": 0.5412560701370239, "learning_rate": 8.717678110645948e-05, "loss": 0.0784, "step": 27150 }, { "grad_norm": 0.43834802508354187, "learning_rate": 8.716572238191279e-05, "loss": 0.0619, "step": 27160 }, { "grad_norm": 0.5005666613578796, "learning_rate": 8.715465959297857e-05, "loss": 0.0733, "step": 27170 }, { "grad_norm": 0.6012097001075745, "learning_rate": 8.714359274086665e-05, "loss": 0.0756, "step": 27180 }, { "grad_norm": 0.5272225737571716, "learning_rate": 8.713252182678726e-05, "loss": 0.0642, "step": 27190 }, { "grad_norm": 0.5659247636795044, "learning_rate": 8.712144685195112e-05, "loss": 0.0837, "step": 27200 }, { "grad_norm": 0.7386958599090576, "learning_rate": 8.711036781756936e-05, "loss": 0.0644, "step": 27210 }, { "grad_norm": 1.0024718046188354, "learning_rate": 8.709928472485357e-05, "loss": 0.0692, "step": 27220 }, { "grad_norm": 0.5036383867263794, "learning_rate": 8.708819757501579e-05, "loss": 0.0637, "step": 27230 }, { "grad_norm": 0.3384692370891571, "learning_rate": 8.707710636926846e-05, "loss": 0.0769, "step": 27240 }, { "grad_norm": 0.5361091494560242, "learning_rate": 8.706601110882455e-05, "loss": 0.0798, "step": 27250 }, { "grad_norm": 0.6664621233940125, "learning_rate": 8.705491179489738e-05, "loss": 0.0596, "step": 27260 }, { "grad_norm": 0.379489541053772, "learning_rate": 8.704380842870077e-05, "loss": 0.0685, "step": 27270 }, { "grad_norm": 0.4902041554450989, "learning_rate": 8.703270101144895e-05, "loss": 0.0622, "step": 27280 }, { "grad_norm": 0.3939966857433319, "learning_rate": 8.702158954435664e-05, "loss": 0.0662, "step": 27290 }, { "grad_norm": 1.7114782333374023, "learning_rate": 8.701047402863896e-05, "loss": 0.0601, "step": 27300 }, { "grad_norm": 0.9471063613891602, "learning_rate": 8.699935446551148e-05, "loss": 0.0705, "step": 27310 }, { "grad_norm": 0.5191165804862976, "learning_rate": 8.698823085619022e-05, "loss": 0.0702, "step": 27320 }, { "grad_norm": 0.5776297450065613, "learning_rate": 8.697710320189166e-05, "loss": 0.053, "step": 27330 }, { "grad_norm": 0.8991086483001709, "learning_rate": 8.696597150383268e-05, "loss": 0.0656, "step": 27340 }, { "grad_norm": 0.5550997257232666, "learning_rate": 8.695483576323063e-05, "loss": 0.0693, "step": 27350 }, { "grad_norm": 0.6490929126739502, "learning_rate": 8.69436959813033e-05, "loss": 0.0719, "step": 27360 }, { "grad_norm": 0.46695518493652344, "learning_rate": 8.693255215926892e-05, "loss": 0.0707, "step": 27370 }, { "grad_norm": 0.6011999249458313, "learning_rate": 8.692140429834617e-05, "loss": 0.0776, "step": 27380 }, { "grad_norm": 0.44916895031929016, "learning_rate": 8.691025239975415e-05, "loss": 0.0629, "step": 27390 }, { "grad_norm": 0.6207895278930664, "learning_rate": 8.689909646471243e-05, "loss": 0.0651, "step": 27400 }, { "grad_norm": 0.5836828947067261, "learning_rate": 8.688793649444099e-05, "loss": 0.0771, "step": 27410 }, { "grad_norm": 0.46242645382881165, "learning_rate": 8.687677249016029e-05, "loss": 0.0627, "step": 27420 }, { "grad_norm": 0.6196951866149902, "learning_rate": 8.686560445309118e-05, "loss": 0.0693, "step": 27430 }, { "grad_norm": 0.9163296222686768, "learning_rate": 8.685443238445499e-05, "loss": 0.0832, "step": 27440 }, { "grad_norm": 0.4669491648674011, "learning_rate": 8.68432562854735e-05, "loss": 0.0843, "step": 27450 }, { "grad_norm": 0.5004749298095703, "learning_rate": 8.683207615736887e-05, "loss": 0.069, "step": 27460 }, { "grad_norm": 0.9491337537765503, "learning_rate": 8.682089200136379e-05, "loss": 0.0846, "step": 27470 }, { "grad_norm": 0.5033673644065857, "learning_rate": 8.680970381868132e-05, "loss": 0.0765, "step": 27480 }, { "grad_norm": 0.5409039258956909, "learning_rate": 8.679851161054498e-05, "loss": 0.0525, "step": 27490 }, { "grad_norm": 0.5014651417732239, "learning_rate": 8.678731537817873e-05, "loss": 0.0768, "step": 27500 }, { "grad_norm": 0.5192517042160034, "learning_rate": 8.677611512280697e-05, "loss": 0.0774, "step": 27510 }, { "grad_norm": 0.7575391530990601, "learning_rate": 8.676491084565457e-05, "loss": 0.0814, "step": 27520 }, { "grad_norm": 0.40020617842674255, "learning_rate": 8.675370254794678e-05, "loss": 0.0827, "step": 27530 }, { "grad_norm": 0.4837707579135895, "learning_rate": 8.674249023090935e-05, "loss": 0.0654, "step": 27540 }, { "grad_norm": 0.6982778906822205, "learning_rate": 8.673127389576843e-05, "loss": 0.0723, "step": 27550 }, { "grad_norm": 0.3691949248313904, "learning_rate": 8.67200535437506e-05, "loss": 0.0619, "step": 27560 }, { "grad_norm": 0.4176788330078125, "learning_rate": 8.670882917608296e-05, "loss": 0.0701, "step": 27570 }, { "grad_norm": 0.4395284354686737, "learning_rate": 8.669760079399292e-05, "loss": 0.0746, "step": 27580 }, { "grad_norm": 0.3964082896709442, "learning_rate": 8.668636839870845e-05, "loss": 0.0666, "step": 27590 }, { "grad_norm": 0.646274745464325, "learning_rate": 8.667513199145789e-05, "loss": 0.0758, "step": 27600 }, { "grad_norm": 0.5871988534927368, "learning_rate": 8.666389157347002e-05, "loss": 0.0776, "step": 27610 }, { "grad_norm": 0.99550861120224, "learning_rate": 8.66526471459741e-05, "loss": 0.0707, "step": 27620 }, { "grad_norm": 0.8744461536407471, "learning_rate": 8.66413987101998e-05, "loss": 0.0694, "step": 27630 }, { "grad_norm": 0.45982879400253296, "learning_rate": 8.663014626737723e-05, "loss": 0.0637, "step": 27640 }, { "grad_norm": 0.6288111805915833, "learning_rate": 8.661888981873691e-05, "loss": 0.0746, "step": 27650 }, { "grad_norm": 0.5576295852661133, "learning_rate": 8.660762936550988e-05, "loss": 0.0779, "step": 27660 }, { "grad_norm": 1.443009614944458, "learning_rate": 8.659636490892753e-05, "loss": 0.0822, "step": 27670 }, { "grad_norm": 0.526928186416626, "learning_rate": 8.658509645022174e-05, "loss": 0.0669, "step": 27680 }, { "grad_norm": 0.4628423750400543, "learning_rate": 8.657382399062481e-05, "loss": 0.0594, "step": 27690 }, { "grad_norm": 1.4257854223251343, "learning_rate": 8.656254753136946e-05, "loss": 0.0734, "step": 27700 }, { "grad_norm": 0.6064934730529785, "learning_rate": 8.655126707368891e-05, "loss": 0.0557, "step": 27710 }, { "grad_norm": 0.5738205313682556, "learning_rate": 8.653998261881672e-05, "loss": 0.0727, "step": 27720 }, { "grad_norm": 0.41760092973709106, "learning_rate": 8.652869416798699e-05, "loss": 0.0593, "step": 27730 }, { "grad_norm": 0.5661826729774475, "learning_rate": 8.651740172243417e-05, "loss": 0.0632, "step": 27740 }, { "grad_norm": 0.4789911210536957, "learning_rate": 8.65061052833932e-05, "loss": 0.0749, "step": 27750 }, { "grad_norm": 0.4827974736690521, "learning_rate": 8.649480485209945e-05, "loss": 0.0796, "step": 27760 }, { "grad_norm": 1.0223990678787231, "learning_rate": 8.64835004297887e-05, "loss": 0.089, "step": 27770 }, { "grad_norm": 0.49414852261543274, "learning_rate": 8.64721920176972e-05, "loss": 0.0682, "step": 27780 }, { "grad_norm": 0.6658696532249451, "learning_rate": 8.646087961706164e-05, "loss": 0.0767, "step": 27790 }, { "grad_norm": 0.5175616145133972, "learning_rate": 8.644956322911908e-05, "loss": 0.0727, "step": 27800 }, { "grad_norm": 0.5085145831108093, "learning_rate": 8.643824285510709e-05, "loss": 0.0744, "step": 27810 }, { "grad_norm": 0.5252597332000732, "learning_rate": 8.642691849626364e-05, "loss": 0.0608, "step": 27820 }, { "grad_norm": 0.5006195902824402, "learning_rate": 8.641559015382717e-05, "loss": 0.0683, "step": 27830 }, { "grad_norm": 0.8188506364822388, "learning_rate": 8.640425782903649e-05, "loss": 0.0625, "step": 27840 }, { "grad_norm": 0.4613055884838104, "learning_rate": 8.639292152313091e-05, "loss": 0.0651, "step": 27850 }, { "grad_norm": 0.6121798753738403, "learning_rate": 8.638158123735015e-05, "loss": 0.0652, "step": 27860 }, { "grad_norm": 0.4059661626815796, "learning_rate": 8.637023697293436e-05, "loss": 0.0537, "step": 27870 }, { "grad_norm": 0.5484378933906555, "learning_rate": 8.635888873112414e-05, "loss": 0.0874, "step": 27880 }, { "grad_norm": 0.4730350077152252, "learning_rate": 8.634753651316052e-05, "loss": 0.0707, "step": 27890 }, { "grad_norm": 0.385906845331192, "learning_rate": 8.633618032028496e-05, "loss": 0.0672, "step": 27900 }, { "grad_norm": 0.5512035489082336, "learning_rate": 8.632482015373934e-05, "loss": 0.0676, "step": 27910 }, { "grad_norm": 0.6521801352500916, "learning_rate": 8.6313456014766e-05, "loss": 0.077, "step": 27920 }, { "grad_norm": 0.5729157328605652, "learning_rate": 8.630208790460771e-05, "loss": 0.0669, "step": 27930 }, { "grad_norm": 0.816809892654419, "learning_rate": 8.629071582450768e-05, "loss": 0.074, "step": 27940 }, { "grad_norm": 0.5526847243309021, "learning_rate": 8.62793397757095e-05, "loss": 0.0726, "step": 27950 }, { "grad_norm": 0.4738641083240509, "learning_rate": 8.626795975945729e-05, "loss": 0.0645, "step": 27960 }, { "grad_norm": 1.1542212963104248, "learning_rate": 8.625657577699551e-05, "loss": 0.0567, "step": 27970 }, { "grad_norm": 0.7229256629943848, "learning_rate": 8.624518782956914e-05, "loss": 0.0651, "step": 27980 }, { "grad_norm": 0.5496752262115479, "learning_rate": 8.62337959184235e-05, "loss": 0.0691, "step": 27990 }, { "grad_norm": 0.539362370967865, "learning_rate": 8.622240004480441e-05, "loss": 0.0801, "step": 28000 }, { "grad_norm": 0.56496661901474, "learning_rate": 8.621100020995814e-05, "loss": 0.069, "step": 28010 }, { "grad_norm": 0.6379109025001526, "learning_rate": 8.619959641513132e-05, "loss": 0.0781, "step": 28020 }, { "grad_norm": 0.640449583530426, "learning_rate": 8.618818866157105e-05, "loss": 0.0714, "step": 28030 }, { "grad_norm": 0.7385073304176331, "learning_rate": 8.617677695052487e-05, "loss": 0.0787, "step": 28040 }, { "grad_norm": 0.4211086928844452, "learning_rate": 8.616536128324078e-05, "loss": 0.0674, "step": 28050 }, { "grad_norm": 0.6049273610115051, "learning_rate": 8.615394166096712e-05, "loss": 0.0679, "step": 28060 }, { "grad_norm": 0.4662015736103058, "learning_rate": 8.614251808495279e-05, "loss": 0.0672, "step": 28070 }, { "grad_norm": 0.5578853487968445, "learning_rate": 8.6131090556447e-05, "loss": 0.057, "step": 28080 }, { "grad_norm": 0.7512295246124268, "learning_rate": 8.611965907669947e-05, "loss": 0.0667, "step": 28090 }, { "grad_norm": 0.5371206402778625, "learning_rate": 8.610822364696034e-05, "loss": 0.0686, "step": 28100 }, { "grad_norm": 0.44162896275520325, "learning_rate": 8.609678426848015e-05, "loss": 0.0649, "step": 28110 }, { "grad_norm": 0.5286576747894287, "learning_rate": 8.60853409425099e-05, "loss": 0.0622, "step": 28120 }, { "grad_norm": 0.5951010584831238, "learning_rate": 8.607389367030104e-05, "loss": 0.0627, "step": 28130 }, { "grad_norm": 0.8545039296150208, "learning_rate": 8.606244245310538e-05, "loss": 0.0645, "step": 28140 }, { "grad_norm": 0.7812464237213135, "learning_rate": 8.605098729217525e-05, "loss": 0.0619, "step": 28150 }, { "grad_norm": 0.6085417866706848, "learning_rate": 8.603952818876335e-05, "loss": 0.0744, "step": 28160 }, { "grad_norm": 0.3676014542579651, "learning_rate": 8.602806514412281e-05, "loss": 0.0605, "step": 28170 }, { "grad_norm": 0.608156144618988, "learning_rate": 8.601659815950726e-05, "loss": 0.079, "step": 28180 }, { "grad_norm": 0.6548241376876831, "learning_rate": 8.600512723617067e-05, "loss": 0.0872, "step": 28190 }, { "grad_norm": 0.6029055714607239, "learning_rate": 8.59936523753675e-05, "loss": 0.0756, "step": 28200 }, { "grad_norm": 0.39133942127227783, "learning_rate": 8.598217357835264e-05, "loss": 0.068, "step": 28210 }, { "grad_norm": 0.5389541387557983, "learning_rate": 8.597069084638135e-05, "loss": 0.0664, "step": 28220 }, { "grad_norm": 0.6057661175727844, "learning_rate": 8.595920418070939e-05, "loss": 0.061, "step": 28230 }, { "grad_norm": 0.5875101685523987, "learning_rate": 8.594771358259295e-05, "loss": 0.0613, "step": 28240 }, { "grad_norm": 0.9211322665214539, "learning_rate": 8.593621905328858e-05, "loss": 0.0757, "step": 28250 }, { "grad_norm": 0.5123424530029297, "learning_rate": 8.592472059405333e-05, "loss": 0.0719, "step": 28260 }, { "grad_norm": 0.5718371272087097, "learning_rate": 8.591321820614464e-05, "loss": 0.0685, "step": 28270 }, { "grad_norm": 0.5357645153999329, "learning_rate": 8.590171189082041e-05, "loss": 0.0847, "step": 28280 }, { "grad_norm": 0.45422518253326416, "learning_rate": 8.589020164933894e-05, "loss": 0.0781, "step": 28290 }, { "grad_norm": 1.145745873451233, "learning_rate": 8.587868748295898e-05, "loss": 0.0656, "step": 28300 }, { "grad_norm": 0.6737880706787109, "learning_rate": 8.586716939293971e-05, "loss": 0.0622, "step": 28310 }, { "grad_norm": 0.6974624395370483, "learning_rate": 8.58556473805407e-05, "loss": 0.0616, "step": 28320 }, { "grad_norm": 0.4765084683895111, "learning_rate": 8.584412144702202e-05, "loss": 0.067, "step": 28330 }, { "grad_norm": 0.5506475567817688, "learning_rate": 8.58325915936441e-05, "loss": 0.0711, "step": 28340 }, { "grad_norm": 0.4881056547164917, "learning_rate": 8.582105782166783e-05, "loss": 0.0591, "step": 28350 }, { "grad_norm": 0.41991978883743286, "learning_rate": 8.580952013235455e-05, "loss": 0.0625, "step": 28360 }, { "grad_norm": 0.8460837006568909, "learning_rate": 8.579797852696596e-05, "loss": 0.0692, "step": 28370 }, { "grad_norm": 0.546844482421875, "learning_rate": 8.578643300676428e-05, "loss": 0.0729, "step": 28380 }, { "grad_norm": 0.43720391392707825, "learning_rate": 8.577488357301209e-05, "loss": 0.0716, "step": 28390 }, { "grad_norm": 0.7940635681152344, "learning_rate": 8.576333022697242e-05, "loss": 0.0608, "step": 28400 }, { "grad_norm": 0.9490814208984375, "learning_rate": 8.575177296990873e-05, "loss": 0.0678, "step": 28410 }, { "grad_norm": 0.5305554866790771, "learning_rate": 8.574021180308489e-05, "loss": 0.0674, "step": 28420 }, { "grad_norm": 0.7469737529754639, "learning_rate": 8.572864672776523e-05, "loss": 0.0767, "step": 28430 }, { "grad_norm": 0.3386784493923187, "learning_rate": 8.571707774521447e-05, "loss": 0.0534, "step": 28440 }, { "grad_norm": 0.4972860813140869, "learning_rate": 8.57055048566978e-05, "loss": 0.0573, "step": 28450 }, { "grad_norm": 0.7231491804122925, "learning_rate": 8.569392806348078e-05, "loss": 0.074, "step": 28460 }, { "grad_norm": 0.6800394654273987, "learning_rate": 8.568234736682947e-05, "loss": 0.0621, "step": 28470 }, { "grad_norm": 0.5854130983352661, "learning_rate": 8.567076276801029e-05, "loss": 0.0914, "step": 28480 }, { "grad_norm": 0.5528398752212524, "learning_rate": 8.565917426829013e-05, "loss": 0.0528, "step": 28490 }, { "grad_norm": 0.7972670793533325, "learning_rate": 8.564758186893628e-05, "loss": 0.0646, "step": 28500 }, { "grad_norm": 0.518192708492279, "learning_rate": 8.563598557121649e-05, "loss": 0.069, "step": 28510 }, { "grad_norm": 0.5081719160079956, "learning_rate": 8.562438537639888e-05, "loss": 0.063, "step": 28520 }, { "grad_norm": 0.5309436321258545, "learning_rate": 8.561278128575206e-05, "loss": 0.0702, "step": 28530 }, { "grad_norm": 0.7774020433425903, "learning_rate": 8.5601173300545e-05, "loss": 0.0708, "step": 28540 }, { "grad_norm": 0.8386213183403015, "learning_rate": 8.558956142204717e-05, "loss": 0.0802, "step": 28550 }, { "grad_norm": 0.624485433101654, "learning_rate": 8.55779456515284e-05, "loss": 0.0756, "step": 28560 }, { "grad_norm": 0.44174501299858093, "learning_rate": 8.556632599025898e-05, "loss": 0.0694, "step": 28570 }, { "grad_norm": 0.4800827205181122, "learning_rate": 8.555470243950964e-05, "loss": 0.0668, "step": 28580 }, { "grad_norm": 0.49145588278770447, "learning_rate": 8.554307500055148e-05, "loss": 0.0647, "step": 28590 }, { "grad_norm": 0.8565353155136108, "learning_rate": 8.553144367465609e-05, "loss": 0.0516, "step": 28600 }, { "grad_norm": 0.5015447735786438, "learning_rate": 8.551980846309544e-05, "loss": 0.0632, "step": 28610 }, { "grad_norm": 0.5863851308822632, "learning_rate": 8.550816936714193e-05, "loss": 0.0648, "step": 28620 }, { "grad_norm": 0.6082820892333984, "learning_rate": 8.549652638806841e-05, "loss": 0.0686, "step": 28630 }, { "grad_norm": 1.1855626106262207, "learning_rate": 8.548487952714812e-05, "loss": 0.0814, "step": 28640 }, { "grad_norm": 0.5416058897972107, "learning_rate": 8.547322878565478e-05, "loss": 0.0554, "step": 28650 }, { "grad_norm": 0.5700041055679321, "learning_rate": 8.546157416486245e-05, "loss": 0.0616, "step": 28660 }, { "grad_norm": 0.48126769065856934, "learning_rate": 8.54499156660457e-05, "loss": 0.0628, "step": 28670 }, { "grad_norm": 0.8457660675048828, "learning_rate": 8.543825329047947e-05, "loss": 0.0484, "step": 28680 }, { "grad_norm": 0.6720700263977051, "learning_rate": 8.542658703943913e-05, "loss": 0.0692, "step": 28690 }, { "grad_norm": 0.5626633167266846, "learning_rate": 8.541491691420051e-05, "loss": 0.0718, "step": 28700 }, { "grad_norm": 0.582580029964447, "learning_rate": 8.54032429160398e-05, "loss": 0.0637, "step": 28710 }, { "grad_norm": 0.5083529949188232, "learning_rate": 8.539156504623369e-05, "loss": 0.0731, "step": 28720 }, { "grad_norm": 0.5397158265113831, "learning_rate": 8.537988330605923e-05, "loss": 0.0887, "step": 28730 }, { "grad_norm": 0.5739240646362305, "learning_rate": 8.536819769679393e-05, "loss": 0.0667, "step": 28740 }, { "grad_norm": 0.4667418897151947, "learning_rate": 8.53565082197157e-05, "loss": 0.0618, "step": 28750 }, { "grad_norm": 0.6889055371284485, "learning_rate": 8.534481487610289e-05, "loss": 0.0824, "step": 28760 }, { "grad_norm": 0.4712006449699402, "learning_rate": 8.533311766723428e-05, "loss": 0.0533, "step": 28770 }, { "grad_norm": 0.4573215842247009, "learning_rate": 8.532141659438901e-05, "loss": 0.0569, "step": 28780 }, { "grad_norm": 0.6912335753440857, "learning_rate": 8.530971165884675e-05, "loss": 0.0627, "step": 28790 }, { "grad_norm": 0.6314652562141418, "learning_rate": 8.529800286188752e-05, "loss": 0.0674, "step": 28800 }, { "grad_norm": 0.6049616932868958, "learning_rate": 8.528629020479175e-05, "loss": 0.0553, "step": 28810 }, { "grad_norm": 0.6139387488365173, "learning_rate": 8.527457368884033e-05, "loss": 0.0605, "step": 28820 }, { "grad_norm": 0.58601313829422, "learning_rate": 8.526285331531458e-05, "loss": 0.0472, "step": 28830 }, { "grad_norm": 0.4688494801521301, "learning_rate": 8.525112908549621e-05, "loss": 0.0556, "step": 28840 }, { "grad_norm": 0.6826665997505188, "learning_rate": 8.523940100066735e-05, "loss": 0.0727, "step": 28850 }, { "grad_norm": 0.3427664637565613, "learning_rate": 8.52276690621106e-05, "loss": 0.059, "step": 28860 }, { "grad_norm": 0.6722236275672913, "learning_rate": 8.521593327110889e-05, "loss": 0.0691, "step": 28870 }, { "grad_norm": 0.5565410852432251, "learning_rate": 8.520419362894569e-05, "loss": 0.078, "step": 28880 }, { "grad_norm": 0.5207546949386597, "learning_rate": 8.51924501369048e-05, "loss": 0.0665, "step": 28890 }, { "grad_norm": 0.4758172929286957, "learning_rate": 8.518070279627047e-05, "loss": 0.0694, "step": 28900 }, { "grad_norm": 0.6478164196014404, "learning_rate": 8.516895160832737e-05, "loss": 0.0693, "step": 28910 }, { "grad_norm": 0.6049181818962097, "learning_rate": 8.515719657436061e-05, "loss": 0.0589, "step": 28920 }, { "grad_norm": 0.44621092081069946, "learning_rate": 8.514543769565568e-05, "loss": 0.071, "step": 28930 }, { "grad_norm": 1.0096254348754883, "learning_rate": 8.513367497349853e-05, "loss": 0.0719, "step": 28940 }, { "grad_norm": 0.5848804116249084, "learning_rate": 8.51219084091755e-05, "loss": 0.0605, "step": 28950 }, { "grad_norm": 0.4521106779575348, "learning_rate": 8.511013800397338e-05, "loss": 0.0667, "step": 28960 }, { "grad_norm": 0.5069073438644409, "learning_rate": 8.509836375917937e-05, "loss": 0.0728, "step": 28970 }, { "grad_norm": 0.6441195011138916, "learning_rate": 8.508658567608104e-05, "loss": 0.0649, "step": 28980 }, { "grad_norm": 0.322795569896698, "learning_rate": 8.507480375596647e-05, "loss": 0.0522, "step": 28990 }, { "grad_norm": 0.5113834738731384, "learning_rate": 8.506301800012408e-05, "loss": 0.0552, "step": 29000 }, { "grad_norm": 0.6853201985359192, "learning_rate": 8.505122840984278e-05, "loss": 0.0629, "step": 29010 }, { "grad_norm": 0.4564480185508728, "learning_rate": 8.503943498641182e-05, "loss": 0.0794, "step": 29020 }, { "grad_norm": 0.7891496419906616, "learning_rate": 8.502763773112095e-05, "loss": 0.0701, "step": 29030 }, { "grad_norm": 0.8083654046058655, "learning_rate": 8.501583664526026e-05, "loss": 0.0645, "step": 29040 }, { "grad_norm": 0.5154508352279663, "learning_rate": 8.500403173012032e-05, "loss": 0.0712, "step": 29050 }, { "grad_norm": 0.5032079815864563, "learning_rate": 8.499222298699211e-05, "loss": 0.0705, "step": 29060 }, { "grad_norm": 0.519109845161438, "learning_rate": 8.498041041716701e-05, "loss": 0.0597, "step": 29070 }, { "grad_norm": 0.7339716553688049, "learning_rate": 8.496859402193681e-05, "loss": 0.0624, "step": 29080 }, { "grad_norm": 1.4874987602233887, "learning_rate": 8.495677380259374e-05, "loss": 0.0677, "step": 29090 }, { "grad_norm": 0.481150358915329, "learning_rate": 8.494494976043045e-05, "loss": 0.073, "step": 29100 }, { "grad_norm": 0.6380738615989685, "learning_rate": 8.493312189673998e-05, "loss": 0.0631, "step": 29110 }, { "grad_norm": 1.2751903533935547, "learning_rate": 8.492129021281584e-05, "loss": 0.0695, "step": 29120 }, { "grad_norm": 0.5515620708465576, "learning_rate": 8.490945470995188e-05, "loss": 0.0665, "step": 29130 }, { "grad_norm": 0.5340465903282166, "learning_rate": 8.489761538944247e-05, "loss": 0.071, "step": 29140 }, { "grad_norm": 0.4730580747127533, "learning_rate": 8.48857722525823e-05, "loss": 0.0614, "step": 29150 }, { "grad_norm": 0.44881170988082886, "learning_rate": 8.487392530066652e-05, "loss": 0.0602, "step": 29160 }, { "grad_norm": 0.5101253986358643, "learning_rate": 8.486207453499069e-05, "loss": 0.0729, "step": 29170 }, { "grad_norm": 0.5868169665336609, "learning_rate": 8.485021995685082e-05, "loss": 0.0801, "step": 29180 }, { "grad_norm": 0.6663827896118164, "learning_rate": 8.483836156754328e-05, "loss": 0.0588, "step": 29190 }, { "grad_norm": 0.8847081065177917, "learning_rate": 8.482649936836491e-05, "loss": 0.071, "step": 29200 }, { "grad_norm": 0.48443475365638733, "learning_rate": 8.481463336061293e-05, "loss": 0.0614, "step": 29210 }, { "grad_norm": 0.41353490948677063, "learning_rate": 8.480276354558496e-05, "loss": 0.0791, "step": 29220 }, { "grad_norm": 0.8479512333869934, "learning_rate": 8.479088992457913e-05, "loss": 0.0666, "step": 29230 }, { "grad_norm": 0.4611201882362366, "learning_rate": 8.477901249889387e-05, "loss": 0.0735, "step": 29240 }, { "grad_norm": 0.5013073682785034, "learning_rate": 8.47671312698281e-05, "loss": 0.0643, "step": 29250 }, { "grad_norm": 0.659766674041748, "learning_rate": 8.475524623868112e-05, "loss": 0.0644, "step": 29260 }, { "grad_norm": 0.5279878973960876, "learning_rate": 8.474335740675266e-05, "loss": 0.0677, "step": 29270 }, { "grad_norm": 0.45807918906211853, "learning_rate": 8.473146477534289e-05, "loss": 0.0697, "step": 29280 }, { "grad_norm": 0.6878936290740967, "learning_rate": 8.471956834575232e-05, "loss": 0.0708, "step": 29290 }, { "grad_norm": 1.043812870979309, "learning_rate": 8.470766811928197e-05, "loss": 0.0626, "step": 29300 }, { "grad_norm": 0.5136010646820068, "learning_rate": 8.469576409723323e-05, "loss": 0.0601, "step": 29310 }, { "grad_norm": 0.7052605748176575, "learning_rate": 8.468385628090788e-05, "loss": 0.0595, "step": 29320 }, { "grad_norm": 0.4684418737888336, "learning_rate": 8.467194467160815e-05, "loss": 0.0659, "step": 29330 }, { "grad_norm": 0.7106634974479675, "learning_rate": 8.466002927063667e-05, "loss": 0.0617, "step": 29340 }, { "grad_norm": 0.44758307933807373, "learning_rate": 8.464811007929651e-05, "loss": 0.0655, "step": 29350 }, { "grad_norm": 0.5410663485527039, "learning_rate": 8.463618709889114e-05, "loss": 0.0624, "step": 29360 }, { "grad_norm": 0.5331364274024963, "learning_rate": 8.462426033072442e-05, "loss": 0.0701, "step": 29370 }, { "grad_norm": 0.5625334978103638, "learning_rate": 8.461232977610061e-05, "loss": 0.0662, "step": 29380 }, { "grad_norm": 0.6808059215545654, "learning_rate": 8.46003954363245e-05, "loss": 0.0555, "step": 29390 }, { "grad_norm": 0.5494192838668823, "learning_rate": 8.458845731270115e-05, "loss": 0.0715, "step": 29400 }, { "grad_norm": 0.6098804473876953, "learning_rate": 8.45765154065361e-05, "loss": 0.0667, "step": 29410 }, { "grad_norm": 0.45356959104537964, "learning_rate": 8.456456971913532e-05, "loss": 0.0591, "step": 29420 }, { "grad_norm": 0.6687137484550476, "learning_rate": 8.455262025180517e-05, "loss": 0.0623, "step": 29430 }, { "grad_norm": 0.6514686942100525, "learning_rate": 8.454066700585242e-05, "loss": 0.0633, "step": 29440 }, { "grad_norm": 0.5188268423080444, "learning_rate": 8.452870998258423e-05, "loss": 0.0751, "step": 29450 }, { "grad_norm": 0.43836894631385803, "learning_rate": 8.451674918330825e-05, "loss": 0.0624, "step": 29460 }, { "grad_norm": 0.4872617721557617, "learning_rate": 8.450478460933246e-05, "loss": 0.0706, "step": 29470 }, { "grad_norm": 0.4714987277984619, "learning_rate": 8.449281626196532e-05, "loss": 0.0786, "step": 29480 }, { "grad_norm": 0.5481548309326172, "learning_rate": 8.448084414251564e-05, "loss": 0.0781, "step": 29490 }, { "grad_norm": 0.4463347792625427, "learning_rate": 8.446886825229271e-05, "loss": 0.0537, "step": 29500 }, { "grad_norm": 0.6097362637519836, "learning_rate": 8.445688859260615e-05, "loss": 0.0714, "step": 29510 }, { "grad_norm": 0.5590413808822632, "learning_rate": 8.444490516476606e-05, "loss": 0.0557, "step": 29520 }, { "grad_norm": 0.5979557037353516, "learning_rate": 8.443291797008293e-05, "loss": 0.0667, "step": 29530 }, { "grad_norm": 0.30076491832733154, "learning_rate": 8.442092700986765e-05, "loss": 0.0713, "step": 29540 }, { "grad_norm": 0.5812272429466248, "learning_rate": 8.440893228543156e-05, "loss": 0.0717, "step": 29550 }, { "grad_norm": 0.9332613945007324, "learning_rate": 8.439693379808638e-05, "loss": 0.0612, "step": 29560 }, { "grad_norm": 0.5712844133377075, "learning_rate": 8.43849315491442e-05, "loss": 0.0687, "step": 29570 }, { "grad_norm": 0.6438945531845093, "learning_rate": 8.437292553991763e-05, "loss": 0.0805, "step": 29580 }, { "grad_norm": 0.7606147527694702, "learning_rate": 8.436091577171959e-05, "loss": 0.0776, "step": 29590 }, { "grad_norm": 0.5720640420913696, "learning_rate": 8.434890224586347e-05, "loss": 0.0592, "step": 29600 }, { "grad_norm": 0.9517804384231567, "learning_rate": 8.433688496366303e-05, "loss": 0.0653, "step": 29610 }, { "grad_norm": 0.46386295557022095, "learning_rate": 8.432486392643248e-05, "loss": 0.0542, "step": 29620 }, { "grad_norm": 0.6257531642913818, "learning_rate": 8.431283913548643e-05, "loss": 0.064, "step": 29630 }, { "grad_norm": 0.5528408288955688, "learning_rate": 8.430081059213985e-05, "loss": 0.068, "step": 29640 }, { "grad_norm": 0.46388503909111023, "learning_rate": 8.428877829770823e-05, "loss": 0.0656, "step": 29650 }, { "grad_norm": 0.7987329959869385, "learning_rate": 8.427674225350735e-05, "loss": 0.0741, "step": 29660 }, { "grad_norm": 1.0030349493026733, "learning_rate": 8.426470246085347e-05, "loss": 0.0839, "step": 29670 }, { "grad_norm": 0.7345337271690369, "learning_rate": 8.425265892106324e-05, "loss": 0.0756, "step": 29680 }, { "grad_norm": 0.49582669138908386, "learning_rate": 8.424061163545374e-05, "loss": 0.0798, "step": 29690 }, { "grad_norm": 0.42934751510620117, "learning_rate": 8.422856060534243e-05, "loss": 0.0675, "step": 29700 }, { "grad_norm": 0.606773316860199, "learning_rate": 8.421650583204718e-05, "loss": 0.069, "step": 29710 }, { "grad_norm": 0.4684344530105591, "learning_rate": 8.420444731688633e-05, "loss": 0.0707, "step": 29720 }, { "grad_norm": 0.4500337541103363, "learning_rate": 8.419238506117852e-05, "loss": 0.0753, "step": 29730 }, { "grad_norm": 0.6313812136650085, "learning_rate": 8.418031906624289e-05, "loss": 0.0638, "step": 29740 }, { "grad_norm": 0.6244277954101562, "learning_rate": 8.416824933339898e-05, "loss": 0.0743, "step": 29750 }, { "grad_norm": 0.5927148461341858, "learning_rate": 8.415617586396667e-05, "loss": 0.0622, "step": 29760 }, { "grad_norm": 0.8396753072738647, "learning_rate": 8.414409865926632e-05, "loss": 0.0659, "step": 29770 }, { "grad_norm": 0.4803055226802826, "learning_rate": 8.413201772061867e-05, "loss": 0.0635, "step": 29780 }, { "grad_norm": 0.5566259622573853, "learning_rate": 8.411993304934488e-05, "loss": 0.0543, "step": 29790 }, { "grad_norm": 0.5581269264221191, "learning_rate": 8.410784464676654e-05, "loss": 0.0608, "step": 29800 }, { "grad_norm": 0.6504049301147461, "learning_rate": 8.409575251420556e-05, "loss": 0.0668, "step": 29810 }, { "grad_norm": 0.410028874874115, "learning_rate": 8.408365665298435e-05, "loss": 0.0573, "step": 29820 }, { "grad_norm": 0.6095439195632935, "learning_rate": 8.40715570644257e-05, "loss": 0.0796, "step": 29830 }, { "grad_norm": 0.4847058057785034, "learning_rate": 8.40594537498528e-05, "loss": 0.0779, "step": 29840 }, { "grad_norm": 0.47739630937576294, "learning_rate": 8.404734671058924e-05, "loss": 0.0682, "step": 29850 }, { "grad_norm": 0.606503963470459, "learning_rate": 8.403523594795902e-05, "loss": 0.0675, "step": 29860 }, { "grad_norm": 0.611824095249176, "learning_rate": 8.402312146328659e-05, "loss": 0.0624, "step": 29870 }, { "grad_norm": 0.5286545753479004, "learning_rate": 8.401100325789675e-05, "loss": 0.0653, "step": 29880 }, { "grad_norm": 0.5504006743431091, "learning_rate": 8.399888133311472e-05, "loss": 0.0583, "step": 29890 }, { "grad_norm": 0.4995923340320587, "learning_rate": 8.398675569026613e-05, "loss": 0.0672, "step": 29900 }, { "grad_norm": 0.49894481897354126, "learning_rate": 8.397462633067705e-05, "loss": 0.0575, "step": 29910 }, { "grad_norm": 0.4664385914802551, "learning_rate": 8.396249325567392e-05, "loss": 0.0707, "step": 29920 }, { "grad_norm": 0.6243371367454529, "learning_rate": 8.395035646658357e-05, "loss": 0.0837, "step": 29930 }, { "grad_norm": 0.7110759615898132, "learning_rate": 8.39382159647333e-05, "loss": 0.0762, "step": 29940 }, { "grad_norm": 0.536725640296936, "learning_rate": 8.392607175145075e-05, "loss": 0.0579, "step": 29950 }, { "grad_norm": 0.6295602321624756, "learning_rate": 8.3913923828064e-05, "loss": 0.0705, "step": 29960 }, { "grad_norm": 0.4556734263896942, "learning_rate": 8.390177219590152e-05, "loss": 0.0503, "step": 29970 }, { "grad_norm": 0.6610842347145081, "learning_rate": 8.388961685629222e-05, "loss": 0.075, "step": 29980 }, { "grad_norm": 0.8374394178390503, "learning_rate": 8.387745781056536e-05, "loss": 0.0613, "step": 29990 }, { "grad_norm": 0.517769455909729, "learning_rate": 8.386529506005065e-05, "loss": 0.053, "step": 30000 }, { "grad_norm": 0.5499053001403809, "learning_rate": 8.38531286060782e-05, "loss": 0.0599, "step": 30010 }, { "grad_norm": 0.6094568371772766, "learning_rate": 8.384095844997849e-05, "loss": 0.0542, "step": 30020 }, { "grad_norm": 0.48168477416038513, "learning_rate": 8.382878459308245e-05, "loss": 0.0539, "step": 30030 }, { "grad_norm": 0.6615855693817139, "learning_rate": 8.381660703672138e-05, "loss": 0.0657, "step": 30040 }, { "grad_norm": 0.5844611525535583, "learning_rate": 8.380442578222702e-05, "loss": 0.0721, "step": 30050 }, { "grad_norm": 0.552383303642273, "learning_rate": 8.379224083093146e-05, "loss": 0.0606, "step": 30060 }, { "grad_norm": 0.4345334470272064, "learning_rate": 8.378005218416727e-05, "loss": 0.0693, "step": 30070 }, { "grad_norm": 0.6083831787109375, "learning_rate": 8.376785984326735e-05, "loss": 0.0662, "step": 30080 }, { "grad_norm": 0.5286577343940735, "learning_rate": 8.375566380956506e-05, "loss": 0.0561, "step": 30090 }, { "grad_norm": 0.7854036092758179, "learning_rate": 8.374346408439411e-05, "loss": 0.072, "step": 30100 }, { "grad_norm": 0.6206765174865723, "learning_rate": 8.373126066908868e-05, "loss": 0.0649, "step": 30110 }, { "grad_norm": 0.5909633040428162, "learning_rate": 8.371905356498326e-05, "loss": 0.0619, "step": 30120 }, { "grad_norm": 0.5383930206298828, "learning_rate": 8.370684277341288e-05, "loss": 0.0565, "step": 30130 }, { "grad_norm": 0.44585511088371277, "learning_rate": 8.369462829571282e-05, "loss": 0.0583, "step": 30140 }, { "grad_norm": 0.9426981210708618, "learning_rate": 8.36824101332189e-05, "loss": 0.0559, "step": 30150 }, { "grad_norm": 0.3708391785621643, "learning_rate": 8.367018828726721e-05, "loss": 0.0566, "step": 30160 }, { "grad_norm": 0.6702141761779785, "learning_rate": 8.365796275919438e-05, "loss": 0.0794, "step": 30170 }, { "grad_norm": 1.762911081314087, "learning_rate": 8.364573355033734e-05, "loss": 0.0552, "step": 30180 }, { "grad_norm": 0.4215933382511139, "learning_rate": 8.363350066203346e-05, "loss": 0.0507, "step": 30190 }, { "grad_norm": 1.032077431678772, "learning_rate": 8.362126409562053e-05, "loss": 0.0693, "step": 30200 }, { "grad_norm": 0.3699759840965271, "learning_rate": 8.360902385243667e-05, "loss": 0.0543, "step": 30210 }, { "grad_norm": 0.5109649300575256, "learning_rate": 8.359677993382052e-05, "loss": 0.0511, "step": 30220 }, { "grad_norm": 0.4130219519138336, "learning_rate": 8.358453234111103e-05, "loss": 0.0555, "step": 30230 }, { "grad_norm": 0.5463927984237671, "learning_rate": 8.357228107564756e-05, "loss": 0.0588, "step": 30240 }, { "grad_norm": 0.5508274435997009, "learning_rate": 8.356002613876993e-05, "loss": 0.0639, "step": 30250 }, { "grad_norm": 0.37658101320266724, "learning_rate": 8.35477675318183e-05, "loss": 0.0539, "step": 30260 }, { "grad_norm": 0.5021398067474365, "learning_rate": 8.353550525613323e-05, "loss": 0.0649, "step": 30270 }, { "grad_norm": 0.566703736782074, "learning_rate": 8.352323931305572e-05, "loss": 0.0703, "step": 30280 }, { "grad_norm": 0.5958095192909241, "learning_rate": 8.351096970392717e-05, "loss": 0.064, "step": 30290 }, { "grad_norm": 0.5864940881729126, "learning_rate": 8.349869643008937e-05, "loss": 0.0634, "step": 30300 }, { "grad_norm": 0.5783952474594116, "learning_rate": 8.348641949288449e-05, "loss": 0.0774, "step": 30310 }, { "grad_norm": 0.47305828332901, "learning_rate": 8.34741388936551e-05, "loss": 0.0677, "step": 30320 }, { "grad_norm": 0.60911625623703, "learning_rate": 8.346185463374423e-05, "loss": 0.079, "step": 30330 }, { "grad_norm": 0.44777876138687134, "learning_rate": 8.344956671449524e-05, "loss": 0.0703, "step": 30340 }, { "grad_norm": 0.6380483508110046, "learning_rate": 8.343727513725192e-05, "loss": 0.0622, "step": 30350 }, { "grad_norm": 0.5600403547286987, "learning_rate": 8.342497990335847e-05, "loss": 0.0651, "step": 30360 }, { "grad_norm": 0.8227300643920898, "learning_rate": 8.341268101415946e-05, "loss": 0.066, "step": 30370 }, { "grad_norm": 0.40049639344215393, "learning_rate": 8.34003784709999e-05, "loss": 0.0502, "step": 30380 }, { "grad_norm": 0.3397454619407654, "learning_rate": 8.338807227522516e-05, "loss": 0.06, "step": 30390 }, { "grad_norm": 0.44787487387657166, "learning_rate": 8.337576242818103e-05, "loss": 0.0634, "step": 30400 }, { "grad_norm": 0.3971112072467804, "learning_rate": 8.336344893121372e-05, "loss": 0.0617, "step": 30410 }, { "grad_norm": 0.42432937026023865, "learning_rate": 8.335113178566977e-05, "loss": 0.0555, "step": 30420 }, { "grad_norm": 0.4857507050037384, "learning_rate": 8.33388109928962e-05, "loss": 0.0635, "step": 30430 }, { "grad_norm": 0.411609023809433, "learning_rate": 8.33264865542404e-05, "loss": 0.0683, "step": 30440 }, { "grad_norm": 0.40056928992271423, "learning_rate": 8.331415847105013e-05, "loss": 0.0618, "step": 30450 }, { "grad_norm": 0.5302758812904358, "learning_rate": 8.330182674467357e-05, "loss": 0.0593, "step": 30460 }, { "grad_norm": 1.48435378074646, "learning_rate": 8.32894913764593e-05, "loss": 0.0622, "step": 30470 }, { "grad_norm": 0.4384503662586212, "learning_rate": 8.327715236775633e-05, "loss": 0.0547, "step": 30480 }, { "grad_norm": 0.5069820880889893, "learning_rate": 8.326480971991398e-05, "loss": 0.0514, "step": 30490 }, { "grad_norm": 0.6470165252685547, "learning_rate": 8.325246343428206e-05, "loss": 0.0696, "step": 30500 }, { "grad_norm": 0.3424203395843506, "learning_rate": 8.324011351221072e-05, "loss": 0.0504, "step": 30510 }, { "grad_norm": 0.5323463678359985, "learning_rate": 8.322775995505057e-05, "loss": 0.0589, "step": 30520 }, { "grad_norm": 0.5443872809410095, "learning_rate": 8.321540276415254e-05, "loss": 0.0676, "step": 30530 }, { "grad_norm": 0.4600512385368347, "learning_rate": 8.320304194086798e-05, "loss": 0.0582, "step": 30540 }, { "grad_norm": 0.5763612389564514, "learning_rate": 8.31906774865487e-05, "loss": 0.0609, "step": 30550 }, { "grad_norm": 0.7713364958763123, "learning_rate": 8.317830940254682e-05, "loss": 0.0669, "step": 30560 }, { "grad_norm": 0.5657011270523071, "learning_rate": 8.316593769021491e-05, "loss": 0.0723, "step": 30570 }, { "grad_norm": 0.48547565937042236, "learning_rate": 8.315356235090592e-05, "loss": 0.0635, "step": 30580 }, { "grad_norm": 1.1192148923873901, "learning_rate": 8.314118338597319e-05, "loss": 0.072, "step": 30590 }, { "grad_norm": 0.45716315507888794, "learning_rate": 8.312880079677048e-05, "loss": 0.0584, "step": 30600 }, { "grad_norm": 0.5209154486656189, "learning_rate": 8.311641458465191e-05, "loss": 0.073, "step": 30610 }, { "grad_norm": 0.960972249507904, "learning_rate": 8.310402475097205e-05, "loss": 0.0669, "step": 30620 }, { "grad_norm": 0.701627790927887, "learning_rate": 8.309163129708581e-05, "loss": 0.0705, "step": 30630 }, { "grad_norm": 0.6368336081504822, "learning_rate": 8.307923422434852e-05, "loss": 0.0554, "step": 30640 }, { "grad_norm": 1.172158122062683, "learning_rate": 8.30668335341159e-05, "loss": 0.0675, "step": 30650 }, { "grad_norm": 0.5836087465286255, "learning_rate": 8.305442922774408e-05, "loss": 0.0671, "step": 30660 }, { "grad_norm": 0.5577280521392822, "learning_rate": 8.304202130658959e-05, "loss": 0.0578, "step": 30670 }, { "grad_norm": 0.4670637547969818, "learning_rate": 8.302960977200931e-05, "loss": 0.0489, "step": 30680 }, { "grad_norm": 0.5427863001823425, "learning_rate": 8.301719462536058e-05, "loss": 0.0678, "step": 30690 }, { "grad_norm": 0.5881525874137878, "learning_rate": 8.300477586800108e-05, "loss": 0.0603, "step": 30700 }, { "grad_norm": 0.3859214782714844, "learning_rate": 8.299235350128892e-05, "loss": 0.0554, "step": 30710 }, { "grad_norm": 0.4319872260093689, "learning_rate": 8.297992752658258e-05, "loss": 0.052, "step": 30720 }, { "grad_norm": 0.4290306866168976, "learning_rate": 8.296749794524094e-05, "loss": 0.0623, "step": 30730 }, { "grad_norm": 0.5252252221107483, "learning_rate": 8.295506475862332e-05, "loss": 0.0542, "step": 30740 }, { "grad_norm": 0.4330235719680786, "learning_rate": 8.294262796808933e-05, "loss": 0.0622, "step": 30750 }, { "grad_norm": 0.6095981001853943, "learning_rate": 8.293018757499909e-05, "loss": 0.0673, "step": 30760 }, { "grad_norm": 1.0861496925354004, "learning_rate": 8.291774358071305e-05, "loss": 0.0648, "step": 30770 }, { "grad_norm": 0.6542552709579468, "learning_rate": 8.290529598659205e-05, "loss": 0.071, "step": 30780 }, { "grad_norm": 0.7915806174278259, "learning_rate": 8.289284479399738e-05, "loss": 0.0641, "step": 30790 }, { "grad_norm": 0.6383407711982727, "learning_rate": 8.288039000429064e-05, "loss": 0.0605, "step": 30800 }, { "grad_norm": 1.4109628200531006, "learning_rate": 8.286793161883388e-05, "loss": 0.0665, "step": 30810 }, { "grad_norm": 0.575689435005188, "learning_rate": 8.285546963898954e-05, "loss": 0.0634, "step": 30820 }, { "grad_norm": 0.6155154705047607, "learning_rate": 8.284300406612044e-05, "loss": 0.0607, "step": 30830 }, { "grad_norm": 0.6455920338630676, "learning_rate": 8.283053490158978e-05, "loss": 0.0561, "step": 30840 }, { "grad_norm": 0.4883652925491333, "learning_rate": 8.28180621467612e-05, "loss": 0.0606, "step": 30850 }, { "grad_norm": 0.5117023587226868, "learning_rate": 8.280558580299867e-05, "loss": 0.0605, "step": 30860 }, { "grad_norm": 0.6480007171630859, "learning_rate": 8.279310587166661e-05, "loss": 0.0629, "step": 30870 }, { "grad_norm": 0.4494709372520447, "learning_rate": 8.278062235412978e-05, "loss": 0.055, "step": 30880 }, { "grad_norm": 0.7173055410385132, "learning_rate": 8.276813525175339e-05, "loss": 0.0679, "step": 30890 }, { "grad_norm": 1.118422269821167, "learning_rate": 8.2755644565903e-05, "loss": 0.0679, "step": 30900 }, { "grad_norm": 0.5473146438598633, "learning_rate": 8.274315029794454e-05, "loss": 0.0564, "step": 30910 }, { "grad_norm": 0.718964159488678, "learning_rate": 8.273065244924443e-05, "loss": 0.066, "step": 30920 }, { "grad_norm": 0.7000981569290161, "learning_rate": 8.271815102116936e-05, "loss": 0.067, "step": 30930 }, { "grad_norm": 0.43798723816871643, "learning_rate": 8.270564601508648e-05, "loss": 0.066, "step": 30940 }, { "grad_norm": 0.4867660403251648, "learning_rate": 8.269313743236333e-05, "loss": 0.0596, "step": 30950 }, { "grad_norm": 0.4009227156639099, "learning_rate": 8.268062527436783e-05, "loss": 0.0523, "step": 30960 }, { "grad_norm": 0.5741157531738281, "learning_rate": 8.266810954246828e-05, "loss": 0.062, "step": 30970 }, { "grad_norm": 0.7315882444381714, "learning_rate": 8.265559023803338e-05, "loss": 0.0472, "step": 30980 }, { "grad_norm": 0.4485141634941101, "learning_rate": 8.264306736243223e-05, "loss": 0.0589, "step": 30990 }, { "grad_norm": 1.2633275985717773, "learning_rate": 8.263054091703432e-05, "loss": 0.069, "step": 31000 }, { "grad_norm": 0.5288218259811401, "learning_rate": 8.26180109032095e-05, "loss": 0.0587, "step": 31010 }, { "grad_norm": 0.8557580709457397, "learning_rate": 8.260547732232807e-05, "loss": 0.0596, "step": 31020 }, { "grad_norm": 0.4539509117603302, "learning_rate": 8.259294017576063e-05, "loss": 0.0583, "step": 31030 }, { "grad_norm": 0.64522385597229, "learning_rate": 8.258039946487828e-05, "loss": 0.0548, "step": 31040 }, { "grad_norm": 0.8944559693336487, "learning_rate": 8.256785519105241e-05, "loss": 0.0535, "step": 31050 }, { "grad_norm": 0.8096775412559509, "learning_rate": 8.255530735565488e-05, "loss": 0.0654, "step": 31060 }, { "grad_norm": 0.7894022464752197, "learning_rate": 8.254275596005787e-05, "loss": 0.0618, "step": 31070 }, { "grad_norm": 0.7123059630393982, "learning_rate": 8.2530201005634e-05, "loss": 0.0743, "step": 31080 }, { "grad_norm": 0.4003221094608307, "learning_rate": 8.251764249375626e-05, "loss": 0.0626, "step": 31090 }, { "grad_norm": 0.6250046491622925, "learning_rate": 8.250508042579803e-05, "loss": 0.0649, "step": 31100 }, { "grad_norm": 0.9196188449859619, "learning_rate": 8.249251480313307e-05, "loss": 0.07, "step": 31110 }, { "grad_norm": 0.7271408438682556, "learning_rate": 8.247994562713555e-05, "loss": 0.0695, "step": 31120 }, { "grad_norm": 0.5263214111328125, "learning_rate": 8.246737289917998e-05, "loss": 0.072, "step": 31130 }, { "grad_norm": 0.5684137344360352, "learning_rate": 8.245479662064135e-05, "loss": 0.0606, "step": 31140 }, { "grad_norm": 0.4926689565181732, "learning_rate": 8.244221679289496e-05, "loss": 0.0561, "step": 31150 }, { "grad_norm": 0.5875319242477417, "learning_rate": 8.242963341731652e-05, "loss": 0.0535, "step": 31160 }, { "grad_norm": 0.5877638459205627, "learning_rate": 8.24170464952821e-05, "loss": 0.0593, "step": 31170 }, { "grad_norm": 0.5238903760910034, "learning_rate": 8.240445602816824e-05, "loss": 0.0558, "step": 31180 }, { "grad_norm": 0.7099484801292419, "learning_rate": 8.239186201735179e-05, "loss": 0.0607, "step": 31190 }, { "grad_norm": 0.3863934874534607, "learning_rate": 8.237926446420998e-05, "loss": 0.0581, "step": 31200 }, { "grad_norm": 0.5471599698066711, "learning_rate": 8.236666337012052e-05, "loss": 0.0577, "step": 31210 }, { "grad_norm": 0.7589980363845825, "learning_rate": 8.23540587364614e-05, "loss": 0.0726, "step": 31220 }, { "grad_norm": 0.5082844495773315, "learning_rate": 8.234145056461107e-05, "loss": 0.0571, "step": 31230 }, { "grad_norm": 0.37938031554222107, "learning_rate": 8.232883885594832e-05, "loss": 0.069, "step": 31240 }, { "grad_norm": 0.5429555773735046, "learning_rate": 8.231622361185236e-05, "loss": 0.0609, "step": 31250 }, { "grad_norm": 0.5933529138565063, "learning_rate": 8.230360483370278e-05, "loss": 0.0858, "step": 31260 }, { "grad_norm": 0.5690203309059143, "learning_rate": 8.229098252287953e-05, "loss": 0.0645, "step": 31270 }, { "grad_norm": 0.535372257232666, "learning_rate": 8.2278356680763e-05, "loss": 0.0752, "step": 31280 }, { "grad_norm": 0.8005028367042542, "learning_rate": 8.22657273087339e-05, "loss": 0.0733, "step": 31290 }, { "grad_norm": 0.6684659123420715, "learning_rate": 8.225309440817336e-05, "loss": 0.0802, "step": 31300 }, { "grad_norm": 0.616405189037323, "learning_rate": 8.224045798046293e-05, "loss": 0.0884, "step": 31310 }, { "grad_norm": 0.4310525357723236, "learning_rate": 8.22278180269845e-05, "loss": 0.0714, "step": 31320 }, { "grad_norm": 0.5264288187026978, "learning_rate": 8.221517454912031e-05, "loss": 0.059, "step": 31330 }, { "grad_norm": 0.46768105030059814, "learning_rate": 8.220252754825308e-05, "loss": 0.0504, "step": 31340 }, { "grad_norm": 0.5139279961585999, "learning_rate": 8.218987702576586e-05, "loss": 0.055, "step": 31350 }, { "grad_norm": 0.5611823201179504, "learning_rate": 8.217722298304207e-05, "loss": 0.0586, "step": 31360 }, { "grad_norm": 0.4646342098712921, "learning_rate": 8.216456542146557e-05, "loss": 0.0555, "step": 31370 }, { "grad_norm": 0.42991217970848083, "learning_rate": 8.215190434242055e-05, "loss": 0.0491, "step": 31380 }, { "grad_norm": 0.7385907173156738, "learning_rate": 8.213923974729161e-05, "loss": 0.0617, "step": 31390 }, { "grad_norm": 0.46081849932670593, "learning_rate": 8.212657163746373e-05, "loss": 0.0562, "step": 31400 }, { "grad_norm": 0.504187822341919, "learning_rate": 8.211390001432227e-05, "loss": 0.0526, "step": 31410 }, { "grad_norm": 0.510650098323822, "learning_rate": 8.210122487925297e-05, "loss": 0.0657, "step": 31420 }, { "grad_norm": 0.6168496608734131, "learning_rate": 8.208854623364202e-05, "loss": 0.0565, "step": 31430 }, { "grad_norm": 0.30468007922172546, "learning_rate": 8.207586407887589e-05, "loss": 0.0541, "step": 31440 }, { "grad_norm": 0.4875885546207428, "learning_rate": 8.206317841634148e-05, "loss": 0.0561, "step": 31450 }, { "grad_norm": 0.4808175563812256, "learning_rate": 8.205048924742609e-05, "loss": 0.0638, "step": 31460 }, { "grad_norm": 0.7239668369293213, "learning_rate": 8.203779657351738e-05, "loss": 0.0617, "step": 31470 }, { "grad_norm": 0.4274426996707916, "learning_rate": 8.20251003960034e-05, "loss": 0.0659, "step": 31480 }, { "grad_norm": 0.4774843752384186, "learning_rate": 8.201240071627258e-05, "loss": 0.066, "step": 31490 }, { "grad_norm": 0.41512274742126465, "learning_rate": 8.199969753571377e-05, "loss": 0.0701, "step": 31500 }, { "grad_norm": 0.6734706163406372, "learning_rate": 8.198699085571615e-05, "loss": 0.0675, "step": 31510 }, { "grad_norm": 0.9975398182868958, "learning_rate": 8.197428067766928e-05, "loss": 0.069, "step": 31520 }, { "grad_norm": 1.2174121141433716, "learning_rate": 8.196156700296316e-05, "loss": 0.0841, "step": 31530 }, { "grad_norm": 0.5969118475914001, "learning_rate": 8.194884983298814e-05, "loss": 0.053, "step": 31540 }, { "grad_norm": 0.37016576528549194, "learning_rate": 8.193612916913491e-05, "loss": 0.0537, "step": 31550 }, { "grad_norm": 1.1339727640151978, "learning_rate": 8.192340501279463e-05, "loss": 0.0547, "step": 31560 }, { "grad_norm": 0.5305784344673157, "learning_rate": 8.191067736535876e-05, "loss": 0.0688, "step": 31570 }, { "grad_norm": 0.5279974937438965, "learning_rate": 8.18979462282192e-05, "loss": 0.0675, "step": 31580 }, { "grad_norm": 0.49890127778053284, "learning_rate": 8.188521160276819e-05, "loss": 0.0653, "step": 31590 }, { "grad_norm": 0.5808979868888855, "learning_rate": 8.187247349039837e-05, "loss": 0.0668, "step": 31600 }, { "grad_norm": 0.5477613210678101, "learning_rate": 8.185973189250278e-05, "loss": 0.0678, "step": 31610 }, { "grad_norm": 0.794893205165863, "learning_rate": 8.184698681047482e-05, "loss": 0.083, "step": 31620 }, { "grad_norm": 0.5380685329437256, "learning_rate": 8.183423824570827e-05, "loss": 0.0682, "step": 31630 }, { "grad_norm": 0.6724821329116821, "learning_rate": 8.182148619959725e-05, "loss": 0.0666, "step": 31640 }, { "grad_norm": 0.527857780456543, "learning_rate": 8.180873067353636e-05, "loss": 0.0661, "step": 31650 }, { "grad_norm": 0.588433027267456, "learning_rate": 8.179597166892052e-05, "loss": 0.0672, "step": 31660 }, { "grad_norm": 0.5250037312507629, "learning_rate": 8.178320918714501e-05, "loss": 0.0596, "step": 31670 }, { "grad_norm": 1.9350576400756836, "learning_rate": 8.177044322960554e-05, "loss": 0.0603, "step": 31680 }, { "grad_norm": 0.518754243850708, "learning_rate": 8.175767379769816e-05, "loss": 0.0668, "step": 31690 }, { "grad_norm": 0.6032214760780334, "learning_rate": 8.174490089281932e-05, "loss": 0.0679, "step": 31700 }, { "grad_norm": 0.47287631034851074, "learning_rate": 8.173212451636584e-05, "loss": 0.0775, "step": 31710 }, { "grad_norm": 0.9743130207061768, "learning_rate": 8.171934466973493e-05, "loss": 0.0773, "step": 31720 }, { "grad_norm": 0.3811646103858948, "learning_rate": 8.170656135432418e-05, "loss": 0.0578, "step": 31730 }, { "grad_norm": 0.5376361608505249, "learning_rate": 8.169377457153155e-05, "loss": 0.0749, "step": 31740 }, { "grad_norm": 0.5909029841423035, "learning_rate": 8.168098432275539e-05, "loss": 0.0713, "step": 31750 }, { "grad_norm": 1.1250215768814087, "learning_rate": 8.166819060939442e-05, "loss": 0.0739, "step": 31760 }, { "grad_norm": 0.5160366892814636, "learning_rate": 8.165539343284772e-05, "loss": 0.0688, "step": 31770 }, { "grad_norm": 1.2382256984710693, "learning_rate": 8.16425927945148e-05, "loss": 0.0801, "step": 31780 }, { "grad_norm": 0.5806798934936523, "learning_rate": 8.162978869579551e-05, "loss": 0.0676, "step": 31790 }, { "grad_norm": 0.47788935899734497, "learning_rate": 8.161698113809007e-05, "loss": 0.0521, "step": 31800 }, { "grad_norm": 0.45056310296058655, "learning_rate": 8.160417012279911e-05, "loss": 0.0832, "step": 31810 }, { "grad_norm": 0.4395137429237366, "learning_rate": 8.159135565132363e-05, "loss": 0.069, "step": 31820 }, { "grad_norm": 1.0290809869766235, "learning_rate": 8.157853772506498e-05, "loss": 0.085, "step": 31830 }, { "grad_norm": 0.6131218075752258, "learning_rate": 8.156571634542494e-05, "loss": 0.0698, "step": 31840 }, { "grad_norm": 1.343503475189209, "learning_rate": 8.15528915138056e-05, "loss": 0.0679, "step": 31850 }, { "grad_norm": 0.4430365562438965, "learning_rate": 8.154006323160949e-05, "loss": 0.0585, "step": 31860 }, { "grad_norm": 0.749911904335022, "learning_rate": 8.152723150023949e-05, "loss": 0.0757, "step": 31870 }, { "grad_norm": 0.5216156840324402, "learning_rate": 8.151439632109886e-05, "loss": 0.0774, "step": 31880 }, { "grad_norm": 0.57146817445755, "learning_rate": 8.150155769559122e-05, "loss": 0.0679, "step": 31890 }, { "grad_norm": 0.6057096719741821, "learning_rate": 8.148871562512058e-05, "loss": 0.0666, "step": 31900 }, { "grad_norm": 0.5322477221488953, "learning_rate": 8.147587011109136e-05, "loss": 0.0798, "step": 31910 }, { "grad_norm": 0.6460737586021423, "learning_rate": 8.14630211549083e-05, "loss": 0.0792, "step": 31920 }, { "grad_norm": 1.0279690027236938, "learning_rate": 8.145016875797655e-05, "loss": 0.0547, "step": 31930 }, { "grad_norm": 0.48147597908973694, "learning_rate": 8.143731292170164e-05, "loss": 0.0678, "step": 31940 }, { "grad_norm": 0.6055989265441895, "learning_rate": 8.142445364748944e-05, "loss": 0.0698, "step": 31950 }, { "grad_norm": 0.5353529453277588, "learning_rate": 8.141159093674624e-05, "loss": 0.0749, "step": 31960 }, { "grad_norm": 0.5546183586120605, "learning_rate": 8.139872479087869e-05, "loss": 0.0603, "step": 31970 }, { "grad_norm": 0.5127283930778503, "learning_rate": 8.13858552112938e-05, "loss": 0.0527, "step": 31980 }, { "grad_norm": 0.45154306292533875, "learning_rate": 8.137298219939895e-05, "loss": 0.0746, "step": 31990 }, { "grad_norm": 0.5479001402854919, "learning_rate": 8.136010575660196e-05, "loss": 0.0521, "step": 32000 }, { "grad_norm": 0.5079127550125122, "learning_rate": 8.134722588431095e-05, "loss": 0.0606, "step": 32010 }, { "grad_norm": 0.4494607150554657, "learning_rate": 8.133434258393444e-05, "loss": 0.0761, "step": 32020 }, { "grad_norm": 0.8304328918457031, "learning_rate": 8.132145585688134e-05, "loss": 0.0697, "step": 32030 }, { "grad_norm": 0.5551819801330566, "learning_rate": 8.130856570456093e-05, "loss": 0.0751, "step": 32040 }, { "grad_norm": 0.4719085991382599, "learning_rate": 8.129567212838283e-05, "loss": 0.0724, "step": 32050 }, { "grad_norm": 0.47447293996810913, "learning_rate": 8.128277512975708e-05, "loss": 0.0754, "step": 32060 }, { "grad_norm": 0.703578531742096, "learning_rate": 8.126987471009408e-05, "loss": 0.0696, "step": 32070 }, { "grad_norm": 0.4397963881492615, "learning_rate": 8.125697087080459e-05, "loss": 0.0601, "step": 32080 }, { "grad_norm": 0.6467996835708618, "learning_rate": 8.124406361329976e-05, "loss": 0.0617, "step": 32090 }, { "grad_norm": 0.521967351436615, "learning_rate": 8.123115293899111e-05, "loss": 0.0561, "step": 32100 }, { "grad_norm": 0.6017237305641174, "learning_rate": 8.121823884929055e-05, "loss": 0.0552, "step": 32110 }, { "grad_norm": 0.6492536067962646, "learning_rate": 8.12053213456103e-05, "loss": 0.0534, "step": 32120 }, { "grad_norm": 0.39640572667121887, "learning_rate": 8.119240042936303e-05, "loss": 0.0443, "step": 32130 }, { "grad_norm": 0.8016153573989868, "learning_rate": 8.117947610196175e-05, "loss": 0.0523, "step": 32140 }, { "grad_norm": 1.2261732816696167, "learning_rate": 8.116654836481982e-05, "loss": 0.0581, "step": 32150 }, { "grad_norm": 1.3638519048690796, "learning_rate": 8.115361721935106e-05, "loss": 0.0716, "step": 32160 }, { "grad_norm": 0.46513259410858154, "learning_rate": 8.114068266696953e-05, "loss": 0.065, "step": 32170 }, { "grad_norm": 0.4701407849788666, "learning_rate": 8.112774470908978e-05, "loss": 0.0455, "step": 32180 }, { "grad_norm": 0.6247933506965637, "learning_rate": 8.111480334712665e-05, "loss": 0.0625, "step": 32190 }, { "grad_norm": 0.5552453994750977, "learning_rate": 8.110185858249542e-05, "loss": 0.0588, "step": 32200 }, { "grad_norm": 0.6669722199440002, "learning_rate": 8.108891041661168e-05, "loss": 0.0587, "step": 32210 }, { "grad_norm": 0.48829492926597595, "learning_rate": 8.107595885089146e-05, "loss": 0.0662, "step": 32220 }, { "grad_norm": 1.6504788398742676, "learning_rate": 8.106300388675108e-05, "loss": 0.0674, "step": 32230 }, { "grad_norm": 0.5056043863296509, "learning_rate": 8.105004552560731e-05, "loss": 0.0643, "step": 32240 }, { "grad_norm": 0.6134249567985535, "learning_rate": 8.103708376887724e-05, "loss": 0.0658, "step": 32250 }, { "grad_norm": 0.4794563353061676, "learning_rate": 8.102411861797836e-05, "loss": 0.0547, "step": 32260 }, { "grad_norm": 0.40978336334228516, "learning_rate": 8.101115007432851e-05, "loss": 0.0521, "step": 32270 }, { "grad_norm": 0.5610289573669434, "learning_rate": 8.09981781393459e-05, "loss": 0.0606, "step": 32280 }, { "grad_norm": 0.758672297000885, "learning_rate": 8.098520281444915e-05, "loss": 0.0573, "step": 32290 }, { "grad_norm": 0.6094570755958557, "learning_rate": 8.09722241010572e-05, "loss": 0.0519, "step": 32300 }, { "grad_norm": 0.41697975993156433, "learning_rate": 8.095924200058939e-05, "loss": 0.0597, "step": 32310 }, { "grad_norm": 0.5962857007980347, "learning_rate": 8.094625651446541e-05, "loss": 0.0527, "step": 32320 }, { "grad_norm": 0.4370649456977844, "learning_rate": 8.093326764410536e-05, "loss": 0.0658, "step": 32330 }, { "grad_norm": 0.4952823221683502, "learning_rate": 8.092027539092966e-05, "loss": 0.0607, "step": 32340 }, { "grad_norm": 0.6265946626663208, "learning_rate": 8.090727975635913e-05, "loss": 0.0574, "step": 32350 }, { "grad_norm": 0.59101802110672, "learning_rate": 8.089428074181497e-05, "loss": 0.0638, "step": 32360 }, { "grad_norm": 0.41195496916770935, "learning_rate": 8.088127834871871e-05, "loss": 0.0508, "step": 32370 }, { "grad_norm": 1.1704909801483154, "learning_rate": 8.086827257849226e-05, "loss": 0.0609, "step": 32380 }, { "grad_norm": 0.6233135461807251, "learning_rate": 8.085526343255795e-05, "loss": 0.0496, "step": 32390 }, { "grad_norm": 0.6596986651420593, "learning_rate": 8.084225091233842e-05, "loss": 0.0723, "step": 32400 }, { "grad_norm": 0.7813766598701477, "learning_rate": 8.082923501925668e-05, "loss": 0.0528, "step": 32410 }, { "grad_norm": 1.0035102367401123, "learning_rate": 8.081621575473617e-05, "loss": 0.0717, "step": 32420 }, { "grad_norm": 0.4942333698272705, "learning_rate": 8.080319312020064e-05, "loss": 0.0636, "step": 32430 }, { "grad_norm": 0.6693204641342163, "learning_rate": 8.079016711707421e-05, "loss": 0.0604, "step": 32440 }, { "grad_norm": 0.5280096530914307, "learning_rate": 8.077713774678139e-05, "loss": 0.0621, "step": 32450 }, { "grad_norm": 0.6655876040458679, "learning_rate": 8.076410501074707e-05, "loss": 0.0531, "step": 32460 }, { "grad_norm": 0.5477615594863892, "learning_rate": 8.075106891039647e-05, "loss": 0.0612, "step": 32470 }, { "grad_norm": 0.5194348692893982, "learning_rate": 8.073802944715523e-05, "loss": 0.0678, "step": 32480 }, { "grad_norm": 0.6094551086425781, "learning_rate": 8.072498662244929e-05, "loss": 0.0577, "step": 32490 }, { "grad_norm": 0.49012431502342224, "learning_rate": 8.0711940437705e-05, "loss": 0.0532, "step": 32500 }, { "grad_norm": 1.0621188879013062, "learning_rate": 8.06988908943491e-05, "loss": 0.0547, "step": 32510 }, { "grad_norm": 0.7268290519714355, "learning_rate": 8.068583799380863e-05, "loss": 0.0478, "step": 32520 }, { "grad_norm": 0.6215039491653442, "learning_rate": 8.067278173751104e-05, "loss": 0.0686, "step": 32530 }, { "grad_norm": 0.6217615604400635, "learning_rate": 8.065972212688417e-05, "loss": 0.0635, "step": 32540 }, { "grad_norm": 1.0388453006744385, "learning_rate": 8.064665916335618e-05, "loss": 0.0777, "step": 32550 }, { "grad_norm": 0.4002118408679962, "learning_rate": 8.063359284835564e-05, "loss": 0.0773, "step": 32560 }, { "grad_norm": 0.5423628091812134, "learning_rate": 8.062052318331142e-05, "loss": 0.0752, "step": 32570 }, { "grad_norm": 0.6198667287826538, "learning_rate": 8.060745016965283e-05, "loss": 0.0645, "step": 32580 }, { "grad_norm": 0.40280845761299133, "learning_rate": 8.059437380880952e-05, "loss": 0.0482, "step": 32590 }, { "grad_norm": 0.44974443316459656, "learning_rate": 8.058129410221146e-05, "loss": 0.0582, "step": 32600 }, { "grad_norm": 0.4788367450237274, "learning_rate": 8.056821105128908e-05, "loss": 0.0594, "step": 32610 }, { "grad_norm": 0.8357924222946167, "learning_rate": 8.05551246574731e-05, "loss": 0.0675, "step": 32620 }, { "grad_norm": 0.5811725854873657, "learning_rate": 8.05420349221946e-05, "loss": 0.0581, "step": 32630 }, { "grad_norm": 0.5583701133728027, "learning_rate": 8.05289418468851e-05, "loss": 0.0463, "step": 32640 }, { "grad_norm": 0.4148089289665222, "learning_rate": 8.051584543297642e-05, "loss": 0.0636, "step": 32650 }, { "grad_norm": 0.38806042075157166, "learning_rate": 8.050274568190074e-05, "loss": 0.0649, "step": 32660 }, { "grad_norm": 0.6643257141113281, "learning_rate": 8.048964259509067e-05, "loss": 0.0608, "step": 32670 }, { "grad_norm": 0.7047713398933411, "learning_rate": 8.047653617397914e-05, "loss": 0.066, "step": 32680 }, { "grad_norm": 0.48849165439605713, "learning_rate": 8.046342641999941e-05, "loss": 0.0596, "step": 32690 }, { "grad_norm": 0.4687907099723816, "learning_rate": 8.045031333458517e-05, "loss": 0.0635, "step": 32700 }, { "grad_norm": 0.5832780599594116, "learning_rate": 8.043719691917047e-05, "loss": 0.0601, "step": 32710 }, { "grad_norm": 0.49425551295280457, "learning_rate": 8.042407717518966e-05, "loss": 0.0581, "step": 32720 }, { "grad_norm": 0.6040213108062744, "learning_rate": 8.041095410407751e-05, "loss": 0.0554, "step": 32730 }, { "grad_norm": 0.4590105712413788, "learning_rate": 8.039782770726913e-05, "loss": 0.0564, "step": 32740 }, { "grad_norm": 0.8560683727264404, "learning_rate": 8.038469798620004e-05, "loss": 0.0625, "step": 32750 }, { "grad_norm": 0.41662052273750305, "learning_rate": 8.037156494230604e-05, "loss": 0.0618, "step": 32760 }, { "grad_norm": 0.4743669629096985, "learning_rate": 8.035842857702338e-05, "loss": 0.0479, "step": 32770 }, { "grad_norm": 0.6763925552368164, "learning_rate": 8.03452888917886e-05, "loss": 0.0672, "step": 32780 }, { "grad_norm": 0.4595227539539337, "learning_rate": 8.033214588803866e-05, "loss": 0.0626, "step": 32790 }, { "grad_norm": 0.5258470773696899, "learning_rate": 8.031899956721083e-05, "loss": 0.0555, "step": 32800 }, { "grad_norm": 0.5597015619277954, "learning_rate": 8.030584993074282e-05, "loss": 0.0622, "step": 32810 }, { "grad_norm": 0.49066388607025146, "learning_rate": 8.02926969800726e-05, "loss": 0.0567, "step": 32820 }, { "grad_norm": 0.5824177861213684, "learning_rate": 8.027954071663859e-05, "loss": 0.0631, "step": 32830 }, { "grad_norm": 0.7405738830566406, "learning_rate": 8.026638114187954e-05, "loss": 0.0641, "step": 32840 }, { "grad_norm": 0.6142674684524536, "learning_rate": 8.025321825723456e-05, "loss": 0.0647, "step": 32850 }, { "grad_norm": 0.5435217618942261, "learning_rate": 8.02400520641431e-05, "loss": 0.0711, "step": 32860 }, { "grad_norm": 0.3672603666782379, "learning_rate": 8.022688256404501e-05, "loss": 0.0562, "step": 32870 }, { "grad_norm": 0.6454229950904846, "learning_rate": 8.02137097583805e-05, "loss": 0.057, "step": 32880 }, { "grad_norm": 0.637563943862915, "learning_rate": 8.02005336485901e-05, "loss": 0.0696, "step": 32890 }, { "grad_norm": 0.7890486121177673, "learning_rate": 8.018735423611476e-05, "loss": 0.0712, "step": 32900 }, { "grad_norm": 1.0303527116775513, "learning_rate": 8.017417152239574e-05, "loss": 0.0615, "step": 32910 }, { "grad_norm": 0.8167387843132019, "learning_rate": 8.01609855088747e-05, "loss": 0.0714, "step": 32920 }, { "grad_norm": 1.229218602180481, "learning_rate": 8.014779619699362e-05, "loss": 0.0603, "step": 32930 }, { "grad_norm": 0.4937235713005066, "learning_rate": 8.013460358819489e-05, "loss": 0.0586, "step": 32940 }, { "grad_norm": 0.6539539098739624, "learning_rate": 8.01214076839212e-05, "loss": 0.0681, "step": 32950 }, { "grad_norm": 0.5080826282501221, "learning_rate": 8.010820848561565e-05, "loss": 0.0569, "step": 32960 }, { "grad_norm": 0.5519569516181946, "learning_rate": 8.009500599472171e-05, "loss": 0.0485, "step": 32970 }, { "grad_norm": 0.8025927543640137, "learning_rate": 8.008180021268314e-05, "loss": 0.062, "step": 32980 }, { "grad_norm": 0.4198330342769623, "learning_rate": 8.006859114094414e-05, "loss": 0.056, "step": 32990 }, { "grad_norm": 0.5176927447319031, "learning_rate": 8.005537878094921e-05, "loss": 0.0642, "step": 33000 }, { "grad_norm": 0.4153161644935608, "learning_rate": 8.004216313414323e-05, "loss": 0.0635, "step": 33010 }, { "grad_norm": 0.3180943429470062, "learning_rate": 8.002894420197149e-05, "loss": 0.0505, "step": 33020 }, { "grad_norm": 0.5241291522979736, "learning_rate": 8.001572198587954e-05, "loss": 0.0525, "step": 33030 }, { "grad_norm": 1.1242049932479858, "learning_rate": 8.000249648731338e-05, "loss": 0.0644, "step": 33040 }, { "grad_norm": 0.5624675154685974, "learning_rate": 7.998926770771928e-05, "loss": 0.0609, "step": 33050 }, { "grad_norm": 0.6938629746437073, "learning_rate": 7.997603564854397e-05, "loss": 0.073, "step": 33060 }, { "grad_norm": 0.4770563244819641, "learning_rate": 7.996280031123448e-05, "loss": 0.0639, "step": 33070 }, { "grad_norm": 0.5485038161277771, "learning_rate": 7.994956169723818e-05, "loss": 0.0745, "step": 33080 }, { "grad_norm": 0.43660321831703186, "learning_rate": 7.993631980800285e-05, "loss": 0.0544, "step": 33090 }, { "grad_norm": 0.5767941474914551, "learning_rate": 7.992307464497659e-05, "loss": 0.0754, "step": 33100 }, { "grad_norm": 0.41070622205734253, "learning_rate": 7.990982620960787e-05, "loss": 0.0494, "step": 33110 }, { "grad_norm": 2.246242046356201, "learning_rate": 7.989657450334554e-05, "loss": 0.0677, "step": 33120 }, { "grad_norm": 0.689734160900116, "learning_rate": 7.988331952763877e-05, "loss": 0.0592, "step": 33130 }, { "grad_norm": 0.46203169226646423, "learning_rate": 7.987006128393709e-05, "loss": 0.0585, "step": 33140 }, { "grad_norm": 0.5870743989944458, "learning_rate": 7.985679977369043e-05, "loss": 0.0504, "step": 33150 }, { "grad_norm": 0.5939298272132874, "learning_rate": 7.984353499834902e-05, "loss": 0.0495, "step": 33160 }, { "grad_norm": 0.4593481421470642, "learning_rate": 7.983026695936351e-05, "loss": 0.058, "step": 33170 }, { "grad_norm": 0.48387280106544495, "learning_rate": 7.981699565818486e-05, "loss": 0.0532, "step": 33180 }, { "grad_norm": 0.5872290730476379, "learning_rate": 7.980372109626437e-05, "loss": 0.0539, "step": 33190 }, { "grad_norm": 0.5347816944122314, "learning_rate": 7.979044327505375e-05, "loss": 0.0486, "step": 33200 }, { "grad_norm": 0.5490474104881287, "learning_rate": 7.977716219600506e-05, "loss": 0.0722, "step": 33210 }, { "grad_norm": 0.6106656193733215, "learning_rate": 7.97638778605707e-05, "loss": 0.0585, "step": 33220 }, { "grad_norm": 0.551535964012146, "learning_rate": 7.975059027020338e-05, "loss": 0.0701, "step": 33230 }, { "grad_norm": 0.6280115246772766, "learning_rate": 7.973729942635623e-05, "loss": 0.059, "step": 33240 }, { "grad_norm": 0.6486293077468872, "learning_rate": 7.972400533048273e-05, "loss": 0.0681, "step": 33250 }, { "grad_norm": 0.377810537815094, "learning_rate": 7.97107079840367e-05, "loss": 0.0659, "step": 33260 }, { "grad_norm": 0.5833755731582642, "learning_rate": 7.969740738847231e-05, "loss": 0.0572, "step": 33270 }, { "grad_norm": 0.6209945678710938, "learning_rate": 7.968410354524411e-05, "loss": 0.0594, "step": 33280 }, { "grad_norm": 0.38595491647720337, "learning_rate": 7.967079645580697e-05, "loss": 0.0492, "step": 33290 }, { "grad_norm": 0.4272417724132538, "learning_rate": 7.965748612161612e-05, "loss": 0.0527, "step": 33300 }, { "grad_norm": 0.6975584030151367, "learning_rate": 7.96441725441272e-05, "loss": 0.0698, "step": 33310 }, { "grad_norm": 0.7723827958106995, "learning_rate": 7.963085572479614e-05, "loss": 0.0604, "step": 33320 }, { "grad_norm": 0.44578924775123596, "learning_rate": 7.961753566507924e-05, "loss": 0.0499, "step": 33330 }, { "grad_norm": 0.4235065281391144, "learning_rate": 7.960421236643316e-05, "loss": 0.0619, "step": 33340 }, { "grad_norm": 0.5128659605979919, "learning_rate": 7.959088583031496e-05, "loss": 0.0582, "step": 33350 }, { "grad_norm": 0.4529941976070404, "learning_rate": 7.957755605818194e-05, "loss": 0.0589, "step": 33360 }, { "grad_norm": 0.38103634119033813, "learning_rate": 7.956422305149185e-05, "loss": 0.0617, "step": 33370 }, { "grad_norm": 0.485334187746048, "learning_rate": 7.95508868117028e-05, "loss": 0.0695, "step": 33380 }, { "grad_norm": 0.5054129958152771, "learning_rate": 7.953754734027318e-05, "loss": 0.0674, "step": 33390 }, { "grad_norm": 0.42235711216926575, "learning_rate": 7.952420463866182e-05, "loss": 0.0599, "step": 33400 }, { "grad_norm": 0.49631819128990173, "learning_rate": 7.951085870832782e-05, "loss": 0.0755, "step": 33410 }, { "grad_norm": 0.5444177389144897, "learning_rate": 7.949750955073067e-05, "loss": 0.0569, "step": 33420 }, { "grad_norm": 0.5024957060813904, "learning_rate": 7.948415716733022e-05, "loss": 0.0623, "step": 33430 }, { "grad_norm": 0.45071646571159363, "learning_rate": 7.947080155958669e-05, "loss": 0.0642, "step": 33440 }, { "grad_norm": 0.5658944845199585, "learning_rate": 7.94574427289606e-05, "loss": 0.0603, "step": 33450 }, { "grad_norm": 0.4615077078342438, "learning_rate": 7.944408067691284e-05, "loss": 0.0561, "step": 33460 }, { "grad_norm": 0.5920742750167847, "learning_rate": 7.943071540490473e-05, "loss": 0.0664, "step": 33470 }, { "grad_norm": 0.479606568813324, "learning_rate": 7.94173469143978e-05, "loss": 0.0701, "step": 33480 }, { "grad_norm": 0.26702457666397095, "learning_rate": 7.940397520685406e-05, "loss": 0.0516, "step": 33490 }, { "grad_norm": 0.41697290539741516, "learning_rate": 7.939060028373577e-05, "loss": 0.0676, "step": 33500 }, { "grad_norm": 0.928337574005127, "learning_rate": 7.937722214650565e-05, "loss": 0.0737, "step": 33510 }, { "grad_norm": 0.3834591805934906, "learning_rate": 7.936384079662666e-05, "loss": 0.0626, "step": 33520 }, { "grad_norm": 0.5959516763687134, "learning_rate": 7.93504562355622e-05, "loss": 0.0571, "step": 33530 }, { "grad_norm": 0.4746156334877014, "learning_rate": 7.933706846477599e-05, "loss": 0.0569, "step": 33540 }, { "grad_norm": 0.44060075283050537, "learning_rate": 7.932367748573206e-05, "loss": 0.0588, "step": 33550 }, { "grad_norm": 0.480275958776474, "learning_rate": 7.931028329989485e-05, "loss": 0.0553, "step": 33560 }, { "grad_norm": 0.5194000601768494, "learning_rate": 7.929688590872913e-05, "loss": 0.078, "step": 33570 }, { "grad_norm": 0.4819481074810028, "learning_rate": 7.928348531370003e-05, "loss": 0.0551, "step": 33580 }, { "grad_norm": 0.6500222682952881, "learning_rate": 7.927008151627297e-05, "loss": 0.0637, "step": 33590 }, { "grad_norm": 0.4228934347629547, "learning_rate": 7.925667451791383e-05, "loss": 0.0571, "step": 33600 }, { "grad_norm": 0.583236813545227, "learning_rate": 7.924326432008874e-05, "loss": 0.0588, "step": 33610 }, { "grad_norm": 0.420858770608902, "learning_rate": 7.922985092426422e-05, "loss": 0.0604, "step": 33620 }, { "grad_norm": 0.7565398216247559, "learning_rate": 7.921643433190717e-05, "loss": 0.0798, "step": 33630 }, { "grad_norm": 1.068621277809143, "learning_rate": 7.920301454448478e-05, "loss": 0.0724, "step": 33640 }, { "grad_norm": 0.5362685322761536, "learning_rate": 7.918959156346461e-05, "loss": 0.0547, "step": 33650 }, { "grad_norm": 1.1006035804748535, "learning_rate": 7.91761653903146e-05, "loss": 0.0561, "step": 33660 }, { "grad_norm": 0.4988742172718048, "learning_rate": 7.916273602650302e-05, "loss": 0.0624, "step": 33670 }, { "grad_norm": 0.5556603074073792, "learning_rate": 7.914930347349847e-05, "loss": 0.0622, "step": 33680 }, { "grad_norm": 0.5494878888130188, "learning_rate": 7.913586773276992e-05, "loss": 0.0644, "step": 33690 }, { "grad_norm": 0.44590726494789124, "learning_rate": 7.912242880578667e-05, "loss": 0.0535, "step": 33700 }, { "grad_norm": 0.4107186198234558, "learning_rate": 7.910898669401839e-05, "loss": 0.0683, "step": 33710 }, { "grad_norm": 0.4134231209754944, "learning_rate": 7.909554139893511e-05, "loss": 0.0595, "step": 33720 }, { "grad_norm": 0.33651888370513916, "learning_rate": 7.908209292200715e-05, "loss": 0.0522, "step": 33730 }, { "grad_norm": 0.7182907462120056, "learning_rate": 7.906864126470523e-05, "loss": 0.0578, "step": 33740 }, { "grad_norm": 1.228399634361267, "learning_rate": 7.905518642850041e-05, "loss": 0.0694, "step": 33750 }, { "grad_norm": 0.4218021631240845, "learning_rate": 7.904172841486409e-05, "loss": 0.0642, "step": 33760 }, { "grad_norm": 0.8018402457237244, "learning_rate": 7.902826722526801e-05, "loss": 0.0661, "step": 33770 }, { "grad_norm": 0.40452927350997925, "learning_rate": 7.901480286118427e-05, "loss": 0.06, "step": 33780 }, { "grad_norm": 0.430255651473999, "learning_rate": 7.900133532408531e-05, "loss": 0.0538, "step": 33790 }, { "grad_norm": 0.5305464863777161, "learning_rate": 7.898786461544395e-05, "loss": 0.0507, "step": 33800 }, { "grad_norm": 1.0992878675460815, "learning_rate": 7.897439073673325e-05, "loss": 0.0712, "step": 33810 }, { "grad_norm": 0.5219168663024902, "learning_rate": 7.896091368942677e-05, "loss": 0.0662, "step": 33820 }, { "grad_norm": 0.7424443364143372, "learning_rate": 7.894743347499832e-05, "loss": 0.0609, "step": 33830 }, { "grad_norm": 0.4373152554035187, "learning_rate": 7.893395009492203e-05, "loss": 0.0669, "step": 33840 }, { "grad_norm": 0.397918701171875, "learning_rate": 7.892046355067248e-05, "loss": 0.0606, "step": 33850 }, { "grad_norm": 0.5054548382759094, "learning_rate": 7.890697384372451e-05, "loss": 0.0525, "step": 33860 }, { "grad_norm": 0.4437214434146881, "learning_rate": 7.889348097555336e-05, "loss": 0.0512, "step": 33870 }, { "grad_norm": 0.5116510987281799, "learning_rate": 7.887998494763455e-05, "loss": 0.0587, "step": 33880 }, { "grad_norm": 0.6502493023872375, "learning_rate": 7.886648576144404e-05, "loss": 0.0689, "step": 33890 }, { "grad_norm": 0.40294063091278076, "learning_rate": 7.885298341845802e-05, "loss": 0.0592, "step": 33900 }, { "grad_norm": 0.5273995995521545, "learning_rate": 7.883947792015311e-05, "loss": 0.0644, "step": 33910 }, { "grad_norm": 0.6000047922134399, "learning_rate": 7.882596926800628e-05, "loss": 0.0842, "step": 33920 }, { "grad_norm": 0.5235999822616577, "learning_rate": 7.881245746349477e-05, "loss": 0.0627, "step": 33930 }, { "grad_norm": 0.786505401134491, "learning_rate": 7.879894250809623e-05, "loss": 0.0605, "step": 33940 }, { "grad_norm": 0.6039375066757202, "learning_rate": 7.878542440328865e-05, "loss": 0.0768, "step": 33950 }, { "grad_norm": 0.40413403511047363, "learning_rate": 7.877190315055031e-05, "loss": 0.0432, "step": 33960 }, { "grad_norm": 0.49289464950561523, "learning_rate": 7.875837875135991e-05, "loss": 0.0535, "step": 33970 }, { "grad_norm": 0.4711160957813263, "learning_rate": 7.874485120719646e-05, "loss": 0.0619, "step": 33980 }, { "grad_norm": 0.5176963806152344, "learning_rate": 7.873132051953928e-05, "loss": 0.0598, "step": 33990 }, { "grad_norm": 1.0009936094284058, "learning_rate": 7.87177866898681e-05, "loss": 0.0678, "step": 34000 }, { "grad_norm": 0.46247145533561707, "learning_rate": 7.870424971966294e-05, "loss": 0.0577, "step": 34010 }, { "grad_norm": 0.37799230217933655, "learning_rate": 7.869070961040419e-05, "loss": 0.0477, "step": 34020 }, { "grad_norm": 0.4449516534805298, "learning_rate": 7.867716636357257e-05, "loss": 0.0647, "step": 34030 }, { "grad_norm": 0.4492364525794983, "learning_rate": 7.866361998064915e-05, "loss": 0.0685, "step": 34040 }, { "grad_norm": 0.6610226035118103, "learning_rate": 7.865007046311534e-05, "loss": 0.057, "step": 34050 }, { "grad_norm": 0.5898067951202393, "learning_rate": 7.86365178124529e-05, "loss": 0.0595, "step": 34060 }, { "grad_norm": 0.5599454045295715, "learning_rate": 7.862296203014394e-05, "loss": 0.0597, "step": 34070 }, { "grad_norm": 0.5245625972747803, "learning_rate": 7.860940311767088e-05, "loss": 0.065, "step": 34080 }, { "grad_norm": 0.5855538845062256, "learning_rate": 7.85958410765165e-05, "loss": 0.0551, "step": 34090 }, { "grad_norm": 0.6009992361068726, "learning_rate": 7.858227590816394e-05, "loss": 0.0672, "step": 34100 }, { "grad_norm": 0.4341205358505249, "learning_rate": 7.856870761409664e-05, "loss": 0.0572, "step": 34110 }, { "grad_norm": 0.6338578462600708, "learning_rate": 7.855513619579846e-05, "loss": 0.0639, "step": 34120 }, { "grad_norm": 0.8993684649467468, "learning_rate": 7.85415616547535e-05, "loss": 0.0581, "step": 34130 }, { "grad_norm": 0.4356965720653534, "learning_rate": 7.852798399244627e-05, "loss": 0.0504, "step": 34140 }, { "grad_norm": 0.5332077145576477, "learning_rate": 7.851440321036161e-05, "loss": 0.0625, "step": 34150 }, { "grad_norm": 0.6253144145011902, "learning_rate": 7.850081930998468e-05, "loss": 0.0555, "step": 34160 }, { "grad_norm": 0.9590504169464111, "learning_rate": 7.8487232292801e-05, "loss": 0.0752, "step": 34170 }, { "grad_norm": 0.7867175340652466, "learning_rate": 7.847364216029642e-05, "loss": 0.0745, "step": 34180 }, { "grad_norm": 0.4882447123527527, "learning_rate": 7.846004891395716e-05, "loss": 0.0621, "step": 34190 }, { "grad_norm": 0.6381310820579529, "learning_rate": 7.844645255526972e-05, "loss": 0.0682, "step": 34200 }, { "grad_norm": 0.5307512879371643, "learning_rate": 7.843285308572101e-05, "loss": 0.0586, "step": 34210 }, { "grad_norm": 0.47732603549957275, "learning_rate": 7.841925050679823e-05, "loss": 0.0464, "step": 34220 }, { "grad_norm": 0.6437689661979675, "learning_rate": 7.840564481998895e-05, "loss": 0.0644, "step": 34230 }, { "grad_norm": 1.1212005615234375, "learning_rate": 7.839203602678105e-05, "loss": 0.0662, "step": 34240 }, { "grad_norm": 0.5811456441879272, "learning_rate": 7.837842412866279e-05, "loss": 0.0696, "step": 34250 }, { "grad_norm": 0.37479445338249207, "learning_rate": 7.836480912712272e-05, "loss": 0.0665, "step": 34260 }, { "grad_norm": 0.6939439177513123, "learning_rate": 7.835119102364976e-05, "loss": 0.052, "step": 34270 }, { "grad_norm": 3.1426992416381836, "learning_rate": 7.83375698197332e-05, "loss": 0.0591, "step": 34280 }, { "grad_norm": 0.9039314389228821, "learning_rate": 7.83239455168626e-05, "loss": 0.0626, "step": 34290 }, { "grad_norm": 0.5268450379371643, "learning_rate": 7.83103181165279e-05, "loss": 0.0674, "step": 34300 }, { "grad_norm": 0.6120365858078003, "learning_rate": 7.829668762021937e-05, "loss": 0.0764, "step": 34310 }, { "grad_norm": 0.6870862245559692, "learning_rate": 7.828305402942764e-05, "loss": 0.0605, "step": 34320 }, { "grad_norm": 0.8220399022102356, "learning_rate": 7.826941734564363e-05, "loss": 0.0542, "step": 34330 }, { "grad_norm": 0.41431954503059387, "learning_rate": 7.825577757035865e-05, "loss": 0.0697, "step": 34340 }, { "grad_norm": 0.4803438186645508, "learning_rate": 7.824213470506431e-05, "loss": 0.0595, "step": 34350 }, { "grad_norm": 0.7436808347702026, "learning_rate": 7.822848875125257e-05, "loss": 0.0639, "step": 34360 }, { "grad_norm": 0.46496298909187317, "learning_rate": 7.821483971041576e-05, "loss": 0.0643, "step": 34370 }, { "grad_norm": 0.47279104590415955, "learning_rate": 7.820118758404649e-05, "loss": 0.0509, "step": 34380 }, { "grad_norm": 0.34669601917266846, "learning_rate": 7.818753237363776e-05, "loss": 0.0582, "step": 34390 }, { "grad_norm": 0.5338414311408997, "learning_rate": 7.817387408068286e-05, "loss": 0.0609, "step": 34400 }, { "grad_norm": 0.3415243327617645, "learning_rate": 7.816021270667544e-05, "loss": 0.0529, "step": 34410 }, { "grad_norm": 0.41161417961120605, "learning_rate": 7.81465482531095e-05, "loss": 0.0551, "step": 34420 }, { "grad_norm": 1.1708122491836548, "learning_rate": 7.813288072147938e-05, "loss": 0.0669, "step": 34430 }, { "grad_norm": 0.6848344802856445, "learning_rate": 7.811921011327972e-05, "loss": 0.0629, "step": 34440 }, { "grad_norm": 0.37544000148773193, "learning_rate": 7.810553643000549e-05, "loss": 0.0572, "step": 34450 }, { "grad_norm": 0.6216089725494385, "learning_rate": 7.809185967315206e-05, "loss": 0.0589, "step": 34460 }, { "grad_norm": 0.40621984004974365, "learning_rate": 7.80781798442151e-05, "loss": 0.0606, "step": 34470 }, { "grad_norm": 0.8093766570091248, "learning_rate": 7.806449694469061e-05, "loss": 0.0513, "step": 34480 }, { "grad_norm": 0.49143433570861816, "learning_rate": 7.805081097607492e-05, "loss": 0.0645, "step": 34490 }, { "grad_norm": 0.615784227848053, "learning_rate": 7.803712193986474e-05, "loss": 0.0604, "step": 34500 }, { "grad_norm": 0.5155499577522278, "learning_rate": 7.802342983755702e-05, "loss": 0.0549, "step": 34510 }, { "grad_norm": 0.6136842966079712, "learning_rate": 7.800973467064918e-05, "loss": 0.0681, "step": 34520 }, { "grad_norm": 0.4172350764274597, "learning_rate": 7.799603644063884e-05, "loss": 0.0568, "step": 34530 }, { "grad_norm": 0.38717958331108093, "learning_rate": 7.798233514902405e-05, "loss": 0.0522, "step": 34540 }, { "grad_norm": 0.7363174557685852, "learning_rate": 7.796863079730318e-05, "loss": 0.0619, "step": 34550 }, { "grad_norm": 0.48778989911079407, "learning_rate": 7.795492338697488e-05, "loss": 0.0576, "step": 34560 }, { "grad_norm": 0.5220041275024414, "learning_rate": 7.794121291953819e-05, "loss": 0.062, "step": 34570 }, { "grad_norm": 0.7365941405296326, "learning_rate": 7.792749939649246e-05, "loss": 0.074, "step": 34580 }, { "grad_norm": 0.42065009474754333, "learning_rate": 7.79137828193374e-05, "loss": 0.0471, "step": 34590 }, { "grad_norm": 0.6304804086685181, "learning_rate": 7.790006318957301e-05, "loss": 0.0702, "step": 34600 }, { "grad_norm": 0.42607927322387695, "learning_rate": 7.788634050869965e-05, "loss": 0.0548, "step": 34610 }, { "grad_norm": 0.6868102550506592, "learning_rate": 7.787261477821803e-05, "loss": 0.0547, "step": 34620 }, { "grad_norm": 0.688524603843689, "learning_rate": 7.785888599962916e-05, "loss": 0.0514, "step": 34630 }, { "grad_norm": 0.8552944660186768, "learning_rate": 7.784515417443439e-05, "loss": 0.0572, "step": 34640 }, { "grad_norm": 0.536510169506073, "learning_rate": 7.783141930413545e-05, "loss": 0.057, "step": 34650 }, { "grad_norm": 0.46897435188293457, "learning_rate": 7.78176813902343e-05, "loss": 0.0675, "step": 34660 }, { "grad_norm": 0.5173875689506531, "learning_rate": 7.780394043423336e-05, "loss": 0.0548, "step": 34670 }, { "grad_norm": 0.5538386702537537, "learning_rate": 7.77901964376353e-05, "loss": 0.0713, "step": 34680 }, { "grad_norm": 0.5375638604164124, "learning_rate": 7.777644940194316e-05, "loss": 0.0563, "step": 34690 }, { "grad_norm": 0.7312757968902588, "learning_rate": 7.776269932866023e-05, "loss": 0.0615, "step": 34700 }, { "grad_norm": 0.4227868914604187, "learning_rate": 7.774894621929026e-05, "loss": 0.0628, "step": 34710 }, { "grad_norm": 0.4625406563282013, "learning_rate": 7.773519007533725e-05, "loss": 0.0748, "step": 34720 }, { "grad_norm": 1.785386562347412, "learning_rate": 7.772143089830556e-05, "loss": 0.0502, "step": 34730 }, { "grad_norm": 0.36965781450271606, "learning_rate": 7.770766868969985e-05, "loss": 0.0541, "step": 34740 }, { "grad_norm": 0.32312893867492676, "learning_rate": 7.769390345102518e-05, "loss": 0.0671, "step": 34750 }, { "grad_norm": 0.7861494421958923, "learning_rate": 7.768013518378683e-05, "loss": 0.0666, "step": 34760 }, { "grad_norm": 0.48542320728302, "learning_rate": 7.766636388949053e-05, "loss": 0.0463, "step": 34770 }, { "grad_norm": 0.49095672369003296, "learning_rate": 7.765258956964229e-05, "loss": 0.0664, "step": 34780 }, { "grad_norm": 0.40110236406326294, "learning_rate": 7.76388122257484e-05, "loss": 0.0615, "step": 34790 }, { "grad_norm": 0.3523491322994232, "learning_rate": 7.762503185931558e-05, "loss": 0.0517, "step": 34800 }, { "grad_norm": 0.47389110922813416, "learning_rate": 7.76112484718508e-05, "loss": 0.0593, "step": 34810 }, { "grad_norm": 0.4659843444824219, "learning_rate": 7.75974620648614e-05, "loss": 0.0644, "step": 34820 }, { "grad_norm": 0.5567617416381836, "learning_rate": 7.758367263985503e-05, "loss": 0.0555, "step": 34830 }, { "grad_norm": 0.5952120423316956, "learning_rate": 7.75698801983397e-05, "loss": 0.0491, "step": 34840 }, { "grad_norm": 0.37037497758865356, "learning_rate": 7.755608474182372e-05, "loss": 0.0652, "step": 34850 }, { "grad_norm": 0.5747390389442444, "learning_rate": 7.754228627181574e-05, "loss": 0.0639, "step": 34860 }, { "grad_norm": 0.3919229209423065, "learning_rate": 7.752848478982476e-05, "loss": 0.0518, "step": 34870 }, { "grad_norm": 0.4561506509780884, "learning_rate": 7.751468029736006e-05, "loss": 0.0538, "step": 34880 }, { "grad_norm": 0.4977453351020813, "learning_rate": 7.750087279593129e-05, "loss": 0.0688, "step": 34890 }, { "grad_norm": 0.36097198724746704, "learning_rate": 7.748706228704843e-05, "loss": 0.0461, "step": 34900 }, { "grad_norm": 0.44683024287223816, "learning_rate": 7.747324877222176e-05, "loss": 0.0655, "step": 34910 }, { "grad_norm": 0.636879026889801, "learning_rate": 7.745943225296188e-05, "loss": 0.0712, "step": 34920 }, { "grad_norm": 0.4575721025466919, "learning_rate": 7.744561273077981e-05, "loss": 0.0442, "step": 34930 }, { "grad_norm": 0.31671419739723206, "learning_rate": 7.743179020718678e-05, "loss": 0.0501, "step": 34940 }, { "grad_norm": 0.5010847449302673, "learning_rate": 7.741796468369443e-05, "loss": 0.055, "step": 34950 }, { "grad_norm": 0.5181582570075989, "learning_rate": 7.740413616181466e-05, "loss": 0.0624, "step": 34960 }, { "grad_norm": 0.5066781640052795, "learning_rate": 7.739030464305978e-05, "loss": 0.063, "step": 34970 }, { "grad_norm": 1.0387446880340576, "learning_rate": 7.737647012894235e-05, "loss": 0.0745, "step": 34980 }, { "grad_norm": 0.7454912066459656, "learning_rate": 7.736263262097532e-05, "loss": 0.0669, "step": 34990 }, { "grad_norm": 0.4979926645755768, "learning_rate": 7.734879212067192e-05, "loss": 0.0642, "step": 35000 }, { "grad_norm": 0.583851158618927, "learning_rate": 7.733494862954573e-05, "loss": 0.0613, "step": 35010 }, { "grad_norm": 0.4980171024799347, "learning_rate": 7.732110214911066e-05, "loss": 0.0472, "step": 35020 }, { "grad_norm": 0.8186679482460022, "learning_rate": 7.730725268088094e-05, "loss": 0.0677, "step": 35030 }, { "grad_norm": 0.3656993806362152, "learning_rate": 7.729340022637112e-05, "loss": 0.0563, "step": 35040 }, { "grad_norm": 0.5556840896606445, "learning_rate": 7.727954478709607e-05, "loss": 0.0572, "step": 35050 }, { "grad_norm": 0.7798149585723877, "learning_rate": 7.726568636457103e-05, "loss": 0.066, "step": 35060 }, { "grad_norm": 0.6220732927322388, "learning_rate": 7.725182496031153e-05, "loss": 0.0716, "step": 35070 }, { "grad_norm": 0.48107439279556274, "learning_rate": 7.723796057583341e-05, "loss": 0.0699, "step": 35080 }, { "grad_norm": 0.8147640824317932, "learning_rate": 7.722409321265287e-05, "loss": 0.0657, "step": 35090 }, { "grad_norm": 0.438907653093338, "learning_rate": 7.721022287228645e-05, "loss": 0.0753, "step": 35100 }, { "grad_norm": 0.6055101156234741, "learning_rate": 7.719634955625094e-05, "loss": 0.0494, "step": 35110 }, { "grad_norm": 0.8030952215194702, "learning_rate": 7.718247326606355e-05, "loss": 0.0545, "step": 35120 }, { "grad_norm": 0.507194995880127, "learning_rate": 7.716859400324175e-05, "loss": 0.0607, "step": 35130 }, { "grad_norm": 0.5560539960861206, "learning_rate": 7.715471176930335e-05, "loss": 0.0762, "step": 35140 }, { "grad_norm": 0.4628041684627533, "learning_rate": 7.714082656576651e-05, "loss": 0.0599, "step": 35150 }, { "grad_norm": 0.5156897902488708, "learning_rate": 7.712693839414968e-05, "loss": 0.0546, "step": 35160 }, { "grad_norm": 0.3711074888706207, "learning_rate": 7.711304725597164e-05, "loss": 0.0468, "step": 35170 }, { "grad_norm": 0.4887494742870331, "learning_rate": 7.709915315275151e-05, "loss": 0.0672, "step": 35180 }, { "grad_norm": 0.4337628185749054, "learning_rate": 7.708525608600876e-05, "loss": 0.0567, "step": 35190 }, { "grad_norm": 0.4100872576236725, "learning_rate": 7.707135605726311e-05, "loss": 0.0629, "step": 35200 }, { "grad_norm": 0.6484190821647644, "learning_rate": 7.705745306803466e-05, "loss": 0.0911, "step": 35210 }, { "grad_norm": 0.6358450651168823, "learning_rate": 7.704354711984383e-05, "loss": 0.0592, "step": 35220 }, { "grad_norm": 0.5662591457366943, "learning_rate": 7.702963821421133e-05, "loss": 0.0574, "step": 35230 }, { "grad_norm": 0.8126494884490967, "learning_rate": 7.701572635265826e-05, "loss": 0.0573, "step": 35240 }, { "grad_norm": 0.518150269985199, "learning_rate": 7.700181153670596e-05, "loss": 0.0459, "step": 35250 }, { "grad_norm": 0.5309092402458191, "learning_rate": 7.698789376787616e-05, "loss": 0.0621, "step": 35260 }, { "grad_norm": 0.930690586566925, "learning_rate": 7.697397304769085e-05, "loss": 0.0628, "step": 35270 }, { "grad_norm": 1.8658279180526733, "learning_rate": 7.696004937767241e-05, "loss": 0.0691, "step": 35280 }, { "grad_norm": 0.3864899277687073, "learning_rate": 7.694612275934352e-05, "loss": 0.0625, "step": 35290 }, { "grad_norm": 0.5071852207183838, "learning_rate": 7.693219319422714e-05, "loss": 0.0643, "step": 35300 }, { "grad_norm": 0.4378501772880554, "learning_rate": 7.69182606838466e-05, "loss": 0.0586, "step": 35310 }, { "grad_norm": 0.5503883361816406, "learning_rate": 7.690432522972558e-05, "loss": 0.0568, "step": 35320 }, { "grad_norm": 0.3689426779747009, "learning_rate": 7.689038683338796e-05, "loss": 0.0555, "step": 35330 }, { "grad_norm": 0.49277374148368835, "learning_rate": 7.687644549635808e-05, "loss": 0.058, "step": 35340 }, { "grad_norm": 1.9758225679397583, "learning_rate": 7.686250122016053e-05, "loss": 0.0502, "step": 35350 }, { "grad_norm": 0.7565080523490906, "learning_rate": 7.684855400632022e-05, "loss": 0.0582, "step": 35360 }, { "grad_norm": 1.0196882486343384, "learning_rate": 7.683460385636243e-05, "loss": 0.0454, "step": 35370 }, { "grad_norm": 0.5047420859336853, "learning_rate": 7.68206507718127e-05, "loss": 0.0612, "step": 35380 }, { "grad_norm": 0.5823501348495483, "learning_rate": 7.680669475419693e-05, "loss": 0.0693, "step": 35390 }, { "grad_norm": 0.4698884189128876, "learning_rate": 7.679273580504132e-05, "loss": 0.0432, "step": 35400 }, { "grad_norm": 0.46181511878967285, "learning_rate": 7.67787739258724e-05, "loss": 0.0676, "step": 35410 }, { "grad_norm": 0.4133344292640686, "learning_rate": 7.676480911821705e-05, "loss": 0.0676, "step": 35420 }, { "grad_norm": 0.452174574136734, "learning_rate": 7.675084138360238e-05, "loss": 0.0573, "step": 35430 }, { "grad_norm": 0.715514063835144, "learning_rate": 7.673687072355592e-05, "loss": 0.0593, "step": 35440 }, { "grad_norm": 0.5544741153717041, "learning_rate": 7.67228971396055e-05, "loss": 0.0525, "step": 35450 }, { "grad_norm": 0.5847429633140564, "learning_rate": 7.670892063327922e-05, "loss": 0.0606, "step": 35460 }, { "grad_norm": 0.7104434370994568, "learning_rate": 7.669494120610552e-05, "loss": 0.0486, "step": 35470 }, { "grad_norm": 0.3293127417564392, "learning_rate": 7.66809588596132e-05, "loss": 0.0459, "step": 35480 }, { "grad_norm": 0.4350021481513977, "learning_rate": 7.666697359533132e-05, "loss": 0.0574, "step": 35490 }, { "grad_norm": 1.4102534055709839, "learning_rate": 7.665298541478932e-05, "loss": 0.0653, "step": 35500 }, { "grad_norm": 0.5813084840774536, "learning_rate": 7.66389943195169e-05, "loss": 0.0588, "step": 35510 }, { "grad_norm": 0.5004801154136658, "learning_rate": 7.66250003110441e-05, "loss": 0.0696, "step": 35520 }, { "grad_norm": 0.5474352240562439, "learning_rate": 7.66110033909013e-05, "loss": 0.0556, "step": 35530 }, { "grad_norm": 0.6413968205451965, "learning_rate": 7.659700356061918e-05, "loss": 0.0635, "step": 35540 }, { "grad_norm": 0.4915740191936493, "learning_rate": 7.658300082172875e-05, "loss": 0.0502, "step": 35550 }, { "grad_norm": 1.146230697631836, "learning_rate": 7.65689951757613e-05, "loss": 0.058, "step": 35560 }, { "grad_norm": 0.4823717176914215, "learning_rate": 7.655498662424849e-05, "loss": 0.0571, "step": 35570 }, { "grad_norm": 0.4911211431026459, "learning_rate": 7.654097516872227e-05, "loss": 0.0529, "step": 35580 }, { "grad_norm": 0.6415998935699463, "learning_rate": 7.65269608107149e-05, "loss": 0.0623, "step": 35590 }, { "grad_norm": 0.47363007068634033, "learning_rate": 7.6512943551759e-05, "loss": 0.0715, "step": 35600 }, { "grad_norm": 0.43738454580307007, "learning_rate": 7.649892339338743e-05, "loss": 0.052, "step": 35610 }, { "grad_norm": 0.6231657266616821, "learning_rate": 7.648490033713344e-05, "loss": 0.0392, "step": 35620 }, { "grad_norm": 0.8600793480873108, "learning_rate": 7.647087438453058e-05, "loss": 0.0644, "step": 35630 }, { "grad_norm": 0.48772895336151123, "learning_rate": 7.64568455371127e-05, "loss": 0.0551, "step": 35640 }, { "grad_norm": 0.6512594819068909, "learning_rate": 7.644281379641396e-05, "loss": 0.0516, "step": 35650 }, { "grad_norm": 0.3697707951068878, "learning_rate": 7.642877916396887e-05, "loss": 0.0559, "step": 35660 }, { "grad_norm": 0.5332645177841187, "learning_rate": 7.641474164131221e-05, "loss": 0.0617, "step": 35670 }, { "grad_norm": 0.5658048391342163, "learning_rate": 7.640070122997913e-05, "loss": 0.0517, "step": 35680 }, { "grad_norm": 0.68595951795578, "learning_rate": 7.638665793150505e-05, "loss": 0.0507, "step": 35690 }, { "grad_norm": 0.49394115805625916, "learning_rate": 7.637261174742574e-05, "loss": 0.0766, "step": 35700 }, { "grad_norm": 0.41276034712791443, "learning_rate": 7.635856267927725e-05, "loss": 0.0452, "step": 35710 }, { "grad_norm": 0.5059139728546143, "learning_rate": 7.634451072859597e-05, "loss": 0.0543, "step": 35720 }, { "grad_norm": 0.6515786647796631, "learning_rate": 7.633045589691863e-05, "loss": 0.0442, "step": 35730 }, { "grad_norm": 0.6946715116500854, "learning_rate": 7.63163981857822e-05, "loss": 0.0506, "step": 35740 }, { "grad_norm": 0.3325376510620117, "learning_rate": 7.630233759672403e-05, "loss": 0.0529, "step": 35750 }, { "grad_norm": 0.49845507740974426, "learning_rate": 7.628827413128177e-05, "loss": 0.0616, "step": 35760 }, { "grad_norm": 0.6764398217201233, "learning_rate": 7.627420779099338e-05, "loss": 0.0603, "step": 35770 }, { "grad_norm": 1.169832468032837, "learning_rate": 7.626013857739711e-05, "loss": 0.066, "step": 35780 }, { "grad_norm": 0.7316238880157471, "learning_rate": 7.624606649203158e-05, "loss": 0.0492, "step": 35790 }, { "grad_norm": 0.40116167068481445, "learning_rate": 7.623199153643569e-05, "loss": 0.0459, "step": 35800 }, { "grad_norm": 1.4793930053710938, "learning_rate": 7.621791371214863e-05, "loss": 0.0655, "step": 35810 }, { "grad_norm": 0.4928237795829773, "learning_rate": 7.620383302070995e-05, "loss": 0.0522, "step": 35820 }, { "grad_norm": 0.5479042530059814, "learning_rate": 7.61897494636595e-05, "loss": 0.0508, "step": 35830 }, { "grad_norm": 1.057695746421814, "learning_rate": 7.617566304253739e-05, "loss": 0.0588, "step": 35840 }, { "grad_norm": 0.4991917014122009, "learning_rate": 7.616157375888416e-05, "loss": 0.0565, "step": 35850 }, { "grad_norm": 0.475472092628479, "learning_rate": 7.614748161424053e-05, "loss": 0.0477, "step": 35860 }, { "grad_norm": 0.550361692905426, "learning_rate": 7.613338661014763e-05, "loss": 0.0512, "step": 35870 }, { "grad_norm": 0.7927120327949524, "learning_rate": 7.611928874814686e-05, "loss": 0.0588, "step": 35880 }, { "grad_norm": 0.5185158252716064, "learning_rate": 7.610518802977996e-05, "loss": 0.0478, "step": 35890 }, { "grad_norm": 0.7046266794204712, "learning_rate": 7.609108445658893e-05, "loss": 0.0535, "step": 35900 }, { "grad_norm": 0.45222657918930054, "learning_rate": 7.607697803011612e-05, "loss": 0.0579, "step": 35910 }, { "grad_norm": 0.610623836517334, "learning_rate": 7.606286875190421e-05, "loss": 0.0617, "step": 35920 }, { "grad_norm": 0.4628243148326874, "learning_rate": 7.604875662349617e-05, "loss": 0.0448, "step": 35930 }, { "grad_norm": 0.3936556875705719, "learning_rate": 7.603464164643525e-05, "loss": 0.063, "step": 35940 }, { "grad_norm": 0.8709924221038818, "learning_rate": 7.602052382226507e-05, "loss": 0.065, "step": 35950 }, { "grad_norm": 0.41415274143218994, "learning_rate": 7.600640315252954e-05, "loss": 0.0595, "step": 35960 }, { "grad_norm": 0.39889469742774963, "learning_rate": 7.599227963877284e-05, "loss": 0.0649, "step": 35970 }, { "grad_norm": 0.6624886393547058, "learning_rate": 7.597815328253954e-05, "loss": 0.083, "step": 35980 }, { "grad_norm": 0.47711095213890076, "learning_rate": 7.596402408537443e-05, "loss": 0.0547, "step": 35990 }, { "grad_norm": 0.5281511545181274, "learning_rate": 7.59498920488227e-05, "loss": 0.065, "step": 36000 }, { "grad_norm": 1.7341952323913574, "learning_rate": 7.593575717442979e-05, "loss": 0.0615, "step": 36010 }, { "grad_norm": 0.6165884733200073, "learning_rate": 7.592161946374147e-05, "loss": 0.0636, "step": 36020 }, { "grad_norm": 0.4597591161727905, "learning_rate": 7.590747891830381e-05, "loss": 0.0534, "step": 36030 }, { "grad_norm": 0.6550778746604919, "learning_rate": 7.58933355396632e-05, "loss": 0.0552, "step": 36040 }, { "grad_norm": 0.4888626039028168, "learning_rate": 7.587918932936636e-05, "loss": 0.0604, "step": 36050 }, { "grad_norm": 0.397521436214447, "learning_rate": 7.586504028896028e-05, "loss": 0.0599, "step": 36060 }, { "grad_norm": 0.6222014427185059, "learning_rate": 7.585088841999228e-05, "loss": 0.0543, "step": 36070 }, { "grad_norm": 0.5935045480728149, "learning_rate": 7.583673372400999e-05, "loss": 0.0622, "step": 36080 }, { "grad_norm": 0.612005889415741, "learning_rate": 7.582257620256134e-05, "loss": 0.0577, "step": 36090 }, { "grad_norm": 0.5640246868133545, "learning_rate": 7.580841585719458e-05, "loss": 0.0631, "step": 36100 }, { "grad_norm": 0.7188739776611328, "learning_rate": 7.579425268945825e-05, "loss": 0.0686, "step": 36110 }, { "grad_norm": 1.3829845190048218, "learning_rate": 7.578008670090127e-05, "loss": 0.054, "step": 36120 }, { "grad_norm": 0.45267316699028015, "learning_rate": 7.576591789307272e-05, "loss": 0.0499, "step": 36130 }, { "grad_norm": 0.3370497524738312, "learning_rate": 7.575174626752216e-05, "loss": 0.0407, "step": 36140 }, { "grad_norm": 1.234474539756775, "learning_rate": 7.573757182579934e-05, "loss": 0.0568, "step": 36150 }, { "grad_norm": 0.3859769105911255, "learning_rate": 7.572339456945435e-05, "loss": 0.0526, "step": 36160 }, { "grad_norm": 0.6392388939857483, "learning_rate": 7.570921450003762e-05, "loss": 0.062, "step": 36170 }, { "grad_norm": 0.4745146632194519, "learning_rate": 7.569503161909983e-05, "loss": 0.0509, "step": 36180 }, { "grad_norm": 0.6571504473686218, "learning_rate": 7.568084592819202e-05, "loss": 0.059, "step": 36190 }, { "grad_norm": 1.7753052711486816, "learning_rate": 7.566665742886551e-05, "loss": 0.0592, "step": 36200 }, { "grad_norm": 0.5702481865882874, "learning_rate": 7.565246612267194e-05, "loss": 0.0581, "step": 36210 }, { "grad_norm": 1.4195789098739624, "learning_rate": 7.563827201116325e-05, "loss": 0.072, "step": 36220 }, { "grad_norm": 0.509972095489502, "learning_rate": 7.562407509589164e-05, "loss": 0.0605, "step": 36230 }, { "grad_norm": 0.4966392517089844, "learning_rate": 7.560987537840973e-05, "loss": 0.0618, "step": 36240 }, { "grad_norm": 0.5433259010314941, "learning_rate": 7.559567286027036e-05, "loss": 0.0532, "step": 36250 }, { "grad_norm": 0.8949564695358276, "learning_rate": 7.558146754302668e-05, "loss": 0.0663, "step": 36260 }, { "grad_norm": 0.6721745133399963, "learning_rate": 7.556725942823217e-05, "loss": 0.0528, "step": 36270 }, { "grad_norm": 0.7002898454666138, "learning_rate": 7.555304851744061e-05, "loss": 0.0682, "step": 36280 }, { "grad_norm": 0.9478493332862854, "learning_rate": 7.553883481220608e-05, "loss": 0.0697, "step": 36290 }, { "grad_norm": 0.6932003498077393, "learning_rate": 7.552461831408298e-05, "loss": 0.0522, "step": 36300 }, { "grad_norm": 0.7543896436691284, "learning_rate": 7.551039902462599e-05, "loss": 0.0548, "step": 36310 }, { "grad_norm": 0.48297426104545593, "learning_rate": 7.549617694539011e-05, "loss": 0.042, "step": 36320 }, { "grad_norm": 0.6416827440261841, "learning_rate": 7.548195207793066e-05, "loss": 0.0582, "step": 36330 }, { "grad_norm": 0.8835012912750244, "learning_rate": 7.546772442380323e-05, "loss": 0.0532, "step": 36340 }, { "grad_norm": 0.49644944071769714, "learning_rate": 7.545349398456376e-05, "loss": 0.073, "step": 36350 }, { "grad_norm": 0.5418319702148438, "learning_rate": 7.543926076176845e-05, "loss": 0.0607, "step": 36360 }, { "grad_norm": 0.3686094582080841, "learning_rate": 7.542502475697385e-05, "loss": 0.0585, "step": 36370 }, { "grad_norm": 0.5086994171142578, "learning_rate": 7.541078597173675e-05, "loss": 0.0638, "step": 36380 }, { "grad_norm": 0.5888609886169434, "learning_rate": 7.539654440761431e-05, "loss": 0.0665, "step": 36390 }, { "grad_norm": 0.484663724899292, "learning_rate": 7.538230006616395e-05, "loss": 0.0518, "step": 36400 }, { "grad_norm": 0.39553454518318176, "learning_rate": 7.536805294894342e-05, "loss": 0.0681, "step": 36410 }, { "grad_norm": 1.1816558837890625, "learning_rate": 7.535380305751076e-05, "loss": 0.0515, "step": 36420 }, { "grad_norm": 0.6519618034362793, "learning_rate": 7.533955039342431e-05, "loss": 0.0639, "step": 36430 }, { "grad_norm": 0.7383256554603577, "learning_rate": 7.532529495824274e-05, "loss": 0.0666, "step": 36440 }, { "grad_norm": 0.6753697395324707, "learning_rate": 7.531103675352497e-05, "loss": 0.0557, "step": 36450 }, { "grad_norm": 0.49353793263435364, "learning_rate": 7.529677578083029e-05, "loss": 0.0633, "step": 36460 }, { "grad_norm": 0.6922683715820312, "learning_rate": 7.528251204171823e-05, "loss": 0.0611, "step": 36470 }, { "grad_norm": 0.4534264802932739, "learning_rate": 7.526824553774867e-05, "loss": 0.0574, "step": 36480 }, { "grad_norm": 0.4750251770019531, "learning_rate": 7.525397627048178e-05, "loss": 0.0589, "step": 36490 }, { "grad_norm": 0.5297956466674805, "learning_rate": 7.523970424147802e-05, "loss": 0.0505, "step": 36500 }, { "grad_norm": 0.4199928343296051, "learning_rate": 7.522542945229813e-05, "loss": 0.0419, "step": 36510 }, { "grad_norm": 0.6494170427322388, "learning_rate": 7.521115190450321e-05, "loss": 0.0539, "step": 36520 }, { "grad_norm": 0.6069225668907166, "learning_rate": 7.519687159965462e-05, "loss": 0.0539, "step": 36530 }, { "grad_norm": 0.6086146235466003, "learning_rate": 7.518258853931403e-05, "loss": 0.0546, "step": 36540 }, { "grad_norm": 0.4379652440547943, "learning_rate": 7.516830272504342e-05, "loss": 0.0453, "step": 36550 }, { "grad_norm": 0.5665156245231628, "learning_rate": 7.515401415840509e-05, "loss": 0.0706, "step": 36560 }, { "grad_norm": 0.6869190335273743, "learning_rate": 7.513972284096155e-05, "loss": 0.0548, "step": 36570 }, { "grad_norm": 0.38430866599082947, "learning_rate": 7.512542877427574e-05, "loss": 0.0588, "step": 36580 }, { "grad_norm": 0.4086233377456665, "learning_rate": 7.51111319599108e-05, "loss": 0.0536, "step": 36590 }, { "grad_norm": 0.43841591477394104, "learning_rate": 7.509683239943024e-05, "loss": 0.0626, "step": 36600 }, { "grad_norm": 0.42750057578086853, "learning_rate": 7.50825300943978e-05, "loss": 0.0686, "step": 36610 }, { "grad_norm": 0.5485020279884338, "learning_rate": 7.50682250463776e-05, "loss": 0.0555, "step": 36620 }, { "grad_norm": 0.4513654112815857, "learning_rate": 7.5053917256934e-05, "loss": 0.0484, "step": 36630 }, { "grad_norm": 0.49288415908813477, "learning_rate": 7.503960672763166e-05, "loss": 0.0559, "step": 36640 }, { "grad_norm": 0.5059831142425537, "learning_rate": 7.502529346003559e-05, "loss": 0.0588, "step": 36650 }, { "grad_norm": 1.7847598791122437, "learning_rate": 7.501097745571106e-05, "loss": 0.0594, "step": 36660 }, { "grad_norm": 0.46339908242225647, "learning_rate": 7.499665871622361e-05, "loss": 0.0669, "step": 36670 }, { "grad_norm": 0.509102463722229, "learning_rate": 7.498233724313917e-05, "loss": 0.0666, "step": 36680 }, { "grad_norm": 0.4944230914115906, "learning_rate": 7.496801303802387e-05, "loss": 0.0656, "step": 36690 }, { "grad_norm": 0.46279120445251465, "learning_rate": 7.495368610244423e-05, "loss": 0.0549, "step": 36700 }, { "grad_norm": 0.7786303758621216, "learning_rate": 7.493935643796697e-05, "loss": 0.0655, "step": 36710 }, { "grad_norm": 0.48886895179748535, "learning_rate": 7.49250240461592e-05, "loss": 0.0602, "step": 36720 }, { "grad_norm": 0.45183897018432617, "learning_rate": 7.491068892858826e-05, "loss": 0.0631, "step": 36730 }, { "grad_norm": 0.5004293918609619, "learning_rate": 7.489635108682184e-05, "loss": 0.0577, "step": 36740 }, { "grad_norm": 0.5507505536079407, "learning_rate": 7.48820105224279e-05, "loss": 0.0517, "step": 36750 }, { "grad_norm": 0.6611539125442505, "learning_rate": 7.48676672369747e-05, "loss": 0.0638, "step": 36760 }, { "grad_norm": 0.6858953237533569, "learning_rate": 7.485332123203079e-05, "loss": 0.0648, "step": 36770 }, { "grad_norm": 0.39147356152534485, "learning_rate": 7.483897250916504e-05, "loss": 0.0627, "step": 36780 }, { "grad_norm": 1.189499855041504, "learning_rate": 7.48246210699466e-05, "loss": 0.0546, "step": 36790 }, { "grad_norm": 0.599064826965332, "learning_rate": 7.481026691594492e-05, "loss": 0.0515, "step": 36800 }, { "grad_norm": 0.37588006258010864, "learning_rate": 7.479591004872975e-05, "loss": 0.0482, "step": 36810 }, { "grad_norm": 0.41520410776138306, "learning_rate": 7.478155046987114e-05, "loss": 0.0585, "step": 36820 }, { "grad_norm": 0.48607462644577026, "learning_rate": 7.476718818093944e-05, "loss": 0.0558, "step": 36830 }, { "grad_norm": 0.3954785466194153, "learning_rate": 7.475282318350527e-05, "loss": 0.0517, "step": 36840 }, { "grad_norm": 0.4427589476108551, "learning_rate": 7.473845547913957e-05, "loss": 0.0456, "step": 36850 }, { "grad_norm": 0.5921582579612732, "learning_rate": 7.472408506941357e-05, "loss": 0.0747, "step": 36860 }, { "grad_norm": 0.45780396461486816, "learning_rate": 7.470971195589881e-05, "loss": 0.0516, "step": 36870 }, { "grad_norm": 0.6689750552177429, "learning_rate": 7.46953361401671e-05, "loss": 0.07, "step": 36880 }, { "grad_norm": 0.7320096492767334, "learning_rate": 7.468095762379055e-05, "loss": 0.0587, "step": 36890 }, { "grad_norm": 0.31893882155418396, "learning_rate": 7.466657640834158e-05, "loss": 0.0672, "step": 36900 }, { "grad_norm": 0.5274337530136108, "learning_rate": 7.46521924953929e-05, "loss": 0.0516, "step": 36910 }, { "grad_norm": 0.550703763961792, "learning_rate": 7.463780588651752e-05, "loss": 0.0546, "step": 36920 }, { "grad_norm": 0.6122044920921326, "learning_rate": 7.462341658328873e-05, "loss": 0.0547, "step": 36930 }, { "grad_norm": 0.4031769633293152, "learning_rate": 7.460902458728012e-05, "loss": 0.0501, "step": 36940 }, { "grad_norm": 0.48309653997421265, "learning_rate": 7.459462990006558e-05, "loss": 0.0483, "step": 36950 }, { "grad_norm": 0.5134316682815552, "learning_rate": 7.458023252321929e-05, "loss": 0.0499, "step": 36960 }, { "grad_norm": 0.6773591041564941, "learning_rate": 7.456583245831573e-05, "loss": 0.0442, "step": 36970 }, { "grad_norm": 0.5578699111938477, "learning_rate": 7.455142970692968e-05, "loss": 0.0743, "step": 36980 }, { "grad_norm": 0.5225330591201782, "learning_rate": 7.453702427063615e-05, "loss": 0.0509, "step": 36990 }, { "grad_norm": 0.6858848333358765, "learning_rate": 7.452261615101057e-05, "loss": 0.0573, "step": 37000 }, { "grad_norm": 1.2206437587738037, "learning_rate": 7.450820534962856e-05, "loss": 0.0645, "step": 37010 }, { "grad_norm": 0.44524964690208435, "learning_rate": 7.449379186806604e-05, "loss": 0.0576, "step": 37020 }, { "grad_norm": 0.5105348229408264, "learning_rate": 7.44793757078993e-05, "loss": 0.0623, "step": 37030 }, { "grad_norm": 0.44775304198265076, "learning_rate": 7.44649568707048e-05, "loss": 0.0529, "step": 37040 }, { "grad_norm": 0.7283190488815308, "learning_rate": 7.445053535805942e-05, "loss": 0.059, "step": 37050 }, { "grad_norm": 0.697671115398407, "learning_rate": 7.443611117154024e-05, "loss": 0.0561, "step": 37060 }, { "grad_norm": 1.571632981300354, "learning_rate": 7.442168431272469e-05, "loss": 0.0567, "step": 37070 }, { "grad_norm": 0.6094165444374084, "learning_rate": 7.440725478319044e-05, "loss": 0.065, "step": 37080 }, { "grad_norm": 0.7834786176681519, "learning_rate": 7.439282258451553e-05, "loss": 0.0622, "step": 37090 }, { "grad_norm": 0.7579284310340881, "learning_rate": 7.43783877182782e-05, "loss": 0.0579, "step": 37100 }, { "grad_norm": 0.37902823090553284, "learning_rate": 7.436395018605707e-05, "loss": 0.0532, "step": 37110 }, { "grad_norm": 0.6358306407928467, "learning_rate": 7.434950998943094e-05, "loss": 0.049, "step": 37120 }, { "grad_norm": 0.4618493914604187, "learning_rate": 7.433506712997902e-05, "loss": 0.0601, "step": 37130 }, { "grad_norm": 0.6714614033699036, "learning_rate": 7.432062160928077e-05, "loss": 0.0566, "step": 37140 }, { "grad_norm": 0.3259819746017456, "learning_rate": 7.430617342891588e-05, "loss": 0.0532, "step": 37150 }, { "grad_norm": 0.4101956784725189, "learning_rate": 7.429172259046442e-05, "loss": 0.055, "step": 37160 }, { "grad_norm": 0.6525675654411316, "learning_rate": 7.42772690955067e-05, "loss": 0.0586, "step": 37170 }, { "grad_norm": 0.5111818909645081, "learning_rate": 7.426281294562332e-05, "loss": 0.0711, "step": 37180 }, { "grad_norm": 0.3166458010673523, "learning_rate": 7.424835414239519e-05, "loss": 0.0671, "step": 37190 }, { "grad_norm": 0.5538613200187683, "learning_rate": 7.423389268740352e-05, "loss": 0.0583, "step": 37200 }, { "grad_norm": 0.7328143119812012, "learning_rate": 7.42194285822298e-05, "loss": 0.052, "step": 37210 }, { "grad_norm": 0.37866583466529846, "learning_rate": 7.420496182845576e-05, "loss": 0.0651, "step": 37220 }, { "grad_norm": 0.5009272694587708, "learning_rate": 7.419049242766348e-05, "loss": 0.07, "step": 37230 }, { "grad_norm": 0.686218798160553, "learning_rate": 7.417602038143533e-05, "loss": 0.0638, "step": 37240 }, { "grad_norm": 0.7754982113838196, "learning_rate": 7.416154569135393e-05, "loss": 0.0661, "step": 37250 }, { "grad_norm": 0.5999303460121155, "learning_rate": 7.414706835900222e-05, "loss": 0.0518, "step": 37260 }, { "grad_norm": 0.5831718444824219, "learning_rate": 7.413258838596345e-05, "loss": 0.052, "step": 37270 }, { "grad_norm": 0.5149441361427307, "learning_rate": 7.411810577382106e-05, "loss": 0.0613, "step": 37280 }, { "grad_norm": 0.5140244364738464, "learning_rate": 7.410362052415889e-05, "loss": 0.0603, "step": 37290 }, { "grad_norm": 0.4695221185684204, "learning_rate": 7.408913263856102e-05, "loss": 0.0629, "step": 37300 }, { "grad_norm": 0.6785141825675964, "learning_rate": 7.407464211861183e-05, "loss": 0.0723, "step": 37310 }, { "grad_norm": 0.4303184449672699, "learning_rate": 7.406014896589597e-05, "loss": 0.0524, "step": 37320 }, { "grad_norm": 0.545362651348114, "learning_rate": 7.404565318199838e-05, "loss": 0.055, "step": 37330 }, { "grad_norm": 0.5125188827514648, "learning_rate": 7.403115476850432e-05, "loss": 0.0751, "step": 37340 }, { "grad_norm": 0.4740261435508728, "learning_rate": 7.401665372699932e-05, "loss": 0.0619, "step": 37350 }, { "grad_norm": 0.7403870820999146, "learning_rate": 7.400215005906916e-05, "loss": 0.0662, "step": 37360 }, { "grad_norm": 0.5401939749717712, "learning_rate": 7.398764376629996e-05, "loss": 0.063, "step": 37370 }, { "grad_norm": 1.063560962677002, "learning_rate": 7.39731348502781e-05, "loss": 0.0618, "step": 37380 }, { "grad_norm": 0.7425305843353271, "learning_rate": 7.395862331259028e-05, "loss": 0.0588, "step": 37390 }, { "grad_norm": 0.6863535642623901, "learning_rate": 7.394410915482341e-05, "loss": 0.065, "step": 37400 }, { "grad_norm": 0.45637914538383484, "learning_rate": 7.392959237856476e-05, "loss": 0.0544, "step": 37410 }, { "grad_norm": 0.600940465927124, "learning_rate": 7.391507298540187e-05, "loss": 0.0543, "step": 37420 }, { "grad_norm": 0.408414751291275, "learning_rate": 7.390055097692257e-05, "loss": 0.0463, "step": 37430 }, { "grad_norm": 0.5301414728164673, "learning_rate": 7.388602635471494e-05, "loss": 0.0617, "step": 37440 }, { "grad_norm": 0.5253419876098633, "learning_rate": 7.387149912036738e-05, "loss": 0.0684, "step": 37450 }, { "grad_norm": 0.5775344967842102, "learning_rate": 7.385696927546854e-05, "loss": 0.0547, "step": 37460 }, { "grad_norm": 0.5126165747642517, "learning_rate": 7.384243682160743e-05, "loss": 0.0845, "step": 37470 }, { "grad_norm": 0.697498619556427, "learning_rate": 7.382790176037329e-05, "loss": 0.0627, "step": 37480 }, { "grad_norm": 0.5908225774765015, "learning_rate": 7.381336409335562e-05, "loss": 0.067, "step": 37490 }, { "grad_norm": 0.8441739678382874, "learning_rate": 7.379882382214426e-05, "loss": 0.0599, "step": 37500 }, { "grad_norm": 0.4884933829307556, "learning_rate": 7.378428094832931e-05, "loss": 0.0602, "step": 37510 }, { "grad_norm": 0.7475184798240662, "learning_rate": 7.376973547350114e-05, "loss": 0.0625, "step": 37520 }, { "grad_norm": 0.5414958596229553, "learning_rate": 7.375518739925046e-05, "loss": 0.0592, "step": 37530 }, { "grad_norm": 0.5225833654403687, "learning_rate": 7.374063672716818e-05, "loss": 0.0483, "step": 37540 }, { "grad_norm": 0.6767761707305908, "learning_rate": 7.372608345884558e-05, "loss": 0.0645, "step": 37550 }, { "grad_norm": 0.85066819190979, "learning_rate": 7.371152759587415e-05, "loss": 0.0651, "step": 37560 }, { "grad_norm": 0.5570716857910156, "learning_rate": 7.369696913984572e-05, "loss": 0.0479, "step": 37570 }, { "grad_norm": 0.4232906699180603, "learning_rate": 7.368240809235237e-05, "loss": 0.0561, "step": 37580 }, { "grad_norm": 0.5549857020378113, "learning_rate": 7.366784445498649e-05, "loss": 0.0684, "step": 37590 }, { "grad_norm": 0.38199982047080994, "learning_rate": 7.36532782293407e-05, "loss": 0.0499, "step": 37600 }, { "grad_norm": 0.38132375478744507, "learning_rate": 7.363870941700797e-05, "loss": 0.0529, "step": 37610 }, { "grad_norm": 0.5317173004150391, "learning_rate": 7.362413801958152e-05, "loss": 0.0476, "step": 37620 }, { "grad_norm": 0.4279378056526184, "learning_rate": 7.360956403865486e-05, "loss": 0.0549, "step": 37630 }, { "grad_norm": 0.5712628364562988, "learning_rate": 7.359498747582177e-05, "loss": 0.0616, "step": 37640 }, { "grad_norm": 0.5231826901435852, "learning_rate": 7.358040833267634e-05, "loss": 0.0518, "step": 37650 }, { "grad_norm": 0.49092766642570496, "learning_rate": 7.356582661081288e-05, "loss": 0.0587, "step": 37660 }, { "grad_norm": 0.5956699252128601, "learning_rate": 7.355124231182607e-05, "loss": 0.0418, "step": 37670 }, { "grad_norm": 0.5572565793991089, "learning_rate": 7.353665543731083e-05, "loss": 0.0508, "step": 37680 }, { "grad_norm": 0.47502848505973816, "learning_rate": 7.352206598886229e-05, "loss": 0.0515, "step": 37690 }, { "grad_norm": 0.36710235476493835, "learning_rate": 7.350747396807601e-05, "loss": 0.05, "step": 37700 }, { "grad_norm": 0.6106243133544922, "learning_rate": 7.349287937654772e-05, "loss": 0.0542, "step": 37710 }, { "grad_norm": 0.42869478464126587, "learning_rate": 7.347828221587345e-05, "loss": 0.0556, "step": 37720 }, { "grad_norm": 0.4361424744129181, "learning_rate": 7.346368248764955e-05, "loss": 0.0514, "step": 37730 }, { "grad_norm": 0.4355437755584717, "learning_rate": 7.34490801934726e-05, "loss": 0.0589, "step": 37740 }, { "grad_norm": 0.5302630662918091, "learning_rate": 7.343447533493947e-05, "loss": 0.074, "step": 37750 }, { "grad_norm": 0.4188370406627655, "learning_rate": 7.341986791364739e-05, "loss": 0.076, "step": 37760 }, { "grad_norm": 0.4088645875453949, "learning_rate": 7.340525793119374e-05, "loss": 0.0515, "step": 37770 }, { "grad_norm": 1.8760899305343628, "learning_rate": 7.33906453891763e-05, "loss": 0.0584, "step": 37780 }, { "grad_norm": 0.826535165309906, "learning_rate": 7.337603028919303e-05, "loss": 0.0536, "step": 37790 }, { "grad_norm": 0.4750617444515228, "learning_rate": 7.336141263284225e-05, "loss": 0.0565, "step": 37800 }, { "grad_norm": 1.869554877281189, "learning_rate": 7.334679242172249e-05, "loss": 0.0632, "step": 37810 }, { "grad_norm": 0.5412811636924744, "learning_rate": 7.33321696574326e-05, "loss": 0.0547, "step": 37820 }, { "grad_norm": 0.4051063656806946, "learning_rate": 7.331754434157174e-05, "loss": 0.0507, "step": 37830 }, { "grad_norm": 0.5817720293998718, "learning_rate": 7.330291647573928e-05, "loss": 0.0594, "step": 37840 }, { "grad_norm": 0.5756624341011047, "learning_rate": 7.32882860615349e-05, "loss": 0.0413, "step": 37850 }, { "grad_norm": 0.4416106343269348, "learning_rate": 7.327365310055858e-05, "loss": 0.0517, "step": 37860 }, { "grad_norm": 0.432732492685318, "learning_rate": 7.325901759441055e-05, "loss": 0.0552, "step": 37870 }, { "grad_norm": 0.46291840076446533, "learning_rate": 7.324437954469132e-05, "loss": 0.0469, "step": 37880 }, { "grad_norm": 0.7555671334266663, "learning_rate": 7.32297389530017e-05, "loss": 0.0483, "step": 37890 }, { "grad_norm": 0.5443809032440186, "learning_rate": 7.321509582094274e-05, "loss": 0.0512, "step": 37900 }, { "grad_norm": 0.3491058051586151, "learning_rate": 7.320045015011582e-05, "loss": 0.064, "step": 37910 }, { "grad_norm": 0.3234527111053467, "learning_rate": 7.318580194212253e-05, "loss": 0.0548, "step": 37920 }, { "grad_norm": 0.45064038038253784, "learning_rate": 7.317115119856483e-05, "loss": 0.0586, "step": 37930 }, { "grad_norm": 0.8581011891365051, "learning_rate": 7.315649792104485e-05, "loss": 0.0537, "step": 37940 }, { "grad_norm": 0.4358278214931488, "learning_rate": 7.314184211116507e-05, "loss": 0.0636, "step": 37950 }, { "grad_norm": 0.7151485085487366, "learning_rate": 7.312718377052823e-05, "loss": 0.0608, "step": 37960 }, { "grad_norm": 0.5541611313819885, "learning_rate": 7.311252290073737e-05, "loss": 0.0532, "step": 37970 }, { "grad_norm": 0.4782249331474304, "learning_rate": 7.309785950339572e-05, "loss": 0.0618, "step": 37980 }, { "grad_norm": 0.492486834526062, "learning_rate": 7.308319358010692e-05, "loss": 0.0528, "step": 37990 }, { "grad_norm": 0.4220012426376343, "learning_rate": 7.306852513247474e-05, "loss": 0.0758, "step": 38000 }, { "grad_norm": 0.5909157991409302, "learning_rate": 7.305385416210336e-05, "loss": 0.0613, "step": 38010 }, { "grad_norm": 0.468192994594574, "learning_rate": 7.303918067059711e-05, "loss": 0.0394, "step": 38020 }, { "grad_norm": 0.37787511944770813, "learning_rate": 7.302450465956075e-05, "loss": 0.049, "step": 38030 }, { "grad_norm": 0.4587491452693939, "learning_rate": 7.300982613059914e-05, "loss": 0.0461, "step": 38040 }, { "grad_norm": 0.7600032687187195, "learning_rate": 7.299514508531757e-05, "loss": 0.0611, "step": 38050 }, { "grad_norm": 0.37249884009361267, "learning_rate": 7.298046152532148e-05, "loss": 0.0511, "step": 38060 }, { "grad_norm": 0.7440681457519531, "learning_rate": 7.296577545221668e-05, "loss": 0.0587, "step": 38070 }, { "grad_norm": 0.43369072675704956, "learning_rate": 7.29510868676092e-05, "loss": 0.0562, "step": 38080 }, { "grad_norm": 0.5555472373962402, "learning_rate": 7.293639577310538e-05, "loss": 0.053, "step": 38090 }, { "grad_norm": 0.5226452350616455, "learning_rate": 7.292170217031179e-05, "loss": 0.0692, "step": 38100 }, { "grad_norm": 0.631184458732605, "learning_rate": 7.290700606083532e-05, "loss": 0.0613, "step": 38110 }, { "grad_norm": 0.4882226288318634, "learning_rate": 7.28923074462831e-05, "loss": 0.0598, "step": 38120 }, { "grad_norm": 0.679345428943634, "learning_rate": 7.287760632826255e-05, "loss": 0.0503, "step": 38130 }, { "grad_norm": 0.5919111967086792, "learning_rate": 7.286290270838139e-05, "loss": 0.0612, "step": 38140 }, { "grad_norm": 0.6116583943367004, "learning_rate": 7.284819658824756e-05, "loss": 0.0509, "step": 38150 }, { "grad_norm": 0.5552591681480408, "learning_rate": 7.283348796946929e-05, "loss": 0.0547, "step": 38160 }, { "grad_norm": 0.4704688489437103, "learning_rate": 7.28187768536551e-05, "loss": 0.0556, "step": 38170 }, { "grad_norm": 0.4753667414188385, "learning_rate": 7.280406324241379e-05, "loss": 0.0625, "step": 38180 }, { "grad_norm": 0.36672258377075195, "learning_rate": 7.278934713735441e-05, "loss": 0.0566, "step": 38190 }, { "grad_norm": 0.6488174200057983, "learning_rate": 7.277462854008629e-05, "loss": 0.0546, "step": 38200 }, { "grad_norm": 0.5308197736740112, "learning_rate": 7.275990745221904e-05, "loss": 0.0494, "step": 38210 }, { "grad_norm": 0.4178028106689453, "learning_rate": 7.274518387536252e-05, "loss": 0.0596, "step": 38220 }, { "grad_norm": 0.6745415329933167, "learning_rate": 7.273045781112689e-05, "loss": 0.0631, "step": 38230 }, { "grad_norm": 0.49219363927841187, "learning_rate": 7.271572926112257e-05, "loss": 0.066, "step": 38240 }, { "grad_norm": 0.7485197186470032, "learning_rate": 7.270099822696024e-05, "loss": 0.0541, "step": 38250 }, { "grad_norm": 0.32358863949775696, "learning_rate": 7.26862647102509e-05, "loss": 0.0509, "step": 38260 }, { "grad_norm": 0.3592887222766876, "learning_rate": 7.267152871260573e-05, "loss": 0.0516, "step": 38270 }, { "grad_norm": 0.7120447754859924, "learning_rate": 7.265679023563629e-05, "loss": 0.0724, "step": 38280 }, { "grad_norm": 0.9185625910758972, "learning_rate": 7.264204928095433e-05, "loss": 0.0679, "step": 38290 }, { "grad_norm": 0.49136170744895935, "learning_rate": 7.262730585017188e-05, "loss": 0.0607, "step": 38300 }, { "grad_norm": 0.45519161224365234, "learning_rate": 7.261255994490131e-05, "loss": 0.0577, "step": 38310 }, { "grad_norm": 0.5350390672683716, "learning_rate": 7.259781156675517e-05, "loss": 0.0581, "step": 38320 }, { "grad_norm": 0.7080080509185791, "learning_rate": 7.258306071734633e-05, "loss": 0.054, "step": 38330 }, { "grad_norm": 0.4466528296470642, "learning_rate": 7.256830739828792e-05, "loss": 0.0581, "step": 38340 }, { "grad_norm": 0.639457106590271, "learning_rate": 7.255355161119336e-05, "loss": 0.0437, "step": 38350 }, { "grad_norm": 0.7525030970573425, "learning_rate": 7.253879335767628e-05, "loss": 0.0486, "step": 38360 }, { "grad_norm": 0.4143257141113281, "learning_rate": 7.252403263935067e-05, "loss": 0.0674, "step": 38370 }, { "grad_norm": 0.6085182428359985, "learning_rate": 7.25092694578307e-05, "loss": 0.064, "step": 38380 }, { "grad_norm": 0.6876429319381714, "learning_rate": 7.249450381473087e-05, "loss": 0.066, "step": 38390 }, { "grad_norm": 0.4766409695148468, "learning_rate": 7.247973571166593e-05, "loss": 0.066, "step": 38400 }, { "grad_norm": 0.4861113131046295, "learning_rate": 7.246496515025089e-05, "loss": 0.0542, "step": 38410 }, { "grad_norm": 0.5884040594100952, "learning_rate": 7.245019213210101e-05, "loss": 0.0562, "step": 38420 }, { "grad_norm": 0.32840821146965027, "learning_rate": 7.24354166588319e-05, "loss": 0.0638, "step": 38430 }, { "grad_norm": 0.5489937663078308, "learning_rate": 7.242063873205935e-05, "loss": 0.0669, "step": 38440 }, { "grad_norm": 0.6330084204673767, "learning_rate": 7.240585835339946e-05, "loss": 0.057, "step": 38450 }, { "grad_norm": 0.3621477782726288, "learning_rate": 7.239107552446857e-05, "loss": 0.0548, "step": 38460 }, { "grad_norm": 0.6073769330978394, "learning_rate": 7.237629024688334e-05, "loss": 0.0598, "step": 38470 }, { "grad_norm": 0.49410566687583923, "learning_rate": 7.236150252226064e-05, "loss": 0.0569, "step": 38480 }, { "grad_norm": 0.6670507192611694, "learning_rate": 7.234671235221765e-05, "loss": 0.0608, "step": 38490 }, { "grad_norm": 0.5975078344345093, "learning_rate": 7.233191973837179e-05, "loss": 0.06, "step": 38500 }, { "grad_norm": 0.7216977477073669, "learning_rate": 7.231712468234078e-05, "loss": 0.0562, "step": 38510 }, { "grad_norm": 0.592130720615387, "learning_rate": 7.230232718574254e-05, "loss": 0.0646, "step": 38520 }, { "grad_norm": 0.29606857895851135, "learning_rate": 7.228752725019535e-05, "loss": 0.0563, "step": 38530 }, { "grad_norm": 0.37002453207969666, "learning_rate": 7.227272487731769e-05, "loss": 0.0497, "step": 38540 }, { "grad_norm": 0.4964624345302582, "learning_rate": 7.225792006872831e-05, "loss": 0.0494, "step": 38550 }, { "grad_norm": 0.9150274395942688, "learning_rate": 7.224311282604628e-05, "loss": 0.0514, "step": 38560 }, { "grad_norm": 0.5605619549751282, "learning_rate": 7.222830315089085e-05, "loss": 0.0564, "step": 38570 }, { "grad_norm": 0.6402880549430847, "learning_rate": 7.22134910448816e-05, "loss": 0.0578, "step": 38580 }, { "grad_norm": 0.5854410529136658, "learning_rate": 7.219867650963839e-05, "loss": 0.0537, "step": 38590 }, { "grad_norm": 0.40108227729797363, "learning_rate": 7.21838595467813e-05, "loss": 0.0449, "step": 38600 }, { "grad_norm": 0.400592565536499, "learning_rate": 7.216904015793067e-05, "loss": 0.0561, "step": 38610 }, { "grad_norm": 0.5125755667686462, "learning_rate": 7.215421834470713e-05, "loss": 0.0553, "step": 38620 }, { "grad_norm": 0.45357897877693176, "learning_rate": 7.213939410873158e-05, "loss": 0.0552, "step": 38630 }, { "grad_norm": 0.6187416315078735, "learning_rate": 7.212456745162519e-05, "loss": 0.0481, "step": 38640 }, { "grad_norm": 0.5395057201385498, "learning_rate": 7.210973837500937e-05, "loss": 0.0623, "step": 38650 }, { "grad_norm": 0.6796927452087402, "learning_rate": 7.209490688050578e-05, "loss": 0.0527, "step": 38660 }, { "grad_norm": 0.6090435981750488, "learning_rate": 7.208007296973641e-05, "loss": 0.0539, "step": 38670 }, { "grad_norm": 0.6089050769805908, "learning_rate": 7.206523664432345e-05, "loss": 0.0727, "step": 38680 }, { "grad_norm": 0.8755207061767578, "learning_rate": 7.205039790588939e-05, "loss": 0.0424, "step": 38690 }, { "grad_norm": 0.4294481575489044, "learning_rate": 7.203555675605697e-05, "loss": 0.0552, "step": 38700 }, { "grad_norm": 0.5730305910110474, "learning_rate": 7.202071319644917e-05, "loss": 0.065, "step": 38710 }, { "grad_norm": 0.47452083230018616, "learning_rate": 7.200586722868932e-05, "loss": 0.0522, "step": 38720 }, { "grad_norm": 0.5983721017837524, "learning_rate": 7.199101885440088e-05, "loss": 0.0669, "step": 38730 }, { "grad_norm": 0.4934983551502228, "learning_rate": 7.197616807520771e-05, "loss": 0.0616, "step": 38740 }, { "grad_norm": 0.49537357687950134, "learning_rate": 7.196131489273381e-05, "loss": 0.0469, "step": 38750 }, { "grad_norm": 0.41193926334381104, "learning_rate": 7.194645930860355e-05, "loss": 0.0489, "step": 38760 }, { "grad_norm": 0.8012397289276123, "learning_rate": 7.19316013244415e-05, "loss": 0.0495, "step": 38770 }, { "grad_norm": 0.41145679354667664, "learning_rate": 7.191674094187248e-05, "loss": 0.0631, "step": 38780 }, { "grad_norm": 0.4979502558708191, "learning_rate": 7.190187816252165e-05, "loss": 0.0437, "step": 38790 }, { "grad_norm": 0.6537152528762817, "learning_rate": 7.188701298801435e-05, "loss": 0.0642, "step": 38800 }, { "grad_norm": 0.4031280279159546, "learning_rate": 7.18721454199762e-05, "loss": 0.0545, "step": 38810 }, { "grad_norm": 1.3864107131958008, "learning_rate": 7.185727546003312e-05, "loss": 0.0574, "step": 38820 }, { "grad_norm": 0.464937299489975, "learning_rate": 7.184240310981126e-05, "loss": 0.0477, "step": 38830 }, { "grad_norm": 0.5869803428649902, "learning_rate": 7.182752837093704e-05, "loss": 0.0666, "step": 38840 }, { "grad_norm": 0.3943650722503662, "learning_rate": 7.181265124503711e-05, "loss": 0.0627, "step": 38850 }, { "grad_norm": 0.5572283864021301, "learning_rate": 7.179777173373847e-05, "loss": 0.0517, "step": 38860 }, { "grad_norm": 0.5626495480537415, "learning_rate": 7.178288983866826e-05, "loss": 0.043, "step": 38870 }, { "grad_norm": 0.4572882354259491, "learning_rate": 7.176800556145397e-05, "loss": 0.0636, "step": 38880 }, { "grad_norm": 0.5245276093482971, "learning_rate": 7.175311890372334e-05, "loss": 0.0545, "step": 38890 }, { "grad_norm": 0.7005877494812012, "learning_rate": 7.17382298671043e-05, "loss": 0.0527, "step": 38900 }, { "grad_norm": 0.5226523876190186, "learning_rate": 7.172333845322515e-05, "loss": 0.0472, "step": 38910 }, { "grad_norm": 0.39918267726898193, "learning_rate": 7.170844466371436e-05, "loss": 0.0573, "step": 38920 }, { "grad_norm": 0.5846613049507141, "learning_rate": 7.16935485002007e-05, "loss": 0.0528, "step": 38930 }, { "grad_norm": 0.6610409617424011, "learning_rate": 7.167864996431319e-05, "loss": 0.0632, "step": 38940 }, { "grad_norm": 0.44073694944381714, "learning_rate": 7.166374905768111e-05, "loss": 0.0644, "step": 38950 }, { "grad_norm": 0.5017037391662598, "learning_rate": 7.164884578193401e-05, "loss": 0.0461, "step": 38960 }, { "grad_norm": 0.5491032600402832, "learning_rate": 7.163394013870168e-05, "loss": 0.0521, "step": 38970 }, { "grad_norm": 0.6638908386230469, "learning_rate": 7.161903212961419e-05, "loss": 0.0567, "step": 38980 }, { "grad_norm": 0.4735158085823059, "learning_rate": 7.160412175630183e-05, "loss": 0.0591, "step": 38990 }, { "grad_norm": 0.29898321628570557, "learning_rate": 7.158920902039521e-05, "loss": 0.0529, "step": 39000 }, { "grad_norm": 0.6623435020446777, "learning_rate": 7.157429392352514e-05, "loss": 0.0537, "step": 39010 }, { "grad_norm": 0.6067583560943604, "learning_rate": 7.155937646732275e-05, "loss": 0.055, "step": 39020 }, { "grad_norm": 0.6057543754577637, "learning_rate": 7.154445665341933e-05, "loss": 0.0518, "step": 39030 }, { "grad_norm": 0.4550634026527405, "learning_rate": 7.152953448344654e-05, "loss": 0.0476, "step": 39040 }, { "grad_norm": 0.9587489366531372, "learning_rate": 7.151460995903624e-05, "loss": 0.0486, "step": 39050 }, { "grad_norm": 0.7014808654785156, "learning_rate": 7.149968308182052e-05, "loss": 0.055, "step": 39060 }, { "grad_norm": 0.664304256439209, "learning_rate": 7.14847538534318e-05, "loss": 0.0619, "step": 39070 }, { "grad_norm": 1.0418120622634888, "learning_rate": 7.14698222755027e-05, "loss": 0.0586, "step": 39080 }, { "grad_norm": 0.45563390851020813, "learning_rate": 7.14548883496661e-05, "loss": 0.0508, "step": 39090 }, { "grad_norm": 0.8574163317680359, "learning_rate": 7.143995207755517e-05, "loss": 0.0603, "step": 39100 }, { "grad_norm": 0.6354643106460571, "learning_rate": 7.142501346080333e-05, "loss": 0.0529, "step": 39110 }, { "grad_norm": 0.8028650879859924, "learning_rate": 7.141007250104421e-05, "loss": 0.0546, "step": 39120 }, { "grad_norm": 2.623835802078247, "learning_rate": 7.139512919991176e-05, "loss": 0.0615, "step": 39130 }, { "grad_norm": 1.123085856437683, "learning_rate": 7.138018355904014e-05, "loss": 0.0594, "step": 39140 }, { "grad_norm": 0.9366180896759033, "learning_rate": 7.13652355800638e-05, "loss": 0.0731, "step": 39150 }, { "grad_norm": 0.643602728843689, "learning_rate": 7.135028526461743e-05, "loss": 0.0587, "step": 39160 }, { "grad_norm": 0.796343207359314, "learning_rate": 7.133533261433594e-05, "loss": 0.0583, "step": 39170 }, { "grad_norm": 0.5666577219963074, "learning_rate": 7.132037763085457e-05, "loss": 0.0548, "step": 39180 }, { "grad_norm": 1.0456825494766235, "learning_rate": 7.130542031580875e-05, "loss": 0.0515, "step": 39190 }, { "grad_norm": 0.42951059341430664, "learning_rate": 7.12904606708342e-05, "loss": 0.0563, "step": 39200 }, { "grad_norm": 0.4897575378417969, "learning_rate": 7.127549869756687e-05, "loss": 0.0587, "step": 39210 }, { "grad_norm": 0.485413521528244, "learning_rate": 7.126053439764299e-05, "loss": 0.0566, "step": 39220 }, { "grad_norm": 0.3138944208621979, "learning_rate": 7.124556777269904e-05, "loss": 0.0576, "step": 39230 }, { "grad_norm": 0.7216468453407288, "learning_rate": 7.123059882437174e-05, "loss": 0.071, "step": 39240 }, { "grad_norm": 0.719224214553833, "learning_rate": 7.121562755429807e-05, "loss": 0.0544, "step": 39250 }, { "grad_norm": 0.4467698931694031, "learning_rate": 7.120065396411528e-05, "loss": 0.0548, "step": 39260 }, { "grad_norm": 0.5990010499954224, "learning_rate": 7.118567805546084e-05, "loss": 0.0514, "step": 39270 }, { "grad_norm": 0.6434754729270935, "learning_rate": 7.117069982997248e-05, "loss": 0.0572, "step": 39280 }, { "grad_norm": 0.4799754023551941, "learning_rate": 7.115571928928825e-05, "loss": 0.0797, "step": 39290 }, { "grad_norm": 0.5287771821022034, "learning_rate": 7.114073643504635e-05, "loss": 0.0554, "step": 39300 }, { "grad_norm": 0.6172722578048706, "learning_rate": 7.11257512688853e-05, "loss": 0.0567, "step": 39310 }, { "grad_norm": 0.4685584604740143, "learning_rate": 7.111076379244384e-05, "loss": 0.0514, "step": 39320 }, { "grad_norm": 0.7924470901489258, "learning_rate": 7.109577400736101e-05, "loss": 0.0627, "step": 39330 }, { "grad_norm": 0.420905202627182, "learning_rate": 7.108078191527605e-05, "loss": 0.0638, "step": 39340 }, { "grad_norm": 0.472347229719162, "learning_rate": 7.106578751782847e-05, "loss": 0.0605, "step": 39350 }, { "grad_norm": 0.3390500545501709, "learning_rate": 7.105079081665803e-05, "loss": 0.0514, "step": 39360 }, { "grad_norm": 0.4891989529132843, "learning_rate": 7.103579181340476e-05, "loss": 0.0744, "step": 39370 }, { "grad_norm": 0.6788622736930847, "learning_rate": 7.102079050970893e-05, "loss": 0.0511, "step": 39380 }, { "grad_norm": 0.3768038749694824, "learning_rate": 7.100578690721105e-05, "loss": 0.0628, "step": 39390 }, { "grad_norm": 0.5539576411247253, "learning_rate": 7.09907810075519e-05, "loss": 0.0625, "step": 39400 }, { "grad_norm": 0.4871456027030945, "learning_rate": 7.097577281237249e-05, "loss": 0.0681, "step": 39410 }, { "grad_norm": 0.5396386384963989, "learning_rate": 7.09607623233141e-05, "loss": 0.0464, "step": 39420 }, { "grad_norm": 0.6219781041145325, "learning_rate": 7.094574954201827e-05, "loss": 0.0581, "step": 39430 }, { "grad_norm": 0.5504164099693298, "learning_rate": 7.093073447012675e-05, "loss": 0.061, "step": 39440 }, { "grad_norm": 0.7102639079093933, "learning_rate": 7.09157171092816e-05, "loss": 0.0572, "step": 39450 }, { "grad_norm": 0.686970055103302, "learning_rate": 7.090069746112504e-05, "loss": 0.064, "step": 39460 }, { "grad_norm": 0.4406135380268097, "learning_rate": 7.088567552729965e-05, "loss": 0.0489, "step": 39470 }, { "grad_norm": 0.4410598874092102, "learning_rate": 7.087065130944818e-05, "loss": 0.065, "step": 39480 }, { "grad_norm": 0.689225435256958, "learning_rate": 7.085562480921366e-05, "loss": 0.0716, "step": 39490 }, { "grad_norm": 0.5130992531776428, "learning_rate": 7.084059602823937e-05, "loss": 0.0611, "step": 39500 }, { "grad_norm": 0.6817694306373596, "learning_rate": 7.082556496816882e-05, "loss": 0.0569, "step": 39510 }, { "grad_norm": 0.49820587038993835, "learning_rate": 7.081053163064582e-05, "loss": 0.0713, "step": 39520 }, { "grad_norm": 0.668481707572937, "learning_rate": 7.079549601731434e-05, "loss": 0.0593, "step": 39530 }, { "grad_norm": 0.5449867844581604, "learning_rate": 7.07804581298187e-05, "loss": 0.055, "step": 39540 }, { "grad_norm": 0.49820196628570557, "learning_rate": 7.07654179698034e-05, "loss": 0.063, "step": 39550 }, { "grad_norm": 0.5387941002845764, "learning_rate": 7.075037553891321e-05, "loss": 0.0683, "step": 39560 }, { "grad_norm": 0.4374488890171051, "learning_rate": 7.073533083879315e-05, "loss": 0.0471, "step": 39570 }, { "grad_norm": 0.5228615999221802, "learning_rate": 7.072028387108849e-05, "loss": 0.056, "step": 39580 }, { "grad_norm": 0.35847511887550354, "learning_rate": 7.070523463744472e-05, "loss": 0.0633, "step": 39590 }, { "grad_norm": 0.31613263487815857, "learning_rate": 7.069018313950763e-05, "loss": 0.0531, "step": 39600 }, { "grad_norm": 0.5225486755371094, "learning_rate": 7.067512937892322e-05, "loss": 0.0568, "step": 39610 }, { "grad_norm": 0.4600284695625305, "learning_rate": 7.066007335733775e-05, "loss": 0.0439, "step": 39620 }, { "grad_norm": 1.6961723566055298, "learning_rate": 7.064501507639772e-05, "loss": 0.0584, "step": 39630 }, { "grad_norm": 0.5279175639152527, "learning_rate": 7.062995453774987e-05, "loss": 0.0729, "step": 39640 }, { "grad_norm": 0.45787957310676575, "learning_rate": 7.061489174304121e-05, "loss": 0.0681, "step": 39650 }, { "grad_norm": 0.34135422110557556, "learning_rate": 7.0599826693919e-05, "loss": 0.0553, "step": 39660 }, { "grad_norm": 0.644995391368866, "learning_rate": 7.05847593920307e-05, "loss": 0.0565, "step": 39670 }, { "grad_norm": 0.8283445835113525, "learning_rate": 7.056968983902406e-05, "loss": 0.0554, "step": 39680 }, { "grad_norm": 0.47763222455978394, "learning_rate": 7.055461803654706e-05, "loss": 0.056, "step": 39690 }, { "grad_norm": 0.5193780064582825, "learning_rate": 7.053954398624794e-05, "loss": 0.0453, "step": 39700 }, { "grad_norm": 0.3767637610435486, "learning_rate": 7.052446768977518e-05, "loss": 0.0517, "step": 39710 }, { "grad_norm": 0.4686013162136078, "learning_rate": 7.050938914877748e-05, "loss": 0.0496, "step": 39720 }, { "grad_norm": 0.5676514506340027, "learning_rate": 7.049430836490382e-05, "loss": 0.0724, "step": 39730 }, { "grad_norm": 0.4909208118915558, "learning_rate": 7.047922533980342e-05, "loss": 0.0532, "step": 39740 }, { "grad_norm": 0.7336732149124146, "learning_rate": 7.046414007512571e-05, "loss": 0.0581, "step": 39750 }, { "grad_norm": 0.5896682143211365, "learning_rate": 7.044905257252042e-05, "loss": 0.0486, "step": 39760 }, { "grad_norm": 0.8360092043876648, "learning_rate": 7.04339628336375e-05, "loss": 0.0619, "step": 39770 }, { "grad_norm": 0.411830872297287, "learning_rate": 7.041887086012711e-05, "loss": 0.0477, "step": 39780 }, { "grad_norm": 0.5927515625953674, "learning_rate": 7.040377665363969e-05, "loss": 0.0529, "step": 39790 }, { "grad_norm": 0.42914846539497375, "learning_rate": 7.038868021582594e-05, "loss": 0.0614, "step": 39800 }, { "grad_norm": 0.453445166349411, "learning_rate": 7.037358154833679e-05, "loss": 0.0515, "step": 39810 }, { "grad_norm": 0.32351261377334595, "learning_rate": 7.035848065282339e-05, "loss": 0.0637, "step": 39820 }, { "grad_norm": 0.40190789103507996, "learning_rate": 7.034337753093714e-05, "loss": 0.0528, "step": 39830 }, { "grad_norm": 0.4374653697013855, "learning_rate": 7.032827218432972e-05, "loss": 0.0587, "step": 39840 }, { "grad_norm": 0.4315715730190277, "learning_rate": 7.031316461465302e-05, "loss": 0.0616, "step": 39850 }, { "grad_norm": 0.5670758485794067, "learning_rate": 7.029805482355915e-05, "loss": 0.055, "step": 39860 }, { "grad_norm": 1.181174635887146, "learning_rate": 7.028294281270055e-05, "loss": 0.0434, "step": 39870 }, { "grad_norm": 0.36878830194473267, "learning_rate": 7.026782858372982e-05, "loss": 0.0499, "step": 39880 }, { "grad_norm": 0.43575751781463623, "learning_rate": 7.025271213829982e-05, "loss": 0.0569, "step": 39890 }, { "grad_norm": 0.45181503891944885, "learning_rate": 7.023759347806366e-05, "loss": 0.0473, "step": 39900 }, { "grad_norm": 0.5341687202453613, "learning_rate": 7.022247260467472e-05, "loss": 0.0587, "step": 39910 }, { "grad_norm": 0.4645366370677948, "learning_rate": 7.020734951978658e-05, "loss": 0.0518, "step": 39920 }, { "grad_norm": 0.612263023853302, "learning_rate": 7.019222422505307e-05, "loss": 0.0571, "step": 39930 }, { "grad_norm": 0.4200182259082794, "learning_rate": 7.017709672212827e-05, "loss": 0.0586, "step": 39940 }, { "grad_norm": 0.7293274998664856, "learning_rate": 7.016196701266652e-05, "loss": 0.0595, "step": 39950 }, { "grad_norm": 1.021011233329773, "learning_rate": 7.014683509832238e-05, "loss": 0.0527, "step": 39960 }, { "grad_norm": 0.5376349091529846, "learning_rate": 7.013170098075063e-05, "loss": 0.0614, "step": 39970 }, { "grad_norm": 0.49690836668014526, "learning_rate": 7.011656466160632e-05, "loss": 0.0683, "step": 39980 }, { "grad_norm": 0.5378111600875854, "learning_rate": 7.010142614254475e-05, "loss": 0.0529, "step": 39990 }, { "grad_norm": 0.3720652163028717, "learning_rate": 7.008628542522147e-05, "loss": 0.0583, "step": 40000 }, { "grad_norm": 0.4630858898162842, "learning_rate": 7.007114251129218e-05, "loss": 0.05, "step": 40010 }, { "grad_norm": 0.6802982687950134, "learning_rate": 7.005599740241293e-05, "loss": 0.0418, "step": 40020 }, { "grad_norm": 0.5097094774246216, "learning_rate": 7.004085010023996e-05, "loss": 0.0592, "step": 40030 }, { "grad_norm": 0.8357699513435364, "learning_rate": 7.002570060642976e-05, "loss": 0.0479, "step": 40040 }, { "grad_norm": 0.5614554286003113, "learning_rate": 7.001054892263903e-05, "loss": 0.0423, "step": 40050 }, { "grad_norm": 0.4779838025569916, "learning_rate": 6.999539505052477e-05, "loss": 0.0645, "step": 40060 }, { "grad_norm": 0.5743301510810852, "learning_rate": 6.998023899174418e-05, "loss": 0.0464, "step": 40070 }, { "grad_norm": 0.5798883438110352, "learning_rate": 6.996508074795467e-05, "loss": 0.0519, "step": 40080 }, { "grad_norm": 0.5317139625549316, "learning_rate": 6.994992032081396e-05, "loss": 0.0642, "step": 40090 }, { "grad_norm": 0.496198832988739, "learning_rate": 6.993475771197995e-05, "loss": 0.0518, "step": 40100 }, { "grad_norm": 0.5728045701980591, "learning_rate": 6.991959292311082e-05, "loss": 0.0475, "step": 40110 }, { "grad_norm": 0.5083343386650085, "learning_rate": 6.990442595586495e-05, "loss": 0.0569, "step": 40120 }, { "grad_norm": 0.7724452614784241, "learning_rate": 6.988925681190098e-05, "loss": 0.0558, "step": 40130 }, { "grad_norm": 1.1838500499725342, "learning_rate": 6.987408549287778e-05, "loss": 0.0659, "step": 40140 }, { "grad_norm": 0.6613211631774902, "learning_rate": 6.985891200045449e-05, "loss": 0.0767, "step": 40150 }, { "grad_norm": 0.6041389107704163, "learning_rate": 6.984373633629045e-05, "loss": 0.069, "step": 40160 }, { "grad_norm": 1.0184037685394287, "learning_rate": 6.98285585020452e-05, "loss": 0.0541, "step": 40170 }, { "grad_norm": 0.36701932549476624, "learning_rate": 6.981337849937864e-05, "loss": 0.0513, "step": 40180 }, { "grad_norm": 0.6545982360839844, "learning_rate": 6.979819632995078e-05, "loss": 0.057, "step": 40190 }, { "grad_norm": 0.2874508500099182, "learning_rate": 6.978301199542193e-05, "loss": 0.0386, "step": 40200 }, { "grad_norm": 1.6431111097335815, "learning_rate": 6.976782549745263e-05, "loss": 0.0496, "step": 40210 }, { "grad_norm": 0.44033461809158325, "learning_rate": 6.975263683770367e-05, "loss": 0.0565, "step": 40220 }, { "grad_norm": 0.38390034437179565, "learning_rate": 6.973744601783604e-05, "loss": 0.0579, "step": 40230 }, { "grad_norm": 0.6840354800224304, "learning_rate": 6.972225303951097e-05, "loss": 0.0522, "step": 40240 }, { "grad_norm": 0.3950800597667694, "learning_rate": 6.970705790438998e-05, "loss": 0.0473, "step": 40250 }, { "grad_norm": 0.5752217769622803, "learning_rate": 6.969186061413477e-05, "loss": 0.0529, "step": 40260 }, { "grad_norm": 0.6237720847129822, "learning_rate": 6.967666117040727e-05, "loss": 0.0654, "step": 40270 }, { "grad_norm": 0.48222616314888, "learning_rate": 6.966145957486972e-05, "loss": 0.055, "step": 40280 }, { "grad_norm": 0.43923017382621765, "learning_rate": 6.964625582918449e-05, "loss": 0.0589, "step": 40290 }, { "grad_norm": 0.32112452387809753, "learning_rate": 6.963104993501425e-05, "loss": 0.0531, "step": 40300 }, { "grad_norm": 0.41697320342063904, "learning_rate": 6.961584189402192e-05, "loss": 0.0441, "step": 40310 }, { "grad_norm": 0.6224372386932373, "learning_rate": 6.96006317078706e-05, "loss": 0.0411, "step": 40320 }, { "grad_norm": 0.56170654296875, "learning_rate": 6.958541937822367e-05, "loss": 0.0491, "step": 40330 }, { "grad_norm": 0.5201998949050903, "learning_rate": 6.957020490674473e-05, "loss": 0.0529, "step": 40340 }, { "grad_norm": 0.4675564467906952, "learning_rate": 6.95549882950976e-05, "loss": 0.0542, "step": 40350 }, { "grad_norm": 0.8393761515617371, "learning_rate": 6.953976954494635e-05, "loss": 0.0417, "step": 40360 }, { "grad_norm": 0.42540982365608215, "learning_rate": 6.952454865795528e-05, "loss": 0.0485, "step": 40370 }, { "grad_norm": 0.782201886177063, "learning_rate": 6.950932563578892e-05, "loss": 0.0585, "step": 40380 }, { "grad_norm": 0.42597508430480957, "learning_rate": 6.949410048011206e-05, "loss": 0.0518, "step": 40390 }, { "grad_norm": 0.4168405830860138, "learning_rate": 6.947887319258966e-05, "loss": 0.0628, "step": 40400 }, { "grad_norm": 0.3544706106185913, "learning_rate": 6.946364377488696e-05, "loss": 0.0547, "step": 40410 }, { "grad_norm": 0.5331676006317139, "learning_rate": 6.944841222866947e-05, "loss": 0.068, "step": 40420 }, { "grad_norm": 0.401706725358963, "learning_rate": 6.943317855560284e-05, "loss": 0.0588, "step": 40430 }, { "grad_norm": 0.7882758378982544, "learning_rate": 6.941794275735302e-05, "loss": 0.0556, "step": 40440 }, { "grad_norm": 0.40070852637290955, "learning_rate": 6.94027048355862e-05, "loss": 0.0455, "step": 40450 }, { "grad_norm": 0.5636683702468872, "learning_rate": 6.938746479196871e-05, "loss": 0.0562, "step": 40460 }, { "grad_norm": 0.6570601463317871, "learning_rate": 6.937222262816724e-05, "loss": 0.051, "step": 40470 }, { "grad_norm": 0.3248845040798187, "learning_rate": 6.935697834584865e-05, "loss": 0.0505, "step": 40480 }, { "grad_norm": 0.5685689449310303, "learning_rate": 6.934173194667998e-05, "loss": 0.0481, "step": 40490 }, { "grad_norm": 0.6096951961517334, "learning_rate": 6.93264834323286e-05, "loss": 0.0555, "step": 40500 }, { "grad_norm": 1.1505533456802368, "learning_rate": 6.931123280446206e-05, "loss": 0.0575, "step": 40510 }, { "grad_norm": 0.47798362374305725, "learning_rate": 6.929598006474811e-05, "loss": 0.0543, "step": 40520 }, { "grad_norm": 0.3564060926437378, "learning_rate": 6.92807252148548e-05, "loss": 0.0678, "step": 40530 }, { "grad_norm": 0.5275061130523682, "learning_rate": 6.926546825645037e-05, "loss": 0.0454, "step": 40540 }, { "grad_norm": 0.3549315333366394, "learning_rate": 6.92502091912033e-05, "loss": 0.0473, "step": 40550 }, { "grad_norm": 0.5171101689338684, "learning_rate": 6.923494802078229e-05, "loss": 0.0613, "step": 40560 }, { "grad_norm": 0.6243270039558411, "learning_rate": 6.921968474685628e-05, "loss": 0.0489, "step": 40570 }, { "grad_norm": 0.6107014417648315, "learning_rate": 6.920441937109445e-05, "loss": 0.0627, "step": 40580 }, { "grad_norm": 0.480296790599823, "learning_rate": 6.91891518951662e-05, "loss": 0.049, "step": 40590 }, { "grad_norm": 0.4589221775531769, "learning_rate": 6.917388232074114e-05, "loss": 0.0445, "step": 40600 }, { "grad_norm": 0.7615552544593811, "learning_rate": 6.915861064948914e-05, "loss": 0.0568, "step": 40610 }, { "grad_norm": 0.345414400100708, "learning_rate": 6.914333688308025e-05, "loss": 0.0422, "step": 40620 }, { "grad_norm": 0.5162860155105591, "learning_rate": 6.912806102318487e-05, "loss": 0.0454, "step": 40630 }, { "grad_norm": 0.3984543979167938, "learning_rate": 6.911278307147347e-05, "loss": 0.0511, "step": 40640 }, { "grad_norm": 0.37315618991851807, "learning_rate": 6.909750302961684e-05, "loss": 0.0542, "step": 40650 }, { "grad_norm": 0.8300755023956299, "learning_rate": 6.9082220899286e-05, "loss": 0.0618, "step": 40660 }, { "grad_norm": 0.8072805404663086, "learning_rate": 6.906693668215219e-05, "loss": 0.0693, "step": 40670 }, { "grad_norm": 1.08358633518219, "learning_rate": 6.905165037988683e-05, "loss": 0.0487, "step": 40680 }, { "grad_norm": 0.5634852647781372, "learning_rate": 6.903636199416164e-05, "loss": 0.0554, "step": 40690 }, { "grad_norm": 0.47805920243263245, "learning_rate": 6.902107152664851e-05, "loss": 0.0548, "step": 40700 }, { "grad_norm": 0.5678945183753967, "learning_rate": 6.90057789790196e-05, "loss": 0.0565, "step": 40710 }, { "grad_norm": 0.48227453231811523, "learning_rate": 6.899048435294728e-05, "loss": 0.0658, "step": 40720 }, { "grad_norm": 0.6038839221000671, "learning_rate": 6.897518765010415e-05, "loss": 0.0589, "step": 40730 }, { "grad_norm": 0.8185408115386963, "learning_rate": 6.895988887216302e-05, "loss": 0.0569, "step": 40740 }, { "grad_norm": 0.48880305886268616, "learning_rate": 6.894458802079694e-05, "loss": 0.0556, "step": 40750 }, { "grad_norm": 0.707940936088562, "learning_rate": 6.892928509767922e-05, "loss": 0.0485, "step": 40760 }, { "grad_norm": 0.6246594786643982, "learning_rate": 6.891398010448333e-05, "loss": 0.0497, "step": 40770 }, { "grad_norm": 0.6734651327133179, "learning_rate": 6.8898673042883e-05, "loss": 0.0657, "step": 40780 }, { "grad_norm": 0.498195081949234, "learning_rate": 6.888336391455222e-05, "loss": 0.0538, "step": 40790 }, { "grad_norm": 0.4346676468849182, "learning_rate": 6.886805272116513e-05, "loss": 0.053, "step": 40800 }, { "grad_norm": 0.5412008762359619, "learning_rate": 6.885273946439617e-05, "loss": 0.0697, "step": 40810 }, { "grad_norm": 0.8476274013519287, "learning_rate": 6.883742414591998e-05, "loss": 0.0705, "step": 40820 }, { "grad_norm": 0.5712675452232361, "learning_rate": 6.88221067674114e-05, "loss": 0.0563, "step": 40830 }, { "grad_norm": 0.5908927321434021, "learning_rate": 6.88067873305455e-05, "loss": 0.0574, "step": 40840 }, { "grad_norm": 0.5973493456840515, "learning_rate": 6.879146583699765e-05, "loss": 0.0527, "step": 40850 }, { "grad_norm": 0.4665515124797821, "learning_rate": 6.877614228844334e-05, "loss": 0.0527, "step": 40860 }, { "grad_norm": 0.45081451535224915, "learning_rate": 6.876081668655832e-05, "loss": 0.0511, "step": 40870 }, { "grad_norm": 0.5582584738731384, "learning_rate": 6.874548903301861e-05, "loss": 0.0608, "step": 40880 }, { "grad_norm": 0.4365890622138977, "learning_rate": 6.873015932950039e-05, "loss": 0.0493, "step": 40890 }, { "grad_norm": 0.6140711903572083, "learning_rate": 6.871482757768012e-05, "loss": 0.057, "step": 40900 }, { "grad_norm": 0.4487750828266144, "learning_rate": 6.869949377923442e-05, "loss": 0.0596, "step": 40910 }, { "grad_norm": 0.5050556063652039, "learning_rate": 6.868415793584022e-05, "loss": 0.0642, "step": 40920 }, { "grad_norm": 0.42553648352622986, "learning_rate": 6.86688200491746e-05, "loss": 0.0489, "step": 40930 }, { "grad_norm": 0.3727019429206848, "learning_rate": 6.865348012091488e-05, "loss": 0.0475, "step": 40940 }, { "grad_norm": 0.49971717596054077, "learning_rate": 6.86381381527386e-05, "loss": 0.0518, "step": 40950 }, { "grad_norm": 0.4912351667881012, "learning_rate": 6.862279414632359e-05, "loss": 0.0484, "step": 40960 }, { "grad_norm": 0.5229288339614868, "learning_rate": 6.860744810334778e-05, "loss": 0.0495, "step": 40970 }, { "grad_norm": 0.4567105174064636, "learning_rate": 6.859210002548943e-05, "loss": 0.0596, "step": 40980 }, { "grad_norm": 0.40684667229652405, "learning_rate": 6.8576749914427e-05, "loss": 0.0509, "step": 40990 }, { "grad_norm": 0.6641053557395935, "learning_rate": 6.85613977718391e-05, "loss": 0.0489, "step": 41000 }, { "grad_norm": 0.47906550765037537, "learning_rate": 6.854604359940464e-05, "loss": 0.0462, "step": 41010 }, { "grad_norm": 0.6138921976089478, "learning_rate": 6.853068739880276e-05, "loss": 0.0574, "step": 41020 }, { "grad_norm": 0.5403135418891907, "learning_rate": 6.851532917171276e-05, "loss": 0.0521, "step": 41030 }, { "grad_norm": 0.5655713081359863, "learning_rate": 6.849996891981418e-05, "loss": 0.0675, "step": 41040 }, { "grad_norm": 0.6918125152587891, "learning_rate": 6.848460664478684e-05, "loss": 0.0571, "step": 41050 }, { "grad_norm": 0.7080461382865906, "learning_rate": 6.846924234831069e-05, "loss": 0.0485, "step": 41060 }, { "grad_norm": 0.8094398975372314, "learning_rate": 6.845387603206597e-05, "loss": 0.0642, "step": 41070 }, { "grad_norm": 0.30576059222221375, "learning_rate": 6.843850769773312e-05, "loss": 0.0471, "step": 41080 }, { "grad_norm": 0.5294889211654663, "learning_rate": 6.842313734699278e-05, "loss": 0.0609, "step": 41090 }, { "grad_norm": 0.4240225851535797, "learning_rate": 6.840776498152584e-05, "loss": 0.0491, "step": 41100 }, { "grad_norm": 0.7294907569885254, "learning_rate": 6.83923906030134e-05, "loss": 0.0456, "step": 41110 }, { "grad_norm": 0.8161594271659851, "learning_rate": 6.837701421313677e-05, "loss": 0.0698, "step": 41120 }, { "grad_norm": 0.6362582445144653, "learning_rate": 6.83616358135775e-05, "loss": 0.0562, "step": 41130 }, { "grad_norm": 0.8278599977493286, "learning_rate": 6.834625540601734e-05, "loss": 0.0711, "step": 41140 }, { "grad_norm": 0.6010816693305969, "learning_rate": 6.833087299213829e-05, "loss": 0.0463, "step": 41150 }, { "grad_norm": 0.5795104503631592, "learning_rate": 6.83154885736225e-05, "loss": 0.0625, "step": 41160 }, { "grad_norm": 0.4804745018482208, "learning_rate": 6.830010215215244e-05, "loss": 0.0573, "step": 41170 }, { "grad_norm": 0.812312126159668, "learning_rate": 6.828471372941072e-05, "loss": 0.0622, "step": 41180 }, { "grad_norm": 0.5198301672935486, "learning_rate": 6.826932330708018e-05, "loss": 0.0475, "step": 41190 }, { "grad_norm": 0.4539756774902344, "learning_rate": 6.825393088684393e-05, "loss": 0.0506, "step": 41200 }, { "grad_norm": 0.45696207880973816, "learning_rate": 6.823853647038524e-05, "loss": 0.0519, "step": 41210 }, { "grad_norm": 0.28815028071403503, "learning_rate": 6.822314005938762e-05, "loss": 0.0449, "step": 41220 }, { "grad_norm": 1.185817837715149, "learning_rate": 6.820774165553479e-05, "loss": 0.064, "step": 41230 }, { "grad_norm": 0.7476150393486023, "learning_rate": 6.819234126051072e-05, "loss": 0.045, "step": 41240 }, { "grad_norm": 0.4927156865596771, "learning_rate": 6.817693887599956e-05, "loss": 0.0659, "step": 41250 }, { "grad_norm": 0.5636172890663147, "learning_rate": 6.816153450368568e-05, "loss": 0.0609, "step": 41260 }, { "grad_norm": 0.46619847416877747, "learning_rate": 6.814612814525372e-05, "loss": 0.0608, "step": 41270 }, { "grad_norm": 0.5010236501693726, "learning_rate": 6.813071980238846e-05, "loss": 0.0546, "step": 41280 }, { "grad_norm": 0.48424723744392395, "learning_rate": 6.811530947677492e-05, "loss": 0.0484, "step": 41290 }, { "grad_norm": 0.5085936188697815, "learning_rate": 6.809989717009839e-05, "loss": 0.0516, "step": 41300 }, { "grad_norm": 0.5770591497421265, "learning_rate": 6.808448288404431e-05, "loss": 0.0639, "step": 41310 }, { "grad_norm": 0.4731644093990326, "learning_rate": 6.806906662029838e-05, "loss": 0.0574, "step": 41320 }, { "grad_norm": 1.398574709892273, "learning_rate": 6.805364838054647e-05, "loss": 0.0591, "step": 41330 }, { "grad_norm": 0.5580778121948242, "learning_rate": 6.803822816647474e-05, "loss": 0.043, "step": 41340 }, { "grad_norm": 0.441405326128006, "learning_rate": 6.802280597976949e-05, "loss": 0.0571, "step": 41350 }, { "grad_norm": 0.5343619585037231, "learning_rate": 6.800738182211727e-05, "loss": 0.0525, "step": 41360 }, { "grad_norm": 0.8990924954414368, "learning_rate": 6.799195569520485e-05, "loss": 0.0595, "step": 41370 }, { "grad_norm": 0.5671149492263794, "learning_rate": 6.79765276007192e-05, "loss": 0.0552, "step": 41380 }, { "grad_norm": 0.5704885721206665, "learning_rate": 6.796109754034753e-05, "loss": 0.0504, "step": 41390 }, { "grad_norm": 0.33515098690986633, "learning_rate": 6.794566551577724e-05, "loss": 0.0472, "step": 41400 }, { "grad_norm": 1.3129478693008423, "learning_rate": 6.793023152869594e-05, "loss": 0.0463, "step": 41410 }, { "grad_norm": 0.560448408126831, "learning_rate": 6.791479558079147e-05, "loss": 0.0483, "step": 41420 }, { "grad_norm": 0.4142099916934967, "learning_rate": 6.789935767375191e-05, "loss": 0.0588, "step": 41430 }, { "grad_norm": 0.35204070806503296, "learning_rate": 6.78839178092655e-05, "loss": 0.0382, "step": 41440 }, { "grad_norm": 0.39315348863601685, "learning_rate": 6.786847598902072e-05, "loss": 0.0545, "step": 41450 }, { "grad_norm": 0.5582247972488403, "learning_rate": 6.785303221470627e-05, "loss": 0.0556, "step": 41460 }, { "grad_norm": 0.4115844666957855, "learning_rate": 6.783758648801108e-05, "loss": 0.0509, "step": 41470 }, { "grad_norm": 0.6455174684524536, "learning_rate": 6.782213881062422e-05, "loss": 0.0552, "step": 41480 }, { "grad_norm": 0.6333045363426208, "learning_rate": 6.780668918423508e-05, "loss": 0.051, "step": 41490 }, { "grad_norm": 0.7574032545089722, "learning_rate": 6.779123761053317e-05, "loss": 0.047, "step": 41500 }, { "grad_norm": 0.36677759885787964, "learning_rate": 6.777578409120826e-05, "loss": 0.0574, "step": 41510 }, { "grad_norm": 0.6100561022758484, "learning_rate": 6.776032862795033e-05, "loss": 0.0564, "step": 41520 }, { "grad_norm": 0.6613268256187439, "learning_rate": 6.774487122244956e-05, "loss": 0.052, "step": 41530 }, { "grad_norm": 0.4883367419242859, "learning_rate": 6.772941187639637e-05, "loss": 0.0378, "step": 41540 }, { "grad_norm": 0.41327401995658875, "learning_rate": 6.771395059148134e-05, "loss": 0.0558, "step": 41550 }, { "grad_norm": 0.594346284866333, "learning_rate": 6.76984873693953e-05, "loss": 0.0501, "step": 41560 }, { "grad_norm": 0.5670834183692932, "learning_rate": 6.768302221182931e-05, "loss": 0.0596, "step": 41570 }, { "grad_norm": 0.7032959461212158, "learning_rate": 6.766755512047457e-05, "loss": 0.0689, "step": 41580 }, { "grad_norm": 0.594152569770813, "learning_rate": 6.765208609702259e-05, "loss": 0.0594, "step": 41590 }, { "grad_norm": 0.5560455918312073, "learning_rate": 6.763661514316499e-05, "loss": 0.0535, "step": 41600 }, { "grad_norm": 0.3349933624267578, "learning_rate": 6.76211422605937e-05, "loss": 0.052, "step": 41610 }, { "grad_norm": 0.3995225429534912, "learning_rate": 6.760566745100076e-05, "loss": 0.0516, "step": 41620 }, { "grad_norm": 0.6480172872543335, "learning_rate": 6.75901907160785e-05, "loss": 0.0583, "step": 41630 }, { "grad_norm": 0.3609640896320343, "learning_rate": 6.757471205751943e-05, "loss": 0.0455, "step": 41640 }, { "grad_norm": 0.5842539072036743, "learning_rate": 6.75592314770163e-05, "loss": 0.0564, "step": 41650 }, { "grad_norm": 0.6153454184532166, "learning_rate": 6.7543748976262e-05, "loss": 0.0599, "step": 41660 }, { "grad_norm": 0.34173306822776794, "learning_rate": 6.752826455694968e-05, "loss": 0.0493, "step": 41670 }, { "grad_norm": 0.4412248432636261, "learning_rate": 6.751277822077271e-05, "loss": 0.0489, "step": 41680 }, { "grad_norm": 0.3588988780975342, "learning_rate": 6.749728996942465e-05, "loss": 0.0392, "step": 41690 }, { "grad_norm": 0.3688163161277771, "learning_rate": 6.748179980459924e-05, "loss": 0.0446, "step": 41700 }, { "grad_norm": 0.3749900162220001, "learning_rate": 6.746630772799052e-05, "loss": 0.0424, "step": 41710 }, { "grad_norm": 0.6368691921234131, "learning_rate": 6.745081374129262e-05, "loss": 0.0516, "step": 41720 }, { "grad_norm": 0.6510012149810791, "learning_rate": 6.743531784619998e-05, "loss": 0.0633, "step": 41730 }, { "grad_norm": 0.7120286822319031, "learning_rate": 6.741982004440719e-05, "loss": 0.0552, "step": 41740 }, { "grad_norm": 0.6188085675239563, "learning_rate": 6.740432033760907e-05, "loss": 0.0574, "step": 41750 }, { "grad_norm": 0.5990220904350281, "learning_rate": 6.738881872750066e-05, "loss": 0.0514, "step": 41760 }, { "grad_norm": 0.3921268582344055, "learning_rate": 6.737331521577715e-05, "loss": 0.059, "step": 41770 }, { "grad_norm": 0.5630009770393372, "learning_rate": 6.735780980413403e-05, "loss": 0.0564, "step": 41780 }, { "grad_norm": 0.5638762712478638, "learning_rate": 6.734230249426692e-05, "loss": 0.056, "step": 41790 }, { "grad_norm": 0.8712558150291443, "learning_rate": 6.732679328787168e-05, "loss": 0.061, "step": 41800 }, { "grad_norm": 0.29271745681762695, "learning_rate": 6.731128218664438e-05, "loss": 0.041, "step": 41810 }, { "grad_norm": 0.9646647572517395, "learning_rate": 6.729576919228129e-05, "loss": 0.0432, "step": 41820 }, { "grad_norm": 0.4042138457298279, "learning_rate": 6.728025430647888e-05, "loss": 0.0454, "step": 41830 }, { "grad_norm": 0.5424306988716125, "learning_rate": 6.726473753093383e-05, "loss": 0.0581, "step": 41840 }, { "grad_norm": 0.4670286178588867, "learning_rate": 6.724921886734305e-05, "loss": 0.0506, "step": 41850 }, { "grad_norm": 0.6472423672676086, "learning_rate": 6.723369831740363e-05, "loss": 0.0609, "step": 41860 }, { "grad_norm": 0.3117123544216156, "learning_rate": 6.721817588281288e-05, "loss": 0.0602, "step": 41870 }, { "grad_norm": 0.5091086030006409, "learning_rate": 6.720265156526828e-05, "loss": 0.0495, "step": 41880 }, { "grad_norm": 0.3506109416484833, "learning_rate": 6.718712536646758e-05, "loss": 0.0476, "step": 41890 }, { "grad_norm": 0.7825108170509338, "learning_rate": 6.71715972881087e-05, "loss": 0.0531, "step": 41900 }, { "grad_norm": 0.4100225865840912, "learning_rate": 6.715606733188974e-05, "loss": 0.0497, "step": 41910 }, { "grad_norm": 0.4748019576072693, "learning_rate": 6.714053549950909e-05, "loss": 0.0441, "step": 41920 }, { "grad_norm": 0.6797329187393188, "learning_rate": 6.71250017926652e-05, "loss": 0.0563, "step": 41930 }, { "grad_norm": 0.5101884603500366, "learning_rate": 6.71094662130569e-05, "loss": 0.0484, "step": 41940 }, { "grad_norm": 0.888299286365509, "learning_rate": 6.709392876238307e-05, "loss": 0.0591, "step": 41950 }, { "grad_norm": 0.8148622512817383, "learning_rate": 6.707838944234289e-05, "loss": 0.0613, "step": 41960 }, { "grad_norm": 0.5565935969352722, "learning_rate": 6.706284825463574e-05, "loss": 0.0616, "step": 41970 }, { "grad_norm": 0.5097442865371704, "learning_rate": 6.704730520096114e-05, "loss": 0.0544, "step": 41980 }, { "grad_norm": 0.6422759890556335, "learning_rate": 6.703176028301888e-05, "loss": 0.0553, "step": 41990 }, { "grad_norm": 0.43551892042160034, "learning_rate": 6.701621350250892e-05, "loss": 0.0533, "step": 42000 }, { "grad_norm": 0.7397832870483398, "learning_rate": 6.700066486113144e-05, "loss": 0.0564, "step": 42010 }, { "grad_norm": 1.0315779447555542, "learning_rate": 6.69851143605868e-05, "loss": 0.0622, "step": 42020 }, { "grad_norm": 0.494615763425827, "learning_rate": 6.696956200257561e-05, "loss": 0.049, "step": 42030 }, { "grad_norm": 0.3764829933643341, "learning_rate": 6.695400778879863e-05, "loss": 0.055, "step": 42040 }, { "grad_norm": 0.8226677179336548, "learning_rate": 6.693845172095683e-05, "loss": 0.0705, "step": 42050 }, { "grad_norm": 0.40309491753578186, "learning_rate": 6.692289380075142e-05, "loss": 0.0526, "step": 42060 }, { "grad_norm": 0.4795338809490204, "learning_rate": 6.690733402988379e-05, "loss": 0.0531, "step": 42070 }, { "grad_norm": 0.5164294242858887, "learning_rate": 6.689177241005553e-05, "loss": 0.0531, "step": 42080 }, { "grad_norm": 0.46976107358932495, "learning_rate": 6.687620894296844e-05, "loss": 0.0503, "step": 42090 }, { "grad_norm": 0.5613430142402649, "learning_rate": 6.686064363032451e-05, "loss": 0.0613, "step": 42100 }, { "grad_norm": 0.6060100197792053, "learning_rate": 6.684507647382596e-05, "loss": 0.0464, "step": 42110 }, { "grad_norm": 0.6364094018936157, "learning_rate": 6.682950747517513e-05, "loss": 0.054, "step": 42120 }, { "grad_norm": 0.5847389101982117, "learning_rate": 6.68139366360747e-05, "loss": 0.056, "step": 42130 }, { "grad_norm": 0.46994340419769287, "learning_rate": 6.679836395822744e-05, "loss": 0.0432, "step": 42140 }, { "grad_norm": 0.5965257287025452, "learning_rate": 6.678278944333633e-05, "loss": 0.0488, "step": 42150 }, { "grad_norm": 0.6096035242080688, "learning_rate": 6.676721309310462e-05, "loss": 0.0455, "step": 42160 }, { "grad_norm": 0.6457878947257996, "learning_rate": 6.67516349092357e-05, "loss": 0.0548, "step": 42170 }, { "grad_norm": 0.407013863325119, "learning_rate": 6.673605489343315e-05, "loss": 0.0413, "step": 42180 }, { "grad_norm": 0.6195597052574158, "learning_rate": 6.672047304740081e-05, "loss": 0.0513, "step": 42190 }, { "grad_norm": 0.5876469016075134, "learning_rate": 6.670488937284268e-05, "loss": 0.0437, "step": 42200 }, { "grad_norm": 0.5582423806190491, "learning_rate": 6.668930387146294e-05, "loss": 0.0493, "step": 42210 }, { "grad_norm": 0.4802088439464569, "learning_rate": 6.667371654496605e-05, "loss": 0.0481, "step": 42220 }, { "grad_norm": 0.5169362425804138, "learning_rate": 6.665812739505659e-05, "loss": 0.0535, "step": 42230 }, { "grad_norm": 0.5598340630531311, "learning_rate": 6.664253642343935e-05, "loss": 0.0554, "step": 42240 }, { "grad_norm": 0.6793485879898071, "learning_rate": 6.662694363181935e-05, "loss": 0.0509, "step": 42250 }, { "grad_norm": 0.5856769680976868, "learning_rate": 6.66113490219018e-05, "loss": 0.0642, "step": 42260 }, { "grad_norm": 0.4049990177154541, "learning_rate": 6.659575259539207e-05, "loss": 0.0466, "step": 42270 }, { "grad_norm": 0.6073353886604309, "learning_rate": 6.65801543539958e-05, "loss": 0.0576, "step": 42280 }, { "grad_norm": 0.5810331106185913, "learning_rate": 6.65645542994188e-05, "loss": 0.0503, "step": 42290 }, { "grad_norm": 0.4744211435317993, "learning_rate": 6.654895243336702e-05, "loss": 0.048, "step": 42300 }, { "grad_norm": 0.5199338793754578, "learning_rate": 6.65333487575467e-05, "loss": 0.0379, "step": 42310 }, { "grad_norm": 1.2635278701782227, "learning_rate": 6.651774327366422e-05, "loss": 0.0429, "step": 42320 }, { "grad_norm": 0.42758291959762573, "learning_rate": 6.650213598342613e-05, "loss": 0.0498, "step": 42330 }, { "grad_norm": 0.38401222229003906, "learning_rate": 6.648652688853931e-05, "loss": 0.0575, "step": 42340 }, { "grad_norm": 2.523935556411743, "learning_rate": 6.647091599071066e-05, "loss": 0.0572, "step": 42350 }, { "grad_norm": 0.4526580572128296, "learning_rate": 6.645530329164743e-05, "loss": 0.0486, "step": 42360 }, { "grad_norm": 0.45477885007858276, "learning_rate": 6.643968879305694e-05, "loss": 0.0495, "step": 42370 }, { "grad_norm": 0.4803876280784607, "learning_rate": 6.642407249664683e-05, "loss": 0.0505, "step": 42380 }, { "grad_norm": 0.507609486579895, "learning_rate": 6.640845440412483e-05, "loss": 0.0703, "step": 42390 }, { "grad_norm": 0.5945781469345093, "learning_rate": 6.639283451719893e-05, "loss": 0.0513, "step": 42400 }, { "grad_norm": 0.47398051619529724, "learning_rate": 6.637721283757729e-05, "loss": 0.0399, "step": 42410 }, { "grad_norm": 0.38163870573043823, "learning_rate": 6.636158936696828e-05, "loss": 0.048, "step": 42420 }, { "grad_norm": 0.8705970048904419, "learning_rate": 6.634596410708047e-05, "loss": 0.0596, "step": 42430 }, { "grad_norm": 0.5194540023803711, "learning_rate": 6.633033705962257e-05, "loss": 0.0512, "step": 42440 }, { "grad_norm": 0.547860324382782, "learning_rate": 6.631470822630359e-05, "loss": 0.0581, "step": 42450 }, { "grad_norm": 0.9156199097633362, "learning_rate": 6.629907760883263e-05, "loss": 0.0518, "step": 42460 }, { "grad_norm": 0.48315978050231934, "learning_rate": 6.628344520891905e-05, "loss": 0.0427, "step": 42470 }, { "grad_norm": 0.4009622037410736, "learning_rate": 6.62678110282724e-05, "loss": 0.0395, "step": 42480 }, { "grad_norm": 0.7353798747062683, "learning_rate": 6.625217506860237e-05, "loss": 0.0563, "step": 42490 }, { "grad_norm": 0.355206161737442, "learning_rate": 6.623653733161892e-05, "loss": 0.0622, "step": 42500 }, { "grad_norm": 0.46121397614479065, "learning_rate": 6.622089781903217e-05, "loss": 0.0538, "step": 42510 }, { "grad_norm": 0.4626241624355316, "learning_rate": 6.620525653255241e-05, "loss": 0.0465, "step": 42520 }, { "grad_norm": 0.5109316110610962, "learning_rate": 6.618961347389015e-05, "loss": 0.0476, "step": 42530 }, { "grad_norm": 0.6220110058784485, "learning_rate": 6.617396864475613e-05, "loss": 0.057, "step": 42540 }, { "grad_norm": 0.4773118793964386, "learning_rate": 6.61583220468612e-05, "loss": 0.0493, "step": 42550 }, { "grad_norm": 0.5350674390792847, "learning_rate": 6.614267368191645e-05, "loss": 0.0514, "step": 42560 }, { "grad_norm": 0.45110633969306946, "learning_rate": 6.61270235516332e-05, "loss": 0.0541, "step": 42570 }, { "grad_norm": 0.3821774125099182, "learning_rate": 6.611137165772287e-05, "loss": 0.0396, "step": 42580 }, { "grad_norm": 0.4213019907474518, "learning_rate": 6.60957180018972e-05, "loss": 0.0522, "step": 42590 }, { "grad_norm": 0.4376075565814972, "learning_rate": 6.608006258586797e-05, "loss": 0.0438, "step": 42600 }, { "grad_norm": 0.6121995449066162, "learning_rate": 6.606440541134731e-05, "loss": 0.0604, "step": 42610 }, { "grad_norm": 0.5987108945846558, "learning_rate": 6.60487464800474e-05, "loss": 0.0698, "step": 42620 }, { "grad_norm": 0.6650046110153198, "learning_rate": 6.603308579368071e-05, "loss": 0.0592, "step": 42630 }, { "grad_norm": 0.6373418569564819, "learning_rate": 6.601742335395987e-05, "loss": 0.0512, "step": 42640 }, { "grad_norm": 0.45028552412986755, "learning_rate": 6.600175916259769e-05, "loss": 0.0549, "step": 42650 }, { "grad_norm": 0.514819324016571, "learning_rate": 6.598609322130718e-05, "loss": 0.0434, "step": 42660 }, { "grad_norm": 0.5519298911094666, "learning_rate": 6.597042553180155e-05, "loss": 0.0446, "step": 42670 }, { "grad_norm": 0.5718034505844116, "learning_rate": 6.59547560957942e-05, "loss": 0.0535, "step": 42680 }, { "grad_norm": 0.49539512395858765, "learning_rate": 6.59390849149987e-05, "loss": 0.0587, "step": 42690 }, { "grad_norm": 0.7180852293968201, "learning_rate": 6.592341199112886e-05, "loss": 0.0471, "step": 42700 }, { "grad_norm": 0.3839721083641052, "learning_rate": 6.590773732589861e-05, "loss": 0.0582, "step": 42710 }, { "grad_norm": 0.46953174471855164, "learning_rate": 6.589206092102213e-05, "loss": 0.0582, "step": 42720 }, { "grad_norm": 0.46437644958496094, "learning_rate": 6.587638277821375e-05, "loss": 0.0372, "step": 42730 }, { "grad_norm": 0.40266719460487366, "learning_rate": 6.586070289918803e-05, "loss": 0.0542, "step": 42740 }, { "grad_norm": 0.4858500361442566, "learning_rate": 6.584502128565968e-05, "loss": 0.0561, "step": 42750 }, { "grad_norm": 0.37793827056884766, "learning_rate": 6.582933793934363e-05, "loss": 0.0495, "step": 42760 }, { "grad_norm": 0.4494979679584503, "learning_rate": 6.581365286195498e-05, "loss": 0.049, "step": 42770 }, { "grad_norm": 0.3793158531188965, "learning_rate": 6.579796605520903e-05, "loss": 0.0472, "step": 42780 }, { "grad_norm": 0.5660494565963745, "learning_rate": 6.578227752082127e-05, "loss": 0.0567, "step": 42790 }, { "grad_norm": 0.4124740958213806, "learning_rate": 6.576658726050735e-05, "loss": 0.0444, "step": 42800 }, { "grad_norm": 0.6928635835647583, "learning_rate": 6.575089527598316e-05, "loss": 0.0434, "step": 42810 }, { "grad_norm": 0.3674830198287964, "learning_rate": 6.573520156896476e-05, "loss": 0.0549, "step": 42820 }, { "grad_norm": 0.5404123067855835, "learning_rate": 6.571950614116835e-05, "loss": 0.0513, "step": 42830 }, { "grad_norm": 0.7890752553939819, "learning_rate": 6.570380899431039e-05, "loss": 0.0752, "step": 42840 }, { "grad_norm": 0.8090819716453552, "learning_rate": 6.568811013010749e-05, "loss": 0.056, "step": 42850 }, { "grad_norm": 0.5128852128982544, "learning_rate": 6.567240955027645e-05, "loss": 0.0564, "step": 42860 }, { "grad_norm": 0.572180986404419, "learning_rate": 6.565670725653427e-05, "loss": 0.0363, "step": 42870 }, { "grad_norm": 0.338092178106308, "learning_rate": 6.564100325059812e-05, "loss": 0.0424, "step": 42880 }, { "grad_norm": 0.5190381407737732, "learning_rate": 6.562529753418538e-05, "loss": 0.0429, "step": 42890 }, { "grad_norm": 0.41396352648735046, "learning_rate": 6.56095901090136e-05, "loss": 0.0427, "step": 42900 }, { "grad_norm": 1.2222031354904175, "learning_rate": 6.55938809768005e-05, "loss": 0.044, "step": 42910 }, { "grad_norm": 1.1216768026351929, "learning_rate": 6.557817013926402e-05, "loss": 0.0569, "step": 42920 }, { "grad_norm": 0.46093231439590454, "learning_rate": 6.556245759812227e-05, "loss": 0.0541, "step": 42930 }, { "grad_norm": 0.5986663103103638, "learning_rate": 6.554674335509357e-05, "loss": 0.0542, "step": 42940 }, { "grad_norm": 0.4158187210559845, "learning_rate": 6.553102741189638e-05, "loss": 0.0467, "step": 42950 }, { "grad_norm": 0.6482424139976501, "learning_rate": 6.55153097702494e-05, "loss": 0.0584, "step": 42960 }, { "grad_norm": 0.5897762775421143, "learning_rate": 6.549959043187146e-05, "loss": 0.0536, "step": 42970 }, { "grad_norm": 0.9079786539077759, "learning_rate": 6.54838693984816e-05, "loss": 0.0651, "step": 42980 }, { "grad_norm": 1.0679830312728882, "learning_rate": 6.546814667179909e-05, "loss": 0.0524, "step": 42990 }, { "grad_norm": 0.6172353625297546, "learning_rate": 6.545242225354328e-05, "loss": 0.0565, "step": 43000 }, { "grad_norm": 0.5656562447547913, "learning_rate": 6.543669614543383e-05, "loss": 0.0602, "step": 43010 }, { "grad_norm": 0.5720323324203491, "learning_rate": 6.542096834919049e-05, "loss": 0.0554, "step": 43020 }, { "grad_norm": 0.6212541460990906, "learning_rate": 6.540523886653324e-05, "loss": 0.0529, "step": 43030 }, { "grad_norm": 0.36575955152511597, "learning_rate": 6.538950769918222e-05, "loss": 0.0583, "step": 43040 }, { "grad_norm": 0.43100449442863464, "learning_rate": 6.537377484885779e-05, "loss": 0.0448, "step": 43050 }, { "grad_norm": 0.5585026144981384, "learning_rate": 6.535804031728045e-05, "loss": 0.0514, "step": 43060 }, { "grad_norm": 0.8096919059753418, "learning_rate": 6.534230410617092e-05, "loss": 0.0609, "step": 43070 }, { "grad_norm": 0.8126828074455261, "learning_rate": 6.532656621725007e-05, "loss": 0.054, "step": 43080 }, { "grad_norm": 0.4991389811038971, "learning_rate": 6.531082665223898e-05, "loss": 0.0643, "step": 43090 }, { "grad_norm": 0.5072961449623108, "learning_rate": 6.529508541285889e-05, "loss": 0.0509, "step": 43100 }, { "grad_norm": 0.8928658962249756, "learning_rate": 6.527934250083128e-05, "loss": 0.0592, "step": 43110 }, { "grad_norm": 0.6208712458610535, "learning_rate": 6.526359791787775e-05, "loss": 0.0517, "step": 43120 }, { "grad_norm": 1.0516979694366455, "learning_rate": 6.524785166572009e-05, "loss": 0.0504, "step": 43130 }, { "grad_norm": 0.640070915222168, "learning_rate": 6.523210374608029e-05, "loss": 0.0556, "step": 43140 }, { "grad_norm": 0.3562425971031189, "learning_rate": 6.521635416068054e-05, "loss": 0.0484, "step": 43150 }, { "grad_norm": 0.5741146802902222, "learning_rate": 6.520060291124317e-05, "loss": 0.072, "step": 43160 }, { "grad_norm": 0.4158153533935547, "learning_rate": 6.518484999949072e-05, "loss": 0.048, "step": 43170 }, { "grad_norm": 0.5598315000534058, "learning_rate": 6.516909542714591e-05, "loss": 0.0611, "step": 43180 }, { "grad_norm": 0.9053007960319519, "learning_rate": 6.515333919593161e-05, "loss": 0.0454, "step": 43190 }, { "grad_norm": 0.594326376914978, "learning_rate": 6.513758130757094e-05, "loss": 0.0443, "step": 43200 }, { "grad_norm": 0.9041376113891602, "learning_rate": 6.512182176378713e-05, "loss": 0.0549, "step": 43210 }, { "grad_norm": 0.9503195881843567, "learning_rate": 6.510606056630362e-05, "loss": 0.0489, "step": 43220 }, { "grad_norm": 0.36313867568969727, "learning_rate": 6.509029771684405e-05, "loss": 0.0458, "step": 43230 }, { "grad_norm": 0.6146364212036133, "learning_rate": 6.50745332171322e-05, "loss": 0.0517, "step": 43240 }, { "grad_norm": 0.45613014698028564, "learning_rate": 6.505876706889207e-05, "loss": 0.0609, "step": 43250 }, { "grad_norm": 0.48915982246398926, "learning_rate": 6.504299927384781e-05, "loss": 0.0537, "step": 43260 }, { "grad_norm": 0.6403818726539612, "learning_rate": 6.502722983372378e-05, "loss": 0.0495, "step": 43270 }, { "grad_norm": 0.7281515598297119, "learning_rate": 6.501145875024446e-05, "loss": 0.0554, "step": 43280 }, { "grad_norm": 0.4536122977733612, "learning_rate": 6.499568602513459e-05, "loss": 0.0553, "step": 43290 }, { "grad_norm": 0.4043738842010498, "learning_rate": 6.497991166011903e-05, "loss": 0.0437, "step": 43300 }, { "grad_norm": 0.451488196849823, "learning_rate": 6.49641356569229e-05, "loss": 0.0484, "step": 43310 }, { "grad_norm": 0.3680741786956787, "learning_rate": 6.494835801727135e-05, "loss": 0.0562, "step": 43320 }, { "grad_norm": 0.6124176979064941, "learning_rate": 6.493257874288986e-05, "loss": 0.0583, "step": 43330 }, { "grad_norm": 0.38830676674842834, "learning_rate": 6.491679783550399e-05, "loss": 0.0519, "step": 43340 }, { "grad_norm": 0.5343138575553894, "learning_rate": 6.490101529683957e-05, "loss": 0.0498, "step": 43350 }, { "grad_norm": 0.3394785225391388, "learning_rate": 6.488523112862248e-05, "loss": 0.0572, "step": 43360 }, { "grad_norm": 0.37289127707481384, "learning_rate": 6.486944533257891e-05, "loss": 0.0477, "step": 43370 }, { "grad_norm": 0.615983247756958, "learning_rate": 6.485365791043517e-05, "loss": 0.0567, "step": 43380 }, { "grad_norm": 0.552243173122406, "learning_rate": 6.483786886391772e-05, "loss": 0.0557, "step": 43390 }, { "grad_norm": 0.31244391202926636, "learning_rate": 6.482207819475323e-05, "loss": 0.0407, "step": 43400 }, { "grad_norm": 0.44523459672927856, "learning_rate": 6.480628590466857e-05, "loss": 0.0411, "step": 43410 }, { "grad_norm": 0.43023252487182617, "learning_rate": 6.479049199539074e-05, "loss": 0.0454, "step": 43420 }, { "grad_norm": 0.5714358687400818, "learning_rate": 6.477469646864693e-05, "loss": 0.0532, "step": 43430 }, { "grad_norm": 0.5554543733596802, "learning_rate": 6.475889932616454e-05, "loss": 0.0536, "step": 43440 }, { "grad_norm": 0.5342839360237122, "learning_rate": 6.474310056967111e-05, "loss": 0.0504, "step": 43450 }, { "grad_norm": 0.829482913017273, "learning_rate": 6.472730020089437e-05, "loss": 0.0604, "step": 43460 }, { "grad_norm": 0.36782798171043396, "learning_rate": 6.471149822156222e-05, "loss": 0.0512, "step": 43470 }, { "grad_norm": 0.8415218591690063, "learning_rate": 6.469569463340275e-05, "loss": 0.0502, "step": 43480 }, { "grad_norm": 0.48978909850120544, "learning_rate": 6.467988943814418e-05, "loss": 0.0509, "step": 43490 }, { "grad_norm": 0.5097572207450867, "learning_rate": 6.4664082637515e-05, "loss": 0.0434, "step": 43500 }, { "grad_norm": 0.5394172072410583, "learning_rate": 6.464827423324377e-05, "loss": 0.0493, "step": 43510 }, { "grad_norm": 0.5671629905700684, "learning_rate": 6.463246422705931e-05, "loss": 0.0471, "step": 43520 }, { "grad_norm": 0.5131101012229919, "learning_rate": 6.461665262069055e-05, "loss": 0.0584, "step": 43530 }, { "grad_norm": 0.5028277635574341, "learning_rate": 6.460083941586665e-05, "loss": 0.051, "step": 43540 }, { "grad_norm": 0.5647649168968201, "learning_rate": 6.45850246143169e-05, "loss": 0.0535, "step": 43550 }, { "grad_norm": 0.5615801215171814, "learning_rate": 6.456920821777077e-05, "loss": 0.0516, "step": 43560 }, { "grad_norm": 0.6194661259651184, "learning_rate": 6.455339022795796e-05, "loss": 0.063, "step": 43570 }, { "grad_norm": 0.35646212100982666, "learning_rate": 6.453757064660825e-05, "loss": 0.0457, "step": 43580 }, { "grad_norm": 0.6182308793067932, "learning_rate": 6.452174947545169e-05, "loss": 0.044, "step": 43590 }, { "grad_norm": 0.5046895146369934, "learning_rate": 6.450592671621842e-05, "loss": 0.045, "step": 43600 }, { "grad_norm": 0.7616259455680847, "learning_rate": 6.449010237063882e-05, "loss": 0.0538, "step": 43610 }, { "grad_norm": 0.5261597037315369, "learning_rate": 6.447427644044343e-05, "loss": 0.0476, "step": 43620 }, { "grad_norm": 0.3135374188423157, "learning_rate": 6.44584489273629e-05, "loss": 0.0464, "step": 43630 }, { "grad_norm": 0.8919675350189209, "learning_rate": 6.444261983312817e-05, "loss": 0.048, "step": 43640 }, { "grad_norm": 0.4413302540779114, "learning_rate": 6.442678915947023e-05, "loss": 0.0549, "step": 43650 }, { "grad_norm": 0.636297881603241, "learning_rate": 6.441095690812032e-05, "loss": 0.0493, "step": 43660 }, { "grad_norm": 0.43964818120002747, "learning_rate": 6.439512308080982e-05, "loss": 0.0535, "step": 43670 }, { "grad_norm": 0.3781420886516571, "learning_rate": 6.43792876792703e-05, "loss": 0.053, "step": 43680 }, { "grad_norm": 0.6342905163764954, "learning_rate": 6.43634507052335e-05, "loss": 0.054, "step": 43690 }, { "grad_norm": 0.6172884702682495, "learning_rate": 6.434761216043135e-05, "loss": 0.0553, "step": 43700 }, { "grad_norm": 0.5255053639411926, "learning_rate": 6.433177204659589e-05, "loss": 0.0502, "step": 43710 }, { "grad_norm": 0.41927802562713623, "learning_rate": 6.431593036545938e-05, "loss": 0.0499, "step": 43720 }, { "grad_norm": 0.5950407981872559, "learning_rate": 6.430008711875426e-05, "loss": 0.0416, "step": 43730 }, { "grad_norm": 0.6702409386634827, "learning_rate": 6.42842423082131e-05, "loss": 0.0554, "step": 43740 }, { "grad_norm": 0.7357397675514221, "learning_rate": 6.426839593556868e-05, "loss": 0.0437, "step": 43750 }, { "grad_norm": 0.32206112146377563, "learning_rate": 6.425254800255392e-05, "loss": 0.0484, "step": 43760 }, { "grad_norm": 0.5031231045722961, "learning_rate": 6.423669851090196e-05, "loss": 0.0487, "step": 43770 }, { "grad_norm": 0.6181116104125977, "learning_rate": 6.422084746234604e-05, "loss": 0.043, "step": 43780 }, { "grad_norm": 0.5146856307983398, "learning_rate": 6.420499485861965e-05, "loss": 0.0359, "step": 43790 }, { "grad_norm": 0.421343058347702, "learning_rate": 6.418914070145636e-05, "loss": 0.0558, "step": 43800 }, { "grad_norm": 0.6472384333610535, "learning_rate": 6.417328499258996e-05, "loss": 0.0421, "step": 43810 }, { "grad_norm": 0.5549188256263733, "learning_rate": 6.415742773375446e-05, "loss": 0.0417, "step": 43820 }, { "grad_norm": 0.44009876251220703, "learning_rate": 6.414156892668393e-05, "loss": 0.0465, "step": 43830 }, { "grad_norm": 0.8457179665565491, "learning_rate": 6.412570857311267e-05, "loss": 0.0571, "step": 43840 }, { "grad_norm": 0.4665989577770233, "learning_rate": 6.410984667477518e-05, "loss": 0.0544, "step": 43850 }, { "grad_norm": 0.6759129762649536, "learning_rate": 6.409398323340607e-05, "loss": 0.0469, "step": 43860 }, { "grad_norm": 0.6665030121803284, "learning_rate": 6.407811825074012e-05, "loss": 0.0579, "step": 43870 }, { "grad_norm": 0.4147345721721649, "learning_rate": 6.406225172851234e-05, "loss": 0.0447, "step": 43880 }, { "grad_norm": 0.5094201564788818, "learning_rate": 6.404638366845786e-05, "loss": 0.0547, "step": 43890 }, { "grad_norm": 0.5674648880958557, "learning_rate": 6.403051407231196e-05, "loss": 0.0611, "step": 43900 }, { "grad_norm": 0.7037814259529114, "learning_rate": 6.401464294181016e-05, "loss": 0.0494, "step": 43910 }, { "grad_norm": 0.49308815598487854, "learning_rate": 6.399877027868808e-05, "loss": 0.0461, "step": 43920 }, { "grad_norm": 0.7889434695243835, "learning_rate": 6.39828960846815e-05, "loss": 0.0553, "step": 43930 }, { "grad_norm": 0.3917890787124634, "learning_rate": 6.396702036152644e-05, "loss": 0.0371, "step": 43940 }, { "grad_norm": 0.319211483001709, "learning_rate": 6.395114311095905e-05, "loss": 0.0603, "step": 43950 }, { "grad_norm": 0.4690377712249756, "learning_rate": 6.393526433471559e-05, "loss": 0.0551, "step": 43960 }, { "grad_norm": 0.3538358211517334, "learning_rate": 6.391938403453262e-05, "loss": 0.0687, "step": 43970 }, { "grad_norm": 0.5600312352180481, "learning_rate": 6.390350221214671e-05, "loss": 0.0545, "step": 43980 }, { "grad_norm": 0.4096710681915283, "learning_rate": 6.38876188692947e-05, "loss": 0.057, "step": 43990 }, { "grad_norm": 0.6828655004501343, "learning_rate": 6.387173400771358e-05, "loss": 0.0511, "step": 44000 }, { "grad_norm": 0.32623523473739624, "learning_rate": 6.385584762914047e-05, "loss": 0.0449, "step": 44010 }, { "grad_norm": 0.2231438010931015, "learning_rate": 6.38399597353127e-05, "loss": 0.0459, "step": 44020 }, { "grad_norm": 0.4451238214969635, "learning_rate": 6.382407032796775e-05, "loss": 0.0368, "step": 44030 }, { "grad_norm": 0.5910465717315674, "learning_rate": 6.380817940884325e-05, "loss": 0.046, "step": 44040 }, { "grad_norm": 1.1112136840820312, "learning_rate": 6.379228697967702e-05, "loss": 0.0537, "step": 44050 }, { "grad_norm": 0.36397743225097656, "learning_rate": 6.3776393042207e-05, "loss": 0.0478, "step": 44060 }, { "grad_norm": 0.6655237674713135, "learning_rate": 6.376049759817137e-05, "loss": 0.05, "step": 44070 }, { "grad_norm": 0.47792646288871765, "learning_rate": 6.37446006493084e-05, "loss": 0.0611, "step": 44080 }, { "grad_norm": 0.47042742371559143, "learning_rate": 6.372870219735656e-05, "loss": 0.0637, "step": 44090 }, { "grad_norm": 0.5395575761795044, "learning_rate": 6.371280224405452e-05, "loss": 0.0551, "step": 44100 }, { "grad_norm": 0.6081150770187378, "learning_rate": 6.369690079114101e-05, "loss": 0.0474, "step": 44110 }, { "grad_norm": 0.5160070061683655, "learning_rate": 6.368099784035504e-05, "loss": 0.0525, "step": 44120 }, { "grad_norm": 0.49799999594688416, "learning_rate": 6.366509339343572e-05, "loss": 0.0435, "step": 44130 }, { "grad_norm": 0.6706750988960266, "learning_rate": 6.364918745212232e-05, "loss": 0.0433, "step": 44140 }, { "grad_norm": 0.8884774446487427, "learning_rate": 6.363328001815431e-05, "loss": 0.0509, "step": 44150 }, { "grad_norm": 0.8254438042640686, "learning_rate": 6.361737109327128e-05, "loss": 0.0534, "step": 44160 }, { "grad_norm": 0.41612061858177185, "learning_rate": 6.360146067921301e-05, "loss": 0.0528, "step": 44170 }, { "grad_norm": 0.6215647459030151, "learning_rate": 6.35855487777195e-05, "loss": 0.0555, "step": 44180 }, { "grad_norm": 0.4039926528930664, "learning_rate": 6.356963539053076e-05, "loss": 0.0494, "step": 44190 }, { "grad_norm": 0.3660399913787842, "learning_rate": 6.355372051938711e-05, "loss": 0.0413, "step": 44200 }, { "grad_norm": 0.4009000062942505, "learning_rate": 6.353780416602894e-05, "loss": 0.0451, "step": 44210 }, { "grad_norm": 0.4479503333568573, "learning_rate": 6.352188633219689e-05, "loss": 0.044, "step": 44220 }, { "grad_norm": 0.43483269214630127, "learning_rate": 6.350596701963166e-05, "loss": 0.0378, "step": 44230 }, { "grad_norm": 0.5424792766571045, "learning_rate": 6.349004623007419e-05, "loss": 0.0431, "step": 44240 }, { "grad_norm": 0.3959353566169739, "learning_rate": 6.347412396526555e-05, "loss": 0.0467, "step": 44250 }, { "grad_norm": 0.6632998585700989, "learning_rate": 6.345820022694696e-05, "loss": 0.0576, "step": 44260 }, { "grad_norm": 1.3002885580062866, "learning_rate": 6.344227501685984e-05, "loss": 0.0547, "step": 44270 }, { "grad_norm": 0.6777743697166443, "learning_rate": 6.342634833674572e-05, "loss": 0.0496, "step": 44280 }, { "grad_norm": 0.5223485827445984, "learning_rate": 6.341042018834635e-05, "loss": 0.0668, "step": 44290 }, { "grad_norm": 0.49102485179901123, "learning_rate": 6.339449057340359e-05, "loss": 0.0521, "step": 44300 }, { "grad_norm": 0.3426542282104492, "learning_rate": 6.337855949365945e-05, "loss": 0.0519, "step": 44310 }, { "grad_norm": 0.6707378625869751, "learning_rate": 6.336262695085619e-05, "loss": 0.0476, "step": 44320 }, { "grad_norm": 0.48578041791915894, "learning_rate": 6.334669294673612e-05, "loss": 0.0523, "step": 44330 }, { "grad_norm": 0.45888951420783997, "learning_rate": 6.333075748304179e-05, "loss": 0.0506, "step": 44340 }, { "grad_norm": 0.4286319315433502, "learning_rate": 6.331482056151585e-05, "loss": 0.0525, "step": 44350 }, { "grad_norm": 0.5095784068107605, "learning_rate": 6.329888218390117e-05, "loss": 0.0522, "step": 44360 }, { "grad_norm": 0.37290292978286743, "learning_rate": 6.328294235194072e-05, "loss": 0.0472, "step": 44370 }, { "grad_norm": 0.4963870048522949, "learning_rate": 6.326700106737765e-05, "loss": 0.053, "step": 44380 }, { "grad_norm": 0.4085451662540436, "learning_rate": 6.325105833195531e-05, "loss": 0.0466, "step": 44390 }, { "grad_norm": 0.47618523240089417, "learning_rate": 6.323511414741715e-05, "loss": 0.0464, "step": 44400 }, { "grad_norm": 0.5811125040054321, "learning_rate": 6.321916851550678e-05, "loss": 0.0667, "step": 44410 }, { "grad_norm": 0.9239470958709717, "learning_rate": 6.320322143796806e-05, "loss": 0.0624, "step": 44420 }, { "grad_norm": 0.46349406242370605, "learning_rate": 6.318727291654488e-05, "loss": 0.0431, "step": 44430 }, { "grad_norm": 0.4535265862941742, "learning_rate": 6.317132295298134e-05, "loss": 0.0469, "step": 44440 }, { "grad_norm": 0.49234384298324585, "learning_rate": 6.315537154902173e-05, "loss": 0.0459, "step": 44450 }, { "grad_norm": 0.6001638770103455, "learning_rate": 6.313941870641048e-05, "loss": 0.0465, "step": 44460 }, { "grad_norm": 0.3812771737575531, "learning_rate": 6.312346442689214e-05, "loss": 0.051, "step": 44470 }, { "grad_norm": 1.370751142501831, "learning_rate": 6.310750871221147e-05, "loss": 0.0464, "step": 44480 }, { "grad_norm": 0.41130390763282776, "learning_rate": 6.309155156411335e-05, "loss": 0.0432, "step": 44490 }, { "grad_norm": 0.4850209951400757, "learning_rate": 6.307559298434284e-05, "loss": 0.0471, "step": 44500 }, { "grad_norm": 0.4098154306411743, "learning_rate": 6.305963297464512e-05, "loss": 0.0515, "step": 44510 }, { "grad_norm": 0.6131619811058044, "learning_rate": 6.304367153676561e-05, "loss": 0.0436, "step": 44520 }, { "grad_norm": 0.5305433869361877, "learning_rate": 6.302770867244978e-05, "loss": 0.0484, "step": 44530 }, { "grad_norm": 0.3839280605316162, "learning_rate": 6.301174438344328e-05, "loss": 0.0438, "step": 44540 }, { "grad_norm": 0.5304619073867798, "learning_rate": 6.299577867149202e-05, "loss": 0.0557, "step": 44550 }, { "grad_norm": 0.3746980130672455, "learning_rate": 6.297981153834192e-05, "loss": 0.053, "step": 44560 }, { "grad_norm": 0.731009840965271, "learning_rate": 6.296384298573916e-05, "loss": 0.0561, "step": 44570 }, { "grad_norm": 0.4055543541908264, "learning_rate": 6.294787301543001e-05, "loss": 0.0795, "step": 44580 }, { "grad_norm": 0.4667966663837433, "learning_rate": 6.293190162916095e-05, "loss": 0.052, "step": 44590 }, { "grad_norm": 0.8626589179039001, "learning_rate": 6.291592882867855e-05, "loss": 0.049, "step": 44600 }, { "grad_norm": 0.42491447925567627, "learning_rate": 6.28999546157296e-05, "loss": 0.0395, "step": 44610 }, { "grad_norm": 0.5008488297462463, "learning_rate": 6.288397899206102e-05, "loss": 0.0463, "step": 44620 }, { "grad_norm": 0.7256494164466858, "learning_rate": 6.286800195941984e-05, "loss": 0.0583, "step": 44630 }, { "grad_norm": 0.41945135593414307, "learning_rate": 6.285202351955334e-05, "loss": 0.0462, "step": 44640 }, { "grad_norm": 0.5135199427604675, "learning_rate": 6.283604367420887e-05, "loss": 0.0578, "step": 44650 }, { "grad_norm": 1.0285885334014893, "learning_rate": 6.282006242513394e-05, "loss": 0.0577, "step": 44660 }, { "grad_norm": 0.670760989189148, "learning_rate": 6.280407977407628e-05, "loss": 0.0572, "step": 44670 }, { "grad_norm": 0.5859763026237488, "learning_rate": 6.27880957227837e-05, "loss": 0.0491, "step": 44680 }, { "grad_norm": 0.518949568271637, "learning_rate": 6.27721102730042e-05, "loss": 0.0531, "step": 44690 }, { "grad_norm": 0.7986701726913452, "learning_rate": 6.275612342648592e-05, "loss": 0.0589, "step": 44700 }, { "grad_norm": 0.5799745321273804, "learning_rate": 6.274013518497716e-05, "loss": 0.0533, "step": 44710 }, { "grad_norm": 0.47105005383491516, "learning_rate": 6.272414555022636e-05, "loss": 0.043, "step": 44720 }, { "grad_norm": 0.5975099802017212, "learning_rate": 6.270815452398215e-05, "loss": 0.0446, "step": 44730 }, { "grad_norm": 0.5047740340232849, "learning_rate": 6.269216210799326e-05, "loss": 0.0415, "step": 44740 }, { "grad_norm": 0.5845452547073364, "learning_rate": 6.26761683040086e-05, "loss": 0.0549, "step": 44750 }, { "grad_norm": 0.3764529824256897, "learning_rate": 6.266017311377723e-05, "loss": 0.0468, "step": 44760 }, { "grad_norm": 0.3416135907173157, "learning_rate": 6.264417653904839e-05, "loss": 0.0631, "step": 44770 }, { "grad_norm": 0.652817964553833, "learning_rate": 6.262817858157139e-05, "loss": 0.0556, "step": 44780 }, { "grad_norm": 0.5553444623947144, "learning_rate": 6.261217924309576e-05, "loss": 0.0679, "step": 44790 }, { "grad_norm": 0.5701849460601807, "learning_rate": 6.259617852537118e-05, "loss": 0.0512, "step": 44800 }, { "grad_norm": 0.7757457494735718, "learning_rate": 6.258017643014747e-05, "loss": 0.051, "step": 44810 }, { "grad_norm": 0.5514477491378784, "learning_rate": 6.256417295917456e-05, "loss": 0.0573, "step": 44820 }, { "grad_norm": 0.43136999011039734, "learning_rate": 6.254816811420258e-05, "loss": 0.0579, "step": 44830 }, { "grad_norm": 0.8782244920730591, "learning_rate": 6.253216189698183e-05, "loss": 0.0496, "step": 44840 }, { "grad_norm": 0.29484400153160095, "learning_rate": 6.251615430926267e-05, "loss": 0.0419, "step": 44850 }, { "grad_norm": 0.4846971929073334, "learning_rate": 6.25001453527957e-05, "loss": 0.0507, "step": 44860 }, { "grad_norm": 0.5793951749801636, "learning_rate": 6.248413502933164e-05, "loss": 0.0526, "step": 44870 }, { "grad_norm": 0.5089234113693237, "learning_rate": 6.246812334062133e-05, "loss": 0.0462, "step": 44880 }, { "grad_norm": 0.689256489276886, "learning_rate": 6.245211028841579e-05, "loss": 0.0589, "step": 44890 }, { "grad_norm": 0.4603942632675171, "learning_rate": 6.24360958744662e-05, "loss": 0.0493, "step": 44900 }, { "grad_norm": 0.5126158595085144, "learning_rate": 6.242008010052387e-05, "loss": 0.0585, "step": 44910 }, { "grad_norm": 0.5459936261177063, "learning_rate": 6.240406296834024e-05, "loss": 0.0507, "step": 44920 }, { "grad_norm": 0.42136886715888977, "learning_rate": 6.238804447966694e-05, "loss": 0.0442, "step": 44930 }, { "grad_norm": 0.5000743865966797, "learning_rate": 6.237202463625573e-05, "loss": 0.0391, "step": 44940 }, { "grad_norm": 0.3501845896244049, "learning_rate": 6.235600343985848e-05, "loss": 0.0458, "step": 44950 }, { "grad_norm": 0.5668272972106934, "learning_rate": 6.233998089222729e-05, "loss": 0.0405, "step": 44960 }, { "grad_norm": 0.3121969997882843, "learning_rate": 6.232395699511433e-05, "loss": 0.0535, "step": 44970 }, { "grad_norm": 0.5356642603874207, "learning_rate": 6.230793175027199e-05, "loss": 0.0507, "step": 44980 }, { "grad_norm": 0.40977612137794495, "learning_rate": 6.22919051594527e-05, "loss": 0.0523, "step": 44990 }, { "grad_norm": 0.8547235727310181, "learning_rate": 6.227587722440917e-05, "loss": 0.0576, "step": 45000 }, { "grad_norm": 0.362745463848114, "learning_rate": 6.225984794689414e-05, "loss": 0.0322, "step": 45010 }, { "grad_norm": 0.6238039135932922, "learning_rate": 6.22438173286606e-05, "loss": 0.0657, "step": 45020 }, { "grad_norm": 0.48741084337234497, "learning_rate": 6.222778537146159e-05, "loss": 0.049, "step": 45030 }, { "grad_norm": 0.47091394662857056, "learning_rate": 6.221175207705037e-05, "loss": 0.0462, "step": 45040 }, { "grad_norm": 0.7407410144805908, "learning_rate": 6.219571744718027e-05, "loss": 0.0514, "step": 45050 }, { "grad_norm": 0.4937594532966614, "learning_rate": 6.217968148360487e-05, "loss": 0.0479, "step": 45060 }, { "grad_norm": 0.5102232694625854, "learning_rate": 6.216364418807782e-05, "loss": 0.0478, "step": 45070 }, { "grad_norm": 0.5110750198364258, "learning_rate": 6.214760556235292e-05, "loss": 0.0409, "step": 45080 }, { "grad_norm": 0.5479971170425415, "learning_rate": 6.213156560818414e-05, "loss": 0.0475, "step": 45090 }, { "grad_norm": 0.6558166146278381, "learning_rate": 6.211552432732559e-05, "loss": 0.0598, "step": 45100 }, { "grad_norm": 0.5508372783660889, "learning_rate": 6.20994817215315e-05, "loss": 0.0555, "step": 45110 }, { "grad_norm": 0.5017266869544983, "learning_rate": 6.208343779255628e-05, "loss": 0.0564, "step": 45120 }, { "grad_norm": 0.6476086974143982, "learning_rate": 6.206739254215449e-05, "loss": 0.0562, "step": 45130 }, { "grad_norm": 0.7238966226577759, "learning_rate": 6.205134597208077e-05, "loss": 0.0701, "step": 45140 }, { "grad_norm": 0.45482003688812256, "learning_rate": 6.203529808409e-05, "loss": 0.052, "step": 45150 }, { "grad_norm": 0.4807889461517334, "learning_rate": 6.201924887993712e-05, "loss": 0.0582, "step": 45160 }, { "grad_norm": 0.5096848011016846, "learning_rate": 6.200319836137725e-05, "loss": 0.0519, "step": 45170 }, { "grad_norm": 0.3479064702987671, "learning_rate": 6.198714653016565e-05, "loss": 0.0461, "step": 45180 }, { "grad_norm": 0.461501806974411, "learning_rate": 6.197109338805774e-05, "loss": 0.0481, "step": 45190 }, { "grad_norm": 0.565261721611023, "learning_rate": 6.195503893680903e-05, "loss": 0.0483, "step": 45200 }, { "grad_norm": 0.8567129969596863, "learning_rate": 6.193898317817524e-05, "loss": 0.0528, "step": 45210 }, { "grad_norm": 1.9285078048706055, "learning_rate": 6.192292611391221e-05, "loss": 0.052, "step": 45220 }, { "grad_norm": 0.3803681433200836, "learning_rate": 6.190686774577591e-05, "loss": 0.0495, "step": 45230 }, { "grad_norm": 0.590234100818634, "learning_rate": 6.189080807552245e-05, "loss": 0.0575, "step": 45240 }, { "grad_norm": 0.6356346607208252, "learning_rate": 6.187474710490809e-05, "loss": 0.0501, "step": 45250 }, { "grad_norm": 0.4645380973815918, "learning_rate": 6.185868483568926e-05, "loss": 0.0499, "step": 45260 }, { "grad_norm": 0.6069352030754089, "learning_rate": 6.184262126962245e-05, "loss": 0.0563, "step": 45270 }, { "grad_norm": 0.5554373860359192, "learning_rate": 6.182655640846442e-05, "loss": 0.0387, "step": 45280 }, { "grad_norm": 0.5057421326637268, "learning_rate": 6.181049025397196e-05, "loss": 0.049, "step": 45290 }, { "grad_norm": 0.39008939266204834, "learning_rate": 6.179442280790202e-05, "loss": 0.0553, "step": 45300 }, { "grad_norm": 0.48666322231292725, "learning_rate": 6.177835407201174e-05, "loss": 0.0566, "step": 45310 }, { "grad_norm": 0.6735435128211975, "learning_rate": 6.176228404805839e-05, "loss": 0.0504, "step": 45320 }, { "grad_norm": 0.47231969237327576, "learning_rate": 6.174621273779932e-05, "loss": 0.0502, "step": 45330 }, { "grad_norm": 0.5310764312744141, "learning_rate": 6.17301401429921e-05, "loss": 0.047, "step": 45340 }, { "grad_norm": 0.2683596611022949, "learning_rate": 6.171406626539436e-05, "loss": 0.053, "step": 45350 }, { "grad_norm": 0.7440733313560486, "learning_rate": 6.169799110676398e-05, "loss": 0.0651, "step": 45360 }, { "grad_norm": 0.43838366866111755, "learning_rate": 6.168191466885885e-05, "loss": 0.0458, "step": 45370 }, { "grad_norm": 0.6181186437606812, "learning_rate": 6.166583695343711e-05, "loss": 0.0536, "step": 45380 }, { "grad_norm": 0.36095112562179565, "learning_rate": 6.164975796225698e-05, "loss": 0.0608, "step": 45390 }, { "grad_norm": 0.46256694197654724, "learning_rate": 6.163367769707683e-05, "loss": 0.0419, "step": 45400 }, { "grad_norm": 0.7710981965065002, "learning_rate": 6.161759615965519e-05, "loss": 0.0495, "step": 45410 }, { "grad_norm": 0.4341622591018677, "learning_rate": 6.16015133517507e-05, "loss": 0.0536, "step": 45420 }, { "grad_norm": 0.5225481390953064, "learning_rate": 6.158542927512214e-05, "loss": 0.0413, "step": 45430 }, { "grad_norm": 0.6532172560691833, "learning_rate": 6.156934393152846e-05, "loss": 0.0521, "step": 45440 }, { "grad_norm": 0.4777357578277588, "learning_rate": 6.15532573227287e-05, "loss": 0.0628, "step": 45450 }, { "grad_norm": 0.7306596040725708, "learning_rate": 6.153716945048212e-05, "loss": 0.0627, "step": 45460 }, { "grad_norm": 0.5765838623046875, "learning_rate": 6.152108031654802e-05, "loss": 0.0587, "step": 45470 }, { "grad_norm": 0.4136897027492523, "learning_rate": 6.15049899226859e-05, "loss": 0.0488, "step": 45480 }, { "grad_norm": 0.4838572144508362, "learning_rate": 6.148889827065537e-05, "loss": 0.0511, "step": 45490 }, { "grad_norm": 0.4068668782711029, "learning_rate": 6.147280536221622e-05, "loss": 0.053, "step": 45500 }, { "grad_norm": 0.7404137253761292, "learning_rate": 6.145671119912832e-05, "loss": 0.0533, "step": 45510 }, { "grad_norm": 0.5481597185134888, "learning_rate": 6.144061578315169e-05, "loss": 0.0478, "step": 45520 }, { "grad_norm": 0.4027351438999176, "learning_rate": 6.142451911604654e-05, "loss": 0.0512, "step": 45530 }, { "grad_norm": 0.4761747717857361, "learning_rate": 6.140842119957315e-05, "loss": 0.0552, "step": 45540 }, { "grad_norm": 0.6810358166694641, "learning_rate": 6.139232203549197e-05, "loss": 0.0475, "step": 45550 }, { "grad_norm": 0.6759286522865295, "learning_rate": 6.137622162556357e-05, "loss": 0.0526, "step": 45560 }, { "grad_norm": 0.3286142647266388, "learning_rate": 6.13601199715487e-05, "loss": 0.0408, "step": 45570 }, { "grad_norm": 0.47679150104522705, "learning_rate": 6.134401707520816e-05, "loss": 0.0543, "step": 45580 }, { "grad_norm": 0.33131271600723267, "learning_rate": 6.132791293830298e-05, "loss": 0.0481, "step": 45590 }, { "grad_norm": 0.391975462436676, "learning_rate": 6.131180756259428e-05, "loss": 0.0488, "step": 45600 }, { "grad_norm": 0.4302375018596649, "learning_rate": 6.129570094984331e-05, "loss": 0.0426, "step": 45610 }, { "grad_norm": 0.770124077796936, "learning_rate": 6.127959310181145e-05, "loss": 0.0617, "step": 45620 }, { "grad_norm": 0.6312257647514343, "learning_rate": 6.126348402026026e-05, "loss": 0.0521, "step": 45630 }, { "grad_norm": 0.4698190987110138, "learning_rate": 6.12473737069514e-05, "loss": 0.0439, "step": 45640 }, { "grad_norm": 0.6978480815887451, "learning_rate": 6.123126216364665e-05, "loss": 0.049, "step": 45650 }, { "grad_norm": 0.60733962059021, "learning_rate": 6.121514939210797e-05, "loss": 0.0383, "step": 45660 }, { "grad_norm": 0.4942692816257477, "learning_rate": 6.119903539409741e-05, "loss": 0.0579, "step": 45670 }, { "grad_norm": 0.2993677258491516, "learning_rate": 6.118292017137716e-05, "loss": 0.0419, "step": 45680 }, { "grad_norm": 0.5800169110298157, "learning_rate": 6.116680372570959e-05, "loss": 0.0491, "step": 45690 }, { "grad_norm": 0.43632590770721436, "learning_rate": 6.115068605885713e-05, "loss": 0.037, "step": 45700 }, { "grad_norm": 0.521248459815979, "learning_rate": 6.113456717258243e-05, "loss": 0.0533, "step": 45710 }, { "grad_norm": 0.3850075304508209, "learning_rate": 6.11184470686482e-05, "loss": 0.0477, "step": 45720 }, { "grad_norm": 0.4967309236526489, "learning_rate": 6.110232574881731e-05, "loss": 0.047, "step": 45730 }, { "grad_norm": 0.7626271843910217, "learning_rate": 6.108620321485277e-05, "loss": 0.0525, "step": 45740 }, { "grad_norm": 0.42464232444763184, "learning_rate": 6.107007946851773e-05, "loss": 0.0533, "step": 45750 }, { "grad_norm": 0.5586748719215393, "learning_rate": 6.105395451157542e-05, "loss": 0.0554, "step": 45760 }, { "grad_norm": 0.36966365575790405, "learning_rate": 6.103782834578928e-05, "loss": 0.0447, "step": 45770 }, { "grad_norm": 0.44568029046058655, "learning_rate": 6.102170097292281e-05, "loss": 0.0481, "step": 45780 }, { "grad_norm": 0.9482623934745789, "learning_rate": 6.1005572394739716e-05, "loss": 0.0574, "step": 45790 }, { "grad_norm": 0.47696998715400696, "learning_rate": 6.0989442613003765e-05, "loss": 0.039, "step": 45800 }, { "grad_norm": 0.44514235854148865, "learning_rate": 6.097331162947888e-05, "loss": 0.0447, "step": 45810 }, { "grad_norm": 0.46227017045021057, "learning_rate": 6.095717944592914e-05, "loss": 0.0524, "step": 45820 }, { "grad_norm": 0.7550222873687744, "learning_rate": 6.094104606411873e-05, "loss": 0.0531, "step": 45830 }, { "grad_norm": 1.3608393669128418, "learning_rate": 6.0924911485811966e-05, "loss": 0.0516, "step": 45840 }, { "grad_norm": 0.7627167701721191, "learning_rate": 6.090877571277331e-05, "loss": 0.0585, "step": 45850 }, { "grad_norm": 0.41321325302124023, "learning_rate": 6.089263874676736e-05, "loss": 0.0429, "step": 45860 }, { "grad_norm": 0.6733171939849854, "learning_rate": 6.0876500589558796e-05, "loss": 0.0545, "step": 45870 }, { "grad_norm": 0.8403815031051636, "learning_rate": 6.086036124291248e-05, "loss": 0.055, "step": 45880 }, { "grad_norm": 0.5056726336479187, "learning_rate": 6.084422070859339e-05, "loss": 0.0483, "step": 45890 }, { "grad_norm": 0.7185901999473572, "learning_rate": 6.082807898836663e-05, "loss": 0.0501, "step": 45900 }, { "grad_norm": 0.4171982407569885, "learning_rate": 6.081193608399742e-05, "loss": 0.0439, "step": 45910 }, { "grad_norm": 0.4988766610622406, "learning_rate": 6.0795791997251164e-05, "loss": 0.0399, "step": 45920 }, { "grad_norm": 0.616547167301178, "learning_rate": 6.0779646729893294e-05, "loss": 0.0418, "step": 45930 }, { "grad_norm": 0.3928239047527313, "learning_rate": 6.0763500283689476e-05, "loss": 0.0396, "step": 45940 }, { "grad_norm": 0.4362286627292633, "learning_rate": 6.0747352660405455e-05, "loss": 0.0499, "step": 45950 }, { "grad_norm": 0.4261334240436554, "learning_rate": 6.073120386180709e-05, "loss": 0.051, "step": 45960 }, { "grad_norm": 0.4673847556114197, "learning_rate": 6.0715053889660425e-05, "loss": 0.0488, "step": 45970 }, { "grad_norm": 0.46818605065345764, "learning_rate": 6.069890274573157e-05, "loss": 0.041, "step": 45980 }, { "grad_norm": 0.6035898923873901, "learning_rate": 6.068275043178679e-05, "loss": 0.0632, "step": 45990 }, { "grad_norm": 1.3680574893951416, "learning_rate": 6.066659694959248e-05, "loss": 0.043, "step": 46000 }, { "grad_norm": 0.3357880115509033, "learning_rate": 6.0650442300915176e-05, "loss": 0.0451, "step": 46010 }, { "grad_norm": 0.3956668972969055, "learning_rate": 6.063428648752152e-05, "loss": 0.0651, "step": 46020 }, { "grad_norm": 0.5431184768676758, "learning_rate": 6.0618129511178266e-05, "loss": 0.04, "step": 46030 }, { "grad_norm": 0.27978649735450745, "learning_rate": 6.060197137365234e-05, "loss": 0.0544, "step": 46040 }, { "grad_norm": 0.6021347641944885, "learning_rate": 6.058581207671077e-05, "loss": 0.044, "step": 46050 }, { "grad_norm": 0.6140959858894348, "learning_rate": 6.056965162212072e-05, "loss": 0.0586, "step": 46060 }, { "grad_norm": 0.33189448714256287, "learning_rate": 6.055349001164943e-05, "loss": 0.0482, "step": 46070 }, { "grad_norm": 0.38199499249458313, "learning_rate": 6.0537327247064347e-05, "loss": 0.0527, "step": 46080 }, { "grad_norm": 0.6426945328712463, "learning_rate": 6.0521163330133014e-05, "loss": 0.0572, "step": 46090 }, { "grad_norm": 0.3979314863681793, "learning_rate": 6.050499826262306e-05, "loss": 0.0469, "step": 46100 }, { "grad_norm": 0.5415053963661194, "learning_rate": 6.0488832046302294e-05, "loss": 0.0581, "step": 46110 }, { "grad_norm": 0.4324088990688324, "learning_rate": 6.0472664682938626e-05, "loss": 0.0442, "step": 46120 }, { "grad_norm": 0.4432258903980255, "learning_rate": 6.045649617430009e-05, "loss": 0.0492, "step": 46130 }, { "grad_norm": 0.4963376522064209, "learning_rate": 6.0440326522154866e-05, "loss": 0.0451, "step": 46140 }, { "grad_norm": 0.46834686398506165, "learning_rate": 6.0424155728271224e-05, "loss": 0.0502, "step": 46150 }, { "grad_norm": 0.29764223098754883, "learning_rate": 6.040798379441758e-05, "loss": 0.0468, "step": 46160 }, { "grad_norm": 0.4211848974227905, "learning_rate": 6.0391810722362485e-05, "loss": 0.0502, "step": 46170 }, { "grad_norm": 0.4727296233177185, "learning_rate": 6.037563651387458e-05, "loss": 0.0451, "step": 46180 }, { "grad_norm": 0.5079716444015503, "learning_rate": 6.0359461170722666e-05, "loss": 0.0409, "step": 46190 }, { "grad_norm": 0.5093175768852234, "learning_rate": 6.034328469467566e-05, "loss": 0.0442, "step": 46200 }, { "grad_norm": 0.552979052066803, "learning_rate": 6.0327107087502596e-05, "loss": 0.0465, "step": 46210 }, { "grad_norm": 0.5436245799064636, "learning_rate": 6.031092835097262e-05, "loss": 0.0445, "step": 46220 }, { "grad_norm": 0.41038545966148376, "learning_rate": 6.0294748486855024e-05, "loss": 0.0479, "step": 46230 }, { "grad_norm": 2.7473270893096924, "learning_rate": 6.0278567496919216e-05, "loss": 0.0567, "step": 46240 }, { "grad_norm": 0.6431238055229187, "learning_rate": 6.026238538293472e-05, "loss": 0.0494, "step": 46250 }, { "grad_norm": 0.4735848903656006, "learning_rate": 6.024620214667118e-05, "loss": 0.0544, "step": 46260 }, { "grad_norm": 0.5438796281814575, "learning_rate": 6.0230017789898384e-05, "loss": 0.0481, "step": 46270 }, { "grad_norm": 3.7800893783569336, "learning_rate": 6.021383231438622e-05, "loss": 0.0447, "step": 46280 }, { "grad_norm": 0.4622669219970703, "learning_rate": 6.0197645721904704e-05, "loss": 0.043, "step": 46290 }, { "grad_norm": 0.5913710594177246, "learning_rate": 6.0181458014224e-05, "loss": 0.0473, "step": 46300 }, { "grad_norm": 0.5830926299095154, "learning_rate": 6.016526919311435e-05, "loss": 0.0508, "step": 46310 }, { "grad_norm": 0.3593362867832184, "learning_rate": 6.014907926034613e-05, "loss": 0.0429, "step": 46320 }, { "grad_norm": 0.5666052103042603, "learning_rate": 6.0132888217689875e-05, "loss": 0.05, "step": 46330 }, { "grad_norm": 0.49761930108070374, "learning_rate": 6.01166960669162e-05, "loss": 0.0526, "step": 46340 }, { "grad_norm": 0.43794217705726624, "learning_rate": 6.0100502809795845e-05, "loss": 0.0477, "step": 46350 }, { "grad_norm": 0.33494099974632263, "learning_rate": 6.00843084480997e-05, "loss": 0.0526, "step": 46360 }, { "grad_norm": 0.32630202174186707, "learning_rate": 6.0068112983598736e-05, "loss": 0.043, "step": 46370 }, { "grad_norm": 0.4151829779148102, "learning_rate": 6.005191641806407e-05, "loss": 0.0421, "step": 46380 }, { "grad_norm": 0.5226445198059082, "learning_rate": 6.003571875326694e-05, "loss": 0.0479, "step": 46390 }, { "grad_norm": 0.6895363926887512, "learning_rate": 6.00195199909787e-05, "loss": 0.0505, "step": 46400 }, { "grad_norm": 0.47908711433410645, "learning_rate": 6.000332013297082e-05, "loss": 0.0472, "step": 46410 }, { "grad_norm": 0.36588624119758606, "learning_rate": 5.998711918101487e-05, "loss": 0.0488, "step": 46420 }, { "grad_norm": 0.2926974296569824, "learning_rate": 5.997091713688261e-05, "loss": 0.0511, "step": 46430 }, { "grad_norm": 0.27891287207603455, "learning_rate": 5.9954714002345836e-05, "loss": 0.0444, "step": 46440 }, { "grad_norm": 0.43990007042884827, "learning_rate": 5.993850977917649e-05, "loss": 0.0537, "step": 46450 }, { "grad_norm": 0.5258760452270508, "learning_rate": 5.992230446914667e-05, "loss": 0.044, "step": 46460 }, { "grad_norm": 0.4101724624633789, "learning_rate": 5.990609807402855e-05, "loss": 0.0504, "step": 46470 }, { "grad_norm": 0.5963112711906433, "learning_rate": 5.988989059559443e-05, "loss": 0.0525, "step": 46480 }, { "grad_norm": 0.6370581388473511, "learning_rate": 5.987368203561675e-05, "loss": 0.0469, "step": 46490 }, { "grad_norm": 0.4657954275608063, "learning_rate": 5.9857472395868055e-05, "loss": 0.0566, "step": 46500 }, { "grad_norm": 0.46872571110725403, "learning_rate": 5.9841261678120983e-05, "loss": 0.0496, "step": 46510 }, { "grad_norm": 0.4659460783004761, "learning_rate": 5.982504988414834e-05, "loss": 0.0382, "step": 46520 }, { "grad_norm": 0.4402855932712555, "learning_rate": 5.9808837015723015e-05, "loss": 0.0571, "step": 46530 }, { "grad_norm": 0.32735294103622437, "learning_rate": 5.9792623074618016e-05, "loss": 0.0472, "step": 46540 }, { "grad_norm": 0.43391314148902893, "learning_rate": 5.977640806260648e-05, "loss": 0.0502, "step": 46550 }, { "grad_norm": 0.3277175724506378, "learning_rate": 5.976019198146166e-05, "loss": 0.0438, "step": 46560 }, { "grad_norm": 0.4982531666755676, "learning_rate": 5.974397483295692e-05, "loss": 0.039, "step": 46570 }, { "grad_norm": 0.42382749915122986, "learning_rate": 5.972775661886575e-05, "loss": 0.0429, "step": 46580 }, { "grad_norm": 0.506373405456543, "learning_rate": 5.971153734096173e-05, "loss": 0.0429, "step": 46590 }, { "grad_norm": 0.4894018769264221, "learning_rate": 5.969531700101859e-05, "loss": 0.0408, "step": 46600 }, { "grad_norm": 0.4647674560546875, "learning_rate": 5.9679095600810155e-05, "loss": 0.0517, "step": 46610 }, { "grad_norm": 0.5164044499397278, "learning_rate": 5.9662873142110384e-05, "loss": 0.0409, "step": 46620 }, { "grad_norm": 0.594279944896698, "learning_rate": 5.964664962669333e-05, "loss": 0.0516, "step": 46630 }, { "grad_norm": 0.5373843312263489, "learning_rate": 5.9630425056333186e-05, "loss": 0.0494, "step": 46640 }, { "grad_norm": 0.30631789565086365, "learning_rate": 5.961419943280422e-05, "loss": 0.0554, "step": 46650 }, { "grad_norm": 0.9968366622924805, "learning_rate": 5.959797275788087e-05, "loss": 0.0447, "step": 46660 }, { "grad_norm": 0.6760167479515076, "learning_rate": 5.958174503333765e-05, "loss": 0.0631, "step": 46670 }, { "grad_norm": 0.45008257031440735, "learning_rate": 5.9565516260949195e-05, "loss": 0.0554, "step": 46680 }, { "grad_norm": 0.5631970167160034, "learning_rate": 5.9549286442490273e-05, "loss": 0.0477, "step": 46690 }, { "grad_norm": 0.4927554428577423, "learning_rate": 5.953305557973572e-05, "loss": 0.0497, "step": 46700 }, { "grad_norm": 0.720245897769928, "learning_rate": 5.9516823674460564e-05, "loss": 0.045, "step": 46710 }, { "grad_norm": 0.5688841938972473, "learning_rate": 5.9500590728439875e-05, "loss": 0.0478, "step": 46720 }, { "grad_norm": 0.38537776470184326, "learning_rate": 5.948435674344886e-05, "loss": 0.0539, "step": 46730 }, { "grad_norm": 0.5890733003616333, "learning_rate": 5.946812172126285e-05, "loss": 0.0484, "step": 46740 }, { "grad_norm": 0.42008501291275024, "learning_rate": 5.945188566365729e-05, "loss": 0.0444, "step": 46750 }, { "grad_norm": 0.4907808005809784, "learning_rate": 5.943564857240773e-05, "loss": 0.0499, "step": 46760 }, { "grad_norm": 0.47315651178359985, "learning_rate": 5.941941044928981e-05, "loss": 0.0478, "step": 46770 }, { "grad_norm": 0.4997006058692932, "learning_rate": 5.940317129607935e-05, "loss": 0.0463, "step": 46780 }, { "grad_norm": 0.4154426157474518, "learning_rate": 5.9386931114552204e-05, "loss": 0.0452, "step": 46790 }, { "grad_norm": 0.6899542212486267, "learning_rate": 5.937068990648438e-05, "loss": 0.045, "step": 46800 }, { "grad_norm": 0.578115701675415, "learning_rate": 5.935444767365199e-05, "loss": 0.0675, "step": 46810 }, { "grad_norm": 0.4549154043197632, "learning_rate": 5.933820441783129e-05, "loss": 0.0578, "step": 46820 }, { "grad_norm": 0.5794663429260254, "learning_rate": 5.932196014079857e-05, "loss": 0.0499, "step": 46830 }, { "grad_norm": 0.42874860763549805, "learning_rate": 5.930571484433032e-05, "loss": 0.0362, "step": 46840 }, { "grad_norm": 0.49567267298698425, "learning_rate": 5.928946853020309e-05, "loss": 0.0478, "step": 46850 }, { "grad_norm": 0.38446030020713806, "learning_rate": 5.927322120019352e-05, "loss": 0.0424, "step": 46860 }, { "grad_norm": 0.6145201325416565, "learning_rate": 5.925697285607844e-05, "loss": 0.0601, "step": 46870 }, { "grad_norm": 0.8904223442077637, "learning_rate": 5.924072349963472e-05, "loss": 0.055, "step": 46880 }, { "grad_norm": 0.39328405261039734, "learning_rate": 5.922447313263938e-05, "loss": 0.0436, "step": 46890 }, { "grad_norm": 0.4097486436367035, "learning_rate": 5.920822175686952e-05, "loss": 0.0549, "step": 46900 }, { "grad_norm": 0.4679967761039734, "learning_rate": 5.919196937410239e-05, "loss": 0.0669, "step": 46910 }, { "grad_norm": 0.45524078607559204, "learning_rate": 5.917571598611529e-05, "loss": 0.0497, "step": 46920 }, { "grad_norm": 0.7863963842391968, "learning_rate": 5.9159461594685686e-05, "loss": 0.0475, "step": 46930 }, { "grad_norm": 0.6132233142852783, "learning_rate": 5.914320620159114e-05, "loss": 0.0478, "step": 46940 }, { "grad_norm": 0.5286682844161987, "learning_rate": 5.912694980860931e-05, "loss": 0.0536, "step": 46950 }, { "grad_norm": 0.8129279017448425, "learning_rate": 5.9110692417517964e-05, "loss": 0.0535, "step": 46960 }, { "grad_norm": 0.5547516942024231, "learning_rate": 5.9094434030095e-05, "loss": 0.0654, "step": 46970 }, { "grad_norm": 0.4970403015613556, "learning_rate": 5.907817464811841e-05, "loss": 0.0427, "step": 46980 }, { "grad_norm": 0.5361076593399048, "learning_rate": 5.906191427336627e-05, "loss": 0.0533, "step": 46990 }, { "grad_norm": 0.7033407688140869, "learning_rate": 5.9045652907616835e-05, "loss": 0.0489, "step": 47000 }, { "grad_norm": 0.5555169582366943, "learning_rate": 5.902939055264838e-05, "loss": 0.0521, "step": 47010 }, { "grad_norm": 0.5989909768104553, "learning_rate": 5.901312721023935e-05, "loss": 0.0539, "step": 47020 }, { "grad_norm": 0.40663471817970276, "learning_rate": 5.8996862882168294e-05, "loss": 0.042, "step": 47030 }, { "grad_norm": 0.45989924669265747, "learning_rate": 5.8980597570213826e-05, "loss": 0.0639, "step": 47040 }, { "grad_norm": 1.2602695226669312, "learning_rate": 5.896433127615471e-05, "loss": 0.0577, "step": 47050 }, { "grad_norm": 0.6829351186752319, "learning_rate": 5.894806400176981e-05, "loss": 0.0416, "step": 47060 }, { "grad_norm": 0.7350665330886841, "learning_rate": 5.893179574883808e-05, "loss": 0.0505, "step": 47070 }, { "grad_norm": 0.4722576439380646, "learning_rate": 5.8915526519138585e-05, "loss": 0.0613, "step": 47080 }, { "grad_norm": 0.7104662656784058, "learning_rate": 5.889925631445053e-05, "loss": 0.0585, "step": 47090 }, { "grad_norm": 0.5040118098258972, "learning_rate": 5.888298513655318e-05, "loss": 0.0379, "step": 47100 }, { "grad_norm": 0.39570754766464233, "learning_rate": 5.886671298722592e-05, "loss": 0.0427, "step": 47110 }, { "grad_norm": 0.685976505279541, "learning_rate": 5.885043986824828e-05, "loss": 0.0437, "step": 47120 }, { "grad_norm": 0.6480796933174133, "learning_rate": 5.883416578139982e-05, "loss": 0.046, "step": 47130 }, { "grad_norm": 0.36578816175460815, "learning_rate": 5.881789072846029e-05, "loss": 0.052, "step": 47140 }, { "grad_norm": 0.7353857159614563, "learning_rate": 5.8801614711209474e-05, "loss": 0.0453, "step": 47150 }, { "grad_norm": 0.43838179111480713, "learning_rate": 5.878533773142732e-05, "loss": 0.0458, "step": 47160 }, { "grad_norm": 0.3717823624610901, "learning_rate": 5.8769059790893834e-05, "loss": 0.0528, "step": 47170 }, { "grad_norm": 0.5673952102661133, "learning_rate": 5.8752780891389145e-05, "loss": 0.06, "step": 47180 }, { "grad_norm": 0.7981929183006287, "learning_rate": 5.873650103469351e-05, "loss": 0.0469, "step": 47190 }, { "grad_norm": 0.2744491398334503, "learning_rate": 5.872022022258725e-05, "loss": 0.0375, "step": 47200 }, { "grad_norm": 0.38457050919532776, "learning_rate": 5.8703938456850814e-05, "loss": 0.0403, "step": 47210 }, { "grad_norm": 0.7950118184089661, "learning_rate": 5.868765573926477e-05, "loss": 0.0566, "step": 47220 }, { "grad_norm": 0.45530036091804504, "learning_rate": 5.867137207160974e-05, "loss": 0.0643, "step": 47230 }, { "grad_norm": 0.4815952479839325, "learning_rate": 5.86550874556665e-05, "loss": 0.0512, "step": 47240 }, { "grad_norm": 0.2721329629421234, "learning_rate": 5.863880189321592e-05, "loss": 0.0443, "step": 47250 }, { "grad_norm": 0.4947353005409241, "learning_rate": 5.862251538603894e-05, "loss": 0.0405, "step": 47260 }, { "grad_norm": 0.3637520670890808, "learning_rate": 5.8606227935916656e-05, "loss": 0.038, "step": 47270 }, { "grad_norm": 0.4022006094455719, "learning_rate": 5.858993954463021e-05, "loss": 0.0475, "step": 47280 }, { "grad_norm": 0.4108688235282898, "learning_rate": 5.8573650213960896e-05, "loss": 0.0481, "step": 47290 }, { "grad_norm": 0.4498973786830902, "learning_rate": 5.8557359945690084e-05, "loss": 0.0493, "step": 47300 }, { "grad_norm": 0.4358050227165222, "learning_rate": 5.8541068741599246e-05, "loss": 0.0553, "step": 47310 }, { "grad_norm": 0.5152832865715027, "learning_rate": 5.8524776603469985e-05, "loss": 0.0412, "step": 47320 }, { "grad_norm": 0.45693913102149963, "learning_rate": 5.850848353308397e-05, "loss": 0.0429, "step": 47330 }, { "grad_norm": 1.0025516748428345, "learning_rate": 5.849218953222297e-05, "loss": 0.0574, "step": 47340 }, { "grad_norm": 0.5523484945297241, "learning_rate": 5.847589460266891e-05, "loss": 0.0471, "step": 47350 }, { "grad_norm": 0.8085948824882507, "learning_rate": 5.845959874620375e-05, "loss": 0.0498, "step": 47360 }, { "grad_norm": 0.3871040344238281, "learning_rate": 5.84433019646096e-05, "loss": 0.0531, "step": 47370 }, { "grad_norm": 1.0040357112884521, "learning_rate": 5.842700425966863e-05, "loss": 0.0587, "step": 47380 }, { "grad_norm": 0.4343281686306, "learning_rate": 5.841070563316315e-05, "loss": 0.0475, "step": 47390 }, { "grad_norm": 0.5648195743560791, "learning_rate": 5.839440608687554e-05, "loss": 0.0471, "step": 47400 }, { "grad_norm": 0.5275601744651794, "learning_rate": 5.837810562258831e-05, "loss": 0.0406, "step": 47410 }, { "grad_norm": 0.5799970626831055, "learning_rate": 5.836180424208405e-05, "loss": 0.0689, "step": 47420 }, { "grad_norm": 0.41189172863960266, "learning_rate": 5.834550194714543e-05, "loss": 0.0472, "step": 47430 }, { "grad_norm": 0.43980687856674194, "learning_rate": 5.832919873955528e-05, "loss": 0.0646, "step": 47440 }, { "grad_norm": 0.6298409104347229, "learning_rate": 5.831289462109648e-05, "loss": 0.0545, "step": 47450 }, { "grad_norm": 0.2963334918022156, "learning_rate": 5.829658959355202e-05, "loss": 0.051, "step": 47460 }, { "grad_norm": 0.4188237190246582, "learning_rate": 5.8280283658704995e-05, "loss": 0.0568, "step": 47470 }, { "grad_norm": 0.5554720163345337, "learning_rate": 5.826397681833859e-05, "loss": 0.0517, "step": 47480 }, { "grad_norm": 0.4187983572483063, "learning_rate": 5.824766907423612e-05, "loss": 0.0519, "step": 47490 }, { "grad_norm": 0.37097448110580444, "learning_rate": 5.8231360428180945e-05, "loss": 0.0481, "step": 47500 }, { "grad_norm": 0.5164404511451721, "learning_rate": 5.821505088195658e-05, "loss": 0.0485, "step": 47510 }, { "grad_norm": 0.5409079194068909, "learning_rate": 5.819874043734661e-05, "loss": 0.0629, "step": 47520 }, { "grad_norm": 0.6226065754890442, "learning_rate": 5.8182429096134695e-05, "loss": 0.0459, "step": 47530 }, { "grad_norm": 0.687758207321167, "learning_rate": 5.816611686010465e-05, "loss": 0.0462, "step": 47540 }, { "grad_norm": 0.4268672466278076, "learning_rate": 5.814980373104033e-05, "loss": 0.0524, "step": 47550 }, { "grad_norm": 0.4592325687408447, "learning_rate": 5.813348971072572e-05, "loss": 0.0593, "step": 47560 }, { "grad_norm": 0.4653462767601013, "learning_rate": 5.811717480094492e-05, "loss": 0.0476, "step": 47570 }, { "grad_norm": 0.48008763790130615, "learning_rate": 5.810085900348209e-05, "loss": 0.0491, "step": 47580 }, { "grad_norm": 0.8519805669784546, "learning_rate": 5.8084542320121483e-05, "loss": 0.0577, "step": 47590 }, { "grad_norm": 0.5466994643211365, "learning_rate": 5.8068224752647497e-05, "loss": 0.0663, "step": 47600 }, { "grad_norm": 0.39379647374153137, "learning_rate": 5.805190630284457e-05, "loss": 0.0356, "step": 47610 }, { "grad_norm": 0.4488788843154907, "learning_rate": 5.803558697249729e-05, "loss": 0.0451, "step": 47620 }, { "grad_norm": 0.4678167998790741, "learning_rate": 5.8019266763390276e-05, "loss": 0.0622, "step": 47630 }, { "grad_norm": 0.3257408142089844, "learning_rate": 5.800294567730833e-05, "loss": 0.045, "step": 47640 }, { "grad_norm": 0.532943606376648, "learning_rate": 5.798662371603628e-05, "loss": 0.0506, "step": 47650 }, { "grad_norm": 0.5835888981819153, "learning_rate": 5.797030088135904e-05, "loss": 0.0488, "step": 47660 }, { "grad_norm": 0.7322531342506409, "learning_rate": 5.7953977175061705e-05, "loss": 0.0464, "step": 47670 }, { "grad_norm": 0.634286642074585, "learning_rate": 5.793765259892938e-05, "loss": 0.0606, "step": 47680 }, { "grad_norm": 0.3058702051639557, "learning_rate": 5.792132715474729e-05, "loss": 0.0395, "step": 47690 }, { "grad_norm": 0.6284377574920654, "learning_rate": 5.790500084430078e-05, "loss": 0.0378, "step": 47700 }, { "grad_norm": 0.3181052803993225, "learning_rate": 5.7888673669375274e-05, "loss": 0.0411, "step": 47710 }, { "grad_norm": 0.32949963212013245, "learning_rate": 5.787234563175625e-05, "loss": 0.043, "step": 47720 }, { "grad_norm": 0.6158754229545593, "learning_rate": 5.7856016733229355e-05, "loss": 0.0526, "step": 47730 }, { "grad_norm": 0.6054035425186157, "learning_rate": 5.7839686975580297e-05, "loss": 0.0626, "step": 47740 }, { "grad_norm": 0.4292980134487152, "learning_rate": 5.782335636059484e-05, "loss": 0.0451, "step": 47750 }, { "grad_norm": 0.4607234299182892, "learning_rate": 5.780702489005889e-05, "loss": 0.0462, "step": 47760 }, { "grad_norm": 0.44120481610298157, "learning_rate": 5.779069256575845e-05, "loss": 0.0496, "step": 47770 }, { "grad_norm": 0.5118167996406555, "learning_rate": 5.7774359389479574e-05, "loss": 0.0395, "step": 47780 }, { "grad_norm": 1.0167169570922852, "learning_rate": 5.775802536300845e-05, "loss": 0.0483, "step": 47790 }, { "grad_norm": 0.3952573537826538, "learning_rate": 5.774169048813134e-05, "loss": 0.0439, "step": 47800 }, { "grad_norm": 0.5644139051437378, "learning_rate": 5.77253547666346e-05, "loss": 0.0455, "step": 47810 }, { "grad_norm": 0.5404062867164612, "learning_rate": 5.770901820030465e-05, "loss": 0.0539, "step": 47820 }, { "grad_norm": 0.4960964024066925, "learning_rate": 5.769268079092809e-05, "loss": 0.0536, "step": 47830 }, { "grad_norm": 0.4037955701351166, "learning_rate": 5.767634254029151e-05, "loss": 0.0357, "step": 47840 }, { "grad_norm": 0.6361731886863708, "learning_rate": 5.7660003450181655e-05, "loss": 0.0759, "step": 47850 }, { "grad_norm": 0.3872086703777313, "learning_rate": 5.764366352238534e-05, "loss": 0.043, "step": 47860 }, { "grad_norm": 0.5290022492408752, "learning_rate": 5.7627322758689474e-05, "loss": 0.0387, "step": 47870 }, { "grad_norm": 0.41642358899116516, "learning_rate": 5.761098116088105e-05, "loss": 0.0464, "step": 47880 }, { "grad_norm": 0.4974634051322937, "learning_rate": 5.759463873074717e-05, "loss": 0.0423, "step": 47890 }, { "grad_norm": 0.544306755065918, "learning_rate": 5.757829547007504e-05, "loss": 0.0436, "step": 47900 }, { "grad_norm": 0.5305526256561279, "learning_rate": 5.756195138065189e-05, "loss": 0.0457, "step": 47910 }, { "grad_norm": 0.6487962603569031, "learning_rate": 5.754560646426511e-05, "loss": 0.049, "step": 47920 }, { "grad_norm": 0.38167819380760193, "learning_rate": 5.752926072270216e-05, "loss": 0.0397, "step": 47930 }, { "grad_norm": 0.5439146757125854, "learning_rate": 5.7512914157750563e-05, "loss": 0.0398, "step": 47940 }, { "grad_norm": 0.5854129195213318, "learning_rate": 5.749656677119798e-05, "loss": 0.0351, "step": 47950 }, { "grad_norm": 0.7839343547821045, "learning_rate": 5.7480218564832125e-05, "loss": 0.0554, "step": 47960 }, { "grad_norm": 0.5993735194206238, "learning_rate": 5.746386954044082e-05, "loss": 0.0397, "step": 47970 }, { "grad_norm": 0.9784262776374817, "learning_rate": 5.744751969981195e-05, "loss": 0.0446, "step": 47980 }, { "grad_norm": 0.4817622900009155, "learning_rate": 5.7431169044733526e-05, "loss": 0.0526, "step": 47990 }, { "grad_norm": 0.5680944323539734, "learning_rate": 5.741481757699364e-05, "loss": 0.0414, "step": 48000 }, { "grad_norm": 0.3447832465171814, "learning_rate": 5.7398465298380434e-05, "loss": 0.0506, "step": 48010 }, { "grad_norm": 0.5553770065307617, "learning_rate": 5.7382112210682193e-05, "loss": 0.0621, "step": 48020 }, { "grad_norm": 0.5259632468223572, "learning_rate": 5.7365758315687266e-05, "loss": 0.048, "step": 48030 }, { "grad_norm": 0.5533468723297119, "learning_rate": 5.734940361518407e-05, "loss": 0.0519, "step": 48040 }, { "grad_norm": 0.9197925329208374, "learning_rate": 5.733304811096116e-05, "loss": 0.0492, "step": 48050 }, { "grad_norm": 0.44074755907058716, "learning_rate": 5.731669180480713e-05, "loss": 0.0585, "step": 48060 }, { "grad_norm": 0.5391954779624939, "learning_rate": 5.730033469851067e-05, "loss": 0.0609, "step": 48070 }, { "grad_norm": 0.7974439263343811, "learning_rate": 5.7283976793860607e-05, "loss": 0.0587, "step": 48080 }, { "grad_norm": 0.307809978723526, "learning_rate": 5.726761809264577e-05, "loss": 0.0486, "step": 48090 }, { "grad_norm": 0.6435874104499817, "learning_rate": 5.7251258596655155e-05, "loss": 0.0454, "step": 48100 }, { "grad_norm": 0.6496708393096924, "learning_rate": 5.72348983076778e-05, "loss": 0.0557, "step": 48110 }, { "grad_norm": 0.5374502539634705, "learning_rate": 5.7218537227502854e-05, "loss": 0.0498, "step": 48120 }, { "grad_norm": 0.3641189932823181, "learning_rate": 5.720217535791951e-05, "loss": 0.0359, "step": 48130 }, { "grad_norm": 1.0916682481765747, "learning_rate": 5.718581270071711e-05, "loss": 0.052, "step": 48140 }, { "grad_norm": 0.5482434630393982, "learning_rate": 5.716944925768505e-05, "loss": 0.0449, "step": 48150 }, { "grad_norm": 0.33587273955345154, "learning_rate": 5.7153085030612786e-05, "loss": 0.0456, "step": 48160 }, { "grad_norm": 0.2973915934562683, "learning_rate": 5.713672002128989e-05, "loss": 0.0583, "step": 48170 }, { "grad_norm": 0.4602092504501343, "learning_rate": 5.712035423150602e-05, "loss": 0.0505, "step": 48180 }, { "grad_norm": 0.5827547311782837, "learning_rate": 5.710398766305094e-05, "loss": 0.0456, "step": 48190 }, { "grad_norm": 0.5673496127128601, "learning_rate": 5.708762031771442e-05, "loss": 0.0412, "step": 48200 }, { "grad_norm": 0.3653818368911743, "learning_rate": 5.707125219728642e-05, "loss": 0.0386, "step": 48210 }, { "grad_norm": 0.543744683265686, "learning_rate": 5.7054883303556905e-05, "loss": 0.0395, "step": 48220 }, { "grad_norm": 0.6775224804878235, "learning_rate": 5.703851363831595e-05, "loss": 0.056, "step": 48230 }, { "grad_norm": 0.49512630701065063, "learning_rate": 5.702214320335374e-05, "loss": 0.0602, "step": 48240 }, { "grad_norm": 0.6971631050109863, "learning_rate": 5.7005772000460514e-05, "loss": 0.052, "step": 48250 }, { "grad_norm": 0.545026957988739, "learning_rate": 5.698940003142656e-05, "loss": 0.0502, "step": 48260 }, { "grad_norm": 0.5086696743965149, "learning_rate": 5.697302729804236e-05, "loss": 0.0545, "step": 48270 }, { "grad_norm": 0.7473211288452148, "learning_rate": 5.695665380209837e-05, "loss": 0.0551, "step": 48280 }, { "grad_norm": 0.5103853940963745, "learning_rate": 5.6940279545385165e-05, "loss": 0.0384, "step": 48290 }, { "grad_norm": 0.8170354962348938, "learning_rate": 5.692390452969344e-05, "loss": 0.0463, "step": 48300 }, { "grad_norm": 0.7093912959098816, "learning_rate": 5.690752875681392e-05, "loss": 0.0436, "step": 48310 }, { "grad_norm": 0.3347548246383667, "learning_rate": 5.6891152228537435e-05, "loss": 0.0459, "step": 48320 }, { "grad_norm": 0.5945118069648743, "learning_rate": 5.687477494665492e-05, "loss": 0.0418, "step": 48330 }, { "grad_norm": 0.34900790452957153, "learning_rate": 5.685839691295733e-05, "loss": 0.0472, "step": 48340 }, { "grad_norm": 0.49403318762779236, "learning_rate": 5.6842018129235786e-05, "loss": 0.0446, "step": 48350 }, { "grad_norm": 0.6709890961647034, "learning_rate": 5.6825638597281404e-05, "loss": 0.037, "step": 48360 }, { "grad_norm": 0.4493904411792755, "learning_rate": 5.680925831888546e-05, "loss": 0.0601, "step": 48370 }, { "grad_norm": 0.40803322196006775, "learning_rate": 5.6792877295839274e-05, "loss": 0.0477, "step": 48380 }, { "grad_norm": 0.609985888004303, "learning_rate": 5.6776495529934224e-05, "loss": 0.0482, "step": 48390 }, { "grad_norm": 0.4836980104446411, "learning_rate": 5.6760113022961824e-05, "loss": 0.0499, "step": 48400 }, { "grad_norm": 0.5570165514945984, "learning_rate": 5.6743729776713617e-05, "loss": 0.057, "step": 48410 }, { "grad_norm": 0.5122767090797424, "learning_rate": 5.672734579298126e-05, "loss": 0.0485, "step": 48420 }, { "grad_norm": 0.39014217257499695, "learning_rate": 5.671096107355649e-05, "loss": 0.0508, "step": 48430 }, { "grad_norm": 0.39820531010627747, "learning_rate": 5.66945756202311e-05, "loss": 0.0411, "step": 48440 }, { "grad_norm": 0.45175281167030334, "learning_rate": 5.667818943479699e-05, "loss": 0.0601, "step": 48450 }, { "grad_norm": 0.5220291018486023, "learning_rate": 5.666180251904612e-05, "loss": 0.0499, "step": 48460 }, { "grad_norm": 0.45527681708335876, "learning_rate": 5.6645414874770555e-05, "loss": 0.0465, "step": 48470 }, { "grad_norm": 1.1902587413787842, "learning_rate": 5.66290265037624e-05, "loss": 0.0573, "step": 48480 }, { "grad_norm": 0.40575653314590454, "learning_rate": 5.661263740781386e-05, "loss": 0.0392, "step": 48490 }, { "grad_norm": 0.5662298798561096, "learning_rate": 5.6596247588717254e-05, "loss": 0.0526, "step": 48500 }, { "grad_norm": 0.75288325548172, "learning_rate": 5.6579857048264926e-05, "loss": 0.0433, "step": 48510 }, { "grad_norm": 0.5349169969558716, "learning_rate": 5.6563465788249314e-05, "loss": 0.0437, "step": 48520 }, { "grad_norm": 0.5455920100212097, "learning_rate": 5.6547073810462956e-05, "loss": 0.0452, "step": 48530 }, { "grad_norm": 0.4368096590042114, "learning_rate": 5.653068111669846e-05, "loss": 0.0473, "step": 48540 }, { "grad_norm": 0.5151993632316589, "learning_rate": 5.651428770874848e-05, "loss": 0.0534, "step": 48550 }, { "grad_norm": 0.38596251606941223, "learning_rate": 5.64978935884058e-05, "loss": 0.0439, "step": 48560 }, { "grad_norm": 0.5033835172653198, "learning_rate": 5.6481498757463244e-05, "loss": 0.0529, "step": 48570 }, { "grad_norm": 0.564872682094574, "learning_rate": 5.646510321771373e-05, "loss": 0.0521, "step": 48580 }, { "grad_norm": 0.5809472799301147, "learning_rate": 5.644870697095024e-05, "loss": 0.0483, "step": 48590 }, { "grad_norm": 0.7463593482971191, "learning_rate": 5.643231001896586e-05, "loss": 0.0418, "step": 48600 }, { "grad_norm": 0.46124517917633057, "learning_rate": 5.6415912363553726e-05, "loss": 0.0716, "step": 48610 }, { "grad_norm": 0.5290727019309998, "learning_rate": 5.639951400650706e-05, "loss": 0.0514, "step": 48620 }, { "grad_norm": 0.49264708161354065, "learning_rate": 5.6383114949619165e-05, "loss": 0.0433, "step": 48630 }, { "grad_norm": 0.5154022574424744, "learning_rate": 5.636671519468342e-05, "loss": 0.0542, "step": 48640 }, { "grad_norm": 0.6815224885940552, "learning_rate": 5.635031474349327e-05, "loss": 0.0487, "step": 48650 }, { "grad_norm": 0.5033145546913147, "learning_rate": 5.6333913597842246e-05, "loss": 0.0406, "step": 48660 }, { "grad_norm": 0.6104846000671387, "learning_rate": 5.6317511759523955e-05, "loss": 0.0524, "step": 48670 }, { "grad_norm": 0.4097028374671936, "learning_rate": 5.630110923033207e-05, "loss": 0.0466, "step": 48680 }, { "grad_norm": 0.5829403400421143, "learning_rate": 5.628470601206036e-05, "loss": 0.049, "step": 48690 }, { "grad_norm": 0.4930555820465088, "learning_rate": 5.626830210650266e-05, "loss": 0.0463, "step": 48700 }, { "grad_norm": 0.4049118459224701, "learning_rate": 5.625189751545285e-05, "loss": 0.0536, "step": 48710 }, { "grad_norm": 0.404489666223526, "learning_rate": 5.6235492240704936e-05, "loss": 0.0347, "step": 48720 }, { "grad_norm": 0.35947203636169434, "learning_rate": 5.621908628405296e-05, "loss": 0.0477, "step": 48730 }, { "grad_norm": 0.7728577256202698, "learning_rate": 5.620267964729106e-05, "loss": 0.0471, "step": 48740 }, { "grad_norm": 0.3683691620826721, "learning_rate": 5.618627233221344e-05, "loss": 0.0382, "step": 48750 }, { "grad_norm": 0.710180938243866, "learning_rate": 5.616986434061438e-05, "loss": 0.0523, "step": 48760 }, { "grad_norm": 0.9313110709190369, "learning_rate": 5.615345567428822e-05, "loss": 0.0473, "step": 48770 }, { "grad_norm": 0.39071962237358093, "learning_rate": 5.613704633502941e-05, "loss": 0.0356, "step": 48780 }, { "grad_norm": 0.664331316947937, "learning_rate": 5.612063632463245e-05, "loss": 0.0454, "step": 48790 }, { "grad_norm": 0.4432460069656372, "learning_rate": 5.610422564489188e-05, "loss": 0.0397, "step": 48800 }, { "grad_norm": 0.434876024723053, "learning_rate": 5.608781429760239e-05, "loss": 0.0418, "step": 48810 }, { "grad_norm": 0.44809895753860474, "learning_rate": 5.607140228455866e-05, "loss": 0.0385, "step": 48820 }, { "grad_norm": 0.5796887278556824, "learning_rate": 5.605498960755553e-05, "loss": 0.0568, "step": 48830 }, { "grad_norm": 0.51520836353302, "learning_rate": 5.603857626838782e-05, "loss": 0.0472, "step": 48840 }, { "grad_norm": 0.5288790464401245, "learning_rate": 5.60221622688505e-05, "loss": 0.043, "step": 48850 }, { "grad_norm": 0.5378863215446472, "learning_rate": 5.6005747610738565e-05, "loss": 0.043, "step": 48860 }, { "grad_norm": 0.4014320373535156, "learning_rate": 5.59893322958471e-05, "loss": 0.0404, "step": 48870 }, { "grad_norm": 0.6608275175094604, "learning_rate": 5.5972916325971256e-05, "loss": 0.0574, "step": 48880 }, { "grad_norm": 0.431910902261734, "learning_rate": 5.595649970290628e-05, "loss": 0.0527, "step": 48890 }, { "grad_norm": 0.4186764061450958, "learning_rate": 5.5940082428447426e-05, "loss": 0.0465, "step": 48900 }, { "grad_norm": 0.5634740591049194, "learning_rate": 5.592366450439012e-05, "loss": 0.0569, "step": 48910 }, { "grad_norm": 0.7804522514343262, "learning_rate": 5.590724593252975e-05, "loss": 0.0535, "step": 48920 }, { "grad_norm": 0.36041003465652466, "learning_rate": 5.589082671466184e-05, "loss": 0.0411, "step": 48930 }, { "grad_norm": 0.4744430184364319, "learning_rate": 5.587440685258199e-05, "loss": 0.0497, "step": 48940 }, { "grad_norm": 0.6013426184654236, "learning_rate": 5.585798634808583e-05, "loss": 0.0483, "step": 48950 }, { "grad_norm": 0.48417744040489197, "learning_rate": 5.584156520296909e-05, "loss": 0.0449, "step": 48960 }, { "grad_norm": 0.5098505616188049, "learning_rate": 5.582514341902757e-05, "loss": 0.0425, "step": 48970 }, { "grad_norm": 0.4749320447444916, "learning_rate": 5.580872099805713e-05, "loss": 0.0419, "step": 48980 }, { "grad_norm": 0.27379053831100464, "learning_rate": 5.5792297941853674e-05, "loss": 0.0447, "step": 48990 }, { "grad_norm": 0.48771560192108154, "learning_rate": 5.5775874252213247e-05, "loss": 0.0462, "step": 49000 }, { "grad_norm": 0.4723840653896332, "learning_rate": 5.575944993093189e-05, "loss": 0.0463, "step": 49010 }, { "grad_norm": 0.5083451867103577, "learning_rate": 5.574302497980574e-05, "loss": 0.0404, "step": 49020 }, { "grad_norm": 0.6814678311347961, "learning_rate": 5.5726599400631005e-05, "loss": 0.0446, "step": 49030 }, { "grad_norm": 0.4167734980583191, "learning_rate": 5.571017319520399e-05, "loss": 0.0506, "step": 49040 }, { "grad_norm": 0.40634968876838684, "learning_rate": 5.5693746365320984e-05, "loss": 0.0434, "step": 49050 }, { "grad_norm": 0.6868711709976196, "learning_rate": 5.567731891277848e-05, "loss": 0.0461, "step": 49060 }, { "grad_norm": 0.37388816475868225, "learning_rate": 5.566089083937288e-05, "loss": 0.0495, "step": 49070 }, { "grad_norm": 0.36426424980163574, "learning_rate": 5.564446214690079e-05, "loss": 0.0474, "step": 49080 }, { "grad_norm": 0.5468394756317139, "learning_rate": 5.56280328371588e-05, "loss": 0.0561, "step": 49090 }, { "grad_norm": 0.4349576532840729, "learning_rate": 5.5611602911943596e-05, "loss": 0.0427, "step": 49100 }, { "grad_norm": 0.6074641942977905, "learning_rate": 5.559517237305194e-05, "loss": 0.053, "step": 49110 }, { "grad_norm": 0.4966033399105072, "learning_rate": 5.557874122228064e-05, "loss": 0.0498, "step": 49120 }, { "grad_norm": 0.6918806433677673, "learning_rate": 5.55623094614266e-05, "loss": 0.0484, "step": 49130 }, { "grad_norm": 0.4954906404018402, "learning_rate": 5.5545877092286743e-05, "loss": 0.0503, "step": 49140 }, { "grad_norm": 0.6508951783180237, "learning_rate": 5.552944411665811e-05, "loss": 0.0381, "step": 49150 }, { "grad_norm": 0.508442759513855, "learning_rate": 5.5513010536337795e-05, "loss": 0.0604, "step": 49160 }, { "grad_norm": 0.47431716322898865, "learning_rate": 5.5496576353122934e-05, "loss": 0.0462, "step": 49170 }, { "grad_norm": 0.5808915495872498, "learning_rate": 5.548014156881074e-05, "loss": 0.0538, "step": 49180 }, { "grad_norm": 0.5187352895736694, "learning_rate": 5.5463706185198494e-05, "loss": 0.0528, "step": 49190 }, { "grad_norm": 0.5780501961708069, "learning_rate": 5.544727020408358e-05, "loss": 0.05, "step": 49200 }, { "grad_norm": 1.0971091985702515, "learning_rate": 5.5430833627263366e-05, "loss": 0.0423, "step": 49210 }, { "grad_norm": 0.5270180106163025, "learning_rate": 5.5414396456535354e-05, "loss": 0.0426, "step": 49220 }, { "grad_norm": 0.49455034732818604, "learning_rate": 5.539795869369709e-05, "loss": 0.0388, "step": 49230 }, { "grad_norm": 0.6609129905700684, "learning_rate": 5.538152034054618e-05, "loss": 0.0438, "step": 49240 }, { "grad_norm": 0.45095935463905334, "learning_rate": 5.536508139888028e-05, "loss": 0.0456, "step": 49250 }, { "grad_norm": 0.5122106075286865, "learning_rate": 5.534864187049716e-05, "loss": 0.0413, "step": 49260 }, { "grad_norm": 0.46261513233184814, "learning_rate": 5.5332201757194615e-05, "loss": 0.0407, "step": 49270 }, { "grad_norm": 0.3942613899707794, "learning_rate": 5.531576106077049e-05, "loss": 0.0395, "step": 49280 }, { "grad_norm": 0.467136025428772, "learning_rate": 5.529931978302272e-05, "loss": 0.0378, "step": 49290 }, { "grad_norm": 0.5149760842323303, "learning_rate": 5.528287792574932e-05, "loss": 0.0405, "step": 49300 }, { "grad_norm": 0.45894619822502136, "learning_rate": 5.5266435490748324e-05, "loss": 0.0395, "step": 49310 }, { "grad_norm": 0.5164324045181274, "learning_rate": 5.524999247981787e-05, "loss": 0.0467, "step": 49320 }, { "grad_norm": 0.8141037225723267, "learning_rate": 5.523354889475613e-05, "loss": 0.0446, "step": 49330 }, { "grad_norm": 0.5192236304283142, "learning_rate": 5.521710473736134e-05, "loss": 0.0532, "step": 49340 }, { "grad_norm": 0.35721662640571594, "learning_rate": 5.5200660009431826e-05, "loss": 0.0433, "step": 49350 }, { "grad_norm": 0.5246466994285583, "learning_rate": 5.518421471276596e-05, "loss": 0.0406, "step": 49360 }, { "grad_norm": 0.38515275716781616, "learning_rate": 5.5167768849162174e-05, "loss": 0.0376, "step": 49370 }, { "grad_norm": 0.3756694197654724, "learning_rate": 5.515132242041893e-05, "loss": 0.0436, "step": 49380 }, { "grad_norm": 0.3951997756958008, "learning_rate": 5.513487542833483e-05, "loss": 0.0453, "step": 49390 }, { "grad_norm": 0.6074905395507812, "learning_rate": 5.5118427874708466e-05, "loss": 0.0435, "step": 49400 }, { "grad_norm": 0.4198646545410156, "learning_rate": 5.510197976133853e-05, "loss": 0.0355, "step": 49410 }, { "grad_norm": 0.6017138361930847, "learning_rate": 5.508553109002376e-05, "loss": 0.0433, "step": 49420 }, { "grad_norm": 0.502245306968689, "learning_rate": 5.5069081862562957e-05, "loss": 0.0423, "step": 49430 }, { "grad_norm": 0.4203532934188843, "learning_rate": 5.5052632080754965e-05, "loss": 0.0439, "step": 49440 }, { "grad_norm": 0.5698729753494263, "learning_rate": 5.503618174639874e-05, "loss": 0.0489, "step": 49450 }, { "grad_norm": 0.5942561626434326, "learning_rate": 5.501973086129325e-05, "loss": 0.0408, "step": 49460 }, { "grad_norm": 0.3384595811367035, "learning_rate": 5.500327942723753e-05, "loss": 0.0366, "step": 49470 }, { "grad_norm": 0.44666042923927307, "learning_rate": 5.498682744603071e-05, "loss": 0.0481, "step": 49480 }, { "grad_norm": 0.8608254790306091, "learning_rate": 5.497037491947194e-05, "loss": 0.0509, "step": 49490 }, { "grad_norm": 0.424793541431427, "learning_rate": 5.4953921849360424e-05, "loss": 0.0533, "step": 49500 }, { "grad_norm": 0.36128485202789307, "learning_rate": 5.493746823749547e-05, "loss": 0.0475, "step": 49510 }, { "grad_norm": 0.5162503123283386, "learning_rate": 5.49210140856764e-05, "loss": 0.0483, "step": 49520 }, { "grad_norm": 0.35802218317985535, "learning_rate": 5.4904559395702647e-05, "loss": 0.0449, "step": 49530 }, { "grad_norm": 0.3535422086715698, "learning_rate": 5.488810416937364e-05, "loss": 0.0482, "step": 49540 }, { "grad_norm": 0.35396841168403625, "learning_rate": 5.4871648408488926e-05, "loss": 0.0488, "step": 49550 }, { "grad_norm": 0.6328856945037842, "learning_rate": 5.485519211484807e-05, "loss": 0.0455, "step": 49560 }, { "grad_norm": 0.5846822261810303, "learning_rate": 5.4838735290250676e-05, "loss": 0.0478, "step": 49570 }, { "grad_norm": 0.5888941884040833, "learning_rate": 5.482227793649649e-05, "loss": 0.0547, "step": 49580 }, { "grad_norm": 0.4770532250404358, "learning_rate": 5.480582005538524e-05, "loss": 0.0512, "step": 49590 }, { "grad_norm": 0.38131314516067505, "learning_rate": 5.478936164871671e-05, "loss": 0.0491, "step": 49600 }, { "grad_norm": 0.4626067280769348, "learning_rate": 5.477290271829081e-05, "loss": 0.0503, "step": 49610 }, { "grad_norm": 0.37018653750419617, "learning_rate": 5.475644326590744e-05, "loss": 0.0399, "step": 49620 }, { "grad_norm": 0.40439411997795105, "learning_rate": 5.473998329336658e-05, "loss": 0.0472, "step": 49630 }, { "grad_norm": 0.3441658914089203, "learning_rate": 5.4723522802468286e-05, "loss": 0.0448, "step": 49640 }, { "grad_norm": 0.6873287558555603, "learning_rate": 5.470706179501264e-05, "loss": 0.0428, "step": 49650 }, { "grad_norm": 0.44616350531578064, "learning_rate": 5.4690600272799776e-05, "loss": 0.0507, "step": 49660 }, { "grad_norm": 0.704075276851654, "learning_rate": 5.467413823762993e-05, "loss": 0.0554, "step": 49670 }, { "grad_norm": 0.4402807354927063, "learning_rate": 5.465767569130335e-05, "loss": 0.0465, "step": 49680 }, { "grad_norm": 0.5624637603759766, "learning_rate": 5.464121263562036e-05, "loss": 0.0466, "step": 49690 }, { "grad_norm": 0.3802633583545685, "learning_rate": 5.4624749072381343e-05, "loss": 0.0456, "step": 49700 }, { "grad_norm": 0.23777399957180023, "learning_rate": 5.460828500338672e-05, "loss": 0.0455, "step": 49710 }, { "grad_norm": 0.4495711922645569, "learning_rate": 5.459182043043698e-05, "loss": 0.0507, "step": 49720 }, { "grad_norm": 0.41022753715515137, "learning_rate": 5.457535535533265e-05, "loss": 0.0437, "step": 49730 }, { "grad_norm": 0.531348466873169, "learning_rate": 5.4558889779874334e-05, "loss": 0.0545, "step": 49740 }, { "grad_norm": 0.4638318419456482, "learning_rate": 5.454242370586269e-05, "loss": 0.0517, "step": 49750 }, { "grad_norm": 0.948058009147644, "learning_rate": 5.452595713509843e-05, "loss": 0.037, "step": 49760 }, { "grad_norm": 0.41913142800331116, "learning_rate": 5.4509490069382274e-05, "loss": 0.0466, "step": 49770 }, { "grad_norm": 0.29618605971336365, "learning_rate": 5.449302251051509e-05, "loss": 0.0406, "step": 49780 }, { "grad_norm": 0.36271971464157104, "learning_rate": 5.44765544602977e-05, "loss": 0.0544, "step": 49790 }, { "grad_norm": 0.3142440617084503, "learning_rate": 5.446008592053107e-05, "loss": 0.0484, "step": 49800 }, { "grad_norm": 0.3558356463909149, "learning_rate": 5.4443616893016135e-05, "loss": 0.0403, "step": 49810 }, { "grad_norm": 0.42762133479118347, "learning_rate": 5.4427147379553934e-05, "loss": 0.0535, "step": 49820 }, { "grad_norm": 0.4837651550769806, "learning_rate": 5.441067738194556e-05, "loss": 0.0403, "step": 49830 }, { "grad_norm": 0.46656304597854614, "learning_rate": 5.439420690199214e-05, "loss": 0.0565, "step": 49840 }, { "grad_norm": 0.3544493317604065, "learning_rate": 5.437773594149484e-05, "loss": 0.0391, "step": 49850 }, { "grad_norm": 0.3432244062423706, "learning_rate": 5.436126450225495e-05, "loss": 0.0483, "step": 49860 }, { "grad_norm": 0.40160417556762695, "learning_rate": 5.434479258607371e-05, "loss": 0.0374, "step": 49870 }, { "grad_norm": 0.3576083779335022, "learning_rate": 5.43283201947525e-05, "loss": 0.0494, "step": 49880 }, { "grad_norm": 0.41644829511642456, "learning_rate": 5.431184733009268e-05, "loss": 0.0523, "step": 49890 }, { "grad_norm": 0.3614446520805359, "learning_rate": 5.4295373993895736e-05, "loss": 0.0508, "step": 49900 }, { "grad_norm": 0.3724859952926636, "learning_rate": 5.4278900187963157e-05, "loss": 0.0515, "step": 49910 }, { "grad_norm": 0.6411356925964355, "learning_rate": 5.426242591409646e-05, "loss": 0.0448, "step": 49920 }, { "grad_norm": 0.3889055550098419, "learning_rate": 5.42459511740973e-05, "loss": 0.0496, "step": 49930 }, { "grad_norm": 0.4864315986633301, "learning_rate": 5.422947596976729e-05, "loss": 0.0451, "step": 49940 }, { "grad_norm": 0.532838761806488, "learning_rate": 5.4213000302908134e-05, "loss": 0.047, "step": 49950 }, { "grad_norm": 0.36312898993492126, "learning_rate": 5.419652417532162e-05, "loss": 0.0415, "step": 49960 }, { "grad_norm": 0.5931374430656433, "learning_rate": 5.4180047588809534e-05, "loss": 0.0405, "step": 49970 }, { "grad_norm": 0.7481992244720459, "learning_rate": 5.4163570545173704e-05, "loss": 0.0438, "step": 49980 }, { "grad_norm": 0.5174508094787598, "learning_rate": 5.414709304621608e-05, "loss": 0.0478, "step": 49990 }, { "grad_norm": 0.5047817230224609, "learning_rate": 5.413061509373858e-05, "loss": 0.0472, "step": 50000 }, { "grad_norm": 0.381878525018692, "learning_rate": 5.411413668954324e-05, "loss": 0.0463, "step": 50010 }, { "grad_norm": 0.4304509162902832, "learning_rate": 5.4097657835432083e-05, "loss": 0.0468, "step": 50020 }, { "grad_norm": 0.4195234477519989, "learning_rate": 5.408117853320723e-05, "loss": 0.0457, "step": 50030 }, { "grad_norm": 0.561201810836792, "learning_rate": 5.406469878467084e-05, "loss": 0.0347, "step": 50040 }, { "grad_norm": 1.0665080547332764, "learning_rate": 5.404821859162509e-05, "loss": 0.0517, "step": 50050 }, { "grad_norm": 0.48400524258613586, "learning_rate": 5.403173795587225e-05, "loss": 0.0407, "step": 50060 }, { "grad_norm": 0.43233999609947205, "learning_rate": 5.4015256879214606e-05, "loss": 0.0493, "step": 50070 }, { "grad_norm": 0.3611893951892853, "learning_rate": 5.39987753634545e-05, "loss": 0.0424, "step": 50080 }, { "grad_norm": 0.3028208017349243, "learning_rate": 5.3982293410394335e-05, "loss": 0.0553, "step": 50090 }, { "grad_norm": 0.42482733726501465, "learning_rate": 5.396581102183655e-05, "loss": 0.0399, "step": 50100 }, { "grad_norm": 0.582209050655365, "learning_rate": 5.3949328199583634e-05, "loss": 0.0471, "step": 50110 }, { "grad_norm": 0.3684747517108917, "learning_rate": 5.393284494543813e-05, "loss": 0.0495, "step": 50120 }, { "grad_norm": 0.34920206665992737, "learning_rate": 5.391636126120262e-05, "loss": 0.0419, "step": 50130 }, { "grad_norm": 0.44185498356819153, "learning_rate": 5.389987714867971e-05, "loss": 0.0423, "step": 50140 }, { "grad_norm": 0.38654986023902893, "learning_rate": 5.3883392609672114e-05, "loss": 0.0475, "step": 50150 }, { "grad_norm": 0.45019522309303284, "learning_rate": 5.3866907645982545e-05, "loss": 0.046, "step": 50160 }, { "grad_norm": 0.6748420596122742, "learning_rate": 5.385042225941375e-05, "loss": 0.0434, "step": 50170 }, { "grad_norm": 0.4088596999645233, "learning_rate": 5.3833936451768585e-05, "loss": 0.0482, "step": 50180 }, { "grad_norm": 0.46616142988204956, "learning_rate": 5.3817450224849884e-05, "loss": 0.0515, "step": 50190 }, { "grad_norm": 0.5777856707572937, "learning_rate": 5.380096358046056e-05, "loss": 0.0542, "step": 50200 }, { "grad_norm": 0.3616344630718231, "learning_rate": 5.378447652040359e-05, "loss": 0.0436, "step": 50210 }, { "grad_norm": 0.3847638964653015, "learning_rate": 5.376798904648195e-05, "loss": 0.0334, "step": 50220 }, { "grad_norm": 0.5633382797241211, "learning_rate": 5.375150116049869e-05, "loss": 0.0367, "step": 50230 }, { "grad_norm": 0.5200110673904419, "learning_rate": 5.373501286425691e-05, "loss": 0.0352, "step": 50240 }, { "grad_norm": 0.391037255525589, "learning_rate": 5.3718524159559726e-05, "loss": 0.0377, "step": 50250 }, { "grad_norm": 0.5456945896148682, "learning_rate": 5.370203504821034e-05, "loss": 0.0514, "step": 50260 }, { "grad_norm": 0.32364436984062195, "learning_rate": 5.368554553201196e-05, "loss": 0.0503, "step": 50270 }, { "grad_norm": 0.4881936311721802, "learning_rate": 5.3669055612767874e-05, "loss": 0.0646, "step": 50280 }, { "grad_norm": 0.77857905626297, "learning_rate": 5.3652565292281376e-05, "loss": 0.046, "step": 50290 }, { "grad_norm": 0.48441189527511597, "learning_rate": 5.3636074572355824e-05, "loss": 0.0528, "step": 50300 }, { "grad_norm": 0.4375592768192291, "learning_rate": 5.361958345479463e-05, "loss": 0.0503, "step": 50310 }, { "grad_norm": 0.3888528048992157, "learning_rate": 5.360309194140123e-05, "loss": 0.0434, "step": 50320 }, { "grad_norm": 0.457904577255249, "learning_rate": 5.3586600033979106e-05, "loss": 0.0444, "step": 50330 }, { "grad_norm": 0.40298402309417725, "learning_rate": 5.357010773433181e-05, "loss": 0.0461, "step": 50340 }, { "grad_norm": 0.4802023768424988, "learning_rate": 5.3553615044262895e-05, "loss": 0.0364, "step": 50350 }, { "grad_norm": 0.3933130204677582, "learning_rate": 5.3537121965575964e-05, "loss": 0.0396, "step": 50360 }, { "grad_norm": 0.47545042634010315, "learning_rate": 5.352062850007471e-05, "loss": 0.0605, "step": 50370 }, { "grad_norm": 0.2485617846250534, "learning_rate": 5.350413464956282e-05, "loss": 0.0397, "step": 50380 }, { "grad_norm": 0.649924099445343, "learning_rate": 5.348764041584403e-05, "loss": 0.052, "step": 50390 }, { "grad_norm": 0.5715463757514954, "learning_rate": 5.347114580072212e-05, "loss": 0.0449, "step": 50400 }, { "grad_norm": 0.4361891448497772, "learning_rate": 5.345465080600093e-05, "loss": 0.0444, "step": 50410 }, { "grad_norm": 0.558407723903656, "learning_rate": 5.3438155433484326e-05, "loss": 0.049, "step": 50420 }, { "grad_norm": 0.4729340970516205, "learning_rate": 5.3421659684976197e-05, "loss": 0.0441, "step": 50430 }, { "grad_norm": 0.5889395475387573, "learning_rate": 5.340516356228052e-05, "loss": 0.0413, "step": 50440 }, { "grad_norm": 0.4635620415210724, "learning_rate": 5.338866706720128e-05, "loss": 0.046, "step": 50450 }, { "grad_norm": 0.5104649066925049, "learning_rate": 5.337217020154249e-05, "loss": 0.0468, "step": 50460 }, { "grad_norm": 0.5393038392066956, "learning_rate": 5.335567296710825e-05, "loss": 0.0468, "step": 50470 }, { "grad_norm": 0.4547474980354309, "learning_rate": 5.333917536570265e-05, "loss": 0.0424, "step": 50480 }, { "grad_norm": 0.5488107204437256, "learning_rate": 5.332267739912986e-05, "loss": 0.047, "step": 50490 }, { "grad_norm": 0.3409190773963928, "learning_rate": 5.330617906919405e-05, "loss": 0.0532, "step": 50500 }, { "grad_norm": 0.5993558168411255, "learning_rate": 5.328968037769949e-05, "loss": 0.0444, "step": 50510 }, { "grad_norm": 0.43039602041244507, "learning_rate": 5.327318132645042e-05, "loss": 0.0463, "step": 50520 }, { "grad_norm": 0.5476148724555969, "learning_rate": 5.325668191725116e-05, "loss": 0.0402, "step": 50530 }, { "grad_norm": 0.5483427047729492, "learning_rate": 5.324018215190606e-05, "loss": 0.0493, "step": 50540 }, { "grad_norm": 0.534451961517334, "learning_rate": 5.3223682032219515e-05, "loss": 0.0431, "step": 50550 }, { "grad_norm": 0.2816149890422821, "learning_rate": 5.320718155999595e-05, "loss": 0.0474, "step": 50560 }, { "grad_norm": 0.457492858171463, "learning_rate": 5.3190680737039835e-05, "loss": 0.0589, "step": 50570 }, { "grad_norm": 0.4282010793685913, "learning_rate": 5.317417956515567e-05, "loss": 0.0407, "step": 50580 }, { "grad_norm": 0.4189492166042328, "learning_rate": 5.315767804614798e-05, "loss": 0.0449, "step": 50590 }, { "grad_norm": 0.3457123041152954, "learning_rate": 5.3141176181821395e-05, "loss": 0.0437, "step": 50600 }, { "grad_norm": 0.31130972504615784, "learning_rate": 5.31246739739805e-05, "loss": 0.0417, "step": 50610 }, { "grad_norm": 0.45957431197166443, "learning_rate": 5.310817142442994e-05, "loss": 0.0596, "step": 50620 }, { "grad_norm": 0.6239164471626282, "learning_rate": 5.309166853497445e-05, "loss": 0.0482, "step": 50630 }, { "grad_norm": 0.5401063561439514, "learning_rate": 5.307516530741873e-05, "loss": 0.0503, "step": 50640 }, { "grad_norm": 0.7773275971412659, "learning_rate": 5.305866174356754e-05, "loss": 0.0512, "step": 50650 }, { "grad_norm": 0.5444313287734985, "learning_rate": 5.304215784522571e-05, "loss": 0.0403, "step": 50660 }, { "grad_norm": 0.5379500389099121, "learning_rate": 5.302565361419808e-05, "loss": 0.0584, "step": 50670 }, { "grad_norm": 0.3400074243545532, "learning_rate": 5.3009149052289507e-05, "loss": 0.0438, "step": 50680 }, { "grad_norm": 0.39036980271339417, "learning_rate": 5.299264416130493e-05, "loss": 0.0463, "step": 50690 }, { "grad_norm": 0.47227880358695984, "learning_rate": 5.297613894304928e-05, "loss": 0.0476, "step": 50700 }, { "grad_norm": 0.57323157787323, "learning_rate": 5.2959633399327534e-05, "loss": 0.0373, "step": 50710 }, { "grad_norm": 0.43214553594589233, "learning_rate": 5.294312753194476e-05, "loss": 0.0409, "step": 50720 }, { "grad_norm": 0.4240562915802002, "learning_rate": 5.292662134270596e-05, "loss": 0.0451, "step": 50730 }, { "grad_norm": 0.492887407541275, "learning_rate": 5.291011483341626e-05, "loss": 0.0418, "step": 50740 }, { "grad_norm": 0.7696982622146606, "learning_rate": 5.2893608005880767e-05, "loss": 0.0524, "step": 50750 }, { "grad_norm": 0.3399224281311035, "learning_rate": 5.287710086190467e-05, "loss": 0.0429, "step": 50760 }, { "grad_norm": 0.33408206701278687, "learning_rate": 5.286059340329314e-05, "loss": 0.0439, "step": 50770 }, { "grad_norm": 0.8034371733665466, "learning_rate": 5.284408563185141e-05, "loss": 0.0519, "step": 50780 }, { "grad_norm": 0.5654700994491577, "learning_rate": 5.282757754938476e-05, "loss": 0.0488, "step": 50790 }, { "grad_norm": 0.6821269989013672, "learning_rate": 5.281106915769849e-05, "loss": 0.0574, "step": 50800 }, { "grad_norm": 0.7488026022911072, "learning_rate": 5.2794560458597897e-05, "loss": 0.0538, "step": 50810 }, { "grad_norm": 0.3921962082386017, "learning_rate": 5.27780514538884e-05, "loss": 0.0376, "step": 50820 }, { "grad_norm": 0.48117223381996155, "learning_rate": 5.2761542145375365e-05, "loss": 0.0494, "step": 50830 }, { "grad_norm": 0.8988431096076965, "learning_rate": 5.274503253486421e-05, "loss": 0.0378, "step": 50840 }, { "grad_norm": 0.49772384762763977, "learning_rate": 5.272852262416046e-05, "loss": 0.0452, "step": 50850 }, { "grad_norm": 0.5343852639198303, "learning_rate": 5.2712012415069555e-05, "loss": 0.0434, "step": 50860 }, { "grad_norm": 0.5729460716247559, "learning_rate": 5.269550190939705e-05, "loss": 0.0499, "step": 50870 }, { "grad_norm": 0.5539380311965942, "learning_rate": 5.267899110894852e-05, "loss": 0.0391, "step": 50880 }, { "grad_norm": 0.34875330328941345, "learning_rate": 5.266248001552955e-05, "loss": 0.043, "step": 50890 }, { "grad_norm": 0.42330852150917053, "learning_rate": 5.264596863094575e-05, "loss": 0.0446, "step": 50900 }, { "grad_norm": 0.7669605612754822, "learning_rate": 5.262945695700282e-05, "loss": 0.0461, "step": 50910 }, { "grad_norm": 0.7332969307899475, "learning_rate": 5.261294499550643e-05, "loss": 0.0428, "step": 50920 }, { "grad_norm": 0.46569499373435974, "learning_rate": 5.25964327482623e-05, "loss": 0.0405, "step": 50930 }, { "grad_norm": 0.45387521386146545, "learning_rate": 5.257992021707617e-05, "loss": 0.0438, "step": 50940 }, { "grad_norm": 0.5439349412918091, "learning_rate": 5.256340740375387e-05, "loss": 0.0455, "step": 50950 }, { "grad_norm": 0.35475948452949524, "learning_rate": 5.254689431010117e-05, "loss": 0.0505, "step": 50960 }, { "grad_norm": 0.36009639501571655, "learning_rate": 5.253038093792395e-05, "loss": 0.0402, "step": 50970 }, { "grad_norm": 0.43698012828826904, "learning_rate": 5.251386728902806e-05, "loss": 0.0437, "step": 50980 }, { "grad_norm": 0.3956384062767029, "learning_rate": 5.2497353365219446e-05, "loss": 0.0461, "step": 50990 }, { "grad_norm": 0.37546566128730774, "learning_rate": 5.2480839168304e-05, "loss": 0.0383, "step": 51000 }, { "grad_norm": 0.8205738663673401, "learning_rate": 5.246432470008773e-05, "loss": 0.0404, "step": 51010 }, { "grad_norm": 0.80839604139328, "learning_rate": 5.2447809962376616e-05, "loss": 0.0533, "step": 51020 }, { "grad_norm": 0.46180275082588196, "learning_rate": 5.243129495697666e-05, "loss": 0.0387, "step": 51030 }, { "grad_norm": 0.9126697778701782, "learning_rate": 5.241477968569397e-05, "loss": 0.0589, "step": 51040 }, { "grad_norm": 0.5022209286689758, "learning_rate": 5.23982641503346e-05, "loss": 0.0514, "step": 51050 }, { "grad_norm": 0.38965773582458496, "learning_rate": 5.238174835270465e-05, "loss": 0.0468, "step": 51060 }, { "grad_norm": 0.43324390053749084, "learning_rate": 5.23652322946103e-05, "loss": 0.036, "step": 51070 }, { "grad_norm": 0.4304284453392029, "learning_rate": 5.234871597785771e-05, "loss": 0.0382, "step": 51080 }, { "grad_norm": 0.8586118817329407, "learning_rate": 5.233219940425306e-05, "loss": 0.0523, "step": 51090 }, { "grad_norm": 1.0571181774139404, "learning_rate": 5.2315682575602576e-05, "loss": 0.0699, "step": 51100 }, { "grad_norm": 0.5076524019241333, "learning_rate": 5.229916549371254e-05, "loss": 0.0397, "step": 51110 }, { "grad_norm": 0.4138611853122711, "learning_rate": 5.228264816038924e-05, "loss": 0.0457, "step": 51120 }, { "grad_norm": 0.9147406816482544, "learning_rate": 5.2266130577438935e-05, "loss": 0.0431, "step": 51130 }, { "grad_norm": 0.3953486979007721, "learning_rate": 5.224961274666802e-05, "loss": 0.0474, "step": 51140 }, { "grad_norm": 0.7372069954872131, "learning_rate": 5.223309466988283e-05, "loss": 0.0491, "step": 51150 }, { "grad_norm": 0.3646392822265625, "learning_rate": 5.221657634888976e-05, "loss": 0.0571, "step": 51160 }, { "grad_norm": 0.4439701437950134, "learning_rate": 5.2200057785495235e-05, "loss": 0.0654, "step": 51170 }, { "grad_norm": 0.6793801188468933, "learning_rate": 5.2183538981505706e-05, "loss": 0.0487, "step": 51180 }, { "grad_norm": 0.32530903816223145, "learning_rate": 5.216701993872762e-05, "loss": 0.0388, "step": 51190 }, { "grad_norm": 0.525151789188385, "learning_rate": 5.215050065896752e-05, "loss": 0.045, "step": 51200 }, { "grad_norm": 0.43801042437553406, "learning_rate": 5.2133981144031875e-05, "loss": 0.0633, "step": 51210 }, { "grad_norm": 0.782669723033905, "learning_rate": 5.211746139572727e-05, "loss": 0.054, "step": 51220 }, { "grad_norm": 0.3136347532272339, "learning_rate": 5.210094141586026e-05, "loss": 0.0393, "step": 51230 }, { "grad_norm": 0.601132333278656, "learning_rate": 5.2084421206237476e-05, "loss": 0.0539, "step": 51240 }, { "grad_norm": 0.5063822269439697, "learning_rate": 5.20679007686655e-05, "loss": 0.0529, "step": 51250 }, { "grad_norm": 0.40110641717910767, "learning_rate": 5.205138010495102e-05, "loss": 0.0525, "step": 51260 }, { "grad_norm": 0.4166516959667206, "learning_rate": 5.20348592169007e-05, "loss": 0.043, "step": 51270 }, { "grad_norm": 0.7549467086791992, "learning_rate": 5.201833810632123e-05, "loss": 0.0423, "step": 51280 }, { "grad_norm": 0.47434160113334656, "learning_rate": 5.200181677501933e-05, "loss": 0.0482, "step": 51290 }, { "grad_norm": 0.33357158303260803, "learning_rate": 5.1985295224801777e-05, "loss": 0.0498, "step": 51300 }, { "grad_norm": 0.5386227369308472, "learning_rate": 5.196877345747531e-05, "loss": 0.0507, "step": 51310 }, { "grad_norm": 0.4932800531387329, "learning_rate": 5.195225147484674e-05, "loss": 0.0495, "step": 51320 }, { "grad_norm": 0.5844507217407227, "learning_rate": 5.193572927872288e-05, "loss": 0.0538, "step": 51330 }, { "grad_norm": 0.5008212327957153, "learning_rate": 5.191920687091059e-05, "loss": 0.045, "step": 51340 }, { "grad_norm": 0.3767785429954529, "learning_rate": 5.1902684253216695e-05, "loss": 0.0417, "step": 51350 }, { "grad_norm": 0.326098769903183, "learning_rate": 5.1886161427448134e-05, "loss": 0.04, "step": 51360 }, { "grad_norm": 0.3613041937351227, "learning_rate": 5.186963839541179e-05, "loss": 0.0369, "step": 51370 }, { "grad_norm": 0.46384626626968384, "learning_rate": 5.185311515891459e-05, "loss": 0.0343, "step": 51380 }, { "grad_norm": 0.7659515738487244, "learning_rate": 5.1836591719763516e-05, "loss": 0.0408, "step": 51390 }, { "grad_norm": 0.48503243923187256, "learning_rate": 5.182006807976554e-05, "loss": 0.0444, "step": 51400 }, { "grad_norm": 0.33451545238494873, "learning_rate": 5.180354424072763e-05, "loss": 0.0373, "step": 51410 }, { "grad_norm": 0.6892234683036804, "learning_rate": 5.178702020445686e-05, "loss": 0.0501, "step": 51420 }, { "grad_norm": 0.38987845182418823, "learning_rate": 5.177049597276024e-05, "loss": 0.0456, "step": 51430 }, { "grad_norm": 0.47172287106513977, "learning_rate": 5.175397154744482e-05, "loss": 0.0445, "step": 51440 }, { "grad_norm": 0.49555638432502747, "learning_rate": 5.1737446930317736e-05, "loss": 0.0353, "step": 51450 }, { "grad_norm": 0.5098652243614197, "learning_rate": 5.172092212318604e-05, "loss": 0.0495, "step": 51460 }, { "grad_norm": 0.4616866707801819, "learning_rate": 5.170439712785692e-05, "loss": 0.0486, "step": 51470 }, { "grad_norm": 0.5038223266601562, "learning_rate": 5.1687871946137467e-05, "loss": 0.0432, "step": 51480 }, { "grad_norm": 0.3881586790084839, "learning_rate": 5.16713465798349e-05, "loss": 0.0413, "step": 51490 }, { "grad_norm": 0.4505019187927246, "learning_rate": 5.165482103075637e-05, "loss": 0.0591, "step": 51500 }, { "grad_norm": 0.3464682102203369, "learning_rate": 5.163829530070909e-05, "loss": 0.0406, "step": 51510 }, { "grad_norm": 0.5327479243278503, "learning_rate": 5.162176939150032e-05, "loss": 0.0449, "step": 51520 }, { "grad_norm": 0.5639435052871704, "learning_rate": 5.160524330493728e-05, "loss": 0.0405, "step": 51530 }, { "grad_norm": 0.3578140437602997, "learning_rate": 5.158871704282724e-05, "loss": 0.0527, "step": 51540 }, { "grad_norm": 0.6872346997261047, "learning_rate": 5.15721906069775e-05, "loss": 0.0513, "step": 51550 }, { "grad_norm": 0.5178585648536682, "learning_rate": 5.155566399919536e-05, "loss": 0.0422, "step": 51560 }, { "grad_norm": 0.5946829915046692, "learning_rate": 5.153913722128813e-05, "loss": 0.046, "step": 51570 }, { "grad_norm": 0.5147249698638916, "learning_rate": 5.1522610275063176e-05, "loss": 0.0483, "step": 51580 }, { "grad_norm": 0.36573150753974915, "learning_rate": 5.150608316232786e-05, "loss": 0.0383, "step": 51590 }, { "grad_norm": 0.5168923735618591, "learning_rate": 5.148955588488955e-05, "loss": 0.0333, "step": 51600 }, { "grad_norm": 0.3597460091114044, "learning_rate": 5.147302844455564e-05, "loss": 0.0327, "step": 51610 }, { "grad_norm": 0.4336993396282196, "learning_rate": 5.145650084313356e-05, "loss": 0.0491, "step": 51620 }, { "grad_norm": 0.5092922449111938, "learning_rate": 5.1439973082430734e-05, "loss": 0.0434, "step": 51630 }, { "grad_norm": 0.37756553292274475, "learning_rate": 5.142344516425461e-05, "loss": 0.0365, "step": 51640 }, { "grad_norm": 0.31485089659690857, "learning_rate": 5.1406917090412673e-05, "loss": 0.0472, "step": 51650 }, { "grad_norm": 0.4852907359600067, "learning_rate": 5.1390388862712404e-05, "loss": 0.0392, "step": 51660 }, { "grad_norm": 0.4344033896923065, "learning_rate": 5.137386048296129e-05, "loss": 0.0399, "step": 51670 }, { "grad_norm": 0.4680997431278229, "learning_rate": 5.135733195296686e-05, "loss": 0.0535, "step": 51680 }, { "grad_norm": 0.2964097559452057, "learning_rate": 5.134080327453664e-05, "loss": 0.042, "step": 51690 }, { "grad_norm": 0.30323171615600586, "learning_rate": 5.1324274449478204e-05, "loss": 0.041, "step": 51700 }, { "grad_norm": 0.3174060583114624, "learning_rate": 5.130774547959909e-05, "loss": 0.0515, "step": 51710 }, { "grad_norm": 0.36869141459465027, "learning_rate": 5.129121636670692e-05, "loss": 0.0475, "step": 51720 }, { "grad_norm": 0.2516610324382782, "learning_rate": 5.1274687112609246e-05, "loss": 0.0288, "step": 51730 }, { "grad_norm": 0.5528287887573242, "learning_rate": 5.125815771911372e-05, "loss": 0.0376, "step": 51740 }, { "grad_norm": 0.6154489517211914, "learning_rate": 5.1241628188027956e-05, "loss": 0.036, "step": 51750 }, { "grad_norm": 0.6126040816307068, "learning_rate": 5.1225098521159585e-05, "loss": 0.0385, "step": 51760 }, { "grad_norm": 0.5299158096313477, "learning_rate": 5.12085687203163e-05, "loss": 0.0363, "step": 51770 }, { "grad_norm": 0.6386337280273438, "learning_rate": 5.119203878730575e-05, "loss": 0.0473, "step": 51780 }, { "grad_norm": 0.5461165904998779, "learning_rate": 5.117550872393564e-05, "loss": 0.0506, "step": 51790 }, { "grad_norm": 0.399615615606308, "learning_rate": 5.1158978532013645e-05, "loss": 0.0528, "step": 51800 }, { "grad_norm": 0.3717665672302246, "learning_rate": 5.114244821334751e-05, "loss": 0.0447, "step": 51810 }, { "grad_norm": 0.3285837471485138, "learning_rate": 5.112591776974497e-05, "loss": 0.0368, "step": 51820 }, { "grad_norm": 0.4730176627635956, "learning_rate": 5.110938720301374e-05, "loss": 0.0376, "step": 51830 }, { "grad_norm": 0.23327045142650604, "learning_rate": 5.1092856514961595e-05, "loss": 0.0462, "step": 51840 }, { "grad_norm": 0.3612969219684601, "learning_rate": 5.107632570739631e-05, "loss": 0.0404, "step": 51850 }, { "grad_norm": 0.5059411525726318, "learning_rate": 5.1059794782125645e-05, "loss": 0.0525, "step": 51860 }, { "grad_norm": 0.6922906637191772, "learning_rate": 5.104326374095744e-05, "loss": 0.0513, "step": 51870 }, { "grad_norm": 0.37159281969070435, "learning_rate": 5.102673258569947e-05, "loss": 0.0424, "step": 51880 }, { "grad_norm": 0.4433794319629669, "learning_rate": 5.101020131815956e-05, "loss": 0.0482, "step": 51890 }, { "grad_norm": 0.6076405644416809, "learning_rate": 5.099366994014556e-05, "loss": 0.0433, "step": 51900 }, { "grad_norm": 0.5729394555091858, "learning_rate": 5.097713845346532e-05, "loss": 0.0485, "step": 51910 }, { "grad_norm": 0.9387565851211548, "learning_rate": 5.096060685992668e-05, "loss": 0.0335, "step": 51920 }, { "grad_norm": 0.7637309432029724, "learning_rate": 5.0944075161337514e-05, "loss": 0.0602, "step": 51930 }, { "grad_norm": 0.6231015920639038, "learning_rate": 5.09275433595057e-05, "loss": 0.0422, "step": 51940 }, { "grad_norm": 0.7586597800254822, "learning_rate": 5.0911011456239157e-05, "loss": 0.0459, "step": 51950 }, { "grad_norm": 0.5204277634620667, "learning_rate": 5.089447945334574e-05, "loss": 0.0479, "step": 51960 }, { "grad_norm": 0.479583740234375, "learning_rate": 5.087794735263343e-05, "loss": 0.037, "step": 51970 }, { "grad_norm": 0.33398309350013733, "learning_rate": 5.086141515591011e-05, "loss": 0.0445, "step": 51980 }, { "grad_norm": 0.5190751552581787, "learning_rate": 5.084488286498371e-05, "loss": 0.0418, "step": 51990 }, { "grad_norm": 0.42459461092948914, "learning_rate": 5.08283504816622e-05, "loss": 0.0597, "step": 52000 }, { "grad_norm": 0.4149712026119232, "learning_rate": 5.081181800775353e-05, "loss": 0.0493, "step": 52010 }, { "grad_norm": 0.37888941168785095, "learning_rate": 5.079528544506565e-05, "loss": 0.0478, "step": 52020 }, { "grad_norm": 0.3494526147842407, "learning_rate": 5.077875279540657e-05, "loss": 0.0374, "step": 52030 }, { "grad_norm": 0.2782885730266571, "learning_rate": 5.076222006058424e-05, "loss": 0.0486, "step": 52040 }, { "grad_norm": 0.43191877007484436, "learning_rate": 5.074568724240667e-05, "loss": 0.0352, "step": 52050 }, { "grad_norm": 0.36602914333343506, "learning_rate": 5.072915434268187e-05, "loss": 0.0431, "step": 52060 }, { "grad_norm": 0.31378936767578125, "learning_rate": 5.071262136321785e-05, "loss": 0.0421, "step": 52070 }, { "grad_norm": 0.42467570304870605, "learning_rate": 5.069608830582262e-05, "loss": 0.0488, "step": 52080 }, { "grad_norm": 0.17976385354995728, "learning_rate": 5.067955517230423e-05, "loss": 0.0474, "step": 52090 }, { "grad_norm": 0.48500263690948486, "learning_rate": 5.06630219644707e-05, "loss": 0.05, "step": 52100 }, { "grad_norm": 0.4239047169685364, "learning_rate": 5.064648868413008e-05, "loss": 0.0324, "step": 52110 }, { "grad_norm": 0.44961270689964294, "learning_rate": 5.062995533309044e-05, "loss": 0.0401, "step": 52120 }, { "grad_norm": 0.42134490609169006, "learning_rate": 5.061342191315983e-05, "loss": 0.0532, "step": 52130 }, { "grad_norm": 0.5015864372253418, "learning_rate": 5.0596888426146325e-05, "loss": 0.0466, "step": 52140 }, { "grad_norm": 0.40723344683647156, "learning_rate": 5.058035487385798e-05, "loss": 0.0369, "step": 52150 }, { "grad_norm": 0.6706097722053528, "learning_rate": 5.0563821258102915e-05, "loss": 0.0442, "step": 52160 }, { "grad_norm": 0.6225513219833374, "learning_rate": 5.05472875806892e-05, "loss": 0.052, "step": 52170 }, { "grad_norm": 0.4787183105945587, "learning_rate": 5.053075384342494e-05, "loss": 0.0464, "step": 52180 }, { "grad_norm": 0.41088810563087463, "learning_rate": 5.0514220048118234e-05, "loss": 0.0425, "step": 52190 }, { "grad_norm": 0.6481543779373169, "learning_rate": 5.049768619657721e-05, "loss": 0.0545, "step": 52200 }, { "grad_norm": 0.3808028995990753, "learning_rate": 5.048115229060996e-05, "loss": 0.0412, "step": 52210 }, { "grad_norm": 0.8000791668891907, "learning_rate": 5.046461833202462e-05, "loss": 0.046, "step": 52220 }, { "grad_norm": 1.0272926092147827, "learning_rate": 5.044808432262932e-05, "loss": 0.0529, "step": 52230 }, { "grad_norm": 0.3482406735420227, "learning_rate": 5.043155026423219e-05, "loss": 0.0474, "step": 52240 }, { "grad_norm": 0.4619297981262207, "learning_rate": 5.041501615864137e-05, "loss": 0.0392, "step": 52250 }, { "grad_norm": 0.30954745411872864, "learning_rate": 5.039848200766502e-05, "loss": 0.0329, "step": 52260 }, { "grad_norm": 0.44456109404563904, "learning_rate": 5.038194781311125e-05, "loss": 0.0486, "step": 52270 }, { "grad_norm": 0.3693494200706482, "learning_rate": 5.0365413576788264e-05, "loss": 0.0342, "step": 52280 }, { "grad_norm": 0.4526633620262146, "learning_rate": 5.0348879300504194e-05, "loss": 0.0408, "step": 52290 }, { "grad_norm": 0.42789947986602783, "learning_rate": 5.033234498606721e-05, "loss": 0.0403, "step": 52300 }, { "grad_norm": 0.6135988831520081, "learning_rate": 5.0315810635285476e-05, "loss": 0.0515, "step": 52310 }, { "grad_norm": 0.5116118788719177, "learning_rate": 5.0299276249967164e-05, "loss": 0.0447, "step": 52320 }, { "grad_norm": 0.5105244517326355, "learning_rate": 5.0282741831920454e-05, "loss": 0.0577, "step": 52330 }, { "grad_norm": 0.7533217668533325, "learning_rate": 5.026620738295351e-05, "loss": 0.0429, "step": 52340 }, { "grad_norm": 0.33907264471054077, "learning_rate": 5.024967290487455e-05, "loss": 0.0379, "step": 52350 }, { "grad_norm": 0.5219067335128784, "learning_rate": 5.023313839949172e-05, "loss": 0.0492, "step": 52360 }, { "grad_norm": 0.35636651515960693, "learning_rate": 5.0216603868613224e-05, "loss": 0.0454, "step": 52370 }, { "grad_norm": 0.5249904990196228, "learning_rate": 5.020006931404726e-05, "loss": 0.0548, "step": 52380 }, { "grad_norm": 0.48793044686317444, "learning_rate": 5.0183534737602014e-05, "loss": 0.0327, "step": 52390 }, { "grad_norm": 0.47378969192504883, "learning_rate": 5.0167000141085686e-05, "loss": 0.0395, "step": 52400 }, { "grad_norm": 0.3725678026676178, "learning_rate": 5.015046552630647e-05, "loss": 0.0423, "step": 52410 }, { "grad_norm": 0.5709235072135925, "learning_rate": 5.013393089507258e-05, "loss": 0.0432, "step": 52420 }, { "grad_norm": 0.677765429019928, "learning_rate": 5.011739624919219e-05, "loss": 0.0396, "step": 52430 }, { "grad_norm": 0.5103296041488647, "learning_rate": 5.0100861590473545e-05, "loss": 0.0485, "step": 52440 }, { "grad_norm": 0.7443496584892273, "learning_rate": 5.008432692072482e-05, "loss": 0.0351, "step": 52450 }, { "grad_norm": 0.8832533955574036, "learning_rate": 5.006779224175423e-05, "loss": 0.0614, "step": 52460 }, { "grad_norm": 0.5372017025947571, "learning_rate": 5.005125755536998e-05, "loss": 0.0493, "step": 52470 }, { "grad_norm": 0.854198694229126, "learning_rate": 5.003472286338029e-05, "loss": 0.051, "step": 52480 }, { "grad_norm": 0.3163140118122101, "learning_rate": 5.001818816759335e-05, "loss": 0.063, "step": 52490 }, { "grad_norm": 0.5064617395401001, "learning_rate": 5.000165346981738e-05, "loss": 0.0462, "step": 52500 }, { "grad_norm": 0.5228025913238525, "learning_rate": 4.9985118771860595e-05, "loss": 0.0432, "step": 52510 }, { "grad_norm": 0.7719451189041138, "learning_rate": 4.996858407553119e-05, "loss": 0.0392, "step": 52520 }, { "grad_norm": 0.31254640221595764, "learning_rate": 4.99520493826374e-05, "loss": 0.0439, "step": 52530 }, { "grad_norm": 0.5287826061248779, "learning_rate": 4.9935514694987384e-05, "loss": 0.0518, "step": 52540 }, { "grad_norm": 0.5569683313369751, "learning_rate": 4.991898001438941e-05, "loss": 0.0421, "step": 52550 }, { "grad_norm": 0.5725201964378357, "learning_rate": 4.990244534265166e-05, "loss": 0.0444, "step": 52560 }, { "grad_norm": 0.57805335521698, "learning_rate": 4.988591068158234e-05, "loss": 0.0516, "step": 52570 }, { "grad_norm": 0.5045009255409241, "learning_rate": 4.986937603298965e-05, "loss": 0.051, "step": 52580 }, { "grad_norm": 0.5723904967308044, "learning_rate": 4.985284139868181e-05, "loss": 0.048, "step": 52590 }, { "grad_norm": 0.38960549235343933, "learning_rate": 4.9836306780467e-05, "loss": 0.0417, "step": 52600 }, { "grad_norm": 0.6516277194023132, "learning_rate": 4.981977218015343e-05, "loss": 0.0444, "step": 52610 }, { "grad_norm": 0.6223212480545044, "learning_rate": 4.980323759954932e-05, "loss": 0.0447, "step": 52620 }, { "grad_norm": 0.5034113526344299, "learning_rate": 4.978670304046284e-05, "loss": 0.038, "step": 52630 }, { "grad_norm": 0.381483256816864, "learning_rate": 4.9770168504702195e-05, "loss": 0.0431, "step": 52640 }, { "grad_norm": 0.5904344916343689, "learning_rate": 4.975363399407559e-05, "loss": 0.0487, "step": 52650 }, { "grad_norm": 0.7174959778785706, "learning_rate": 4.9737099510391194e-05, "loss": 0.0498, "step": 52660 }, { "grad_norm": 0.38261479139328003, "learning_rate": 4.972056505545719e-05, "loss": 0.0471, "step": 52670 }, { "grad_norm": 0.4060916602611542, "learning_rate": 4.9704030631081775e-05, "loss": 0.0439, "step": 52680 }, { "grad_norm": 0.3931119441986084, "learning_rate": 4.968749623907313e-05, "loss": 0.0345, "step": 52690 }, { "grad_norm": 0.37976565957069397, "learning_rate": 4.967096188123943e-05, "loss": 0.0498, "step": 52700 }, { "grad_norm": 0.5602021217346191, "learning_rate": 4.965442755938884e-05, "loss": 0.0449, "step": 52710 }, { "grad_norm": 0.452862024307251, "learning_rate": 4.963789327532954e-05, "loss": 0.0453, "step": 52720 }, { "grad_norm": 0.4274156093597412, "learning_rate": 4.962135903086968e-05, "loss": 0.0504, "step": 52730 }, { "grad_norm": 0.40208756923675537, "learning_rate": 4.960482482781741e-05, "loss": 0.0411, "step": 52740 }, { "grad_norm": 0.6050032377243042, "learning_rate": 4.958829066798092e-05, "loss": 0.0447, "step": 52750 }, { "grad_norm": 0.3799433410167694, "learning_rate": 4.957175655316835e-05, "loss": 0.0426, "step": 52760 }, { "grad_norm": 0.36539947986602783, "learning_rate": 4.9555222485187834e-05, "loss": 0.0456, "step": 52770 }, { "grad_norm": 0.2998180389404297, "learning_rate": 4.953868846584752e-05, "loss": 0.0367, "step": 52780 }, { "grad_norm": 0.4832945466041565, "learning_rate": 4.9522154496955544e-05, "loss": 0.0464, "step": 52790 }, { "grad_norm": 0.3240458369255066, "learning_rate": 4.950562058032001e-05, "loss": 0.0561, "step": 52800 }, { "grad_norm": 0.5054057836532593, "learning_rate": 4.948908671774909e-05, "loss": 0.0465, "step": 52810 }, { "grad_norm": 0.41116565465927124, "learning_rate": 4.947255291105087e-05, "loss": 0.0467, "step": 52820 }, { "grad_norm": 0.5252784490585327, "learning_rate": 4.945601916203348e-05, "loss": 0.0468, "step": 52830 }, { "grad_norm": 0.4344319999217987, "learning_rate": 4.9439485472505016e-05, "loss": 0.0409, "step": 52840 }, { "grad_norm": 0.8672839999198914, "learning_rate": 4.942295184427356e-05, "loss": 0.0461, "step": 52850 }, { "grad_norm": 0.49171772599220276, "learning_rate": 4.9406418279147234e-05, "loss": 0.0415, "step": 52860 }, { "grad_norm": 0.567480742931366, "learning_rate": 4.9389884778934116e-05, "loss": 0.0423, "step": 52870 }, { "grad_norm": 0.6103688478469849, "learning_rate": 4.937335134544228e-05, "loss": 0.0401, "step": 52880 }, { "grad_norm": 0.30370157957077026, "learning_rate": 4.935681798047979e-05, "loss": 0.0639, "step": 52890 }, { "grad_norm": 0.3999970555305481, "learning_rate": 4.934028468585473e-05, "loss": 0.0435, "step": 52900 }, { "grad_norm": 0.4162345230579376, "learning_rate": 4.932375146337514e-05, "loss": 0.0489, "step": 52910 }, { "grad_norm": 0.4124011993408203, "learning_rate": 4.930721831484908e-05, "loss": 0.0455, "step": 52920 }, { "grad_norm": 0.8315998315811157, "learning_rate": 4.929068524208456e-05, "loss": 0.0544, "step": 52930 }, { "grad_norm": 0.5047398209571838, "learning_rate": 4.927415224688965e-05, "loss": 0.0534, "step": 52940 }, { "grad_norm": 0.28556060791015625, "learning_rate": 4.9257619331072375e-05, "loss": 0.0497, "step": 52950 }, { "grad_norm": 0.8332564830780029, "learning_rate": 4.924108649644072e-05, "loss": 0.0502, "step": 52960 }, { "grad_norm": 0.48906660079956055, "learning_rate": 4.922455374480271e-05, "loss": 0.0636, "step": 52970 }, { "grad_norm": 0.5046120285987854, "learning_rate": 4.920802107796634e-05, "loss": 0.0448, "step": 52980 }, { "grad_norm": 0.7609884142875671, "learning_rate": 4.9191488497739564e-05, "loss": 0.0553, "step": 52990 }, { "grad_norm": 0.9125698804855347, "learning_rate": 4.917495600593042e-05, "loss": 0.0489, "step": 53000 }, { "grad_norm": 0.4139840304851532, "learning_rate": 4.915842360434684e-05, "loss": 0.0447, "step": 53010 }, { "grad_norm": 1.0943423509597778, "learning_rate": 4.914189129479679e-05, "loss": 0.0584, "step": 53020 }, { "grad_norm": 0.3273319602012634, "learning_rate": 4.912535907908822e-05, "loss": 0.0448, "step": 53030 }, { "grad_norm": 0.8128108978271484, "learning_rate": 4.910882695902907e-05, "loss": 0.0482, "step": 53040 }, { "grad_norm": 0.55907142162323, "learning_rate": 4.9092294936427234e-05, "loss": 0.0416, "step": 53050 }, { "grad_norm": 0.44975894689559937, "learning_rate": 4.9075763013090685e-05, "loss": 0.0437, "step": 53060 }, { "grad_norm": 0.5416720509529114, "learning_rate": 4.9059231190827305e-05, "loss": 0.039, "step": 53070 }, { "grad_norm": 0.3637123703956604, "learning_rate": 4.904269947144499e-05, "loss": 0.0528, "step": 53080 }, { "grad_norm": 0.6352793574333191, "learning_rate": 4.902616785675161e-05, "loss": 0.049, "step": 53090 }, { "grad_norm": 0.4821160137653351, "learning_rate": 4.900963634855505e-05, "loss": 0.0496, "step": 53100 }, { "grad_norm": 0.4639512002468109, "learning_rate": 4.8993104948663173e-05, "loss": 0.034, "step": 53110 }, { "grad_norm": 0.4229156970977783, "learning_rate": 4.8976573658883823e-05, "loss": 0.0422, "step": 53120 }, { "grad_norm": 0.4410575330257416, "learning_rate": 4.896004248102484e-05, "loss": 0.0445, "step": 53130 }, { "grad_norm": 0.47413256764411926, "learning_rate": 4.8943511416894045e-05, "loss": 0.0488, "step": 53140 }, { "grad_norm": 1.0456355810165405, "learning_rate": 4.892698046829926e-05, "loss": 0.0359, "step": 53150 }, { "grad_norm": 0.5776230692863464, "learning_rate": 4.8910449637048275e-05, "loss": 0.0514, "step": 53160 }, { "grad_norm": 0.5212626457214355, "learning_rate": 4.8893918924948884e-05, "loss": 0.0469, "step": 53170 }, { "grad_norm": 0.3970619738101959, "learning_rate": 4.8877388333808835e-05, "loss": 0.0535, "step": 53180 }, { "grad_norm": 0.41896477341651917, "learning_rate": 4.8860857865435925e-05, "loss": 0.0468, "step": 53190 }, { "grad_norm": 0.42143216729164124, "learning_rate": 4.884432752163789e-05, "loss": 0.0427, "step": 53200 }, { "grad_norm": 0.3958907425403595, "learning_rate": 4.8827797304222466e-05, "loss": 0.0418, "step": 53210 }, { "grad_norm": 0.3397925794124603, "learning_rate": 4.8811267214997366e-05, "loss": 0.0411, "step": 53220 }, { "grad_norm": 0.47148242592811584, "learning_rate": 4.879473725577029e-05, "loss": 0.0483, "step": 53230 }, { "grad_norm": 0.4756767749786377, "learning_rate": 4.877820742834891e-05, "loss": 0.042, "step": 53240 }, { "grad_norm": 1.6585825681686401, "learning_rate": 4.8761677734540956e-05, "loss": 0.0477, "step": 53250 }, { "grad_norm": 0.7980702519416809, "learning_rate": 4.874514817615407e-05, "loss": 0.0449, "step": 53260 }, { "grad_norm": 0.47217845916748047, "learning_rate": 4.872861875499588e-05, "loss": 0.0503, "step": 53270 }, { "grad_norm": 0.6126418709754944, "learning_rate": 4.871208947287404e-05, "loss": 0.0435, "step": 53280 }, { "grad_norm": 0.596141517162323, "learning_rate": 4.869556033159615e-05, "loss": 0.0558, "step": 53290 }, { "grad_norm": 0.33801573514938354, "learning_rate": 4.8679031332969835e-05, "loss": 0.0386, "step": 53300 }, { "grad_norm": 0.2578372359275818, "learning_rate": 4.8662502478802635e-05, "loss": 0.0649, "step": 53310 }, { "grad_norm": 0.5153038501739502, "learning_rate": 4.8645973770902176e-05, "loss": 0.0421, "step": 53320 }, { "grad_norm": 0.46566057205200195, "learning_rate": 4.862944521107599e-05, "loss": 0.0304, "step": 53330 }, { "grad_norm": 0.6011937856674194, "learning_rate": 4.861291680113162e-05, "loss": 0.0493, "step": 53340 }, { "grad_norm": 0.513589084148407, "learning_rate": 4.8596388542876563e-05, "loss": 0.0384, "step": 53350 }, { "grad_norm": 0.7629355192184448, "learning_rate": 4.857986043811836e-05, "loss": 0.0458, "step": 53360 }, { "grad_norm": 0.47065743803977966, "learning_rate": 4.856333248866448e-05, "loss": 0.0436, "step": 53370 }, { "grad_norm": 0.6170979738235474, "learning_rate": 4.85468046963224e-05, "loss": 0.0395, "step": 53380 }, { "grad_norm": 0.6398842334747314, "learning_rate": 4.853027706289957e-05, "loss": 0.0376, "step": 53390 }, { "grad_norm": 0.3901926577091217, "learning_rate": 4.851374959020344e-05, "loss": 0.0402, "step": 53400 }, { "grad_norm": 0.3582763373851776, "learning_rate": 4.849722228004141e-05, "loss": 0.0376, "step": 53410 }, { "grad_norm": 0.7741436958312988, "learning_rate": 4.848069513422091e-05, "loss": 0.0493, "step": 53420 }, { "grad_norm": 0.4100683927536011, "learning_rate": 4.8464168154549294e-05, "loss": 0.0395, "step": 53430 }, { "grad_norm": 0.5644972920417786, "learning_rate": 4.844764134283392e-05, "loss": 0.0451, "step": 53440 }, { "grad_norm": 0.4685809910297394, "learning_rate": 4.843111470088218e-05, "loss": 0.0519, "step": 53450 }, { "grad_norm": 0.7295382022857666, "learning_rate": 4.841458823050137e-05, "loss": 0.0431, "step": 53460 }, { "grad_norm": 0.5851431488990784, "learning_rate": 4.8398061933498816e-05, "loss": 0.0347, "step": 53470 }, { "grad_norm": 0.7111433744430542, "learning_rate": 4.83815358116818e-05, "loss": 0.0386, "step": 53480 }, { "grad_norm": 0.4550940692424774, "learning_rate": 4.8365009866857584e-05, "loss": 0.04, "step": 53490 }, { "grad_norm": 0.42597147822380066, "learning_rate": 4.834848410083342e-05, "loss": 0.0353, "step": 53500 }, { "grad_norm": 0.6320055723190308, "learning_rate": 4.833195851541657e-05, "loss": 0.0474, "step": 53510 }, { "grad_norm": 0.41880160570144653, "learning_rate": 4.831543311241423e-05, "loss": 0.0463, "step": 53520 }, { "grad_norm": 0.5532740354537964, "learning_rate": 4.829890789363359e-05, "loss": 0.0584, "step": 53530 }, { "grad_norm": 0.39632877707481384, "learning_rate": 4.828238286088182e-05, "loss": 0.0423, "step": 53540 }, { "grad_norm": 0.31124815344810486, "learning_rate": 4.826585801596609e-05, "loss": 0.044, "step": 53550 }, { "grad_norm": 0.37831366062164307, "learning_rate": 4.8249333360693486e-05, "loss": 0.0402, "step": 53560 }, { "grad_norm": 0.33749935030937195, "learning_rate": 4.8232808896871185e-05, "loss": 0.0339, "step": 53570 }, { "grad_norm": 0.27714240550994873, "learning_rate": 4.821628462630624e-05, "loss": 0.0431, "step": 53580 }, { "grad_norm": 0.4836026430130005, "learning_rate": 4.8199760550805735e-05, "loss": 0.0391, "step": 53590 }, { "grad_norm": 0.5243932604789734, "learning_rate": 4.818323667217669e-05, "loss": 0.0443, "step": 53600 }, { "grad_norm": 0.7347685098648071, "learning_rate": 4.8166712992226165e-05, "loss": 0.0452, "step": 53610 }, { "grad_norm": 0.45879465341567993, "learning_rate": 4.8150189512761154e-05, "loss": 0.0525, "step": 53620 }, { "grad_norm": 0.44785353541374207, "learning_rate": 4.813366623558862e-05, "loss": 0.0442, "step": 53630 }, { "grad_norm": 0.5020239949226379, "learning_rate": 4.811714316251554e-05, "loss": 0.039, "step": 53640 }, { "grad_norm": 0.6126184463500977, "learning_rate": 4.810062029534886e-05, "loss": 0.0409, "step": 53650 }, { "grad_norm": 0.4480721056461334, "learning_rate": 4.8084097635895505e-05, "loss": 0.0459, "step": 53660 }, { "grad_norm": 0.5775603652000427, "learning_rate": 4.8067575185962335e-05, "loss": 0.0441, "step": 53670 }, { "grad_norm": 0.4812808334827423, "learning_rate": 4.805105294735625e-05, "loss": 0.0461, "step": 53680 }, { "grad_norm": 0.4738266170024872, "learning_rate": 4.803453092188406e-05, "loss": 0.043, "step": 53690 }, { "grad_norm": 0.6148887872695923, "learning_rate": 4.801800911135263e-05, "loss": 0.0424, "step": 53700 }, { "grad_norm": 0.7341668605804443, "learning_rate": 4.8001487517568744e-05, "loss": 0.0518, "step": 53710 }, { "grad_norm": 0.9229778051376343, "learning_rate": 4.798496614233918e-05, "loss": 0.0499, "step": 53720 }, { "grad_norm": 0.40781277418136597, "learning_rate": 4.796844498747069e-05, "loss": 0.0545, "step": 53730 }, { "grad_norm": 0.468002051115036, "learning_rate": 4.795192405477e-05, "loss": 0.036, "step": 53740 }, { "grad_norm": 0.6467915773391724, "learning_rate": 4.79354033460438e-05, "loss": 0.0531, "step": 53750 }, { "grad_norm": 0.41982126235961914, "learning_rate": 4.79188828630988e-05, "loss": 0.0518, "step": 53760 }, { "grad_norm": 0.45192456245422363, "learning_rate": 4.790236260774165e-05, "loss": 0.048, "step": 53770 }, { "grad_norm": 1.0005589723587036, "learning_rate": 4.788584258177896e-05, "loss": 0.0457, "step": 53780 }, { "grad_norm": 0.4845554828643799, "learning_rate": 4.7869322787017356e-05, "loss": 0.04, "step": 53790 }, { "grad_norm": 0.4961632490158081, "learning_rate": 4.78528032252634e-05, "loss": 0.0504, "step": 53800 }, { "grad_norm": 0.4138760566711426, "learning_rate": 4.783628389832366e-05, "loss": 0.0433, "step": 53810 }, { "grad_norm": 0.5750702619552612, "learning_rate": 4.7819764808004644e-05, "loss": 0.0497, "step": 53820 }, { "grad_norm": 0.34908634424209595, "learning_rate": 4.780324595611289e-05, "loss": 0.0419, "step": 53830 }, { "grad_norm": 0.4027157425880432, "learning_rate": 4.7786727344454854e-05, "loss": 0.0556, "step": 53840 }, { "grad_norm": 0.4978964328765869, "learning_rate": 4.777020897483697e-05, "loss": 0.0468, "step": 53850 }, { "grad_norm": 0.3933253884315491, "learning_rate": 4.775369084906569e-05, "loss": 0.0417, "step": 53860 }, { "grad_norm": 0.40104663372039795, "learning_rate": 4.773717296894742e-05, "loss": 0.0544, "step": 53870 }, { "grad_norm": 0.7793435454368591, "learning_rate": 4.772065533628848e-05, "loss": 0.0457, "step": 53880 }, { "grad_norm": 0.4011037051677704, "learning_rate": 4.7704137952895265e-05, "loss": 0.0431, "step": 53890 }, { "grad_norm": 0.5432129502296448, "learning_rate": 4.768762082057407e-05, "loss": 0.0392, "step": 53900 }, { "grad_norm": 0.3951883316040039, "learning_rate": 4.7671103941131186e-05, "loss": 0.042, "step": 53910 }, { "grad_norm": 0.4220900535583496, "learning_rate": 4.765458731637287e-05, "loss": 0.0418, "step": 53920 }, { "grad_norm": 0.3563253581523895, "learning_rate": 4.763807094810537e-05, "loss": 0.0524, "step": 53930 }, { "grad_norm": 0.4236431419849396, "learning_rate": 4.762155483813487e-05, "loss": 0.055, "step": 53940 }, { "grad_norm": 0.6195166707038879, "learning_rate": 4.760503898826754e-05, "loss": 0.0519, "step": 53950 }, { "grad_norm": 0.4139939248561859, "learning_rate": 4.758852340030957e-05, "loss": 0.0523, "step": 53960 }, { "grad_norm": 0.6358659863471985, "learning_rate": 4.757200807606705e-05, "loss": 0.0441, "step": 53970 }, { "grad_norm": 0.6834442019462585, "learning_rate": 4.755549301734607e-05, "loss": 0.0421, "step": 53980 }, { "grad_norm": 0.3585640490055084, "learning_rate": 4.753897822595271e-05, "loss": 0.0558, "step": 53990 }, { "grad_norm": 0.5229442715644836, "learning_rate": 4.7522463703692994e-05, "loss": 0.0406, "step": 54000 }, { "grad_norm": 0.3902454972267151, "learning_rate": 4.75059494523729e-05, "loss": 0.0422, "step": 54010 }, { "grad_norm": 0.4001680016517639, "learning_rate": 4.748943547379843e-05, "loss": 0.0431, "step": 54020 }, { "grad_norm": 0.3058525621891022, "learning_rate": 4.7472921769775535e-05, "loss": 0.0591, "step": 54030 }, { "grad_norm": 0.4188573360443115, "learning_rate": 4.745640834211011e-05, "loss": 0.0557, "step": 54040 }, { "grad_norm": 0.36257633566856384, "learning_rate": 4.7439895192608054e-05, "loss": 0.0447, "step": 54050 }, { "grad_norm": 0.4650050699710846, "learning_rate": 4.7423382323075186e-05, "loss": 0.042, "step": 54060 }, { "grad_norm": 0.48040854930877686, "learning_rate": 4.740686973531737e-05, "loss": 0.0391, "step": 54070 }, { "grad_norm": 0.6231541037559509, "learning_rate": 4.7390357431140375e-05, "loss": 0.0444, "step": 54080 }, { "grad_norm": 0.4447092115879059, "learning_rate": 4.737384541234998e-05, "loss": 0.0516, "step": 54090 }, { "grad_norm": 0.40897929668426514, "learning_rate": 4.735733368075189e-05, "loss": 0.0326, "step": 54100 }, { "grad_norm": 0.4334278106689453, "learning_rate": 4.734082223815182e-05, "loss": 0.0438, "step": 54110 }, { "grad_norm": 0.3895878493785858, "learning_rate": 4.7324311086355436e-05, "loss": 0.0365, "step": 54120 }, { "grad_norm": 0.33365893363952637, "learning_rate": 4.730780022716836e-05, "loss": 0.0405, "step": 54130 }, { "grad_norm": 0.44043904542922974, "learning_rate": 4.729128966239619e-05, "loss": 0.0342, "step": 54140 }, { "grad_norm": 0.9594855308532715, "learning_rate": 4.7274779393844536e-05, "loss": 0.0355, "step": 54150 }, { "grad_norm": 0.9373764395713806, "learning_rate": 4.725826942331891e-05, "loss": 0.062, "step": 54160 }, { "grad_norm": 0.9005420804023743, "learning_rate": 4.724175975262481e-05, "loss": 0.0476, "step": 54170 }, { "grad_norm": 0.5975368618965149, "learning_rate": 4.722525038356772e-05, "loss": 0.0421, "step": 54180 }, { "grad_norm": 0.5159618258476257, "learning_rate": 4.7208741317953075e-05, "loss": 0.049, "step": 54190 }, { "grad_norm": 0.5080665349960327, "learning_rate": 4.7192232557586264e-05, "loss": 0.053, "step": 54200 }, { "grad_norm": 0.5796300768852234, "learning_rate": 4.717572410427269e-05, "loss": 0.0464, "step": 54210 }, { "grad_norm": 0.6716300845146179, "learning_rate": 4.71592159598177e-05, "loss": 0.053, "step": 54220 }, { "grad_norm": 0.659532904624939, "learning_rate": 4.714270812602657e-05, "loss": 0.0375, "step": 54230 }, { "grad_norm": 0.36540332436561584, "learning_rate": 4.712620060470458e-05, "loss": 0.0678, "step": 54240 }, { "grad_norm": 0.5331284999847412, "learning_rate": 4.7109693397656986e-05, "loss": 0.0503, "step": 54250 }, { "grad_norm": 0.915444552898407, "learning_rate": 4.709318650668894e-05, "loss": 0.0429, "step": 54260 }, { "grad_norm": 0.5412611961364746, "learning_rate": 4.7076679933605676e-05, "loss": 0.0456, "step": 54270 }, { "grad_norm": 0.5202363729476929, "learning_rate": 4.70601736802123e-05, "loss": 0.0367, "step": 54280 }, { "grad_norm": 0.4616893231868744, "learning_rate": 4.704366774831391e-05, "loss": 0.0529, "step": 54290 }, { "grad_norm": 0.5295690894126892, "learning_rate": 4.702716213971557e-05, "loss": 0.0411, "step": 54300 }, { "grad_norm": 0.7271756529808044, "learning_rate": 4.70106568562223e-05, "loss": 0.0485, "step": 54310 }, { "grad_norm": 0.46879640221595764, "learning_rate": 4.6994151899639105e-05, "loss": 0.0408, "step": 54320 }, { "grad_norm": 0.4871346652507782, "learning_rate": 4.697764727177093e-05, "loss": 0.0659, "step": 54330 }, { "grad_norm": 0.5718227028846741, "learning_rate": 4.696114297442272e-05, "loss": 0.0452, "step": 54340 }, { "grad_norm": 0.521608829498291, "learning_rate": 4.694463900939933e-05, "loss": 0.0402, "step": 54350 }, { "grad_norm": 0.41972416639328003, "learning_rate": 4.692813537850564e-05, "loss": 0.0441, "step": 54360 }, { "grad_norm": 0.26113131642341614, "learning_rate": 4.6911632083546445e-05, "loss": 0.0342, "step": 54370 }, { "grad_norm": 0.3314322531223297, "learning_rate": 4.6895129126326525e-05, "loss": 0.047, "step": 54380 }, { "grad_norm": 0.4473766088485718, "learning_rate": 4.687862650865059e-05, "loss": 0.0395, "step": 54390 }, { "grad_norm": 0.43458104133605957, "learning_rate": 4.686212423232339e-05, "loss": 0.0366, "step": 54400 }, { "grad_norm": 0.4962931275367737, "learning_rate": 4.6845622299149565e-05, "loss": 0.0533, "step": 54410 }, { "grad_norm": 0.4088002145290375, "learning_rate": 4.682912071093374e-05, "loss": 0.0387, "step": 54420 }, { "grad_norm": 0.5586636662483215, "learning_rate": 4.681261946948052e-05, "loss": 0.0506, "step": 54430 }, { "grad_norm": 0.37582364678382874, "learning_rate": 4.679611857659443e-05, "loss": 0.0428, "step": 54440 }, { "grad_norm": 0.2621993124485016, "learning_rate": 4.6779618034079975e-05, "loss": 0.0449, "step": 54450 }, { "grad_norm": 0.6217311024665833, "learning_rate": 4.6763117843741675e-05, "loss": 0.0489, "step": 54460 }, { "grad_norm": 0.48774686455726624, "learning_rate": 4.674661800738394e-05, "loss": 0.0404, "step": 54470 }, { "grad_norm": 0.4302182197570801, "learning_rate": 4.673011852681117e-05, "loss": 0.0364, "step": 54480 }, { "grad_norm": 0.3500242233276367, "learning_rate": 4.671361940382772e-05, "loss": 0.0471, "step": 54490 }, { "grad_norm": 0.40139660239219666, "learning_rate": 4.669712064023791e-05, "loss": 0.0437, "step": 54500 }, { "grad_norm": 0.4810378849506378, "learning_rate": 4.668062223784602e-05, "loss": 0.0365, "step": 54510 }, { "grad_norm": 0.5222028493881226, "learning_rate": 4.6664124198456274e-05, "loss": 0.0454, "step": 54520 }, { "grad_norm": 0.4204753041267395, "learning_rate": 4.6647626523872905e-05, "loss": 0.0505, "step": 54530 }, { "grad_norm": 0.325713574886322, "learning_rate": 4.6631129215900066e-05, "loss": 0.0449, "step": 54540 }, { "grad_norm": 0.8103700876235962, "learning_rate": 4.661463227634187e-05, "loss": 0.0687, "step": 54550 }, { "grad_norm": 0.5362070202827454, "learning_rate": 4.659813570700238e-05, "loss": 0.0373, "step": 54560 }, { "grad_norm": 0.5663995742797852, "learning_rate": 4.658163950968568e-05, "loss": 0.0355, "step": 54570 }, { "grad_norm": 0.40995439887046814, "learning_rate": 4.6565143686195726e-05, "loss": 0.0457, "step": 54580 }, { "grad_norm": 0.4815084636211395, "learning_rate": 4.654864823833649e-05, "loss": 0.0386, "step": 54590 }, { "grad_norm": 0.5066435933113098, "learning_rate": 4.65321531679119e-05, "loss": 0.0411, "step": 54600 }, { "grad_norm": 0.4076790511608124, "learning_rate": 4.6515658476725834e-05, "loss": 0.0509, "step": 54610 }, { "grad_norm": 0.614454448223114, "learning_rate": 4.649916416658212e-05, "loss": 0.0484, "step": 54620 }, { "grad_norm": 0.6228865385055542, "learning_rate": 4.648267023928455e-05, "loss": 0.0469, "step": 54630 }, { "grad_norm": 1.7725458145141602, "learning_rate": 4.646617669663687e-05, "loss": 0.041, "step": 54640 }, { "grad_norm": 0.669338047504425, "learning_rate": 4.644968354044277e-05, "loss": 0.0378, "step": 54650 }, { "grad_norm": 0.8445534706115723, "learning_rate": 4.643319077250596e-05, "loss": 0.0495, "step": 54660 }, { "grad_norm": 0.9856851696968079, "learning_rate": 4.6416698394630055e-05, "loss": 0.0377, "step": 54670 }, { "grad_norm": 0.6766400933265686, "learning_rate": 4.640020640861862e-05, "loss": 0.0431, "step": 54680 }, { "grad_norm": 0.35810142755508423, "learning_rate": 4.6383714816275195e-05, "loss": 0.0432, "step": 54690 }, { "grad_norm": 0.33558693528175354, "learning_rate": 4.636722361940329e-05, "loss": 0.035, "step": 54700 }, { "grad_norm": 0.4463180899620056, "learning_rate": 4.635073281980632e-05, "loss": 0.0486, "step": 54710 }, { "grad_norm": 0.2732292413711548, "learning_rate": 4.633424241928775e-05, "loss": 0.0533, "step": 54720 }, { "grad_norm": 0.4070128798484802, "learning_rate": 4.6317752419650914e-05, "loss": 0.0366, "step": 54730 }, { "grad_norm": 0.5911746621131897, "learning_rate": 4.630126282269914e-05, "loss": 0.0502, "step": 54740 }, { "grad_norm": 0.37128737568855286, "learning_rate": 4.628477363023569e-05, "loss": 0.0398, "step": 54750 }, { "grad_norm": 0.4005960524082184, "learning_rate": 4.626828484406383e-05, "loss": 0.0427, "step": 54760 }, { "grad_norm": 0.45321035385131836, "learning_rate": 4.625179646598669e-05, "loss": 0.0542, "step": 54770 }, { "grad_norm": 0.3855955898761749, "learning_rate": 4.6235308497807487e-05, "loss": 0.0416, "step": 54780 }, { "grad_norm": 0.5030990839004517, "learning_rate": 4.6218820941329276e-05, "loss": 0.0545, "step": 54790 }, { "grad_norm": 0.3020719885826111, "learning_rate": 4.620233379835513e-05, "loss": 0.034, "step": 54800 }, { "grad_norm": 0.3894881010055542, "learning_rate": 4.618584707068804e-05, "loss": 0.045, "step": 54810 }, { "grad_norm": 0.43821293115615845, "learning_rate": 4.616936076013099e-05, "loss": 0.0465, "step": 54820 }, { "grad_norm": 0.5475499033927917, "learning_rate": 4.615287486848688e-05, "loss": 0.0442, "step": 54830 }, { "grad_norm": 0.5873856544494629, "learning_rate": 4.6136389397558584e-05, "loss": 0.0521, "step": 54840 }, { "grad_norm": 0.41298040747642517, "learning_rate": 4.6119904349148935e-05, "loss": 0.0349, "step": 54850 }, { "grad_norm": 0.6107140779495239, "learning_rate": 4.6103419725060726e-05, "loss": 0.0372, "step": 54860 }, { "grad_norm": 0.41029834747314453, "learning_rate": 4.608693552709667e-05, "loss": 0.0294, "step": 54870 }, { "grad_norm": 0.5519348382949829, "learning_rate": 4.607045175705947e-05, "loss": 0.0461, "step": 54880 }, { "grad_norm": 0.49418044090270996, "learning_rate": 4.605396841675176e-05, "loss": 0.0425, "step": 54890 }, { "grad_norm": 0.44952213764190674, "learning_rate": 4.6037485507976094e-05, "loss": 0.0462, "step": 54900 }, { "grad_norm": 0.3855116665363312, "learning_rate": 4.602100303253509e-05, "loss": 0.0412, "step": 54910 }, { "grad_norm": 0.5460203886032104, "learning_rate": 4.600452099223122e-05, "loss": 0.0565, "step": 54920 }, { "grad_norm": 0.3007640540599823, "learning_rate": 4.598803938886691e-05, "loss": 0.0407, "step": 54930 }, { "grad_norm": 0.5069179534912109, "learning_rate": 4.597155822424459e-05, "loss": 0.0411, "step": 54940 }, { "grad_norm": 0.3712540864944458, "learning_rate": 4.59550775001666e-05, "loss": 0.0512, "step": 54950 }, { "grad_norm": 0.32668259739875793, "learning_rate": 4.5938597218435234e-05, "loss": 0.0446, "step": 54960 }, { "grad_norm": 0.8727319836616516, "learning_rate": 4.592211738085279e-05, "loss": 0.0643, "step": 54970 }, { "grad_norm": 0.4995763897895813, "learning_rate": 4.590563798922146e-05, "loss": 0.0334, "step": 54980 }, { "grad_norm": 0.37962475419044495, "learning_rate": 4.5889159045343404e-05, "loss": 0.0492, "step": 54990 }, { "grad_norm": 0.4341720938682556, "learning_rate": 4.5872680551020734e-05, "loss": 0.0419, "step": 55000 }, { "grad_norm": 0.6223703026771545, "learning_rate": 4.585620250805552e-05, "loss": 0.0332, "step": 55010 }, { "grad_norm": 0.8668262362480164, "learning_rate": 4.583972491824975e-05, "loss": 0.0401, "step": 55020 }, { "grad_norm": 0.4011925756931305, "learning_rate": 4.582324778340541e-05, "loss": 0.0482, "step": 55030 }, { "grad_norm": 0.4464052617549896, "learning_rate": 4.5806771105324424e-05, "loss": 0.051, "step": 55040 }, { "grad_norm": 0.4335315227508545, "learning_rate": 4.579029488580865e-05, "loss": 0.0463, "step": 55050 }, { "grad_norm": 0.2946293354034424, "learning_rate": 4.5773819126659876e-05, "loss": 0.0314, "step": 55060 }, { "grad_norm": 0.3594527542591095, "learning_rate": 4.575734382967991e-05, "loss": 0.0394, "step": 55070 }, { "grad_norm": 0.46168628334999084, "learning_rate": 4.574086899667044e-05, "loss": 0.0446, "step": 55080 }, { "grad_norm": 0.5819087624549866, "learning_rate": 4.5724394629433106e-05, "loss": 0.0482, "step": 55090 }, { "grad_norm": 0.30615314841270447, "learning_rate": 4.570792072976958e-05, "loss": 0.0415, "step": 55100 }, { "grad_norm": 0.8307952880859375, "learning_rate": 4.569144729948138e-05, "loss": 0.0478, "step": 55110 }, { "grad_norm": 0.6066034436225891, "learning_rate": 4.5674974340370033e-05, "loss": 0.0409, "step": 55120 }, { "grad_norm": 0.41518181562423706, "learning_rate": 4.565850185423699e-05, "loss": 0.0423, "step": 55130 }, { "grad_norm": 0.4359135031700134, "learning_rate": 4.564202984288365e-05, "loss": 0.0441, "step": 55140 }, { "grad_norm": 0.7026261687278748, "learning_rate": 4.5625558308111356e-05, "loss": 0.0387, "step": 55150 }, { "grad_norm": 0.39164382219314575, "learning_rate": 4.560908725172145e-05, "loss": 0.0485, "step": 55160 }, { "grad_norm": 0.5171120166778564, "learning_rate": 4.559261667551516e-05, "loss": 0.0474, "step": 55170 }, { "grad_norm": 0.5838128328323364, "learning_rate": 4.5576146581293685e-05, "loss": 0.0516, "step": 55180 }, { "grad_norm": 0.5648307800292969, "learning_rate": 4.5559676970858164e-05, "loss": 0.0463, "step": 55190 }, { "grad_norm": 0.582490086555481, "learning_rate": 4.5543207846009705e-05, "loss": 0.0477, "step": 55200 }, { "grad_norm": 0.3466402590274811, "learning_rate": 4.552673920854933e-05, "loss": 0.0561, "step": 55210 }, { "grad_norm": 0.7833617925643921, "learning_rate": 4.5510271060278006e-05, "loss": 0.0425, "step": 55220 }, { "grad_norm": 0.48337075114250183, "learning_rate": 4.5493803402996704e-05, "loss": 0.0489, "step": 55230 }, { "grad_norm": 0.31441205739974976, "learning_rate": 4.5477336238506295e-05, "loss": 0.0419, "step": 55240 }, { "grad_norm": 0.47331470251083374, "learning_rate": 4.546086956860759e-05, "loss": 0.043, "step": 55250 }, { "grad_norm": 0.6075094938278198, "learning_rate": 4.5444403395101374e-05, "loss": 0.0384, "step": 55260 }, { "grad_norm": 0.7429170608520508, "learning_rate": 4.542793771978834e-05, "loss": 0.0396, "step": 55270 }, { "grad_norm": 0.7002943754196167, "learning_rate": 4.541147254446916e-05, "loss": 0.0488, "step": 55280 }, { "grad_norm": 0.42012929916381836, "learning_rate": 4.539500787094445e-05, "loss": 0.0461, "step": 55290 }, { "grad_norm": 0.5916547179222107, "learning_rate": 4.537854370101475e-05, "loss": 0.0479, "step": 55300 }, { "grad_norm": 0.32269713282585144, "learning_rate": 4.5362080036480566e-05, "loss": 0.0379, "step": 55310 }, { "grad_norm": 0.48635968565940857, "learning_rate": 4.534561687914235e-05, "loss": 0.0406, "step": 55320 }, { "grad_norm": 0.40047675371170044, "learning_rate": 4.5329154230800466e-05, "loss": 0.0418, "step": 55330 }, { "grad_norm": 0.2287757396697998, "learning_rate": 4.531269209325525e-05, "loss": 0.0391, "step": 55340 }, { "grad_norm": 0.5311124920845032, "learning_rate": 4.529623046830696e-05, "loss": 0.0329, "step": 55350 }, { "grad_norm": 1.023780345916748, "learning_rate": 4.5279769357755866e-05, "loss": 0.0477, "step": 55360 }, { "grad_norm": 0.7614895701408386, "learning_rate": 4.5263308763402084e-05, "loss": 0.038, "step": 55370 }, { "grad_norm": 0.2775678038597107, "learning_rate": 4.524684868704574e-05, "loss": 0.0413, "step": 55380 }, { "grad_norm": 0.4498484134674072, "learning_rate": 4.523038913048688e-05, "loss": 0.0426, "step": 55390 }, { "grad_norm": 2.2082724571228027, "learning_rate": 4.521393009552549e-05, "loss": 0.038, "step": 55400 }, { "grad_norm": 0.36529451608657837, "learning_rate": 4.5197471583961484e-05, "loss": 0.0469, "step": 55410 }, { "grad_norm": 0.2512979209423065, "learning_rate": 4.51810135975948e-05, "loss": 0.0387, "step": 55420 }, { "grad_norm": 0.5686495900154114, "learning_rate": 4.516455613822521e-05, "loss": 0.0403, "step": 55430 }, { "grad_norm": 0.26537665724754333, "learning_rate": 4.514809920765249e-05, "loss": 0.0438, "step": 55440 }, { "grad_norm": 0.5627830624580383, "learning_rate": 4.5131642807676356e-05, "loss": 0.0398, "step": 55450 }, { "grad_norm": 0.8792743682861328, "learning_rate": 4.511518694009644e-05, "loss": 0.0459, "step": 55460 }, { "grad_norm": 0.419757604598999, "learning_rate": 4.509873160671231e-05, "loss": 0.0516, "step": 55470 }, { "grad_norm": 0.43542709946632385, "learning_rate": 4.508227680932356e-05, "loss": 0.0368, "step": 55480 }, { "grad_norm": 0.48551449179649353, "learning_rate": 4.5065822549729605e-05, "loss": 0.0366, "step": 55490 }, { "grad_norm": 0.38343971967697144, "learning_rate": 4.504936882972989e-05, "loss": 0.0389, "step": 55500 }, { "grad_norm": 0.40364211797714233, "learning_rate": 4.503291565112374e-05, "loss": 0.0317, "step": 55510 }, { "grad_norm": 0.4940299391746521, "learning_rate": 4.501646301571048e-05, "loss": 0.0443, "step": 55520 }, { "grad_norm": 0.6281993985176086, "learning_rate": 4.500001092528933e-05, "loss": 0.0477, "step": 55530 }, { "grad_norm": 0.29915714263916016, "learning_rate": 4.4983559381659455e-05, "loss": 0.0376, "step": 55540 }, { "grad_norm": 0.4161156415939331, "learning_rate": 4.496710838661999e-05, "loss": 0.0415, "step": 55550 }, { "grad_norm": 0.4620263874530792, "learning_rate": 4.495065794196999e-05, "loss": 0.0484, "step": 55560 }, { "grad_norm": 0.4036382734775543, "learning_rate": 4.4934208049508455e-05, "loss": 0.0384, "step": 55570 }, { "grad_norm": 0.4426104426383972, "learning_rate": 4.4917758711034303e-05, "loss": 0.041, "step": 55580 }, { "grad_norm": 0.40239056944847107, "learning_rate": 4.490130992834642e-05, "loss": 0.0438, "step": 55590 }, { "grad_norm": 0.641100287437439, "learning_rate": 4.488486170324359e-05, "loss": 0.0615, "step": 55600 }, { "grad_norm": 0.5354111194610596, "learning_rate": 4.486841403752463e-05, "loss": 0.0379, "step": 55610 }, { "grad_norm": 0.3444244861602783, "learning_rate": 4.4851966932988174e-05, "loss": 0.033, "step": 55620 }, { "grad_norm": 0.5782412886619568, "learning_rate": 4.4835520391432894e-05, "loss": 0.0413, "step": 55630 }, { "grad_norm": 0.4662707448005676, "learning_rate": 4.481907441465733e-05, "loss": 0.0476, "step": 55640 }, { "grad_norm": 0.4175097644329071, "learning_rate": 4.480262900446e-05, "loss": 0.0447, "step": 55650 }, { "grad_norm": 0.3201026916503906, "learning_rate": 4.478618416263933e-05, "loss": 0.0465, "step": 55660 }, { "grad_norm": 0.6157488226890564, "learning_rate": 4.476973989099373e-05, "loss": 0.0403, "step": 55670 }, { "grad_norm": 0.5797641277313232, "learning_rate": 4.475329619132153e-05, "loss": 0.0505, "step": 55680 }, { "grad_norm": 0.5343760848045349, "learning_rate": 4.4736853065420956e-05, "loss": 0.0612, "step": 55690 }, { "grad_norm": 0.46442797780036926, "learning_rate": 4.472041051509023e-05, "loss": 0.0408, "step": 55700 }, { "grad_norm": 0.3908461034297943, "learning_rate": 4.470396854212746e-05, "loss": 0.0431, "step": 55710 }, { "grad_norm": 0.5683252215385437, "learning_rate": 4.468752714833072e-05, "loss": 0.0419, "step": 55720 }, { "grad_norm": 0.3499385714530945, "learning_rate": 4.467108633549803e-05, "loss": 0.0344, "step": 55730 }, { "grad_norm": 0.6388146281242371, "learning_rate": 4.4654646105427335e-05, "loss": 0.0539, "step": 55740 }, { "grad_norm": 0.5011349320411682, "learning_rate": 4.463820645991651e-05, "loss": 0.0389, "step": 55750 }, { "grad_norm": 0.37491849064826965, "learning_rate": 4.462176740076334e-05, "loss": 0.0325, "step": 55760 }, { "grad_norm": 0.46381616592407227, "learning_rate": 4.4605328929765624e-05, "loss": 0.0549, "step": 55770 }, { "grad_norm": 0.4448367953300476, "learning_rate": 4.458889104872102e-05, "loss": 0.0403, "step": 55780 }, { "grad_norm": 0.6751968860626221, "learning_rate": 4.457245375942715e-05, "loss": 0.0486, "step": 55790 }, { "grad_norm": 0.4136602580547333, "learning_rate": 4.455601706368157e-05, "loss": 0.0451, "step": 55800 }, { "grad_norm": 0.824173092842102, "learning_rate": 4.453958096328179e-05, "loss": 0.0394, "step": 55810 }, { "grad_norm": 0.6051200032234192, "learning_rate": 4.452314546002523e-05, "loss": 0.0495, "step": 55820 }, { "grad_norm": 0.3499579429626465, "learning_rate": 4.4506710555709246e-05, "loss": 0.0513, "step": 55830 }, { "grad_norm": 0.4721122086048126, "learning_rate": 4.449027625213114e-05, "loss": 0.0389, "step": 55840 }, { "grad_norm": 0.3905397355556488, "learning_rate": 4.4473842551088135e-05, "loss": 0.0381, "step": 55850 }, { "grad_norm": 0.7098143696784973, "learning_rate": 4.445740945437738e-05, "loss": 0.0436, "step": 55860 }, { "grad_norm": 0.41251376271247864, "learning_rate": 4.444097696379602e-05, "loss": 0.0383, "step": 55870 }, { "grad_norm": 0.3562302589416504, "learning_rate": 4.442454508114106e-05, "loss": 0.0385, "step": 55880 }, { "grad_norm": 0.9008588194847107, "learning_rate": 4.4408113808209465e-05, "loss": 0.0519, "step": 55890 }, { "grad_norm": 0.39043283462524414, "learning_rate": 4.439168314679813e-05, "loss": 0.0458, "step": 55900 }, { "grad_norm": 2.2347583770751953, "learning_rate": 4.43752530987039e-05, "loss": 0.0409, "step": 55910 }, { "grad_norm": 0.45314159989356995, "learning_rate": 4.4358823665723515e-05, "loss": 0.0344, "step": 55920 }, { "grad_norm": 0.6607480049133301, "learning_rate": 4.434239484965371e-05, "loss": 0.0434, "step": 55930 }, { "grad_norm": 0.5248647332191467, "learning_rate": 4.4325966652291103e-05, "loss": 0.0457, "step": 55940 }, { "grad_norm": 0.4423140287399292, "learning_rate": 4.430953907543225e-05, "loss": 0.0368, "step": 55950 }, { "grad_norm": 0.5806746482849121, "learning_rate": 4.4293112120873645e-05, "loss": 0.0354, "step": 55960 }, { "grad_norm": 0.7137953639030457, "learning_rate": 4.427668579041172e-05, "loss": 0.0327, "step": 55970 }, { "grad_norm": 0.4581570327281952, "learning_rate": 4.4260260085842816e-05, "loss": 0.0463, "step": 55980 }, { "grad_norm": 0.6618936657905579, "learning_rate": 4.4243835008963256e-05, "loss": 0.0368, "step": 55990 }, { "grad_norm": 0.339620977640152, "learning_rate": 4.422741056156925e-05, "loss": 0.0339, "step": 56000 }, { "grad_norm": 0.4027942717075348, "learning_rate": 4.421098674545693e-05, "loss": 0.0384, "step": 56010 }, { "grad_norm": 0.6105218529701233, "learning_rate": 4.419456356242241e-05, "loss": 0.0434, "step": 56020 }, { "grad_norm": 0.35352757573127747, "learning_rate": 4.4178141014261685e-05, "loss": 0.0412, "step": 56030 }, { "grad_norm": 0.4881362020969391, "learning_rate": 4.416171910277071e-05, "loss": 0.0403, "step": 56040 }, { "grad_norm": 0.5324360132217407, "learning_rate": 4.414529782974533e-05, "loss": 0.0378, "step": 56050 }, { "grad_norm": 0.36473411321640015, "learning_rate": 4.41288771969814e-05, "loss": 0.0384, "step": 56060 }, { "grad_norm": 0.4942525029182434, "learning_rate": 4.411245720627464e-05, "loss": 0.0419, "step": 56070 }, { "grad_norm": 0.4351188838481903, "learning_rate": 4.40960378594207e-05, "loss": 0.0415, "step": 56080 }, { "grad_norm": 0.7215514779090881, "learning_rate": 4.407961915821519e-05, "loss": 0.0481, "step": 56090 }, { "grad_norm": 0.3163788616657257, "learning_rate": 4.406320110445362e-05, "loss": 0.0381, "step": 56100 }, { "grad_norm": 0.47855621576309204, "learning_rate": 4.404678369993144e-05, "loss": 0.0373, "step": 56110 }, { "grad_norm": 0.5689253211021423, "learning_rate": 4.403036694644406e-05, "loss": 0.0484, "step": 56120 }, { "grad_norm": 0.34256038069725037, "learning_rate": 4.4013950845786764e-05, "loss": 0.0425, "step": 56130 }, { "grad_norm": 0.4440479576587677, "learning_rate": 4.399753539975482e-05, "loss": 0.0564, "step": 56140 }, { "grad_norm": 0.3809376060962677, "learning_rate": 4.398112061014337e-05, "loss": 0.0423, "step": 56150 }, { "grad_norm": 0.31667977571487427, "learning_rate": 4.396470647874753e-05, "loss": 0.0447, "step": 56160 }, { "grad_norm": 0.6888822317123413, "learning_rate": 4.394829300736229e-05, "loss": 0.0312, "step": 56170 }, { "grad_norm": 1.081640601158142, "learning_rate": 4.393188019778265e-05, "loss": 0.0454, "step": 56180 }, { "grad_norm": 0.22309482097625732, "learning_rate": 4.391546805180347e-05, "loss": 0.0429, "step": 56190 }, { "grad_norm": 0.686524510383606, "learning_rate": 4.389905657121955e-05, "loss": 0.0404, "step": 56200 }, { "grad_norm": 0.5613188147544861, "learning_rate": 4.3882645757825635e-05, "loss": 0.0395, "step": 56210 }, { "grad_norm": 0.282699316740036, "learning_rate": 4.386623561341637e-05, "loss": 0.0353, "step": 56220 }, { "grad_norm": 1.2598379850387573, "learning_rate": 4.384982613978637e-05, "loss": 0.0441, "step": 56230 }, { "grad_norm": 0.47667378187179565, "learning_rate": 4.383341733873012e-05, "loss": 0.0374, "step": 56240 }, { "grad_norm": 0.48757627606391907, "learning_rate": 4.381700921204209e-05, "loss": 0.0419, "step": 56250 }, { "grad_norm": 1.5277410745620728, "learning_rate": 4.380060176151663e-05, "loss": 0.0372, "step": 56260 }, { "grad_norm": 0.35414281487464905, "learning_rate": 4.378419498894805e-05, "loss": 0.0498, "step": 56270 }, { "grad_norm": 0.44639837741851807, "learning_rate": 4.376778889613056e-05, "loss": 0.0441, "step": 56280 }, { "grad_norm": 0.382022887468338, "learning_rate": 4.375138348485831e-05, "loss": 0.0491, "step": 56290 }, { "grad_norm": 0.4356871545314789, "learning_rate": 4.3734978756925346e-05, "loss": 0.0372, "step": 56300 }, { "grad_norm": 0.46291959285736084, "learning_rate": 4.3718574714125706e-05, "loss": 0.0463, "step": 56310 }, { "grad_norm": 0.508655846118927, "learning_rate": 4.370217135825329e-05, "loss": 0.0479, "step": 56320 }, { "grad_norm": 0.44095781445503235, "learning_rate": 4.368576869110194e-05, "loss": 0.0417, "step": 56330 }, { "grad_norm": 0.4252943992614746, "learning_rate": 4.366936671446544e-05, "loss": 0.0447, "step": 56340 }, { "grad_norm": 0.5258497595787048, "learning_rate": 4.3652965430137474e-05, "loss": 0.0438, "step": 56350 }, { "grad_norm": 0.4096919894218445, "learning_rate": 4.3636564839911646e-05, "loss": 0.0324, "step": 56360 }, { "grad_norm": 0.37449511885643005, "learning_rate": 4.3620164945581545e-05, "loss": 0.0376, "step": 56370 }, { "grad_norm": 0.5544416904449463, "learning_rate": 4.360376574894061e-05, "loss": 0.0428, "step": 56380 }, { "grad_norm": 0.3348428010940552, "learning_rate": 4.358736725178224e-05, "loss": 0.0398, "step": 56390 }, { "grad_norm": 0.5323742032051086, "learning_rate": 4.357096945589974e-05, "loss": 0.0479, "step": 56400 }, { "grad_norm": 0.45429766178131104, "learning_rate": 4.355457236308636e-05, "loss": 0.0384, "step": 56410 }, { "grad_norm": 0.45484861731529236, "learning_rate": 4.353817597513526e-05, "loss": 0.0371, "step": 56420 }, { "grad_norm": 0.5304962396621704, "learning_rate": 4.352178029383948e-05, "loss": 0.0332, "step": 56430 }, { "grad_norm": 0.4562714993953705, "learning_rate": 4.3505385320992105e-05, "loss": 0.0465, "step": 56440 }, { "grad_norm": 0.4138622283935547, "learning_rate": 4.348899105838602e-05, "loss": 0.0335, "step": 56450 }, { "grad_norm": 0.6248294115066528, "learning_rate": 4.3472597507814087e-05, "loss": 0.04, "step": 56460 }, { "grad_norm": 0.6628825068473816, "learning_rate": 4.3456204671069066e-05, "loss": 0.0438, "step": 56470 }, { "grad_norm": 0.4059847295284271, "learning_rate": 4.343981254994367e-05, "loss": 0.0394, "step": 56480 }, { "grad_norm": 0.38488152623176575, "learning_rate": 4.34234211462305e-05, "loss": 0.0414, "step": 56490 }, { "grad_norm": 0.4169490933418274, "learning_rate": 4.340703046172213e-05, "loss": 0.0405, "step": 56500 }, { "grad_norm": 0.21841777861118317, "learning_rate": 4.339064049821097e-05, "loss": 0.0507, "step": 56510 }, { "grad_norm": 0.4798460304737091, "learning_rate": 4.3374251257489446e-05, "loss": 0.0461, "step": 56520 }, { "grad_norm": 0.5723845362663269, "learning_rate": 4.335786274134984e-05, "loss": 0.0505, "step": 56530 }, { "grad_norm": 0.48316600918769836, "learning_rate": 4.334147495158438e-05, "loss": 0.0484, "step": 56540 }, { "grad_norm": 0.48241275548934937, "learning_rate": 4.33250878899852e-05, "loss": 0.0497, "step": 56550 }, { "grad_norm": 0.42796680331230164, "learning_rate": 4.330870155834437e-05, "loss": 0.0418, "step": 56560 }, { "grad_norm": 0.4787188470363617, "learning_rate": 4.3292315958453884e-05, "loss": 0.0424, "step": 56570 }, { "grad_norm": 1.0225106477737427, "learning_rate": 4.3275931092105646e-05, "loss": 0.0423, "step": 56580 }, { "grad_norm": 1.4284859895706177, "learning_rate": 4.325954696109147e-05, "loss": 0.0506, "step": 56590 }, { "grad_norm": 0.5301995873451233, "learning_rate": 4.32431635672031e-05, "loss": 0.0466, "step": 56600 }, { "grad_norm": 0.5338922142982483, "learning_rate": 4.3226780912232214e-05, "loss": 0.047, "step": 56610 }, { "grad_norm": 0.29333099722862244, "learning_rate": 4.321039899797036e-05, "loss": 0.0439, "step": 56620 }, { "grad_norm": 0.34843355417251587, "learning_rate": 4.319401782620908e-05, "loss": 0.0402, "step": 56630 }, { "grad_norm": 0.3846861720085144, "learning_rate": 4.317763739873978e-05, "loss": 0.0388, "step": 56640 }, { "grad_norm": 0.6215417981147766, "learning_rate": 4.31612577173538e-05, "loss": 0.0402, "step": 56650 }, { "grad_norm": 0.6165185570716858, "learning_rate": 4.314487878384239e-05, "loss": 0.0467, "step": 56660 }, { "grad_norm": 0.5683592557907104, "learning_rate": 4.3128500599996736e-05, "loss": 0.0486, "step": 56670 }, { "grad_norm": 0.5101858377456665, "learning_rate": 4.31121231676079e-05, "loss": 0.0395, "step": 56680 }, { "grad_norm": 0.5711920857429504, "learning_rate": 4.309574648846694e-05, "loss": 0.0489, "step": 56690 }, { "grad_norm": 0.8533223271369934, "learning_rate": 4.3079370564364755e-05, "loss": 0.054, "step": 56700 }, { "grad_norm": 0.37841013073921204, "learning_rate": 4.306299539709221e-05, "loss": 0.0478, "step": 56710 }, { "grad_norm": 0.3110511898994446, "learning_rate": 4.3046620988440046e-05, "loss": 0.0424, "step": 56720 }, { "grad_norm": 0.4802863895893097, "learning_rate": 4.303024734019897e-05, "loss": 0.0467, "step": 56730 }, { "grad_norm": 0.6967442035675049, "learning_rate": 4.3013874454159565e-05, "loss": 0.0464, "step": 56740 }, { "grad_norm": 0.8454430103302002, "learning_rate": 4.299750233211233e-05, "loss": 0.0505, "step": 56750 }, { "grad_norm": 0.48849472403526306, "learning_rate": 4.2981130975847715e-05, "loss": 0.0374, "step": 56760 }, { "grad_norm": 0.5467495918273926, "learning_rate": 4.2964760387156076e-05, "loss": 0.0449, "step": 56770 }, { "grad_norm": 0.5602744817733765, "learning_rate": 4.294839056782767e-05, "loss": 0.0437, "step": 56780 }, { "grad_norm": 0.5134314894676208, "learning_rate": 4.293202151965266e-05, "loss": 0.0507, "step": 56790 }, { "grad_norm": 0.3882778286933899, "learning_rate": 4.291565324442115e-05, "loss": 0.0376, "step": 56800 }, { "grad_norm": 0.4332055449485779, "learning_rate": 4.2899285743923135e-05, "loss": 0.0428, "step": 56810 }, { "grad_norm": 0.3396044373512268, "learning_rate": 4.288291901994857e-05, "loss": 0.0349, "step": 56820 }, { "grad_norm": 0.5108222365379333, "learning_rate": 4.286655307428729e-05, "loss": 0.0396, "step": 56830 }, { "grad_norm": 0.5511975884437561, "learning_rate": 4.285018790872903e-05, "loss": 0.0479, "step": 56840 }, { "grad_norm": 0.5930910110473633, "learning_rate": 4.283382352506347e-05, "loss": 0.0534, "step": 56850 }, { "grad_norm": 0.3271118104457855, "learning_rate": 4.2817459925080205e-05, "loss": 0.062, "step": 56860 }, { "grad_norm": 0.5143041014671326, "learning_rate": 4.28010971105687e-05, "loss": 0.0448, "step": 56870 }, { "grad_norm": 0.6994273066520691, "learning_rate": 4.278473508331841e-05, "loss": 0.0436, "step": 56880 }, { "grad_norm": 0.5484669804573059, "learning_rate": 4.276837384511864e-05, "loss": 0.0456, "step": 56890 }, { "grad_norm": 0.2917720377445221, "learning_rate": 4.275201339775864e-05, "loss": 0.0366, "step": 56900 }, { "grad_norm": 0.5229719281196594, "learning_rate": 4.273565374302756e-05, "loss": 0.0447, "step": 56910 }, { "grad_norm": 0.5488879680633545, "learning_rate": 4.2719294882714454e-05, "loss": 0.0382, "step": 56920 }, { "grad_norm": 0.38956916332244873, "learning_rate": 4.270293681860831e-05, "loss": 0.0401, "step": 56930 }, { "grad_norm": 0.9761305451393127, "learning_rate": 4.268657955249802e-05, "loss": 0.0404, "step": 56940 }, { "grad_norm": 0.7112489938735962, "learning_rate": 4.267022308617241e-05, "loss": 0.0352, "step": 56950 }, { "grad_norm": 0.8223727345466614, "learning_rate": 4.265386742142018e-05, "loss": 0.0487, "step": 56960 }, { "grad_norm": 0.5555541515350342, "learning_rate": 4.263751256002996e-05, "loss": 0.037, "step": 56970 }, { "grad_norm": 0.4953358471393585, "learning_rate": 4.26211585037903e-05, "loss": 0.0499, "step": 56980 }, { "grad_norm": 0.5370742678642273, "learning_rate": 4.2604805254489656e-05, "loss": 0.0459, "step": 56990 }, { "grad_norm": 1.5001437664031982, "learning_rate": 4.258845281391638e-05, "loss": 0.0401, "step": 57000 }, { "grad_norm": 0.3853273391723633, "learning_rate": 4.257210118385876e-05, "loss": 0.0426, "step": 57010 }, { "grad_norm": 0.5795443058013916, "learning_rate": 4.2555750366104994e-05, "loss": 0.0475, "step": 57020 }, { "grad_norm": 0.4134390354156494, "learning_rate": 4.2539400362443184e-05, "loss": 0.0378, "step": 57030 }, { "grad_norm": 0.541935920715332, "learning_rate": 4.252305117466133e-05, "loss": 0.0549, "step": 57040 }, { "grad_norm": 0.4437970519065857, "learning_rate": 4.2506702804547356e-05, "loss": 0.0451, "step": 57050 }, { "grad_norm": 0.7537166476249695, "learning_rate": 4.249035525388907e-05, "loss": 0.0409, "step": 57060 }, { "grad_norm": 0.6067507863044739, "learning_rate": 4.247400852447427e-05, "loss": 0.0442, "step": 57070 }, { "grad_norm": 0.4042118489742279, "learning_rate": 4.245766261809059e-05, "loss": 0.0457, "step": 57080 }, { "grad_norm": 0.33421337604522705, "learning_rate": 4.244131753652558e-05, "loss": 0.0379, "step": 57090 }, { "grad_norm": 0.3684192895889282, "learning_rate": 4.242497328156672e-05, "loss": 0.0366, "step": 57100 }, { "grad_norm": 0.542325496673584, "learning_rate": 4.240862985500139e-05, "loss": 0.0478, "step": 57110 }, { "grad_norm": 0.29562944173812866, "learning_rate": 4.2392287258616894e-05, "loss": 0.0346, "step": 57120 }, { "grad_norm": 0.3823264241218567, "learning_rate": 4.2375945494200396e-05, "loss": 0.0304, "step": 57130 }, { "grad_norm": 0.4447661340236664, "learning_rate": 4.2359604563539054e-05, "loss": 0.0482, "step": 57140 }, { "grad_norm": 0.2852008044719696, "learning_rate": 4.2343264468419885e-05, "loss": 0.0403, "step": 57150 }, { "grad_norm": 0.5654098987579346, "learning_rate": 4.232692521062979e-05, "loss": 0.0311, "step": 57160 }, { "grad_norm": 0.5719883441925049, "learning_rate": 4.231058679195562e-05, "loss": 0.0374, "step": 57170 }, { "grad_norm": 0.4862084984779358, "learning_rate": 4.229424921418411e-05, "loss": 0.0378, "step": 57180 }, { "grad_norm": 0.461331307888031, "learning_rate": 4.2277912479101916e-05, "loss": 0.044, "step": 57190 }, { "grad_norm": 0.6380120515823364, "learning_rate": 4.226157658849562e-05, "loss": 0.04, "step": 57200 }, { "grad_norm": 0.3709696829319, "learning_rate": 4.2245241544151664e-05, "loss": 0.0413, "step": 57210 }, { "grad_norm": 0.5405169725418091, "learning_rate": 4.2228907347856425e-05, "loss": 0.0546, "step": 57220 }, { "grad_norm": 0.5443733334541321, "learning_rate": 4.2212574001396214e-05, "loss": 0.0401, "step": 57230 }, { "grad_norm": 0.5687107443809509, "learning_rate": 4.21962415065572e-05, "loss": 0.04, "step": 57240 }, { "grad_norm": 0.5414342880249023, "learning_rate": 4.217990986512548e-05, "loss": 0.0384, "step": 57250 }, { "grad_norm": 0.39932939410209656, "learning_rate": 4.216357907888704e-05, "loss": 0.0388, "step": 57260 }, { "grad_norm": 0.5472037196159363, "learning_rate": 4.2147249149627824e-05, "loss": 0.0421, "step": 57270 }, { "grad_norm": 0.411886066198349, "learning_rate": 4.213092007913364e-05, "loss": 0.0404, "step": 57280 }, { "grad_norm": 0.2752096951007843, "learning_rate": 4.2114591869190205e-05, "loss": 0.0468, "step": 57290 }, { "grad_norm": 0.5421522855758667, "learning_rate": 4.209826452158315e-05, "loss": 0.0479, "step": 57300 }, { "grad_norm": 0.6241905093193054, "learning_rate": 4.2081938038098e-05, "loss": 0.0421, "step": 57310 }, { "grad_norm": 0.3913899064064026, "learning_rate": 4.2065612420520185e-05, "loss": 0.0343, "step": 57320 }, { "grad_norm": 0.28278058767318726, "learning_rate": 4.204928767063509e-05, "loss": 0.0398, "step": 57330 }, { "grad_norm": 0.3390452563762665, "learning_rate": 4.2032963790227944e-05, "loss": 0.0356, "step": 57340 }, { "grad_norm": 0.2784368395805359, "learning_rate": 4.2016640781083887e-05, "loss": 0.0353, "step": 57350 }, { "grad_norm": 0.4884042739868164, "learning_rate": 4.2000318644988004e-05, "loss": 0.0598, "step": 57360 }, { "grad_norm": 0.44283241033554077, "learning_rate": 4.198399738372524e-05, "loss": 0.0374, "step": 57370 }, { "grad_norm": 0.3621896207332611, "learning_rate": 4.196767699908045e-05, "loss": 0.0373, "step": 57380 }, { "grad_norm": 0.5570928454399109, "learning_rate": 4.195135749283845e-05, "loss": 0.0418, "step": 57390 }, { "grad_norm": 0.3598417639732361, "learning_rate": 4.1935038866783894e-05, "loss": 0.0333, "step": 57400 }, { "grad_norm": 0.3250925540924072, "learning_rate": 4.1918721122701355e-05, "loss": 0.0358, "step": 57410 }, { "grad_norm": 0.651942253112793, "learning_rate": 4.190240426237533e-05, "loss": 0.0586, "step": 57420 }, { "grad_norm": 0.2937028706073761, "learning_rate": 4.188608828759019e-05, "loss": 0.0388, "step": 57430 }, { "grad_norm": 0.46338215470314026, "learning_rate": 4.1869773200130246e-05, "loss": 0.0677, "step": 57440 }, { "grad_norm": 0.7273699641227722, "learning_rate": 4.185345900177967e-05, "loss": 0.0494, "step": 57450 }, { "grad_norm": 0.2799966335296631, "learning_rate": 4.183714569432258e-05, "loss": 0.0457, "step": 57460 }, { "grad_norm": 0.4189551770687103, "learning_rate": 4.182083327954295e-05, "loss": 0.04, "step": 57470 }, { "grad_norm": 0.3513379693031311, "learning_rate": 4.1804521759224714e-05, "loss": 0.0398, "step": 57480 }, { "grad_norm": 0.2739327549934387, "learning_rate": 4.178821113515165e-05, "loss": 0.0413, "step": 57490 }, { "grad_norm": 0.471042275428772, "learning_rate": 4.1771901409107476e-05, "loss": 0.0458, "step": 57500 }, { "grad_norm": 0.5288673639297485, "learning_rate": 4.175559258287577e-05, "loss": 0.0371, "step": 57510 }, { "grad_norm": 0.6219507455825806, "learning_rate": 4.173928465824008e-05, "loss": 0.0361, "step": 57520 }, { "grad_norm": 0.49352818727493286, "learning_rate": 4.17229776369838e-05, "loss": 0.0425, "step": 57530 }, { "grad_norm": 0.7017719149589539, "learning_rate": 4.170667152089025e-05, "loss": 0.0535, "step": 57540 }, { "grad_norm": 0.42152437567710876, "learning_rate": 4.1690366311742626e-05, "loss": 0.0421, "step": 57550 }, { "grad_norm": 0.7042508721351624, "learning_rate": 4.1674062011324055e-05, "loss": 0.0412, "step": 57560 }, { "grad_norm": 0.31805500388145447, "learning_rate": 4.165775862141752e-05, "loss": 0.0349, "step": 57570 }, { "grad_norm": 0.5508809685707092, "learning_rate": 4.164145614380598e-05, "loss": 0.0373, "step": 57580 }, { "grad_norm": 0.3120655119419098, "learning_rate": 4.162515458027223e-05, "loss": 0.0346, "step": 57590 }, { "grad_norm": 0.47544002532958984, "learning_rate": 4.160885393259899e-05, "loss": 0.0461, "step": 57600 }, { "grad_norm": 0.4448734521865845, "learning_rate": 4.159255420256886e-05, "loss": 0.0465, "step": 57610 }, { "grad_norm": 0.7198603749275208, "learning_rate": 4.157625539196436e-05, "loss": 0.0376, "step": 57620 }, { "grad_norm": 0.3118217885494232, "learning_rate": 4.155995750256791e-05, "loss": 0.0467, "step": 57630 }, { "grad_norm": 0.3861401081085205, "learning_rate": 4.15436605361618e-05, "loss": 0.0475, "step": 57640 }, { "grad_norm": 0.3637198507785797, "learning_rate": 4.152736449452827e-05, "loss": 0.0529, "step": 57650 }, { "grad_norm": 0.41080227494239807, "learning_rate": 4.151106937944943e-05, "loss": 0.0421, "step": 57660 }, { "grad_norm": 0.5824777483940125, "learning_rate": 4.149477519270727e-05, "loss": 0.0383, "step": 57670 }, { "grad_norm": 0.6721168756484985, "learning_rate": 4.147848193608371e-05, "loss": 0.054, "step": 57680 }, { "grad_norm": 0.45058244466781616, "learning_rate": 4.146218961136056e-05, "loss": 0.0396, "step": 57690 }, { "grad_norm": 0.36226487159729004, "learning_rate": 4.14458982203195e-05, "loss": 0.0479, "step": 57700 }, { "grad_norm": 0.3405909538269043, "learning_rate": 4.1429607764742165e-05, "loss": 0.0469, "step": 57710 }, { "grad_norm": 0.5423495769500732, "learning_rate": 4.141331824641003e-05, "loss": 0.0481, "step": 57720 }, { "grad_norm": 0.4799780547618866, "learning_rate": 4.139702966710452e-05, "loss": 0.0569, "step": 57730 }, { "grad_norm": 0.39486169815063477, "learning_rate": 4.13807420286069e-05, "loss": 0.0444, "step": 57740 }, { "grad_norm": 0.3841056227684021, "learning_rate": 4.136445533269837e-05, "loss": 0.0463, "step": 57750 }, { "grad_norm": 0.33868271112442017, "learning_rate": 4.134816958116003e-05, "loss": 0.031, "step": 57760 }, { "grad_norm": 0.49898725748062134, "learning_rate": 4.133188477577283e-05, "loss": 0.0424, "step": 57770 }, { "grad_norm": 0.39966505765914917, "learning_rate": 4.131560091831771e-05, "loss": 0.0383, "step": 57780 }, { "grad_norm": 1.07423734664917, "learning_rate": 4.1299318010575405e-05, "loss": 0.0362, "step": 57790 }, { "grad_norm": 0.8988545536994934, "learning_rate": 4.128303605432661e-05, "loss": 0.0648, "step": 57800 }, { "grad_norm": 0.4673173725605011, "learning_rate": 4.126675505135188e-05, "loss": 0.0397, "step": 57810 }, { "grad_norm": 1.2881559133529663, "learning_rate": 4.125047500343169e-05, "loss": 0.0435, "step": 57820 }, { "grad_norm": 0.5522606372833252, "learning_rate": 4.1234195912346376e-05, "loss": 0.0421, "step": 57830 }, { "grad_norm": 0.40630435943603516, "learning_rate": 4.1217917779876235e-05, "loss": 0.0441, "step": 57840 }, { "grad_norm": 0.3565405309200287, "learning_rate": 4.120164060780141e-05, "loss": 0.0512, "step": 57850 }, { "grad_norm": 0.45356619358062744, "learning_rate": 4.1185364397901935e-05, "loss": 0.0438, "step": 57860 }, { "grad_norm": 0.37798774242401123, "learning_rate": 4.116908915195776e-05, "loss": 0.0395, "step": 57870 }, { "grad_norm": 0.4549140930175781, "learning_rate": 4.115281487174872e-05, "loss": 0.0364, "step": 57880 }, { "grad_norm": 0.7134655714035034, "learning_rate": 4.113654155905452e-05, "loss": 0.0372, "step": 57890 }, { "grad_norm": 0.45543697476387024, "learning_rate": 4.112026921565483e-05, "loss": 0.0402, "step": 57900 }, { "grad_norm": 0.5141527056694031, "learning_rate": 4.1103997843329147e-05, "loss": 0.0373, "step": 57910 }, { "grad_norm": 0.4257833957672119, "learning_rate": 4.108772744385689e-05, "loss": 0.0527, "step": 57920 }, { "grad_norm": 0.42783278226852417, "learning_rate": 4.1071458019017346e-05, "loss": 0.0352, "step": 57930 }, { "grad_norm": 0.4432189166545868, "learning_rate": 4.1055189570589734e-05, "loss": 0.0419, "step": 57940 }, { "grad_norm": 0.4226747155189514, "learning_rate": 4.103892210035315e-05, "loss": 0.0319, "step": 57950 }, { "grad_norm": 1.492741346359253, "learning_rate": 4.102265561008656e-05, "loss": 0.047, "step": 57960 }, { "grad_norm": 0.4907660186290741, "learning_rate": 4.100639010156885e-05, "loss": 0.0352, "step": 57970 }, { "grad_norm": 0.6517298817634583, "learning_rate": 4.099012557657882e-05, "loss": 0.0468, "step": 57980 }, { "grad_norm": 0.4564547836780548, "learning_rate": 4.09738620368951e-05, "loss": 0.0407, "step": 57990 }, { "grad_norm": 0.2500320374965668, "learning_rate": 4.095759948429626e-05, "loss": 0.0385, "step": 58000 }, { "grad_norm": 0.5338078737258911, "learning_rate": 4.094133792056075e-05, "loss": 0.0452, "step": 58010 }, { "grad_norm": 0.28786852955818176, "learning_rate": 4.092507734746688e-05, "loss": 0.0477, "step": 58020 }, { "grad_norm": 0.4206133782863617, "learning_rate": 4.090881776679293e-05, "loss": 0.037, "step": 58030 }, { "grad_norm": 0.4233384132385254, "learning_rate": 4.0892559180316994e-05, "loss": 0.0383, "step": 58040 }, { "grad_norm": 0.5610241293907166, "learning_rate": 4.08763015898171e-05, "loss": 0.0376, "step": 58050 }, { "grad_norm": 0.47179505228996277, "learning_rate": 4.0860044997071145e-05, "loss": 0.0463, "step": 58060 }, { "grad_norm": 0.554297149181366, "learning_rate": 4.084378940385693e-05, "loss": 0.0437, "step": 58070 }, { "grad_norm": 0.48922276496887207, "learning_rate": 4.082753481195211e-05, "loss": 0.0389, "step": 58080 }, { "grad_norm": 0.3903045952320099, "learning_rate": 4.0811281223134324e-05, "loss": 0.0389, "step": 58090 }, { "grad_norm": 0.4253093898296356, "learning_rate": 4.0795028639181e-05, "loss": 0.0455, "step": 58100 }, { "grad_norm": 1.2765823602676392, "learning_rate": 4.077877706186951e-05, "loss": 0.0453, "step": 58110 }, { "grad_norm": 0.4460606575012207, "learning_rate": 4.076252649297709e-05, "loss": 0.0393, "step": 58120 }, { "grad_norm": 0.44401460886001587, "learning_rate": 4.07462769342809e-05, "loss": 0.0412, "step": 58130 }, { "grad_norm": 0.5316068530082703, "learning_rate": 4.073002838755793e-05, "loss": 0.0494, "step": 58140 }, { "grad_norm": 0.6552265882492065, "learning_rate": 4.0713780854585115e-05, "loss": 0.0468, "step": 58150 }, { "grad_norm": 0.519473671913147, "learning_rate": 4.069753433713927e-05, "loss": 0.0528, "step": 58160 }, { "grad_norm": 0.5011873245239258, "learning_rate": 4.068128883699709e-05, "loss": 0.036, "step": 58170 }, { "grad_norm": 0.5682612061500549, "learning_rate": 4.0665044355935155e-05, "loss": 0.0367, "step": 58180 }, { "grad_norm": 0.3519219160079956, "learning_rate": 4.064880089572993e-05, "loss": 0.0433, "step": 58190 }, { "grad_norm": 0.5062103271484375, "learning_rate": 4.063255845815779e-05, "loss": 0.0423, "step": 58200 }, { "grad_norm": 0.32775968313217163, "learning_rate": 4.0616317044994945e-05, "loss": 0.044, "step": 58210 }, { "grad_norm": 0.36263930797576904, "learning_rate": 4.0600076658017585e-05, "loss": 0.0357, "step": 58220 }, { "grad_norm": 0.47146666049957275, "learning_rate": 4.0583837299001706e-05, "loss": 0.0407, "step": 58230 }, { "grad_norm": 0.3684443235397339, "learning_rate": 4.056759896972322e-05, "loss": 0.0358, "step": 58240 }, { "grad_norm": 0.4230232536792755, "learning_rate": 4.055136167195793e-05, "loss": 0.0408, "step": 58250 }, { "grad_norm": 0.3422093987464905, "learning_rate": 4.053512540748152e-05, "loss": 0.0405, "step": 58260 }, { "grad_norm": 0.6722171306610107, "learning_rate": 4.051889017806955e-05, "loss": 0.0458, "step": 58270 }, { "grad_norm": 0.5867396593093872, "learning_rate": 4.0502655985497516e-05, "loss": 0.0408, "step": 58280 }, { "grad_norm": 0.5809038281440735, "learning_rate": 4.048642283154074e-05, "loss": 0.0542, "step": 58290 }, { "grad_norm": 0.31925830245018005, "learning_rate": 4.0470190717974464e-05, "loss": 0.0409, "step": 58300 }, { "grad_norm": 0.3967719078063965, "learning_rate": 4.045395964657381e-05, "loss": 0.042, "step": 58310 }, { "grad_norm": 0.5420838594436646, "learning_rate": 4.043772961911376e-05, "loss": 0.0481, "step": 58320 }, { "grad_norm": 0.4009375274181366, "learning_rate": 4.042150063736924e-05, "loss": 0.0456, "step": 58330 }, { "grad_norm": 0.48289385437965393, "learning_rate": 4.0405272703114976e-05, "loss": 0.0383, "step": 58340 }, { "grad_norm": 0.5723260045051575, "learning_rate": 4.038904581812569e-05, "loss": 0.048, "step": 58350 }, { "grad_norm": 0.4695133566856384, "learning_rate": 4.03728199841759e-05, "loss": 0.036, "step": 58360 }, { "grad_norm": 0.3134785592556, "learning_rate": 4.0356595203040044e-05, "loss": 0.0464, "step": 58370 }, { "grad_norm": 0.3188597559928894, "learning_rate": 4.0340371476492426e-05, "loss": 0.0474, "step": 58380 }, { "grad_norm": 0.47871461510658264, "learning_rate": 4.032414880630727e-05, "loss": 0.0399, "step": 58390 }, { "grad_norm": 0.5239722728729248, "learning_rate": 4.0307927194258646e-05, "loss": 0.0369, "step": 58400 }, { "grad_norm": 0.4247687757015228, "learning_rate": 4.029170664212054e-05, "loss": 0.0408, "step": 58410 }, { "grad_norm": 0.3464969992637634, "learning_rate": 4.0275487151666794e-05, "loss": 0.0494, "step": 58420 }, { "grad_norm": 0.5503040552139282, "learning_rate": 4.025926872467116e-05, "loss": 0.0449, "step": 58430 }, { "grad_norm": 0.4143165349960327, "learning_rate": 4.024305136290725e-05, "loss": 0.0471, "step": 58440 }, { "grad_norm": 0.37890493869781494, "learning_rate": 4.022683506814858e-05, "loss": 0.0491, "step": 58450 }, { "grad_norm": 0.5150409936904907, "learning_rate": 4.021061984216853e-05, "loss": 0.0493, "step": 58460 }, { "grad_norm": 0.6367658376693726, "learning_rate": 4.019440568674035e-05, "loss": 0.0399, "step": 58470 }, { "grad_norm": 0.7260384559631348, "learning_rate": 4.017819260363724e-05, "loss": 0.0401, "step": 58480 }, { "grad_norm": 0.5805335640907288, "learning_rate": 4.016198059463223e-05, "loss": 0.0556, "step": 58490 }, { "grad_norm": 0.23630385100841522, "learning_rate": 4.0145769661498233e-05, "loss": 0.0413, "step": 58500 }, { "grad_norm": 0.9768731594085693, "learning_rate": 4.012955980600804e-05, "loss": 0.0422, "step": 58510 }, { "grad_norm": 0.6484410166740417, "learning_rate": 4.0113351029934355e-05, "loss": 0.0443, "step": 58520 }, { "grad_norm": 0.3654024302959442, "learning_rate": 4.0097143335049714e-05, "loss": 0.0403, "step": 58530 }, { "grad_norm": 0.36753588914871216, "learning_rate": 4.0080936723126606e-05, "loss": 0.0306, "step": 58540 }, { "grad_norm": 0.37887275218963623, "learning_rate": 4.006473119593735e-05, "loss": 0.0401, "step": 58550 }, { "grad_norm": 0.4877026379108429, "learning_rate": 4.004852675525415e-05, "loss": 0.0431, "step": 58560 }, { "grad_norm": 0.40229541063308716, "learning_rate": 4.00323234028491e-05, "loss": 0.0397, "step": 58570 }, { "grad_norm": 0.4411360025405884, "learning_rate": 4.0016121140494186e-05, "loss": 0.0415, "step": 58580 }, { "grad_norm": 0.40087050199508667, "learning_rate": 3.999991996996122e-05, "loss": 0.042, "step": 58590 }, { "grad_norm": 0.36551615595817566, "learning_rate": 3.9983719893022e-05, "loss": 0.0347, "step": 58600 }, { "grad_norm": 0.5155504941940308, "learning_rate": 3.996752091144812e-05, "loss": 0.0393, "step": 58610 }, { "grad_norm": 0.5500565767288208, "learning_rate": 3.9951323027011055e-05, "loss": 0.0407, "step": 58620 }, { "grad_norm": 0.4260997474193573, "learning_rate": 3.993512624148219e-05, "loss": 0.0401, "step": 58630 }, { "grad_norm": 0.4727810025215149, "learning_rate": 3.99189305566328e-05, "loss": 0.0393, "step": 58640 }, { "grad_norm": 0.29091155529022217, "learning_rate": 3.990273597423401e-05, "loss": 0.0477, "step": 58650 }, { "grad_norm": 0.3905208110809326, "learning_rate": 3.988654249605681e-05, "loss": 0.0455, "step": 58660 }, { "grad_norm": 0.49521303176879883, "learning_rate": 3.987035012387213e-05, "loss": 0.0396, "step": 58670 }, { "grad_norm": 0.4818011224269867, "learning_rate": 3.9854158859450735e-05, "loss": 0.0354, "step": 58680 }, { "grad_norm": 0.37437787652015686, "learning_rate": 3.983796870456328e-05, "loss": 0.0423, "step": 58690 }, { "grad_norm": 0.3224804699420929, "learning_rate": 3.982177966098028e-05, "loss": 0.0363, "step": 58700 }, { "grad_norm": 0.49219757318496704, "learning_rate": 3.980559173047216e-05, "loss": 0.05, "step": 58710 }, { "grad_norm": 0.38220295310020447, "learning_rate": 3.9789404914809176e-05, "loss": 0.0441, "step": 58720 }, { "grad_norm": 0.350754052400589, "learning_rate": 3.9773219215761544e-05, "loss": 0.0423, "step": 58730 }, { "grad_norm": 0.34756261110305786, "learning_rate": 3.975703463509928e-05, "loss": 0.0313, "step": 58740 }, { "grad_norm": 1.8475209474563599, "learning_rate": 3.974085117459231e-05, "loss": 0.0363, "step": 58750 }, { "grad_norm": 0.3804871737957001, "learning_rate": 3.972466883601043e-05, "loss": 0.0491, "step": 58760 }, { "grad_norm": 0.29179927706718445, "learning_rate": 3.9708487621123316e-05, "loss": 0.0268, "step": 58770 }, { "grad_norm": 0.42908158898353577, "learning_rate": 3.969230753170051e-05, "loss": 0.05, "step": 58780 }, { "grad_norm": 0.4797123372554779, "learning_rate": 3.967612856951146e-05, "loss": 0.0368, "step": 58790 }, { "grad_norm": 0.5141481161117554, "learning_rate": 3.9659950736325476e-05, "loss": 0.036, "step": 58800 }, { "grad_norm": 1.129671573638916, "learning_rate": 3.964377403391174e-05, "loss": 0.0769, "step": 58810 }, { "grad_norm": 0.6973875164985657, "learning_rate": 3.9627598464039303e-05, "loss": 0.0625, "step": 58820 }, { "grad_norm": 0.27511343359947205, "learning_rate": 3.9611424028477096e-05, "loss": 0.0342, "step": 58830 }, { "grad_norm": 0.7697016000747681, "learning_rate": 3.959525072899394e-05, "loss": 0.0437, "step": 58840 }, { "grad_norm": 0.4017467200756073, "learning_rate": 3.9579078567358506e-05, "loss": 0.0371, "step": 58850 }, { "grad_norm": 0.5614559054374695, "learning_rate": 3.956290754533939e-05, "loss": 0.0425, "step": 58860 }, { "grad_norm": 0.6370790004730225, "learning_rate": 3.954673766470501e-05, "loss": 0.0469, "step": 58870 }, { "grad_norm": 0.3238096535205841, "learning_rate": 3.9530568927223674e-05, "loss": 0.0376, "step": 58880 }, { "grad_norm": 0.38420623540878296, "learning_rate": 3.95144013346636e-05, "loss": 0.0482, "step": 58890 }, { "grad_norm": 0.6699463129043579, "learning_rate": 3.949823488879282e-05, "loss": 0.0313, "step": 58900 }, { "grad_norm": 0.34141215682029724, "learning_rate": 3.9482069591379276e-05, "loss": 0.0452, "step": 58910 }, { "grad_norm": 0.35898637771606445, "learning_rate": 3.9465905444190784e-05, "loss": 0.0348, "step": 58920 }, { "grad_norm": 0.31171804666519165, "learning_rate": 3.944974244899505e-05, "loss": 0.0446, "step": 58930 }, { "grad_norm": 0.4021263122558594, "learning_rate": 3.9433580607559616e-05, "loss": 0.0374, "step": 58940 }, { "grad_norm": 0.373426616191864, "learning_rate": 3.941741992165193e-05, "loss": 0.0388, "step": 58950 }, { "grad_norm": 0.43366920948028564, "learning_rate": 3.940126039303928e-05, "loss": 0.0411, "step": 58960 }, { "grad_norm": 0.41307932138442993, "learning_rate": 3.938510202348886e-05, "loss": 0.0421, "step": 58970 }, { "grad_norm": 0.3650587201118469, "learning_rate": 3.93689448147677e-05, "loss": 0.0382, "step": 58980 }, { "grad_norm": 0.33539828658103943, "learning_rate": 3.935278876864278e-05, "loss": 0.0385, "step": 58990 }, { "grad_norm": 0.37343716621398926, "learning_rate": 3.933663388688087e-05, "loss": 0.0447, "step": 59000 }, { "grad_norm": 0.6417221426963806, "learning_rate": 3.9320480171248645e-05, "loss": 0.0364, "step": 59010 }, { "grad_norm": 0.5496613383293152, "learning_rate": 3.930432762351265e-05, "loss": 0.0446, "step": 59020 }, { "grad_norm": 0.3891635537147522, "learning_rate": 3.928817624543931e-05, "loss": 0.0438, "step": 59030 }, { "grad_norm": 0.3588433265686035, "learning_rate": 3.9272026038794885e-05, "loss": 0.0346, "step": 59040 }, { "grad_norm": 0.5560519099235535, "learning_rate": 3.9255877005345586e-05, "loss": 0.0412, "step": 59050 }, { "grad_norm": 0.7673013806343079, "learning_rate": 3.923972914685743e-05, "loss": 0.0469, "step": 59060 }, { "grad_norm": 0.3690527379512787, "learning_rate": 3.922358246509631e-05, "loss": 0.0385, "step": 59070 }, { "grad_norm": 0.4819459617137909, "learning_rate": 3.920743696182801e-05, "loss": 0.0448, "step": 59080 }, { "grad_norm": 0.3881509304046631, "learning_rate": 3.9191292638818164e-05, "loss": 0.0373, "step": 59090 }, { "grad_norm": 0.6865499019622803, "learning_rate": 3.9175149497832316e-05, "loss": 0.042, "step": 59100 }, { "grad_norm": 0.4080233871936798, "learning_rate": 3.915900754063584e-05, "loss": 0.0378, "step": 59110 }, { "grad_norm": 0.4097467064857483, "learning_rate": 3.9142866768994014e-05, "loss": 0.039, "step": 59120 }, { "grad_norm": 0.41412410140037537, "learning_rate": 3.912672718467193e-05, "loss": 0.0431, "step": 59130 }, { "grad_norm": 0.40495583415031433, "learning_rate": 3.911058878943463e-05, "loss": 0.0417, "step": 59140 }, { "grad_norm": 0.3685939908027649, "learning_rate": 3.9094451585046974e-05, "loss": 0.0329, "step": 59150 }, { "grad_norm": 0.35306259989738464, "learning_rate": 3.90783155732737e-05, "loss": 0.042, "step": 59160 }, { "grad_norm": 0.3434554934501648, "learning_rate": 3.9062180755879373e-05, "loss": 0.0373, "step": 59170 }, { "grad_norm": 0.9411885142326355, "learning_rate": 3.904604713462855e-05, "loss": 0.045, "step": 59180 }, { "grad_norm": 0.2971648871898651, "learning_rate": 3.902991471128554e-05, "loss": 0.0253, "step": 59190 }, { "grad_norm": 0.2807689607143402, "learning_rate": 3.901378348761456e-05, "loss": 0.0334, "step": 59200 }, { "grad_norm": 0.3116926848888397, "learning_rate": 3.899765346537971e-05, "loss": 0.0332, "step": 59210 }, { "grad_norm": 0.33155637979507446, "learning_rate": 3.898152464634493e-05, "loss": 0.032, "step": 59220 }, { "grad_norm": 0.5797122716903687, "learning_rate": 3.896539703227402e-05, "loss": 0.0379, "step": 59230 }, { "grad_norm": 0.587940514087677, "learning_rate": 3.894927062493073e-05, "loss": 0.0391, "step": 59240 }, { "grad_norm": 0.8716708421707153, "learning_rate": 3.893314542607859e-05, "loss": 0.0457, "step": 59250 }, { "grad_norm": 0.4505367577075958, "learning_rate": 3.891702143748103e-05, "loss": 0.0392, "step": 59260 }, { "grad_norm": 0.4265744984149933, "learning_rate": 3.890089866090133e-05, "loss": 0.0413, "step": 59270 }, { "grad_norm": 0.3966297209262848, "learning_rate": 3.888477709810267e-05, "loss": 0.0352, "step": 59280 }, { "grad_norm": 0.39744389057159424, "learning_rate": 3.886865675084806e-05, "loss": 0.0396, "step": 59290 }, { "grad_norm": 0.44981470704078674, "learning_rate": 3.885253762090043e-05, "loss": 0.042, "step": 59300 }, { "grad_norm": 0.5398343801498413, "learning_rate": 3.8836419710022534e-05, "loss": 0.0502, "step": 59310 }, { "grad_norm": 0.5092706680297852, "learning_rate": 3.882030301997698e-05, "loss": 0.0424, "step": 59320 }, { "grad_norm": 0.5282815098762512, "learning_rate": 3.880418755252629e-05, "loss": 0.0374, "step": 59330 }, { "grad_norm": 0.31376758217811584, "learning_rate": 3.87880733094328e-05, "loss": 0.0474, "step": 59340 }, { "grad_norm": 0.5253139138221741, "learning_rate": 3.877196029245877e-05, "loss": 0.0338, "step": 59350 }, { "grad_norm": 0.4322883188724518, "learning_rate": 3.875584850336627e-05, "loss": 0.053, "step": 59360 }, { "grad_norm": 0.6976908445358276, "learning_rate": 3.873973794391728e-05, "loss": 0.0352, "step": 59370 }, { "grad_norm": 0.3044375479221344, "learning_rate": 3.872362861587361e-05, "loss": 0.0307, "step": 59380 }, { "grad_norm": 0.454603374004364, "learning_rate": 3.870752052099698e-05, "loss": 0.0417, "step": 59390 }, { "grad_norm": 0.4573655426502228, "learning_rate": 3.8691413661048916e-05, "loss": 0.046, "step": 59400 }, { "grad_norm": 0.8760120868682861, "learning_rate": 3.867530803779085e-05, "loss": 0.0399, "step": 59410 }, { "grad_norm": 0.3236280083656311, "learning_rate": 3.8659203652984054e-05, "loss": 0.0382, "step": 59420 }, { "grad_norm": 0.4317496418952942, "learning_rate": 3.864310050838972e-05, "loss": 0.0391, "step": 59430 }, { "grad_norm": 0.3107442855834961, "learning_rate": 3.8626998605768835e-05, "loss": 0.0394, "step": 59440 }, { "grad_norm": 0.4387635588645935, "learning_rate": 3.8610897946882294e-05, "loss": 0.0471, "step": 59450 }, { "grad_norm": 0.6591599583625793, "learning_rate": 3.859479853349082e-05, "loss": 0.0502, "step": 59460 }, { "grad_norm": 1.4935065507888794, "learning_rate": 3.857870036735503e-05, "loss": 0.0423, "step": 59470 }, { "grad_norm": 0.7019153833389282, "learning_rate": 3.856260345023538e-05, "loss": 0.0471, "step": 59480 }, { "grad_norm": 0.303830623626709, "learning_rate": 3.854650778389224e-05, "loss": 0.0315, "step": 59490 }, { "grad_norm": 0.4741123616695404, "learning_rate": 3.8530413370085786e-05, "loss": 0.0336, "step": 59500 }, { "grad_norm": 0.2921995222568512, "learning_rate": 3.851432021057608e-05, "loss": 0.0336, "step": 59510 }, { "grad_norm": 0.4112580418586731, "learning_rate": 3.849822830712306e-05, "loss": 0.0429, "step": 59520 }, { "grad_norm": 0.45182114839553833, "learning_rate": 3.848213766148649e-05, "loss": 0.0375, "step": 59530 }, { "grad_norm": 0.6493595838546753, "learning_rate": 3.846604827542602e-05, "loss": 0.0367, "step": 59540 }, { "grad_norm": 0.6705962419509888, "learning_rate": 3.844996015070117e-05, "loss": 0.0431, "step": 59550 }, { "grad_norm": 0.4613913297653198, "learning_rate": 3.843387328907132e-05, "loss": 0.0365, "step": 59560 }, { "grad_norm": 0.4378148317337036, "learning_rate": 3.8417787692295696e-05, "loss": 0.047, "step": 59570 }, { "grad_norm": 0.3838958740234375, "learning_rate": 3.84017033621334e-05, "loss": 0.037, "step": 59580 }, { "grad_norm": 0.22274772822856903, "learning_rate": 3.838562030034338e-05, "loss": 0.0306, "step": 59590 }, { "grad_norm": 0.34201428294181824, "learning_rate": 3.836953850868447e-05, "loss": 0.0349, "step": 59600 }, { "grad_norm": 0.28470101952552795, "learning_rate": 3.835345798891533e-05, "loss": 0.0304, "step": 59610 }, { "grad_norm": 0.40751948952674866, "learning_rate": 3.833737874279452e-05, "loss": 0.0382, "step": 59620 }, { "grad_norm": 0.3969137966632843, "learning_rate": 3.8321300772080433e-05, "loss": 0.0254, "step": 59630 }, { "grad_norm": 0.9508298635482788, "learning_rate": 3.8305224078531345e-05, "loss": 0.0423, "step": 59640 }, { "grad_norm": 0.5086135864257812, "learning_rate": 3.828914866390537e-05, "loss": 0.0508, "step": 59650 }, { "grad_norm": 0.39771968126296997, "learning_rate": 3.827307452996048e-05, "loss": 0.0419, "step": 59660 }, { "grad_norm": 0.4681413471698761, "learning_rate": 3.825700167845454e-05, "loss": 0.049, "step": 59670 }, { "grad_norm": 0.38078954815864563, "learning_rate": 3.8240930111145214e-05, "loss": 0.0489, "step": 59680 }, { "grad_norm": 0.440396249294281, "learning_rate": 3.8224859829790116e-05, "loss": 0.0351, "step": 59690 }, { "grad_norm": 4.359667778015137, "learning_rate": 3.820879083614664e-05, "loss": 0.0539, "step": 59700 }, { "grad_norm": 0.5053653120994568, "learning_rate": 3.819272313197208e-05, "loss": 0.0437, "step": 59710 }, { "grad_norm": 9.581748962402344, "learning_rate": 3.817665671902355e-05, "loss": 0.0371, "step": 59720 }, { "grad_norm": 0.42528921365737915, "learning_rate": 3.816059159905808e-05, "loss": 0.0461, "step": 59730 }, { "grad_norm": 0.6317689418792725, "learning_rate": 3.8144527773832476e-05, "loss": 0.041, "step": 59740 }, { "grad_norm": 0.46459442377090454, "learning_rate": 3.812846524510352e-05, "loss": 0.0296, "step": 59750 }, { "grad_norm": 0.3753311336040497, "learning_rate": 3.811240401462775e-05, "loss": 0.0466, "step": 59760 }, { "grad_norm": 0.5013509392738342, "learning_rate": 3.809634408416162e-05, "loss": 0.0408, "step": 59770 }, { "grad_norm": 0.4425811469554901, "learning_rate": 3.8080285455461394e-05, "loss": 0.0399, "step": 59780 }, { "grad_norm": 0.6774542331695557, "learning_rate": 3.806422813028323e-05, "loss": 0.0339, "step": 59790 }, { "grad_norm": 0.5156309604644775, "learning_rate": 3.8048172110383114e-05, "loss": 0.0356, "step": 59800 }, { "grad_norm": 0.26462289690971375, "learning_rate": 3.8032117397516944e-05, "loss": 0.0407, "step": 59810 }, { "grad_norm": 0.5135742425918579, "learning_rate": 3.801606399344042e-05, "loss": 0.0453, "step": 59820 }, { "grad_norm": 0.35924530029296875, "learning_rate": 3.800001189990913e-05, "loss": 0.0422, "step": 59830 }, { "grad_norm": 0.4267888069152832, "learning_rate": 3.798396111867847e-05, "loss": 0.0376, "step": 59840 }, { "grad_norm": 0.7110745906829834, "learning_rate": 3.796791165150378e-05, "loss": 0.042, "step": 59850 }, { "grad_norm": 0.5539724230766296, "learning_rate": 3.795186350014018e-05, "loss": 0.0396, "step": 59860 }, { "grad_norm": 0.31351491808891296, "learning_rate": 3.793581666634266e-05, "loss": 0.0375, "step": 59870 }, { "grad_norm": 0.6255069971084595, "learning_rate": 3.791977115186609e-05, "loss": 0.0424, "step": 59880 }, { "grad_norm": 0.4254037141799927, "learning_rate": 3.79037269584652e-05, "loss": 0.0349, "step": 59890 }, { "grad_norm": 0.5565292239189148, "learning_rate": 3.7887684087894545e-05, "loss": 0.0436, "step": 59900 }, { "grad_norm": 0.31960228085517883, "learning_rate": 3.787164254190854e-05, "loss": 0.0406, "step": 59910 }, { "grad_norm": 0.3031725287437439, "learning_rate": 3.7855602322261473e-05, "loss": 0.0379, "step": 59920 }, { "grad_norm": 0.3369615077972412, "learning_rate": 3.783956343070746e-05, "loss": 0.0475, "step": 59930 }, { "grad_norm": 0.9283498525619507, "learning_rate": 3.7823525869000526e-05, "loss": 0.0423, "step": 59940 }, { "grad_norm": 0.5398930907249451, "learning_rate": 3.78074896388945e-05, "loss": 0.04, "step": 59950 }, { "grad_norm": 0.5693472623825073, "learning_rate": 3.7791454742143075e-05, "loss": 0.0488, "step": 59960 }, { "grad_norm": 0.49606671929359436, "learning_rate": 3.7775421180499795e-05, "loss": 0.036, "step": 59970 }, { "grad_norm": 0.30193090438842773, "learning_rate": 3.7759388955718086e-05, "loss": 0.0436, "step": 59980 }, { "grad_norm": 0.2923566997051239, "learning_rate": 3.774335806955117e-05, "loss": 0.0411, "step": 59990 }, { "grad_norm": 0.3770946264266968, "learning_rate": 3.7727328523752215e-05, "loss": 0.0321, "step": 60000 }, { "grad_norm": 0.529253363609314, "learning_rate": 3.771130032007415e-05, "loss": 0.0383, "step": 60010 }, { "grad_norm": 0.5163702964782715, "learning_rate": 3.769527346026982e-05, "loss": 0.0317, "step": 60020 }, { "grad_norm": 0.4191693663597107, "learning_rate": 3.767924794609188e-05, "loss": 0.0369, "step": 60030 }, { "grad_norm": 0.4877806305885315, "learning_rate": 3.7663223779292854e-05, "loss": 0.0414, "step": 60040 }, { "grad_norm": 0.32235345244407654, "learning_rate": 3.764720096162512e-05, "loss": 0.043, "step": 60050 }, { "grad_norm": 0.558406412601471, "learning_rate": 3.763117949484092e-05, "loss": 0.0361, "step": 60060 }, { "grad_norm": 0.8377243876457214, "learning_rate": 3.7615159380692344e-05, "loss": 0.0463, "step": 60070 }, { "grad_norm": 0.5863171815872192, "learning_rate": 3.759914062093132e-05, "loss": 0.0302, "step": 60080 }, { "grad_norm": 0.46456649899482727, "learning_rate": 3.7583123217309615e-05, "loss": 0.0405, "step": 60090 }, { "grad_norm": 0.40293997526168823, "learning_rate": 3.7567107171578904e-05, "loss": 0.0457, "step": 60100 }, { "grad_norm": 0.9943296313285828, "learning_rate": 3.755109248549066e-05, "loss": 0.0478, "step": 60110 }, { "grad_norm": 0.5489662289619446, "learning_rate": 3.75350791607962e-05, "loss": 0.0358, "step": 60120 }, { "grad_norm": 0.722744345664978, "learning_rate": 3.751906719924676e-05, "loss": 0.0394, "step": 60130 }, { "grad_norm": 0.33182796835899353, "learning_rate": 3.750305660259337e-05, "loss": 0.0355, "step": 60140 }, { "grad_norm": 0.44525784254074097, "learning_rate": 3.748704737258693e-05, "loss": 0.0384, "step": 60150 }, { "grad_norm": 0.4198225140571594, "learning_rate": 3.747103951097816e-05, "loss": 0.0559, "step": 60160 }, { "grad_norm": 0.7177860736846924, "learning_rate": 3.745503301951768e-05, "loss": 0.0436, "step": 60170 }, { "grad_norm": 0.5196357369422913, "learning_rate": 3.743902789995592e-05, "loss": 0.0401, "step": 60180 }, { "grad_norm": 0.33387961983680725, "learning_rate": 3.7423024154043183e-05, "loss": 0.0362, "step": 60190 }, { "grad_norm": 0.6851571798324585, "learning_rate": 3.740702178352963e-05, "loss": 0.0361, "step": 60200 }, { "grad_norm": 3.057244300842285, "learning_rate": 3.739102079016523e-05, "loss": 0.0506, "step": 60210 }, { "grad_norm": 0.46162331104278564, "learning_rate": 3.7375021175699846e-05, "loss": 0.0326, "step": 60220 }, { "grad_norm": 0.4929504692554474, "learning_rate": 3.735902294188317e-05, "loss": 0.039, "step": 60230 }, { "grad_norm": 0.5413495302200317, "learning_rate": 3.734302609046473e-05, "loss": 0.0368, "step": 60240 }, { "grad_norm": 0.5325109958648682, "learning_rate": 3.732703062319391e-05, "loss": 0.0456, "step": 60250 }, { "grad_norm": 1.190597414970398, "learning_rate": 3.731103654181999e-05, "loss": 0.0521, "step": 60260 }, { "grad_norm": 0.5609275102615356, "learning_rate": 3.729504384809203e-05, "loss": 0.0378, "step": 60270 }, { "grad_norm": 0.39088988304138184, "learning_rate": 3.727905254375898e-05, "loss": 0.0373, "step": 60280 }, { "grad_norm": 0.4444648325443268, "learning_rate": 3.7263062630569614e-05, "loss": 0.0368, "step": 60290 }, { "grad_norm": 0.68224036693573, "learning_rate": 3.724707411027256e-05, "loss": 0.0409, "step": 60300 }, { "grad_norm": 0.3988424241542816, "learning_rate": 3.723108698461631e-05, "loss": 0.0348, "step": 60310 }, { "grad_norm": 1.0142840147018433, "learning_rate": 3.72151012553492e-05, "loss": 0.0429, "step": 60320 }, { "grad_norm": 0.39047354459762573, "learning_rate": 3.719911692421939e-05, "loss": 0.0287, "step": 60330 }, { "grad_norm": 0.47054150700569153, "learning_rate": 3.71831339929749e-05, "loss": 0.0369, "step": 60340 }, { "grad_norm": 0.42812854051589966, "learning_rate": 3.716715246336362e-05, "loss": 0.036, "step": 60350 }, { "grad_norm": 0.32454079389572144, "learning_rate": 3.715117233713324e-05, "loss": 0.0312, "step": 60360 }, { "grad_norm": 0.3509271740913391, "learning_rate": 3.713519361603135e-05, "loss": 0.0359, "step": 60370 }, { "grad_norm": 0.35295093059539795, "learning_rate": 3.711921630180532e-05, "loss": 0.0456, "step": 60380 }, { "grad_norm": 0.35789167881011963, "learning_rate": 3.710324039620245e-05, "loss": 0.039, "step": 60390 }, { "grad_norm": 0.48087435960769653, "learning_rate": 3.708726590096982e-05, "loss": 0.0424, "step": 60400 }, { "grad_norm": 0.4328816533088684, "learning_rate": 3.707129281785437e-05, "loss": 0.0362, "step": 60410 }, { "grad_norm": 0.5454679727554321, "learning_rate": 3.705532114860291e-05, "loss": 0.0528, "step": 60420 }, { "grad_norm": 0.4553897976875305, "learning_rate": 3.703935089496207e-05, "loss": 0.0596, "step": 60430 }, { "grad_norm": 0.34599384665489197, "learning_rate": 3.702338205867829e-05, "loss": 0.036, "step": 60440 }, { "grad_norm": 0.4716150164604187, "learning_rate": 3.7007414641497976e-05, "loss": 0.0424, "step": 60450 }, { "grad_norm": 0.5349326729774475, "learning_rate": 3.699144864516726e-05, "loss": 0.0367, "step": 60460 }, { "grad_norm": 0.5439233779907227, "learning_rate": 3.6975484071432153e-05, "loss": 0.041, "step": 60470 }, { "grad_norm": 5.213205337524414, "learning_rate": 3.695952092203854e-05, "loss": 0.0439, "step": 60480 }, { "grad_norm": 0.5109161138534546, "learning_rate": 3.69435591987321e-05, "loss": 0.0471, "step": 60490 }, { "grad_norm": 1.4602775573730469, "learning_rate": 3.6927598903258374e-05, "loss": 0.0388, "step": 60500 }, { "grad_norm": 0.42610403895378113, "learning_rate": 3.691164003736281e-05, "loss": 0.0278, "step": 60510 }, { "grad_norm": 0.366473913192749, "learning_rate": 3.689568260279061e-05, "loss": 0.0369, "step": 60520 }, { "grad_norm": 0.37815746665000916, "learning_rate": 3.687972660128686e-05, "loss": 0.0372, "step": 60530 }, { "grad_norm": 0.6722061634063721, "learning_rate": 3.686377203459648e-05, "loss": 0.0437, "step": 60540 }, { "grad_norm": 0.382449746131897, "learning_rate": 3.6847818904464226e-05, "loss": 0.0405, "step": 60550 }, { "grad_norm": 0.32062584161758423, "learning_rate": 3.683186721263474e-05, "loss": 0.0364, "step": 60560 }, { "grad_norm": 0.494525283575058, "learning_rate": 3.6815916960852435e-05, "loss": 0.0431, "step": 60570 }, { "grad_norm": 1.2209802865982056, "learning_rate": 3.679996815086165e-05, "loss": 0.0412, "step": 60580 }, { "grad_norm": 0.4606516361236572, "learning_rate": 3.678402078440649e-05, "loss": 0.0434, "step": 60590 }, { "grad_norm": 0.46977826952934265, "learning_rate": 3.676807486323096e-05, "loss": 0.0405, "step": 60600 }, { "grad_norm": 0.6325879096984863, "learning_rate": 3.6752130389078864e-05, "loss": 0.0385, "step": 60610 }, { "grad_norm": 0.37139275670051575, "learning_rate": 3.673618736369388e-05, "loss": 0.0338, "step": 60620 }, { "grad_norm": 0.4021625816822052, "learning_rate": 3.6720245788819474e-05, "loss": 0.0366, "step": 60630 }, { "grad_norm": 0.5632616281509399, "learning_rate": 3.6704305666199044e-05, "loss": 0.0484, "step": 60640 }, { "grad_norm": 0.8381553292274475, "learning_rate": 3.668836699757576e-05, "loss": 0.0381, "step": 60650 }, { "grad_norm": 0.678442656993866, "learning_rate": 3.667242978469265e-05, "loss": 0.0453, "step": 60660 }, { "grad_norm": 0.6338963508605957, "learning_rate": 3.6656494029292575e-05, "loss": 0.0502, "step": 60670 }, { "grad_norm": 0.4237475097179413, "learning_rate": 3.664055973311825e-05, "loss": 0.0403, "step": 60680 }, { "grad_norm": 0.3971220850944519, "learning_rate": 3.662462689791221e-05, "loss": 0.0348, "step": 60690 }, { "grad_norm": 0.29949066042900085, "learning_rate": 3.660869552541689e-05, "loss": 0.0309, "step": 60700 }, { "grad_norm": 0.6033753156661987, "learning_rate": 3.659276561737448e-05, "loss": 0.0388, "step": 60710 }, { "grad_norm": 0.49982574582099915, "learning_rate": 3.657683717552708e-05, "loss": 0.0366, "step": 60720 }, { "grad_norm": 0.5253174304962158, "learning_rate": 3.6560910201616574e-05, "loss": 0.0307, "step": 60730 }, { "grad_norm": 0.29629313945770264, "learning_rate": 3.654498469738472e-05, "loss": 0.0409, "step": 60740 }, { "grad_norm": 0.5030363202095032, "learning_rate": 3.6529060664573126e-05, "loss": 0.0493, "step": 60750 }, { "grad_norm": 0.3519474267959595, "learning_rate": 3.6513138104923176e-05, "loss": 0.0391, "step": 60760 }, { "grad_norm": 0.4434361457824707, "learning_rate": 3.6497217020176176e-05, "loss": 0.0366, "step": 60770 }, { "grad_norm": 0.47608649730682373, "learning_rate": 3.648129741207323e-05, "loss": 0.0422, "step": 60780 }, { "grad_norm": 0.3504972457885742, "learning_rate": 3.646537928235527e-05, "loss": 0.033, "step": 60790 }, { "grad_norm": 0.3842223286628723, "learning_rate": 3.6449462632763075e-05, "loss": 0.0377, "step": 60800 }, { "grad_norm": 0.4485498368740082, "learning_rate": 3.643354746503729e-05, "loss": 0.0333, "step": 60810 }, { "grad_norm": 0.5743110775947571, "learning_rate": 3.6417633780918345e-05, "loss": 0.0325, "step": 60820 }, { "grad_norm": 0.6600126028060913, "learning_rate": 3.640172158214655e-05, "loss": 0.0457, "step": 60830 }, { "grad_norm": 0.7139086723327637, "learning_rate": 3.6385810870462035e-05, "loss": 0.0321, "step": 60840 }, { "grad_norm": 0.5624899864196777, "learning_rate": 3.636990164760479e-05, "loss": 0.0349, "step": 60850 }, { "grad_norm": 0.5184348225593567, "learning_rate": 3.6353993915314596e-05, "loss": 0.0381, "step": 60860 }, { "grad_norm": 0.5716805458068848, "learning_rate": 3.6338087675331124e-05, "loss": 0.0316, "step": 60870 }, { "grad_norm": 0.3845066726207733, "learning_rate": 3.632218292939383e-05, "loss": 0.0295, "step": 60880 }, { "grad_norm": 0.4857446253299713, "learning_rate": 3.6306279679242033e-05, "loss": 0.0338, "step": 60890 }, { "grad_norm": 0.3548418879508972, "learning_rate": 3.629037792661492e-05, "loss": 0.0413, "step": 60900 }, { "grad_norm": 0.4130878150463104, "learning_rate": 3.627447767325147e-05, "loss": 0.0406, "step": 60910 }, { "grad_norm": 0.8693224787712097, "learning_rate": 3.62585789208905e-05, "loss": 0.0353, "step": 60920 }, { "grad_norm": 0.3775380253791809, "learning_rate": 3.624268167127068e-05, "loss": 0.0588, "step": 60930 }, { "grad_norm": 0.3155551552772522, "learning_rate": 3.622678592613051e-05, "loss": 0.0377, "step": 60940 }, { "grad_norm": 0.6535478830337524, "learning_rate": 3.6210891687208304e-05, "loss": 0.0431, "step": 60950 }, { "grad_norm": 0.7644729018211365, "learning_rate": 3.619499895624228e-05, "loss": 0.0365, "step": 60960 }, { "grad_norm": 0.48533761501312256, "learning_rate": 3.61791077349704e-05, "loss": 0.0384, "step": 60970 }, { "grad_norm": 0.4874257743358612, "learning_rate": 3.616321802513053e-05, "loss": 0.0352, "step": 60980 }, { "grad_norm": 0.5494328737258911, "learning_rate": 3.6147329828460325e-05, "loss": 0.0487, "step": 60990 }, { "grad_norm": 0.4460183382034302, "learning_rate": 3.6131443146697306e-05, "loss": 0.0352, "step": 61000 }, { "grad_norm": 0.40698155760765076, "learning_rate": 3.611555798157879e-05, "loss": 0.035, "step": 61010 }, { "grad_norm": 0.5623607039451599, "learning_rate": 3.6099674334842e-05, "loss": 0.0407, "step": 61020 }, { "grad_norm": 0.34470298886299133, "learning_rate": 3.6083792208223934e-05, "loss": 0.0471, "step": 61030 }, { "grad_norm": 0.39068830013275146, "learning_rate": 3.606791160346142e-05, "loss": 0.0426, "step": 61040 }, { "grad_norm": 0.33863765001296997, "learning_rate": 3.6052032522291134e-05, "loss": 0.0413, "step": 61050 }, { "grad_norm": 0.6397621631622314, "learning_rate": 3.603615496644962e-05, "loss": 0.0526, "step": 61060 }, { "grad_norm": 0.5351033210754395, "learning_rate": 3.60202789376732e-05, "loss": 0.0412, "step": 61070 }, { "grad_norm": 0.4298609793186188, "learning_rate": 3.6004404437698036e-05, "loss": 0.0396, "step": 61080 }, { "grad_norm": 0.4681175947189331, "learning_rate": 3.5988531468260184e-05, "loss": 0.0402, "step": 61090 }, { "grad_norm": 0.48093181848526, "learning_rate": 3.597266003109546e-05, "loss": 0.0345, "step": 61100 }, { "grad_norm": 0.3657045066356659, "learning_rate": 3.5956790127939557e-05, "loss": 0.0356, "step": 61110 }, { "grad_norm": 0.41837677359580994, "learning_rate": 3.594092176052796e-05, "loss": 0.0402, "step": 61120 }, { "grad_norm": 0.3577243983745575, "learning_rate": 3.592505493059603e-05, "loss": 0.0375, "step": 61130 }, { "grad_norm": 0.44411131739616394, "learning_rate": 3.5909189639878905e-05, "loss": 0.0429, "step": 61140 }, { "grad_norm": 0.2861658036708832, "learning_rate": 3.5893325890111626e-05, "loss": 0.0343, "step": 61150 }, { "grad_norm": 0.4583679139614105, "learning_rate": 3.5877463683029034e-05, "loss": 0.0486, "step": 61160 }, { "grad_norm": 0.24609845876693726, "learning_rate": 3.586160302036578e-05, "loss": 0.0383, "step": 61170 }, { "grad_norm": 0.3242384195327759, "learning_rate": 3.584574390385636e-05, "loss": 0.0351, "step": 61180 }, { "grad_norm": 0.29832321405410767, "learning_rate": 3.582988633523511e-05, "loss": 0.046, "step": 61190 }, { "grad_norm": 0.3453856110572815, "learning_rate": 3.581403031623616e-05, "loss": 0.0438, "step": 61200 }, { "grad_norm": 0.4495278298854828, "learning_rate": 3.579817584859354e-05, "loss": 0.0411, "step": 61210 }, { "grad_norm": 0.43104371428489685, "learning_rate": 3.578232293404107e-05, "loss": 0.0426, "step": 61220 }, { "grad_norm": 0.5745663046836853, "learning_rate": 3.5766471574312385e-05, "loss": 0.0338, "step": 61230 }, { "grad_norm": 0.4597088098526001, "learning_rate": 3.5750621771140964e-05, "loss": 0.0468, "step": 61240 }, { "grad_norm": 0.3935905694961548, "learning_rate": 3.5734773526260115e-05, "loss": 0.0356, "step": 61250 }, { "grad_norm": 0.39462152123451233, "learning_rate": 3.5718926841402996e-05, "loss": 0.0382, "step": 61260 }, { "grad_norm": 0.38761064410209656, "learning_rate": 3.5703081718302555e-05, "loss": 0.0418, "step": 61270 }, { "grad_norm": 0.44027772545814514, "learning_rate": 3.568723815869162e-05, "loss": 0.0397, "step": 61280 }, { "grad_norm": 0.3577360510826111, "learning_rate": 3.5671396164302777e-05, "loss": 0.0604, "step": 61290 }, { "grad_norm": 0.5771890878677368, "learning_rate": 3.565555573686853e-05, "loss": 0.0302, "step": 61300 }, { "grad_norm": 0.39983561635017395, "learning_rate": 3.563971687812113e-05, "loss": 0.0345, "step": 61310 }, { "grad_norm": 0.4801652133464813, "learning_rate": 3.5623879589792706e-05, "loss": 0.0362, "step": 61320 }, { "grad_norm": 0.2720334827899933, "learning_rate": 3.5608043873615185e-05, "loss": 0.046, "step": 61330 }, { "grad_norm": 0.3318169116973877, "learning_rate": 3.559220973132035e-05, "loss": 0.0314, "step": 61340 }, { "grad_norm": 0.7199392318725586, "learning_rate": 3.557637716463981e-05, "loss": 0.0419, "step": 61350 }, { "grad_norm": 0.5999705791473389, "learning_rate": 3.556054617530497e-05, "loss": 0.0404, "step": 61360 }, { "grad_norm": 0.36495548486709595, "learning_rate": 3.5544716765047106e-05, "loss": 0.0355, "step": 61370 }, { "grad_norm": 0.4340447187423706, "learning_rate": 3.5528888935597274e-05, "loss": 0.0383, "step": 61380 }, { "grad_norm": 0.42099830508232117, "learning_rate": 3.551306268868636e-05, "loss": 0.0379, "step": 61390 }, { "grad_norm": 0.29554420709609985, "learning_rate": 3.5497238026045166e-05, "loss": 0.0366, "step": 61400 }, { "grad_norm": 0.40167492628097534, "learning_rate": 3.548141494940421e-05, "loss": 0.0428, "step": 61410 }, { "grad_norm": 0.5331840515136719, "learning_rate": 3.546559346049388e-05, "loss": 0.0423, "step": 61420 }, { "grad_norm": 0.4540238678455353, "learning_rate": 3.544977356104441e-05, "loss": 0.0379, "step": 61430 }, { "grad_norm": 0.41831207275390625, "learning_rate": 3.543395525278582e-05, "loss": 0.0409, "step": 61440 }, { "grad_norm": 0.47336676716804504, "learning_rate": 3.5418138537447985e-05, "loss": 0.0402, "step": 61450 }, { "grad_norm": 0.3025481402873993, "learning_rate": 3.5402323416760574e-05, "loss": 0.032, "step": 61460 }, { "grad_norm": 0.29052215814590454, "learning_rate": 3.538650989245316e-05, "loss": 0.0346, "step": 61470 }, { "grad_norm": 0.5457688570022583, "learning_rate": 3.537069796625504e-05, "loss": 0.0299, "step": 61480 }, { "grad_norm": 0.5370224118232727, "learning_rate": 3.53548876398954e-05, "loss": 0.0413, "step": 61490 }, { "grad_norm": 0.5879839658737183, "learning_rate": 3.5339078915103214e-05, "loss": 0.0378, "step": 61500 }, { "grad_norm": 0.44503962993621826, "learning_rate": 3.532327179360733e-05, "loss": 0.0478, "step": 61510 }, { "grad_norm": 0.704454779624939, "learning_rate": 3.530746627713636e-05, "loss": 0.0356, "step": 61520 }, { "grad_norm": 0.4300774335861206, "learning_rate": 3.5291662367418805e-05, "loss": 0.037, "step": 61530 }, { "grad_norm": 0.47822967171669006, "learning_rate": 3.527586006618293e-05, "loss": 0.0569, "step": 61540 }, { "grad_norm": 1.7245901823043823, "learning_rate": 3.526005937515687e-05, "loss": 0.0576, "step": 61550 }, { "grad_norm": 0.3427745997905731, "learning_rate": 3.524426029606856e-05, "loss": 0.0536, "step": 61560 }, { "grad_norm": 0.32719990611076355, "learning_rate": 3.5228462830645746e-05, "loss": 0.0367, "step": 61570 }, { "grad_norm": 0.5689682960510254, "learning_rate": 3.5212666980616036e-05, "loss": 0.0376, "step": 61580 }, { "grad_norm": 0.4274023473262787, "learning_rate": 3.519687274770681e-05, "loss": 0.0454, "step": 61590 }, { "grad_norm": 0.35767778754234314, "learning_rate": 3.518108013364535e-05, "loss": 0.0512, "step": 61600 }, { "grad_norm": 0.4861077070236206, "learning_rate": 3.516528914015868e-05, "loss": 0.0421, "step": 61610 }, { "grad_norm": 0.5114796161651611, "learning_rate": 3.5149499768973694e-05, "loss": 0.0537, "step": 61620 }, { "grad_norm": 0.2781449556350708, "learning_rate": 3.513371202181708e-05, "loss": 0.0378, "step": 61630 }, { "grad_norm": 0.5991443991661072, "learning_rate": 3.511792590041537e-05, "loss": 0.0393, "step": 61640 }, { "grad_norm": 0.6183175444602966, "learning_rate": 3.510214140649488e-05, "loss": 0.0465, "step": 61650 }, { "grad_norm": 0.38040271401405334, "learning_rate": 3.508635854178184e-05, "loss": 0.0458, "step": 61660 }, { "grad_norm": 0.5634583830833435, "learning_rate": 3.50705773080022e-05, "loss": 0.0491, "step": 61670 }, { "grad_norm": 0.43790262937545776, "learning_rate": 3.505479770688178e-05, "loss": 0.0417, "step": 61680 }, { "grad_norm": 0.35846146941185, "learning_rate": 3.503901974014622e-05, "loss": 0.0358, "step": 61690 }, { "grad_norm": 0.5791066288948059, "learning_rate": 3.502324340952095e-05, "loss": 0.0427, "step": 61700 }, { "grad_norm": 0.3874073922634125, "learning_rate": 3.500746871673126e-05, "loss": 0.0525, "step": 61710 }, { "grad_norm": 0.35594701766967773, "learning_rate": 3.499169566350227e-05, "loss": 0.0543, "step": 61720 }, { "grad_norm": 0.34631723165512085, "learning_rate": 3.4975924251558876e-05, "loss": 0.0316, "step": 61730 }, { "grad_norm": 0.3269028961658478, "learning_rate": 3.496015448262582e-05, "loss": 0.0396, "step": 61740 }, { "grad_norm": 0.5773088932037354, "learning_rate": 3.494438635842764e-05, "loss": 0.0386, "step": 61750 }, { "grad_norm": 0.30596494674682617, "learning_rate": 3.492861988068874e-05, "loss": 0.0357, "step": 61760 }, { "grad_norm": 0.303986132144928, "learning_rate": 3.491285505113332e-05, "loss": 0.0511, "step": 61770 }, { "grad_norm": 0.36209896206855774, "learning_rate": 3.4897091871485376e-05, "loss": 0.051, "step": 61780 }, { "grad_norm": 0.35472413897514343, "learning_rate": 3.488133034346875e-05, "loss": 0.0368, "step": 61790 }, { "grad_norm": 0.6810617446899414, "learning_rate": 3.486557046880713e-05, "loss": 0.0337, "step": 61800 }, { "grad_norm": 0.39574795961380005, "learning_rate": 3.484981224922395e-05, "loss": 0.0468, "step": 61810 }, { "grad_norm": 0.35346493124961853, "learning_rate": 3.4834055686442544e-05, "loss": 0.0274, "step": 61820 }, { "grad_norm": 0.6780319809913635, "learning_rate": 3.4818300782186e-05, "loss": 0.0426, "step": 61830 }, { "grad_norm": 0.4858217239379883, "learning_rate": 3.4802547538177236e-05, "loss": 0.0348, "step": 61840 }, { "grad_norm": 0.4750465452671051, "learning_rate": 3.478679595613905e-05, "loss": 0.034, "step": 61850 }, { "grad_norm": 0.3175145387649536, "learning_rate": 3.4771046037793975e-05, "loss": 0.0356, "step": 61860 }, { "grad_norm": 0.3475787937641144, "learning_rate": 3.4755297784864416e-05, "loss": 0.0323, "step": 61870 }, { "grad_norm": 0.5404379963874817, "learning_rate": 3.473955119907257e-05, "loss": 0.0371, "step": 61880 }, { "grad_norm": 0.4474603533744812, "learning_rate": 3.472380628214046e-05, "loss": 0.0374, "step": 61890 }, { "grad_norm": 0.42227742075920105, "learning_rate": 3.4708063035789906e-05, "loss": 0.0369, "step": 61900 }, { "grad_norm": 0.44567060470581055, "learning_rate": 3.4692321461742614e-05, "loss": 0.0387, "step": 61910 }, { "grad_norm": 0.4124995768070221, "learning_rate": 3.4676581561720023e-05, "loss": 0.0352, "step": 61920 }, { "grad_norm": 0.34305813908576965, "learning_rate": 3.466084333744344e-05, "loss": 0.032, "step": 61930 }, { "grad_norm": 0.4891577959060669, "learning_rate": 3.4645106790633964e-05, "loss": 0.045, "step": 61940 }, { "grad_norm": 0.6201066374778748, "learning_rate": 3.4629371923012526e-05, "loss": 0.0486, "step": 61950 }, { "grad_norm": 0.43962711095809937, "learning_rate": 3.4613638736299856e-05, "loss": 0.0426, "step": 61960 }, { "grad_norm": 0.30685487389564514, "learning_rate": 3.459790723221652e-05, "loss": 0.0241, "step": 61970 }, { "grad_norm": 0.6994437575340271, "learning_rate": 3.45821774124829e-05, "loss": 0.0451, "step": 61980 }, { "grad_norm": 0.5144402980804443, "learning_rate": 3.456644927881918e-05, "loss": 0.037, "step": 61990 }, { "grad_norm": 0.3588210940361023, "learning_rate": 3.4550722832945354e-05, "loss": 0.0414, "step": 62000 }, { "grad_norm": 0.5890398025512695, "learning_rate": 3.453499807658126e-05, "loss": 0.0371, "step": 62010 }, { "grad_norm": 0.3377654552459717, "learning_rate": 3.451927501144653e-05, "loss": 0.0298, "step": 62020 }, { "grad_norm": 0.4690794050693512, "learning_rate": 3.450355363926059e-05, "loss": 0.0534, "step": 62030 }, { "grad_norm": 0.5734063386917114, "learning_rate": 3.448783396174273e-05, "loss": 0.0484, "step": 62040 }, { "grad_norm": 0.5581666231155396, "learning_rate": 3.4472115980612034e-05, "loss": 0.038, "step": 62050 }, { "grad_norm": 0.5343737602233887, "learning_rate": 3.4456399697587396e-05, "loss": 0.0396, "step": 62060 }, { "grad_norm": 0.5803978443145752, "learning_rate": 3.4440685114387506e-05, "loss": 0.0472, "step": 62070 }, { "grad_norm": 0.49427324533462524, "learning_rate": 3.442497223273091e-05, "loss": 0.0364, "step": 62080 }, { "grad_norm": 0.8267617225646973, "learning_rate": 3.4409261054335926e-05, "loss": 0.0486, "step": 62090 }, { "grad_norm": 0.42213037610054016, "learning_rate": 3.439355158092069e-05, "loss": 0.0488, "step": 62100 }, { "grad_norm": 0.42022985219955444, "learning_rate": 3.437784381420322e-05, "loss": 0.044, "step": 62110 }, { "grad_norm": 0.32372617721557617, "learning_rate": 3.436213775590126e-05, "loss": 0.0329, "step": 62120 }, { "grad_norm": 0.43139955401420593, "learning_rate": 3.4346433407732405e-05, "loss": 0.0383, "step": 62130 }, { "grad_norm": 0.5574665665626526, "learning_rate": 3.433073077141405e-05, "loss": 0.0256, "step": 62140 }, { "grad_norm": 0.9036703109741211, "learning_rate": 3.431502984866343e-05, "loss": 0.0423, "step": 62150 }, { "grad_norm": 0.4267119765281677, "learning_rate": 3.429933064119753e-05, "loss": 0.057, "step": 62160 }, { "grad_norm": 0.8315793871879578, "learning_rate": 3.4283633150733255e-05, "loss": 0.0383, "step": 62170 }, { "grad_norm": 0.47734516859054565, "learning_rate": 3.4267937378987225e-05, "loss": 0.0453, "step": 62180 }, { "grad_norm": 0.41822341084480286, "learning_rate": 3.425224332767592e-05, "loss": 0.0324, "step": 62190 }, { "grad_norm": 0.6462672352790833, "learning_rate": 3.4236550998515595e-05, "loss": 0.0506, "step": 62200 }, { "grad_norm": 0.5219871997833252, "learning_rate": 3.422086039322235e-05, "loss": 0.0374, "step": 62210 }, { "grad_norm": 0.45412909984588623, "learning_rate": 3.420517151351209e-05, "loss": 0.0409, "step": 62220 }, { "grad_norm": 0.3456297814846039, "learning_rate": 3.418948436110053e-05, "loss": 0.0423, "step": 62230 }, { "grad_norm": 0.3529307246208191, "learning_rate": 3.417379893770319e-05, "loss": 0.0401, "step": 62240 }, { "grad_norm": 0.5402876734733582, "learning_rate": 3.41581152450354e-05, "loss": 0.0319, "step": 62250 }, { "grad_norm": 0.32081857323646545, "learning_rate": 3.414243328481231e-05, "loss": 0.0384, "step": 62260 }, { "grad_norm": 0.3779853880405426, "learning_rate": 3.412675305874888e-05, "loss": 0.0428, "step": 62270 }, { "grad_norm": 0.6666389107704163, "learning_rate": 3.4111074568559876e-05, "loss": 0.0385, "step": 62280 }, { "grad_norm": 0.5214768648147583, "learning_rate": 3.4095397815959846e-05, "loss": 0.0469, "step": 62290 }, { "grad_norm": 0.6194337010383606, "learning_rate": 3.407972280266322e-05, "loss": 0.0408, "step": 62300 }, { "grad_norm": 0.41123199462890625, "learning_rate": 3.4064049530384166e-05, "loss": 0.0399, "step": 62310 }, { "grad_norm": 1.410632848739624, "learning_rate": 3.40483780008367e-05, "loss": 0.0365, "step": 62320 }, { "grad_norm": 0.41981983184814453, "learning_rate": 3.4032708215734635e-05, "loss": 0.0379, "step": 62330 }, { "grad_norm": 0.5020826458930969, "learning_rate": 3.4017040176791584e-05, "loss": 0.0403, "step": 62340 }, { "grad_norm": 0.49603715538978577, "learning_rate": 3.4001373885720976e-05, "loss": 0.0504, "step": 62350 }, { "grad_norm": 0.3487834632396698, "learning_rate": 3.398570934423608e-05, "loss": 0.0391, "step": 62360 }, { "grad_norm": 0.4156249463558197, "learning_rate": 3.397004655404994e-05, "loss": 0.0522, "step": 62370 }, { "grad_norm": 0.3716617226600647, "learning_rate": 3.395438551687541e-05, "loss": 0.0417, "step": 62380 }, { "grad_norm": 0.5752809643745422, "learning_rate": 3.393872623442515e-05, "loss": 0.0389, "step": 62390 }, { "grad_norm": 0.5155162811279297, "learning_rate": 3.392306870841164e-05, "loss": 0.0421, "step": 62400 }, { "grad_norm": 0.4690198302268982, "learning_rate": 3.390741294054715e-05, "loss": 0.0379, "step": 62410 }, { "grad_norm": 0.549967885017395, "learning_rate": 3.38917589325438e-05, "loss": 0.035, "step": 62420 }, { "grad_norm": 0.5485597848892212, "learning_rate": 3.387610668611349e-05, "loss": 0.0473, "step": 62430 }, { "grad_norm": 0.6636345386505127, "learning_rate": 3.38604562029679e-05, "loss": 0.046, "step": 62440 }, { "grad_norm": 0.2989060878753662, "learning_rate": 3.384480748481855e-05, "loss": 0.038, "step": 62450 }, { "grad_norm": 0.3742018938064575, "learning_rate": 3.382916053337676e-05, "loss": 0.0393, "step": 62460 }, { "grad_norm": 0.3199940621852875, "learning_rate": 3.381351535035366e-05, "loss": 0.0489, "step": 62470 }, { "grad_norm": 0.48271673917770386, "learning_rate": 3.379787193746019e-05, "loss": 0.0411, "step": 62480 }, { "grad_norm": 0.363344669342041, "learning_rate": 3.378223029640709e-05, "loss": 0.0357, "step": 62490 }, { "grad_norm": 0.39625054597854614, "learning_rate": 3.376659042890489e-05, "loss": 0.044, "step": 62500 }, { "grad_norm": 0.39014679193496704, "learning_rate": 3.375095233666397e-05, "loss": 0.0391, "step": 62510 }, { "grad_norm": 0.31938406825065613, "learning_rate": 3.373531602139447e-05, "loss": 0.0368, "step": 62520 }, { "grad_norm": 0.44884440302848816, "learning_rate": 3.371968148480635e-05, "loss": 0.0353, "step": 62530 }, { "grad_norm": 0.8604061007499695, "learning_rate": 3.370404872860937e-05, "loss": 0.0396, "step": 62540 }, { "grad_norm": 0.35317322611808777, "learning_rate": 3.3688417754513145e-05, "loss": 0.0392, "step": 62550 }, { "grad_norm": 0.394390344619751, "learning_rate": 3.367278856422703e-05, "loss": 0.0381, "step": 62560 }, { "grad_norm": 0.859866738319397, "learning_rate": 3.365716115946021e-05, "loss": 0.0533, "step": 62570 }, { "grad_norm": 0.6603788733482361, "learning_rate": 3.3641535541921666e-05, "loss": 0.0593, "step": 62580 }, { "grad_norm": 0.6879206895828247, "learning_rate": 3.36259117133202e-05, "loss": 0.041, "step": 62590 }, { "grad_norm": 0.4910539388656616, "learning_rate": 3.36102896753644e-05, "loss": 0.0302, "step": 62600 }, { "grad_norm": 0.5250205397605896, "learning_rate": 3.35946694297627e-05, "loss": 0.0311, "step": 62610 }, { "grad_norm": 0.2914700210094452, "learning_rate": 3.357905097822328e-05, "loss": 0.0418, "step": 62620 }, { "grad_norm": 0.7699577808380127, "learning_rate": 3.356343432245416e-05, "loss": 0.0392, "step": 62630 }, { "grad_norm": 0.6731919050216675, "learning_rate": 3.354781946416315e-05, "loss": 0.0578, "step": 62640 }, { "grad_norm": 0.4881246089935303, "learning_rate": 3.3532206405057864e-05, "loss": 0.0456, "step": 62650 }, { "grad_norm": 0.6440869569778442, "learning_rate": 3.351659514684573e-05, "loss": 0.0386, "step": 62660 }, { "grad_norm": 1.0569941997528076, "learning_rate": 3.350098569123395e-05, "loss": 0.0362, "step": 62670 }, { "grad_norm": 0.3875059187412262, "learning_rate": 3.348537803992958e-05, "loss": 0.0351, "step": 62680 }, { "grad_norm": 0.3147045075893402, "learning_rate": 3.346977219463945e-05, "loss": 0.0372, "step": 62690 }, { "grad_norm": 0.4492529630661011, "learning_rate": 3.345416815707017e-05, "loss": 0.0326, "step": 62700 }, { "grad_norm": 0.707522988319397, "learning_rate": 3.3438565928928185e-05, "loss": 0.0409, "step": 62710 }, { "grad_norm": 0.5547756552696228, "learning_rate": 3.3422965511919736e-05, "loss": 0.0359, "step": 62720 }, { "grad_norm": 0.3925575017929077, "learning_rate": 3.3407366907750846e-05, "loss": 0.0344, "step": 62730 }, { "grad_norm": 0.5905942916870117, "learning_rate": 3.339177011812737e-05, "loss": 0.0492, "step": 62740 }, { "grad_norm": 0.4138709008693695, "learning_rate": 3.3376175144754935e-05, "loss": 0.0326, "step": 62750 }, { "grad_norm": 0.443746954202652, "learning_rate": 3.3360581989339004e-05, "loss": 0.0402, "step": 62760 }, { "grad_norm": 0.23009715974330902, "learning_rate": 3.3344990653584804e-05, "loss": 0.039, "step": 62770 }, { "grad_norm": 0.36365124583244324, "learning_rate": 3.332940113919739e-05, "loss": 0.0448, "step": 62780 }, { "grad_norm": 0.4444813132286072, "learning_rate": 3.331381344788161e-05, "loss": 0.0345, "step": 62790 }, { "grad_norm": 0.442907452583313, "learning_rate": 3.3298227581342076e-05, "loss": 0.0403, "step": 62800 }, { "grad_norm": 0.6205840706825256, "learning_rate": 3.328264354128329e-05, "loss": 0.0511, "step": 62810 }, { "grad_norm": 0.458831787109375, "learning_rate": 3.3267061329409465e-05, "loss": 0.0462, "step": 62820 }, { "grad_norm": 0.3671416938304901, "learning_rate": 3.325148094742466e-05, "loss": 0.0429, "step": 62830 }, { "grad_norm": 0.7582951784133911, "learning_rate": 3.323590239703271e-05, "loss": 0.0382, "step": 62840 }, { "grad_norm": 0.3460932672023773, "learning_rate": 3.322032567993728e-05, "loss": 0.0358, "step": 62850 }, { "grad_norm": 0.5010699033737183, "learning_rate": 3.3204750797841774e-05, "loss": 0.0379, "step": 62860 }, { "grad_norm": 0.3272773325443268, "learning_rate": 3.31891777524495e-05, "loss": 0.0362, "step": 62870 }, { "grad_norm": 1.3359204530715942, "learning_rate": 3.317360654546347e-05, "loss": 0.0431, "step": 62880 }, { "grad_norm": 0.37617921829223633, "learning_rate": 3.3158037178586534e-05, "loss": 0.0489, "step": 62890 }, { "grad_norm": 1.6023647785186768, "learning_rate": 3.3142469653521316e-05, "loss": 0.0388, "step": 62900 }, { "grad_norm": 0.8865647912025452, "learning_rate": 3.3126903971970284e-05, "loss": 0.0483, "step": 62910 }, { "grad_norm": 1.3669577836990356, "learning_rate": 3.311134013563564e-05, "loss": 0.0401, "step": 62920 }, { "grad_norm": 0.5735477209091187, "learning_rate": 3.309577814621947e-05, "loss": 0.0349, "step": 62930 }, { "grad_norm": 0.36090853810310364, "learning_rate": 3.308021800542359e-05, "loss": 0.0346, "step": 62940 }, { "grad_norm": 0.45197004079818726, "learning_rate": 3.306465971494963e-05, "loss": 0.0379, "step": 62950 }, { "grad_norm": 0.7110024690628052, "learning_rate": 3.304910327649901e-05, "loss": 0.0414, "step": 62960 }, { "grad_norm": 0.2595452070236206, "learning_rate": 3.303354869177297e-05, "loss": 0.0342, "step": 62970 }, { "grad_norm": 0.42763224244117737, "learning_rate": 3.301799596247255e-05, "loss": 0.036, "step": 62980 }, { "grad_norm": 0.3293132781982422, "learning_rate": 3.300244509029854e-05, "loss": 0.0342, "step": 62990 }, { "grad_norm": 0.4364747703075409, "learning_rate": 3.2986896076951585e-05, "loss": 0.035, "step": 63000 }, { "grad_norm": 0.3932030200958252, "learning_rate": 3.297134892413211e-05, "loss": 0.0474, "step": 63010 }, { "grad_norm": 0.4604928195476532, "learning_rate": 3.295580363354031e-05, "loss": 0.0462, "step": 63020 }, { "grad_norm": 0.5148892998695374, "learning_rate": 3.2940260206876194e-05, "loss": 0.0346, "step": 63030 }, { "grad_norm": 0.49739235639572144, "learning_rate": 3.292471864583958e-05, "loss": 0.0393, "step": 63040 }, { "grad_norm": 0.3387305736541748, "learning_rate": 3.290917895213004e-05, "loss": 0.0368, "step": 63050 }, { "grad_norm": 0.46358099579811096, "learning_rate": 3.289364112744701e-05, "loss": 0.0415, "step": 63060 }, { "grad_norm": 0.2749583423137665, "learning_rate": 3.287810517348967e-05, "loss": 0.0405, "step": 63070 }, { "grad_norm": 0.407722532749176, "learning_rate": 3.2862571091957014e-05, "loss": 0.0347, "step": 63080 }, { "grad_norm": 0.31576940417289734, "learning_rate": 3.284703888454781e-05, "loss": 0.0351, "step": 63090 }, { "grad_norm": 0.5500407218933105, "learning_rate": 3.2831508552960646e-05, "loss": 0.0266, "step": 63100 }, { "grad_norm": 0.5513317584991455, "learning_rate": 3.281598009889387e-05, "loss": 0.0409, "step": 63110 }, { "grad_norm": 0.46281105279922485, "learning_rate": 3.28004535240457e-05, "loss": 0.0492, "step": 63120 }, { "grad_norm": 0.42448922991752625, "learning_rate": 3.278492883011407e-05, "loss": 0.04, "step": 63130 }, { "grad_norm": 0.25239986181259155, "learning_rate": 3.276940601879675e-05, "loss": 0.0424, "step": 63140 }, { "grad_norm": 0.6004579663276672, "learning_rate": 3.275388509179127e-05, "loss": 0.0492, "step": 63150 }, { "grad_norm": 0.4641512930393219, "learning_rate": 3.2738366050794985e-05, "loss": 0.0495, "step": 63160 }, { "grad_norm": 0.3933624029159546, "learning_rate": 3.2722848897505034e-05, "loss": 0.0361, "step": 63170 }, { "grad_norm": 0.4573124051094055, "learning_rate": 3.2707333633618356e-05, "loss": 0.0344, "step": 63180 }, { "grad_norm": 0.5001767873764038, "learning_rate": 3.269182026083167e-05, "loss": 0.036, "step": 63190 }, { "grad_norm": 0.33489179611206055, "learning_rate": 3.2676308780841516e-05, "loss": 0.0416, "step": 63200 }, { "grad_norm": 0.32752254605293274, "learning_rate": 3.266079919534416e-05, "loss": 0.0372, "step": 63210 }, { "grad_norm": 0.3615362346172333, "learning_rate": 3.264529150603575e-05, "loss": 0.0339, "step": 63220 }, { "grad_norm": 0.5492894649505615, "learning_rate": 3.262978571461217e-05, "loss": 0.0351, "step": 63230 }, { "grad_norm": 0.5382152199745178, "learning_rate": 3.26142818227691e-05, "loss": 0.0482, "step": 63240 }, { "grad_norm": 0.6032724380493164, "learning_rate": 3.259877983220203e-05, "loss": 0.0447, "step": 63250 }, { "grad_norm": 0.6075158715248108, "learning_rate": 3.2583279744606246e-05, "loss": 0.0416, "step": 63260 }, { "grad_norm": 0.30751898884773254, "learning_rate": 3.25677815616768e-05, "loss": 0.0265, "step": 63270 }, { "grad_norm": 1.0582482814788818, "learning_rate": 3.255228528510856e-05, "loss": 0.0494, "step": 63280 }, { "grad_norm": 0.4748672544956207, "learning_rate": 3.253679091659617e-05, "loss": 0.0344, "step": 63290 }, { "grad_norm": 0.5571048855781555, "learning_rate": 3.252129845783404e-05, "loss": 0.0396, "step": 63300 }, { "grad_norm": 0.3225969970226288, "learning_rate": 3.2505807910516456e-05, "loss": 0.0309, "step": 63310 }, { "grad_norm": 0.6332171559333801, "learning_rate": 3.249031927633742e-05, "loss": 0.05, "step": 63320 }, { "grad_norm": 0.3700563907623291, "learning_rate": 3.2474832556990737e-05, "loss": 0.034, "step": 63330 }, { "grad_norm": 0.38576388359069824, "learning_rate": 3.245934775417002e-05, "loss": 0.0408, "step": 63340 }, { "grad_norm": 0.9344939589500427, "learning_rate": 3.244386486956866e-05, "loss": 0.0426, "step": 63350 }, { "grad_norm": 0.6430748701095581, "learning_rate": 3.2428383904879846e-05, "loss": 0.0436, "step": 63360 }, { "grad_norm": 0.42047491669654846, "learning_rate": 3.241290486179652e-05, "loss": 0.0557, "step": 63370 }, { "grad_norm": 0.3221622109413147, "learning_rate": 3.239742774201151e-05, "loss": 0.0321, "step": 63380 }, { "grad_norm": 0.298469215631485, "learning_rate": 3.238195254721733e-05, "loss": 0.0328, "step": 63390 }, { "grad_norm": 0.510536789894104, "learning_rate": 3.236647927910633e-05, "loss": 0.0434, "step": 63400 }, { "grad_norm": 0.25994205474853516, "learning_rate": 3.2351007939370646e-05, "loss": 0.035, "step": 63410 }, { "grad_norm": 0.6108907461166382, "learning_rate": 3.2335538529702184e-05, "loss": 0.0423, "step": 63420 }, { "grad_norm": 0.5254331231117249, "learning_rate": 3.232007105179268e-05, "loss": 0.042, "step": 63430 }, { "grad_norm": 0.4957745671272278, "learning_rate": 3.230460550733363e-05, "loss": 0.0295, "step": 63440 }, { "grad_norm": 0.43342217803001404, "learning_rate": 3.2289141898016315e-05, "loss": 0.0336, "step": 63450 }, { "grad_norm": 0.32966411113739014, "learning_rate": 3.227368022553181e-05, "loss": 0.0411, "step": 63460 }, { "grad_norm": 0.3752328157424927, "learning_rate": 3.225822049157099e-05, "loss": 0.0334, "step": 63470 }, { "grad_norm": 0.4403052031993866, "learning_rate": 3.2242762697824494e-05, "loss": 0.0288, "step": 63480 }, { "grad_norm": 1.1369937658309937, "learning_rate": 3.222730684598279e-05, "loss": 0.0427, "step": 63490 }, { "grad_norm": 0.43038710951805115, "learning_rate": 3.221185293773605e-05, "loss": 0.0313, "step": 63500 }, { "grad_norm": 0.36770328879356384, "learning_rate": 3.2196400974774354e-05, "loss": 0.0302, "step": 63510 }, { "grad_norm": 0.39372915029525757, "learning_rate": 3.2180950958787485e-05, "loss": 0.0384, "step": 63520 }, { "grad_norm": 0.2551364600658417, "learning_rate": 3.2165502891465024e-05, "loss": 0.029, "step": 63530 }, { "grad_norm": 0.5052998661994934, "learning_rate": 3.215005677449636e-05, "loss": 0.0405, "step": 63540 }, { "grad_norm": 0.38021090626716614, "learning_rate": 3.2134612609570644e-05, "loss": 0.0293, "step": 63550 }, { "grad_norm": 0.31363940238952637, "learning_rate": 3.211917039837682e-05, "loss": 0.035, "step": 63560 }, { "grad_norm": 0.4406931400299072, "learning_rate": 3.210373014260365e-05, "loss": 0.0525, "step": 63570 }, { "grad_norm": 0.38785895705223083, "learning_rate": 3.208829184393965e-05, "loss": 0.0325, "step": 63580 }, { "grad_norm": 0.5552996397018433, "learning_rate": 3.207285550407312e-05, "loss": 0.0478, "step": 63590 }, { "grad_norm": 0.40065333247184753, "learning_rate": 3.205742112469217e-05, "loss": 0.032, "step": 63600 }, { "grad_norm": 0.5313113927841187, "learning_rate": 3.204198870748465e-05, "loss": 0.0434, "step": 63610 }, { "grad_norm": 0.47664013504981995, "learning_rate": 3.202655825413824e-05, "loss": 0.0359, "step": 63620 }, { "grad_norm": 0.3149162232875824, "learning_rate": 3.2011129766340416e-05, "loss": 0.0538, "step": 63630 }, { "grad_norm": 0.44002997875213623, "learning_rate": 3.1995703245778384e-05, "loss": 0.0297, "step": 63640 }, { "grad_norm": 0.27471673488616943, "learning_rate": 3.1980278694139186e-05, "loss": 0.0332, "step": 63650 }, { "grad_norm": 0.4750097692012787, "learning_rate": 3.196485611310962e-05, "loss": 0.0347, "step": 63660 }, { "grad_norm": 0.5570342540740967, "learning_rate": 3.1949435504376257e-05, "loss": 0.0439, "step": 63670 }, { "grad_norm": 0.7382997274398804, "learning_rate": 3.1934016869625505e-05, "loss": 0.0323, "step": 63680 }, { "grad_norm": 0.7847108840942383, "learning_rate": 3.191860021054349e-05, "loss": 0.0506, "step": 63690 }, { "grad_norm": 0.47262394428253174, "learning_rate": 3.190318552881618e-05, "loss": 0.0474, "step": 63700 }, { "grad_norm": 0.8811744451522827, "learning_rate": 3.1887772826129285e-05, "loss": 0.0385, "step": 63710 }, { "grad_norm": 1.1569583415985107, "learning_rate": 3.1872362104168335e-05, "loss": 0.0435, "step": 63720 }, { "grad_norm": 0.41856998205184937, "learning_rate": 3.185695336461861e-05, "loss": 0.0382, "step": 63730 }, { "grad_norm": 0.5569615364074707, "learning_rate": 3.184154660916519e-05, "loss": 0.0306, "step": 63740 }, { "grad_norm": 0.4376163184642792, "learning_rate": 3.1826141839492906e-05, "loss": 0.0347, "step": 63750 }, { "grad_norm": 0.6515703201293945, "learning_rate": 3.181073905728644e-05, "loss": 0.0333, "step": 63760 }, { "grad_norm": 0.43124011158943176, "learning_rate": 3.1795338264230214e-05, "loss": 0.0231, "step": 63770 }, { "grad_norm": 0.6045215129852295, "learning_rate": 3.1779939462008426e-05, "loss": 0.0413, "step": 63780 }, { "grad_norm": 0.5285158753395081, "learning_rate": 3.1764542652305066e-05, "loss": 0.0323, "step": 63790 }, { "grad_norm": 0.4305359125137329, "learning_rate": 3.17491478368039e-05, "loss": 0.0407, "step": 63800 }, { "grad_norm": 0.5952514410018921, "learning_rate": 3.173375501718847e-05, "loss": 0.0355, "step": 63810 }, { "grad_norm": 0.3560195565223694, "learning_rate": 3.171836419514215e-05, "loss": 0.0332, "step": 63820 }, { "grad_norm": 0.6136110424995422, "learning_rate": 3.170297537234803e-05, "loss": 0.0569, "step": 63830 }, { "grad_norm": 0.5030789971351624, "learning_rate": 3.168758855048901e-05, "loss": 0.0327, "step": 63840 }, { "grad_norm": 0.37048619985580444, "learning_rate": 3.1672203731247785e-05, "loss": 0.0361, "step": 63850 }, { "grad_norm": 0.4280499517917633, "learning_rate": 3.16568209163068e-05, "loss": 0.0414, "step": 63860 }, { "grad_norm": 0.37841278314590454, "learning_rate": 3.1641440107348295e-05, "loss": 0.0435, "step": 63870 }, { "grad_norm": 0.42375504970550537, "learning_rate": 3.162606130605429e-05, "loss": 0.0354, "step": 63880 }, { "grad_norm": 0.4084790349006653, "learning_rate": 3.16106845141066e-05, "loss": 0.0336, "step": 63890 }, { "grad_norm": 0.4143826961517334, "learning_rate": 3.159530973318682e-05, "loss": 0.0323, "step": 63900 }, { "grad_norm": 0.42259567975997925, "learning_rate": 3.1579936964976286e-05, "loss": 0.0388, "step": 63910 }, { "grad_norm": 0.44627314805984497, "learning_rate": 3.156456621115615e-05, "loss": 0.0447, "step": 63920 }, { "grad_norm": 0.4577433168888092, "learning_rate": 3.154919747340734e-05, "loss": 0.0303, "step": 63930 }, { "grad_norm": 0.6460750699043274, "learning_rate": 3.153383075341054e-05, "loss": 0.0357, "step": 63940 }, { "grad_norm": 0.45521295070648193, "learning_rate": 3.1518466052846264e-05, "loss": 0.0371, "step": 63950 }, { "grad_norm": 0.8319339752197266, "learning_rate": 3.150310337339474e-05, "loss": 0.0377, "step": 63960 }, { "grad_norm": 0.5708317756652832, "learning_rate": 3.148774271673603e-05, "loss": 0.0346, "step": 63970 }, { "grad_norm": 0.306084007024765, "learning_rate": 3.147238408454995e-05, "loss": 0.0477, "step": 63980 }, { "grad_norm": 0.2652300000190735, "learning_rate": 3.1457027478516086e-05, "loss": 0.0391, "step": 63990 }, { "grad_norm": 0.46827781200408936, "learning_rate": 3.144167290031382e-05, "loss": 0.0379, "step": 64000 }, { "grad_norm": 0.5657390356063843, "learning_rate": 3.142632035162227e-05, "loss": 0.04, "step": 64010 }, { "grad_norm": 0.41172510385513306, "learning_rate": 3.141096983412044e-05, "loss": 0.0286, "step": 64020 }, { "grad_norm": 0.3683262765407562, "learning_rate": 3.139562134948699e-05, "loss": 0.0374, "step": 64030 }, { "grad_norm": 0.5642575025558472, "learning_rate": 3.1380274899400424e-05, "loss": 0.037, "step": 64040 }, { "grad_norm": 0.4082626700401306, "learning_rate": 3.1364930485539e-05, "loss": 0.0372, "step": 64050 }, { "grad_norm": 0.4535948634147644, "learning_rate": 3.134958810958076e-05, "loss": 0.0302, "step": 64060 }, { "grad_norm": 0.5086895227432251, "learning_rate": 3.13342477732035e-05, "loss": 0.0388, "step": 64070 }, { "grad_norm": 0.6742973327636719, "learning_rate": 3.131890947808487e-05, "loss": 0.0306, "step": 64080 }, { "grad_norm": 0.3458559513092041, "learning_rate": 3.13035732259022e-05, "loss": 0.0422, "step": 64090 }, { "grad_norm": 0.41713517904281616, "learning_rate": 3.1288239018332656e-05, "loss": 0.0323, "step": 64100 }, { "grad_norm": 0.45026782155036926, "learning_rate": 3.1272906857053164e-05, "loss": 0.0333, "step": 64110 }, { "grad_norm": 0.5822532176971436, "learning_rate": 3.125757674374042e-05, "loss": 0.0349, "step": 64120 }, { "grad_norm": 0.3302503824234009, "learning_rate": 3.124224868007088e-05, "loss": 0.0361, "step": 64130 }, { "grad_norm": 0.39023348689079285, "learning_rate": 3.1226922667720854e-05, "loss": 0.033, "step": 64140 }, { "grad_norm": 0.6497423648834229, "learning_rate": 3.121159870836633e-05, "loss": 0.0477, "step": 64150 }, { "grad_norm": 0.43945759534835815, "learning_rate": 3.119627680368312e-05, "loss": 0.0346, "step": 64160 }, { "grad_norm": 0.30283838510513306, "learning_rate": 3.1180956955346816e-05, "loss": 0.0355, "step": 64170 }, { "grad_norm": 0.4191688895225525, "learning_rate": 3.116563916503276e-05, "loss": 0.0383, "step": 64180 }, { "grad_norm": 0.46081456542015076, "learning_rate": 3.1150323434416095e-05, "loss": 0.0382, "step": 64190 }, { "grad_norm": 0.3347524106502533, "learning_rate": 3.1135009765171696e-05, "loss": 0.0374, "step": 64200 }, { "grad_norm": 0.36470794677734375, "learning_rate": 3.111969815897429e-05, "loss": 0.0409, "step": 64210 }, { "grad_norm": 0.36247435212135315, "learning_rate": 3.1104388617498295e-05, "loss": 0.0337, "step": 64220 }, { "grad_norm": 0.3863159120082855, "learning_rate": 3.1089081142417966e-05, "loss": 0.0423, "step": 64230 }, { "grad_norm": 0.38589608669281006, "learning_rate": 3.1073775735407285e-05, "loss": 0.0371, "step": 64240 }, { "grad_norm": 0.4954494833946228, "learning_rate": 3.1058472398140035e-05, "loss": 0.0391, "step": 64250 }, { "grad_norm": 0.4079478681087494, "learning_rate": 3.1043171132289736e-05, "loss": 0.0342, "step": 64260 }, { "grad_norm": 0.30967795848846436, "learning_rate": 3.1027871939529774e-05, "loss": 0.0283, "step": 64270 }, { "grad_norm": 0.646320104598999, "learning_rate": 3.1012574821533214e-05, "loss": 0.044, "step": 64280 }, { "grad_norm": 0.5675082802772522, "learning_rate": 3.099727977997291e-05, "loss": 0.0365, "step": 64290 }, { "grad_norm": 0.23236288130283356, "learning_rate": 3.098198681652154e-05, "loss": 0.0502, "step": 64300 }, { "grad_norm": 0.36088624596595764, "learning_rate": 3.0966695932851475e-05, "loss": 0.0279, "step": 64310 }, { "grad_norm": 0.5112788677215576, "learning_rate": 3.0951407130634926e-05, "loss": 0.0429, "step": 64320 }, { "grad_norm": 0.3727450370788574, "learning_rate": 3.0936120411543864e-05, "loss": 0.0358, "step": 64330 }, { "grad_norm": 0.5709592700004578, "learning_rate": 3.0920835777250015e-05, "loss": 0.0393, "step": 64340 }, { "grad_norm": 0.5500376224517822, "learning_rate": 3.090555322942488e-05, "loss": 0.0429, "step": 64350 }, { "grad_norm": 0.48119550943374634, "learning_rate": 3.089027276973973e-05, "loss": 0.0352, "step": 64360 }, { "grad_norm": 0.3915725648403168, "learning_rate": 3.0874994399865615e-05, "loss": 0.0336, "step": 64370 }, { "grad_norm": 0.5620823502540588, "learning_rate": 3.085971812147337e-05, "loss": 0.0346, "step": 64380 }, { "grad_norm": 0.32016435265541077, "learning_rate": 3.0844443936233557e-05, "loss": 0.039, "step": 64390 }, { "grad_norm": 0.5363320708274841, "learning_rate": 3.082917184581657e-05, "loss": 0.0366, "step": 64400 }, { "grad_norm": 0.29584965109825134, "learning_rate": 3.0813901851892513e-05, "loss": 0.0385, "step": 64410 }, { "grad_norm": 0.6576417088508606, "learning_rate": 3.0798633956131316e-05, "loss": 0.0437, "step": 64420 }, { "grad_norm": 0.30366525053977966, "learning_rate": 3.078336816020264e-05, "loss": 0.0318, "step": 64430 }, { "grad_norm": 0.8236488699913025, "learning_rate": 3.0768104465775935e-05, "loss": 0.0461, "step": 64440 }, { "grad_norm": 0.5869563221931458, "learning_rate": 3.0752842874520396e-05, "loss": 0.0342, "step": 64450 }, { "grad_norm": 0.4894031584262848, "learning_rate": 3.073758338810503e-05, "loss": 0.0353, "step": 64460 }, { "grad_norm": 0.4267702102661133, "learning_rate": 3.072232600819859e-05, "loss": 0.0385, "step": 64470 }, { "grad_norm": 0.341938853263855, "learning_rate": 3.070707073646959e-05, "loss": 0.0405, "step": 64480 }, { "grad_norm": 0.3842465281486511, "learning_rate": 3.069181757458633e-05, "loss": 0.0395, "step": 64490 }, { "grad_norm": 0.3676340579986572, "learning_rate": 3.067656652421687e-05, "loss": 0.0358, "step": 64500 }, { "grad_norm": 0.2822681665420532, "learning_rate": 3.0661317587029025e-05, "loss": 0.042, "step": 64510 }, { "grad_norm": 0.5024269223213196, "learning_rate": 3.064607076469043e-05, "loss": 0.033, "step": 64520 }, { "grad_norm": 0.4704747498035431, "learning_rate": 3.063082605886845e-05, "loss": 0.0384, "step": 64530 }, { "grad_norm": 0.505531370639801, "learning_rate": 3.061558347123019e-05, "loss": 0.0373, "step": 64540 }, { "grad_norm": 0.24295055866241455, "learning_rate": 3.060034300344259e-05, "loss": 0.038, "step": 64550 }, { "grad_norm": 0.5080150961875916, "learning_rate": 3.058510465717231e-05, "loss": 0.044, "step": 64560 }, { "grad_norm": 0.32808637619018555, "learning_rate": 3.0569868434085804e-05, "loss": 0.0277, "step": 64570 }, { "grad_norm": 0.5225654244422913, "learning_rate": 3.055463433584924e-05, "loss": 0.0277, "step": 64580 }, { "grad_norm": 0.5416894555091858, "learning_rate": 3.0539402364128656e-05, "loss": 0.0355, "step": 64590 }, { "grad_norm": 0.37942788004875183, "learning_rate": 3.052417252058977e-05, "loss": 0.0414, "step": 64600 }, { "grad_norm": 0.43071940541267395, "learning_rate": 3.0508944806898092e-05, "loss": 0.0457, "step": 64610 }, { "grad_norm": 0.3531651198863983, "learning_rate": 3.04937192247189e-05, "loss": 0.0494, "step": 64620 }, { "grad_norm": 1.1353250741958618, "learning_rate": 3.047849577571725e-05, "loss": 0.0457, "step": 64630 }, { "grad_norm": 0.3718309700489044, "learning_rate": 3.046327446155795e-05, "loss": 0.0345, "step": 64640 }, { "grad_norm": 0.39249780774116516, "learning_rate": 3.0448055283905587e-05, "loss": 0.0336, "step": 64650 }, { "grad_norm": 0.41811689734458923, "learning_rate": 3.0432838244424494e-05, "loss": 0.0318, "step": 64660 }, { "grad_norm": 0.3242288827896118, "learning_rate": 3.04176233447788e-05, "loss": 0.0262, "step": 64670 }, { "grad_norm": 0.3060383200645447, "learning_rate": 3.040241058663238e-05, "loss": 0.0311, "step": 64680 }, { "grad_norm": 0.4007227122783661, "learning_rate": 3.038719997164887e-05, "loss": 0.0446, "step": 64690 }, { "grad_norm": 0.2703786790370941, "learning_rate": 3.0371991501491683e-05, "loss": 0.0366, "step": 64700 }, { "grad_norm": 0.36400532722473145, "learning_rate": 3.0356785177823975e-05, "loss": 0.0392, "step": 64710 }, { "grad_norm": 0.2940483093261719, "learning_rate": 3.0341581002308726e-05, "loss": 0.0352, "step": 64720 }, { "grad_norm": 0.41402801871299744, "learning_rate": 3.0326378976608627e-05, "loss": 0.0342, "step": 64730 }, { "grad_norm": 0.28473156690597534, "learning_rate": 3.0311179102386146e-05, "loss": 0.0505, "step": 64740 }, { "grad_norm": 0.820848286151886, "learning_rate": 3.0295981381303513e-05, "loss": 0.0411, "step": 64750 }, { "grad_norm": 0.4249012768268585, "learning_rate": 3.0280785815022727e-05, "loss": 0.0363, "step": 64760 }, { "grad_norm": 0.6123962998390198, "learning_rate": 3.0265592405205544e-05, "loss": 0.0368, "step": 64770 }, { "grad_norm": 1.6438441276550293, "learning_rate": 3.0250401153513517e-05, "loss": 0.0325, "step": 64780 }, { "grad_norm": 1.7697336673736572, "learning_rate": 3.0235212061607936e-05, "loss": 0.0473, "step": 64790 }, { "grad_norm": 0.22308054566383362, "learning_rate": 3.0220025131149842e-05, "loss": 0.0334, "step": 64800 }, { "grad_norm": 0.4702512323856354, "learning_rate": 3.020484036380007e-05, "loss": 0.0317, "step": 64810 }, { "grad_norm": 0.42123112082481384, "learning_rate": 3.0189657761219188e-05, "loss": 0.046, "step": 64820 }, { "grad_norm": 0.2559159994125366, "learning_rate": 3.017447732506753e-05, "loss": 0.0499, "step": 64830 }, { "grad_norm": 0.33207812905311584, "learning_rate": 3.015929905700524e-05, "loss": 0.0315, "step": 64840 }, { "grad_norm": 0.3175274431705475, "learning_rate": 3.014412295869218e-05, "loss": 0.0329, "step": 64850 }, { "grad_norm": 0.6588858366012573, "learning_rate": 3.0128949031787977e-05, "loss": 0.0442, "step": 64860 }, { "grad_norm": 0.38947802782058716, "learning_rate": 3.0113777277952022e-05, "loss": 0.0412, "step": 64870 }, { "grad_norm": 0.5233381390571594, "learning_rate": 3.0098607698843496e-05, "loss": 0.0348, "step": 64880 }, { "grad_norm": 0.4404936730861664, "learning_rate": 3.0083440296121312e-05, "loss": 0.0353, "step": 64890 }, { "grad_norm": 0.33105021715164185, "learning_rate": 3.0068275071444142e-05, "loss": 0.0314, "step": 64900 }, { "grad_norm": 0.5452907681465149, "learning_rate": 3.0053112026470442e-05, "loss": 0.0444, "step": 64910 }, { "grad_norm": 0.4289010167121887, "learning_rate": 3.0037951162858434e-05, "loss": 0.0342, "step": 64920 }, { "grad_norm": 0.3686385154724121, "learning_rate": 3.002279248226607e-05, "loss": 0.034, "step": 64930 }, { "grad_norm": 0.3670361042022705, "learning_rate": 3.000763598635109e-05, "loss": 0.038, "step": 64940 }, { "grad_norm": 0.5277934074401855, "learning_rate": 2.999248167677098e-05, "loss": 0.0515, "step": 64950 }, { "grad_norm": 0.5485573410987854, "learning_rate": 2.9977329555182976e-05, "loss": 0.0434, "step": 64960 }, { "grad_norm": 0.37288251519203186, "learning_rate": 2.9962179623244124e-05, "loss": 0.0438, "step": 64970 }, { "grad_norm": 0.5051500201225281, "learning_rate": 2.9947031882611175e-05, "loss": 0.0366, "step": 64980 }, { "grad_norm": 0.5036627650260925, "learning_rate": 2.9931886334940685e-05, "loss": 0.0359, "step": 64990 }, { "grad_norm": 0.2789117991924286, "learning_rate": 2.9916742981888923e-05, "loss": 0.0335, "step": 65000 }, { "grad_norm": 0.5275129079818726, "learning_rate": 2.990160182511196e-05, "loss": 0.0458, "step": 65010 }, { "grad_norm": 0.7054137587547302, "learning_rate": 2.9886462866265574e-05, "loss": 0.0374, "step": 65020 }, { "grad_norm": 0.3543451428413391, "learning_rate": 2.9871326107005392e-05, "loss": 0.0282, "step": 65030 }, { "grad_norm": 0.6115793585777283, "learning_rate": 2.9856191548986735e-05, "loss": 0.0452, "step": 65040 }, { "grad_norm": 0.30284255743026733, "learning_rate": 2.9841059193864672e-05, "loss": 0.0405, "step": 65050 }, { "grad_norm": 0.3646521270275116, "learning_rate": 2.982592904329407e-05, "loss": 0.0336, "step": 65060 }, { "grad_norm": 0.17903585731983185, "learning_rate": 2.981080109892954e-05, "loss": 0.0281, "step": 65070 }, { "grad_norm": 0.6635757684707642, "learning_rate": 2.9795675362425434e-05, "loss": 0.0334, "step": 65080 }, { "grad_norm": 0.38851112127304077, "learning_rate": 2.97805518354359e-05, "loss": 0.0307, "step": 65090 }, { "grad_norm": 0.7486721277236938, "learning_rate": 2.9765430519614825e-05, "loss": 0.0319, "step": 65100 }, { "grad_norm": 0.5689735412597656, "learning_rate": 2.975031141661584e-05, "loss": 0.0364, "step": 65110 }, { "grad_norm": 0.34352993965148926, "learning_rate": 2.9735194528092347e-05, "loss": 0.0356, "step": 65120 }, { "grad_norm": 0.38038918375968933, "learning_rate": 2.9720079855697525e-05, "loss": 0.0525, "step": 65130 }, { "grad_norm": 1.5511703491210938, "learning_rate": 2.970496740108427e-05, "loss": 0.0525, "step": 65140 }, { "grad_norm": 0.40684574842453003, "learning_rate": 2.968985716590527e-05, "loss": 0.0308, "step": 65150 }, { "grad_norm": 0.3070218563079834, "learning_rate": 2.9674749151812953e-05, "loss": 0.034, "step": 65160 }, { "grad_norm": 0.4591968059539795, "learning_rate": 2.965964336045951e-05, "loss": 0.0371, "step": 65170 }, { "grad_norm": 2.7552666664123535, "learning_rate": 2.9644539793496894e-05, "loss": 0.0333, "step": 65180 }, { "grad_norm": 0.5875915884971619, "learning_rate": 2.9629438452576807e-05, "loss": 0.0386, "step": 65190 }, { "grad_norm": 0.48316895961761475, "learning_rate": 2.9614339339350698e-05, "loss": 0.032, "step": 65200 }, { "grad_norm": 0.8596364259719849, "learning_rate": 2.959924245546979e-05, "loss": 0.0347, "step": 65210 }, { "grad_norm": 0.41960081458091736, "learning_rate": 2.9584147802585038e-05, "loss": 0.0532, "step": 65220 }, { "grad_norm": 0.40467768907546997, "learning_rate": 2.956905538234721e-05, "loss": 0.0368, "step": 65230 }, { "grad_norm": 0.4947377145290375, "learning_rate": 2.9553965196406764e-05, "loss": 0.0416, "step": 65240 }, { "grad_norm": 0.587742805480957, "learning_rate": 2.9538877246413943e-05, "loss": 0.0292, "step": 65250 }, { "grad_norm": 0.40432050824165344, "learning_rate": 2.952379153401874e-05, "loss": 0.0373, "step": 65260 }, { "grad_norm": 0.47484391927719116, "learning_rate": 2.9508708060870914e-05, "loss": 0.0325, "step": 65270 }, { "grad_norm": 0.3799184560775757, "learning_rate": 2.9493626828619948e-05, "loss": 0.0342, "step": 65280 }, { "grad_norm": 0.4437927007675171, "learning_rate": 2.9478547838915137e-05, "loss": 0.0321, "step": 65290 }, { "grad_norm": 0.6842339038848877, "learning_rate": 2.946347109340548e-05, "loss": 0.0351, "step": 65300 }, { "grad_norm": 0.5021464228630066, "learning_rate": 2.944839659373976e-05, "loss": 0.0337, "step": 65310 }, { "grad_norm": 0.6618209481239319, "learning_rate": 2.943332434156647e-05, "loss": 0.0412, "step": 65320 }, { "grad_norm": 0.4105689227581024, "learning_rate": 2.9418254338533912e-05, "loss": 0.0422, "step": 65330 }, { "grad_norm": 0.914428174495697, "learning_rate": 2.940318658629011e-05, "loss": 0.0397, "step": 65340 }, { "grad_norm": 0.4192323088645935, "learning_rate": 2.938812108648287e-05, "loss": 0.0413, "step": 65350 }, { "grad_norm": 0.4107896089553833, "learning_rate": 2.937305784075972e-05, "loss": 0.0316, "step": 65360 }, { "grad_norm": 0.6274029612541199, "learning_rate": 2.9357996850767937e-05, "loss": 0.0332, "step": 65370 }, { "grad_norm": 0.4293578267097473, "learning_rate": 2.9342938118154595e-05, "loss": 0.0307, "step": 65380 }, { "grad_norm": 0.5590438842773438, "learning_rate": 2.9327881644566485e-05, "loss": 0.0452, "step": 65390 }, { "grad_norm": 0.5308656096458435, "learning_rate": 2.9312827431650157e-05, "loss": 0.0354, "step": 65400 }, { "grad_norm": 0.2500970959663391, "learning_rate": 2.92977754810519e-05, "loss": 0.0508, "step": 65410 }, { "grad_norm": 0.40250125527381897, "learning_rate": 2.928272579441781e-05, "loss": 0.0379, "step": 65420 }, { "grad_norm": 0.40231937170028687, "learning_rate": 2.9267678373393676e-05, "loss": 0.0452, "step": 65430 }, { "grad_norm": 0.28378498554229736, "learning_rate": 2.925263321962507e-05, "loss": 0.041, "step": 65440 }, { "grad_norm": 0.4152117669582367, "learning_rate": 2.9237590334757292e-05, "loss": 0.0347, "step": 65450 }, { "grad_norm": 0.4667902886867523, "learning_rate": 2.922254972043543e-05, "loss": 0.0391, "step": 65460 }, { "grad_norm": 1.157322883605957, "learning_rate": 2.9207511378304265e-05, "loss": 0.0391, "step": 65470 }, { "grad_norm": 0.5402024388313293, "learning_rate": 2.9192475310008428e-05, "loss": 0.0487, "step": 65480 }, { "grad_norm": 0.34961241483688354, "learning_rate": 2.917744151719219e-05, "loss": 0.0314, "step": 65490 }, { "grad_norm": 0.6302024722099304, "learning_rate": 2.916241000149964e-05, "loss": 0.0388, "step": 65500 }, { "grad_norm": 0.40429121255874634, "learning_rate": 2.9147380764574622e-05, "loss": 0.0303, "step": 65510 }, { "grad_norm": 0.6744664311408997, "learning_rate": 2.9132353808060674e-05, "loss": 0.0498, "step": 65520 }, { "grad_norm": 0.3572118878364563, "learning_rate": 2.9117329133601134e-05, "loss": 0.036, "step": 65530 }, { "grad_norm": 0.32450422644615173, "learning_rate": 2.9102306742839093e-05, "loss": 0.0379, "step": 65540 }, { "grad_norm": 0.3037191927433014, "learning_rate": 2.9087286637417375e-05, "loss": 0.0344, "step": 65550 }, { "grad_norm": 0.4600870907306671, "learning_rate": 2.9072268818978537e-05, "loss": 0.0403, "step": 65560 }, { "grad_norm": 0.3701547682285309, "learning_rate": 2.905725328916493e-05, "loss": 0.0333, "step": 65570 }, { "grad_norm": 0.5883277058601379, "learning_rate": 2.9042240049618608e-05, "loss": 0.0394, "step": 65580 }, { "grad_norm": 0.47826579213142395, "learning_rate": 2.9027229101981413e-05, "loss": 0.0341, "step": 65590 }, { "grad_norm": 0.33552372455596924, "learning_rate": 2.901222044789488e-05, "loss": 0.0419, "step": 65600 }, { "grad_norm": 0.40327557921409607, "learning_rate": 2.89972140890004e-05, "loss": 0.0315, "step": 65610 }, { "grad_norm": 0.6026955842971802, "learning_rate": 2.8982210026938988e-05, "loss": 0.0425, "step": 65620 }, { "grad_norm": 0.5193939208984375, "learning_rate": 2.896720826335151e-05, "loss": 0.0569, "step": 65630 }, { "grad_norm": 0.42097076773643494, "learning_rate": 2.8952208799878493e-05, "loss": 0.0442, "step": 65640 }, { "grad_norm": 0.41354745626449585, "learning_rate": 2.89372116381603e-05, "loss": 0.0331, "step": 65650 }, { "grad_norm": 0.49781760573387146, "learning_rate": 2.8922216779836932e-05, "loss": 0.0444, "step": 65660 }, { "grad_norm": 0.5392994284629822, "learning_rate": 2.8907224226548285e-05, "loss": 0.0307, "step": 65670 }, { "grad_norm": 0.44399186968803406, "learning_rate": 2.8892233979933868e-05, "loss": 0.0348, "step": 65680 }, { "grad_norm": 0.689238965511322, "learning_rate": 2.8877246041633026e-05, "loss": 0.0409, "step": 65690 }, { "grad_norm": 0.6684481501579285, "learning_rate": 2.886226041328478e-05, "loss": 0.0363, "step": 65700 }, { "grad_norm": 0.43229302763938904, "learning_rate": 2.8847277096527957e-05, "loss": 0.0316, "step": 65710 }, { "grad_norm": 0.3940562307834625, "learning_rate": 2.8832296093001105e-05, "loss": 0.0282, "step": 65720 }, { "grad_norm": 0.3045719265937805, "learning_rate": 2.8817317404342547e-05, "loss": 0.0318, "step": 65730 }, { "grad_norm": 0.4097434878349304, "learning_rate": 2.880234103219029e-05, "loss": 0.0453, "step": 65740 }, { "grad_norm": 0.7352378964424133, "learning_rate": 2.878736697818215e-05, "loss": 0.03, "step": 65750 }, { "grad_norm": 0.6421294808387756, "learning_rate": 2.877239524395568e-05, "loss": 0.0368, "step": 65760 }, { "grad_norm": 0.2919195592403412, "learning_rate": 2.8757425831148134e-05, "loss": 0.0396, "step": 65770 }, { "grad_norm": 0.42271947860717773, "learning_rate": 2.874245874139658e-05, "loss": 0.0401, "step": 65780 }, { "grad_norm": 0.43199583888053894, "learning_rate": 2.8727493976337737e-05, "loss": 0.025, "step": 65790 }, { "grad_norm": 0.5994628071784973, "learning_rate": 2.8712531537608218e-05, "loss": 0.0296, "step": 65800 }, { "grad_norm": 0.5101528763771057, "learning_rate": 2.8697571426844217e-05, "loss": 0.0416, "step": 65810 }, { "grad_norm": 0.5095633268356323, "learning_rate": 2.86826136456818e-05, "loss": 0.0329, "step": 65820 }, { "grad_norm": 0.3396972119808197, "learning_rate": 2.8667658195756685e-05, "loss": 0.0446, "step": 65830 }, { "grad_norm": 0.5346297025680542, "learning_rate": 2.8652705078704412e-05, "loss": 0.0494, "step": 65840 }, { "grad_norm": 0.40235084295272827, "learning_rate": 2.863775429616018e-05, "loss": 0.0449, "step": 65850 }, { "grad_norm": 0.4067859947681427, "learning_rate": 2.862280584975906e-05, "loss": 0.0323, "step": 65860 }, { "grad_norm": 0.32436996698379517, "learning_rate": 2.8607859741135735e-05, "loss": 0.0415, "step": 65870 }, { "grad_norm": 0.3374960422515869, "learning_rate": 2.8592915971924723e-05, "loss": 0.045, "step": 65880 }, { "grad_norm": 0.43964147567749023, "learning_rate": 2.8577974543760212e-05, "loss": 0.0397, "step": 65890 }, { "grad_norm": 0.4524308145046234, "learning_rate": 2.8563035458276222e-05, "loss": 0.0331, "step": 65900 }, { "grad_norm": 0.25981801748275757, "learning_rate": 2.8548098717106424e-05, "loss": 0.0358, "step": 65910 }, { "grad_norm": 0.4342934489250183, "learning_rate": 2.85331643218843e-05, "loss": 0.0369, "step": 65920 }, { "grad_norm": 0.48884138464927673, "learning_rate": 2.8518232274243046e-05, "loss": 0.0431, "step": 65930 }, { "grad_norm": 1.135780930519104, "learning_rate": 2.8503302575815638e-05, "loss": 0.0567, "step": 65940 }, { "grad_norm": 0.25905120372772217, "learning_rate": 2.8488375228234716e-05, "loss": 0.0393, "step": 65950 }, { "grad_norm": 0.5961889028549194, "learning_rate": 2.8473450233132738e-05, "loss": 0.0387, "step": 65960 }, { "grad_norm": 0.37292248010635376, "learning_rate": 2.8458527592141898e-05, "loss": 0.0359, "step": 65970 }, { "grad_norm": 0.3305107057094574, "learning_rate": 2.8443607306894073e-05, "loss": 0.0415, "step": 65980 }, { "grad_norm": 0.3177517056465149, "learning_rate": 2.8428689379020946e-05, "loss": 0.0399, "step": 65990 }, { "grad_norm": 0.34341397881507874, "learning_rate": 2.841377381015391e-05, "loss": 0.0413, "step": 66000 }, { "grad_norm": 0.3595815598964691, "learning_rate": 2.8398860601924144e-05, "loss": 0.0323, "step": 66010 }, { "grad_norm": 0.38291168212890625, "learning_rate": 2.838394975596248e-05, "loss": 0.0324, "step": 66020 }, { "grad_norm": 0.6459179520606995, "learning_rate": 2.83690412738996e-05, "loss": 0.0497, "step": 66030 }, { "grad_norm": 0.30750736594200134, "learning_rate": 2.835413515736579e-05, "loss": 0.0412, "step": 66040 }, { "grad_norm": 0.4755784273147583, "learning_rate": 2.8339231407991267e-05, "loss": 0.0351, "step": 66050 }, { "grad_norm": 0.572659432888031, "learning_rate": 2.832433002740581e-05, "loss": 0.0332, "step": 66060 }, { "grad_norm": 0.562131404876709, "learning_rate": 2.8309431017239047e-05, "loss": 0.029, "step": 66070 }, { "grad_norm": 0.3331749439239502, "learning_rate": 2.829453437912028e-05, "loss": 0.0375, "step": 66080 }, { "grad_norm": 0.3979825973510742, "learning_rate": 2.8279640114678617e-05, "loss": 0.0301, "step": 66090 }, { "grad_norm": 0.5262713432312012, "learning_rate": 2.8264748225542836e-05, "loss": 0.0458, "step": 66100 }, { "grad_norm": 0.24880148470401764, "learning_rate": 2.8249858713341516e-05, "loss": 0.0316, "step": 66110 }, { "grad_norm": 0.5503616333007812, "learning_rate": 2.823497157970294e-05, "loss": 0.0442, "step": 66120 }, { "grad_norm": 0.371297150850296, "learning_rate": 2.8220086826255164e-05, "loss": 0.027, "step": 66130 }, { "grad_norm": 0.4962557554244995, "learning_rate": 2.8205204454625934e-05, "loss": 0.0381, "step": 66140 }, { "grad_norm": 0.40119415521621704, "learning_rate": 2.8190324466442798e-05, "loss": 0.0351, "step": 66150 }, { "grad_norm": 0.6231469511985779, "learning_rate": 2.8175446863332965e-05, "loss": 0.0334, "step": 66160 }, { "grad_norm": 0.4395422041416168, "learning_rate": 2.816057164692345e-05, "loss": 0.0387, "step": 66170 }, { "grad_norm": 0.25446584820747375, "learning_rate": 2.8145698818840992e-05, "loss": 0.0455, "step": 66180 }, { "grad_norm": 0.3775192201137543, "learning_rate": 2.8130828380712065e-05, "loss": 0.0345, "step": 66190 }, { "grad_norm": 0.5732877254486084, "learning_rate": 2.811596033416285e-05, "loss": 0.0378, "step": 66200 }, { "grad_norm": 0.4770174026489258, "learning_rate": 2.8101094680819307e-05, "loss": 0.0366, "step": 66210 }, { "grad_norm": 0.8991924524307251, "learning_rate": 2.808623142230715e-05, "loss": 0.0325, "step": 66220 }, { "grad_norm": 0.4221303164958954, "learning_rate": 2.8071370560251757e-05, "loss": 0.0307, "step": 66230 }, { "grad_norm": 0.5201526880264282, "learning_rate": 2.805651209627831e-05, "loss": 0.0411, "step": 66240 }, { "grad_norm": 0.463098406791687, "learning_rate": 2.804165603201171e-05, "loss": 0.0356, "step": 66250 }, { "grad_norm": 0.5449079871177673, "learning_rate": 2.802680236907661e-05, "loss": 0.0357, "step": 66260 }, { "grad_norm": 0.4001524746417999, "learning_rate": 2.8011951109097344e-05, "loss": 0.0307, "step": 66270 }, { "grad_norm": 0.3428606390953064, "learning_rate": 2.7997102253698072e-05, "loss": 0.0378, "step": 66280 }, { "grad_norm": 0.38993290066719055, "learning_rate": 2.7982255804502588e-05, "loss": 0.034, "step": 66290 }, { "grad_norm": 0.3510727882385254, "learning_rate": 2.7967411763134516e-05, "loss": 0.0415, "step": 66300 }, { "grad_norm": 0.42657652497291565, "learning_rate": 2.7952570131217166e-05, "loss": 0.0301, "step": 66310 }, { "grad_norm": 0.4529920816421509, "learning_rate": 2.7937730910373618e-05, "loss": 0.0382, "step": 66320 }, { "grad_norm": 0.25089454650878906, "learning_rate": 2.7922894102226628e-05, "loss": 0.0292, "step": 66330 }, { "grad_norm": 0.3656826913356781, "learning_rate": 2.790805970839876e-05, "loss": 0.0324, "step": 66340 }, { "grad_norm": 0.32006126642227173, "learning_rate": 2.789322773051225e-05, "loss": 0.0262, "step": 66350 }, { "grad_norm": 0.4576055109500885, "learning_rate": 2.7878398170189124e-05, "loss": 0.0307, "step": 66360 }, { "grad_norm": 0.390049546957016, "learning_rate": 2.7863571029051112e-05, "loss": 0.0357, "step": 66370 }, { "grad_norm": 0.32244864106178284, "learning_rate": 2.784874630871971e-05, "loss": 0.0406, "step": 66380 }, { "grad_norm": 0.9388384222984314, "learning_rate": 2.7833924010816086e-05, "loss": 0.0361, "step": 66390 }, { "grad_norm": 0.35327786207199097, "learning_rate": 2.7819104136961226e-05, "loss": 0.0366, "step": 66400 }, { "grad_norm": 0.17599332332611084, "learning_rate": 2.780428668877577e-05, "loss": 0.0425, "step": 66410 }, { "grad_norm": 0.4103219509124756, "learning_rate": 2.7789471667880152e-05, "loss": 0.0399, "step": 66420 }, { "grad_norm": 0.40241241455078125, "learning_rate": 2.7774659075894523e-05, "loss": 0.0419, "step": 66430 }, { "grad_norm": 0.5025670528411865, "learning_rate": 2.7759848914438778e-05, "loss": 0.0556, "step": 66440 }, { "grad_norm": 0.46143680810928345, "learning_rate": 2.7745041185132496e-05, "loss": 0.0425, "step": 66450 }, { "grad_norm": 0.43058741092681885, "learning_rate": 2.773023588959506e-05, "loss": 0.0423, "step": 66460 }, { "grad_norm": 0.4109157621860504, "learning_rate": 2.7715433029445555e-05, "loss": 0.0383, "step": 66470 }, { "grad_norm": 0.4242524206638336, "learning_rate": 2.7700632606302766e-05, "loss": 0.0383, "step": 66480 }, { "grad_norm": 0.438224732875824, "learning_rate": 2.768583462178528e-05, "loss": 0.0371, "step": 66490 }, { "grad_norm": 0.537908673286438, "learning_rate": 2.767103907751136e-05, "loss": 0.0353, "step": 66500 }, { "grad_norm": 0.6113970875740051, "learning_rate": 2.7656245975099065e-05, "loss": 0.0344, "step": 66510 }, { "grad_norm": 0.4429483711719513, "learning_rate": 2.7641455316166093e-05, "loss": 0.0374, "step": 66520 }, { "grad_norm": 0.3125934302806854, "learning_rate": 2.7626667102329973e-05, "loss": 0.037, "step": 66530 }, { "grad_norm": 0.32249003648757935, "learning_rate": 2.761188133520788e-05, "loss": 0.0331, "step": 66540 }, { "grad_norm": 0.7078824639320374, "learning_rate": 2.7597098016416784e-05, "loss": 0.0371, "step": 66550 }, { "grad_norm": 0.35567548871040344, "learning_rate": 2.7582317147573368e-05, "loss": 0.0312, "step": 66560 }, { "grad_norm": 0.387093186378479, "learning_rate": 2.7567538730294064e-05, "loss": 0.0398, "step": 66570 }, { "grad_norm": 0.8442152738571167, "learning_rate": 2.7552762766194973e-05, "loss": 0.0451, "step": 66580 }, { "grad_norm": 0.4458977282047272, "learning_rate": 2.753798925689202e-05, "loss": 0.0423, "step": 66590 }, { "grad_norm": 0.4423471689224243, "learning_rate": 2.7523218204000774e-05, "loss": 0.0388, "step": 66600 }, { "grad_norm": 0.3720455467700958, "learning_rate": 2.7508449609136612e-05, "loss": 0.0307, "step": 66610 }, { "grad_norm": 0.514685869216919, "learning_rate": 2.749368347391454e-05, "loss": 0.0469, "step": 66620 }, { "grad_norm": 0.9140571355819702, "learning_rate": 2.7478919799949443e-05, "loss": 0.0473, "step": 66630 }, { "grad_norm": 0.4637454152107239, "learning_rate": 2.7464158588855803e-05, "loss": 0.0343, "step": 66640 }, { "grad_norm": 0.37545260787010193, "learning_rate": 2.7449399842247918e-05, "loss": 0.0326, "step": 66650 }, { "grad_norm": 0.34806138277053833, "learning_rate": 2.7434643561739736e-05, "loss": 0.0353, "step": 66660 }, { "grad_norm": 0.5639461874961853, "learning_rate": 2.7419889748945016e-05, "loss": 0.0322, "step": 66670 }, { "grad_norm": 0.3094526529312134, "learning_rate": 2.7405138405477203e-05, "loss": 0.0354, "step": 66680 }, { "grad_norm": 0.3601575195789337, "learning_rate": 2.73903895329495e-05, "loss": 0.0354, "step": 66690 }, { "grad_norm": 0.6089060306549072, "learning_rate": 2.737564313297478e-05, "loss": 0.042, "step": 66700 }, { "grad_norm": 0.38611817359924316, "learning_rate": 2.736089920716572e-05, "loss": 0.0325, "step": 66710 }, { "grad_norm": 0.3546663522720337, "learning_rate": 2.7346157757134706e-05, "loss": 0.0374, "step": 66720 }, { "grad_norm": 0.47272562980651855, "learning_rate": 2.733141878449379e-05, "loss": 0.0304, "step": 66730 }, { "grad_norm": 1.4641305208206177, "learning_rate": 2.7316682290854844e-05, "loss": 0.0421, "step": 66740 }, { "grad_norm": 0.42316070199012756, "learning_rate": 2.7301948277829414e-05, "loss": 0.0396, "step": 66750 }, { "grad_norm": 0.7762956023216248, "learning_rate": 2.7287216747028814e-05, "loss": 0.0433, "step": 66760 }, { "grad_norm": 0.603532612323761, "learning_rate": 2.7272487700064025e-05, "loss": 0.0371, "step": 66770 }, { "grad_norm": 0.845112144947052, "learning_rate": 2.725776113854583e-05, "loss": 0.0479, "step": 66780 }, { "grad_norm": 0.32182034850120544, "learning_rate": 2.7243037064084664e-05, "loss": 0.0354, "step": 66790 }, { "grad_norm": 0.5682206153869629, "learning_rate": 2.7228315478290766e-05, "loss": 0.0446, "step": 66800 }, { "grad_norm": 0.7174062728881836, "learning_rate": 2.721359638277402e-05, "loss": 0.0365, "step": 66810 }, { "grad_norm": 0.43895968794822693, "learning_rate": 2.7198879779144148e-05, "loss": 0.0296, "step": 66820 }, { "grad_norm": 0.3080295920372009, "learning_rate": 2.7184165669010485e-05, "loss": 0.0348, "step": 66830 }, { "grad_norm": 0.5437845587730408, "learning_rate": 2.716945405398218e-05, "loss": 0.0362, "step": 66840 }, { "grad_norm": 0.5410961508750916, "learning_rate": 2.715474493566803e-05, "loss": 0.0361, "step": 66850 }, { "grad_norm": 0.4099293649196625, "learning_rate": 2.7140038315676652e-05, "loss": 0.0282, "step": 66860 }, { "grad_norm": 0.27383744716644287, "learning_rate": 2.7125334195616275e-05, "loss": 0.0607, "step": 66870 }, { "grad_norm": 0.40572646260261536, "learning_rate": 2.7110632577094997e-05, "loss": 0.0372, "step": 66880 }, { "grad_norm": 0.49299320578575134, "learning_rate": 2.709593346172051e-05, "loss": 0.0455, "step": 66890 }, { "grad_norm": 0.49111706018447876, "learning_rate": 2.708123685110032e-05, "loss": 0.0366, "step": 66900 }, { "grad_norm": 0.2999686598777771, "learning_rate": 2.7066542746841588e-05, "loss": 0.0233, "step": 66910 }, { "grad_norm": 0.7243971228599548, "learning_rate": 2.7051851150551267e-05, "loss": 0.0385, "step": 66920 }, { "grad_norm": 0.40914762020111084, "learning_rate": 2.7037162063835996e-05, "loss": 0.0322, "step": 66930 }, { "grad_norm": 0.9054360389709473, "learning_rate": 2.702247548830218e-05, "loss": 0.0431, "step": 66940 }, { "grad_norm": 0.6953735947608948, "learning_rate": 2.7007791425555883e-05, "loss": 0.0343, "step": 66950 }, { "grad_norm": 0.31583940982818604, "learning_rate": 2.6993109877202945e-05, "loss": 0.0343, "step": 66960 }, { "grad_norm": 0.5787689685821533, "learning_rate": 2.6978430844848935e-05, "loss": 0.0364, "step": 66970 }, { "grad_norm": 0.437777578830719, "learning_rate": 2.69637543300991e-05, "loss": 0.0325, "step": 66980 }, { "grad_norm": 0.30985400080680847, "learning_rate": 2.694908033455848e-05, "loss": 0.0413, "step": 66990 }, { "grad_norm": 0.4781862795352936, "learning_rate": 2.6934408859831733e-05, "loss": 0.0414, "step": 67000 }, { "grad_norm": 0.5277650356292725, "learning_rate": 2.69197399075234e-05, "loss": 0.0409, "step": 67010 }, { "grad_norm": 0.3998471200466156, "learning_rate": 2.6905073479237584e-05, "loss": 0.037, "step": 67020 }, { "grad_norm": 0.6382278203964233, "learning_rate": 2.6890409576578235e-05, "loss": 0.0427, "step": 67030 }, { "grad_norm": 0.8067014813423157, "learning_rate": 2.687574820114892e-05, "loss": 0.0373, "step": 67040 }, { "grad_norm": 0.7726795077323914, "learning_rate": 2.686108935455305e-05, "loss": 0.0418, "step": 67050 }, { "grad_norm": 0.4054168462753296, "learning_rate": 2.6846433038393603e-05, "loss": 0.0357, "step": 67060 }, { "grad_norm": 0.47236207127571106, "learning_rate": 2.6831779254273482e-05, "loss": 0.0362, "step": 67070 }, { "grad_norm": 1.4569472074508667, "learning_rate": 2.6817128003795122e-05, "loss": 0.0372, "step": 67080 }, { "grad_norm": 0.35102424025535583, "learning_rate": 2.680247928856081e-05, "loss": 0.0323, "step": 67090 }, { "grad_norm": 0.44475510716438293, "learning_rate": 2.6787833110172466e-05, "loss": 0.0434, "step": 67100 }, { "grad_norm": 0.25344330072402954, "learning_rate": 2.6773189470231807e-05, "loss": 0.0339, "step": 67110 }, { "grad_norm": 0.44843554496765137, "learning_rate": 2.675854837034022e-05, "loss": 0.035, "step": 67120 }, { "grad_norm": 0.44023314118385315, "learning_rate": 2.6743909812098833e-05, "loss": 0.0343, "step": 67130 }, { "grad_norm": 0.3380180299282074, "learning_rate": 2.67292737971085e-05, "loss": 0.0369, "step": 67140 }, { "grad_norm": 0.3790590763092041, "learning_rate": 2.671464032696982e-05, "loss": 0.041, "step": 67150 }, { "grad_norm": 0.5222645401954651, "learning_rate": 2.6700009403283055e-05, "loss": 0.0406, "step": 67160 }, { "grad_norm": 0.5009134411811829, "learning_rate": 2.6685381027648214e-05, "loss": 0.0412, "step": 67170 }, { "grad_norm": 0.5150229334831238, "learning_rate": 2.6670755201665086e-05, "loss": 0.0459, "step": 67180 }, { "grad_norm": 0.4471222758293152, "learning_rate": 2.6656131926933066e-05, "loss": 0.041, "step": 67190 }, { "grad_norm": 0.8448491096496582, "learning_rate": 2.664151120505136e-05, "loss": 0.035, "step": 67200 }, { "grad_norm": 0.33631160855293274, "learning_rate": 2.6626893037618873e-05, "loss": 0.0352, "step": 67210 }, { "grad_norm": 0.5634744763374329, "learning_rate": 2.6612277426234243e-05, "loss": 0.0573, "step": 67220 }, { "grad_norm": 0.49690350890159607, "learning_rate": 2.659766437249577e-05, "loss": 0.0549, "step": 67230 }, { "grad_norm": 0.4365033805370331, "learning_rate": 2.6583053878001557e-05, "loss": 0.0373, "step": 67240 }, { "grad_norm": 0.6591017842292786, "learning_rate": 2.6568445944349323e-05, "loss": 0.0406, "step": 67250 }, { "grad_norm": 0.4899609088897705, "learning_rate": 2.6553840573136657e-05, "loss": 0.0345, "step": 67260 }, { "grad_norm": 0.5103960037231445, "learning_rate": 2.6539237765960706e-05, "loss": 0.0423, "step": 67270 }, { "grad_norm": 1.016613483428955, "learning_rate": 2.6524637524418465e-05, "loss": 0.0389, "step": 67280 }, { "grad_norm": 0.3041157126426697, "learning_rate": 2.651003985010655e-05, "loss": 0.027, "step": 67290 }, { "grad_norm": 0.4340198040008545, "learning_rate": 2.649544474462138e-05, "loss": 0.0412, "step": 67300 }, { "grad_norm": 0.4253090023994446, "learning_rate": 2.6480852209559015e-05, "loss": 0.0398, "step": 67310 }, { "grad_norm": 0.47325751185417175, "learning_rate": 2.6466262246515296e-05, "loss": 0.0296, "step": 67320 }, { "grad_norm": 0.4410945773124695, "learning_rate": 2.6451674857085746e-05, "loss": 0.043, "step": 67330 }, { "grad_norm": 0.6659220457077026, "learning_rate": 2.6437090042865655e-05, "loss": 0.0519, "step": 67340 }, { "grad_norm": 0.2612423300743103, "learning_rate": 2.6422507805449947e-05, "loss": 0.0328, "step": 67350 }, { "grad_norm": 0.37204819917678833, "learning_rate": 2.6407928146433357e-05, "loss": 0.0342, "step": 67360 }, { "grad_norm": 0.5112460255622864, "learning_rate": 2.639335106741026e-05, "loss": 0.0323, "step": 67370 }, { "grad_norm": 0.39881259202957153, "learning_rate": 2.6378776569974796e-05, "loss": 0.0433, "step": 67380 }, { "grad_norm": 0.3452412784099579, "learning_rate": 2.6364204655720814e-05, "loss": 0.0324, "step": 67390 }, { "grad_norm": 0.4035796523094177, "learning_rate": 2.6349635326241894e-05, "loss": 0.0383, "step": 67400 }, { "grad_norm": 0.457736074924469, "learning_rate": 2.633506858313127e-05, "loss": 0.0398, "step": 67410 }, { "grad_norm": 0.2665964663028717, "learning_rate": 2.6320504427981975e-05, "loss": 0.0366, "step": 67420 }, { "grad_norm": 0.3126240074634552, "learning_rate": 2.6305942862386734e-05, "loss": 0.0339, "step": 67430 }, { "grad_norm": 0.543601930141449, "learning_rate": 2.629138388793793e-05, "loss": 0.0344, "step": 67440 }, { "grad_norm": 0.39487263560295105, "learning_rate": 2.6276827506227743e-05, "loss": 0.0376, "step": 67450 }, { "grad_norm": 0.3526124060153961, "learning_rate": 2.6262273718848037e-05, "loss": 0.0329, "step": 67460 }, { "grad_norm": 0.37553727626800537, "learning_rate": 2.6247722527390407e-05, "loss": 0.0299, "step": 67470 }, { "grad_norm": 0.35532429814338684, "learning_rate": 2.6233173933446104e-05, "loss": 0.039, "step": 67480 }, { "grad_norm": 0.5925508737564087, "learning_rate": 2.6218627938606188e-05, "loss": 0.0358, "step": 67490 }, { "grad_norm": 0.35659554600715637, "learning_rate": 2.620408454446135e-05, "loss": 0.0407, "step": 67500 }, { "grad_norm": 0.24549631774425507, "learning_rate": 2.6189543752602046e-05, "loss": 0.0296, "step": 67510 }, { "grad_norm": 0.811522901058197, "learning_rate": 2.6175005564618444e-05, "loss": 0.0381, "step": 67520 }, { "grad_norm": 0.3497166037559509, "learning_rate": 2.6160469982100428e-05, "loss": 0.0381, "step": 67530 }, { "grad_norm": 0.5658373832702637, "learning_rate": 2.6145937006637555e-05, "loss": 0.0568, "step": 67540 }, { "grad_norm": 0.36212384700775146, "learning_rate": 2.6131406639819166e-05, "loss": 0.0388, "step": 67550 }, { "grad_norm": 0.39581364393234253, "learning_rate": 2.611687888323424e-05, "loss": 0.0308, "step": 67560 }, { "grad_norm": 0.5505203008651733, "learning_rate": 2.610235373847153e-05, "loss": 0.0406, "step": 67570 }, { "grad_norm": 0.5918718576431274, "learning_rate": 2.6087831207119494e-05, "loss": 0.0456, "step": 67580 }, { "grad_norm": 0.38442301750183105, "learning_rate": 2.60733112907663e-05, "loss": 0.045, "step": 67590 }, { "grad_norm": 0.5232082605361938, "learning_rate": 2.6058793990999786e-05, "loss": 0.0406, "step": 67600 }, { "grad_norm": 0.548121988773346, "learning_rate": 2.6044279309407593e-05, "loss": 0.0359, "step": 67610 }, { "grad_norm": 0.4723670184612274, "learning_rate": 2.6029767247576976e-05, "loss": 0.0434, "step": 67620 }, { "grad_norm": 0.48885223269462585, "learning_rate": 2.601525780709497e-05, "loss": 0.0312, "step": 67630 }, { "grad_norm": 0.35950520634651184, "learning_rate": 2.6000750989548318e-05, "loss": 0.0359, "step": 67640 }, { "grad_norm": 0.4114455282688141, "learning_rate": 2.5986246796523462e-05, "loss": 0.0279, "step": 67650 }, { "grad_norm": 0.3146851360797882, "learning_rate": 2.5971745229606546e-05, "loss": 0.0357, "step": 67660 }, { "grad_norm": 0.533349871635437, "learning_rate": 2.5957246290383442e-05, "loss": 0.0437, "step": 67670 }, { "grad_norm": 0.5879355072975159, "learning_rate": 2.594274998043975e-05, "loss": 0.0375, "step": 67680 }, { "grad_norm": 0.3345901370048523, "learning_rate": 2.5928256301360733e-05, "loss": 0.0337, "step": 67690 }, { "grad_norm": 3.7781200408935547, "learning_rate": 2.591376525473142e-05, "loss": 0.0317, "step": 67700 }, { "grad_norm": 0.4312129318714142, "learning_rate": 2.5899276842136523e-05, "loss": 0.0376, "step": 67710 }, { "grad_norm": 0.26383763551712036, "learning_rate": 2.5884791065160495e-05, "loss": 0.0391, "step": 67720 }, { "grad_norm": 0.4020245373249054, "learning_rate": 2.5870307925387436e-05, "loss": 0.0306, "step": 67730 }, { "grad_norm": 0.5634621381759644, "learning_rate": 2.585582742440125e-05, "loss": 0.029, "step": 67740 }, { "grad_norm": 0.48562827706336975, "learning_rate": 2.5841349563785455e-05, "loss": 0.0418, "step": 67750 }, { "grad_norm": 0.14908088743686676, "learning_rate": 2.5826874345123355e-05, "loss": 0.0299, "step": 67760 }, { "grad_norm": 0.33570826053619385, "learning_rate": 2.581240176999792e-05, "loss": 0.0289, "step": 67770 }, { "grad_norm": 0.6584428548812866, "learning_rate": 2.579793183999189e-05, "loss": 0.0366, "step": 67780 }, { "grad_norm": 0.48016348481178284, "learning_rate": 2.5783464556687627e-05, "loss": 0.0389, "step": 67790 }, { "grad_norm": 0.3115565776824951, "learning_rate": 2.576899992166729e-05, "loss": 0.0307, "step": 67800 }, { "grad_norm": 0.2582833170890808, "learning_rate": 2.5754537936512668e-05, "loss": 0.0286, "step": 67810 }, { "grad_norm": 0.6867373585700989, "learning_rate": 2.574007860280535e-05, "loss": 0.0473, "step": 67820 }, { "grad_norm": 0.33134227991104126, "learning_rate": 2.5725621922126518e-05, "loss": 0.0322, "step": 67830 }, { "grad_norm": 0.4367195665836334, "learning_rate": 2.571116789605722e-05, "loss": 0.035, "step": 67840 }, { "grad_norm": 0.2848113179206848, "learning_rate": 2.5696716526178054e-05, "loss": 0.0409, "step": 67850 }, { "grad_norm": 0.5048987865447998, "learning_rate": 2.568226781406946e-05, "loss": 0.0413, "step": 67860 }, { "grad_norm": 0.4343222379684448, "learning_rate": 2.5667821761311474e-05, "loss": 0.0484, "step": 67870 }, { "grad_norm": 0.4492608606815338, "learning_rate": 2.5653378369483917e-05, "loss": 0.0277, "step": 67880 }, { "grad_norm": 0.39382269978523254, "learning_rate": 2.5638937640166292e-05, "loss": 0.0305, "step": 67890 }, { "grad_norm": 0.35870787501335144, "learning_rate": 2.5624499574937842e-05, "loss": 0.0304, "step": 67900 }, { "grad_norm": 0.37010252475738525, "learning_rate": 2.5610064175377456e-05, "loss": 0.0357, "step": 67910 }, { "grad_norm": 0.3746695816516876, "learning_rate": 2.5595631443063782e-05, "loss": 0.0359, "step": 67920 }, { "grad_norm": 0.29390791058540344, "learning_rate": 2.5581201379575177e-05, "loss": 0.0322, "step": 67930 }, { "grad_norm": 0.3901103734970093, "learning_rate": 2.5566773986489655e-05, "loss": 0.0386, "step": 67940 }, { "grad_norm": 0.39899954199790955, "learning_rate": 2.5552349265385e-05, "loss": 0.0514, "step": 67950 }, { "grad_norm": 0.5162866115570068, "learning_rate": 2.5537927217838675e-05, "loss": 0.0274, "step": 67960 }, { "grad_norm": 0.4790458381175995, "learning_rate": 2.5523507845427864e-05, "loss": 0.0392, "step": 67970 }, { "grad_norm": 0.44951528310775757, "learning_rate": 2.5509091149729413e-05, "loss": 0.036, "step": 67980 }, { "grad_norm": 1.368597388267517, "learning_rate": 2.5494677132319962e-05, "loss": 0.0413, "step": 67990 }, { "grad_norm": 0.5704799890518188, "learning_rate": 2.548026579477575e-05, "loss": 0.041, "step": 68000 }, { "grad_norm": 0.43182262778282166, "learning_rate": 2.5465857138672834e-05, "loss": 0.036, "step": 68010 }, { "grad_norm": 0.404437780380249, "learning_rate": 2.5451451165586848e-05, "loss": 0.0428, "step": 68020 }, { "grad_norm": 0.44917163252830505, "learning_rate": 2.543704787709329e-05, "loss": 0.0316, "step": 68030 }, { "grad_norm": 0.8640981316566467, "learning_rate": 2.542264727476723e-05, "loss": 0.0303, "step": 68040 }, { "grad_norm": 0.36956310272216797, "learning_rate": 2.5408249360183535e-05, "loss": 0.049, "step": 68050 }, { "grad_norm": 0.7302695512771606, "learning_rate": 2.5393854134916694e-05, "loss": 0.0331, "step": 68060 }, { "grad_norm": 0.4405753016471863, "learning_rate": 2.5379461600540993e-05, "loss": 0.0383, "step": 68070 }, { "grad_norm": 0.2924351394176483, "learning_rate": 2.5365071758630317e-05, "loss": 0.0321, "step": 68080 }, { "grad_norm": 0.36598366498947144, "learning_rate": 2.535068461075839e-05, "loss": 0.0315, "step": 68090 }, { "grad_norm": 0.49849334359169006, "learning_rate": 2.5336300158498516e-05, "loss": 0.0364, "step": 68100 }, { "grad_norm": 0.6775906085968018, "learning_rate": 2.5321918403423793e-05, "loss": 0.0388, "step": 68110 }, { "grad_norm": 0.3789496123790741, "learning_rate": 2.5307539347106957e-05, "loss": 0.0327, "step": 68120 }, { "grad_norm": 0.45114803314208984, "learning_rate": 2.529316299112049e-05, "loss": 0.0365, "step": 68130 }, { "grad_norm": 0.5943936705589294, "learning_rate": 2.5278789337036567e-05, "loss": 0.0327, "step": 68140 }, { "grad_norm": 0.43533021211624146, "learning_rate": 2.52644183864271e-05, "loss": 0.0343, "step": 68150 }, { "grad_norm": 0.4964613616466522, "learning_rate": 2.525005014086363e-05, "loss": 0.0386, "step": 68160 }, { "grad_norm": 0.4673908054828644, "learning_rate": 2.5235684601917465e-05, "loss": 0.0449, "step": 68170 }, { "grad_norm": 0.4489840567111969, "learning_rate": 2.5221321771159618e-05, "loss": 0.0399, "step": 68180 }, { "grad_norm": 0.6762032508850098, "learning_rate": 2.5206961650160743e-05, "loss": 0.0409, "step": 68190 }, { "grad_norm": 0.2737928628921509, "learning_rate": 2.519260424049129e-05, "loss": 0.0354, "step": 68200 }, { "grad_norm": 0.3712921142578125, "learning_rate": 2.5178249543721304e-05, "loss": 0.0487, "step": 68210 }, { "grad_norm": 0.25291013717651367, "learning_rate": 2.5163897561420668e-05, "loss": 0.0324, "step": 68220 }, { "grad_norm": 0.5363196730613708, "learning_rate": 2.514954829515883e-05, "loss": 0.0463, "step": 68230 }, { "grad_norm": 0.413440465927124, "learning_rate": 2.513520174650505e-05, "loss": 0.039, "step": 68240 }, { "grad_norm": 0.622256875038147, "learning_rate": 2.5120857917028207e-05, "loss": 0.0347, "step": 68250 }, { "grad_norm": 0.4179241359233856, "learning_rate": 2.5106516808296953e-05, "loss": 0.0251, "step": 68260 }, { "grad_norm": 0.32453081011772156, "learning_rate": 2.5092178421879552e-05, "loss": 0.0378, "step": 68270 }, { "grad_norm": 0.4439229965209961, "learning_rate": 2.5077842759344116e-05, "loss": 0.0499, "step": 68280 }, { "grad_norm": 0.33990633487701416, "learning_rate": 2.5063509822258304e-05, "loss": 0.0386, "step": 68290 }, { "grad_norm": 0.6679622530937195, "learning_rate": 2.5049179612189576e-05, "loss": 0.0352, "step": 68300 }, { "grad_norm": 0.5196717977523804, "learning_rate": 2.5034852130705043e-05, "loss": 0.0423, "step": 68310 }, { "grad_norm": 0.38314002752304077, "learning_rate": 2.5020527379371572e-05, "loss": 0.0408, "step": 68320 }, { "grad_norm": 0.9274930357933044, "learning_rate": 2.500620535975564e-05, "loss": 0.0334, "step": 68330 }, { "grad_norm": 2.2074923515319824, "learning_rate": 2.4991886073423515e-05, "loss": 0.0429, "step": 68340 }, { "grad_norm": 0.27665629982948303, "learning_rate": 2.4977569521941134e-05, "loss": 0.03, "step": 68350 }, { "grad_norm": 0.33134615421295166, "learning_rate": 2.4963255706874143e-05, "loss": 0.0363, "step": 68360 }, { "grad_norm": 0.4863337576389313, "learning_rate": 2.4948944629787853e-05, "loss": 0.0381, "step": 68370 }, { "grad_norm": 0.5404918789863586, "learning_rate": 2.493463629224731e-05, "loss": 0.0451, "step": 68380 }, { "grad_norm": 0.5757125616073608, "learning_rate": 2.492033069581728e-05, "loss": 0.0392, "step": 68390 }, { "grad_norm": 0.3882896602153778, "learning_rate": 2.4906027842062137e-05, "loss": 0.0331, "step": 68400 }, { "grad_norm": 0.4449980556964874, "learning_rate": 2.48917277325461e-05, "loss": 0.0382, "step": 68410 }, { "grad_norm": 0.32007065415382385, "learning_rate": 2.487743036883295e-05, "loss": 0.0288, "step": 68420 }, { "grad_norm": 0.27553072571754456, "learning_rate": 2.486313575248626e-05, "loss": 0.0295, "step": 68430 }, { "grad_norm": 0.4800097346305847, "learning_rate": 2.4848843885069235e-05, "loss": 0.0369, "step": 68440 }, { "grad_norm": 0.4309438169002533, "learning_rate": 2.483455476814484e-05, "loss": 0.041, "step": 68450 }, { "grad_norm": 0.381570428609848, "learning_rate": 2.4820268403275672e-05, "loss": 0.0349, "step": 68460 }, { "grad_norm": 0.5231472849845886, "learning_rate": 2.480598479202413e-05, "loss": 0.0379, "step": 68470 }, { "grad_norm": 0.9612829685211182, "learning_rate": 2.4791703935952193e-05, "loss": 0.0437, "step": 68480 }, { "grad_norm": 0.4674387574195862, "learning_rate": 2.477742583662163e-05, "loss": 0.0375, "step": 68490 }, { "grad_norm": 0.4188739061355591, "learning_rate": 2.476315049559384e-05, "loss": 0.0304, "step": 68500 }, { "grad_norm": 0.6931771636009216, "learning_rate": 2.4748877914429987e-05, "loss": 0.0435, "step": 68510 }, { "grad_norm": 0.3195819556713104, "learning_rate": 2.4734608094690857e-05, "loss": 0.0409, "step": 68520 }, { "grad_norm": 0.2938946783542633, "learning_rate": 2.4720341037937006e-05, "loss": 0.0483, "step": 68530 }, { "grad_norm": 0.41605469584465027, "learning_rate": 2.4706076745728645e-05, "loss": 0.0419, "step": 68540 }, { "grad_norm": 0.35033631324768066, "learning_rate": 2.469181521962573e-05, "loss": 0.039, "step": 68550 }, { "grad_norm": 0.5081701278686523, "learning_rate": 2.467755646118783e-05, "loss": 0.0267, "step": 68560 }, { "grad_norm": 0.5822009444236755, "learning_rate": 2.46633004719743e-05, "loss": 0.0372, "step": 68570 }, { "grad_norm": 0.43320581316947937, "learning_rate": 2.4649047253544123e-05, "loss": 0.045, "step": 68580 }, { "grad_norm": 0.22841516137123108, "learning_rate": 2.4634796807456016e-05, "loss": 0.0311, "step": 68590 }, { "grad_norm": 0.33230477571487427, "learning_rate": 2.46205491352684e-05, "loss": 0.0366, "step": 68600 }, { "grad_norm": 0.28140679001808167, "learning_rate": 2.4606304238539375e-05, "loss": 0.0305, "step": 68610 }, { "grad_norm": 0.9154761433601379, "learning_rate": 2.4592062118826753e-05, "loss": 0.0378, "step": 68620 }, { "grad_norm": 0.42079949378967285, "learning_rate": 2.4577822777688004e-05, "loss": 0.0246, "step": 68630 }, { "grad_norm": 0.5162123441696167, "learning_rate": 2.4563586216680347e-05, "loss": 0.0358, "step": 68640 }, { "grad_norm": 0.3529819846153259, "learning_rate": 2.4549352437360622e-05, "loss": 0.0266, "step": 68650 }, { "grad_norm": 0.40594664216041565, "learning_rate": 2.4535121441285493e-05, "loss": 0.042, "step": 68660 }, { "grad_norm": 1.088731050491333, "learning_rate": 2.4520893230011172e-05, "loss": 0.0356, "step": 68670 }, { "grad_norm": 0.2544908821582794, "learning_rate": 2.450666780509368e-05, "loss": 0.033, "step": 68680 }, { "grad_norm": 0.4054378569126129, "learning_rate": 2.4492445168088656e-05, "loss": 0.0331, "step": 68690 }, { "grad_norm": 0.41718897223472595, "learning_rate": 2.447822532055149e-05, "loss": 0.0438, "step": 68700 }, { "grad_norm": 0.5671228170394897, "learning_rate": 2.4464008264037212e-05, "loss": 0.0398, "step": 68710 }, { "grad_norm": 0.5572513937950134, "learning_rate": 2.4449794000100605e-05, "loss": 0.0478, "step": 68720 }, { "grad_norm": 0.3297123610973358, "learning_rate": 2.4435582530296108e-05, "loss": 0.0356, "step": 68730 }, { "grad_norm": 0.6005448698997498, "learning_rate": 2.4421373856177887e-05, "loss": 0.0378, "step": 68740 }, { "grad_norm": 0.5827931761741638, "learning_rate": 2.440716797929975e-05, "loss": 0.0301, "step": 68750 }, { "grad_norm": 0.8443384766578674, "learning_rate": 2.439296490121526e-05, "loss": 0.0307, "step": 68760 }, { "grad_norm": 0.37415701150894165, "learning_rate": 2.4378764623477608e-05, "loss": 0.0317, "step": 68770 }, { "grad_norm": 0.3109992742538452, "learning_rate": 2.436456714763974e-05, "loss": 0.0325, "step": 68780 }, { "grad_norm": 0.4595986604690552, "learning_rate": 2.4350372475254262e-05, "loss": 0.0388, "step": 68790 }, { "grad_norm": 0.6818254590034485, "learning_rate": 2.433618060787351e-05, "loss": 0.0374, "step": 68800 }, { "grad_norm": 0.3158138394355774, "learning_rate": 2.432199154704944e-05, "loss": 0.0378, "step": 68810 }, { "grad_norm": 0.5088898539543152, "learning_rate": 2.430780529433377e-05, "loss": 0.0415, "step": 68820 }, { "grad_norm": 0.4012693762779236, "learning_rate": 2.42936218512779e-05, "loss": 0.0375, "step": 68830 }, { "grad_norm": 0.3121536374092102, "learning_rate": 2.427944121943288e-05, "loss": 0.0328, "step": 68840 }, { "grad_norm": 0.3830702602863312, "learning_rate": 2.4265263400349497e-05, "loss": 0.0384, "step": 68850 }, { "grad_norm": 0.5786198973655701, "learning_rate": 2.4251088395578213e-05, "loss": 0.038, "step": 68860 }, { "grad_norm": 0.666135847568512, "learning_rate": 2.4236916206669214e-05, "loss": 0.043, "step": 68870 }, { "grad_norm": 0.8257836699485779, "learning_rate": 2.4222746835172304e-05, "loss": 0.0457, "step": 68880 }, { "grad_norm": 0.5259509086608887, "learning_rate": 2.4208580282637065e-05, "loss": 0.0393, "step": 68890 }, { "grad_norm": 0.48847901821136475, "learning_rate": 2.4194416550612687e-05, "loss": 0.0338, "step": 68900 }, { "grad_norm": 0.26596716046333313, "learning_rate": 2.418025564064812e-05, "loss": 0.0303, "step": 68910 }, { "grad_norm": 0.6054120659828186, "learning_rate": 2.4166097554291978e-05, "loss": 0.0287, "step": 68920 }, { "grad_norm": 0.2506900727748871, "learning_rate": 2.4151942293092583e-05, "loss": 0.0304, "step": 68930 }, { "grad_norm": 0.19776318967342377, "learning_rate": 2.41377898585979e-05, "loss": 0.0283, "step": 68940 }, { "grad_norm": 0.5848667025566101, "learning_rate": 2.4123640252355652e-05, "loss": 0.0363, "step": 68950 }, { "grad_norm": 0.4451708197593689, "learning_rate": 2.410949347591318e-05, "loss": 0.0403, "step": 68960 }, { "grad_norm": 0.4526781439781189, "learning_rate": 2.4095349530817578e-05, "loss": 0.0557, "step": 68970 }, { "grad_norm": 0.4793946146965027, "learning_rate": 2.4081208418615603e-05, "loss": 0.0367, "step": 68980 }, { "grad_norm": 0.6572749018669128, "learning_rate": 2.4067070140853725e-05, "loss": 0.0475, "step": 68990 }, { "grad_norm": 0.30528122186660767, "learning_rate": 2.405293469907805e-05, "loss": 0.0378, "step": 69000 }, { "grad_norm": 0.53287672996521, "learning_rate": 2.4038802094834446e-05, "loss": 0.0365, "step": 69010 }, { "grad_norm": 0.31036892533302307, "learning_rate": 2.4024672329668395e-05, "loss": 0.0382, "step": 69020 }, { "grad_norm": 0.38906851410865784, "learning_rate": 2.4010545405125125e-05, "loss": 0.0291, "step": 69030 }, { "grad_norm": 0.6102487444877625, "learning_rate": 2.3996421322749528e-05, "loss": 0.0414, "step": 69040 }, { "grad_norm": 0.4253525733947754, "learning_rate": 2.3982300084086224e-05, "loss": 0.0377, "step": 69050 }, { "grad_norm": 0.5323421359062195, "learning_rate": 2.396818169067944e-05, "loss": 0.0322, "step": 69060 }, { "grad_norm": 0.3983986973762512, "learning_rate": 2.3954066144073178e-05, "loss": 0.0374, "step": 69070 }, { "grad_norm": 0.5902035236358643, "learning_rate": 2.39399534458111e-05, "loss": 0.042, "step": 69080 }, { "grad_norm": 0.3920324742794037, "learning_rate": 2.392584359743651e-05, "loss": 0.0403, "step": 69090 }, { "grad_norm": 0.4270356595516205, "learning_rate": 2.3911736600492463e-05, "loss": 0.0258, "step": 69100 }, { "grad_norm": 0.5832449793815613, "learning_rate": 2.3897632456521672e-05, "loss": 0.0335, "step": 69110 }, { "grad_norm": 0.3559775948524475, "learning_rate": 2.388353116706658e-05, "loss": 0.0443, "step": 69120 }, { "grad_norm": 0.4433788061141968, "learning_rate": 2.3869432733669232e-05, "loss": 0.0362, "step": 69130 }, { "grad_norm": 0.4476642906665802, "learning_rate": 2.385533715787145e-05, "loss": 0.0406, "step": 69140 }, { "grad_norm": 0.32491111755371094, "learning_rate": 2.3841244441214673e-05, "loss": 0.0292, "step": 69150 }, { "grad_norm": 0.5262219309806824, "learning_rate": 2.3827154585240073e-05, "loss": 0.0382, "step": 69160 }, { "grad_norm": 0.3747137486934662, "learning_rate": 2.3813067591488498e-05, "loss": 0.0484, "step": 69170 }, { "grad_norm": 0.7083399891853333, "learning_rate": 2.3798983461500494e-05, "loss": 0.0476, "step": 69180 }, { "grad_norm": 0.9525315165519714, "learning_rate": 2.3784902196816256e-05, "loss": 0.0433, "step": 69190 }, { "grad_norm": 0.4035129249095917, "learning_rate": 2.3770823798975716e-05, "loss": 0.0372, "step": 69200 }, { "grad_norm": 0.9505653381347656, "learning_rate": 2.3756748269518424e-05, "loss": 0.0389, "step": 69210 }, { "grad_norm": 0.3677304685115814, "learning_rate": 2.374267560998371e-05, "loss": 0.043, "step": 69220 }, { "grad_norm": 0.48200052976608276, "learning_rate": 2.3728605821910477e-05, "loss": 0.0352, "step": 69230 }, { "grad_norm": 0.7088202238082886, "learning_rate": 2.3714538906837453e-05, "loss": 0.0398, "step": 69240 }, { "grad_norm": 0.3963700234889984, "learning_rate": 2.3700474866302913e-05, "loss": 0.0342, "step": 69250 }, { "grad_norm": 0.33692970871925354, "learning_rate": 2.3686413701844917e-05, "loss": 0.0308, "step": 69260 }, { "grad_norm": 0.39231836795806885, "learning_rate": 2.367235541500114e-05, "loss": 0.0299, "step": 69270 }, { "grad_norm": 0.25255098938941956, "learning_rate": 2.3658300007308993e-05, "loss": 0.0344, "step": 69280 }, { "grad_norm": 0.4770420789718628, "learning_rate": 2.364424748030555e-05, "loss": 0.0297, "step": 69290 }, { "grad_norm": 0.47729477286338806, "learning_rate": 2.3630197835527595e-05, "loss": 0.041, "step": 69300 }, { "grad_norm": 0.41516217589378357, "learning_rate": 2.3616151074511546e-05, "loss": 0.0409, "step": 69310 }, { "grad_norm": 0.32775673270225525, "learning_rate": 2.3602107198793545e-05, "loss": 0.0373, "step": 69320 }, { "grad_norm": 0.4810989499092102, "learning_rate": 2.3588066209909427e-05, "loss": 0.0342, "step": 69330 }, { "grad_norm": 0.5569334030151367, "learning_rate": 2.3574028109394662e-05, "loss": 0.0352, "step": 69340 }, { "grad_norm": 0.2652190029621124, "learning_rate": 2.3559992898784454e-05, "loss": 0.0366, "step": 69350 }, { "grad_norm": 0.5119271278381348, "learning_rate": 2.354596057961367e-05, "loss": 0.0291, "step": 69360 }, { "grad_norm": 0.3780069351196289, "learning_rate": 2.353193115341688e-05, "loss": 0.0408, "step": 69370 }, { "grad_norm": 0.284055233001709, "learning_rate": 2.3517904621728294e-05, "loss": 0.0272, "step": 69380 }, { "grad_norm": 0.3381030559539795, "learning_rate": 2.3503880986081855e-05, "loss": 0.0356, "step": 69390 }, { "grad_norm": 0.4690668284893036, "learning_rate": 2.3489860248011136e-05, "loss": 0.037, "step": 69400 }, { "grad_norm": 0.3125351071357727, "learning_rate": 2.3475842409049464e-05, "loss": 0.0372, "step": 69410 }, { "grad_norm": 0.43061643838882446, "learning_rate": 2.346182747072975e-05, "loss": 0.0422, "step": 69420 }, { "grad_norm": 0.33831894397735596, "learning_rate": 2.3447815434584718e-05, "loss": 0.0309, "step": 69430 }, { "grad_norm": 0.6328626275062561, "learning_rate": 2.3433806302146656e-05, "loss": 0.0295, "step": 69440 }, { "grad_norm": 0.4919946491718292, "learning_rate": 2.3419800074947612e-05, "loss": 0.0384, "step": 69450 }, { "grad_norm": 0.49974143505096436, "learning_rate": 2.3405796754519248e-05, "loss": 0.0412, "step": 69460 }, { "grad_norm": 0.6312850117683411, "learning_rate": 2.3391796342392992e-05, "loss": 0.0415, "step": 69470 }, { "grad_norm": 0.3812008798122406, "learning_rate": 2.337779884009984e-05, "loss": 0.0402, "step": 69480 }, { "grad_norm": 0.2965957224369049, "learning_rate": 2.3363804249170624e-05, "loss": 0.0322, "step": 69490 }, { "grad_norm": 0.4608844816684723, "learning_rate": 2.3349812571135703e-05, "loss": 0.034, "step": 69500 }, { "grad_norm": 0.21793629229068756, "learning_rate": 2.333582380752523e-05, "loss": 0.0266, "step": 69510 }, { "grad_norm": 0.4314562678337097, "learning_rate": 2.3321837959868963e-05, "loss": 0.0284, "step": 69520 }, { "grad_norm": 0.5148586630821228, "learning_rate": 2.3307855029696383e-05, "loss": 0.032, "step": 69530 }, { "grad_norm": 0.43264344334602356, "learning_rate": 2.3293875018536648e-05, "loss": 0.042, "step": 69540 }, { "grad_norm": 0.4000200927257538, "learning_rate": 2.327989792791861e-05, "loss": 0.0305, "step": 69550 }, { "grad_norm": 0.8665324449539185, "learning_rate": 2.326592375937074e-05, "loss": 0.033, "step": 69560 }, { "grad_norm": 0.4448479413986206, "learning_rate": 2.3251952514421248e-05, "loss": 0.0301, "step": 69570 }, { "grad_norm": 0.49177441000938416, "learning_rate": 2.323798419459804e-05, "loss": 0.0336, "step": 69580 }, { "grad_norm": 0.37518852949142456, "learning_rate": 2.3224018801428625e-05, "loss": 0.0313, "step": 69590 }, { "grad_norm": 0.4539162516593933, "learning_rate": 2.3210056336440273e-05, "loss": 0.0252, "step": 69600 }, { "grad_norm": 0.36466020345687866, "learning_rate": 2.3196096801159844e-05, "loss": 0.0323, "step": 69610 }, { "grad_norm": 0.4963677227497101, "learning_rate": 2.3182140197114012e-05, "loss": 0.0357, "step": 69620 }, { "grad_norm": 0.43799328804016113, "learning_rate": 2.3168186525828985e-05, "loss": 0.044, "step": 69630 }, { "grad_norm": 0.4254468083381653, "learning_rate": 2.3154235788830757e-05, "loss": 0.0326, "step": 69640 }, { "grad_norm": 0.3312326967716217, "learning_rate": 2.3140287987644927e-05, "loss": 0.0401, "step": 69650 }, { "grad_norm": 0.3161779046058655, "learning_rate": 2.3126343123796833e-05, "loss": 0.0342, "step": 69660 }, { "grad_norm": 0.5031960010528564, "learning_rate": 2.3112401198811418e-05, "loss": 0.0336, "step": 69670 }, { "grad_norm": 0.5404233932495117, "learning_rate": 2.3098462214213418e-05, "loss": 0.0392, "step": 69680 }, { "grad_norm": 0.3210906982421875, "learning_rate": 2.308452617152712e-05, "loss": 0.0353, "step": 69690 }, { "grad_norm": 0.6861296892166138, "learning_rate": 2.3070593072276603e-05, "loss": 0.0385, "step": 69700 }, { "grad_norm": 0.3827893137931824, "learning_rate": 2.3056662917985518e-05, "loss": 0.0312, "step": 69710 }, { "grad_norm": 0.4078470468521118, "learning_rate": 2.3042735710177283e-05, "loss": 0.0428, "step": 69720 }, { "grad_norm": 0.28236353397369385, "learning_rate": 2.302881145037492e-05, "loss": 0.0284, "step": 69730 }, { "grad_norm": 0.2951601445674896, "learning_rate": 2.3014890140101192e-05, "loss": 0.0314, "step": 69740 }, { "grad_norm": 0.3270411789417267, "learning_rate": 2.30009717808785e-05, "loss": 0.0267, "step": 69750 }, { "grad_norm": 0.2803352475166321, "learning_rate": 2.2987056374228967e-05, "loss": 0.0336, "step": 69760 }, { "grad_norm": 0.30350983142852783, "learning_rate": 2.297314392167431e-05, "loss": 0.042, "step": 69770 }, { "grad_norm": 0.359563410282135, "learning_rate": 2.295923442473601e-05, "loss": 0.0333, "step": 69780 }, { "grad_norm": 0.40719515085220337, "learning_rate": 2.294532788493519e-05, "loss": 0.0288, "step": 69790 }, { "grad_norm": 0.5886693000793457, "learning_rate": 2.293142430379262e-05, "loss": 0.0363, "step": 69800 }, { "grad_norm": 0.6095537543296814, "learning_rate": 2.291752368282879e-05, "loss": 0.0372, "step": 69810 }, { "grad_norm": 0.3270566761493683, "learning_rate": 2.290362602356385e-05, "loss": 0.0325, "step": 69820 }, { "grad_norm": 0.829785943031311, "learning_rate": 2.288973132751765e-05, "loss": 0.0403, "step": 69830 }, { "grad_norm": 0.41360193490982056, "learning_rate": 2.287583959620965e-05, "loss": 0.0436, "step": 69840 }, { "grad_norm": 0.3163119852542877, "learning_rate": 2.2861950831159074e-05, "loss": 0.03, "step": 69850 }, { "grad_norm": 0.695131242275238, "learning_rate": 2.284806503388472e-05, "loss": 0.0336, "step": 69860 }, { "grad_norm": 0.4888617992401123, "learning_rate": 2.283418220590518e-05, "loss": 0.0347, "step": 69870 }, { "grad_norm": 0.711777925491333, "learning_rate": 2.282030234873862e-05, "loss": 0.0426, "step": 69880 }, { "grad_norm": 0.6454881429672241, "learning_rate": 2.2806425463902948e-05, "loss": 0.0241, "step": 69890 }, { "grad_norm": 0.48345229029655457, "learning_rate": 2.279255155291568e-05, "loss": 0.0393, "step": 69900 }, { "grad_norm": 0.2907424569129944, "learning_rate": 2.277868061729409e-05, "loss": 0.0267, "step": 69910 }, { "grad_norm": 0.485837459564209, "learning_rate": 2.276481265855504e-05, "loss": 0.0269, "step": 69920 }, { "grad_norm": 0.36619141697883606, "learning_rate": 2.2750947678215128e-05, "loss": 0.0308, "step": 69930 }, { "grad_norm": 0.622041642665863, "learning_rate": 2.2737085677790616e-05, "loss": 0.0304, "step": 69940 }, { "grad_norm": 0.3667881488800049, "learning_rate": 2.2723226658797443e-05, "loss": 0.0316, "step": 69950 }, { "grad_norm": 0.37358707189559937, "learning_rate": 2.270937062275117e-05, "loss": 0.0288, "step": 69960 }, { "grad_norm": 0.24810464680194855, "learning_rate": 2.2695517571167113e-05, "loss": 0.0269, "step": 69970 }, { "grad_norm": 0.3633517622947693, "learning_rate": 2.268166750556019e-05, "loss": 0.0451, "step": 69980 }, { "grad_norm": 0.21071994304656982, "learning_rate": 2.266782042744503e-05, "loss": 0.0255, "step": 69990 }, { "grad_norm": 0.5582051277160645, "learning_rate": 2.2653976338335935e-05, "loss": 0.0414, "step": 70000 }, { "grad_norm": 0.7788136005401611, "learning_rate": 2.2640135239746895e-05, "loss": 0.0398, "step": 70010 }, { "grad_norm": 0.6126223206520081, "learning_rate": 2.2626297133191515e-05, "loss": 0.0347, "step": 70020 }, { "grad_norm": 0.45680367946624756, "learning_rate": 2.261246202018312e-05, "loss": 0.0321, "step": 70030 }, { "grad_norm": 0.281281977891922, "learning_rate": 2.2598629902234718e-05, "loss": 0.0256, "step": 70040 }, { "grad_norm": 0.362781822681427, "learning_rate": 2.258480078085894e-05, "loss": 0.0343, "step": 70050 }, { "grad_norm": 0.9508247375488281, "learning_rate": 2.2570974657568126e-05, "loss": 0.0383, "step": 70060 }, { "grad_norm": 0.292711079120636, "learning_rate": 2.2557151533874286e-05, "loss": 0.0351, "step": 70070 }, { "grad_norm": 0.4164251685142517, "learning_rate": 2.2543331411289116e-05, "loss": 0.0296, "step": 70080 }, { "grad_norm": 0.38210439682006836, "learning_rate": 2.252951429132392e-05, "loss": 0.0436, "step": 70090 }, { "grad_norm": 0.49337995052337646, "learning_rate": 2.2515700175489762e-05, "loss": 0.0404, "step": 70100 }, { "grad_norm": 0.28995853662490845, "learning_rate": 2.2501889065297288e-05, "loss": 0.0296, "step": 70110 }, { "grad_norm": 0.33843910694122314, "learning_rate": 2.2488080962256887e-05, "loss": 0.0347, "step": 70120 }, { "grad_norm": 0.4445640444755554, "learning_rate": 2.2474275867878585e-05, "loss": 0.0463, "step": 70130 }, { "grad_norm": 0.4636029303073883, "learning_rate": 2.246047378367211e-05, "loss": 0.0502, "step": 70140 }, { "grad_norm": 0.3146933913230896, "learning_rate": 2.2446674711146798e-05, "loss": 0.0301, "step": 70150 }, { "grad_norm": 0.38869380950927734, "learning_rate": 2.2432878651811734e-05, "loss": 0.034, "step": 70160 }, { "grad_norm": 0.31942805647850037, "learning_rate": 2.2419085607175594e-05, "loss": 0.0399, "step": 70170 }, { "grad_norm": 0.3165651559829712, "learning_rate": 2.240529557874678e-05, "loss": 0.0324, "step": 70180 }, { "grad_norm": 0.4535594880580902, "learning_rate": 2.2391508568033358e-05, "loss": 0.0326, "step": 70190 }, { "grad_norm": 0.42887309193611145, "learning_rate": 2.237772457654307e-05, "loss": 0.0413, "step": 70200 }, { "grad_norm": 0.36083146929740906, "learning_rate": 2.236394360578327e-05, "loss": 0.0423, "step": 70210 }, { "grad_norm": 0.40491196513175964, "learning_rate": 2.2350165657261073e-05, "loss": 0.0379, "step": 70220 }, { "grad_norm": 0.487050324678421, "learning_rate": 2.2336390732483164e-05, "loss": 0.0369, "step": 70230 }, { "grad_norm": 0.6001296043395996, "learning_rate": 2.2322618832955976e-05, "loss": 0.045, "step": 70240 }, { "grad_norm": 0.5828686952590942, "learning_rate": 2.2308849960185578e-05, "loss": 0.0351, "step": 70250 }, { "grad_norm": 0.431253582239151, "learning_rate": 2.229508411567774e-05, "loss": 0.0346, "step": 70260 }, { "grad_norm": 0.3820704519748688, "learning_rate": 2.2281321300937823e-05, "loss": 0.0468, "step": 70270 }, { "grad_norm": 0.3530586063861847, "learning_rate": 2.2267561517470937e-05, "loss": 0.0318, "step": 70280 }, { "grad_norm": 0.8344801068305969, "learning_rate": 2.2253804766781845e-05, "loss": 0.0347, "step": 70290 }, { "grad_norm": 0.45066753029823303, "learning_rate": 2.2240051050374933e-05, "loss": 0.0341, "step": 70300 }, { "grad_norm": 0.5200132131576538, "learning_rate": 2.22263003697543e-05, "loss": 0.0363, "step": 70310 }, { "grad_norm": 0.3036136031150818, "learning_rate": 2.2212552726423702e-05, "loss": 0.0355, "step": 70320 }, { "grad_norm": 0.3282565772533417, "learning_rate": 2.2198808121886578e-05, "loss": 0.0369, "step": 70330 }, { "grad_norm": 0.3651430010795593, "learning_rate": 2.218506655764598e-05, "loss": 0.0319, "step": 70340 }, { "grad_norm": 1.1397496461868286, "learning_rate": 2.2171328035204704e-05, "loss": 0.0294, "step": 70350 }, { "grad_norm": 0.2463533729314804, "learning_rate": 2.2157592556065133e-05, "loss": 0.0284, "step": 70360 }, { "grad_norm": 0.417609840631485, "learning_rate": 2.2143860121729382e-05, "loss": 0.0283, "step": 70370 }, { "grad_norm": 0.27099713683128357, "learning_rate": 2.2130130733699206e-05, "loss": 0.0342, "step": 70380 }, { "grad_norm": 0.45819616317749023, "learning_rate": 2.2116404393476052e-05, "loss": 0.0378, "step": 70390 }, { "grad_norm": 0.5297889709472656, "learning_rate": 2.2102681102560967e-05, "loss": 0.0317, "step": 70400 }, { "grad_norm": 0.5791276097297668, "learning_rate": 2.208896086245476e-05, "loss": 0.0296, "step": 70410 }, { "grad_norm": 0.5668978095054626, "learning_rate": 2.2075243674657808e-05, "loss": 0.0374, "step": 70420 }, { "grad_norm": 0.36879536509513855, "learning_rate": 2.206152954067024e-05, "loss": 0.0265, "step": 70430 }, { "grad_norm": 0.27608153223991394, "learning_rate": 2.2047818461991758e-05, "loss": 0.0404, "step": 70440 }, { "grad_norm": 0.45952877402305603, "learning_rate": 2.2034110440121864e-05, "loss": 0.0317, "step": 70450 }, { "grad_norm": 0.3924961984157562, "learning_rate": 2.2020405476559585e-05, "loss": 0.0324, "step": 70460 }, { "grad_norm": 0.4117904007434845, "learning_rate": 2.2006703572803722e-05, "loss": 0.0355, "step": 70470 }, { "grad_norm": 0.2687905430793762, "learning_rate": 2.1993004730352646e-05, "loss": 0.0377, "step": 70480 }, { "grad_norm": 0.2504916191101074, "learning_rate": 2.1979308950704464e-05, "loss": 0.0254, "step": 70490 }, { "grad_norm": 0.29081350564956665, "learning_rate": 2.1965616235356923e-05, "loss": 0.0443, "step": 70500 }, { "grad_norm": 0.36818546056747437, "learning_rate": 2.1951926585807465e-05, "loss": 0.0295, "step": 70510 }, { "grad_norm": 0.2524694800376892, "learning_rate": 2.193824000355313e-05, "loss": 0.0361, "step": 70520 }, { "grad_norm": 0.5657184720039368, "learning_rate": 2.192455649009067e-05, "loss": 0.0395, "step": 70530 }, { "grad_norm": 0.4936959147453308, "learning_rate": 2.1910876046916523e-05, "loss": 0.0364, "step": 70540 }, { "grad_norm": 0.21360193192958832, "learning_rate": 2.1897198675526725e-05, "loss": 0.0286, "step": 70550 }, { "grad_norm": 0.4289056062698364, "learning_rate": 2.1883524377417024e-05, "loss": 0.032, "step": 70560 }, { "grad_norm": 0.4046931564807892, "learning_rate": 2.1869853154082825e-05, "loss": 0.0324, "step": 70570 }, { "grad_norm": 0.3724876642227173, "learning_rate": 2.1856185007019215e-05, "loss": 0.0355, "step": 70580 }, { "grad_norm": 0.4049377739429474, "learning_rate": 2.1842519937720874e-05, "loss": 0.0389, "step": 70590 }, { "grad_norm": 0.4612596333026886, "learning_rate": 2.182885794768224e-05, "loss": 0.0472, "step": 70600 }, { "grad_norm": 0.4096294939517975, "learning_rate": 2.1815199038397326e-05, "loss": 0.0385, "step": 70610 }, { "grad_norm": 0.634284257888794, "learning_rate": 2.1801543211359893e-05, "loss": 0.0348, "step": 70620 }, { "grad_norm": 0.34171703457832336, "learning_rate": 2.178789046806326e-05, "loss": 0.0396, "step": 70630 }, { "grad_norm": 0.5745972394943237, "learning_rate": 2.1774240810000547e-05, "loss": 0.0309, "step": 70640 }, { "grad_norm": 0.4590560495853424, "learning_rate": 2.17605942386644e-05, "loss": 0.0462, "step": 70650 }, { "grad_norm": 0.40545031428337097, "learning_rate": 2.174695075554723e-05, "loss": 0.0462, "step": 70660 }, { "grad_norm": 0.5015695691108704, "learning_rate": 2.173331036214103e-05, "loss": 0.0447, "step": 70670 }, { "grad_norm": 0.6254236698150635, "learning_rate": 2.171967305993753e-05, "loss": 0.0331, "step": 70680 }, { "grad_norm": 0.4439327120780945, "learning_rate": 2.1706038850428024e-05, "loss": 0.0348, "step": 70690 }, { "grad_norm": 0.4472942650318146, "learning_rate": 2.169240773510361e-05, "loss": 0.0267, "step": 70700 }, { "grad_norm": 0.440244197845459, "learning_rate": 2.1678779715454906e-05, "loss": 0.0416, "step": 70710 }, { "grad_norm": 0.3624189794063568, "learning_rate": 2.166515479297229e-05, "loss": 0.0332, "step": 70720 }, { "grad_norm": 0.4339163899421692, "learning_rate": 2.1651532969145732e-05, "loss": 0.0308, "step": 70730 }, { "grad_norm": 0.27111223340034485, "learning_rate": 2.1637914245464903e-05, "loss": 0.0394, "step": 70740 }, { "grad_norm": 0.37127742171287537, "learning_rate": 2.1624298623419133e-05, "loss": 0.0312, "step": 70750 }, { "grad_norm": 0.5660806894302368, "learning_rate": 2.1610686104497413e-05, "loss": 0.0338, "step": 70760 }, { "grad_norm": 0.56965571641922, "learning_rate": 2.1597076690188366e-05, "loss": 0.0374, "step": 70770 }, { "grad_norm": 0.3940293788909912, "learning_rate": 2.1583470381980315e-05, "loss": 0.044, "step": 70780 }, { "grad_norm": 0.44290655851364136, "learning_rate": 2.156986718136123e-05, "loss": 0.0412, "step": 70790 }, { "grad_norm": 0.377445787191391, "learning_rate": 2.155626708981871e-05, "loss": 0.0298, "step": 70800 }, { "grad_norm": 0.46984925866127014, "learning_rate": 2.1542670108840075e-05, "loss": 0.0369, "step": 70810 }, { "grad_norm": 0.5572429895401001, "learning_rate": 2.152907623991221e-05, "loss": 0.0397, "step": 70820 }, { "grad_norm": 0.3530787527561188, "learning_rate": 2.1515485484521802e-05, "loss": 0.0298, "step": 70830 }, { "grad_norm": 0.3725702464580536, "learning_rate": 2.150189784415506e-05, "loss": 0.0316, "step": 70840 }, { "grad_norm": 0.2842971980571747, "learning_rate": 2.148831332029794e-05, "loss": 0.0287, "step": 70850 }, { "grad_norm": 1.2060573101043701, "learning_rate": 2.147473191443599e-05, "loss": 0.0316, "step": 70860 }, { "grad_norm": 0.20833033323287964, "learning_rate": 2.1461153628054492e-05, "loss": 0.0291, "step": 70870 }, { "grad_norm": 0.3148657977581024, "learning_rate": 2.1447578462638284e-05, "loss": 0.0347, "step": 70880 }, { "grad_norm": 0.9379934668540955, "learning_rate": 2.143400641967201e-05, "loss": 0.0267, "step": 70890 }, { "grad_norm": 0.24672146141529083, "learning_rate": 2.1420437500639823e-05, "loss": 0.0272, "step": 70900 }, { "grad_norm": 0.8700012564659119, "learning_rate": 2.1406871707025646e-05, "loss": 0.0353, "step": 70910 }, { "grad_norm": 0.45151615142822266, "learning_rate": 2.1393309040312963e-05, "loss": 0.0468, "step": 70920 }, { "grad_norm": 0.4796859920024872, "learning_rate": 2.137974950198502e-05, "loss": 0.0362, "step": 70930 }, { "grad_norm": 0.8538246750831604, "learning_rate": 2.1366193093524617e-05, "loss": 0.0261, "step": 70940 }, { "grad_norm": 0.39046385884284973, "learning_rate": 2.1352639816414288e-05, "loss": 0.0332, "step": 70950 }, { "grad_norm": 0.30611512064933777, "learning_rate": 2.1339089672136202e-05, "loss": 0.0335, "step": 70960 }, { "grad_norm": 0.23841595649719238, "learning_rate": 2.1325542662172193e-05, "loss": 0.0311, "step": 70970 }, { "grad_norm": 0.4700140655040741, "learning_rate": 2.1311998788003712e-05, "loss": 0.0272, "step": 70980 }, { "grad_norm": 0.32976484298706055, "learning_rate": 2.1298458051111907e-05, "loss": 0.0356, "step": 70990 }, { "grad_norm": 1.055729866027832, "learning_rate": 2.1284920452977602e-05, "loss": 0.0307, "step": 71000 }, { "grad_norm": 0.6246014833450317, "learning_rate": 2.12713859950812e-05, "loss": 0.0373, "step": 71010 }, { "grad_norm": 0.3003600239753723, "learning_rate": 2.1257854678902832e-05, "loss": 0.031, "step": 71020 }, { "grad_norm": 0.517074704170227, "learning_rate": 2.124432650592227e-05, "loss": 0.0414, "step": 71030 }, { "grad_norm": 0.6838014721870422, "learning_rate": 2.1230801477618937e-05, "loss": 0.0429, "step": 71040 }, { "grad_norm": 0.33716946840286255, "learning_rate": 2.121727959547189e-05, "loss": 0.0333, "step": 71050 }, { "grad_norm": 0.2408987283706665, "learning_rate": 2.1203760860959886e-05, "loss": 0.0306, "step": 71060 }, { "grad_norm": 0.5191987752914429, "learning_rate": 2.119024527556127e-05, "loss": 0.0375, "step": 71070 }, { "grad_norm": 0.4201778471469879, "learning_rate": 2.117673284075415e-05, "loss": 0.0273, "step": 71080 }, { "grad_norm": 0.4115526080131531, "learning_rate": 2.1163223558016165e-05, "loss": 0.0299, "step": 71090 }, { "grad_norm": 0.3766961097717285, "learning_rate": 2.1149717428824722e-05, "loss": 0.0377, "step": 71100 }, { "grad_norm": 0.5999358892440796, "learning_rate": 2.1136214454656784e-05, "loss": 0.0419, "step": 71110 }, { "grad_norm": 0.888485848903656, "learning_rate": 2.1122714636989056e-05, "loss": 0.0377, "step": 71120 }, { "grad_norm": 0.9129758477210999, "learning_rate": 2.1109217977297823e-05, "loss": 0.0388, "step": 71130 }, { "grad_norm": 0.39209291338920593, "learning_rate": 2.1095724477059077e-05, "loss": 0.0264, "step": 71140 }, { "grad_norm": 0.42366302013397217, "learning_rate": 2.1082234137748434e-05, "loss": 0.0269, "step": 71150 }, { "grad_norm": 0.37521791458129883, "learning_rate": 2.106874696084122e-05, "loss": 0.0376, "step": 71160 }, { "grad_norm": 0.4156193435192108, "learning_rate": 2.1055262947812315e-05, "loss": 0.0364, "step": 71170 }, { "grad_norm": 0.31503260135650635, "learning_rate": 2.104178210013636e-05, "loss": 0.0462, "step": 71180 }, { "grad_norm": 0.6658599376678467, "learning_rate": 2.1028304419287554e-05, "loss": 0.0373, "step": 71190 }, { "grad_norm": 0.7437023520469666, "learning_rate": 2.101482990673982e-05, "loss": 0.0338, "step": 71200 }, { "grad_norm": 0.3905823230743408, "learning_rate": 2.1001358563966707e-05, "loss": 0.0401, "step": 71210 }, { "grad_norm": 1.134381651878357, "learning_rate": 2.098789039244144e-05, "loss": 0.0356, "step": 71220 }, { "grad_norm": 0.28006845712661743, "learning_rate": 2.097442539363685e-05, "loss": 0.0319, "step": 71230 }, { "grad_norm": 0.3613852262496948, "learning_rate": 2.0960963569025448e-05, "loss": 0.0337, "step": 71240 }, { "grad_norm": 0.307969868183136, "learning_rate": 2.0947504920079435e-05, "loss": 0.0269, "step": 71250 }, { "grad_norm": 0.3349522054195404, "learning_rate": 2.093404944827059e-05, "loss": 0.0313, "step": 71260 }, { "grad_norm": 0.5158030986785889, "learning_rate": 2.0920597155070394e-05, "loss": 0.0364, "step": 71270 }, { "grad_norm": 0.7109646201133728, "learning_rate": 2.090714804194997e-05, "loss": 0.0415, "step": 71280 }, { "grad_norm": 0.1998654454946518, "learning_rate": 2.0893702110380114e-05, "loss": 0.029, "step": 71290 }, { "grad_norm": 0.44651302695274353, "learning_rate": 2.0880259361831216e-05, "loss": 0.0351, "step": 71300 }, { "grad_norm": 0.4045860171318054, "learning_rate": 2.0866819797773397e-05, "loss": 0.0307, "step": 71310 }, { "grad_norm": 0.33917734026908875, "learning_rate": 2.085338341967634e-05, "loss": 0.0314, "step": 71320 }, { "grad_norm": 0.48777514696121216, "learning_rate": 2.083995022900946e-05, "loss": 0.0327, "step": 71330 }, { "grad_norm": 0.5071365833282471, "learning_rate": 2.0826520227241768e-05, "loss": 0.0297, "step": 71340 }, { "grad_norm": 0.4658636450767517, "learning_rate": 2.081309341584199e-05, "loss": 0.037, "step": 71350 }, { "grad_norm": 0.2547292411327362, "learning_rate": 2.0799669796278413e-05, "loss": 0.032, "step": 71360 }, { "grad_norm": 0.5048912167549133, "learning_rate": 2.078624937001906e-05, "loss": 0.0457, "step": 71370 }, { "grad_norm": 1.2802050113677979, "learning_rate": 2.0772832138531544e-05, "loss": 0.0604, "step": 71380 }, { "grad_norm": 0.36243221163749695, "learning_rate": 2.0759418103283158e-05, "loss": 0.0304, "step": 71390 }, { "grad_norm": 0.7594977617263794, "learning_rate": 2.0746007265740846e-05, "loss": 0.0304, "step": 71400 }, { "grad_norm": 0.679904580116272, "learning_rate": 2.0732599627371215e-05, "loss": 0.0442, "step": 71410 }, { "grad_norm": 0.3820117115974426, "learning_rate": 2.0719195189640468e-05, "loss": 0.0308, "step": 71420 }, { "grad_norm": 0.685255765914917, "learning_rate": 2.0705793954014528e-05, "loss": 0.0359, "step": 71430 }, { "grad_norm": 0.2969588339328766, "learning_rate": 2.06923959219589e-05, "loss": 0.0292, "step": 71440 }, { "grad_norm": 0.4743745028972626, "learning_rate": 2.0679001094938793e-05, "loss": 0.029, "step": 71450 }, { "grad_norm": 0.31300175189971924, "learning_rate": 2.066560947441904e-05, "loss": 0.0338, "step": 71460 }, { "grad_norm": 0.3789512515068054, "learning_rate": 2.0652221061864152e-05, "loss": 0.0267, "step": 71470 }, { "grad_norm": 0.42033764719963074, "learning_rate": 2.0638835858738226e-05, "loss": 0.0334, "step": 71480 }, { "grad_norm": 0.3206658363342285, "learning_rate": 2.0625453866505068e-05, "loss": 0.029, "step": 71490 }, { "grad_norm": 0.33003899455070496, "learning_rate": 2.0612075086628123e-05, "loss": 0.0337, "step": 71500 }, { "grad_norm": 0.4143100082874298, "learning_rate": 2.059869952057045e-05, "loss": 0.0327, "step": 71510 }, { "grad_norm": 0.4965742528438568, "learning_rate": 2.0585327169794793e-05, "loss": 0.0342, "step": 71520 }, { "grad_norm": 0.39574456214904785, "learning_rate": 2.0571958035763534e-05, "loss": 0.0302, "step": 71530 }, { "grad_norm": 0.5488799810409546, "learning_rate": 2.0558592119938715e-05, "loss": 0.0363, "step": 71540 }, { "grad_norm": 0.24200399219989777, "learning_rate": 2.054522942378198e-05, "loss": 0.0249, "step": 71550 }, { "grad_norm": 0.8250519037246704, "learning_rate": 2.053186994875469e-05, "loss": 0.0353, "step": 71560 }, { "grad_norm": 0.4656200706958771, "learning_rate": 2.0518513696317775e-05, "loss": 0.0415, "step": 71570 }, { "grad_norm": 0.3738733232021332, "learning_rate": 2.050516066793188e-05, "loss": 0.0291, "step": 71580 }, { "grad_norm": 0.2823602259159088, "learning_rate": 2.0491810865057276e-05, "loss": 0.0294, "step": 71590 }, { "grad_norm": 0.5510146617889404, "learning_rate": 2.0478464289153882e-05, "loss": 0.0337, "step": 71600 }, { "grad_norm": 0.5014829635620117, "learning_rate": 2.0465120941681232e-05, "loss": 0.03, "step": 71610 }, { "grad_norm": 0.39228910207748413, "learning_rate": 2.0451780824098566e-05, "loss": 0.035, "step": 71620 }, { "grad_norm": 0.27715927362442017, "learning_rate": 2.043844393786471e-05, "loss": 0.0339, "step": 71630 }, { "grad_norm": 0.9663355350494385, "learning_rate": 2.0425110284438193e-05, "loss": 0.0358, "step": 71640 }, { "grad_norm": 0.3333267271518707, "learning_rate": 2.041177986527712e-05, "loss": 0.0319, "step": 71650 }, { "grad_norm": 0.6090872883796692, "learning_rate": 2.0398452681839347e-05, "loss": 0.0283, "step": 71660 }, { "grad_norm": 0.36691394448280334, "learning_rate": 2.0385128735582272e-05, "loss": 0.0443, "step": 71670 }, { "grad_norm": 0.34705042839050293, "learning_rate": 2.0371808027963003e-05, "loss": 0.0317, "step": 71680 }, { "grad_norm": 0.43896493315696716, "learning_rate": 2.035849056043825e-05, "loss": 0.0374, "step": 71690 }, { "grad_norm": 0.6423818469047546, "learning_rate": 2.03451763344644e-05, "loss": 0.0334, "step": 71700 }, { "grad_norm": 0.5120998620986938, "learning_rate": 2.0331865351497482e-05, "loss": 0.0428, "step": 71710 }, { "grad_norm": 0.5105448365211487, "learning_rate": 2.0318557612993185e-05, "loss": 0.0364, "step": 71720 }, { "grad_norm": 0.4403060972690582, "learning_rate": 2.0305253120406782e-05, "loss": 0.0339, "step": 71730 }, { "grad_norm": 0.3201821446418762, "learning_rate": 2.029195187519325e-05, "loss": 0.0297, "step": 71740 }, { "grad_norm": 0.34823983907699585, "learning_rate": 2.0278653878807217e-05, "loss": 0.0314, "step": 71750 }, { "grad_norm": 0.2684277594089508, "learning_rate": 2.0265359132702892e-05, "loss": 0.0427, "step": 71760 }, { "grad_norm": 0.34677526354789734, "learning_rate": 2.0252067638334182e-05, "loss": 0.0292, "step": 71770 }, { "grad_norm": 0.38665008544921875, "learning_rate": 2.023877939715464e-05, "loss": 0.0448, "step": 71780 }, { "grad_norm": 0.4465872347354889, "learning_rate": 2.0225494410617448e-05, "loss": 0.0379, "step": 71790 }, { "grad_norm": 0.29126623272895813, "learning_rate": 2.02122126801754e-05, "loss": 0.0451, "step": 71800 }, { "grad_norm": 0.33850178122520447, "learning_rate": 2.0198934207281018e-05, "loss": 0.0372, "step": 71810 }, { "grad_norm": 0.32648754119873047, "learning_rate": 2.0185658993386357e-05, "loss": 0.0301, "step": 71820 }, { "grad_norm": 0.5463224649429321, "learning_rate": 2.0172387039943226e-05, "loss": 0.0347, "step": 71830 }, { "grad_norm": 0.22983026504516602, "learning_rate": 2.0159118348402968e-05, "loss": 0.0374, "step": 71840 }, { "grad_norm": 0.41188839077949524, "learning_rate": 2.0145852920216697e-05, "loss": 0.0334, "step": 71850 }, { "grad_norm": 0.5489484667778015, "learning_rate": 2.0132590756835045e-05, "loss": 0.0356, "step": 71860 }, { "grad_norm": 0.37243714928627014, "learning_rate": 2.0119331859708386e-05, "loss": 0.0444, "step": 71870 }, { "grad_norm": 0.7202033996582031, "learning_rate": 2.0106076230286654e-05, "loss": 0.0352, "step": 71880 }, { "grad_norm": 0.8189578056335449, "learning_rate": 2.0092823870019488e-05, "loss": 0.0338, "step": 71890 }, { "grad_norm": 0.2964939773082733, "learning_rate": 2.0079574780356115e-05, "loss": 0.0348, "step": 71900 }, { "grad_norm": 0.3820931613445282, "learning_rate": 2.0066328962745486e-05, "loss": 0.0353, "step": 71910 }, { "grad_norm": 0.251338928937912, "learning_rate": 2.00530864186361e-05, "loss": 0.0295, "step": 71920 }, { "grad_norm": 0.32504335045814514, "learning_rate": 2.003984714947618e-05, "loss": 0.0406, "step": 71930 }, { "grad_norm": 0.32790952920913696, "learning_rate": 2.0026611156713515e-05, "loss": 0.0409, "step": 71940 }, { "grad_norm": 0.3754132091999054, "learning_rate": 2.0013378441795584e-05, "loss": 0.0345, "step": 71950 }, { "grad_norm": 0.40491557121276855, "learning_rate": 2.0000149006169506e-05, "loss": 0.0301, "step": 71960 }, { "grad_norm": 0.41502681374549866, "learning_rate": 1.9986922851282048e-05, "loss": 0.0334, "step": 71970 }, { "grad_norm": 0.4029884338378906, "learning_rate": 1.997369997857956e-05, "loss": 0.0319, "step": 71980 }, { "grad_norm": 0.48329785466194153, "learning_rate": 1.9960480389508103e-05, "loss": 0.0315, "step": 71990 }, { "grad_norm": 0.3895886540412903, "learning_rate": 1.9947264085513363e-05, "loss": 0.0497, "step": 72000 }, { "grad_norm": 0.5038635730743408, "learning_rate": 1.993405106804062e-05, "loss": 0.0363, "step": 72010 }, { "grad_norm": 0.5632842183113098, "learning_rate": 1.9920841338534867e-05, "loss": 0.0402, "step": 72020 }, { "grad_norm": 0.5010889768600464, "learning_rate": 1.9907634898440646e-05, "loss": 0.027, "step": 72030 }, { "grad_norm": 0.28745320439338684, "learning_rate": 1.989443174920227e-05, "loss": 0.029, "step": 72040 }, { "grad_norm": 0.5155184864997864, "learning_rate": 1.988123189226355e-05, "loss": 0.0435, "step": 72050 }, { "grad_norm": 0.6858323216438293, "learning_rate": 1.986803532906805e-05, "loss": 0.0374, "step": 72060 }, { "grad_norm": 0.3199061453342438, "learning_rate": 1.985484206105889e-05, "loss": 0.0321, "step": 72070 }, { "grad_norm": 0.2831757962703705, "learning_rate": 1.9841652089678887e-05, "loss": 0.031, "step": 72080 }, { "grad_norm": 0.3034980297088623, "learning_rate": 1.9828465416370435e-05, "loss": 0.0453, "step": 72090 }, { "grad_norm": 0.462159663438797, "learning_rate": 1.9815282042575685e-05, "loss": 0.0339, "step": 72100 }, { "grad_norm": 0.22347457706928253, "learning_rate": 1.9802101969736285e-05, "loss": 0.0407, "step": 72110 }, { "grad_norm": 0.4449322521686554, "learning_rate": 1.9788925199293633e-05, "loss": 0.0381, "step": 72120 }, { "grad_norm": 0.2544189989566803, "learning_rate": 1.977575173268868e-05, "loss": 0.031, "step": 72130 }, { "grad_norm": 0.4546494781970978, "learning_rate": 1.9762581571362094e-05, "loss": 0.0367, "step": 72140 }, { "grad_norm": 0.8095129728317261, "learning_rate": 1.9749414716754112e-05, "loss": 0.0397, "step": 72150 }, { "grad_norm": 0.26098906993865967, "learning_rate": 1.973625117030465e-05, "loss": 0.0313, "step": 72160 }, { "grad_norm": 0.6780800819396973, "learning_rate": 1.972309093345326e-05, "loss": 0.0306, "step": 72170 }, { "grad_norm": 0.34743255376815796, "learning_rate": 1.970993400763915e-05, "loss": 0.0359, "step": 72180 }, { "grad_norm": 0.510452926158905, "learning_rate": 1.969678039430109e-05, "loss": 0.0421, "step": 72190 }, { "grad_norm": 0.4199560582637787, "learning_rate": 1.9683630094877564e-05, "loss": 0.0268, "step": 72200 }, { "grad_norm": 0.32380175590515137, "learning_rate": 1.9670483110806696e-05, "loss": 0.0315, "step": 72210 }, { "grad_norm": 0.36936700344085693, "learning_rate": 1.9657339443526172e-05, "loss": 0.0402, "step": 72220 }, { "grad_norm": 0.28536364436149597, "learning_rate": 1.964419909447339e-05, "loss": 0.0387, "step": 72230 }, { "grad_norm": 0.5073961019515991, "learning_rate": 1.963106206508536e-05, "loss": 0.0366, "step": 72240 }, { "grad_norm": 0.29698923230171204, "learning_rate": 1.9617928356798738e-05, "loss": 0.0283, "step": 72250 }, { "grad_norm": 0.3316662609577179, "learning_rate": 1.960479797104977e-05, "loss": 0.0318, "step": 72260 }, { "grad_norm": 0.881115734577179, "learning_rate": 1.9591670909274418e-05, "loss": 0.0373, "step": 72270 }, { "grad_norm": 0.21618640422821045, "learning_rate": 1.9578547172908184e-05, "loss": 0.0343, "step": 72280 }, { "grad_norm": 0.603603184223175, "learning_rate": 1.9565426763386325e-05, "loss": 0.0328, "step": 72290 }, { "grad_norm": 0.3312125504016876, "learning_rate": 1.9552309682143622e-05, "loss": 0.0358, "step": 72300 }, { "grad_norm": 0.24171237647533417, "learning_rate": 1.9539195930614574e-05, "loss": 0.0341, "step": 72310 }, { "grad_norm": 0.41124117374420166, "learning_rate": 1.952608551023324e-05, "loss": 0.0369, "step": 72320 }, { "grad_norm": 0.5282498598098755, "learning_rate": 1.9512978422433397e-05, "loss": 0.0338, "step": 72330 }, { "grad_norm": 0.27906373143196106, "learning_rate": 1.9499874668648376e-05, "loss": 0.0234, "step": 72340 }, { "grad_norm": 0.3338567316532135, "learning_rate": 1.9486774250311213e-05, "loss": 0.0278, "step": 72350 }, { "grad_norm": 0.6240041851997375, "learning_rate": 1.9473677168854537e-05, "loss": 0.046, "step": 72360 }, { "grad_norm": 0.3336770534515381, "learning_rate": 1.946058342571065e-05, "loss": 0.0391, "step": 72370 }, { "grad_norm": 0.31281253695487976, "learning_rate": 1.9447493022311424e-05, "loss": 0.0381, "step": 72380 }, { "grad_norm": 0.4203544557094574, "learning_rate": 1.9434405960088442e-05, "loss": 0.0299, "step": 72390 }, { "grad_norm": 0.4474563002586365, "learning_rate": 1.9421322240472857e-05, "loss": 0.0262, "step": 72400 }, { "grad_norm": 0.5601793527603149, "learning_rate": 1.9408241864895494e-05, "loss": 0.0347, "step": 72410 }, { "grad_norm": 0.4742075502872467, "learning_rate": 1.939516483478681e-05, "loss": 0.0305, "step": 72420 }, { "grad_norm": 0.8860805034637451, "learning_rate": 1.9382091151576902e-05, "loss": 0.0356, "step": 72430 }, { "grad_norm": 0.34209904074668884, "learning_rate": 1.9369020816695454e-05, "loss": 0.0255, "step": 72440 }, { "grad_norm": 0.37169763445854187, "learning_rate": 1.935595383157184e-05, "loss": 0.0331, "step": 72450 }, { "grad_norm": 0.4126783013343811, "learning_rate": 1.9342890197635057e-05, "loss": 0.0269, "step": 72460 }, { "grad_norm": 0.6445587873458862, "learning_rate": 1.9329829916313685e-05, "loss": 0.0327, "step": 72470 }, { "grad_norm": 0.30902934074401855, "learning_rate": 1.9316772989036013e-05, "loss": 0.0381, "step": 72480 }, { "grad_norm": 0.3886934518814087, "learning_rate": 1.930371941722991e-05, "loss": 0.0347, "step": 72490 }, { "grad_norm": 0.43732360005378723, "learning_rate": 1.929066920232292e-05, "loss": 0.0322, "step": 72500 }, { "grad_norm": 0.5082778334617615, "learning_rate": 1.9277622345742152e-05, "loss": 0.0342, "step": 72510 }, { "grad_norm": 0.37946245074272156, "learning_rate": 1.926457884891444e-05, "loss": 0.0289, "step": 72520 }, { "grad_norm": 0.5166339874267578, "learning_rate": 1.9251538713266147e-05, "loss": 0.0347, "step": 72530 }, { "grad_norm": 0.4820222556591034, "learning_rate": 1.9238501940223358e-05, "loss": 0.0295, "step": 72540 }, { "grad_norm": 0.5530960559844971, "learning_rate": 1.922546853121175e-05, "loss": 0.0392, "step": 72550 }, { "grad_norm": 0.3795310854911804, "learning_rate": 1.9212438487656647e-05, "loss": 0.032, "step": 72560 }, { "grad_norm": 0.2718299627304077, "learning_rate": 1.9199411810982965e-05, "loss": 0.0303, "step": 72570 }, { "grad_norm": 0.26212096214294434, "learning_rate": 1.918638850261532e-05, "loss": 0.0299, "step": 72580 }, { "grad_norm": 0.6043633222579956, "learning_rate": 1.9173368563977885e-05, "loss": 0.0339, "step": 72590 }, { "grad_norm": 0.2874279022216797, "learning_rate": 1.9160351996494513e-05, "loss": 0.034, "step": 72600 }, { "grad_norm": 0.3984951674938202, "learning_rate": 1.9147338801588677e-05, "loss": 0.0325, "step": 72610 }, { "grad_norm": 0.4894947409629822, "learning_rate": 1.913432898068351e-05, "loss": 0.036, "step": 72620 }, { "grad_norm": 0.3257395625114441, "learning_rate": 1.912132253520169e-05, "loss": 0.0417, "step": 72630 }, { "grad_norm": 0.4302758574485779, "learning_rate": 1.910831946656564e-05, "loss": 0.0395, "step": 72640 }, { "grad_norm": 0.37011268734931946, "learning_rate": 1.909531977619731e-05, "loss": 0.0456, "step": 72650 }, { "grad_norm": 0.333058625459671, "learning_rate": 1.908232346551834e-05, "loss": 0.0294, "step": 72660 }, { "grad_norm": 0.4369058907032013, "learning_rate": 1.9069330535949993e-05, "loss": 0.0426, "step": 72670 }, { "grad_norm": 0.5412392020225525, "learning_rate": 1.9056340988913167e-05, "loss": 0.0247, "step": 72680 }, { "grad_norm": 0.44650596380233765, "learning_rate": 1.9043354825828357e-05, "loss": 0.0337, "step": 72690 }, { "grad_norm": 0.18111376464366913, "learning_rate": 1.9030372048115724e-05, "loss": 0.0314, "step": 72700 }, { "grad_norm": 0.4124235510826111, "learning_rate": 1.9017392657195056e-05, "loss": 0.0376, "step": 72710 }, { "grad_norm": 0.4837549030780792, "learning_rate": 1.9004416654485724e-05, "loss": 0.0366, "step": 72720 }, { "grad_norm": 0.3942064642906189, "learning_rate": 1.8991444041406785e-05, "loss": 0.0362, "step": 72730 }, { "grad_norm": 0.3762359619140625, "learning_rate": 1.8978474819376912e-05, "loss": 0.0406, "step": 72740 }, { "grad_norm": 0.37666890025138855, "learning_rate": 1.8965508989814406e-05, "loss": 0.0357, "step": 72750 }, { "grad_norm": 0.36923450231552124, "learning_rate": 1.8952546554137158e-05, "loss": 0.0315, "step": 72760 }, { "grad_norm": 0.3954443037509918, "learning_rate": 1.8939587513762757e-05, "loss": 0.0274, "step": 72770 }, { "grad_norm": 0.23627401888370514, "learning_rate": 1.892663187010835e-05, "loss": 0.0243, "step": 72780 }, { "grad_norm": 0.2971550524234772, "learning_rate": 1.891367962459077e-05, "loss": 0.0362, "step": 72790 }, { "grad_norm": 0.5105854868888855, "learning_rate": 1.890073077862644e-05, "loss": 0.0413, "step": 72800 }, { "grad_norm": 0.3933841288089752, "learning_rate": 1.888778533363146e-05, "loss": 0.028, "step": 72810 }, { "grad_norm": 0.43118369579315186, "learning_rate": 1.8874843291021478e-05, "loss": 0.0331, "step": 72820 }, { "grad_norm": 0.43134215474128723, "learning_rate": 1.8861904652211853e-05, "loss": 0.0335, "step": 72830 }, { "grad_norm": 0.3845098316669464, "learning_rate": 1.8848969418617507e-05, "loss": 0.0321, "step": 72840 }, { "grad_norm": 0.4498218894004822, "learning_rate": 1.8836037591653044e-05, "loss": 0.0362, "step": 72850 }, { "grad_norm": 0.517379641532898, "learning_rate": 1.882310917273262e-05, "loss": 0.0393, "step": 72860 }, { "grad_norm": 0.4270981252193451, "learning_rate": 1.881018416327014e-05, "loss": 0.033, "step": 72870 }, { "grad_norm": 0.3856559991836548, "learning_rate": 1.8797262564679014e-05, "loss": 0.0353, "step": 72880 }, { "grad_norm": 0.4370626211166382, "learning_rate": 1.8784344378372355e-05, "loss": 0.0282, "step": 72890 }, { "grad_norm": 0.37383151054382324, "learning_rate": 1.8771429605762843e-05, "loss": 0.0302, "step": 72900 }, { "grad_norm": 0.3288102447986603, "learning_rate": 1.8758518248262835e-05, "loss": 0.036, "step": 72910 }, { "grad_norm": 0.3669873774051666, "learning_rate": 1.8745610307284305e-05, "loss": 0.0446, "step": 72920 }, { "grad_norm": 0.36513054370880127, "learning_rate": 1.8732705784238862e-05, "loss": 0.033, "step": 72930 }, { "grad_norm": 0.46410173177719116, "learning_rate": 1.8719804680537682e-05, "loss": 0.0344, "step": 72940 }, { "grad_norm": 0.5019000768661499, "learning_rate": 1.870690699759164e-05, "loss": 0.0312, "step": 72950 }, { "grad_norm": 0.4691159129142761, "learning_rate": 1.8694012736811213e-05, "loss": 0.0441, "step": 72960 }, { "grad_norm": 0.5711354613304138, "learning_rate": 1.8681121899606463e-05, "loss": 0.0256, "step": 72970 }, { "grad_norm": 0.4115159213542938, "learning_rate": 1.8668234487387144e-05, "loss": 0.0356, "step": 72980 }, { "grad_norm": 0.3232820928096771, "learning_rate": 1.8655350501562592e-05, "loss": 0.0369, "step": 72990 }, { "grad_norm": 0.6032773852348328, "learning_rate": 1.8642469943541796e-05, "loss": 0.0424, "step": 73000 }, { "grad_norm": 0.27783873677253723, "learning_rate": 1.8629592814733328e-05, "loss": 0.0281, "step": 73010 }, { "grad_norm": 0.7919202446937561, "learning_rate": 1.8616719116545438e-05, "loss": 0.0422, "step": 73020 }, { "grad_norm": 0.3375498652458191, "learning_rate": 1.8603848850385952e-05, "loss": 0.0274, "step": 73030 }, { "grad_norm": 0.5252235531806946, "learning_rate": 1.859098201766236e-05, "loss": 0.03, "step": 73040 }, { "grad_norm": 0.29132452607154846, "learning_rate": 1.8578118619781726e-05, "loss": 0.0282, "step": 73050 }, { "grad_norm": 0.19076305627822876, "learning_rate": 1.856525865815083e-05, "loss": 0.0237, "step": 73060 }, { "grad_norm": 0.3344949781894684, "learning_rate": 1.8552402134175966e-05, "loss": 0.0286, "step": 73070 }, { "grad_norm": 0.3027186393737793, "learning_rate": 1.8539549049263143e-05, "loss": 0.0323, "step": 73080 }, { "grad_norm": 0.3067774176597595, "learning_rate": 1.8526699404817917e-05, "loss": 0.0316, "step": 73090 }, { "grad_norm": 0.34673935174942017, "learning_rate": 1.8513853202245546e-05, "loss": 0.0274, "step": 73100 }, { "grad_norm": 0.4503212869167328, "learning_rate": 1.850101044295081e-05, "loss": 0.0352, "step": 73110 }, { "grad_norm": 0.3780038058757782, "learning_rate": 1.848817112833825e-05, "loss": 0.0222, "step": 73120 }, { "grad_norm": 0.4702514111995697, "learning_rate": 1.8475335259811905e-05, "loss": 0.0273, "step": 73130 }, { "grad_norm": 0.7326634526252747, "learning_rate": 1.8462502838775508e-05, "loss": 0.0389, "step": 73140 }, { "grad_norm": 0.30315759778022766, "learning_rate": 1.8449673866632378e-05, "loss": 0.0371, "step": 73150 }, { "grad_norm": 0.30037158727645874, "learning_rate": 1.8436848344785473e-05, "loss": 0.0413, "step": 73160 }, { "grad_norm": 0.3564680218696594, "learning_rate": 1.8424026274637378e-05, "loss": 0.0302, "step": 73170 }, { "grad_norm": 0.4072701036930084, "learning_rate": 1.8411207657590312e-05, "loss": 0.0407, "step": 73180 }, { "grad_norm": 0.29930445551872253, "learning_rate": 1.839839249504607e-05, "loss": 0.036, "step": 73190 }, { "grad_norm": 0.33638086915016174, "learning_rate": 1.8385580788406108e-05, "loss": 0.0368, "step": 73200 }, { "grad_norm": 0.42916935682296753, "learning_rate": 1.837277253907152e-05, "loss": 0.0342, "step": 73210 }, { "grad_norm": 0.3039076626300812, "learning_rate": 1.8359967748442957e-05, "loss": 0.035, "step": 73220 }, { "grad_norm": 0.3505226969718933, "learning_rate": 1.8347166417920774e-05, "loss": 0.0333, "step": 73230 }, { "grad_norm": 0.40179309248924255, "learning_rate": 1.833436854890484e-05, "loss": 0.037, "step": 73240 }, { "grad_norm": 0.6586581468582153, "learning_rate": 1.8321574142794796e-05, "loss": 0.0319, "step": 73250 }, { "grad_norm": 1.6291638612747192, "learning_rate": 1.8308783200989748e-05, "loss": 0.0308, "step": 73260 }, { "grad_norm": 0.7108073830604553, "learning_rate": 1.829599572488855e-05, "loss": 0.0361, "step": 73270 }, { "grad_norm": 0.23519229888916016, "learning_rate": 1.828321171588957e-05, "loss": 0.0304, "step": 73280 }, { "grad_norm": 0.57506263256073, "learning_rate": 1.827043117539089e-05, "loss": 0.0428, "step": 73290 }, { "grad_norm": 0.16499395668506622, "learning_rate": 1.8257654104790116e-05, "loss": 0.0239, "step": 73300 }, { "grad_norm": 0.4771544337272644, "learning_rate": 1.82448805054846e-05, "loss": 0.0297, "step": 73310 }, { "grad_norm": 0.3638787865638733, "learning_rate": 1.823211037887119e-05, "loss": 0.0421, "step": 73320 }, { "grad_norm": 1.0439324378967285, "learning_rate": 1.821934372634645e-05, "loss": 0.0401, "step": 73330 }, { "grad_norm": 0.3846302926540375, "learning_rate": 1.8206580549306472e-05, "loss": 0.0298, "step": 73340 }, { "grad_norm": 0.25536927580833435, "learning_rate": 1.819382084914707e-05, "loss": 0.033, "step": 73350 }, { "grad_norm": 0.37067362666130066, "learning_rate": 1.8181064627263577e-05, "loss": 0.0237, "step": 73360 }, { "grad_norm": 0.33928364515304565, "learning_rate": 1.8168311885051022e-05, "loss": 0.0381, "step": 73370 }, { "grad_norm": 0.46438488364219666, "learning_rate": 1.815556262390402e-05, "loss": 0.0317, "step": 73380 }, { "grad_norm": 0.4534183442592621, "learning_rate": 1.8142816845216826e-05, "loss": 0.0359, "step": 73390 }, { "grad_norm": 0.3178975284099579, "learning_rate": 1.8130074550383276e-05, "loss": 0.0306, "step": 73400 }, { "grad_norm": 0.33007195591926575, "learning_rate": 1.8117335740796848e-05, "loss": 0.0252, "step": 73410 }, { "grad_norm": 0.5347126722335815, "learning_rate": 1.810460041785067e-05, "loss": 0.0329, "step": 73420 }, { "grad_norm": 0.6561334133148193, "learning_rate": 1.809186858293742e-05, "loss": 0.0469, "step": 73430 }, { "grad_norm": 0.40285658836364746, "learning_rate": 1.8079140237449448e-05, "loss": 0.0272, "step": 73440 }, { "grad_norm": 0.6235220432281494, "learning_rate": 1.8066415382778707e-05, "loss": 0.0344, "step": 73450 }, { "grad_norm": 0.4026564657688141, "learning_rate": 1.8053694020316786e-05, "loss": 0.0325, "step": 73460 }, { "grad_norm": 0.31240764260292053, "learning_rate": 1.8040976151454837e-05, "loss": 0.0353, "step": 73470 }, { "grad_norm": 0.43261900544166565, "learning_rate": 1.8028261777583704e-05, "loss": 0.0439, "step": 73480 }, { "grad_norm": 0.4060170650482178, "learning_rate": 1.8015550900093758e-05, "loss": 0.0397, "step": 73490 }, { "grad_norm": 0.5063987970352173, "learning_rate": 1.800284352037512e-05, "loss": 0.0286, "step": 73500 }, { "grad_norm": 0.27545881271362305, "learning_rate": 1.799013963981738e-05, "loss": 0.0393, "step": 73510 }, { "grad_norm": 0.39945363998413086, "learning_rate": 1.797743925980987e-05, "loss": 0.0294, "step": 73520 }, { "grad_norm": 0.5305570363998413, "learning_rate": 1.7964742381741435e-05, "loss": 0.0355, "step": 73530 }, { "grad_norm": 0.7473589181900024, "learning_rate": 1.7952049007000627e-05, "loss": 0.0452, "step": 73540 }, { "grad_norm": 0.37231069803237915, "learning_rate": 1.7939359136975535e-05, "loss": 0.0267, "step": 73550 }, { "grad_norm": 0.39976727962493896, "learning_rate": 1.7926672773053932e-05, "loss": 0.0302, "step": 73560 }, { "grad_norm": 0.37206411361694336, "learning_rate": 1.791398991662317e-05, "loss": 0.033, "step": 73570 }, { "grad_norm": 0.24363049864768982, "learning_rate": 1.7901310569070244e-05, "loss": 0.0336, "step": 73580 }, { "grad_norm": 0.581592857837677, "learning_rate": 1.7888634731781724e-05, "loss": 0.0301, "step": 73590 }, { "grad_norm": 0.8384050130844116, "learning_rate": 1.7875962406143844e-05, "loss": 0.0256, "step": 73600 }, { "grad_norm": 0.34399768710136414, "learning_rate": 1.7863293593542403e-05, "loss": 0.04, "step": 73610 }, { "grad_norm": 0.3207768201828003, "learning_rate": 1.7850628295362853e-05, "loss": 0.0269, "step": 73620 }, { "grad_norm": 0.3410850167274475, "learning_rate": 1.7837966512990263e-05, "loss": 0.0356, "step": 73630 }, { "grad_norm": 0.3571338951587677, "learning_rate": 1.782530824780931e-05, "loss": 0.0336, "step": 73640 }, { "grad_norm": 0.3300788104534149, "learning_rate": 1.7812653501204263e-05, "loss": 0.0314, "step": 73650 }, { "grad_norm": 0.348217636346817, "learning_rate": 1.780000227455903e-05, "loss": 0.0343, "step": 73660 }, { "grad_norm": 0.5966875553131104, "learning_rate": 1.778735456925716e-05, "loss": 0.0355, "step": 73670 }, { "grad_norm": 0.2816707193851471, "learning_rate": 1.7774710386681737e-05, "loss": 0.0283, "step": 73680 }, { "grad_norm": 0.3135494887828827, "learning_rate": 1.7762069728215545e-05, "loss": 0.0356, "step": 73690 }, { "grad_norm": 0.3815406560897827, "learning_rate": 1.7749432595240932e-05, "loss": 0.0346, "step": 73700 }, { "grad_norm": 0.37935835123062134, "learning_rate": 1.773679898913991e-05, "loss": 0.041, "step": 73710 }, { "grad_norm": 0.47362127900123596, "learning_rate": 1.7724168911294015e-05, "loss": 0.0359, "step": 73720 }, { "grad_norm": 0.9359452128410339, "learning_rate": 1.7711542363084504e-05, "loss": 0.0307, "step": 73730 }, { "grad_norm": 0.32757896184921265, "learning_rate": 1.7698919345892158e-05, "loss": 0.0271, "step": 73740 }, { "grad_norm": 0.30078020691871643, "learning_rate": 1.768629986109743e-05, "loss": 0.0365, "step": 73750 }, { "grad_norm": 0.49797481298446655, "learning_rate": 1.7673683910080357e-05, "loss": 0.0289, "step": 73760 }, { "grad_norm": 0.2489888221025467, "learning_rate": 1.7661071494220638e-05, "loss": 0.0326, "step": 73770 }, { "grad_norm": 0.1796896606683731, "learning_rate": 1.7648462614897498e-05, "loss": 0.027, "step": 73780 }, { "grad_norm": 0.4614333510398865, "learning_rate": 1.7635857273489863e-05, "loss": 0.0272, "step": 73790 }, { "grad_norm": 0.6069244146347046, "learning_rate": 1.7623255471376198e-05, "loss": 0.0347, "step": 73800 }, { "grad_norm": 0.5256844758987427, "learning_rate": 1.7610657209934646e-05, "loss": 0.046, "step": 73810 }, { "grad_norm": 0.27373266220092773, "learning_rate": 1.7598062490542915e-05, "loss": 0.0352, "step": 73820 }, { "grad_norm": 0.34484779834747314, "learning_rate": 1.7585471314578377e-05, "loss": 0.0373, "step": 73830 }, { "grad_norm": 0.4453219771385193, "learning_rate": 1.7572883683417945e-05, "loss": 0.041, "step": 73840 }, { "grad_norm": 0.5181744694709778, "learning_rate": 1.7560299598438212e-05, "loss": 0.041, "step": 73850 }, { "grad_norm": 0.2578631341457367, "learning_rate": 1.7547719061015334e-05, "loss": 0.0311, "step": 73860 }, { "grad_norm": 0.2733179032802582, "learning_rate": 1.75351420725251e-05, "loss": 0.0269, "step": 73870 }, { "grad_norm": 0.297873318195343, "learning_rate": 1.7522568634342923e-05, "loss": 0.031, "step": 73880 }, { "grad_norm": 0.30145564675331116, "learning_rate": 1.7509998747843825e-05, "loss": 0.029, "step": 73890 }, { "grad_norm": 0.3064759075641632, "learning_rate": 1.7497432414402403e-05, "loss": 0.0334, "step": 73900 }, { "grad_norm": 0.48703742027282715, "learning_rate": 1.7484869635392903e-05, "loss": 0.0268, "step": 73910 }, { "grad_norm": 0.42637866735458374, "learning_rate": 1.747231041218919e-05, "loss": 0.0265, "step": 73920 }, { "grad_norm": 0.3841066360473633, "learning_rate": 1.7459754746164692e-05, "loss": 0.0321, "step": 73930 }, { "grad_norm": 1.172614574432373, "learning_rate": 1.744720263869249e-05, "loss": 0.0411, "step": 73940 }, { "grad_norm": 0.6170775294303894, "learning_rate": 1.7434654091145263e-05, "loss": 0.0294, "step": 73950 }, { "grad_norm": 0.42742806673049927, "learning_rate": 1.742210910489532e-05, "loss": 0.0286, "step": 73960 }, { "grad_norm": 0.15908238291740417, "learning_rate": 1.7409567681314532e-05, "loss": 0.0294, "step": 73970 }, { "grad_norm": 0.2679498791694641, "learning_rate": 1.7397029821774434e-05, "loss": 0.0342, "step": 73980 }, { "grad_norm": 0.5151761770248413, "learning_rate": 1.7384495527646126e-05, "loss": 0.0328, "step": 73990 }, { "grad_norm": 0.3280373513698578, "learning_rate": 1.737196480030035e-05, "loss": 0.0325, "step": 74000 }, { "grad_norm": 0.3728993535041809, "learning_rate": 1.7359437641107447e-05, "loss": 0.036, "step": 74010 }, { "grad_norm": 0.3777846395969391, "learning_rate": 1.7346914051437387e-05, "loss": 0.0274, "step": 74020 }, { "grad_norm": 0.3555569052696228, "learning_rate": 1.733439403265969e-05, "loss": 0.0282, "step": 74030 }, { "grad_norm": 0.37493476271629333, "learning_rate": 1.732187758614357e-05, "loss": 0.0407, "step": 74040 }, { "grad_norm": 0.3759414553642273, "learning_rate": 1.730936471325777e-05, "loss": 0.0321, "step": 74050 }, { "grad_norm": 0.285277783870697, "learning_rate": 1.7296855415370718e-05, "loss": 0.0254, "step": 74060 }, { "grad_norm": 0.45976051688194275, "learning_rate": 1.728434969385035e-05, "loss": 0.027, "step": 74070 }, { "grad_norm": 0.8194001913070679, "learning_rate": 1.7271847550064346e-05, "loss": 0.0328, "step": 74080 }, { "grad_norm": 0.3803919851779938, "learning_rate": 1.725934898537987e-05, "loss": 0.0349, "step": 74090 }, { "grad_norm": 0.588758647441864, "learning_rate": 1.7246854001163783e-05, "loss": 0.0247, "step": 74100 }, { "grad_norm": 0.48805904388427734, "learning_rate": 1.7234362598782476e-05, "loss": 0.0333, "step": 74110 }, { "grad_norm": 0.8091663122177124, "learning_rate": 1.722187477960202e-05, "loss": 0.0369, "step": 74120 }, { "grad_norm": 0.7435299754142761, "learning_rate": 1.7209390544988057e-05, "loss": 0.0367, "step": 74130 }, { "grad_norm": 0.32382136583328247, "learning_rate": 1.7196909896305856e-05, "loss": 0.0424, "step": 74140 }, { "grad_norm": 0.35459521412849426, "learning_rate": 1.718443283492026e-05, "loss": 0.0331, "step": 74150 }, { "grad_norm": 0.3620699346065521, "learning_rate": 1.7171959362195746e-05, "loss": 0.0361, "step": 74160 }, { "grad_norm": 0.23301342129707336, "learning_rate": 1.715948947949642e-05, "loss": 0.0335, "step": 74170 }, { "grad_norm": 0.36395540833473206, "learning_rate": 1.714702318818593e-05, "loss": 0.0414, "step": 74180 }, { "grad_norm": 0.2548806369304657, "learning_rate": 1.7134560489627592e-05, "loss": 0.0279, "step": 74190 }, { "grad_norm": 0.3867047429084778, "learning_rate": 1.712210138518431e-05, "loss": 0.0247, "step": 74200 }, { "grad_norm": 0.32638978958129883, "learning_rate": 1.71096458762186e-05, "loss": 0.0379, "step": 74210 }, { "grad_norm": 0.34791630506515503, "learning_rate": 1.7097193964092552e-05, "loss": 0.0421, "step": 74220 }, { "grad_norm": 0.3674801290035248, "learning_rate": 1.708474565016792e-05, "loss": 0.0335, "step": 74230 }, { "grad_norm": 0.37715837359428406, "learning_rate": 1.7072300935806e-05, "loss": 0.0353, "step": 74240 }, { "grad_norm": 0.4789426624774933, "learning_rate": 1.705985982236776e-05, "loss": 0.0339, "step": 74250 }, { "grad_norm": 0.39602863788604736, "learning_rate": 1.7047422311213697e-05, "loss": 0.0264, "step": 74260 }, { "grad_norm": 0.49162524938583374, "learning_rate": 1.703498840370401e-05, "loss": 0.037, "step": 74270 }, { "grad_norm": 0.4529948830604553, "learning_rate": 1.702255810119842e-05, "loss": 0.0274, "step": 74280 }, { "grad_norm": 0.40782788395881653, "learning_rate": 1.7010131405056305e-05, "loss": 0.0262, "step": 74290 }, { "grad_norm": 0.2879769802093506, "learning_rate": 1.6997708316636613e-05, "loss": 0.0289, "step": 74300 }, { "grad_norm": 0.3644479811191559, "learning_rate": 1.6985288837297932e-05, "loss": 0.0331, "step": 74310 }, { "grad_norm": 0.6063867807388306, "learning_rate": 1.6972872968398394e-05, "loss": 0.0346, "step": 74320 }, { "grad_norm": 0.3817354440689087, "learning_rate": 1.6960460711295845e-05, "loss": 0.0334, "step": 74330 }, { "grad_norm": 0.3522472381591797, "learning_rate": 1.6948052067347623e-05, "loss": 0.0263, "step": 74340 }, { "grad_norm": 0.5371964573860168, "learning_rate": 1.693564703791075e-05, "loss": 0.0318, "step": 74350 }, { "grad_norm": 0.184176966547966, "learning_rate": 1.6923245624341793e-05, "loss": 0.0357, "step": 74360 }, { "grad_norm": 0.32363229990005493, "learning_rate": 1.6910847827996962e-05, "loss": 0.0259, "step": 74370 }, { "grad_norm": 0.3431397080421448, "learning_rate": 1.689845365023206e-05, "loss": 0.0273, "step": 74380 }, { "grad_norm": 0.22409699857234955, "learning_rate": 1.6886063092402505e-05, "loss": 0.0274, "step": 74390 }, { "grad_norm": 2.3421247005462646, "learning_rate": 1.6873676155863317e-05, "loss": 0.047, "step": 74400 }, { "grad_norm": 0.26053479313850403, "learning_rate": 1.6861292841969085e-05, "loss": 0.0275, "step": 74410 }, { "grad_norm": 0.4774027466773987, "learning_rate": 1.684891315207406e-05, "loss": 0.0315, "step": 74420 }, { "grad_norm": 0.3992598354816437, "learning_rate": 1.6836537087532034e-05, "loss": 0.0324, "step": 74430 }, { "grad_norm": 0.28147855401039124, "learning_rate": 1.6824164649696465e-05, "loss": 0.0315, "step": 74440 }, { "grad_norm": 0.7703085541725159, "learning_rate": 1.681179583992034e-05, "loss": 0.0286, "step": 74450 }, { "grad_norm": 0.31842318177223206, "learning_rate": 1.679943065955636e-05, "loss": 0.027, "step": 74460 }, { "grad_norm": 0.5746163725852966, "learning_rate": 1.678706910995671e-05, "loss": 0.0278, "step": 74470 }, { "grad_norm": 0.3282768726348877, "learning_rate": 1.6774711192473262e-05, "loss": 0.0313, "step": 74480 }, { "grad_norm": 0.3634437024593353, "learning_rate": 1.676235690845743e-05, "loss": 0.0257, "step": 74490 }, { "grad_norm": 0.49045300483703613, "learning_rate": 1.6750006259260285e-05, "loss": 0.0375, "step": 74500 }, { "grad_norm": 0.5623490214347839, "learning_rate": 1.673765924623244e-05, "loss": 0.0407, "step": 74510 }, { "grad_norm": 0.3550797998905182, "learning_rate": 1.67253158707242e-05, "loss": 0.0242, "step": 74520 }, { "grad_norm": 0.46432745456695557, "learning_rate": 1.671297613408537e-05, "loss": 0.0299, "step": 74530 }, { "grad_norm": 0.7322523593902588, "learning_rate": 1.6700640037665444e-05, "loss": 0.0364, "step": 74540 }, { "grad_norm": 0.4924595057964325, "learning_rate": 1.668830758281344e-05, "loss": 0.0341, "step": 74550 }, { "grad_norm": 0.34306129813194275, "learning_rate": 1.6675978770878052e-05, "loss": 0.0307, "step": 74560 }, { "grad_norm": 0.27947330474853516, "learning_rate": 1.6663653603207506e-05, "loss": 0.0304, "step": 74570 }, { "grad_norm": 0.3884986639022827, "learning_rate": 1.6651332081149685e-05, "loss": 0.0384, "step": 74580 }, { "grad_norm": 0.5912228226661682, "learning_rate": 1.6639014206052046e-05, "loss": 0.0291, "step": 74590 }, { "grad_norm": 0.3110737204551697, "learning_rate": 1.6626699979261657e-05, "loss": 0.0265, "step": 74600 }, { "grad_norm": 0.37522467970848083, "learning_rate": 1.6614389402125197e-05, "loss": 0.0396, "step": 74610 }, { "grad_norm": 0.28552302718162537, "learning_rate": 1.66020824759889e-05, "loss": 0.0246, "step": 74620 }, { "grad_norm": 0.4539881944656372, "learning_rate": 1.6589779202198675e-05, "loss": 0.0298, "step": 74630 }, { "grad_norm": 0.5238872766494751, "learning_rate": 1.6577479582099924e-05, "loss": 0.0379, "step": 74640 }, { "grad_norm": 0.611741840839386, "learning_rate": 1.6565183617037788e-05, "loss": 0.0406, "step": 74650 }, { "grad_norm": 0.39279401302337646, "learning_rate": 1.6552891308356894e-05, "loss": 0.0315, "step": 74660 }, { "grad_norm": 0.5079959034919739, "learning_rate": 1.6540602657401526e-05, "loss": 0.0402, "step": 74670 }, { "grad_norm": 0.3512642979621887, "learning_rate": 1.6528317665515537e-05, "loss": 0.042, "step": 74680 }, { "grad_norm": 0.9606490731239319, "learning_rate": 1.6516036334042416e-05, "loss": 0.0481, "step": 74690 }, { "grad_norm": 0.3385048806667328, "learning_rate": 1.650375866432519e-05, "loss": 0.0479, "step": 74700 }, { "grad_norm": 0.36457934975624084, "learning_rate": 1.649148465770659e-05, "loss": 0.0372, "step": 74710 }, { "grad_norm": 0.6815648078918457, "learning_rate": 1.647921431552883e-05, "loss": 0.0411, "step": 74720 }, { "grad_norm": 1.6505767107009888, "learning_rate": 1.646694763913381e-05, "loss": 0.0398, "step": 74730 }, { "grad_norm": 0.3059283196926117, "learning_rate": 1.6454684629862966e-05, "loss": 0.0325, "step": 74740 }, { "grad_norm": 0.724825382232666, "learning_rate": 1.6442425289057388e-05, "loss": 0.0303, "step": 74750 }, { "grad_norm": 0.3830198645591736, "learning_rate": 1.6430169618057718e-05, "loss": 0.0323, "step": 74760 }, { "grad_norm": 0.19371503591537476, "learning_rate": 1.641791761820422e-05, "loss": 0.0277, "step": 74770 }, { "grad_norm": 0.3742530345916748, "learning_rate": 1.640566929083676e-05, "loss": 0.031, "step": 74780 }, { "grad_norm": 0.5581526756286621, "learning_rate": 1.6393424637294823e-05, "loss": 0.0385, "step": 74790 }, { "grad_norm": 0.36077064275741577, "learning_rate": 1.638118365891742e-05, "loss": 0.0287, "step": 74800 }, { "grad_norm": 0.6116725206375122, "learning_rate": 1.636894635704323e-05, "loss": 0.0371, "step": 74810 }, { "grad_norm": 0.3006126880645752, "learning_rate": 1.6356712733010517e-05, "loss": 0.0353, "step": 74820 }, { "grad_norm": 0.7759559750556946, "learning_rate": 1.6344482788157096e-05, "loss": 0.0306, "step": 74830 }, { "grad_norm": 0.3882923722267151, "learning_rate": 1.6332256523820444e-05, "loss": 0.0265, "step": 74840 }, { "grad_norm": 0.8136546611785889, "learning_rate": 1.6320033941337597e-05, "loss": 0.0401, "step": 74850 }, { "grad_norm": 0.43436846137046814, "learning_rate": 1.6307815042045216e-05, "loss": 0.0308, "step": 74860 }, { "grad_norm": 0.493198424577713, "learning_rate": 1.6295599827279513e-05, "loss": 0.0297, "step": 74870 }, { "grad_norm": 0.42551881074905396, "learning_rate": 1.6283388298376356e-05, "loss": 0.0312, "step": 74880 }, { "grad_norm": 0.3293335437774658, "learning_rate": 1.6271180456671125e-05, "loss": 0.0306, "step": 74890 }, { "grad_norm": 0.33861878514289856, "learning_rate": 1.625897630349893e-05, "loss": 0.0257, "step": 74900 }, { "grad_norm": 0.39260461926460266, "learning_rate": 1.6246775840194335e-05, "loss": 0.0313, "step": 74910 }, { "grad_norm": 0.2539617419242859, "learning_rate": 1.6234579068091604e-05, "loss": 0.0254, "step": 74920 }, { "grad_norm": 0.5329747200012207, "learning_rate": 1.6222385988524524e-05, "loss": 0.0339, "step": 74930 }, { "grad_norm": 0.4007266163825989, "learning_rate": 1.6210196602826544e-05, "loss": 0.0351, "step": 74940 }, { "grad_norm": 0.22121089696884155, "learning_rate": 1.6198010912330646e-05, "loss": 0.0317, "step": 74950 }, { "grad_norm": 0.5358795523643494, "learning_rate": 1.6185828918369456e-05, "loss": 0.0368, "step": 74960 }, { "grad_norm": 0.5923806428909302, "learning_rate": 1.617365062227517e-05, "loss": 0.0319, "step": 74970 }, { "grad_norm": 0.24733597040176392, "learning_rate": 1.616147602537962e-05, "loss": 0.0414, "step": 74980 }, { "grad_norm": 0.3101884722709656, "learning_rate": 1.614930512901415e-05, "loss": 0.0309, "step": 74990 }, { "grad_norm": 0.40587592124938965, "learning_rate": 1.6137137934509804e-05, "loss": 0.0225, "step": 75000 }, { "grad_norm": 0.3240866959095001, "learning_rate": 1.6124974443197123e-05, "loss": 0.0315, "step": 75010 }, { "grad_norm": 0.3381856381893158, "learning_rate": 1.6112814656406306e-05, "loss": 0.0369, "step": 75020 }, { "grad_norm": 0.46242421865463257, "learning_rate": 1.6100658575467125e-05, "loss": 0.0372, "step": 75030 }, { "grad_norm": 0.4039727449417114, "learning_rate": 1.6088506201708974e-05, "loss": 0.0262, "step": 75040 }, { "grad_norm": 0.5103972554206848, "learning_rate": 1.6076357536460785e-05, "loss": 0.0379, "step": 75050 }, { "grad_norm": 0.36920395493507385, "learning_rate": 1.606421258105113e-05, "loss": 0.0272, "step": 75060 }, { "grad_norm": 0.3883899450302124, "learning_rate": 1.6052071336808183e-05, "loss": 0.034, "step": 75070 }, { "grad_norm": 0.3187628388404846, "learning_rate": 1.6039933805059658e-05, "loss": 0.0373, "step": 75080 }, { "grad_norm": 0.37342238426208496, "learning_rate": 1.6027799987132915e-05, "loss": 0.0378, "step": 75090 }, { "grad_norm": 0.3307923376560211, "learning_rate": 1.6015669884354888e-05, "loss": 0.0425, "step": 75100 }, { "grad_norm": 0.45659223198890686, "learning_rate": 1.6003543498052125e-05, "loss": 0.0362, "step": 75110 }, { "grad_norm": 0.5577834248542786, "learning_rate": 1.5991420829550715e-05, "loss": 0.0312, "step": 75120 }, { "grad_norm": 1.1129517555236816, "learning_rate": 1.5979301880176405e-05, "loss": 0.0439, "step": 75130 }, { "grad_norm": 0.4385179579257965, "learning_rate": 1.5967186651254485e-05, "loss": 0.0351, "step": 75140 }, { "grad_norm": 0.5039545893669128, "learning_rate": 1.5955075144109853e-05, "loss": 0.0321, "step": 75150 }, { "grad_norm": 0.45372095704078674, "learning_rate": 1.594296736006703e-05, "loss": 0.0313, "step": 75160 }, { "grad_norm": 0.25480392575263977, "learning_rate": 1.59308633004501e-05, "loss": 0.0389, "step": 75170 }, { "grad_norm": 0.4221808910369873, "learning_rate": 1.5918762966582724e-05, "loss": 0.0397, "step": 75180 }, { "grad_norm": 0.45260128378868103, "learning_rate": 1.5906666359788203e-05, "loss": 0.0354, "step": 75190 }, { "grad_norm": 0.26342058181762695, "learning_rate": 1.589457348138938e-05, "loss": 0.0293, "step": 75200 }, { "grad_norm": 0.4594433307647705, "learning_rate": 1.588248433270872e-05, "loss": 0.0351, "step": 75210 }, { "grad_norm": 0.4667721390724182, "learning_rate": 1.587039891506828e-05, "loss": 0.0294, "step": 75220 }, { "grad_norm": 0.24245287477970123, "learning_rate": 1.5858317229789715e-05, "loss": 0.0346, "step": 75230 }, { "grad_norm": 0.5262159109115601, "learning_rate": 1.584623927819423e-05, "loss": 0.0284, "step": 75240 }, { "grad_norm": 0.731086254119873, "learning_rate": 1.583416506160269e-05, "loss": 0.0421, "step": 75250 }, { "grad_norm": 0.4356614947319031, "learning_rate": 1.5822094581335472e-05, "loss": 0.0397, "step": 75260 }, { "grad_norm": 0.24439728260040283, "learning_rate": 1.5810027838712604e-05, "loss": 0.0365, "step": 75270 }, { "grad_norm": 0.3870351314544678, "learning_rate": 1.5797964835053687e-05, "loss": 0.0296, "step": 75280 }, { "grad_norm": 0.21145711839199066, "learning_rate": 1.5785905571677933e-05, "loss": 0.0289, "step": 75290 }, { "grad_norm": 0.5596063137054443, "learning_rate": 1.5773850049904086e-05, "loss": 0.0338, "step": 75300 }, { "grad_norm": 0.4636214077472687, "learning_rate": 1.5761798271050543e-05, "loss": 0.0366, "step": 75310 }, { "grad_norm": 0.4082252085208893, "learning_rate": 1.5749750236435277e-05, "loss": 0.0316, "step": 75320 }, { "grad_norm": 0.6046430468559265, "learning_rate": 1.5737705947375815e-05, "loss": 0.0326, "step": 75330 }, { "grad_norm": 0.4065287113189697, "learning_rate": 1.5725665405189322e-05, "loss": 0.0347, "step": 75340 }, { "grad_norm": 0.7049311399459839, "learning_rate": 1.571362861119253e-05, "loss": 0.0337, "step": 75350 }, { "grad_norm": 0.3031873404979706, "learning_rate": 1.570159556670177e-05, "loss": 0.0312, "step": 75360 }, { "grad_norm": 0.2524651288986206, "learning_rate": 1.5689566273032946e-05, "loss": 0.0331, "step": 75370 }, { "grad_norm": 0.3966018259525299, "learning_rate": 1.5677540731501583e-05, "loss": 0.0374, "step": 75380 }, { "grad_norm": 0.38418054580688477, "learning_rate": 1.5665518943422746e-05, "loss": 0.0314, "step": 75390 }, { "grad_norm": 0.3868996500968933, "learning_rate": 1.5653500910111134e-05, "loss": 0.0349, "step": 75400 }, { "grad_norm": 0.385625958442688, "learning_rate": 1.5641486632881023e-05, "loss": 0.0332, "step": 75410 }, { "grad_norm": 0.48782214522361755, "learning_rate": 1.56294761130463e-05, "loss": 0.0296, "step": 75420 }, { "grad_norm": 0.3928978145122528, "learning_rate": 1.5617469351920373e-05, "loss": 0.0379, "step": 75430 }, { "grad_norm": 0.32246705889701843, "learning_rate": 1.5605466350816315e-05, "loss": 0.0326, "step": 75440 }, { "grad_norm": 0.3942696452140808, "learning_rate": 1.559346711104674e-05, "loss": 0.0338, "step": 75450 }, { "grad_norm": 0.475241482257843, "learning_rate": 1.5581471633923883e-05, "loss": 0.0365, "step": 75460 }, { "grad_norm": 0.45478737354278564, "learning_rate": 1.5569479920759505e-05, "loss": 0.0366, "step": 75470 }, { "grad_norm": 0.3572903871536255, "learning_rate": 1.5557491972865072e-05, "loss": 0.0321, "step": 75480 }, { "grad_norm": 0.3397614657878876, "learning_rate": 1.554550779155152e-05, "loss": 0.0257, "step": 75490 }, { "grad_norm": 0.3453049659729004, "learning_rate": 1.5533527378129453e-05, "loss": 0.0227, "step": 75500 }, { "grad_norm": 0.43384072184562683, "learning_rate": 1.552155073390899e-05, "loss": 0.0362, "step": 75510 }, { "grad_norm": 0.699255645275116, "learning_rate": 1.5509577860199903e-05, "loss": 0.0464, "step": 75520 }, { "grad_norm": 0.30617469549179077, "learning_rate": 1.5497608758311525e-05, "loss": 0.0351, "step": 75530 }, { "grad_norm": 0.37340664863586426, "learning_rate": 1.5485643429552797e-05, "loss": 0.032, "step": 75540 }, { "grad_norm": 0.4379703998565674, "learning_rate": 1.5473681875232197e-05, "loss": 0.0346, "step": 75550 }, { "grad_norm": 0.37095510959625244, "learning_rate": 1.5461724096657837e-05, "loss": 0.0292, "step": 75560 }, { "grad_norm": 0.5844066143035889, "learning_rate": 1.5449770095137416e-05, "loss": 0.03, "step": 75570 }, { "grad_norm": 0.39913010597229004, "learning_rate": 1.5437819871978178e-05, "loss": 0.0291, "step": 75580 }, { "grad_norm": 0.6496866941452026, "learning_rate": 1.5425873428486992e-05, "loss": 0.0353, "step": 75590 }, { "grad_norm": 0.2711491584777832, "learning_rate": 1.5413930765970312e-05, "loss": 0.0352, "step": 75600 }, { "grad_norm": 0.3927268385887146, "learning_rate": 1.5401991885734175e-05, "loss": 0.0396, "step": 75610 }, { "grad_norm": 0.41977012157440186, "learning_rate": 1.5390056789084174e-05, "loss": 0.0292, "step": 75620 }, { "grad_norm": 0.5743744969367981, "learning_rate": 1.5378125477325545e-05, "loss": 0.0284, "step": 75630 }, { "grad_norm": 0.22817282378673553, "learning_rate": 1.5366197951763046e-05, "loss": 0.0271, "step": 75640 }, { "grad_norm": 0.6387767195701599, "learning_rate": 1.5354274213701075e-05, "loss": 0.0336, "step": 75650 }, { "grad_norm": 0.35419589281082153, "learning_rate": 1.5342354264443566e-05, "loss": 0.031, "step": 75660 }, { "grad_norm": 0.4255877435207367, "learning_rate": 1.533043810529411e-05, "loss": 0.0353, "step": 75670 }, { "grad_norm": 0.41599923372268677, "learning_rate": 1.5318525737555807e-05, "loss": 0.0402, "step": 75680 }, { "grad_norm": 0.5430196523666382, "learning_rate": 1.53066171625314e-05, "loss": 0.0367, "step": 75690 }, { "grad_norm": 0.518233597278595, "learning_rate": 1.529471238152317e-05, "loss": 0.0304, "step": 75700 }, { "grad_norm": 0.4533887803554535, "learning_rate": 1.5282811395833023e-05, "loss": 0.0309, "step": 75710 }, { "grad_norm": 0.4654386639595032, "learning_rate": 1.5270914206762397e-05, "loss": 0.0316, "step": 75720 }, { "grad_norm": 0.37397658824920654, "learning_rate": 1.525902081561241e-05, "loss": 0.0311, "step": 75730 }, { "grad_norm": 0.4443455934524536, "learning_rate": 1.5247131223683664e-05, "loss": 0.0362, "step": 75740 }, { "grad_norm": 0.5367082357406616, "learning_rate": 1.5235245432276408e-05, "loss": 0.0318, "step": 75750 }, { "grad_norm": 0.4294389486312866, "learning_rate": 1.5223363442690424e-05, "loss": 0.0344, "step": 75760 }, { "grad_norm": 0.5490804314613342, "learning_rate": 1.5211485256225133e-05, "loss": 0.0305, "step": 75770 }, { "grad_norm": 0.36340421438217163, "learning_rate": 1.519961087417951e-05, "loss": 0.0258, "step": 75780 }, { "grad_norm": 0.2910956144332886, "learning_rate": 1.5187740297852137e-05, "loss": 0.0328, "step": 75790 }, { "grad_norm": 0.2502272129058838, "learning_rate": 1.517587352854113e-05, "loss": 0.0224, "step": 75800 }, { "grad_norm": 0.8485552072525024, "learning_rate": 1.516401056754423e-05, "loss": 0.0357, "step": 75810 }, { "grad_norm": 0.45486027002334595, "learning_rate": 1.5152151416158783e-05, "loss": 0.0349, "step": 75820 }, { "grad_norm": 0.31631550192832947, "learning_rate": 1.5140296075681642e-05, "loss": 0.0363, "step": 75830 }, { "grad_norm": 0.4088296592235565, "learning_rate": 1.5128444547409331e-05, "loss": 0.0288, "step": 75840 }, { "grad_norm": 0.4331344664096832, "learning_rate": 1.5116596832637864e-05, "loss": 0.0294, "step": 75850 }, { "grad_norm": 0.3170097768306732, "learning_rate": 1.5104752932662952e-05, "loss": 0.037, "step": 75860 }, { "grad_norm": 0.47106698155403137, "learning_rate": 1.5092912848779778e-05, "loss": 0.0379, "step": 75870 }, { "grad_norm": 0.430159330368042, "learning_rate": 1.5081076582283194e-05, "loss": 0.0304, "step": 75880 }, { "grad_norm": 0.36121344566345215, "learning_rate": 1.5069244134467553e-05, "loss": 0.0264, "step": 75890 }, { "grad_norm": 0.23112784326076508, "learning_rate": 1.5057415506626876e-05, "loss": 0.0352, "step": 75900 }, { "grad_norm": 0.5670849084854126, "learning_rate": 1.5045590700054668e-05, "loss": 0.0328, "step": 75910 }, { "grad_norm": 1.1277996301651, "learning_rate": 1.5033769716044144e-05, "loss": 0.0306, "step": 75920 }, { "grad_norm": 2.0110361576080322, "learning_rate": 1.5021952555887975e-05, "loss": 0.0295, "step": 75930 }, { "grad_norm": 0.25148120522499084, "learning_rate": 1.50101392208785e-05, "loss": 0.0303, "step": 75940 }, { "grad_norm": 0.46068599820137024, "learning_rate": 1.4998329712307579e-05, "loss": 0.0239, "step": 75950 }, { "grad_norm": 0.4257188141345978, "learning_rate": 1.4986524031466708e-05, "loss": 0.0364, "step": 75960 }, { "grad_norm": 0.3621141016483307, "learning_rate": 1.4974722179646916e-05, "loss": 0.0212, "step": 75970 }, { "grad_norm": 0.5405293703079224, "learning_rate": 1.4962924158138847e-05, "loss": 0.0431, "step": 75980 }, { "grad_norm": 0.4035557210445404, "learning_rate": 1.4951129968232714e-05, "loss": 0.0325, "step": 75990 }, { "grad_norm": 0.37929996848106384, "learning_rate": 1.4939339611218329e-05, "loss": 0.0338, "step": 76000 }, { "grad_norm": 0.455359548330307, "learning_rate": 1.4927553088385032e-05, "loss": 0.0323, "step": 76010 }, { "grad_norm": 0.3724784255027771, "learning_rate": 1.4915770401021801e-05, "loss": 0.0299, "step": 76020 }, { "grad_norm": 0.3688282072544098, "learning_rate": 1.4903991550417185e-05, "loss": 0.046, "step": 76030 }, { "grad_norm": 0.4114021360874176, "learning_rate": 1.4892216537859277e-05, "loss": 0.0423, "step": 76040 }, { "grad_norm": 0.4807833731174469, "learning_rate": 1.4880445364635775e-05, "loss": 0.0308, "step": 76050 }, { "grad_norm": 0.4099549353122711, "learning_rate": 1.4868678032033973e-05, "loss": 0.0356, "step": 76060 }, { "grad_norm": 0.419985294342041, "learning_rate": 1.4856914541340739e-05, "loss": 0.0302, "step": 76070 }, { "grad_norm": 0.46440553665161133, "learning_rate": 1.4845154893842472e-05, "loss": 0.0296, "step": 76080 }, { "grad_norm": 0.5214619636535645, "learning_rate": 1.4833399090825234e-05, "loss": 0.0422, "step": 76090 }, { "grad_norm": 0.49224361777305603, "learning_rate": 1.4821647133574562e-05, "loss": 0.0399, "step": 76100 }, { "grad_norm": 0.502582311630249, "learning_rate": 1.4809899023375706e-05, "loss": 0.0286, "step": 76110 }, { "grad_norm": 0.4631495475769043, "learning_rate": 1.4798154761513366e-05, "loss": 0.0366, "step": 76120 }, { "grad_norm": 0.40706923604011536, "learning_rate": 1.4786414349271916e-05, "loss": 0.0337, "step": 76130 }, { "grad_norm": 0.48089903593063354, "learning_rate": 1.4774677787935226e-05, "loss": 0.0324, "step": 76140 }, { "grad_norm": 0.5379176139831543, "learning_rate": 1.4762945078786838e-05, "loss": 0.0319, "step": 76150 }, { "grad_norm": 0.4283744990825653, "learning_rate": 1.4751216223109777e-05, "loss": 0.0262, "step": 76160 }, { "grad_norm": 0.4815729856491089, "learning_rate": 1.4739491222186714e-05, "loss": 0.037, "step": 76170 }, { "grad_norm": 0.5145862102508545, "learning_rate": 1.4727770077299879e-05, "loss": 0.0342, "step": 76180 }, { "grad_norm": 0.3962850272655487, "learning_rate": 1.4716052789731094e-05, "loss": 0.0332, "step": 76190 }, { "grad_norm": 0.5505253076553345, "learning_rate": 1.470433936076171e-05, "loss": 0.0319, "step": 76200 }, { "grad_norm": 0.6175615787506104, "learning_rate": 1.4692629791672719e-05, "loss": 0.0285, "step": 76210 }, { "grad_norm": 0.5496824979782104, "learning_rate": 1.468092408374464e-05, "loss": 0.0292, "step": 76220 }, { "grad_norm": 0.45710617303848267, "learning_rate": 1.4669222238257602e-05, "loss": 0.0447, "step": 76230 }, { "grad_norm": 0.2943735122680664, "learning_rate": 1.4657524256491301e-05, "loss": 0.0245, "step": 76240 }, { "grad_norm": 0.500881016254425, "learning_rate": 1.464583013972502e-05, "loss": 0.0287, "step": 76250 }, { "grad_norm": 0.24715681374073029, "learning_rate": 1.4634139889237591e-05, "loss": 0.0291, "step": 76260 }, { "grad_norm": 0.3662310540676117, "learning_rate": 1.4622453506307448e-05, "loss": 0.0345, "step": 76270 }, { "grad_norm": 0.43589940667152405, "learning_rate": 1.461077099221262e-05, "loss": 0.0285, "step": 76280 }, { "grad_norm": 0.6404316425323486, "learning_rate": 1.4599092348230652e-05, "loss": 0.0333, "step": 76290 }, { "grad_norm": 0.5476266741752625, "learning_rate": 1.4587417575638724e-05, "loss": 0.0254, "step": 76300 }, { "grad_norm": 1.1430050134658813, "learning_rate": 1.4575746675713563e-05, "loss": 0.0329, "step": 76310 }, { "grad_norm": 0.4552570581436157, "learning_rate": 1.4564079649731504e-05, "loss": 0.0329, "step": 76320 }, { "grad_norm": 0.6580346822738647, "learning_rate": 1.4552416498968402e-05, "loss": 0.0326, "step": 76330 }, { "grad_norm": 0.49766287207603455, "learning_rate": 1.4540757224699747e-05, "loss": 0.0259, "step": 76340 }, { "grad_norm": 0.2873188853263855, "learning_rate": 1.452910182820056e-05, "loss": 0.0327, "step": 76350 }, { "grad_norm": 0.5255569219589233, "learning_rate": 1.4517450310745468e-05, "loss": 0.0289, "step": 76360 }, { "grad_norm": 0.3848136365413666, "learning_rate": 1.450580267360866e-05, "loss": 0.0281, "step": 76370 }, { "grad_norm": 0.20043884217739105, "learning_rate": 1.4494158918063922e-05, "loss": 0.0295, "step": 76380 }, { "grad_norm": 0.5082846283912659, "learning_rate": 1.4482519045384568e-05, "loss": 0.0291, "step": 76390 }, { "grad_norm": 0.39540764689445496, "learning_rate": 1.4470883056843548e-05, "loss": 0.0292, "step": 76400 }, { "grad_norm": 0.44265663623809814, "learning_rate": 1.4459250953713322e-05, "loss": 0.0376, "step": 76410 }, { "grad_norm": 0.4092405438423157, "learning_rate": 1.444762273726598e-05, "loss": 0.0311, "step": 76420 }, { "grad_norm": 0.44672757387161255, "learning_rate": 1.443599840877316e-05, "loss": 0.0316, "step": 76430 }, { "grad_norm": 0.32984989881515503, "learning_rate": 1.4424377969506103e-05, "loss": 0.0305, "step": 76440 }, { "grad_norm": 0.5635491013526917, "learning_rate": 1.4412761420735565e-05, "loss": 0.0376, "step": 76450 }, { "grad_norm": 0.5571357011795044, "learning_rate": 1.4401148763731953e-05, "loss": 0.0346, "step": 76460 }, { "grad_norm": 0.5560318231582642, "learning_rate": 1.4389539999765173e-05, "loss": 0.0325, "step": 76470 }, { "grad_norm": 0.49667423963546753, "learning_rate": 1.4377935130104758e-05, "loss": 0.0289, "step": 76480 }, { "grad_norm": 0.30126842856407166, "learning_rate": 1.4366334156019801e-05, "loss": 0.034, "step": 76490 }, { "grad_norm": 0.5518956184387207, "learning_rate": 1.435473707877898e-05, "loss": 0.0336, "step": 76500 }, { "grad_norm": 0.36121559143066406, "learning_rate": 1.4343143899650508e-05, "loss": 0.0368, "step": 76510 }, { "grad_norm": 0.33791935443878174, "learning_rate": 1.4331554619902204e-05, "loss": 0.0324, "step": 76520 }, { "grad_norm": 0.3590395152568817, "learning_rate": 1.4319969240801484e-05, "loss": 0.0252, "step": 76530 }, { "grad_norm": 0.4457295835018158, "learning_rate": 1.430838776361526e-05, "loss": 0.0396, "step": 76540 }, { "grad_norm": 0.333015501499176, "learning_rate": 1.4296810189610093e-05, "loss": 0.0394, "step": 76550 }, { "grad_norm": 0.3977106511592865, "learning_rate": 1.4285236520052087e-05, "loss": 0.03, "step": 76560 }, { "grad_norm": 0.3839241862297058, "learning_rate": 1.4273666756206933e-05, "loss": 0.0287, "step": 76570 }, { "grad_norm": 0.47471296787261963, "learning_rate": 1.4262100899339848e-05, "loss": 0.0332, "step": 76580 }, { "grad_norm": 0.28953465819358826, "learning_rate": 1.4250538950715696e-05, "loss": 0.0272, "step": 76590 }, { "grad_norm": 0.5418914556503296, "learning_rate": 1.4238980911598842e-05, "loss": 0.0412, "step": 76600 }, { "grad_norm": 0.31442907452583313, "learning_rate": 1.4227426783253267e-05, "loss": 0.0334, "step": 76610 }, { "grad_norm": 0.4102906286716461, "learning_rate": 1.4215876566942521e-05, "loss": 0.0325, "step": 76620 }, { "grad_norm": 0.5822196006774902, "learning_rate": 1.4204330263929721e-05, "loss": 0.0287, "step": 76630 }, { "grad_norm": 0.783858060836792, "learning_rate": 1.4192787875477537e-05, "loss": 0.0406, "step": 76640 }, { "grad_norm": 0.22978660464286804, "learning_rate": 1.4181249402848246e-05, "loss": 0.0296, "step": 76650 }, { "grad_norm": 0.46448197960853577, "learning_rate": 1.4169714847303656e-05, "loss": 0.0319, "step": 76660 }, { "grad_norm": 0.5867120027542114, "learning_rate": 1.4158184210105197e-05, "loss": 0.0354, "step": 76670 }, { "grad_norm": 0.29764771461486816, "learning_rate": 1.4146657492513787e-05, "loss": 0.0311, "step": 76680 }, { "grad_norm": 0.2305128574371338, "learning_rate": 1.4135134695790047e-05, "loss": 0.0247, "step": 76690 }, { "grad_norm": 0.4869626462459564, "learning_rate": 1.4123615821194037e-05, "loss": 0.035, "step": 76700 }, { "grad_norm": 0.43876489996910095, "learning_rate": 1.4112100869985473e-05, "loss": 0.0312, "step": 76710 }, { "grad_norm": 0.25749003887176514, "learning_rate": 1.4100589843423589e-05, "loss": 0.0326, "step": 76720 }, { "grad_norm": 0.3267850875854492, "learning_rate": 1.408908274276723e-05, "loss": 0.0355, "step": 76730 }, { "grad_norm": 0.2813234329223633, "learning_rate": 1.4077579569274786e-05, "loss": 0.0309, "step": 76740 }, { "grad_norm": 0.39831244945526123, "learning_rate": 1.4066080324204245e-05, "loss": 0.0305, "step": 76750 }, { "grad_norm": 0.3912231922149658, "learning_rate": 1.405458500881312e-05, "loss": 0.0361, "step": 76760 }, { "grad_norm": 0.4124385416507721, "learning_rate": 1.4043093624358538e-05, "loss": 0.0298, "step": 76770 }, { "grad_norm": 0.2872268557548523, "learning_rate": 1.4031606172097189e-05, "loss": 0.05, "step": 76780 }, { "grad_norm": 0.40665361285209656, "learning_rate": 1.40201226532853e-05, "loss": 0.0401, "step": 76790 }, { "grad_norm": 0.28299108147621155, "learning_rate": 1.4008643069178701e-05, "loss": 0.0284, "step": 76800 }, { "grad_norm": 0.38819587230682373, "learning_rate": 1.399716742103278e-05, "loss": 0.0326, "step": 76810 }, { "grad_norm": 0.3810846209526062, "learning_rate": 1.398569571010252e-05, "loss": 0.0296, "step": 76820 }, { "grad_norm": 0.4500865936279297, "learning_rate": 1.3974227937642414e-05, "loss": 0.0355, "step": 76830 }, { "grad_norm": 0.5751137137413025, "learning_rate": 1.3962764104906595e-05, "loss": 0.0241, "step": 76840 }, { "grad_norm": 0.34527653455734253, "learning_rate": 1.3951304213148696e-05, "loss": 0.0354, "step": 76850 }, { "grad_norm": 0.3435710668563843, "learning_rate": 1.3939848263621979e-05, "loss": 0.0286, "step": 76860 }, { "grad_norm": 1.2148741483688354, "learning_rate": 1.3928396257579219e-05, "loss": 0.039, "step": 76870 }, { "grad_norm": 0.4386667311191559, "learning_rate": 1.3916948196272833e-05, "loss": 0.0412, "step": 76880 }, { "grad_norm": 0.90641850233078, "learning_rate": 1.3905504080954734e-05, "loss": 0.0451, "step": 76890 }, { "grad_norm": 0.37057942152023315, "learning_rate": 1.3894063912876454e-05, "loss": 0.0339, "step": 76900 }, { "grad_norm": 0.44191861152648926, "learning_rate": 1.3882627693289046e-05, "loss": 0.0255, "step": 76910 }, { "grad_norm": 0.38233664631843567, "learning_rate": 1.3871195423443183e-05, "loss": 0.03, "step": 76920 }, { "grad_norm": 0.31733450293540955, "learning_rate": 1.3859767104589044e-05, "loss": 0.0297, "step": 76930 }, { "grad_norm": 0.6018064022064209, "learning_rate": 1.3848342737976467e-05, "loss": 0.0304, "step": 76940 }, { "grad_norm": 0.4130078852176666, "learning_rate": 1.3836922324854757e-05, "loss": 0.0332, "step": 76950 }, { "grad_norm": 1.0135068893432617, "learning_rate": 1.382550586647287e-05, "loss": 0.0377, "step": 76960 }, { "grad_norm": 0.23842479288578033, "learning_rate": 1.3814093364079257e-05, "loss": 0.0321, "step": 76970 }, { "grad_norm": 0.7295520901679993, "learning_rate": 1.3802684818921991e-05, "loss": 0.0348, "step": 76980 }, { "grad_norm": 0.2514590919017792, "learning_rate": 1.3791280232248687e-05, "loss": 0.0345, "step": 76990 }, { "grad_norm": 0.36893430352211, "learning_rate": 1.3779879605306555e-05, "loss": 0.0342, "step": 77000 }, { "grad_norm": 0.37348583340644836, "learning_rate": 1.376848293934232e-05, "loss": 0.0357, "step": 77010 }, { "grad_norm": 0.269765704870224, "learning_rate": 1.3757090235602316e-05, "loss": 0.0264, "step": 77020 }, { "grad_norm": 0.33613160252571106, "learning_rate": 1.3745701495332447e-05, "loss": 0.0329, "step": 77030 }, { "grad_norm": 0.3877478837966919, "learning_rate": 1.3734316719778145e-05, "loss": 0.0267, "step": 77040 }, { "grad_norm": 0.4548453688621521, "learning_rate": 1.3722935910184454e-05, "loss": 0.0281, "step": 77050 }, { "grad_norm": 0.39058926701545715, "learning_rate": 1.3711559067795921e-05, "loss": 0.0317, "step": 77060 }, { "grad_norm": 0.3938356637954712, "learning_rate": 1.3700186193856756e-05, "loss": 0.038, "step": 77070 }, { "grad_norm": 0.32762694358825684, "learning_rate": 1.368881728961064e-05, "loss": 0.0275, "step": 77080 }, { "grad_norm": 0.37664857506752014, "learning_rate": 1.3677452356300885e-05, "loss": 0.0328, "step": 77090 }, { "grad_norm": 0.48764437437057495, "learning_rate": 1.3666091395170321e-05, "loss": 0.0381, "step": 77100 }, { "grad_norm": 0.4336792826652527, "learning_rate": 1.3654734407461384e-05, "loss": 0.0345, "step": 77110 }, { "grad_norm": 0.6055566072463989, "learning_rate": 1.3643381394416022e-05, "loss": 0.0289, "step": 77120 }, { "grad_norm": 0.4803113043308258, "learning_rate": 1.3632032357275838e-05, "loss": 0.0347, "step": 77130 }, { "grad_norm": 0.29966801404953003, "learning_rate": 1.3620687297281898e-05, "loss": 0.0301, "step": 77140 }, { "grad_norm": 0.7385135889053345, "learning_rate": 1.3609346215674917e-05, "loss": 0.0228, "step": 77150 }, { "grad_norm": 0.5503154993057251, "learning_rate": 1.3598009113695109e-05, "loss": 0.0401, "step": 77160 }, { "grad_norm": 0.25325578451156616, "learning_rate": 1.358667599258231e-05, "loss": 0.0293, "step": 77170 }, { "grad_norm": 0.3923078775405884, "learning_rate": 1.357534685357586e-05, "loss": 0.0378, "step": 77180 }, { "grad_norm": 0.3918105661869049, "learning_rate": 1.3564021697914715e-05, "loss": 0.0322, "step": 77190 }, { "grad_norm": 0.2744505703449249, "learning_rate": 1.3552700526837381e-05, "loss": 0.0289, "step": 77200 }, { "grad_norm": 0.3168521523475647, "learning_rate": 1.3541383341581925e-05, "loss": 0.0278, "step": 77210 }, { "grad_norm": 0.5327057838439941, "learning_rate": 1.3530070143385965e-05, "loss": 0.026, "step": 77220 }, { "grad_norm": 0.3730027675628662, "learning_rate": 1.3518760933486702e-05, "loss": 0.0241, "step": 77230 }, { "grad_norm": 0.41549190878868103, "learning_rate": 1.3507455713120908e-05, "loss": 0.0308, "step": 77240 }, { "grad_norm": 0.35091808438301086, "learning_rate": 1.3496154483524876e-05, "loss": 0.0263, "step": 77250 }, { "grad_norm": 0.636139988899231, "learning_rate": 1.3484857245934506e-05, "loss": 0.0412, "step": 77260 }, { "grad_norm": 0.2746312618255615, "learning_rate": 1.3473564001585248e-05, "loss": 0.0356, "step": 77270 }, { "grad_norm": 0.3798198997974396, "learning_rate": 1.3462274751712129e-05, "loss": 0.0414, "step": 77280 }, { "grad_norm": 0.3479653298854828, "learning_rate": 1.3450989497549698e-05, "loss": 0.0283, "step": 77290 }, { "grad_norm": 0.5748348832130432, "learning_rate": 1.3439708240332115e-05, "loss": 0.0385, "step": 77300 }, { "grad_norm": 0.447162002325058, "learning_rate": 1.3428430981293043e-05, "loss": 0.0342, "step": 77310 }, { "grad_norm": 0.30806511640548706, "learning_rate": 1.3417157721665808e-05, "loss": 0.0323, "step": 77320 }, { "grad_norm": 0.29822757840156555, "learning_rate": 1.3405888462683186e-05, "loss": 0.0255, "step": 77330 }, { "grad_norm": 0.41047823429107666, "learning_rate": 1.3394623205577595e-05, "loss": 0.0322, "step": 77340 }, { "grad_norm": 0.716524064540863, "learning_rate": 1.3383361951580963e-05, "loss": 0.0297, "step": 77350 }, { "grad_norm": 0.5887596011161804, "learning_rate": 1.337210470192483e-05, "loss": 0.0316, "step": 77360 }, { "grad_norm": 0.40504974126815796, "learning_rate": 1.3360851457840245e-05, "loss": 0.0318, "step": 77370 }, { "grad_norm": 0.29464054107666016, "learning_rate": 1.3349602220557856e-05, "loss": 0.0255, "step": 77380 }, { "grad_norm": 0.6713521480560303, "learning_rate": 1.3338356991307865e-05, "loss": 0.0375, "step": 77390 }, { "grad_norm": 0.627291738986969, "learning_rate": 1.3327115771320048e-05, "loss": 0.0332, "step": 77400 }, { "grad_norm": 0.3093265891075134, "learning_rate": 1.3315878561823697e-05, "loss": 0.0253, "step": 77410 }, { "grad_norm": 0.8927424550056458, "learning_rate": 1.330464536404773e-05, "loss": 0.0326, "step": 77420 }, { "grad_norm": 0.40985241532325745, "learning_rate": 1.3293416179220563e-05, "loss": 0.0365, "step": 77430 }, { "grad_norm": 0.6786230206489563, "learning_rate": 1.328219100857021e-05, "loss": 0.0314, "step": 77440 }, { "grad_norm": 0.29654461145401, "learning_rate": 1.327096985332425e-05, "loss": 0.031, "step": 77450 }, { "grad_norm": 0.4892543852329254, "learning_rate": 1.3259752714709816e-05, "loss": 0.0301, "step": 77460 }, { "grad_norm": 0.5166285037994385, "learning_rate": 1.3248539593953569e-05, "loss": 0.0301, "step": 77470 }, { "grad_norm": 0.6515269875526428, "learning_rate": 1.323733049228178e-05, "loss": 0.042, "step": 77480 }, { "grad_norm": 0.5104919672012329, "learning_rate": 1.3226125410920276e-05, "loss": 0.0351, "step": 77490 }, { "grad_norm": 0.49352845549583435, "learning_rate": 1.3214924351094388e-05, "loss": 0.0316, "step": 77500 }, { "grad_norm": 0.3114616274833679, "learning_rate": 1.3203727314029069e-05, "loss": 0.0276, "step": 77510 }, { "grad_norm": 0.5138165354728699, "learning_rate": 1.3192534300948811e-05, "loss": 0.0329, "step": 77520 }, { "grad_norm": 0.402109295129776, "learning_rate": 1.3181345313077675e-05, "loss": 0.0367, "step": 77530 }, { "grad_norm": 0.536109447479248, "learning_rate": 1.3170160351639243e-05, "loss": 0.0344, "step": 77540 }, { "grad_norm": 0.4811096787452698, "learning_rate": 1.3158979417856721e-05, "loss": 0.0389, "step": 77550 }, { "grad_norm": 0.49908944964408875, "learning_rate": 1.31478025129528e-05, "loss": 0.0311, "step": 77560 }, { "grad_norm": 0.28281423449516296, "learning_rate": 1.3136629638149794e-05, "loss": 0.0284, "step": 77570 }, { "grad_norm": 0.4422109127044678, "learning_rate": 1.3125460794669547e-05, "loss": 0.031, "step": 77580 }, { "grad_norm": 0.5318630933761597, "learning_rate": 1.3114295983733483e-05, "loss": 0.029, "step": 77590 }, { "grad_norm": 0.38476815819740295, "learning_rate": 1.3103135206562534e-05, "loss": 0.0384, "step": 77600 }, { "grad_norm": 0.42732781171798706, "learning_rate": 1.3091978464377264e-05, "loss": 0.0396, "step": 77610 }, { "grad_norm": 0.2941015362739563, "learning_rate": 1.3080825758397724e-05, "loss": 0.033, "step": 77620 }, { "grad_norm": 1.0151783227920532, "learning_rate": 1.3069677089843568e-05, "loss": 0.0331, "step": 77630 }, { "grad_norm": 0.32923150062561035, "learning_rate": 1.305853245993401e-05, "loss": 0.0347, "step": 77640 }, { "grad_norm": 0.3008784353733063, "learning_rate": 1.3047391869887809e-05, "loss": 0.0328, "step": 77650 }, { "grad_norm": 0.3631792962551117, "learning_rate": 1.3036255320923263e-05, "loss": 0.0285, "step": 77660 }, { "grad_norm": 0.4061983525753021, "learning_rate": 1.3025122814258273e-05, "loss": 0.0289, "step": 77670 }, { "grad_norm": 0.7667822241783142, "learning_rate": 1.3013994351110249e-05, "loss": 0.0346, "step": 77680 }, { "grad_norm": 0.49632880091667175, "learning_rate": 1.3002869932696194e-05, "loss": 0.0331, "step": 77690 }, { "grad_norm": 0.4871128499507904, "learning_rate": 1.2991749560232664e-05, "loss": 0.0335, "step": 77700 }, { "grad_norm": 0.2326987385749817, "learning_rate": 1.2980633234935768e-05, "loss": 0.0223, "step": 77710 }, { "grad_norm": 0.37232211232185364, "learning_rate": 1.2969520958021153e-05, "loss": 0.0318, "step": 77720 }, { "grad_norm": 0.3740556240081787, "learning_rate": 1.295841273070405e-05, "loss": 0.0277, "step": 77730 }, { "grad_norm": 0.49150848388671875, "learning_rate": 1.2947308554199255e-05, "loss": 0.0298, "step": 77740 }, { "grad_norm": 0.5652151107788086, "learning_rate": 1.2936208429721075e-05, "loss": 0.0422, "step": 77750 }, { "grad_norm": 0.2975734770298004, "learning_rate": 1.2925112358483416e-05, "loss": 0.026, "step": 77760 }, { "grad_norm": 0.8555594086647034, "learning_rate": 1.2914020341699728e-05, "loss": 0.033, "step": 77770 }, { "grad_norm": 0.4987689256668091, "learning_rate": 1.2902932380583032e-05, "loss": 0.0278, "step": 77780 }, { "grad_norm": 0.4589618146419525, "learning_rate": 1.2891848476345863e-05, "loss": 0.0342, "step": 77790 }, { "grad_norm": 0.4019864499568939, "learning_rate": 1.2880768630200368e-05, "loss": 0.0352, "step": 77800 }, { "grad_norm": 0.34534478187561035, "learning_rate": 1.286969284335819e-05, "loss": 0.0261, "step": 77810 }, { "grad_norm": 0.30804556608200073, "learning_rate": 1.2858621117030578e-05, "loss": 0.0243, "step": 77820 }, { "grad_norm": 0.46301546692848206, "learning_rate": 1.2847553452428324e-05, "loss": 0.023, "step": 77830 }, { "grad_norm": 0.24913442134857178, "learning_rate": 1.283648985076178e-05, "loss": 0.023, "step": 77840 }, { "grad_norm": 0.5013895034790039, "learning_rate": 1.2825430313240817e-05, "loss": 0.0375, "step": 77850 }, { "grad_norm": 0.6326391696929932, "learning_rate": 1.281437484107491e-05, "loss": 0.0332, "step": 77860 }, { "grad_norm": 0.2529362142086029, "learning_rate": 1.2803323435473058e-05, "loss": 0.0302, "step": 77870 }, { "grad_norm": 0.4292986989021301, "learning_rate": 1.2792276097643841e-05, "loss": 0.0283, "step": 77880 }, { "grad_norm": 0.5684289336204529, "learning_rate": 1.2781232828795342e-05, "loss": 0.0323, "step": 77890 }, { "grad_norm": 0.3750408887863159, "learning_rate": 1.2770193630135292e-05, "loss": 0.0347, "step": 77900 }, { "grad_norm": 0.5206466317176819, "learning_rate": 1.2759158502870877e-05, "loss": 0.0372, "step": 77910 }, { "grad_norm": 0.3087124228477478, "learning_rate": 1.2748127448208913e-05, "loss": 0.0285, "step": 77920 }, { "grad_norm": 0.22019906342029572, "learning_rate": 1.2737100467355706e-05, "loss": 0.0368, "step": 77930 }, { "grad_norm": 0.426224946975708, "learning_rate": 1.2726077561517169e-05, "loss": 0.0365, "step": 77940 }, { "grad_norm": 0.35718783736228943, "learning_rate": 1.2715058731898755e-05, "loss": 0.0285, "step": 77950 }, { "grad_norm": 0.5415353775024414, "learning_rate": 1.2704043979705471e-05, "loss": 0.0386, "step": 77960 }, { "grad_norm": 0.5520304441452026, "learning_rate": 1.269303330614185e-05, "loss": 0.0317, "step": 77970 }, { "grad_norm": 0.5798884630203247, "learning_rate": 1.2682026712412015e-05, "loss": 0.0332, "step": 77980 }, { "grad_norm": 0.36550337076187134, "learning_rate": 1.2671024199719644e-05, "loss": 0.0246, "step": 77990 }, { "grad_norm": 0.551375687122345, "learning_rate": 1.2660025769267936e-05, "loss": 0.0326, "step": 78000 }, { "grad_norm": 0.7065132260322571, "learning_rate": 1.2649031422259667e-05, "loss": 0.0311, "step": 78010 }, { "grad_norm": 0.5613954067230225, "learning_rate": 1.2638041159897163e-05, "loss": 0.0309, "step": 78020 }, { "grad_norm": 0.3602008521556854, "learning_rate": 1.262705498338232e-05, "loss": 0.0365, "step": 78030 }, { "grad_norm": 0.34703007340431213, "learning_rate": 1.2616072893916542e-05, "loss": 0.0392, "step": 78040 }, { "grad_norm": 0.29140156507492065, "learning_rate": 1.2605094892700836e-05, "loss": 0.0323, "step": 78050 }, { "grad_norm": 0.3751204013824463, "learning_rate": 1.259412098093572e-05, "loss": 0.0329, "step": 78060 }, { "grad_norm": 0.22891353070735931, "learning_rate": 1.258315115982131e-05, "loss": 0.0301, "step": 78070 }, { "grad_norm": 0.5353444218635559, "learning_rate": 1.2572185430557203e-05, "loss": 0.0266, "step": 78080 }, { "grad_norm": 0.481802374124527, "learning_rate": 1.2561223794342659e-05, "loss": 0.0292, "step": 78090 }, { "grad_norm": 0.42501384019851685, "learning_rate": 1.255026625237637e-05, "loss": 0.0362, "step": 78100 }, { "grad_norm": 0.40607747435569763, "learning_rate": 1.2539312805856678e-05, "loss": 0.0407, "step": 78110 }, { "grad_norm": 0.6586433053016663, "learning_rate": 1.25283634559814e-05, "loss": 0.0336, "step": 78120 }, { "grad_norm": 0.24217848479747772, "learning_rate": 1.2517418203947972e-05, "loss": 0.032, "step": 78130 }, { "grad_norm": 0.5083265900611877, "learning_rate": 1.2506477050953308e-05, "loss": 0.0284, "step": 78140 }, { "grad_norm": 0.4020407497882843, "learning_rate": 1.2495539998193972e-05, "loss": 0.0322, "step": 78150 }, { "grad_norm": 0.5027636885643005, "learning_rate": 1.2484607046865975e-05, "loss": 0.029, "step": 78160 }, { "grad_norm": 0.427699476480484, "learning_rate": 1.2473678198164968e-05, "loss": 0.0276, "step": 78170 }, { "grad_norm": 0.5240817070007324, "learning_rate": 1.2462753453286075e-05, "loss": 0.0327, "step": 78180 }, { "grad_norm": 0.2678886950016022, "learning_rate": 1.2451832813424031e-05, "loss": 0.0528, "step": 78190 }, { "grad_norm": 0.5415598750114441, "learning_rate": 1.2440916279773095e-05, "loss": 0.0291, "step": 78200 }, { "grad_norm": 0.28740164637565613, "learning_rate": 1.2430003853527101e-05, "loss": 0.0384, "step": 78210 }, { "grad_norm": 0.5175172686576843, "learning_rate": 1.2419095535879383e-05, "loss": 0.0427, "step": 78220 }, { "grad_norm": 0.23371219635009766, "learning_rate": 1.2408191328022878e-05, "loss": 0.0331, "step": 78230 }, { "grad_norm": 0.5344889760017395, "learning_rate": 1.2397291231150066e-05, "loss": 0.0294, "step": 78240 }, { "grad_norm": 0.6240077614784241, "learning_rate": 1.2386395246452937e-05, "loss": 0.034, "step": 78250 }, { "grad_norm": 0.2653518319129944, "learning_rate": 1.2375503375123083e-05, "loss": 0.031, "step": 78260 }, { "grad_norm": 0.21900911629199982, "learning_rate": 1.2364615618351583e-05, "loss": 0.0426, "step": 78270 }, { "grad_norm": 0.3436826765537262, "learning_rate": 1.2353731977329169e-05, "loss": 0.0312, "step": 78280 }, { "grad_norm": 0.42166072130203247, "learning_rate": 1.2342852453246007e-05, "loss": 0.0221, "step": 78290 }, { "grad_norm": 0.3384476602077484, "learning_rate": 1.2331977047291898e-05, "loss": 0.0338, "step": 78300 }, { "grad_norm": 0.46177351474761963, "learning_rate": 1.2321105760656137e-05, "loss": 0.0296, "step": 78310 }, { "grad_norm": 0.24191035330295563, "learning_rate": 1.2310238594527613e-05, "loss": 0.0356, "step": 78320 }, { "grad_norm": 0.8349399566650391, "learning_rate": 1.2299375550094693e-05, "loss": 0.0318, "step": 78330 }, { "grad_norm": 0.7484542727470398, "learning_rate": 1.2288516628545421e-05, "loss": 0.0313, "step": 78340 }, { "grad_norm": 0.361909419298172, "learning_rate": 1.2277661831067255e-05, "loss": 0.0335, "step": 78350 }, { "grad_norm": 0.2817022204399109, "learning_rate": 1.2266811158847286e-05, "loss": 0.0338, "step": 78360 }, { "grad_norm": 0.3173084855079651, "learning_rate": 1.2255964613072107e-05, "loss": 0.0243, "step": 78370 }, { "grad_norm": 0.4693877100944519, "learning_rate": 1.2245122194927905e-05, "loss": 0.0237, "step": 78380 }, { "grad_norm": 0.4126788079738617, "learning_rate": 1.2234283905600364e-05, "loss": 0.035, "step": 78390 }, { "grad_norm": 0.5418547987937927, "learning_rate": 1.222344974627475e-05, "loss": 0.0413, "step": 78400 }, { "grad_norm": 0.3937681317329407, "learning_rate": 1.2212619718135875e-05, "loss": 0.0285, "step": 78410 }, { "grad_norm": 0.35224518179893494, "learning_rate": 1.2201793822368118e-05, "loss": 0.0395, "step": 78420 }, { "grad_norm": 0.505609929561615, "learning_rate": 1.2190972060155337e-05, "loss": 0.0297, "step": 78430 }, { "grad_norm": 0.27959465980529785, "learning_rate": 1.2180154432681013e-05, "loss": 0.0297, "step": 78440 }, { "grad_norm": 0.33334261178970337, "learning_rate": 1.2169340941128148e-05, "loss": 0.0331, "step": 78450 }, { "grad_norm": 0.35433533787727356, "learning_rate": 1.2158531586679267e-05, "loss": 0.0253, "step": 78460 }, { "grad_norm": 0.5992754697799683, "learning_rate": 1.2147726370516477e-05, "loss": 0.0277, "step": 78470 }, { "grad_norm": 0.2345949113368988, "learning_rate": 1.2136925293821422e-05, "loss": 0.0375, "step": 78480 }, { "grad_norm": 0.47471943497657776, "learning_rate": 1.2126128357775306e-05, "loss": 0.0329, "step": 78490 }, { "grad_norm": 0.5369955897331238, "learning_rate": 1.2115335563558838e-05, "loss": 0.0363, "step": 78500 }, { "grad_norm": 0.3524308502674103, "learning_rate": 1.2104546912352327e-05, "loss": 0.0325, "step": 78510 }, { "grad_norm": 0.4909350872039795, "learning_rate": 1.2093762405335557e-05, "loss": 0.0244, "step": 78520 }, { "grad_norm": 0.4153166115283966, "learning_rate": 1.2082982043687974e-05, "loss": 0.0308, "step": 78530 }, { "grad_norm": 0.513816773891449, "learning_rate": 1.207220582858845e-05, "loss": 0.0306, "step": 78540 }, { "grad_norm": 0.42291393876075745, "learning_rate": 1.206143376121549e-05, "loss": 0.0273, "step": 78550 }, { "grad_norm": 0.8528031706809998, "learning_rate": 1.2050665842747078e-05, "loss": 0.0268, "step": 78560 }, { "grad_norm": 0.5171395540237427, "learning_rate": 1.2039902074360804e-05, "loss": 0.0295, "step": 78570 }, { "grad_norm": 0.524366557598114, "learning_rate": 1.2029142457233755e-05, "loss": 0.0347, "step": 78580 }, { "grad_norm": 0.5853226780891418, "learning_rate": 1.2018386992542601e-05, "loss": 0.0337, "step": 78590 }, { "grad_norm": 0.2868247926235199, "learning_rate": 1.2007635681463542e-05, "loss": 0.0378, "step": 78600 }, { "grad_norm": 0.18401159346103668, "learning_rate": 1.1996888525172334e-05, "loss": 0.0304, "step": 78610 }, { "grad_norm": 0.35588493943214417, "learning_rate": 1.198614552484425e-05, "loss": 0.0248, "step": 78620 }, { "grad_norm": 0.4670783579349518, "learning_rate": 1.197540668165415e-05, "loss": 0.0309, "step": 78630 }, { "grad_norm": 0.6113635301589966, "learning_rate": 1.1964671996776395e-05, "loss": 0.0402, "step": 78640 }, { "grad_norm": 0.5278029441833496, "learning_rate": 1.1953941471384922e-05, "loss": 0.0463, "step": 78650 }, { "grad_norm": 0.43730616569519043, "learning_rate": 1.1943215106653205e-05, "loss": 0.0351, "step": 78660 }, { "grad_norm": 0.39767399430274963, "learning_rate": 1.193249290375429e-05, "loss": 0.0333, "step": 78670 }, { "grad_norm": 0.36640575528144836, "learning_rate": 1.1921774863860696e-05, "loss": 0.0343, "step": 78680 }, { "grad_norm": 0.420082151889801, "learning_rate": 1.1911060988144556e-05, "loss": 0.0357, "step": 78690 }, { "grad_norm": 0.369106262922287, "learning_rate": 1.1900351277777533e-05, "loss": 0.0302, "step": 78700 }, { "grad_norm": 0.47922182083129883, "learning_rate": 1.1889645733930798e-05, "loss": 0.0331, "step": 78710 }, { "grad_norm": 0.6326553225517273, "learning_rate": 1.1878944357775107e-05, "loss": 0.0312, "step": 78720 }, { "grad_norm": 0.30846914649009705, "learning_rate": 1.1868247150480749e-05, "loss": 0.0293, "step": 78730 }, { "grad_norm": 0.5154182314872742, "learning_rate": 1.1857554113217567e-05, "loss": 0.0302, "step": 78740 }, { "grad_norm": 0.29992398619651794, "learning_rate": 1.1846865247154903e-05, "loss": 0.032, "step": 78750 }, { "grad_norm": 0.25675123929977417, "learning_rate": 1.1836180553461711e-05, "loss": 0.0302, "step": 78760 }, { "grad_norm": 0.555175244808197, "learning_rate": 1.1825500033306419e-05, "loss": 0.0438, "step": 78770 }, { "grad_norm": 0.46058258414268494, "learning_rate": 1.1814823687857052e-05, "loss": 0.0344, "step": 78780 }, { "grad_norm": 0.4373789131641388, "learning_rate": 1.1804151518281153e-05, "loss": 0.0327, "step": 78790 }, { "grad_norm": 0.4095974266529083, "learning_rate": 1.179348352574583e-05, "loss": 0.041, "step": 78800 }, { "grad_norm": 0.4486576318740845, "learning_rate": 1.1782819711417697e-05, "loss": 0.0349, "step": 78810 }, { "grad_norm": 0.2743287682533264, "learning_rate": 1.1772160076462951e-05, "loss": 0.0253, "step": 78820 }, { "grad_norm": 0.417934387922287, "learning_rate": 1.1761504622047293e-05, "loss": 0.0366, "step": 78830 }, { "grad_norm": 0.16091011464595795, "learning_rate": 1.1750853349335999e-05, "loss": 0.0315, "step": 78840 }, { "grad_norm": 0.5577698349952698, "learning_rate": 1.1740206259493874e-05, "loss": 0.0341, "step": 78850 }, { "grad_norm": 0.36728718876838684, "learning_rate": 1.1729563353685286e-05, "loss": 0.0272, "step": 78860 }, { "grad_norm": 0.5234214663505554, "learning_rate": 1.1718924633074091e-05, "loss": 0.0316, "step": 78870 }, { "grad_norm": 0.3385205864906311, "learning_rate": 1.1708290098823765e-05, "loss": 0.0312, "step": 78880 }, { "grad_norm": 0.4448518753051758, "learning_rate": 1.1697659752097245e-05, "loss": 0.0309, "step": 78890 }, { "grad_norm": 0.4101680815219879, "learning_rate": 1.1687033594057068e-05, "loss": 0.0433, "step": 78900 }, { "grad_norm": 0.3372408449649811, "learning_rate": 1.1676411625865297e-05, "loss": 0.0365, "step": 78910 }, { "grad_norm": 0.5237115621566772, "learning_rate": 1.1665793848683542e-05, "loss": 0.038, "step": 78920 }, { "grad_norm": 0.5932615995407104, "learning_rate": 1.1655180263672927e-05, "loss": 0.0449, "step": 78930 }, { "grad_norm": 0.3815513849258423, "learning_rate": 1.1644570871994142e-05, "loss": 0.0322, "step": 78940 }, { "grad_norm": 0.49120619893074036, "learning_rate": 1.163396567480744e-05, "loss": 0.0267, "step": 78950 }, { "grad_norm": 0.528894305229187, "learning_rate": 1.1623364673272552e-05, "loss": 0.0337, "step": 78960 }, { "grad_norm": 0.347263902425766, "learning_rate": 1.1612767868548807e-05, "loss": 0.0302, "step": 78970 }, { "grad_norm": 0.48234495520591736, "learning_rate": 1.1602175261795057e-05, "loss": 0.0326, "step": 78980 }, { "grad_norm": 0.39979287981987, "learning_rate": 1.1591586854169696e-05, "loss": 0.0367, "step": 78990 }, { "grad_norm": 0.26593565940856934, "learning_rate": 1.158100264683064e-05, "loss": 0.0263, "step": 79000 }, { "grad_norm": 0.40837329626083374, "learning_rate": 1.1570422640935386e-05, "loss": 0.035, "step": 79010 }, { "grad_norm": 0.3230157196521759, "learning_rate": 1.155984683764092e-05, "loss": 0.0424, "step": 79020 }, { "grad_norm": 0.523739755153656, "learning_rate": 1.1549275238103812e-05, "loss": 0.0313, "step": 79030 }, { "grad_norm": 0.3993072807788849, "learning_rate": 1.1538707843480152e-05, "loss": 0.0342, "step": 79040 }, { "grad_norm": 0.36732420325279236, "learning_rate": 1.1528144654925593e-05, "loss": 0.0362, "step": 79050 }, { "grad_norm": 0.3575275242328644, "learning_rate": 1.1517585673595277e-05, "loss": 0.029, "step": 79060 }, { "grad_norm": 0.30872032046318054, "learning_rate": 1.150703090064395e-05, "loss": 0.0353, "step": 79070 }, { "grad_norm": 0.35226312279701233, "learning_rate": 1.1496480337225835e-05, "loss": 0.0285, "step": 79080 }, { "grad_norm": 0.27329879999160767, "learning_rate": 1.1485933984494751e-05, "loss": 0.0261, "step": 79090 }, { "grad_norm": 0.47439950704574585, "learning_rate": 1.1475391843604e-05, "loss": 0.0306, "step": 79100 }, { "grad_norm": 0.4345412254333496, "learning_rate": 1.1464853915706503e-05, "loss": 0.0403, "step": 79110 }, { "grad_norm": 0.46880021691322327, "learning_rate": 1.1454320201954626e-05, "loss": 0.0294, "step": 79120 }, { "grad_norm": 0.2963392436504364, "learning_rate": 1.1443790703500356e-05, "loss": 0.0336, "step": 79130 }, { "grad_norm": 0.36123088002204895, "learning_rate": 1.1433265421495154e-05, "loss": 0.0354, "step": 79140 }, { "grad_norm": 0.3391476571559906, "learning_rate": 1.1422744357090065e-05, "loss": 0.0387, "step": 79150 }, { "grad_norm": 0.4771658778190613, "learning_rate": 1.1412227511435647e-05, "loss": 0.0367, "step": 79160 }, { "grad_norm": 0.40251290798187256, "learning_rate": 1.1401714885682025e-05, "loss": 0.0294, "step": 79170 }, { "grad_norm": 0.34784767031669617, "learning_rate": 1.1391206480978823e-05, "loss": 0.0249, "step": 79180 }, { "grad_norm": 0.23967164754867554, "learning_rate": 1.1380702298475227e-05, "loss": 0.0348, "step": 79190 }, { "grad_norm": 0.36515581607818604, "learning_rate": 1.1370202339319985e-05, "loss": 0.028, "step": 79200 }, { "grad_norm": 0.3133273422718048, "learning_rate": 1.1359706604661319e-05, "loss": 0.04, "step": 79210 }, { "grad_norm": 0.5475024580955505, "learning_rate": 1.1349215095647042e-05, "loss": 0.0414, "step": 79220 }, { "grad_norm": 0.3239758610725403, "learning_rate": 1.1338727813424493e-05, "loss": 0.0263, "step": 79230 }, { "grad_norm": 0.34874165058135986, "learning_rate": 1.132824475914056e-05, "loss": 0.0288, "step": 79240 }, { "grad_norm": 0.7554962038993835, "learning_rate": 1.131776593394162e-05, "loss": 0.0258, "step": 79250 }, { "grad_norm": 0.46811261773109436, "learning_rate": 1.1307291338973657e-05, "loss": 0.0372, "step": 79260 }, { "grad_norm": 0.18299081921577454, "learning_rate": 1.1296820975382121e-05, "loss": 0.0277, "step": 79270 }, { "grad_norm": 0.395002543926239, "learning_rate": 1.1286354844312074e-05, "loss": 0.0403, "step": 79280 }, { "grad_norm": 0.36939045786857605, "learning_rate": 1.1275892946908023e-05, "loss": 0.0339, "step": 79290 }, { "grad_norm": 0.44417092204093933, "learning_rate": 1.126543528431413e-05, "loss": 0.0462, "step": 79300 }, { "grad_norm": 0.37203872203826904, "learning_rate": 1.125498185767398e-05, "loss": 0.0277, "step": 79310 }, { "grad_norm": 0.29547902941703796, "learning_rate": 1.1244532668130781e-05, "loss": 0.0409, "step": 79320 }, { "grad_norm": 0.29762035608291626, "learning_rate": 1.1234087716827203e-05, "loss": 0.033, "step": 79330 }, { "grad_norm": 0.4046403169631958, "learning_rate": 1.1223647004905524e-05, "loss": 0.0433, "step": 79340 }, { "grad_norm": 1.0108513832092285, "learning_rate": 1.1213210533507485e-05, "loss": 0.0463, "step": 79350 }, { "grad_norm": 0.3300681710243225, "learning_rate": 1.1202778303774447e-05, "loss": 0.0296, "step": 79360 }, { "grad_norm": 0.8850659132003784, "learning_rate": 1.1192350316847234e-05, "loss": 0.0348, "step": 79370 }, { "grad_norm": 0.4761461317539215, "learning_rate": 1.1181926573866258e-05, "loss": 0.0371, "step": 79380 }, { "grad_norm": 0.41509902477264404, "learning_rate": 1.1171507075971416e-05, "loss": 0.0357, "step": 79390 }, { "grad_norm": 0.265775591135025, "learning_rate": 1.1161091824302177e-05, "loss": 0.0286, "step": 79400 }, { "grad_norm": 0.4270344376564026, "learning_rate": 1.115068081999755e-05, "loss": 0.0382, "step": 79410 }, { "grad_norm": 0.5005027055740356, "learning_rate": 1.1140274064196071e-05, "loss": 0.0349, "step": 79420 }, { "grad_norm": 0.31442415714263916, "learning_rate": 1.112987155803578e-05, "loss": 0.03, "step": 79430 }, { "grad_norm": 0.5569708943367004, "learning_rate": 1.1119473302654298e-05, "loss": 0.0341, "step": 79440 }, { "grad_norm": 1.1892499923706055, "learning_rate": 1.1109079299188769e-05, "loss": 0.0365, "step": 79450 }, { "grad_norm": 0.43412235379219055, "learning_rate": 1.1098689548775847e-05, "loss": 0.0345, "step": 79460 }, { "grad_norm": 0.5527733564376831, "learning_rate": 1.1088304052551762e-05, "loss": 0.0314, "step": 79470 }, { "grad_norm": 0.21386684477329254, "learning_rate": 1.1077922811652215e-05, "loss": 0.0257, "step": 79480 }, { "grad_norm": 0.5985848307609558, "learning_rate": 1.1067545827212534e-05, "loss": 0.0302, "step": 79490 }, { "grad_norm": 0.6603556275367737, "learning_rate": 1.1057173100367496e-05, "loss": 0.027, "step": 79500 }, { "grad_norm": 0.36256974935531616, "learning_rate": 1.1046804632251474e-05, "loss": 0.0314, "step": 79510 }, { "grad_norm": 0.5792319774627686, "learning_rate": 1.1036440423998318e-05, "loss": 0.0252, "step": 79520 }, { "grad_norm": 0.5120801329612732, "learning_rate": 1.1026080476741469e-05, "loss": 0.0338, "step": 79530 }, { "grad_norm": 0.4320048689842224, "learning_rate": 1.1015724791613836e-05, "loss": 0.0238, "step": 79540 }, { "grad_norm": 0.46462690830230713, "learning_rate": 1.1005373369747951e-05, "loss": 0.0361, "step": 79550 }, { "grad_norm": 0.7025175094604492, "learning_rate": 1.09950262122758e-05, "loss": 0.0283, "step": 79560 }, { "grad_norm": 0.40179958939552307, "learning_rate": 1.0984683320328948e-05, "loss": 0.0325, "step": 79570 }, { "grad_norm": 0.23259390890598297, "learning_rate": 1.0974344695038458e-05, "loss": 0.0305, "step": 79580 }, { "grad_norm": 0.34292900562286377, "learning_rate": 1.0964010337534968e-05, "loss": 0.0358, "step": 79590 }, { "grad_norm": 0.5549555420875549, "learning_rate": 1.0953680248948611e-05, "loss": 0.0355, "step": 79600 }, { "grad_norm": 0.21614347398281097, "learning_rate": 1.0943354430409069e-05, "loss": 0.0277, "step": 79610 }, { "grad_norm": 0.9502535462379456, "learning_rate": 1.0933032883045574e-05, "loss": 0.0279, "step": 79620 }, { "grad_norm": 0.6202038526535034, "learning_rate": 1.0922715607986872e-05, "loss": 0.0441, "step": 79630 }, { "grad_norm": 0.2933672070503235, "learning_rate": 1.0912402606361234e-05, "loss": 0.0317, "step": 79640 }, { "grad_norm": 0.32647332549095154, "learning_rate": 1.090209387929647e-05, "loss": 0.0228, "step": 79650 }, { "grad_norm": 0.5886098146438599, "learning_rate": 1.0891789427919957e-05, "loss": 0.0282, "step": 79660 }, { "grad_norm": 0.2683091163635254, "learning_rate": 1.088148925335854e-05, "loss": 0.0242, "step": 79670 }, { "grad_norm": 0.44065508246421814, "learning_rate": 1.087119335673864e-05, "loss": 0.0381, "step": 79680 }, { "grad_norm": 0.4224929213523865, "learning_rate": 1.0860901739186208e-05, "loss": 0.0261, "step": 79690 }, { "grad_norm": 0.5682497620582581, "learning_rate": 1.085061440182673e-05, "loss": 0.0293, "step": 79700 }, { "grad_norm": 0.25116512179374695, "learning_rate": 1.0840331345785193e-05, "loss": 0.0292, "step": 79710 }, { "grad_norm": 0.3061070144176483, "learning_rate": 1.083005257218615e-05, "loss": 0.0275, "step": 79720 }, { "grad_norm": 0.3677244186401367, "learning_rate": 1.0819778082153642e-05, "loss": 0.0259, "step": 79730 }, { "grad_norm": 0.4521821439266205, "learning_rate": 1.0809507876811325e-05, "loss": 0.0372, "step": 79740 }, { "grad_norm": 0.2751452624797821, "learning_rate": 1.0799241957282292e-05, "loss": 0.0314, "step": 79750 }, { "grad_norm": 0.3833693265914917, "learning_rate": 1.078898032468923e-05, "loss": 0.0311, "step": 79760 }, { "grad_norm": 0.4376145303249359, "learning_rate": 1.0778722980154315e-05, "loss": 0.0337, "step": 79770 }, { "grad_norm": 0.37773647904396057, "learning_rate": 1.0768469924799302e-05, "loss": 0.0279, "step": 79780 }, { "grad_norm": 0.5657116174697876, "learning_rate": 1.0758221159745419e-05, "loss": 0.0374, "step": 79790 }, { "grad_norm": 0.47665706276893616, "learning_rate": 1.074797668611347e-05, "loss": 0.0276, "step": 79800 }, { "grad_norm": 0.22019170224666595, "learning_rate": 1.0737736505023777e-05, "loss": 0.0313, "step": 79810 }, { "grad_norm": 1.1801568269729614, "learning_rate": 1.0727500617596203e-05, "loss": 0.0372, "step": 79820 }, { "grad_norm": 0.6531089544296265, "learning_rate": 1.0717269024950104e-05, "loss": 0.0289, "step": 79830 }, { "grad_norm": 0.312494695186615, "learning_rate": 1.070704172820442e-05, "loss": 0.035, "step": 79840 }, { "grad_norm": 0.2467990517616272, "learning_rate": 1.0696818728477559e-05, "loss": 0.031, "step": 79850 }, { "grad_norm": 0.23635445535182953, "learning_rate": 1.0686600026887511e-05, "loss": 0.0352, "step": 79860 }, { "grad_norm": 0.48986512422561646, "learning_rate": 1.0676385624551777e-05, "loss": 0.0368, "step": 79870 }, { "grad_norm": 0.4054761528968811, "learning_rate": 1.0666175522587402e-05, "loss": 0.0352, "step": 79880 }, { "grad_norm": 0.2538701593875885, "learning_rate": 1.0655969722110925e-05, "loss": 0.0259, "step": 79890 }, { "grad_norm": 0.6337994337081909, "learning_rate": 1.0645768224238439e-05, "loss": 0.0302, "step": 79900 }, { "grad_norm": 0.47749394178390503, "learning_rate": 1.063557103008559e-05, "loss": 0.0305, "step": 79910 }, { "grad_norm": 0.1714038848876953, "learning_rate": 1.0625378140767494e-05, "loss": 0.0253, "step": 79920 }, { "grad_norm": 0.3782270848751068, "learning_rate": 1.0615189557398846e-05, "loss": 0.0239, "step": 79930 }, { "grad_norm": 0.42705902457237244, "learning_rate": 1.060500528109386e-05, "loss": 0.0329, "step": 79940 }, { "grad_norm": 0.369065523147583, "learning_rate": 1.0594825312966272e-05, "loss": 0.0282, "step": 79950 }, { "grad_norm": 0.4409780502319336, "learning_rate": 1.0584649654129336e-05, "loss": 0.034, "step": 79960 }, { "grad_norm": 0.25364115834236145, "learning_rate": 1.0574478305695863e-05, "loss": 0.0187, "step": 79970 }, { "grad_norm": 0.35015952587127686, "learning_rate": 1.0564311268778154e-05, "loss": 0.0326, "step": 79980 }, { "grad_norm": 0.4001951217651367, "learning_rate": 1.0554148544488073e-05, "loss": 0.0337, "step": 79990 }, { "grad_norm": 0.61025071144104, "learning_rate": 1.0543990133937004e-05, "loss": 0.0352, "step": 80000 }, { "grad_norm": 0.50816410779953, "learning_rate": 1.0533836038235868e-05, "loss": 0.0319, "step": 80010 }, { "grad_norm": 0.5057185292243958, "learning_rate": 1.052368625849507e-05, "loss": 0.026, "step": 80020 }, { "grad_norm": 0.4164060950279236, "learning_rate": 1.05135407958246e-05, "loss": 0.0232, "step": 80030 }, { "grad_norm": 0.24733009934425354, "learning_rate": 1.0503399651333934e-05, "loss": 0.0353, "step": 80040 }, { "grad_norm": 0.47762587666511536, "learning_rate": 1.04932628261321e-05, "loss": 0.0303, "step": 80050 }, { "grad_norm": 0.3397800624370575, "learning_rate": 1.0483130321327644e-05, "loss": 0.0285, "step": 80060 }, { "grad_norm": 0.226708322763443, "learning_rate": 1.0473002138028654e-05, "loss": 0.0252, "step": 80070 }, { "grad_norm": 0.5665169358253479, "learning_rate": 1.0462878277342713e-05, "loss": 0.0286, "step": 80080 }, { "grad_norm": 0.36146819591522217, "learning_rate": 1.0452758740376972e-05, "loss": 0.029, "step": 80090 }, { "grad_norm": 0.2712392508983612, "learning_rate": 1.0442643528238067e-05, "loss": 0.0272, "step": 80100 }, { "grad_norm": 0.3864806890487671, "learning_rate": 1.0432532642032188e-05, "loss": 0.0311, "step": 80110 }, { "grad_norm": 0.23628340661525726, "learning_rate": 1.042242608286505e-05, "loss": 0.0333, "step": 80120 }, { "grad_norm": 0.7321776747703552, "learning_rate": 1.041232385184191e-05, "loss": 0.0305, "step": 80130 }, { "grad_norm": 0.520633339881897, "learning_rate": 1.0402225950067495e-05, "loss": 0.0298, "step": 80140 }, { "grad_norm": 0.3101503252983093, "learning_rate": 1.0392132378646119e-05, "loss": 0.0301, "step": 80150 }, { "grad_norm": 0.5474132299423218, "learning_rate": 1.0382043138681607e-05, "loss": 0.0372, "step": 80160 }, { "grad_norm": 0.4373818635940552, "learning_rate": 1.0371958231277284e-05, "loss": 0.0193, "step": 80170 }, { "grad_norm": 0.28529512882232666, "learning_rate": 1.0361877657536024e-05, "loss": 0.0308, "step": 80180 }, { "grad_norm": 0.4196888506412506, "learning_rate": 1.035180141856023e-05, "loss": 0.0384, "step": 80190 }, { "grad_norm": 0.37331971526145935, "learning_rate": 1.0341729515451836e-05, "loss": 0.028, "step": 80200 }, { "grad_norm": 0.3553535044193268, "learning_rate": 1.0331661949312265e-05, "loss": 0.0314, "step": 80210 }, { "grad_norm": 0.5899164080619812, "learning_rate": 1.0321598721242514e-05, "loss": 0.036, "step": 80220 }, { "grad_norm": 0.3649804890155792, "learning_rate": 1.0311539832343064e-05, "loss": 0.0351, "step": 80230 }, { "grad_norm": 0.3520263433456421, "learning_rate": 1.0301485283713942e-05, "loss": 0.0321, "step": 80240 }, { "grad_norm": 0.43870899081230164, "learning_rate": 1.0291435076454703e-05, "loss": 0.0361, "step": 80250 }, { "grad_norm": 0.4178590476512909, "learning_rate": 1.028138921166444e-05, "loss": 0.0329, "step": 80260 }, { "grad_norm": 0.3507283926010132, "learning_rate": 1.0271347690441719e-05, "loss": 0.0249, "step": 80270 }, { "grad_norm": 0.31510069966316223, "learning_rate": 1.0261310513884698e-05, "loss": 0.0327, "step": 80280 }, { "grad_norm": 0.48763445019721985, "learning_rate": 1.0251277683090998e-05, "loss": 0.0284, "step": 80290 }, { "grad_norm": 0.3625052571296692, "learning_rate": 1.0241249199157826e-05, "loss": 0.039, "step": 80300 }, { "grad_norm": 0.23275190591812134, "learning_rate": 1.0231225063181837e-05, "loss": 0.0254, "step": 80310 }, { "grad_norm": 0.5363839268684387, "learning_rate": 1.0221205276259316e-05, "loss": 0.0311, "step": 80320 }, { "grad_norm": 0.5219233632087708, "learning_rate": 1.0211189839485958e-05, "loss": 0.0363, "step": 80330 }, { "grad_norm": 0.6898338794708252, "learning_rate": 1.0201178753957075e-05, "loss": 0.0357, "step": 80340 }, { "grad_norm": 0.45505672693252563, "learning_rate": 1.0191172020767432e-05, "loss": 0.0324, "step": 80350 }, { "grad_norm": 0.30982261896133423, "learning_rate": 1.0181169641011362e-05, "loss": 0.0246, "step": 80360 }, { "grad_norm": 0.40812405943870544, "learning_rate": 1.0171171615782721e-05, "loss": 0.0344, "step": 80370 }, { "grad_norm": 0.37519359588623047, "learning_rate": 1.0161177946174866e-05, "loss": 0.0382, "step": 80380 }, { "grad_norm": 0.35036784410476685, "learning_rate": 1.0151188633280706e-05, "loss": 0.0273, "step": 80390 }, { "grad_norm": 0.40563374757766724, "learning_rate": 1.0141203678192635e-05, "loss": 0.0367, "step": 80400 }, { "grad_norm": 0.6435106992721558, "learning_rate": 1.0131223082002612e-05, "loss": 0.036, "step": 80410 }, { "grad_norm": 0.30691805481910706, "learning_rate": 1.012124684580208e-05, "loss": 0.0249, "step": 80420 }, { "grad_norm": 0.29357266426086426, "learning_rate": 1.0111274970682039e-05, "loss": 0.0301, "step": 80430 }, { "grad_norm": 0.44687479734420776, "learning_rate": 1.010130745773299e-05, "loss": 0.032, "step": 80440 }, { "grad_norm": 0.43375951051712036, "learning_rate": 1.0091344308044986e-05, "loss": 0.027, "step": 80450 }, { "grad_norm": 0.31292951107025146, "learning_rate": 1.0081385522707554e-05, "loss": 0.0245, "step": 80460 }, { "grad_norm": 0.48404261469841003, "learning_rate": 1.0071431102809803e-05, "loss": 0.031, "step": 80470 }, { "grad_norm": 0.268764853477478, "learning_rate": 1.0061481049440297e-05, "loss": 0.0396, "step": 80480 }, { "grad_norm": 0.34068259596824646, "learning_rate": 1.0051535363687187e-05, "loss": 0.0221, "step": 80490 }, { "grad_norm": 0.4648759961128235, "learning_rate": 1.0041594046638087e-05, "loss": 0.0296, "step": 80500 }, { "grad_norm": 0.34393757581710815, "learning_rate": 1.0031657099380204e-05, "loss": 0.0259, "step": 80510 }, { "grad_norm": 0.352963387966156, "learning_rate": 1.0021724523000204e-05, "loss": 0.0321, "step": 80520 }, { "grad_norm": 0.3018799424171448, "learning_rate": 1.001179631858431e-05, "loss": 0.0334, "step": 80530 }, { "grad_norm": 0.37369751930236816, "learning_rate": 1.0001872487218245e-05, "loss": 0.0308, "step": 80540 }, { "grad_norm": 0.20414596796035767, "learning_rate": 9.991953029987272e-06, "loss": 0.0238, "step": 80550 }, { "grad_norm": 0.39887535572052, "learning_rate": 9.982037947976147e-06, "loss": 0.0367, "step": 80560 }, { "grad_norm": 0.43409666419029236, "learning_rate": 9.972127242269208e-06, "loss": 0.0347, "step": 80570 }, { "grad_norm": 0.2932397723197937, "learning_rate": 9.962220913950238e-06, "loss": 0.0416, "step": 80580 }, { "grad_norm": 0.424906462430954, "learning_rate": 9.952318964102591e-06, "loss": 0.0432, "step": 80590 }, { "grad_norm": 0.7074607610702515, "learning_rate": 9.942421393809148e-06, "loss": 0.0369, "step": 80600 }, { "grad_norm": 0.3149346709251404, "learning_rate": 9.932528204152264e-06, "loss": 0.0305, "step": 80610 }, { "grad_norm": 0.6767365336418152, "learning_rate": 9.922639396213856e-06, "loss": 0.0531, "step": 80620 }, { "grad_norm": 0.5722572803497314, "learning_rate": 9.912754971075344e-06, "loss": 0.0402, "step": 80630 }, { "grad_norm": 0.36866191029548645, "learning_rate": 9.902874929817696e-06, "loss": 0.0293, "step": 80640 }, { "grad_norm": 0.26298487186431885, "learning_rate": 9.892999273521342e-06, "loss": 0.0281, "step": 80650 }, { "grad_norm": 0.38386034965515137, "learning_rate": 9.883128003266302e-06, "loss": 0.0308, "step": 80660 }, { "grad_norm": 0.416878879070282, "learning_rate": 9.87326112013206e-06, "loss": 0.0271, "step": 80670 }, { "grad_norm": 0.47638174891471863, "learning_rate": 9.863398625197662e-06, "loss": 0.0363, "step": 80680 }, { "grad_norm": 0.6762905120849609, "learning_rate": 9.853540519541615e-06, "loss": 0.0273, "step": 80690 }, { "grad_norm": 0.7409453392028809, "learning_rate": 9.843686804242053e-06, "loss": 0.0442, "step": 80700 }, { "grad_norm": 0.41248342394828796, "learning_rate": 9.833837480376506e-06, "loss": 0.0373, "step": 80710 }, { "grad_norm": 0.363180547952652, "learning_rate": 9.82399254902212e-06, "loss": 0.0408, "step": 80720 }, { "grad_norm": 0.47584137320518494, "learning_rate": 9.814152011255494e-06, "loss": 0.0294, "step": 80730 }, { "grad_norm": 0.3733151853084564, "learning_rate": 9.804315868152797e-06, "loss": 0.0291, "step": 80740 }, { "grad_norm": 0.444070041179657, "learning_rate": 9.79448412078966e-06, "loss": 0.0307, "step": 80750 }, { "grad_norm": 0.5193219780921936, "learning_rate": 9.784656770241312e-06, "loss": 0.0333, "step": 80760 }, { "grad_norm": 0.3618786334991455, "learning_rate": 9.77483381758243e-06, "loss": 0.0221, "step": 80770 }, { "grad_norm": 0.553356409072876, "learning_rate": 9.76501526388725e-06, "loss": 0.0341, "step": 80780 }, { "grad_norm": 0.2296314239501953, "learning_rate": 9.7552011102295e-06, "loss": 0.0448, "step": 80790 }, { "grad_norm": 0.31971943378448486, "learning_rate": 9.745391357682449e-06, "loss": 0.0298, "step": 80800 }, { "grad_norm": 0.13195596635341644, "learning_rate": 9.735586007318875e-06, "loss": 0.0446, "step": 80810 }, { "grad_norm": 0.26933708786964417, "learning_rate": 9.725785060211096e-06, "loss": 0.0303, "step": 80820 }, { "grad_norm": 0.5649427175521851, "learning_rate": 9.715988517430896e-06, "loss": 0.0363, "step": 80830 }, { "grad_norm": 0.45273569226264954, "learning_rate": 9.706196380049625e-06, "loss": 0.0227, "step": 80840 }, { "grad_norm": 0.34319576621055603, "learning_rate": 9.69640864913815e-06, "loss": 0.0312, "step": 80850 }, { "grad_norm": 0.3091708719730377, "learning_rate": 9.686625325766818e-06, "loss": 0.027, "step": 80860 }, { "grad_norm": 0.2706906795501709, "learning_rate": 9.67684641100554e-06, "loss": 0.0306, "step": 80870 }, { "grad_norm": 0.523475706577301, "learning_rate": 9.667071905923692e-06, "loss": 0.0365, "step": 80880 }, { "grad_norm": 0.6070840358734131, "learning_rate": 9.657301811590242e-06, "loss": 0.0286, "step": 80890 }, { "grad_norm": 0.42655953764915466, "learning_rate": 9.647536129073603e-06, "loss": 0.0296, "step": 80900 }, { "grad_norm": 0.35573285818099976, "learning_rate": 9.637774859441751e-06, "loss": 0.0248, "step": 80910 }, { "grad_norm": 0.6795060634613037, "learning_rate": 9.628018003762152e-06, "loss": 0.0255, "step": 80920 }, { "grad_norm": 0.814237117767334, "learning_rate": 9.618265563101813e-06, "loss": 0.0323, "step": 80930 }, { "grad_norm": 0.3884563744068146, "learning_rate": 9.608517538527211e-06, "loss": 0.0283, "step": 80940 }, { "grad_norm": 0.3776196539402008, "learning_rate": 9.598773931104431e-06, "loss": 0.034, "step": 80950 }, { "grad_norm": 0.9727234244346619, "learning_rate": 9.589034741898983e-06, "loss": 0.03, "step": 80960 }, { "grad_norm": 0.4542592763900757, "learning_rate": 9.579299971975947e-06, "loss": 0.0282, "step": 80970 }, { "grad_norm": 0.34462910890579224, "learning_rate": 9.569569622399882e-06, "loss": 0.0261, "step": 80980 }, { "grad_norm": 0.30612531304359436, "learning_rate": 9.559843694234916e-06, "loss": 0.0249, "step": 80990 }, { "grad_norm": 0.42304089665412903, "learning_rate": 9.550122188544625e-06, "loss": 0.0327, "step": 81000 }, { "grad_norm": 0.4837791919708252, "learning_rate": 9.540405106392152e-06, "loss": 0.0295, "step": 81010 }, { "grad_norm": 0.34442055225372314, "learning_rate": 9.53069244884015e-06, "loss": 0.0347, "step": 81020 }, { "grad_norm": 0.399904727935791, "learning_rate": 9.52098421695079e-06, "loss": 0.0258, "step": 81030 }, { "grad_norm": 0.622924268245697, "learning_rate": 9.511280411785728e-06, "loss": 0.0272, "step": 81040 }, { "grad_norm": 0.3627953827381134, "learning_rate": 9.50158103440616e-06, "loss": 0.035, "step": 81050 }, { "grad_norm": 0.2963894307613373, "learning_rate": 9.491886085872815e-06, "loss": 0.0275, "step": 81060 }, { "grad_norm": 0.3405081331729889, "learning_rate": 9.482195567245895e-06, "loss": 0.0306, "step": 81070 }, { "grad_norm": 0.3482404947280884, "learning_rate": 9.472509479585151e-06, "loss": 0.0392, "step": 81080 }, { "grad_norm": 0.4390363395214081, "learning_rate": 9.462827823949833e-06, "loss": 0.0326, "step": 81090 }, { "grad_norm": 1.0503106117248535, "learning_rate": 9.453150601398735e-06, "loss": 0.0313, "step": 81100 }, { "grad_norm": 0.35411354899406433, "learning_rate": 9.44347781299011e-06, "loss": 0.0273, "step": 81110 }, { "grad_norm": 0.3186589479446411, "learning_rate": 9.43380945978179e-06, "loss": 0.0357, "step": 81120 }, { "grad_norm": 0.4498881995677948, "learning_rate": 9.424145542831047e-06, "loss": 0.0237, "step": 81130 }, { "grad_norm": 0.37288910150527954, "learning_rate": 9.414486063194778e-06, "loss": 0.0302, "step": 81140 }, { "grad_norm": 0.840711236000061, "learning_rate": 9.404831021929273e-06, "loss": 0.0374, "step": 81150 }, { "grad_norm": 0.4229736328125, "learning_rate": 9.395180420090427e-06, "loss": 0.0323, "step": 81160 }, { "grad_norm": 0.2932587265968323, "learning_rate": 9.385534258733591e-06, "loss": 0.0233, "step": 81170 }, { "grad_norm": 0.4275265336036682, "learning_rate": 9.375892538913677e-06, "loss": 0.0266, "step": 81180 }, { "grad_norm": 0.24137888848781586, "learning_rate": 9.36625526168507e-06, "loss": 0.0296, "step": 81190 }, { "grad_norm": 0.33239057660102844, "learning_rate": 9.356622428101686e-06, "loss": 0.0274, "step": 81200 }, { "grad_norm": 0.33999645709991455, "learning_rate": 9.346994039216972e-06, "loss": 0.0281, "step": 81210 }, { "grad_norm": 0.20815637707710266, "learning_rate": 9.337370096083886e-06, "loss": 0.0222, "step": 81220 }, { "grad_norm": 0.36045369505882263, "learning_rate": 9.327750599754854e-06, "loss": 0.0327, "step": 81230 }, { "grad_norm": 0.25571581721305847, "learning_rate": 9.318135551281875e-06, "loss": 0.0325, "step": 81240 }, { "grad_norm": 0.32836923003196716, "learning_rate": 9.308524951716424e-06, "loss": 0.0347, "step": 81250 }, { "grad_norm": 0.26295503973960876, "learning_rate": 9.298918802109497e-06, "loss": 0.0303, "step": 81260 }, { "grad_norm": 0.7370248436927795, "learning_rate": 9.289317103511625e-06, "loss": 0.0305, "step": 81270 }, { "grad_norm": 0.46367570757865906, "learning_rate": 9.279719856972835e-06, "loss": 0.0291, "step": 81280 }, { "grad_norm": 0.3510417342185974, "learning_rate": 9.270127063542645e-06, "loss": 0.0364, "step": 81290 }, { "grad_norm": 0.3652357757091522, "learning_rate": 9.260538724270124e-06, "loss": 0.0291, "step": 81300 }, { "grad_norm": 0.5306499004364014, "learning_rate": 9.250954840203852e-06, "loss": 0.0267, "step": 81310 }, { "grad_norm": 0.4280114471912384, "learning_rate": 9.241375412391878e-06, "loss": 0.0306, "step": 81320 }, { "grad_norm": 0.25157222151756287, "learning_rate": 9.23180044188181e-06, "loss": 0.0281, "step": 81330 }, { "grad_norm": 0.34394800662994385, "learning_rate": 9.222229929720755e-06, "loss": 0.0264, "step": 81340 }, { "grad_norm": 0.37054356932640076, "learning_rate": 9.212663876955336e-06, "loss": 0.0457, "step": 81350 }, { "grad_norm": 0.43639859557151794, "learning_rate": 9.20310228463166e-06, "loss": 0.0404, "step": 81360 }, { "grad_norm": 0.3916729688644409, "learning_rate": 9.19354515379539e-06, "loss": 0.0292, "step": 81370 }, { "grad_norm": 0.43753641843795776, "learning_rate": 9.183992485491666e-06, "loss": 0.0274, "step": 81380 }, { "grad_norm": 0.7531238198280334, "learning_rate": 9.174444280765154e-06, "loss": 0.0383, "step": 81390 }, { "grad_norm": 0.375394344329834, "learning_rate": 9.164900540660032e-06, "loss": 0.0263, "step": 81400 }, { "grad_norm": 0.302064448595047, "learning_rate": 9.155361266220008e-06, "loss": 0.0336, "step": 81410 }, { "grad_norm": 0.5916458368301392, "learning_rate": 9.145826458488254e-06, "loss": 0.0294, "step": 81420 }, { "grad_norm": 0.25700533390045166, "learning_rate": 9.136296118507509e-06, "loss": 0.033, "step": 81430 }, { "grad_norm": 0.5094600319862366, "learning_rate": 9.126770247319966e-06, "loss": 0.0304, "step": 81440 }, { "grad_norm": 0.3967740535736084, "learning_rate": 9.11724884596738e-06, "loss": 0.0174, "step": 81450 }, { "grad_norm": 0.49931037425994873, "learning_rate": 9.107731915490996e-06, "loss": 0.0284, "step": 81460 }, { "grad_norm": 0.41709014773368835, "learning_rate": 9.098219456931579e-06, "loss": 0.0238, "step": 81470 }, { "grad_norm": 0.4919070303440094, "learning_rate": 9.088711471329376e-06, "loss": 0.0367, "step": 81480 }, { "grad_norm": 2.7839930057525635, "learning_rate": 9.079207959724184e-06, "loss": 0.0403, "step": 81490 }, { "grad_norm": 0.31632715463638306, "learning_rate": 9.069708923155279e-06, "loss": 0.0324, "step": 81500 }, { "grad_norm": 0.5480285882949829, "learning_rate": 9.060214362661468e-06, "loss": 0.0318, "step": 81510 }, { "grad_norm": 0.27032962441444397, "learning_rate": 9.050724279281064e-06, "loss": 0.0308, "step": 81520 }, { "grad_norm": 0.43811771273612976, "learning_rate": 9.041238674051894e-06, "loss": 0.033, "step": 81530 }, { "grad_norm": 0.5482845902442932, "learning_rate": 9.031757548011277e-06, "loss": 0.0437, "step": 81540 }, { "grad_norm": 1.3113116025924683, "learning_rate": 9.022280902196051e-06, "loss": 0.0307, "step": 81550 }, { "grad_norm": 0.296308696269989, "learning_rate": 9.012808737642592e-06, "loss": 0.0273, "step": 81560 }, { "grad_norm": 0.3720758557319641, "learning_rate": 9.003341055386739e-06, "loss": 0.0366, "step": 81570 }, { "grad_norm": 0.4227730333805084, "learning_rate": 8.993877856463862e-06, "loss": 0.0269, "step": 81580 }, { "grad_norm": 0.2225104421377182, "learning_rate": 8.984419141908856e-06, "loss": 0.0238, "step": 81590 }, { "grad_norm": 0.23203925788402557, "learning_rate": 8.974964912756117e-06, "loss": 0.0308, "step": 81600 }, { "grad_norm": 0.6150721907615662, "learning_rate": 8.965515170039523e-06, "loss": 0.0351, "step": 81610 }, { "grad_norm": 0.37267622351646423, "learning_rate": 8.956069914792514e-06, "loss": 0.0238, "step": 81620 }, { "grad_norm": 0.39516204595565796, "learning_rate": 8.946629148047969e-06, "loss": 0.0306, "step": 81630 }, { "grad_norm": 0.4493127167224884, "learning_rate": 8.93719287083834e-06, "loss": 0.0362, "step": 81640 }, { "grad_norm": 0.4101414680480957, "learning_rate": 8.927761084195569e-06, "loss": 0.0323, "step": 81650 }, { "grad_norm": 0.3728317618370056, "learning_rate": 8.918333789151096e-06, "loss": 0.0268, "step": 81660 }, { "grad_norm": 0.37166884541511536, "learning_rate": 8.90891098673587e-06, "loss": 0.0321, "step": 81670 }, { "grad_norm": 0.21381233632564545, "learning_rate": 8.89949267798037e-06, "loss": 0.0245, "step": 81680 }, { "grad_norm": 0.2548106908798218, "learning_rate": 8.890078863914542e-06, "loss": 0.03, "step": 81690 }, { "grad_norm": 0.2889532744884491, "learning_rate": 8.880669545567894e-06, "loss": 0.0299, "step": 81700 }, { "grad_norm": 0.44837266206741333, "learning_rate": 8.871264723969381e-06, "loss": 0.0314, "step": 81710 }, { "grad_norm": 0.7392489314079285, "learning_rate": 8.86186440014754e-06, "loss": 0.0427, "step": 81720 }, { "grad_norm": 0.6341367959976196, "learning_rate": 8.852468575130351e-06, "loss": 0.0338, "step": 81730 }, { "grad_norm": 0.3065580129623413, "learning_rate": 8.843077249945347e-06, "loss": 0.0268, "step": 81740 }, { "grad_norm": 0.8278543949127197, "learning_rate": 8.833690425619523e-06, "loss": 0.0314, "step": 81750 }, { "grad_norm": 0.29542744159698486, "learning_rate": 8.82430810317942e-06, "loss": 0.0341, "step": 81760 }, { "grad_norm": 0.31439366936683655, "learning_rate": 8.814930283651074e-06, "loss": 0.0279, "step": 81770 }, { "grad_norm": 0.23276080191135406, "learning_rate": 8.805556968060048e-06, "loss": 0.0283, "step": 81780 }, { "grad_norm": 0.38529250025749207, "learning_rate": 8.796188157431362e-06, "loss": 0.0355, "step": 81790 }, { "grad_norm": 0.38161954283714294, "learning_rate": 8.786823852789595e-06, "loss": 0.0233, "step": 81800 }, { "grad_norm": 0.5155854225158691, "learning_rate": 8.777464055158813e-06, "loss": 0.0361, "step": 81810 }, { "grad_norm": 0.21103832125663757, "learning_rate": 8.768108765562578e-06, "loss": 0.0319, "step": 81820 }, { "grad_norm": 0.3543279767036438, "learning_rate": 8.758757985023975e-06, "loss": 0.0326, "step": 81830 }, { "grad_norm": 0.2511395514011383, "learning_rate": 8.749411714565598e-06, "loss": 0.0225, "step": 81840 }, { "grad_norm": 0.4460477828979492, "learning_rate": 8.74006995520954e-06, "loss": 0.0373, "step": 81850 }, { "grad_norm": 0.4880077540874481, "learning_rate": 8.730732707977397e-06, "loss": 0.0349, "step": 81860 }, { "grad_norm": 0.5205329656600952, "learning_rate": 8.721399973890281e-06, "loss": 0.0294, "step": 81870 }, { "grad_norm": 0.3254295289516449, "learning_rate": 8.712071753968792e-06, "loss": 0.0314, "step": 81880 }, { "grad_norm": 0.6474516987800598, "learning_rate": 8.702748049233072e-06, "loss": 0.035, "step": 81890 }, { "grad_norm": 0.29604673385620117, "learning_rate": 8.693428860702718e-06, "loss": 0.0249, "step": 81900 }, { "grad_norm": 0.41202688217163086, "learning_rate": 8.684114189396898e-06, "loss": 0.0285, "step": 81910 }, { "grad_norm": 0.20489022135734558, "learning_rate": 8.674804036334222e-06, "loss": 0.0309, "step": 81920 }, { "grad_norm": 0.3834124207496643, "learning_rate": 8.665498402532857e-06, "loss": 0.0235, "step": 81930 }, { "grad_norm": 0.506354033946991, "learning_rate": 8.656197289010432e-06, "loss": 0.0348, "step": 81940 }, { "grad_norm": 0.4593655467033386, "learning_rate": 8.646900696784122e-06, "loss": 0.026, "step": 81950 }, { "grad_norm": 0.6062418818473816, "learning_rate": 8.637608626870553e-06, "loss": 0.0304, "step": 81960 }, { "grad_norm": 0.36238688230514526, "learning_rate": 8.628321080285944e-06, "loss": 0.0293, "step": 81970 }, { "grad_norm": 0.7032416462898254, "learning_rate": 8.619038058045925e-06, "loss": 0.042, "step": 81980 }, { "grad_norm": 0.3582271337509155, "learning_rate": 8.609759561165704e-06, "loss": 0.0255, "step": 81990 }, { "grad_norm": 0.5952491164207458, "learning_rate": 8.600485590659934e-06, "loss": 0.0324, "step": 82000 }, { "grad_norm": 0.6201675534248352, "learning_rate": 8.591216147542813e-06, "loss": 0.0261, "step": 82010 }, { "grad_norm": 0.27879461646080017, "learning_rate": 8.581951232828045e-06, "loss": 0.0477, "step": 82020 }, { "grad_norm": 0.39266133308410645, "learning_rate": 8.572690847528825e-06, "loss": 0.0274, "step": 82030 }, { "grad_norm": 0.5646823048591614, "learning_rate": 8.563434992657838e-06, "loss": 0.03, "step": 82040 }, { "grad_norm": 0.5082781314849854, "learning_rate": 8.554183669227307e-06, "loss": 0.0275, "step": 82050 }, { "grad_norm": 0.6347107887268066, "learning_rate": 8.544936878248937e-06, "loss": 0.0327, "step": 82060 }, { "grad_norm": 0.4141652584075928, "learning_rate": 8.535694620733942e-06, "loss": 0.0314, "step": 82070 }, { "grad_norm": 0.37744125723838806, "learning_rate": 8.526456897693053e-06, "loss": 0.0286, "step": 82080 }, { "grad_norm": 0.5342274308204651, "learning_rate": 8.517223710136456e-06, "loss": 0.033, "step": 82090 }, { "grad_norm": 0.723568320274353, "learning_rate": 8.507995059073931e-06, "loss": 0.0303, "step": 82100 }, { "grad_norm": 0.40307748317718506, "learning_rate": 8.498770945514673e-06, "loss": 0.0381, "step": 82110 }, { "grad_norm": 0.28050798177719116, "learning_rate": 8.489551370467436e-06, "loss": 0.025, "step": 82120 }, { "grad_norm": 0.29038137197494507, "learning_rate": 8.480336334940442e-06, "loss": 0.0323, "step": 82130 }, { "grad_norm": 0.3616011142730713, "learning_rate": 8.471125839941458e-06, "loss": 0.026, "step": 82140 }, { "grad_norm": 0.6923468708992004, "learning_rate": 8.461919886477682e-06, "loss": 0.0323, "step": 82150 }, { "grad_norm": 0.6034029722213745, "learning_rate": 8.452718475555927e-06, "loss": 0.0313, "step": 82160 }, { "grad_norm": 0.2255084067583084, "learning_rate": 8.443521608182398e-06, "loss": 0.0251, "step": 82170 }, { "grad_norm": 0.26350006461143494, "learning_rate": 8.43432928536288e-06, "loss": 0.0339, "step": 82180 }, { "grad_norm": 0.47308146953582764, "learning_rate": 8.425141508102607e-06, "loss": 0.0318, "step": 82190 }, { "grad_norm": 0.26048940420150757, "learning_rate": 8.415958277406365e-06, "loss": 0.0269, "step": 82200 }, { "grad_norm": 0.2765001952648163, "learning_rate": 8.406779594278392e-06, "loss": 0.0322, "step": 82210 }, { "grad_norm": 0.6023885607719421, "learning_rate": 8.397605459722468e-06, "loss": 0.0285, "step": 82220 }, { "grad_norm": 0.4876737594604492, "learning_rate": 8.388435874741863e-06, "loss": 0.0351, "step": 82230 }, { "grad_norm": 0.4295814633369446, "learning_rate": 8.37927084033936e-06, "loss": 0.0326, "step": 82240 }, { "grad_norm": 0.5029999613761902, "learning_rate": 8.370110357517208e-06, "loss": 0.0314, "step": 82250 }, { "grad_norm": 0.30002906918525696, "learning_rate": 8.360954427277201e-06, "loss": 0.0328, "step": 82260 }, { "grad_norm": 0.17304781079292297, "learning_rate": 8.351803050620627e-06, "loss": 0.0326, "step": 82270 }, { "grad_norm": 0.6191108822822571, "learning_rate": 8.34265622854824e-06, "loss": 0.0319, "step": 82280 }, { "grad_norm": 0.34377479553222656, "learning_rate": 8.333513962060335e-06, "loss": 0.0457, "step": 82290 }, { "grad_norm": 0.3409910202026367, "learning_rate": 8.324376252156701e-06, "loss": 0.0186, "step": 82300 }, { "grad_norm": 0.3415186107158661, "learning_rate": 8.315243099836633e-06, "loss": 0.0286, "step": 82310 }, { "grad_norm": 0.5951178073883057, "learning_rate": 8.3061145060989e-06, "loss": 0.0394, "step": 82320 }, { "grad_norm": 0.34900274872779846, "learning_rate": 8.296990471941806e-06, "loss": 0.0375, "step": 82330 }, { "grad_norm": 0.47214049100875854, "learning_rate": 8.287870998363112e-06, "loss": 0.0359, "step": 82340 }, { "grad_norm": 0.27872851490974426, "learning_rate": 8.278756086360156e-06, "loss": 0.0359, "step": 82350 }, { "grad_norm": 0.3348315358161926, "learning_rate": 8.269645736929693e-06, "loss": 0.0289, "step": 82360 }, { "grad_norm": 0.4247249960899353, "learning_rate": 8.260539951068048e-06, "loss": 0.0322, "step": 82370 }, { "grad_norm": 0.4444196820259094, "learning_rate": 8.251438729770983e-06, "loss": 0.0406, "step": 82380 }, { "grad_norm": 0.34633007645606995, "learning_rate": 8.242342074033826e-06, "loss": 0.0311, "step": 82390 }, { "grad_norm": 0.4108966588973999, "learning_rate": 8.233249984851344e-06, "loss": 0.0346, "step": 82400 }, { "grad_norm": 0.6728194952011108, "learning_rate": 8.224162463217849e-06, "loss": 0.0414, "step": 82410 }, { "grad_norm": 0.3536645472049713, "learning_rate": 8.215079510127139e-06, "loss": 0.0274, "step": 82420 }, { "grad_norm": 0.49468526244163513, "learning_rate": 8.206001126572521e-06, "loss": 0.0358, "step": 82430 }, { "grad_norm": 0.5294720530509949, "learning_rate": 8.196927313546776e-06, "loss": 0.0354, "step": 82440 }, { "grad_norm": 0.22915728390216827, "learning_rate": 8.18785807204222e-06, "loss": 0.036, "step": 82450 }, { "grad_norm": 0.4642317295074463, "learning_rate": 8.178793403050627e-06, "loss": 0.0264, "step": 82460 }, { "grad_norm": 0.32250770926475525, "learning_rate": 8.16973330756331e-06, "loss": 0.0319, "step": 82470 }, { "grad_norm": 0.4807886481285095, "learning_rate": 8.16067778657107e-06, "loss": 0.0332, "step": 82480 }, { "grad_norm": 0.7072001099586487, "learning_rate": 8.151626841064214e-06, "loss": 0.026, "step": 82490 }, { "grad_norm": 0.7729419469833374, "learning_rate": 8.142580472032518e-06, "loss": 0.0316, "step": 82500 }, { "grad_norm": 0.5208311080932617, "learning_rate": 8.133538680465285e-06, "loss": 0.0245, "step": 82510 }, { "grad_norm": 0.4892098903656006, "learning_rate": 8.124501467351336e-06, "loss": 0.0346, "step": 82520 }, { "grad_norm": 0.6883528828620911, "learning_rate": 8.11546883367893e-06, "loss": 0.0247, "step": 82530 }, { "grad_norm": 0.2484077364206314, "learning_rate": 8.106440780435882e-06, "loss": 0.0287, "step": 82540 }, { "grad_norm": 0.2114873230457306, "learning_rate": 8.097417308609485e-06, "loss": 0.0338, "step": 82550 }, { "grad_norm": 0.2891177833080292, "learning_rate": 8.088398419186544e-06, "loss": 0.03, "step": 82560 }, { "grad_norm": 0.5546797513961792, "learning_rate": 8.079384113153321e-06, "loss": 0.0355, "step": 82570 }, { "grad_norm": 0.3799384832382202, "learning_rate": 8.070374391495644e-06, "loss": 0.0244, "step": 82580 }, { "grad_norm": 0.4001219570636749, "learning_rate": 8.061369255198764e-06, "loss": 0.0246, "step": 82590 }, { "grad_norm": 0.5201351046562195, "learning_rate": 8.052368705247493e-06, "loss": 0.041, "step": 82600 }, { "grad_norm": 0.7115673422813416, "learning_rate": 8.043372742626114e-06, "loss": 0.0475, "step": 82610 }, { "grad_norm": 0.960486114025116, "learning_rate": 8.034381368318422e-06, "loss": 0.0343, "step": 82620 }, { "grad_norm": 0.5451896786689758, "learning_rate": 8.025394583307677e-06, "loss": 0.0305, "step": 82630 }, { "grad_norm": 0.6668903231620789, "learning_rate": 8.016412388576683e-06, "loss": 0.0352, "step": 82640 }, { "grad_norm": 0.5247187614440918, "learning_rate": 8.007434785107699e-06, "loss": 0.0252, "step": 82650 }, { "grad_norm": 0.6588650941848755, "learning_rate": 7.998461773882509e-06, "loss": 0.0337, "step": 82660 }, { "grad_norm": 0.5118951797485352, "learning_rate": 7.989493355882393e-06, "loss": 0.0286, "step": 82670 }, { "grad_norm": 0.41740626096725464, "learning_rate": 7.980529532088137e-06, "loss": 0.0319, "step": 82680 }, { "grad_norm": 0.5344963073730469, "learning_rate": 7.97157030347998e-06, "loss": 0.0308, "step": 82690 }, { "grad_norm": 0.38466009497642517, "learning_rate": 7.96261567103772e-06, "loss": 0.0237, "step": 82700 }, { "grad_norm": 0.6263942718505859, "learning_rate": 7.953665635740593e-06, "loss": 0.0246, "step": 82710 }, { "grad_norm": 0.27666568756103516, "learning_rate": 7.94472019856738e-06, "loss": 0.0262, "step": 82720 }, { "grad_norm": 0.3858231008052826, "learning_rate": 7.935779360496337e-06, "loss": 0.0297, "step": 82730 }, { "grad_norm": 0.5350705981254578, "learning_rate": 7.92684312250523e-06, "loss": 0.0384, "step": 82740 }, { "grad_norm": 0.3322950303554535, "learning_rate": 7.917911485571289e-06, "loss": 0.0268, "step": 82750 }, { "grad_norm": 0.7217255234718323, "learning_rate": 7.90898445067128e-06, "loss": 0.0328, "step": 82760 }, { "grad_norm": 0.6020310521125793, "learning_rate": 7.900062018781462e-06, "loss": 0.0353, "step": 82770 }, { "grad_norm": 0.36141639947891235, "learning_rate": 7.891144190877553e-06, "loss": 0.0262, "step": 82780 }, { "grad_norm": 0.20201896131038666, "learning_rate": 7.882230967934806e-06, "loss": 0.0266, "step": 82790 }, { "grad_norm": 0.3429388403892517, "learning_rate": 7.873322350927959e-06, "loss": 0.0324, "step": 82800 }, { "grad_norm": 0.8686011433601379, "learning_rate": 7.864418340831253e-06, "loss": 0.0276, "step": 82810 }, { "grad_norm": 0.3299062252044678, "learning_rate": 7.855518938618395e-06, "loss": 0.036, "step": 82820 }, { "grad_norm": 0.39842504262924194, "learning_rate": 7.846624145262633e-06, "loss": 0.03, "step": 82830 }, { "grad_norm": 0.22991809248924255, "learning_rate": 7.837733961736672e-06, "loss": 0.032, "step": 82840 }, { "grad_norm": 0.40255334973335266, "learning_rate": 7.828848389012727e-06, "loss": 0.0267, "step": 82850 }, { "grad_norm": 0.29522934556007385, "learning_rate": 7.819967428062524e-06, "loss": 0.0291, "step": 82860 }, { "grad_norm": 0.48412182927131653, "learning_rate": 7.811091079857274e-06, "loss": 0.036, "step": 82870 }, { "grad_norm": 0.4968198537826538, "learning_rate": 7.802219345367667e-06, "loss": 0.0273, "step": 82880 }, { "grad_norm": 0.6770652532577515, "learning_rate": 7.793352225563916e-06, "loss": 0.026, "step": 82890 }, { "grad_norm": 0.5109937787055969, "learning_rate": 7.7844897214157e-06, "loss": 0.0314, "step": 82900 }, { "grad_norm": 0.29536035656929016, "learning_rate": 7.775631833892228e-06, "loss": 0.0403, "step": 82910 }, { "grad_norm": 0.25354984402656555, "learning_rate": 7.76677856396215e-06, "loss": 0.0244, "step": 82920 }, { "grad_norm": 0.46499741077423096, "learning_rate": 7.757929912593697e-06, "loss": 0.0283, "step": 82930 }, { "grad_norm": 0.5825230479240417, "learning_rate": 7.749085880754509e-06, "loss": 0.0328, "step": 82940 }, { "grad_norm": 0.34994229674339294, "learning_rate": 7.740246469411777e-06, "loss": 0.0364, "step": 82950 }, { "grad_norm": 0.3725908100605011, "learning_rate": 7.731411679532142e-06, "loss": 0.0309, "step": 82960 }, { "grad_norm": 0.3737712800502777, "learning_rate": 7.722581512081779e-06, "loss": 0.0395, "step": 82970 }, { "grad_norm": 0.30735284090042114, "learning_rate": 7.713755968026331e-06, "loss": 0.0272, "step": 82980 }, { "grad_norm": 0.8117489814758301, "learning_rate": 7.704935048330969e-06, "loss": 0.0324, "step": 82990 }, { "grad_norm": 0.4757767617702484, "learning_rate": 7.696118753960313e-06, "loss": 0.0295, "step": 83000 }, { "grad_norm": 0.23892559111118317, "learning_rate": 7.687307085878503e-06, "loss": 0.0281, "step": 83010 }, { "grad_norm": 0.2379651665687561, "learning_rate": 7.678500045049186e-06, "loss": 0.0282, "step": 83020 }, { "grad_norm": 1.579521894454956, "learning_rate": 7.669697632435463e-06, "loss": 0.0339, "step": 83030 }, { "grad_norm": 0.36688053607940674, "learning_rate": 7.660899848999963e-06, "loss": 0.0238, "step": 83040 }, { "grad_norm": 0.3272823095321655, "learning_rate": 7.652106695704803e-06, "loss": 0.0311, "step": 83050 }, { "grad_norm": 0.41302624344825745, "learning_rate": 7.643318173511598e-06, "loss": 0.0244, "step": 83060 }, { "grad_norm": 0.5154572129249573, "learning_rate": 7.634534283381423e-06, "loss": 0.0331, "step": 83070 }, { "grad_norm": 0.39453113079071045, "learning_rate": 7.625755026274889e-06, "loss": 0.0303, "step": 83080 }, { "grad_norm": 0.4237464666366577, "learning_rate": 7.616980403152069e-06, "loss": 0.0333, "step": 83090 }, { "grad_norm": 0.4615374803543091, "learning_rate": 7.608210414972561e-06, "loss": 0.0315, "step": 83100 }, { "grad_norm": 0.3921409845352173, "learning_rate": 7.599445062695404e-06, "loss": 0.0369, "step": 83110 }, { "grad_norm": 0.37608757615089417, "learning_rate": 7.590684347279214e-06, "loss": 0.0319, "step": 83120 }, { "grad_norm": 0.312028706073761, "learning_rate": 7.581928269682004e-06, "loss": 0.0263, "step": 83130 }, { "grad_norm": 0.27875956892967224, "learning_rate": 7.573176830861361e-06, "loss": 0.0242, "step": 83140 }, { "grad_norm": 0.37084177136421204, "learning_rate": 7.5644300317743e-06, "loss": 0.0344, "step": 83150 }, { "grad_norm": 0.4045082628726959, "learning_rate": 7.555687873377376e-06, "loss": 0.0365, "step": 83160 }, { "grad_norm": 0.45808619260787964, "learning_rate": 7.546950356626592e-06, "loss": 0.0378, "step": 83170 }, { "grad_norm": 0.6246862411499023, "learning_rate": 7.538217482477516e-06, "loss": 0.0329, "step": 83180 }, { "grad_norm": 0.410194993019104, "learning_rate": 7.529489251885119e-06, "loss": 0.032, "step": 83190 }, { "grad_norm": 0.3061020076274872, "learning_rate": 7.520765665803941e-06, "loss": 0.0278, "step": 83200 }, { "grad_norm": 0.2311195433139801, "learning_rate": 7.512046725187949e-06, "loss": 0.0273, "step": 83210 }, { "grad_norm": 0.581929624080658, "learning_rate": 7.503332430990645e-06, "loss": 0.0259, "step": 83220 }, { "grad_norm": 0.4165242314338684, "learning_rate": 7.494622784165017e-06, "loss": 0.035, "step": 83230 }, { "grad_norm": 0.31450459361076355, "learning_rate": 7.485917785663543e-06, "loss": 0.0376, "step": 83240 }, { "grad_norm": 0.5186268091201782, "learning_rate": 7.477217436438172e-06, "loss": 0.0269, "step": 83250 }, { "grad_norm": 0.4509114623069763, "learning_rate": 7.46852173744037e-06, "loss": 0.0216, "step": 83260 }, { "grad_norm": 0.9062982797622681, "learning_rate": 7.4598306896211004e-06, "loss": 0.0377, "step": 83270 }, { "grad_norm": 0.509322464466095, "learning_rate": 7.451144293930773e-06, "loss": 0.0297, "step": 83280 }, { "grad_norm": 0.17331534624099731, "learning_rate": 7.4424625513193455e-06, "loss": 0.0324, "step": 83290 }, { "grad_norm": 0.45616424083709717, "learning_rate": 7.433785462736209e-06, "loss": 0.0304, "step": 83300 }, { "grad_norm": 0.32183247804641724, "learning_rate": 7.425113029130315e-06, "loss": 0.0286, "step": 83310 }, { "grad_norm": 0.3846820890903473, "learning_rate": 7.416445251450044e-06, "loss": 0.0297, "step": 83320 }, { "grad_norm": 0.3264414966106415, "learning_rate": 7.4077821306433035e-06, "loss": 0.036, "step": 83330 }, { "grad_norm": 0.33217114210128784, "learning_rate": 7.399123667657465e-06, "loss": 0.0245, "step": 83340 }, { "grad_norm": 0.5244583487510681, "learning_rate": 7.390469863439426e-06, "loss": 0.0229, "step": 83350 }, { "grad_norm": 1.0625144243240356, "learning_rate": 7.381820718935512e-06, "loss": 0.0322, "step": 83360 }, { "grad_norm": 0.37739095091819763, "learning_rate": 7.373176235091633e-06, "loss": 0.0364, "step": 83370 }, { "grad_norm": 0.28827765583992004, "learning_rate": 7.364536412853101e-06, "loss": 0.0273, "step": 83380 }, { "grad_norm": 1.8219462633132935, "learning_rate": 7.355901253164776e-06, "loss": 0.0307, "step": 83390 }, { "grad_norm": 0.3196762204170227, "learning_rate": 7.34727075697097e-06, "loss": 0.025, "step": 83400 }, { "grad_norm": 0.30658668279647827, "learning_rate": 7.3386449252155095e-06, "loss": 0.0309, "step": 83410 }, { "grad_norm": 1.2798972129821777, "learning_rate": 7.330023758841692e-06, "loss": 0.0343, "step": 83420 }, { "grad_norm": 0.6018044352531433, "learning_rate": 7.321407258792323e-06, "loss": 0.0341, "step": 83430 }, { "grad_norm": 0.4600321650505066, "learning_rate": 7.312795426009694e-06, "loss": 0.029, "step": 83440 }, { "grad_norm": 0.4696866571903229, "learning_rate": 7.304188261435585e-06, "loss": 0.0279, "step": 83450 }, { "grad_norm": 0.5278300046920776, "learning_rate": 7.295585766011242e-06, "loss": 0.0286, "step": 83460 }, { "grad_norm": 0.6078324317932129, "learning_rate": 7.286987940677436e-06, "loss": 0.039, "step": 83470 }, { "grad_norm": 0.3307481110095978, "learning_rate": 7.278394786374426e-06, "loss": 0.0248, "step": 83480 }, { "grad_norm": 0.24755994975566864, "learning_rate": 7.269806304041915e-06, "loss": 0.027, "step": 83490 }, { "grad_norm": 0.6274604201316833, "learning_rate": 7.261222494619147e-06, "loss": 0.0266, "step": 83500 }, { "grad_norm": 0.499528169631958, "learning_rate": 7.252643359044825e-06, "loss": 0.0348, "step": 83510 }, { "grad_norm": 0.4749901294708252, "learning_rate": 7.244068898257172e-06, "loss": 0.022, "step": 83520 }, { "grad_norm": 0.18037036061286926, "learning_rate": 7.2354991131938495e-06, "loss": 0.0249, "step": 83530 }, { "grad_norm": 0.34815362095832825, "learning_rate": 7.2269340047920574e-06, "loss": 0.0196, "step": 83540 }, { "grad_norm": 0.7553690671920776, "learning_rate": 7.218373573988435e-06, "loss": 0.0318, "step": 83550 }, { "grad_norm": 0.42770975828170776, "learning_rate": 7.209817821719178e-06, "loss": 0.0409, "step": 83560 }, { "grad_norm": 0.19294506311416626, "learning_rate": 7.201266748919894e-06, "loss": 0.0327, "step": 83570 }, { "grad_norm": 0.39874207973480225, "learning_rate": 7.19272035652575e-06, "loss": 0.0277, "step": 83580 }, { "grad_norm": 0.39338862895965576, "learning_rate": 7.184178645471334e-06, "loss": 0.0301, "step": 83590 }, { "grad_norm": 0.7256516218185425, "learning_rate": 7.175641616690775e-06, "loss": 0.0343, "step": 83600 }, { "grad_norm": 0.49605637788772583, "learning_rate": 7.167109271117661e-06, "loss": 0.0333, "step": 83610 }, { "grad_norm": 0.6632750034332275, "learning_rate": 7.158581609685072e-06, "loss": 0.0374, "step": 83620 }, { "grad_norm": 0.2033250480890274, "learning_rate": 7.150058633325591e-06, "loss": 0.0272, "step": 83630 }, { "grad_norm": 0.6788641214370728, "learning_rate": 7.141540342971287e-06, "loss": 0.0376, "step": 83640 }, { "grad_norm": 0.29740744829177856, "learning_rate": 7.133026739553677e-06, "loss": 0.0247, "step": 83650 }, { "grad_norm": 0.3096070885658264, "learning_rate": 7.124517824003829e-06, "loss": 0.0368, "step": 83660 }, { "grad_norm": 0.6000601649284363, "learning_rate": 7.116013597252236e-06, "loss": 0.0259, "step": 83670 }, { "grad_norm": 0.39306220412254333, "learning_rate": 7.107514060228921e-06, "loss": 0.0311, "step": 83680 }, { "grad_norm": 0.3219170868396759, "learning_rate": 7.099019213863378e-06, "loss": 0.0295, "step": 83690 }, { "grad_norm": 0.4654454290866852, "learning_rate": 7.090529059084611e-06, "loss": 0.0208, "step": 83700 }, { "grad_norm": 0.2788471579551697, "learning_rate": 7.082043596821058e-06, "loss": 0.0237, "step": 83710 }, { "grad_norm": 0.41645607352256775, "learning_rate": 7.07356282800069e-06, "loss": 0.0334, "step": 83720 }, { "grad_norm": 0.5280946493148804, "learning_rate": 7.0650867535509645e-06, "loss": 0.0266, "step": 83730 }, { "grad_norm": 0.4483402371406555, "learning_rate": 7.0566153743987895e-06, "loss": 0.0252, "step": 83740 }, { "grad_norm": 0.2800077497959137, "learning_rate": 7.04814869147059e-06, "loss": 0.0368, "step": 83750 }, { "grad_norm": 0.42863547801971436, "learning_rate": 7.0396867056922725e-06, "loss": 0.0353, "step": 83760 }, { "grad_norm": 0.31240782141685486, "learning_rate": 7.031229417989244e-06, "loss": 0.0311, "step": 83770 }, { "grad_norm": 0.4412001967430115, "learning_rate": 7.022776829286348e-06, "loss": 0.0393, "step": 83780 }, { "grad_norm": 0.5562940239906311, "learning_rate": 7.014328940507975e-06, "loss": 0.0328, "step": 83790 }, { "grad_norm": 0.4255578815937042, "learning_rate": 7.005885752577945e-06, "loss": 0.0343, "step": 83800 }, { "grad_norm": 0.544852614402771, "learning_rate": 6.997447266419615e-06, "loss": 0.0372, "step": 83810 }, { "grad_norm": 0.41346266865730286, "learning_rate": 6.989013482955797e-06, "loss": 0.0382, "step": 83820 }, { "grad_norm": 0.6417983770370483, "learning_rate": 6.980584403108808e-06, "loss": 0.0294, "step": 83830 }, { "grad_norm": 0.5178214311599731, "learning_rate": 6.972160027800417e-06, "loss": 0.0285, "step": 83840 }, { "grad_norm": 0.4145227074623108, "learning_rate": 6.963740357951931e-06, "loss": 0.033, "step": 83850 }, { "grad_norm": 0.4144970774650574, "learning_rate": 6.95532539448408e-06, "loss": 0.0358, "step": 83860 }, { "grad_norm": 0.3328557014465332, "learning_rate": 6.946915138317128e-06, "loss": 0.0305, "step": 83870 }, { "grad_norm": 0.21290238201618195, "learning_rate": 6.9385095903708065e-06, "loss": 0.0361, "step": 83880 }, { "grad_norm": 0.48243746161460876, "learning_rate": 6.930108751564346e-06, "loss": 0.0362, "step": 83890 }, { "grad_norm": 0.35167545080184937, "learning_rate": 6.921712622816428e-06, "loss": 0.0304, "step": 83900 }, { "grad_norm": 0.4442530870437622, "learning_rate": 6.913321205045259e-06, "loss": 0.0351, "step": 83910 }, { "grad_norm": 0.5587146878242493, "learning_rate": 6.904934499168491e-06, "loss": 0.0225, "step": 83920 }, { "grad_norm": 0.2905813455581665, "learning_rate": 6.896552506103299e-06, "loss": 0.0331, "step": 83930 }, { "grad_norm": 0.47458475828170776, "learning_rate": 6.888175226766313e-06, "loss": 0.0429, "step": 83940 }, { "grad_norm": 0.39572644233703613, "learning_rate": 6.879802662073686e-06, "loss": 0.0318, "step": 83950 }, { "grad_norm": 0.3149591386318207, "learning_rate": 6.871434812940991e-06, "loss": 0.0387, "step": 83960 }, { "grad_norm": 0.3260420262813568, "learning_rate": 6.863071680283345e-06, "loss": 0.0225, "step": 83970 }, { "grad_norm": 0.36898791790008545, "learning_rate": 6.854713265015339e-06, "loss": 0.0449, "step": 83980 }, { "grad_norm": 0.8407918810844421, "learning_rate": 6.8463595680510025e-06, "loss": 0.0388, "step": 83990 }, { "grad_norm": 0.49485865235328674, "learning_rate": 6.838010590303911e-06, "loss": 0.0318, "step": 84000 }, { "grad_norm": 0.6718322038650513, "learning_rate": 6.829666332687085e-06, "loss": 0.0256, "step": 84010 }, { "grad_norm": 0.5350461602210999, "learning_rate": 6.821326796113059e-06, "loss": 0.0266, "step": 84020 }, { "grad_norm": 0.3414265811443329, "learning_rate": 6.812991981493799e-06, "loss": 0.0278, "step": 84030 }, { "grad_norm": 0.5584354400634766, "learning_rate": 6.804661889740821e-06, "loss": 0.037, "step": 84040 }, { "grad_norm": 0.7040855884552002, "learning_rate": 6.7963365217650625e-06, "loss": 0.0357, "step": 84050 }, { "grad_norm": 0.3221740126609802, "learning_rate": 6.788015878476983e-06, "loss": 0.031, "step": 84060 }, { "grad_norm": 0.5445693731307983, "learning_rate": 6.779699960786523e-06, "loss": 0.038, "step": 84070 }, { "grad_norm": 1.0180621147155762, "learning_rate": 6.771388769603104e-06, "loss": 0.0316, "step": 84080 }, { "grad_norm": 0.4610811769962311, "learning_rate": 6.763082305835605e-06, "loss": 0.0461, "step": 84090 }, { "grad_norm": 0.3462503254413605, "learning_rate": 6.754780570392433e-06, "loss": 0.0272, "step": 84100 }, { "grad_norm": 0.5002638697624207, "learning_rate": 6.7464835641814215e-06, "loss": 0.0359, "step": 84110 }, { "grad_norm": 0.1800455003976822, "learning_rate": 6.7381912881099515e-06, "loss": 0.0232, "step": 84120 }, { "grad_norm": 0.587763786315918, "learning_rate": 6.72990374308482e-06, "loss": 0.0376, "step": 84130 }, { "grad_norm": 0.4426119029521942, "learning_rate": 6.721620930012379e-06, "loss": 0.0349, "step": 84140 }, { "grad_norm": 0.42284223437309265, "learning_rate": 6.713342849798393e-06, "loss": 0.0275, "step": 84150 }, { "grad_norm": 0.5936816334724426, "learning_rate": 6.705069503348166e-06, "loss": 0.0314, "step": 84160 }, { "grad_norm": 0.7095282077789307, "learning_rate": 6.696800891566435e-06, "loss": 0.0368, "step": 84170 }, { "grad_norm": 0.3408166170120239, "learning_rate": 6.688537015357449e-06, "loss": 0.0286, "step": 84180 }, { "grad_norm": 0.44472822546958923, "learning_rate": 6.680277875624941e-06, "loss": 0.0277, "step": 84190 }, { "grad_norm": 0.41806039214134216, "learning_rate": 6.672023473272121e-06, "loss": 0.0335, "step": 84200 }, { "grad_norm": 0.46339187026023865, "learning_rate": 6.663773809201667e-06, "loss": 0.0394, "step": 84210 }, { "grad_norm": 0.36739206314086914, "learning_rate": 6.655528884315748e-06, "loss": 0.0266, "step": 84220 }, { "grad_norm": 0.4570995569229126, "learning_rate": 6.647288699516036e-06, "loss": 0.0397, "step": 84230 }, { "grad_norm": 0.661577045917511, "learning_rate": 6.639053255703642e-06, "loss": 0.0351, "step": 84240 }, { "grad_norm": 0.5012019872665405, "learning_rate": 6.6308225537791925e-06, "loss": 0.0352, "step": 84250 }, { "grad_norm": 0.293447345495224, "learning_rate": 6.622596594642777e-06, "loss": 0.028, "step": 84260 }, { "grad_norm": 0.9558806419372559, "learning_rate": 6.614375379193999e-06, "loss": 0.0263, "step": 84270 }, { "grad_norm": 1.8908326625823975, "learning_rate": 6.606158908331889e-06, "loss": 0.0275, "step": 84280 }, { "grad_norm": 0.3022083640098572, "learning_rate": 6.597947182955005e-06, "loss": 0.0375, "step": 84290 }, { "grad_norm": 0.3998284935951233, "learning_rate": 6.589740203961359e-06, "loss": 0.0369, "step": 84300 }, { "grad_norm": 1.079028844833374, "learning_rate": 6.581537972248464e-06, "loss": 0.0419, "step": 84310 }, { "grad_norm": 0.7784219980239868, "learning_rate": 6.573340488713275e-06, "loss": 0.0361, "step": 84320 }, { "grad_norm": 0.6832865476608276, "learning_rate": 6.565147754252304e-06, "loss": 0.0372, "step": 84330 }, { "grad_norm": 0.3519715964794159, "learning_rate": 6.556959769761462e-06, "loss": 0.0322, "step": 84340 }, { "grad_norm": 1.0919626951217651, "learning_rate": 6.548776536136192e-06, "loss": 0.0301, "step": 84350 }, { "grad_norm": 0.6583736538887024, "learning_rate": 6.54059805427138e-06, "loss": 0.0294, "step": 84360 }, { "grad_norm": 0.37737026810646057, "learning_rate": 6.532424325061437e-06, "loss": 0.0463, "step": 84370 }, { "grad_norm": 0.5265069603919983, "learning_rate": 6.5242553494001935e-06, "loss": 0.0274, "step": 84380 }, { "grad_norm": 0.4373796582221985, "learning_rate": 6.516091128181046e-06, "loss": 0.0266, "step": 84390 }, { "grad_norm": 0.43105053901672363, "learning_rate": 6.507931662296785e-06, "loss": 0.0312, "step": 84400 }, { "grad_norm": 0.49014613032341003, "learning_rate": 6.4997769526397354e-06, "loss": 0.0313, "step": 84410 }, { "grad_norm": 0.33954644203186035, "learning_rate": 6.491627000101663e-06, "loss": 0.0285, "step": 84420 }, { "grad_norm": 0.5127135515213013, "learning_rate": 6.4834818055738545e-06, "loss": 0.0302, "step": 84430 }, { "grad_norm": 0.2998824119567871, "learning_rate": 6.4753413699470465e-06, "loss": 0.0232, "step": 84440 }, { "grad_norm": 0.6263128519058228, "learning_rate": 6.467205694111478e-06, "loss": 0.0286, "step": 84450 }, { "grad_norm": 0.27965816855430603, "learning_rate": 6.459074778956836e-06, "loss": 0.0309, "step": 84460 }, { "grad_norm": 0.2863745391368866, "learning_rate": 6.450948625372311e-06, "loss": 0.0266, "step": 84470 }, { "grad_norm": 0.34820276498794556, "learning_rate": 6.442827234246585e-06, "loss": 0.0215, "step": 84480 }, { "grad_norm": 0.3554556667804718, "learning_rate": 6.43471060646777e-06, "loss": 0.0333, "step": 84490 }, { "grad_norm": 0.15502749383449554, "learning_rate": 6.426598742923512e-06, "loss": 0.0267, "step": 84500 }, { "grad_norm": 0.6190333962440491, "learning_rate": 6.418491644500885e-06, "loss": 0.0358, "step": 84510 }, { "grad_norm": 0.34799909591674805, "learning_rate": 6.410389312086512e-06, "loss": 0.0325, "step": 84520 }, { "grad_norm": 0.35590478777885437, "learning_rate": 6.4022917465664135e-06, "loss": 0.0347, "step": 84530 }, { "grad_norm": 0.8902786374092102, "learning_rate": 6.3941989488261575e-06, "loss": 0.0388, "step": 84540 }, { "grad_norm": 0.2730090320110321, "learning_rate": 6.386110919750726e-06, "loss": 0.0306, "step": 84550 }, { "grad_norm": 0.3086160123348236, "learning_rate": 6.378027660224651e-06, "loss": 0.0246, "step": 84560 }, { "grad_norm": 0.6524681448936462, "learning_rate": 6.369949171131856e-06, "loss": 0.0268, "step": 84570 }, { "grad_norm": 0.3517143726348877, "learning_rate": 6.361875453355848e-06, "loss": 0.0246, "step": 84580 }, { "grad_norm": 0.7647067308425903, "learning_rate": 6.353806507779525e-06, "loss": 0.031, "step": 84590 }, { "grad_norm": 0.3516765534877777, "learning_rate": 6.3457423352853044e-06, "loss": 0.0335, "step": 84600 }, { "grad_norm": 0.7427707314491272, "learning_rate": 6.337682936755062e-06, "loss": 0.0302, "step": 84610 }, { "grad_norm": 0.3689185380935669, "learning_rate": 6.329628313070179e-06, "loss": 0.0301, "step": 84620 }, { "grad_norm": 0.31845739483833313, "learning_rate": 6.321578465111477e-06, "loss": 0.0317, "step": 84630 }, { "grad_norm": 0.3010735809803009, "learning_rate": 6.31353339375928e-06, "loss": 0.0282, "step": 84640 }, { "grad_norm": 0.3418896496295929, "learning_rate": 6.305493099893395e-06, "loss": 0.0196, "step": 84650 }, { "grad_norm": 1.9664562940597534, "learning_rate": 6.297457584393096e-06, "loss": 0.0303, "step": 84660 }, { "grad_norm": 0.3428502678871155, "learning_rate": 6.289426848137126e-06, "loss": 0.0295, "step": 84670 }, { "grad_norm": 0.44518017768859863, "learning_rate": 6.281400892003713e-06, "loss": 0.0251, "step": 84680 }, { "grad_norm": 0.33818376064300537, "learning_rate": 6.273379716870581e-06, "loss": 0.0287, "step": 84690 }, { "grad_norm": 0.37434545159339905, "learning_rate": 6.265363323614887e-06, "loss": 0.029, "step": 84700 }, { "grad_norm": 0.7088225483894348, "learning_rate": 6.257351713113308e-06, "loss": 0.0259, "step": 84710 }, { "grad_norm": 0.6560187935829163, "learning_rate": 6.249344886241981e-06, "loss": 0.0362, "step": 84720 }, { "grad_norm": 0.3547614514827728, "learning_rate": 6.2413428438765295e-06, "loss": 0.0245, "step": 84730 }, { "grad_norm": 0.20611074566841125, "learning_rate": 6.233345586892026e-06, "loss": 0.0269, "step": 84740 }, { "grad_norm": 0.23639386892318726, "learning_rate": 6.225353116163052e-06, "loss": 0.0219, "step": 84750 }, { "grad_norm": 0.358268141746521, "learning_rate": 6.217365432563627e-06, "loss": 0.0279, "step": 84760 }, { "grad_norm": 0.36054590344429016, "learning_rate": 6.209382536967318e-06, "loss": 0.027, "step": 84770 }, { "grad_norm": 1.1159591674804688, "learning_rate": 6.201404430247082e-06, "loss": 0.0361, "step": 84780 }, { "grad_norm": 0.3273460566997528, "learning_rate": 6.1934311132754165e-06, "loss": 0.0239, "step": 84790 }, { "grad_norm": 0.8925826549530029, "learning_rate": 6.185462586924257e-06, "loss": 0.0364, "step": 84800 }, { "grad_norm": 0.5006183385848999, "learning_rate": 6.177498852065039e-06, "loss": 0.0326, "step": 84810 }, { "grad_norm": 0.27618592977523804, "learning_rate": 6.169539909568655e-06, "loss": 0.0295, "step": 84820 }, { "grad_norm": 0.33159080147743225, "learning_rate": 6.161585760305488e-06, "loss": 0.0364, "step": 84830 }, { "grad_norm": 0.5462019443511963, "learning_rate": 6.153636405145391e-06, "loss": 0.0337, "step": 84840 }, { "grad_norm": 1.0410159826278687, "learning_rate": 6.145691844957702e-06, "loss": 0.0313, "step": 84850 }, { "grad_norm": 0.2780713438987732, "learning_rate": 6.1377520806112145e-06, "loss": 0.0276, "step": 84860 }, { "grad_norm": 0.3567277193069458, "learning_rate": 6.129817112974218e-06, "loss": 0.0256, "step": 84870 }, { "grad_norm": 0.40853506326675415, "learning_rate": 6.121886942914457e-06, "loss": 0.0234, "step": 84880 }, { "grad_norm": 0.6173173785209656, "learning_rate": 6.11396157129917e-06, "loss": 0.0297, "step": 84890 }, { "grad_norm": 0.4000313878059387, "learning_rate": 6.106040998995066e-06, "loss": 0.0337, "step": 84900 }, { "grad_norm": 0.2561114728450775, "learning_rate": 6.098125226868329e-06, "loss": 0.0313, "step": 84910 }, { "grad_norm": 0.3328644931316376, "learning_rate": 6.090214255784604e-06, "loss": 0.0301, "step": 84920 }, { "grad_norm": 0.46801280975341797, "learning_rate": 6.0823080866090346e-06, "loss": 0.0348, "step": 84930 }, { "grad_norm": 0.4094822108745575, "learning_rate": 6.074406720206233e-06, "loss": 0.0247, "step": 84940 }, { "grad_norm": 0.530306875705719, "learning_rate": 6.066510157440259e-06, "loss": 0.0258, "step": 84950 }, { "grad_norm": 0.2463712841272354, "learning_rate": 6.058618399174676e-06, "loss": 0.0285, "step": 84960 }, { "grad_norm": 0.42841747403144836, "learning_rate": 6.050731446272528e-06, "loss": 0.0367, "step": 84970 }, { "grad_norm": 0.9311162233352661, "learning_rate": 6.042849299596321e-06, "loss": 0.0333, "step": 84980 }, { "grad_norm": 0.4201088547706604, "learning_rate": 6.034971960008013e-06, "loss": 0.0316, "step": 84990 }, { "grad_norm": 0.4974554777145386, "learning_rate": 6.027099428369082e-06, "loss": 0.0289, "step": 85000 }, { "grad_norm": 0.6234527826309204, "learning_rate": 6.019231705540435e-06, "loss": 0.0327, "step": 85010 }, { "grad_norm": 0.42346495389938354, "learning_rate": 6.011368792382488e-06, "loss": 0.0272, "step": 85020 }, { "grad_norm": 0.251156210899353, "learning_rate": 6.00351068975511e-06, "loss": 0.0282, "step": 85030 }, { "grad_norm": 0.7389996647834778, "learning_rate": 5.9956573985176674e-06, "loss": 0.0343, "step": 85040 }, { "grad_norm": 0.33437424898147583, "learning_rate": 5.987808919528959e-06, "loss": 0.0302, "step": 85050 }, { "grad_norm": 0.43196171522140503, "learning_rate": 5.979965253647307e-06, "loss": 0.0398, "step": 85060 }, { "grad_norm": 0.5914658904075623, "learning_rate": 5.9721264017304655e-06, "loss": 0.0292, "step": 85070 }, { "grad_norm": 0.33890676498413086, "learning_rate": 5.964292364635682e-06, "loss": 0.0353, "step": 85080 }, { "grad_norm": 0.30634915828704834, "learning_rate": 5.956463143219682e-06, "loss": 0.0327, "step": 85090 }, { "grad_norm": 0.3477575480937958, "learning_rate": 5.948638738338663e-06, "loss": 0.0307, "step": 85100 }, { "grad_norm": 0.6648527979850769, "learning_rate": 5.940819150848271e-06, "loss": 0.033, "step": 85110 }, { "grad_norm": 0.2010272741317749, "learning_rate": 5.933004381603663e-06, "loss": 0.0298, "step": 85120 }, { "grad_norm": 0.3729061186313629, "learning_rate": 5.92519443145943e-06, "loss": 0.0288, "step": 85130 }, { "grad_norm": 0.42195215821266174, "learning_rate": 5.917389301269671e-06, "loss": 0.0262, "step": 85140 }, { "grad_norm": 0.3758542835712433, "learning_rate": 5.9095889918879435e-06, "loss": 0.024, "step": 85150 }, { "grad_norm": 0.30271005630493164, "learning_rate": 5.901793504167286e-06, "loss": 0.0214, "step": 85160 }, { "grad_norm": 0.49549415707588196, "learning_rate": 5.894002838960172e-06, "loss": 0.0296, "step": 85170 }, { "grad_norm": 0.47373396158218384, "learning_rate": 5.886216997118604e-06, "loss": 0.0254, "step": 85180 }, { "grad_norm": 0.4225498139858246, "learning_rate": 5.878435979494023e-06, "loss": 0.0317, "step": 85190 }, { "grad_norm": 0.442721426486969, "learning_rate": 5.870659786937344e-06, "loss": 0.0328, "step": 85200 }, { "grad_norm": 0.5064247250556946, "learning_rate": 5.862888420298962e-06, "loss": 0.0298, "step": 85210 }, { "grad_norm": 0.5041894912719727, "learning_rate": 5.855121880428743e-06, "loss": 0.0244, "step": 85220 }, { "grad_norm": 0.5472580790519714, "learning_rate": 5.84736016817603e-06, "loss": 0.0296, "step": 85230 }, { "grad_norm": 0.3000398874282837, "learning_rate": 5.839603284389622e-06, "loss": 0.0252, "step": 85240 }, { "grad_norm": 0.39174699783325195, "learning_rate": 5.831851229917812e-06, "loss": 0.027, "step": 85250 }, { "grad_norm": 0.3420279026031494, "learning_rate": 5.824104005608338e-06, "loss": 0.0274, "step": 85260 }, { "grad_norm": 0.2931084930896759, "learning_rate": 5.81636161230843e-06, "loss": 0.032, "step": 85270 }, { "grad_norm": 0.4204750657081604, "learning_rate": 5.808624050864791e-06, "loss": 0.0333, "step": 85280 }, { "grad_norm": 0.303489089012146, "learning_rate": 5.800891322123592e-06, "loss": 0.0283, "step": 85290 }, { "grad_norm": 0.20023970305919647, "learning_rate": 5.793163426930454e-06, "loss": 0.0243, "step": 85300 }, { "grad_norm": 0.3297566771507263, "learning_rate": 5.785440366130512e-06, "loss": 0.0253, "step": 85310 }, { "grad_norm": 0.2313379943370819, "learning_rate": 5.777722140568326e-06, "loss": 0.0319, "step": 85320 }, { "grad_norm": 0.4756508767604828, "learning_rate": 5.7700087510879685e-06, "loss": 0.0359, "step": 85330 }, { "grad_norm": 0.5906140208244324, "learning_rate": 5.762300198532933e-06, "loss": 0.0262, "step": 85340 }, { "grad_norm": 0.24643288552761078, "learning_rate": 5.7545964837462606e-06, "loss": 0.0256, "step": 85350 }, { "grad_norm": 0.535200834274292, "learning_rate": 5.746897607570384e-06, "loss": 0.0288, "step": 85360 }, { "grad_norm": 0.3181234300136566, "learning_rate": 5.7392035708472626e-06, "loss": 0.0266, "step": 85370 }, { "grad_norm": 0.978473961353302, "learning_rate": 5.731514374418278e-06, "loss": 0.0306, "step": 85380 }, { "grad_norm": 0.34840530157089233, "learning_rate": 5.7238300191243325e-06, "loss": 0.0318, "step": 85390 }, { "grad_norm": 0.6852123141288757, "learning_rate": 5.716150505805762e-06, "loss": 0.0309, "step": 85400 }, { "grad_norm": 0.8428255915641785, "learning_rate": 5.708475835302407e-06, "loss": 0.0261, "step": 85410 }, { "grad_norm": 0.27314427495002747, "learning_rate": 5.70080600845353e-06, "loss": 0.0366, "step": 85420 }, { "grad_norm": 0.6407908797264099, "learning_rate": 5.693141026097909e-06, "loss": 0.0268, "step": 85430 }, { "grad_norm": 0.7968100309371948, "learning_rate": 5.6854808890737745e-06, "loss": 0.035, "step": 85440 }, { "grad_norm": 0.3345113694667816, "learning_rate": 5.6778255982188225e-06, "loss": 0.0323, "step": 85450 }, { "grad_norm": 0.915276288986206, "learning_rate": 5.670175154370217e-06, "loss": 0.034, "step": 85460 }, { "grad_norm": 0.3364563286304474, "learning_rate": 5.662529558364616e-06, "loss": 0.0339, "step": 85470 }, { "grad_norm": 0.5979358553886414, "learning_rate": 5.654888811038128e-06, "loss": 0.0277, "step": 85480 }, { "grad_norm": 0.5371049642562866, "learning_rate": 5.647252913226314e-06, "loss": 0.0245, "step": 85490 }, { "grad_norm": 0.42464011907577515, "learning_rate": 5.639621865764255e-06, "loss": 0.0379, "step": 85500 }, { "grad_norm": 0.5007175803184509, "learning_rate": 5.63199566948644e-06, "loss": 0.0306, "step": 85510 }, { "grad_norm": 0.3987308740615845, "learning_rate": 5.6243743252268835e-06, "loss": 0.0307, "step": 85520 }, { "grad_norm": 0.5129225850105286, "learning_rate": 5.616757833819009e-06, "loss": 0.0323, "step": 85530 }, { "grad_norm": 0.3136216998100281, "learning_rate": 5.609146196095794e-06, "loss": 0.0256, "step": 85540 }, { "grad_norm": 0.38490381836891174, "learning_rate": 5.6015394128896005e-06, "loss": 0.0241, "step": 85550 }, { "grad_norm": 0.30706676840782166, "learning_rate": 5.593937485032308e-06, "loss": 0.0227, "step": 85560 }, { "grad_norm": 0.28183895349502563, "learning_rate": 5.586340413355246e-06, "loss": 0.0301, "step": 85570 }, { "grad_norm": 0.28188037872314453, "learning_rate": 5.578748198689227e-06, "loss": 0.0301, "step": 85580 }, { "grad_norm": 0.3941172659397125, "learning_rate": 5.571160841864504e-06, "loss": 0.0278, "step": 85590 }, { "grad_norm": 0.3859092891216278, "learning_rate": 5.563578343710846e-06, "loss": 0.0251, "step": 85600 }, { "grad_norm": 0.40134164690971375, "learning_rate": 5.556000705057446e-06, "loss": 0.0257, "step": 85610 }, { "grad_norm": 0.26479020714759827, "learning_rate": 5.548427926732996e-06, "loss": 0.0267, "step": 85620 }, { "grad_norm": 0.9633787870407104, "learning_rate": 5.540860009565629e-06, "loss": 0.0237, "step": 85630 }, { "grad_norm": 0.38731735944747925, "learning_rate": 5.533296954382966e-06, "loss": 0.033, "step": 85640 }, { "grad_norm": 0.25786668062210083, "learning_rate": 5.525738762012095e-06, "loss": 0.0279, "step": 85650 }, { "grad_norm": 1.0641834735870361, "learning_rate": 5.518185433279577e-06, "loss": 0.0359, "step": 85660 }, { "grad_norm": 0.2896344065666199, "learning_rate": 5.510636969011407e-06, "loss": 0.0346, "step": 85670 }, { "grad_norm": 0.24508295953273773, "learning_rate": 5.503093370033096e-06, "loss": 0.0289, "step": 85680 }, { "grad_norm": 0.8395662307739258, "learning_rate": 5.495554637169603e-06, "loss": 0.0328, "step": 85690 }, { "grad_norm": 0.9626721739768982, "learning_rate": 5.488020771245328e-06, "loss": 0.0384, "step": 85700 }, { "grad_norm": 0.35320237278938293, "learning_rate": 5.480491773084195e-06, "loss": 0.0327, "step": 85710 }, { "grad_norm": 0.2788803279399872, "learning_rate": 5.472967643509525e-06, "loss": 0.0297, "step": 85720 }, { "grad_norm": 0.29112568497657776, "learning_rate": 5.46544838334419e-06, "loss": 0.0262, "step": 85730 }, { "grad_norm": 0.3023100197315216, "learning_rate": 5.4579339934104525e-06, "loss": 0.0271, "step": 85740 }, { "grad_norm": 0.4056192636489868, "learning_rate": 5.450424474530091e-06, "loss": 0.0477, "step": 85750 }, { "grad_norm": 0.4540095925331116, "learning_rate": 5.442919827524323e-06, "loss": 0.0269, "step": 85760 }, { "grad_norm": 0.33257925510406494, "learning_rate": 5.435420053213863e-06, "loss": 0.0271, "step": 85770 }, { "grad_norm": 0.4061655104160309, "learning_rate": 5.427925152418845e-06, "loss": 0.0401, "step": 85780 }, { "grad_norm": 0.5541518926620483, "learning_rate": 5.420435125958934e-06, "loss": 0.033, "step": 85790 }, { "grad_norm": 0.2936902642250061, "learning_rate": 5.412949974653209e-06, "loss": 0.0253, "step": 85800 }, { "grad_norm": 0.4451318085193634, "learning_rate": 5.405469699320248e-06, "loss": 0.0271, "step": 85810 }, { "grad_norm": 0.3942384421825409, "learning_rate": 5.397994300778064e-06, "loss": 0.0224, "step": 85820 }, { "grad_norm": 0.7150826454162598, "learning_rate": 5.39052377984417e-06, "loss": 0.0323, "step": 85830 }, { "grad_norm": 0.31487351655960083, "learning_rate": 5.383058137335523e-06, "loss": 0.0302, "step": 85840 }, { "grad_norm": 0.29997915029525757, "learning_rate": 5.375597374068553e-06, "loss": 0.035, "step": 85850 }, { "grad_norm": 0.41517651081085205, "learning_rate": 5.368141490859163e-06, "loss": 0.026, "step": 85860 }, { "grad_norm": 0.3985568881034851, "learning_rate": 5.360690488522729e-06, "loss": 0.0341, "step": 85870 }, { "grad_norm": 0.5219780802726746, "learning_rate": 5.353244367874055e-06, "loss": 0.0316, "step": 85880 }, { "grad_norm": 0.4561072885990143, "learning_rate": 5.345803129727456e-06, "loss": 0.0293, "step": 85890 }, { "grad_norm": 0.24186639487743378, "learning_rate": 5.338366774896697e-06, "loss": 0.0169, "step": 85900 }, { "grad_norm": 0.5183665752410889, "learning_rate": 5.330935304194995e-06, "loss": 0.0313, "step": 85910 }, { "grad_norm": 0.44249215722084045, "learning_rate": 5.323508718435044e-06, "loss": 0.0314, "step": 85920 }, { "grad_norm": 0.28404566645622253, "learning_rate": 5.316087018429011e-06, "loss": 0.0257, "step": 85930 }, { "grad_norm": 0.2561192214488983, "learning_rate": 5.30867020498853e-06, "loss": 0.0218, "step": 85940 }, { "grad_norm": 0.41778749227523804, "learning_rate": 5.30125827892467e-06, "loss": 0.0295, "step": 85950 }, { "grad_norm": 0.5967745184898376, "learning_rate": 5.293851241048015e-06, "loss": 0.0416, "step": 85960 }, { "grad_norm": 0.400568425655365, "learning_rate": 5.286449092168555e-06, "loss": 0.0333, "step": 85970 }, { "grad_norm": 0.6228575706481934, "learning_rate": 5.2790518330958105e-06, "loss": 0.0365, "step": 85980 }, { "grad_norm": 0.31789737939834595, "learning_rate": 5.271659464638717e-06, "loss": 0.026, "step": 85990 }, { "grad_norm": 0.5234777331352234, "learning_rate": 5.264271987605701e-06, "loss": 0.0296, "step": 86000 }, { "grad_norm": 0.24169118702411652, "learning_rate": 5.256889402804632e-06, "loss": 0.0276, "step": 86010 }, { "grad_norm": 0.4305945932865143, "learning_rate": 5.249511711042881e-06, "loss": 0.0295, "step": 86020 }, { "grad_norm": 0.3628580868244171, "learning_rate": 5.242138913127237e-06, "loss": 0.0239, "step": 86030 }, { "grad_norm": 0.5948741436004639, "learning_rate": 5.234771009863987e-06, "loss": 0.0309, "step": 86040 }, { "grad_norm": 0.50229811668396, "learning_rate": 5.227408002058881e-06, "loss": 0.0359, "step": 86050 }, { "grad_norm": 0.5673587322235107, "learning_rate": 5.220049890517126e-06, "loss": 0.0364, "step": 86060 }, { "grad_norm": 0.4380342960357666, "learning_rate": 5.212696676043377e-06, "loss": 0.0251, "step": 86070 }, { "grad_norm": 0.49056753516197205, "learning_rate": 5.205348359441798e-06, "loss": 0.0316, "step": 86080 }, { "grad_norm": 0.41355210542678833, "learning_rate": 5.198004941515966e-06, "loss": 0.0337, "step": 86090 }, { "grad_norm": 0.6053926348686218, "learning_rate": 5.190666423068952e-06, "loss": 0.0284, "step": 86100 }, { "grad_norm": 0.4616785943508148, "learning_rate": 5.183332804903284e-06, "loss": 0.0223, "step": 86110 }, { "grad_norm": 0.33603882789611816, "learning_rate": 5.176004087820979e-06, "loss": 0.0279, "step": 86120 }, { "grad_norm": 0.45438501238822937, "learning_rate": 5.168680272623455e-06, "loss": 0.0323, "step": 86130 }, { "grad_norm": 0.31646445393562317, "learning_rate": 5.161361360111661e-06, "loss": 0.027, "step": 86140 }, { "grad_norm": 0.21559208631515503, "learning_rate": 5.154047351085984e-06, "loss": 0.0204, "step": 86150 }, { "grad_norm": 0.44370603561401367, "learning_rate": 5.1467382463462385e-06, "loss": 0.0325, "step": 86160 }, { "grad_norm": 0.9719023704528809, "learning_rate": 5.139434046691788e-06, "loss": 0.0328, "step": 86170 }, { "grad_norm": 0.44261109828948975, "learning_rate": 5.13213475292137e-06, "loss": 0.0299, "step": 86180 }, { "grad_norm": 0.3758021295070648, "learning_rate": 5.12484036583325e-06, "loss": 0.0224, "step": 86190 }, { "grad_norm": 0.3029994070529938, "learning_rate": 5.117550886225109e-06, "loss": 0.0242, "step": 86200 }, { "grad_norm": 0.706887423992157, "learning_rate": 5.11026631489413e-06, "loss": 0.0285, "step": 86210 }, { "grad_norm": 0.5014220476150513, "learning_rate": 5.102986652636926e-06, "loss": 0.0343, "step": 86220 }, { "grad_norm": 0.35610225796699524, "learning_rate": 5.095711900249606e-06, "loss": 0.025, "step": 86230 }, { "grad_norm": 0.4489193856716156, "learning_rate": 5.088442058527715e-06, "loss": 0.0282, "step": 86240 }, { "grad_norm": 0.31663867831230164, "learning_rate": 5.08117712826629e-06, "loss": 0.0291, "step": 86250 }, { "grad_norm": 0.4825368821620941, "learning_rate": 5.073917110259785e-06, "loss": 0.0334, "step": 86260 }, { "grad_norm": 0.38023486733436584, "learning_rate": 5.066662005302175e-06, "loss": 0.0224, "step": 86270 }, { "grad_norm": 0.46832871437072754, "learning_rate": 5.059411814186838e-06, "loss": 0.0295, "step": 86280 }, { "grad_norm": 0.4010372459888458, "learning_rate": 5.05216653770666e-06, "loss": 0.0352, "step": 86290 }, { "grad_norm": 0.3745596706867218, "learning_rate": 5.044926176653969e-06, "loss": 0.027, "step": 86300 }, { "grad_norm": 0.4802546799182892, "learning_rate": 5.037690731820571e-06, "loss": 0.0284, "step": 86310 }, { "grad_norm": 0.4308428466320038, "learning_rate": 5.030460203997711e-06, "loss": 0.0268, "step": 86320 }, { "grad_norm": 0.5613656640052795, "learning_rate": 5.0232345939761174e-06, "loss": 0.0264, "step": 86330 }, { "grad_norm": 0.4131912589073181, "learning_rate": 5.016013902545957e-06, "loss": 0.0312, "step": 86340 }, { "grad_norm": 0.6987098455429077, "learning_rate": 5.008798130496889e-06, "loss": 0.0393, "step": 86350 }, { "grad_norm": 0.556537926197052, "learning_rate": 5.001587278618008e-06, "loss": 0.0331, "step": 86360 }, { "grad_norm": 0.39700213074684143, "learning_rate": 4.994381347697891e-06, "loss": 0.0261, "step": 86370 }, { "grad_norm": 0.3035520911216736, "learning_rate": 4.987180338524572e-06, "loss": 0.026, "step": 86380 }, { "grad_norm": 0.472915381193161, "learning_rate": 4.9799842518855275e-06, "loss": 0.0317, "step": 86390 }, { "grad_norm": 0.580980658531189, "learning_rate": 4.9727930885677285e-06, "loss": 0.0349, "step": 86400 }, { "grad_norm": 0.251234233379364, "learning_rate": 4.965606849357573e-06, "loss": 0.0298, "step": 86410 }, { "grad_norm": 0.30103421211242676, "learning_rate": 4.9584255350409445e-06, "loss": 0.0257, "step": 86420 }, { "grad_norm": 0.3122096657752991, "learning_rate": 4.951249146403175e-06, "loss": 0.0333, "step": 86430 }, { "grad_norm": 0.37053629755973816, "learning_rate": 4.9440776842290815e-06, "loss": 0.0303, "step": 86440 }, { "grad_norm": 0.37648388743400574, "learning_rate": 4.936911149302903e-06, "loss": 0.0349, "step": 86450 }, { "grad_norm": 0.6066662669181824, "learning_rate": 4.929749542408379e-06, "loss": 0.0301, "step": 86460 }, { "grad_norm": 0.2753843069076538, "learning_rate": 4.922592864328673e-06, "loss": 0.0265, "step": 86470 }, { "grad_norm": 1.1303473711013794, "learning_rate": 4.915441115846436e-06, "loss": 0.0442, "step": 86480 }, { "grad_norm": 0.3363534212112427, "learning_rate": 4.908294297743782e-06, "loss": 0.0342, "step": 86490 }, { "grad_norm": 0.33223527669906616, "learning_rate": 4.901152410802273e-06, "loss": 0.0267, "step": 86500 }, { "grad_norm": 0.8422300815582275, "learning_rate": 4.894015455802919e-06, "loss": 0.0411, "step": 86510 }, { "grad_norm": 0.5922983884811401, "learning_rate": 4.8868834335262346e-06, "loss": 0.0378, "step": 86520 }, { "grad_norm": 0.29149189591407776, "learning_rate": 4.87975634475214e-06, "loss": 0.0262, "step": 86530 }, { "grad_norm": 0.3879345655441284, "learning_rate": 4.8726341902600614e-06, "loss": 0.0324, "step": 86540 }, { "grad_norm": 0.45075225830078125, "learning_rate": 4.865516970828843e-06, "loss": 0.0312, "step": 86550 }, { "grad_norm": 0.9609775543212891, "learning_rate": 4.85840468723685e-06, "loss": 0.0336, "step": 86560 }, { "grad_norm": 0.34399524331092834, "learning_rate": 4.85129734026184e-06, "loss": 0.0277, "step": 86570 }, { "grad_norm": 0.39132222533226013, "learning_rate": 4.844194930681073e-06, "loss": 0.0203, "step": 86580 }, { "grad_norm": 0.28608909249305725, "learning_rate": 4.837097459271267e-06, "loss": 0.0394, "step": 86590 }, { "grad_norm": 0.49343833327293396, "learning_rate": 4.830004926808573e-06, "loss": 0.0259, "step": 86600 }, { "grad_norm": 0.40752077102661133, "learning_rate": 4.822917334068627e-06, "loss": 0.0273, "step": 86610 }, { "grad_norm": 1.0229508876800537, "learning_rate": 4.81583468182652e-06, "loss": 0.022, "step": 86620 }, { "grad_norm": 0.21034368872642517, "learning_rate": 4.808756970856809e-06, "loss": 0.0326, "step": 86630 }, { "grad_norm": 0.3363710641860962, "learning_rate": 4.801684201933476e-06, "loss": 0.0237, "step": 86640 }, { "grad_norm": 0.2837977111339569, "learning_rate": 4.794616375830019e-06, "loss": 0.0251, "step": 86650 }, { "grad_norm": 0.4117758274078369, "learning_rate": 4.787553493319341e-06, "loss": 0.0267, "step": 86660 }, { "grad_norm": 0.17952562868595123, "learning_rate": 4.780495555173836e-06, "loss": 0.0227, "step": 86670 }, { "grad_norm": 0.3210162818431854, "learning_rate": 4.7734425621653475e-06, "loss": 0.0299, "step": 86680 }, { "grad_norm": 0.30636733770370483, "learning_rate": 4.766394515065198e-06, "loss": 0.022, "step": 86690 }, { "grad_norm": 0.3556784391403198, "learning_rate": 4.75935141464412e-06, "loss": 0.0302, "step": 86700 }, { "grad_norm": 0.5591493248939514, "learning_rate": 4.75231326167237e-06, "loss": 0.0259, "step": 86710 }, { "grad_norm": 0.3478275239467621, "learning_rate": 4.745280056919599e-06, "loss": 0.0252, "step": 86720 }, { "grad_norm": 0.39996761083602905, "learning_rate": 4.738251801154975e-06, "loss": 0.0251, "step": 86730 }, { "grad_norm": 0.20243193209171295, "learning_rate": 4.731228495147061e-06, "loss": 0.0296, "step": 86740 }, { "grad_norm": 0.58418208360672, "learning_rate": 4.7242101396639595e-06, "loss": 0.0299, "step": 86750 }, { "grad_norm": 0.36072513461112976, "learning_rate": 4.717196735473151e-06, "loss": 0.0273, "step": 86760 }, { "grad_norm": 0.32726848125457764, "learning_rate": 4.710188283341643e-06, "loss": 0.0316, "step": 86770 }, { "grad_norm": 0.4622173607349396, "learning_rate": 4.703184784035835e-06, "loss": 0.0375, "step": 86780 }, { "grad_norm": 0.3674147427082062, "learning_rate": 4.696186238321642e-06, "loss": 0.029, "step": 86790 }, { "grad_norm": 0.41588935256004333, "learning_rate": 4.689192646964408e-06, "loss": 0.0348, "step": 86800 }, { "grad_norm": 0.6829690337181091, "learning_rate": 4.682204010728952e-06, "loss": 0.0315, "step": 86810 }, { "grad_norm": 0.36306890845298767, "learning_rate": 4.67522033037952e-06, "loss": 0.0191, "step": 86820 }, { "grad_norm": 0.2824445366859436, "learning_rate": 4.668241606679846e-06, "loss": 0.0334, "step": 86830 }, { "grad_norm": 0.3053497076034546, "learning_rate": 4.66126784039313e-06, "loss": 0.0272, "step": 86840 }, { "grad_norm": 0.36540213227272034, "learning_rate": 4.6542990322819836e-06, "loss": 0.0353, "step": 86850 }, { "grad_norm": 0.23449048399925232, "learning_rate": 4.64733518310852e-06, "loss": 0.0239, "step": 86860 }, { "grad_norm": 0.47150325775146484, "learning_rate": 4.640376293634291e-06, "loss": 0.0275, "step": 86870 }, { "grad_norm": 0.5854222774505615, "learning_rate": 4.633422364620327e-06, "loss": 0.0361, "step": 86880 }, { "grad_norm": 0.7989470362663269, "learning_rate": 4.6264733968270745e-06, "loss": 0.0363, "step": 86890 }, { "grad_norm": 0.4765584468841553, "learning_rate": 4.619529391014482e-06, "loss": 0.0298, "step": 86900 }, { "grad_norm": 0.3034853935241699, "learning_rate": 4.61259034794192e-06, "loss": 0.0273, "step": 86910 }, { "grad_norm": 0.6056811809539795, "learning_rate": 4.605656268368247e-06, "loss": 0.0333, "step": 86920 }, { "grad_norm": 0.29057642817497253, "learning_rate": 4.598727153051735e-06, "loss": 0.0286, "step": 86930 }, { "grad_norm": 0.3092264235019684, "learning_rate": 4.591803002750178e-06, "loss": 0.0336, "step": 86940 }, { "grad_norm": 0.31401848793029785, "learning_rate": 4.58488381822077e-06, "loss": 0.0273, "step": 86950 }, { "grad_norm": 1.6055505275726318, "learning_rate": 4.5779696002201946e-06, "loss": 0.0313, "step": 86960 }, { "grad_norm": 0.38060104846954346, "learning_rate": 4.571060349504563e-06, "loss": 0.0285, "step": 86970 }, { "grad_norm": 0.31776100397109985, "learning_rate": 4.564156066829478e-06, "loss": 0.0261, "step": 86980 }, { "grad_norm": 0.21287566423416138, "learning_rate": 4.557256752949957e-06, "loss": 0.0342, "step": 86990 }, { "grad_norm": 0.5036935210227966, "learning_rate": 4.550362408620534e-06, "loss": 0.0266, "step": 87000 }, { "grad_norm": 0.18240441381931305, "learning_rate": 4.543473034595136e-06, "loss": 0.0288, "step": 87010 }, { "grad_norm": 0.6281523704528809, "learning_rate": 4.536588631627192e-06, "loss": 0.0335, "step": 87020 }, { "grad_norm": 0.4408918023109436, "learning_rate": 4.5297092004695565e-06, "loss": 0.031, "step": 87030 }, { "grad_norm": 0.30441659688949585, "learning_rate": 4.522834741874555e-06, "loss": 0.029, "step": 87040 }, { "grad_norm": 0.4608466625213623, "learning_rate": 4.515965256593974e-06, "loss": 0.0307, "step": 87050 }, { "grad_norm": 0.3413713872432709, "learning_rate": 4.509100745379058e-06, "loss": 0.0308, "step": 87060 }, { "grad_norm": 0.2131984680891037, "learning_rate": 4.502241208980479e-06, "loss": 0.028, "step": 87070 }, { "grad_norm": 0.6873742938041687, "learning_rate": 4.495386648148398e-06, "loss": 0.0276, "step": 87080 }, { "grad_norm": 0.3023506700992584, "learning_rate": 4.488537063632431e-06, "loss": 0.0337, "step": 87090 }, { "grad_norm": 0.4956401586532593, "learning_rate": 4.481692456181608e-06, "loss": 0.0327, "step": 87100 }, { "grad_norm": 0.3723278343677521, "learning_rate": 4.474852826544479e-06, "loss": 0.0277, "step": 87110 }, { "grad_norm": 0.4899342358112335, "learning_rate": 4.468018175468974e-06, "loss": 0.0267, "step": 87120 }, { "grad_norm": 0.36238524317741394, "learning_rate": 4.461188503702568e-06, "loss": 0.0389, "step": 87130 }, { "grad_norm": 0.8621293902397156, "learning_rate": 4.454363811992107e-06, "loss": 0.0392, "step": 87140 }, { "grad_norm": 0.4297279119491577, "learning_rate": 4.447544101083956e-06, "loss": 0.029, "step": 87150 }, { "grad_norm": 1.1638997793197632, "learning_rate": 4.440729371723884e-06, "loss": 0.0351, "step": 87160 }, { "grad_norm": 0.31864938139915466, "learning_rate": 4.433919624657162e-06, "loss": 0.0228, "step": 87170 }, { "grad_norm": 0.6358467936515808, "learning_rate": 4.427114860628462e-06, "loss": 0.0318, "step": 87180 }, { "grad_norm": 0.3644872009754181, "learning_rate": 4.420315080381982e-06, "loss": 0.0273, "step": 87190 }, { "grad_norm": 0.3008538484573364, "learning_rate": 4.4135202846613045e-06, "loss": 0.0316, "step": 87200 }, { "grad_norm": 0.4387553036212921, "learning_rate": 4.406730474209525e-06, "loss": 0.0294, "step": 87210 }, { "grad_norm": 0.3927527666091919, "learning_rate": 4.399945649769138e-06, "loss": 0.0355, "step": 87220 }, { "grad_norm": 0.4826967716217041, "learning_rate": 4.393165812082151e-06, "loss": 0.0259, "step": 87230 }, { "grad_norm": 0.2969164252281189, "learning_rate": 4.386390961889969e-06, "loss": 0.0359, "step": 87240 }, { "grad_norm": 0.5560754537582397, "learning_rate": 4.3796210999334895e-06, "loss": 0.0217, "step": 87250 }, { "grad_norm": 0.39468714594841003, "learning_rate": 4.372856226953065e-06, "loss": 0.028, "step": 87260 }, { "grad_norm": 0.5296738147735596, "learning_rate": 4.366096343688486e-06, "loss": 0.0271, "step": 87270 }, { "grad_norm": 0.4254184663295746, "learning_rate": 4.359341450878995e-06, "loss": 0.0224, "step": 87280 }, { "grad_norm": 0.46179208159446716, "learning_rate": 4.352591549263302e-06, "loss": 0.0318, "step": 87290 }, { "grad_norm": 0.347013384103775, "learning_rate": 4.345846639579582e-06, "loss": 0.0275, "step": 87300 }, { "grad_norm": 0.6633291840553284, "learning_rate": 4.339106722565417e-06, "loss": 0.0291, "step": 87310 }, { "grad_norm": 0.43175992369651794, "learning_rate": 4.332371798957896e-06, "loss": 0.0366, "step": 87320 }, { "grad_norm": 0.38459521532058716, "learning_rate": 4.3256418694935295e-06, "loss": 0.0355, "step": 87330 }, { "grad_norm": 0.330127477645874, "learning_rate": 4.3189169349083125e-06, "loss": 0.023, "step": 87340 }, { "grad_norm": 0.37591123580932617, "learning_rate": 4.312196995937645e-06, "loss": 0.0307, "step": 87350 }, { "grad_norm": 0.3319598436355591, "learning_rate": 4.305482053316434e-06, "loss": 0.0324, "step": 87360 }, { "grad_norm": 0.43673691153526306, "learning_rate": 4.2987721077789925e-06, "loss": 0.0294, "step": 87370 }, { "grad_norm": 0.28429850935935974, "learning_rate": 4.2920671600591325e-06, "loss": 0.0275, "step": 87380 }, { "grad_norm": 0.3241718113422394, "learning_rate": 4.285367210890084e-06, "loss": 0.0281, "step": 87390 }, { "grad_norm": 0.20573608577251434, "learning_rate": 4.278672261004551e-06, "loss": 0.0319, "step": 87400 }, { "grad_norm": 0.36203789710998535, "learning_rate": 4.2719823111346704e-06, "loss": 0.0329, "step": 87410 }, { "grad_norm": 0.34010404348373413, "learning_rate": 4.265297362012061e-06, "loss": 0.0211, "step": 87420 }, { "grad_norm": 0.28078001737594604, "learning_rate": 4.258617414367766e-06, "loss": 0.0272, "step": 87430 }, { "grad_norm": 0.28576013445854187, "learning_rate": 4.251942468932296e-06, "loss": 0.0255, "step": 87440 }, { "grad_norm": 0.5573794841766357, "learning_rate": 4.2452725264356175e-06, "loss": 0.023, "step": 87450 }, { "grad_norm": 0.9460787773132324, "learning_rate": 4.2386075876071575e-06, "loss": 0.0291, "step": 87460 }, { "grad_norm": 0.3854362964630127, "learning_rate": 4.231947653175755e-06, "loss": 0.0327, "step": 87470 }, { "grad_norm": 0.6388932466506958, "learning_rate": 4.225292723869761e-06, "loss": 0.0353, "step": 87480 }, { "grad_norm": 0.2735166549682617, "learning_rate": 4.218642800416916e-06, "loss": 0.0312, "step": 87490 }, { "grad_norm": 0.5830062031745911, "learning_rate": 4.211997883544472e-06, "loss": 0.0221, "step": 87500 }, { "grad_norm": 0.32873445749282837, "learning_rate": 4.205357973979096e-06, "loss": 0.0305, "step": 87510 }, { "grad_norm": 0.27891111373901367, "learning_rate": 4.198723072446925e-06, "loss": 0.0246, "step": 87520 }, { "grad_norm": 0.4782174229621887, "learning_rate": 4.192093179673534e-06, "loss": 0.027, "step": 87530 }, { "grad_norm": 0.3407703936100006, "learning_rate": 4.185468296383959e-06, "loss": 0.0321, "step": 87540 }, { "grad_norm": 0.6634054183959961, "learning_rate": 4.1788484233027005e-06, "loss": 0.0383, "step": 87550 }, { "grad_norm": 0.6618199944496155, "learning_rate": 4.172233561153677e-06, "loss": 0.0235, "step": 87560 }, { "grad_norm": 0.21058692038059235, "learning_rate": 4.165623710660288e-06, "loss": 0.0279, "step": 87570 }, { "grad_norm": 0.2804737389087677, "learning_rate": 4.1590188725453835e-06, "loss": 0.0269, "step": 87580 }, { "grad_norm": 1.3932936191558838, "learning_rate": 4.152419047531259e-06, "loss": 0.0314, "step": 87590 }, { "grad_norm": 0.3017428517341614, "learning_rate": 4.145824236339651e-06, "loss": 0.02, "step": 87600 }, { "grad_norm": 0.37253686785697937, "learning_rate": 4.1392344396917695e-06, "loss": 0.0196, "step": 87610 }, { "grad_norm": 0.2764524817466736, "learning_rate": 4.132649658308252e-06, "loss": 0.0268, "step": 87620 }, { "grad_norm": 0.458932489156723, "learning_rate": 4.126069892909213e-06, "loss": 0.0219, "step": 87630 }, { "grad_norm": 0.285500168800354, "learning_rate": 4.119495144214191e-06, "loss": 0.0291, "step": 87640 }, { "grad_norm": 0.2924859821796417, "learning_rate": 4.112925412942214e-06, "loss": 0.0269, "step": 87650 }, { "grad_norm": 0.45181357860565186, "learning_rate": 4.106360699811712e-06, "loss": 0.0294, "step": 87660 }, { "grad_norm": 0.3120811879634857, "learning_rate": 4.099801005540616e-06, "loss": 0.026, "step": 87670 }, { "grad_norm": 0.4148661494255066, "learning_rate": 4.093246330846256e-06, "loss": 0.0244, "step": 87680 }, { "grad_norm": 0.40885257720947266, "learning_rate": 4.086696676445467e-06, "loss": 0.0268, "step": 87690 }, { "grad_norm": 0.5978702306747437, "learning_rate": 4.0801520430544914e-06, "loss": 0.0213, "step": 87700 }, { "grad_norm": 0.4183332026004791, "learning_rate": 4.073612431389062e-06, "loss": 0.0263, "step": 87710 }, { "grad_norm": 0.42220601439476013, "learning_rate": 4.067077842164319e-06, "loss": 0.0237, "step": 87720 }, { "grad_norm": 0.45956072211265564, "learning_rate": 4.0605482760948945e-06, "loss": 0.0362, "step": 87730 }, { "grad_norm": 0.38794055581092834, "learning_rate": 4.054023733894829e-06, "loss": 0.0294, "step": 87740 }, { "grad_norm": 0.8624640703201294, "learning_rate": 4.047504216277653e-06, "loss": 0.0316, "step": 87750 }, { "grad_norm": 0.3423224985599518, "learning_rate": 4.040989723956323e-06, "loss": 0.0296, "step": 87760 }, { "grad_norm": 0.523064374923706, "learning_rate": 4.03448025764327e-06, "loss": 0.0407, "step": 87770 }, { "grad_norm": 0.3018878698348999, "learning_rate": 4.027975818050339e-06, "loss": 0.0292, "step": 87780 }, { "grad_norm": 0.4864671528339386, "learning_rate": 4.021476405888858e-06, "loss": 0.0291, "step": 87790 }, { "grad_norm": 0.4621662199497223, "learning_rate": 4.01498202186959e-06, "loss": 0.0248, "step": 87800 }, { "grad_norm": 0.5516571402549744, "learning_rate": 4.008492666702751e-06, "loss": 0.0299, "step": 87810 }, { "grad_norm": 0.32858726382255554, "learning_rate": 4.002008341098001e-06, "loss": 0.0352, "step": 87820 }, { "grad_norm": 0.35954758524894714, "learning_rate": 3.995529045764462e-06, "loss": 0.0236, "step": 87830 }, { "grad_norm": 0.3022671341896057, "learning_rate": 3.989054781410712e-06, "loss": 0.0248, "step": 87840 }, { "grad_norm": 1.2706960439682007, "learning_rate": 3.982585548744744e-06, "loss": 0.0304, "step": 87850 }, { "grad_norm": 0.6217344999313354, "learning_rate": 3.9761213484740435e-06, "loss": 0.0269, "step": 87860 }, { "grad_norm": 0.3956950902938843, "learning_rate": 3.9696621813055055e-06, "loss": 0.0251, "step": 87870 }, { "grad_norm": 0.18367525935173035, "learning_rate": 3.9632080479455044e-06, "loss": 0.0323, "step": 87880 }, { "grad_norm": 0.571371853351593, "learning_rate": 3.956758949099859e-06, "loss": 0.0312, "step": 87890 }, { "grad_norm": 0.3364224135875702, "learning_rate": 3.9503148854738385e-06, "loss": 0.0304, "step": 87900 }, { "grad_norm": 0.7414334416389465, "learning_rate": 3.943875857772139e-06, "loss": 0.0401, "step": 87910 }, { "grad_norm": 0.5069018006324768, "learning_rate": 3.9374418666989375e-06, "loss": 0.0322, "step": 87920 }, { "grad_norm": 0.43959537148475647, "learning_rate": 3.931012912957838e-06, "loss": 0.027, "step": 87930 }, { "grad_norm": 0.2555628716945648, "learning_rate": 3.9245889972519045e-06, "loss": 0.026, "step": 87940 }, { "grad_norm": 0.3846839964389801, "learning_rate": 3.9181701202836265e-06, "loss": 0.0328, "step": 87950 }, { "grad_norm": 0.46034353971481323, "learning_rate": 3.911756282755003e-06, "loss": 0.0282, "step": 87960 }, { "grad_norm": 0.5499197244644165, "learning_rate": 3.905347485367411e-06, "loss": 0.0302, "step": 87970 }, { "grad_norm": 0.2732591927051544, "learning_rate": 3.898943728821725e-06, "loss": 0.0208, "step": 87980 }, { "grad_norm": 0.30658307671546936, "learning_rate": 3.892545013818227e-06, "loss": 0.026, "step": 87990 }, { "grad_norm": 0.34969887137413025, "learning_rate": 3.886151341056687e-06, "loss": 0.0273, "step": 88000 }, { "grad_norm": 0.4189677834510803, "learning_rate": 3.8797627112363066e-06, "loss": 0.0313, "step": 88010 }, { "grad_norm": 0.3577994704246521, "learning_rate": 3.873379125055743e-06, "loss": 0.0242, "step": 88020 }, { "grad_norm": 0.507621705532074, "learning_rate": 3.867000583213082e-06, "loss": 0.0266, "step": 88030 }, { "grad_norm": 0.3292138874530792, "learning_rate": 3.860627086405871e-06, "loss": 0.0333, "step": 88040 }, { "grad_norm": 0.5576412677764893, "learning_rate": 3.854258635331126e-06, "loss": 0.0309, "step": 88050 }, { "grad_norm": 0.2961186468601227, "learning_rate": 3.8478952306852665e-06, "loss": 0.0279, "step": 88060 }, { "grad_norm": 0.38332343101501465, "learning_rate": 3.841536873164198e-06, "loss": 0.0297, "step": 88070 }, { "grad_norm": 0.31079772114753723, "learning_rate": 3.8351835634632626e-06, "loss": 0.0265, "step": 88080 }, { "grad_norm": 1.8316771984100342, "learning_rate": 3.828835302277245e-06, "loss": 0.0339, "step": 88090 }, { "grad_norm": 0.8086695075035095, "learning_rate": 3.822492090300378e-06, "loss": 0.0376, "step": 88100 }, { "grad_norm": 0.36931610107421875, "learning_rate": 3.8161539282263566e-06, "loss": 0.0238, "step": 88110 }, { "grad_norm": 0.6711835265159607, "learning_rate": 3.8098208167482995e-06, "loss": 0.0264, "step": 88120 }, { "grad_norm": 0.8873575925827026, "learning_rate": 3.803492756558796e-06, "loss": 0.0306, "step": 88130 }, { "grad_norm": 0.3373478949069977, "learning_rate": 3.797169748349855e-06, "loss": 0.0305, "step": 88140 }, { "grad_norm": 0.4737894535064697, "learning_rate": 3.790851792812977e-06, "loss": 0.0258, "step": 88150 }, { "grad_norm": 0.1454281508922577, "learning_rate": 3.784538890639072e-06, "loss": 0.0225, "step": 88160 }, { "grad_norm": 0.3359827697277069, "learning_rate": 3.7782310425185153e-06, "loss": 0.0323, "step": 88170 }, { "grad_norm": 0.906344473361969, "learning_rate": 3.77192824914111e-06, "loss": 0.0186, "step": 88180 }, { "grad_norm": 0.3487299382686615, "learning_rate": 3.7656305111961333e-06, "loss": 0.0301, "step": 88190 }, { "grad_norm": 0.4481259882450104, "learning_rate": 3.7593378293722782e-06, "loss": 0.0363, "step": 88200 }, { "grad_norm": 0.42839759588241577, "learning_rate": 3.7530502043577333e-06, "loss": 0.0286, "step": 88210 }, { "grad_norm": 0.4897910952568054, "learning_rate": 3.746767636840076e-06, "loss": 0.0259, "step": 88220 }, { "grad_norm": 0.255216121673584, "learning_rate": 3.74049012750638e-06, "loss": 0.0327, "step": 88230 }, { "grad_norm": 0.766637921333313, "learning_rate": 3.734217677043128e-06, "loss": 0.0206, "step": 88240 }, { "grad_norm": 0.45051315426826477, "learning_rate": 3.727950286136267e-06, "loss": 0.031, "step": 88250 }, { "grad_norm": 0.34329450130462646, "learning_rate": 3.7216879554711925e-06, "loss": 0.0272, "step": 88260 }, { "grad_norm": 0.5068956017494202, "learning_rate": 3.715430685732757e-06, "loss": 0.031, "step": 88270 }, { "grad_norm": 0.6659764051437378, "learning_rate": 3.7091784776052243e-06, "loss": 0.0256, "step": 88280 }, { "grad_norm": 0.41217589378356934, "learning_rate": 3.7029313317723356e-06, "loss": 0.0277, "step": 88290 }, { "grad_norm": 0.24653564393520355, "learning_rate": 3.6966892489172787e-06, "loss": 0.032, "step": 88300 }, { "grad_norm": 0.664855420589447, "learning_rate": 3.6904522297226575e-06, "loss": 0.0307, "step": 88310 }, { "grad_norm": 0.3241491913795471, "learning_rate": 3.6842202748705647e-06, "loss": 0.0277, "step": 88320 }, { "grad_norm": 0.38759922981262207, "learning_rate": 3.677993385042494e-06, "loss": 0.0315, "step": 88330 }, { "grad_norm": 0.4431806802749634, "learning_rate": 3.67177156091944e-06, "loss": 0.0314, "step": 88340 }, { "grad_norm": 0.3468986749649048, "learning_rate": 3.6655548031817866e-06, "loss": 0.0216, "step": 88350 }, { "grad_norm": 0.5518692135810852, "learning_rate": 3.6593431125094e-06, "loss": 0.0323, "step": 88360 }, { "grad_norm": 0.44140714406967163, "learning_rate": 3.6531364895815766e-06, "loss": 0.0247, "step": 88370 }, { "grad_norm": 0.5516344308853149, "learning_rate": 3.6469349350770677e-06, "loss": 0.0261, "step": 88380 }, { "grad_norm": 0.5352041721343994, "learning_rate": 3.6407384496740525e-06, "loss": 0.0322, "step": 88390 }, { "grad_norm": 0.6150928139686584, "learning_rate": 3.6345470340501954e-06, "loss": 0.0331, "step": 88400 }, { "grad_norm": 0.4768381714820862, "learning_rate": 3.6283606888825537e-06, "loss": 0.0263, "step": 88410 }, { "grad_norm": 0.5080942511558533, "learning_rate": 3.622179414847676e-06, "loss": 0.0312, "step": 88420 }, { "grad_norm": 0.6507163047790527, "learning_rate": 3.616003212621527e-06, "loss": 0.0329, "step": 88430 }, { "grad_norm": 0.37784135341644287, "learning_rate": 3.609832082879533e-06, "loss": 0.0245, "step": 88440 }, { "grad_norm": 0.2324192076921463, "learning_rate": 3.6036660262965485e-06, "loss": 0.0324, "step": 88450 }, { "grad_norm": 0.3035406768321991, "learning_rate": 3.597505043546895e-06, "loss": 0.0249, "step": 88460 }, { "grad_norm": 0.4914097785949707, "learning_rate": 3.5913491353043227e-06, "loss": 0.0206, "step": 88470 }, { "grad_norm": 0.4039199948310852, "learning_rate": 3.5851983022420375e-06, "loss": 0.0314, "step": 88480 }, { "grad_norm": 0.5848303437232971, "learning_rate": 3.5790525450326797e-06, "loss": 0.0224, "step": 88490 }, { "grad_norm": 0.33439916372299194, "learning_rate": 3.572911864348344e-06, "loss": 0.025, "step": 88500 }, { "grad_norm": 0.3727011978626251, "learning_rate": 3.5667762608605717e-06, "loss": 0.0291, "step": 88510 }, { "grad_norm": 0.391817569732666, "learning_rate": 3.5606457352403255e-06, "loss": 0.0268, "step": 88520 }, { "grad_norm": 0.6736997961997986, "learning_rate": 3.554520288158042e-06, "loss": 0.0228, "step": 88530 }, { "grad_norm": 0.18906627595424652, "learning_rate": 3.5483999202835903e-06, "loss": 0.0269, "step": 88540 }, { "grad_norm": 0.4336428642272949, "learning_rate": 3.5422846322862913e-06, "loss": 0.0264, "step": 88550 }, { "grad_norm": 0.18354621529579163, "learning_rate": 3.5361744248348925e-06, "loss": 0.0237, "step": 88560 }, { "grad_norm": 0.2859600782394409, "learning_rate": 3.5300692985976047e-06, "loss": 0.0242, "step": 88570 }, { "grad_norm": 0.4543491303920746, "learning_rate": 3.52396925424206e-06, "loss": 0.023, "step": 88580 }, { "grad_norm": 0.45747318863868713, "learning_rate": 3.5178742924353747e-06, "loss": 0.0267, "step": 88590 }, { "grad_norm": 0.3127526342868805, "learning_rate": 3.511784413844066e-06, "loss": 0.0309, "step": 88600 }, { "grad_norm": 0.41395100951194763, "learning_rate": 3.5056996191341287e-06, "loss": 0.0409, "step": 88610 }, { "grad_norm": 0.46022745966911316, "learning_rate": 3.4996199089709692e-06, "loss": 0.0305, "step": 88620 }, { "grad_norm": 0.20847953855991364, "learning_rate": 3.4935452840194725e-06, "loss": 0.0292, "step": 88630 }, { "grad_norm": 0.5126996636390686, "learning_rate": 3.48747574494393e-06, "loss": 0.0298, "step": 88640 }, { "grad_norm": 0.5549286603927612, "learning_rate": 3.4814112924081154e-06, "loss": 0.0307, "step": 88650 }, { "grad_norm": 0.7855004668235779, "learning_rate": 3.475351927075221e-06, "loss": 0.0276, "step": 88660 }, { "grad_norm": 0.4035508334636688, "learning_rate": 3.469297649607894e-06, "loss": 0.0221, "step": 88670 }, { "grad_norm": 0.6212500333786011, "learning_rate": 3.463248460668217e-06, "loss": 0.0268, "step": 88680 }, { "grad_norm": 0.5280065536499023, "learning_rate": 3.457204360917726e-06, "loss": 0.0327, "step": 88690 }, { "grad_norm": 0.5413952469825745, "learning_rate": 3.4511653510173827e-06, "loss": 0.0256, "step": 88700 }, { "grad_norm": 0.2471849024295807, "learning_rate": 3.4451314316276074e-06, "loss": 0.0268, "step": 88710 }, { "grad_norm": 0.5185454487800598, "learning_rate": 3.4391026034082674e-06, "loss": 0.0351, "step": 88720 }, { "grad_norm": 0.4121955633163452, "learning_rate": 3.433078867018674e-06, "loss": 0.0277, "step": 88730 }, { "grad_norm": 0.6240317225456238, "learning_rate": 3.427060223117551e-06, "loss": 0.0279, "step": 88740 }, { "grad_norm": 0.6081361174583435, "learning_rate": 3.4210466723631042e-06, "loss": 0.0279, "step": 88750 }, { "grad_norm": 0.3203197121620178, "learning_rate": 3.415038215412969e-06, "loss": 0.0191, "step": 88760 }, { "grad_norm": 0.558540403842926, "learning_rate": 3.4090348529242088e-06, "loss": 0.0358, "step": 88770 }, { "grad_norm": 0.8514910936355591, "learning_rate": 3.4030365855533487e-06, "loss": 0.0382, "step": 88780 }, { "grad_norm": 0.5278912782669067, "learning_rate": 3.397043413956347e-06, "loss": 0.0403, "step": 88790 }, { "grad_norm": 0.22942592203617096, "learning_rate": 3.3910553387886237e-06, "loss": 0.0246, "step": 88800 }, { "grad_norm": 0.40629005432128906, "learning_rate": 3.3850723607049996e-06, "loss": 0.0331, "step": 88810 }, { "grad_norm": 0.34373000264167786, "learning_rate": 3.3790944803597902e-06, "loss": 0.0271, "step": 88820 }, { "grad_norm": 0.258480042219162, "learning_rate": 3.373121698406706e-06, "loss": 0.0336, "step": 88830 }, { "grad_norm": 0.4597350060939789, "learning_rate": 3.3671540154989302e-06, "loss": 0.0233, "step": 88840 }, { "grad_norm": 0.5968821048736572, "learning_rate": 3.361191432289079e-06, "loss": 0.0329, "step": 88850 }, { "grad_norm": 0.5740665793418884, "learning_rate": 3.3552339494292207e-06, "loss": 0.0361, "step": 88860 }, { "grad_norm": 0.8384631872177124, "learning_rate": 3.349281567570839e-06, "loss": 0.033, "step": 88870 }, { "grad_norm": 0.5778340101242065, "learning_rate": 3.3433342873648965e-06, "loss": 0.0304, "step": 88880 }, { "grad_norm": 0.36886441707611084, "learning_rate": 3.337392109461762e-06, "loss": 0.0255, "step": 88890 }, { "grad_norm": 0.3008195459842682, "learning_rate": 3.331455034511266e-06, "loss": 0.0267, "step": 88900 }, { "grad_norm": 0.2790750563144684, "learning_rate": 3.325523063162689e-06, "loss": 0.0301, "step": 88910 }, { "grad_norm": 0.5853429436683655, "learning_rate": 3.3195961960647393e-06, "loss": 0.0245, "step": 88920 }, { "grad_norm": 0.33197513222694397, "learning_rate": 3.31367443386556e-06, "loss": 0.0181, "step": 88930 }, { "grad_norm": 0.2294185310602188, "learning_rate": 3.3077577772127556e-06, "loss": 0.0251, "step": 88940 }, { "grad_norm": 0.36292555928230286, "learning_rate": 3.301846226753358e-06, "loss": 0.026, "step": 88950 }, { "grad_norm": 0.48547783493995667, "learning_rate": 3.2959397831338444e-06, "loss": 0.0314, "step": 88960 }, { "grad_norm": 0.42678236961364746, "learning_rate": 3.2900384470001432e-06, "loss": 0.0273, "step": 88970 }, { "grad_norm": 0.346590131521225, "learning_rate": 3.28414221899761e-06, "loss": 0.026, "step": 88980 }, { "grad_norm": 0.6750046610832214, "learning_rate": 3.278251099771046e-06, "loss": 0.0352, "step": 88990 }, { "grad_norm": 0.43257781863212585, "learning_rate": 3.2723650899646906e-06, "loss": 0.0336, "step": 89000 }, { "grad_norm": 0.4413951337337494, "learning_rate": 3.2664841902222466e-06, "loss": 0.0284, "step": 89010 }, { "grad_norm": 0.3073276877403259, "learning_rate": 3.2606084011868208e-06, "loss": 0.0341, "step": 89020 }, { "grad_norm": 0.29577767848968506, "learning_rate": 3.254737723500989e-06, "loss": 0.0246, "step": 89030 }, { "grad_norm": 0.5251016616821289, "learning_rate": 3.2488721578067595e-06, "loss": 0.0255, "step": 89040 }, { "grad_norm": 0.274250864982605, "learning_rate": 3.243011704745591e-06, "loss": 0.024, "step": 89050 }, { "grad_norm": 0.5467252135276794, "learning_rate": 3.237156364958355e-06, "loss": 0.0291, "step": 89060 }, { "grad_norm": 0.4195893108844757, "learning_rate": 3.231306139085394e-06, "loss": 0.0284, "step": 89070 }, { "grad_norm": 0.6986241936683655, "learning_rate": 3.2254610277664797e-06, "loss": 0.0408, "step": 89080 }, { "grad_norm": 0.38418304920196533, "learning_rate": 3.2196210316408183e-06, "loss": 0.027, "step": 89090 }, { "grad_norm": 0.27407386898994446, "learning_rate": 3.2137861513470644e-06, "loss": 0.0241, "step": 89100 }, { "grad_norm": 0.3047579526901245, "learning_rate": 3.2079563875233252e-06, "loss": 0.0289, "step": 89110 }, { "grad_norm": 0.369203120470047, "learning_rate": 3.2021317408071184e-06, "loss": 0.0219, "step": 89120 }, { "grad_norm": 0.2686046063899994, "learning_rate": 3.196312211835434e-06, "loss": 0.0292, "step": 89130 }, { "grad_norm": 0.49832046031951904, "learning_rate": 3.1904978012446686e-06, "loss": 0.0197, "step": 89140 }, { "grad_norm": 0.5498576760292053, "learning_rate": 3.184688509670691e-06, "loss": 0.0225, "step": 89150 }, { "grad_norm": 0.2523723542690277, "learning_rate": 3.178884337748783e-06, "loss": 0.0284, "step": 89160 }, { "grad_norm": 0.43332427740097046, "learning_rate": 3.173085286113697e-06, "loss": 0.0314, "step": 89170 }, { "grad_norm": 0.435036838054657, "learning_rate": 3.1672913553995986e-06, "loss": 0.0356, "step": 89180 }, { "grad_norm": 0.1753174215555191, "learning_rate": 3.1615025462401138e-06, "loss": 0.0172, "step": 89190 }, { "grad_norm": 0.5645332932472229, "learning_rate": 3.1557188592682808e-06, "loss": 0.0326, "step": 89200 }, { "grad_norm": 0.32862627506256104, "learning_rate": 3.1499402951166045e-06, "loss": 0.0227, "step": 89210 }, { "grad_norm": 0.2835714817047119, "learning_rate": 3.144166854417019e-06, "loss": 0.0224, "step": 89220 }, { "grad_norm": 0.5220147967338562, "learning_rate": 3.1383985378009073e-06, "loss": 0.0264, "step": 89230 }, { "grad_norm": 0.35360369086265564, "learning_rate": 3.1326353458990654e-06, "loss": 0.029, "step": 89240 }, { "grad_norm": 0.44686779379844666, "learning_rate": 3.126877279341761e-06, "loss": 0.0258, "step": 89250 }, { "grad_norm": 0.3231358826160431, "learning_rate": 3.1211243387586964e-06, "loss": 0.0288, "step": 89260 }, { "grad_norm": 0.43557918071746826, "learning_rate": 3.115376524778979e-06, "loss": 0.0346, "step": 89270 }, { "grad_norm": 0.18202988803386688, "learning_rate": 3.109633838031195e-06, "loss": 0.0309, "step": 89280 }, { "grad_norm": 0.23380811512470245, "learning_rate": 3.103896279143359e-06, "loss": 0.0288, "step": 89290 }, { "grad_norm": 0.7142724990844727, "learning_rate": 3.098163848742924e-06, "loss": 0.035, "step": 89300 }, { "grad_norm": 0.3578644096851349, "learning_rate": 3.0924365474567675e-06, "loss": 0.0309, "step": 89310 }, { "grad_norm": 0.41802340745925903, "learning_rate": 3.086714375911237e-06, "loss": 0.0296, "step": 89320 }, { "grad_norm": 0.39068737626075745, "learning_rate": 3.0809973347320775e-06, "loss": 0.0323, "step": 89330 }, { "grad_norm": 0.4007256329059601, "learning_rate": 3.075285424544516e-06, "loss": 0.0336, "step": 89340 }, { "grad_norm": 0.9171292185783386, "learning_rate": 3.069578645973176e-06, "loss": 0.03, "step": 89350 }, { "grad_norm": 0.7036662697792053, "learning_rate": 3.0638769996421735e-06, "loss": 0.0338, "step": 89360 }, { "grad_norm": 0.7939574122428894, "learning_rate": 3.058180486175005e-06, "loss": 0.0322, "step": 89370 }, { "grad_norm": 0.20538708567619324, "learning_rate": 3.052489106194645e-06, "loss": 0.0223, "step": 89380 }, { "grad_norm": 0.3272094130516052, "learning_rate": 3.0468028603234946e-06, "loss": 0.0306, "step": 89390 }, { "grad_norm": 0.39979684352874756, "learning_rate": 3.04112174918339e-06, "loss": 0.0221, "step": 89400 }, { "grad_norm": 0.4930323362350464, "learning_rate": 3.0354457733955954e-06, "loss": 0.0319, "step": 89410 }, { "grad_norm": 0.2876621186733246, "learning_rate": 3.029774933580859e-06, "loss": 0.0221, "step": 89420 }, { "grad_norm": 0.7259917259216309, "learning_rate": 3.0241092303593065e-06, "loss": 0.0251, "step": 89430 }, { "grad_norm": 0.4831485450267792, "learning_rate": 3.018448664350554e-06, "loss": 0.03, "step": 89440 }, { "grad_norm": 0.31779518723487854, "learning_rate": 3.0127932361736055e-06, "loss": 0.0305, "step": 89450 }, { "grad_norm": 0.5556570291519165, "learning_rate": 3.00714294644695e-06, "loss": 0.026, "step": 89460 }, { "grad_norm": 0.6349818110466003, "learning_rate": 3.0014977957884828e-06, "loss": 0.0321, "step": 89470 }, { "grad_norm": 0.39176037907600403, "learning_rate": 2.99585778481557e-06, "loss": 0.025, "step": 89480 }, { "grad_norm": 0.318013072013855, "learning_rate": 2.990222914144969e-06, "loss": 0.033, "step": 89490 }, { "grad_norm": 0.2898317873477936, "learning_rate": 2.984593184392914e-06, "loss": 0.031, "step": 89500 }, { "grad_norm": 0.26747167110443115, "learning_rate": 2.978968596175069e-06, "loss": 0.0294, "step": 89510 }, { "grad_norm": 0.4516150653362274, "learning_rate": 2.9733491501065137e-06, "loss": 0.0278, "step": 89520 }, { "grad_norm": 0.3133315443992615, "learning_rate": 2.9677348468018005e-06, "loss": 0.0264, "step": 89530 }, { "grad_norm": 0.21623043715953827, "learning_rate": 2.9621256868748726e-06, "loss": 0.0265, "step": 89540 }, { "grad_norm": 0.24304386973381042, "learning_rate": 2.956521670939183e-06, "loss": 0.0289, "step": 89550 }, { "grad_norm": 0.15418604016304016, "learning_rate": 2.950922799607542e-06, "loss": 0.0291, "step": 89560 }, { "grad_norm": 0.4848038852214813, "learning_rate": 2.9453290734922535e-06, "loss": 0.0275, "step": 89570 }, { "grad_norm": 0.44060376286506653, "learning_rate": 2.939740493205029e-06, "loss": 0.0408, "step": 89580 }, { "grad_norm": 0.1871744692325592, "learning_rate": 2.9341570593570345e-06, "loss": 0.026, "step": 89590 }, { "grad_norm": 0.46071019768714905, "learning_rate": 2.928578772558843e-06, "loss": 0.0232, "step": 89600 }, { "grad_norm": 0.5368497967720032, "learning_rate": 2.923005633420528e-06, "loss": 0.0317, "step": 89610 }, { "grad_norm": 0.32051950693130493, "learning_rate": 2.9174376425515237e-06, "loss": 0.0311, "step": 89620 }, { "grad_norm": 1.1477690935134888, "learning_rate": 2.91187480056076e-06, "loss": 0.0313, "step": 89630 }, { "grad_norm": 0.5942862033843994, "learning_rate": 2.9063171080565664e-06, "loss": 0.0274, "step": 89640 }, { "grad_norm": 0.40067192912101746, "learning_rate": 2.900764565646741e-06, "loss": 0.025, "step": 89650 }, { "grad_norm": 0.2729856073856354, "learning_rate": 2.895217173938475e-06, "loss": 0.0303, "step": 89660 }, { "grad_norm": 0.7000411152839661, "learning_rate": 2.8896749335384445e-06, "loss": 0.0262, "step": 89670 }, { "grad_norm": 0.26025885343551636, "learning_rate": 2.884137845052731e-06, "loss": 0.036, "step": 89680 }, { "grad_norm": 0.3924007713794708, "learning_rate": 2.8786059090868734e-06, "loss": 0.0282, "step": 89690 }, { "grad_norm": 0.54819256067276, "learning_rate": 2.87307912624582e-06, "loss": 0.0425, "step": 89700 }, { "grad_norm": 0.28297126293182373, "learning_rate": 2.867557497133977e-06, "loss": 0.0348, "step": 89710 }, { "grad_norm": 0.5042292475700378, "learning_rate": 2.862041022355194e-06, "loss": 0.0269, "step": 89720 }, { "grad_norm": 0.8108548521995544, "learning_rate": 2.8565297025127226e-06, "loss": 0.0358, "step": 89730 }, { "grad_norm": 0.3788870871067047, "learning_rate": 2.8510235382092853e-06, "loss": 0.0256, "step": 89740 }, { "grad_norm": 0.43291348218917847, "learning_rate": 2.845522530047029e-06, "loss": 0.0274, "step": 89750 }, { "grad_norm": 0.615673840045929, "learning_rate": 2.8400266786275387e-06, "loss": 0.0287, "step": 89760 }, { "grad_norm": 1.030240535736084, "learning_rate": 2.834535984551817e-06, "loss": 0.0352, "step": 89770 }, { "grad_norm": 0.7719262838363647, "learning_rate": 2.8290504484203397e-06, "loss": 0.0396, "step": 89780 }, { "grad_norm": 0.35403701663017273, "learning_rate": 2.8235700708329705e-06, "loss": 0.029, "step": 89790 }, { "grad_norm": 0.5829920172691345, "learning_rate": 2.818094852389058e-06, "loss": 0.0234, "step": 89800 }, { "grad_norm": 0.3009374439716339, "learning_rate": 2.8126247936873516e-06, "loss": 0.0237, "step": 89810 }, { "grad_norm": 0.39370307326316833, "learning_rate": 2.8071598953260614e-06, "loss": 0.0284, "step": 89820 }, { "grad_norm": 0.40026044845581055, "learning_rate": 2.8017001579028035e-06, "loss": 0.0264, "step": 89830 }, { "grad_norm": 0.3012928068637848, "learning_rate": 2.7962455820146617e-06, "loss": 0.0352, "step": 89840 }, { "grad_norm": 0.4523451328277588, "learning_rate": 2.7907961682581253e-06, "loss": 0.0274, "step": 89850 }, { "grad_norm": 0.31641414761543274, "learning_rate": 2.7853519172291453e-06, "loss": 0.0299, "step": 89860 }, { "grad_norm": 0.4003739058971405, "learning_rate": 2.7799128295230904e-06, "loss": 0.0289, "step": 89870 }, { "grad_norm": 0.2226843684911728, "learning_rate": 2.774478905734784e-06, "loss": 0.0293, "step": 89880 }, { "grad_norm": 0.5116366147994995, "learning_rate": 2.7690501464584516e-06, "loss": 0.0305, "step": 89890 }, { "grad_norm": 1.7826313972473145, "learning_rate": 2.763626552287796e-06, "loss": 0.031, "step": 89900 }, { "grad_norm": 0.41332152485847473, "learning_rate": 2.7582081238159083e-06, "loss": 0.0297, "step": 89910 }, { "grad_norm": 0.23539772629737854, "learning_rate": 2.7527948616353548e-06, "loss": 0.0312, "step": 89920 }, { "grad_norm": 0.2780146598815918, "learning_rate": 2.7473867663381226e-06, "loss": 0.0246, "step": 89930 }, { "grad_norm": 0.2260240763425827, "learning_rate": 2.7419838385156383e-06, "loss": 0.0254, "step": 89940 }, { "grad_norm": 0.6841291189193726, "learning_rate": 2.7365860787587407e-06, "loss": 0.0332, "step": 89950 }, { "grad_norm": 0.8461763858795166, "learning_rate": 2.7311934876577295e-06, "loss": 0.0364, "step": 89960 }, { "grad_norm": 0.347769170999527, "learning_rate": 2.725806065802339e-06, "loss": 0.0265, "step": 89970 }, { "grad_norm": 0.3333933353424072, "learning_rate": 2.720423813781714e-06, "loss": 0.0364, "step": 89980 }, { "grad_norm": 0.4423607885837555, "learning_rate": 2.715046732184462e-06, "loss": 0.0252, "step": 89990 }, { "grad_norm": 0.46521836519241333, "learning_rate": 2.7096748215986013e-06, "loss": 0.0242, "step": 90000 }, { "grad_norm": 1.0583930015563965, "learning_rate": 2.704308082611612e-06, "loss": 0.034, "step": 90010 }, { "grad_norm": 0.37659746408462524, "learning_rate": 2.698946515810374e-06, "loss": 0.0273, "step": 90020 }, { "grad_norm": 0.8651430010795593, "learning_rate": 2.6935901217812367e-06, "loss": 0.0332, "step": 90030 }, { "grad_norm": 0.4129065275192261, "learning_rate": 2.688238901109952e-06, "loss": 0.0279, "step": 90040 }, { "grad_norm": 0.5607964992523193, "learning_rate": 2.68289285438173e-06, "loss": 0.0394, "step": 90050 }, { "grad_norm": 0.29241615533828735, "learning_rate": 2.6775519821812033e-06, "loss": 0.0291, "step": 90060 }, { "grad_norm": 0.34927311539649963, "learning_rate": 2.672216285092449e-06, "loss": 0.0332, "step": 90070 }, { "grad_norm": 0.4094199240207672, "learning_rate": 2.666885763698962e-06, "loss": 0.0248, "step": 90080 }, { "grad_norm": 0.5893336534500122, "learning_rate": 2.661560418583692e-06, "loss": 0.0399, "step": 90090 }, { "grad_norm": 0.30548250675201416, "learning_rate": 2.656240250328995e-06, "loss": 0.0285, "step": 90100 }, { "grad_norm": 0.4228748679161072, "learning_rate": 2.650925259516679e-06, "loss": 0.0387, "step": 90110 }, { "grad_norm": 0.18173861503601074, "learning_rate": 2.645615446727995e-06, "loss": 0.0213, "step": 90120 }, { "grad_norm": 0.5803213119506836, "learning_rate": 2.6403108125436115e-06, "loss": 0.0232, "step": 90130 }, { "grad_norm": 0.3239869475364685, "learning_rate": 2.6350113575436264e-06, "loss": 0.0243, "step": 90140 }, { "grad_norm": 0.43107691407203674, "learning_rate": 2.629717082307598e-06, "loss": 0.0251, "step": 90150 }, { "grad_norm": 0.3435346782207489, "learning_rate": 2.6244279874144793e-06, "loss": 0.0256, "step": 90160 }, { "grad_norm": 0.17019982635974884, "learning_rate": 2.6191440734426862e-06, "loss": 0.0228, "step": 90170 }, { "grad_norm": 0.5230004787445068, "learning_rate": 2.613865340970062e-06, "loss": 0.029, "step": 90180 }, { "grad_norm": 0.4473726749420166, "learning_rate": 2.608591790573889e-06, "loss": 0.0313, "step": 90190 }, { "grad_norm": 0.4365460276603699, "learning_rate": 2.603323422830861e-06, "loss": 0.0306, "step": 90200 }, { "grad_norm": 0.3748740255832672, "learning_rate": 2.5980602383171183e-06, "loss": 0.035, "step": 90210 }, { "grad_norm": 0.4364773631095886, "learning_rate": 2.5928022376082496e-06, "loss": 0.0333, "step": 90220 }, { "grad_norm": 0.6970659494400024, "learning_rate": 2.5875494212792505e-06, "loss": 0.0248, "step": 90230 }, { "grad_norm": 0.34006473422050476, "learning_rate": 2.5823017899045565e-06, "loss": 0.0305, "step": 90240 }, { "grad_norm": 0.24562589824199677, "learning_rate": 2.577059344058047e-06, "loss": 0.0349, "step": 90250 }, { "grad_norm": 0.4125911593437195, "learning_rate": 2.5718220843130415e-06, "loss": 0.0252, "step": 90260 }, { "grad_norm": 0.4616628587245941, "learning_rate": 2.5665900112422535e-06, "loss": 0.0316, "step": 90270 }, { "grad_norm": 0.6356438398361206, "learning_rate": 2.561363125417876e-06, "loss": 0.0425, "step": 90280 }, { "grad_norm": 0.35308167338371277, "learning_rate": 2.556141427411496e-06, "loss": 0.0212, "step": 90290 }, { "grad_norm": 0.2633346915245056, "learning_rate": 2.550924917794156e-06, "loss": 0.0291, "step": 90300 }, { "grad_norm": 0.1997852474451065, "learning_rate": 2.545713597136329e-06, "loss": 0.0223, "step": 90310 }, { "grad_norm": 0.16400110721588135, "learning_rate": 2.5405074660079298e-06, "loss": 0.0244, "step": 90320 }, { "grad_norm": 0.30131885409355164, "learning_rate": 2.5353065249782647e-06, "loss": 0.0288, "step": 90330 }, { "grad_norm": 0.3374761641025543, "learning_rate": 2.530110774616129e-06, "loss": 0.0294, "step": 90340 }, { "grad_norm": 0.35216161608695984, "learning_rate": 2.524920215489701e-06, "loss": 0.0295, "step": 90350 }, { "grad_norm": 0.24316322803497314, "learning_rate": 2.519734848166627e-06, "loss": 0.0272, "step": 90360 }, { "grad_norm": 0.2798274755477905, "learning_rate": 2.514554673213948e-06, "loss": 0.0267, "step": 90370 }, { "grad_norm": 0.3168470859527588, "learning_rate": 2.5093796911981936e-06, "loss": 0.0234, "step": 90380 }, { "grad_norm": 0.2980189621448517, "learning_rate": 2.5042099026852672e-06, "loss": 0.0359, "step": 90390 }, { "grad_norm": 0.33856528997421265, "learning_rate": 2.499045308240544e-06, "loss": 0.0256, "step": 90400 }, { "grad_norm": 0.22157259285449982, "learning_rate": 2.4938859084288003e-06, "loss": 0.0215, "step": 90410 }, { "grad_norm": 0.31892406940460205, "learning_rate": 2.4887317038142734e-06, "loss": 0.0295, "step": 90420 }, { "grad_norm": 0.3255150616168976, "learning_rate": 2.4835826949606124e-06, "loss": 0.0264, "step": 90430 }, { "grad_norm": 0.6179909110069275, "learning_rate": 2.4784388824309167e-06, "loss": 0.0328, "step": 90440 }, { "grad_norm": 0.3807172179222107, "learning_rate": 2.473300266787687e-06, "loss": 0.0283, "step": 90450 }, { "grad_norm": 0.5125951766967773, "learning_rate": 2.46816684859289e-06, "loss": 0.0339, "step": 90460 }, { "grad_norm": 0.3594886064529419, "learning_rate": 2.4630386284079043e-06, "loss": 0.0243, "step": 90470 }, { "grad_norm": 0.4377657473087311, "learning_rate": 2.4579156067935427e-06, "loss": 0.0258, "step": 90480 }, { "grad_norm": 0.309745728969574, "learning_rate": 2.452797784310046e-06, "loss": 0.0322, "step": 90490 }, { "grad_norm": 0.5587414503097534, "learning_rate": 2.447685161517105e-06, "loss": 0.0295, "step": 90500 }, { "grad_norm": 0.40330395102500916, "learning_rate": 2.4425777389738224e-06, "loss": 0.0222, "step": 90510 }, { "grad_norm": 0.5818418860435486, "learning_rate": 2.43747551723873e-06, "loss": 0.0224, "step": 90520 }, { "grad_norm": 0.25498801469802856, "learning_rate": 2.4323784968698136e-06, "loss": 0.0429, "step": 90530 }, { "grad_norm": 0.670996367931366, "learning_rate": 2.4272866784244565e-06, "loss": 0.0315, "step": 90540 }, { "grad_norm": 0.3243281841278076, "learning_rate": 2.422200062459518e-06, "loss": 0.0258, "step": 90550 }, { "grad_norm": 0.9804198145866394, "learning_rate": 2.41711864953123e-06, "loss": 0.036, "step": 90560 }, { "grad_norm": 1.1701427698135376, "learning_rate": 2.412042440195322e-06, "loss": 0.037, "step": 90570 }, { "grad_norm": 0.7455783486366272, "learning_rate": 2.4069714350068993e-06, "loss": 0.032, "step": 90580 }, { "grad_norm": 0.42721763253211975, "learning_rate": 2.4019056345205294e-06, "loss": 0.0256, "step": 90590 }, { "grad_norm": 0.47447454929351807, "learning_rate": 2.396845039290191e-06, "loss": 0.0286, "step": 90600 }, { "grad_norm": 0.38714689016342163, "learning_rate": 2.3917896498693194e-06, "loss": 0.0356, "step": 90610 }, { "grad_norm": 0.4826337993144989, "learning_rate": 2.3867394668107335e-06, "loss": 0.0331, "step": 90620 }, { "grad_norm": 0.1533401906490326, "learning_rate": 2.3816944906667527e-06, "loss": 0.0262, "step": 90630 }, { "grad_norm": 0.6517766714096069, "learning_rate": 2.3766547219890634e-06, "loss": 0.0282, "step": 90640 }, { "grad_norm": 0.4038297235965729, "learning_rate": 2.371620161328819e-06, "loss": 0.0259, "step": 90650 }, { "grad_norm": 0.49857714772224426, "learning_rate": 2.366590809236574e-06, "loss": 0.0311, "step": 90660 }, { "grad_norm": 0.5434819459915161, "learning_rate": 2.3615666662623494e-06, "loss": 0.029, "step": 90670 }, { "grad_norm": 0.6533682942390442, "learning_rate": 2.3565477329555664e-06, "loss": 0.0279, "step": 90680 }, { "grad_norm": 0.5791686773300171, "learning_rate": 2.3515340098650975e-06, "loss": 0.0337, "step": 90690 }, { "grad_norm": 0.7673231363296509, "learning_rate": 2.3465254975392258e-06, "loss": 0.0396, "step": 90700 }, { "grad_norm": 0.2320631891489029, "learning_rate": 2.3415221965256805e-06, "loss": 0.041, "step": 90710 }, { "grad_norm": 0.4175221621990204, "learning_rate": 2.3365241073716237e-06, "loss": 0.0299, "step": 90720 }, { "grad_norm": 0.24884770810604095, "learning_rate": 2.3315312306236183e-06, "loss": 0.0265, "step": 90730 }, { "grad_norm": 0.45786339044570923, "learning_rate": 2.326543566827699e-06, "loss": 0.0339, "step": 90740 }, { "grad_norm": 0.26123541593551636, "learning_rate": 2.3215611165292862e-06, "loss": 0.0313, "step": 90750 }, { "grad_norm": 0.38258346915245056, "learning_rate": 2.3165838802732765e-06, "loss": 0.0299, "step": 90760 }, { "grad_norm": 0.2803923487663269, "learning_rate": 2.311611858603957e-06, "loss": 0.0262, "step": 90770 }, { "grad_norm": 0.6804590821266174, "learning_rate": 2.3066450520650763e-06, "loss": 0.0334, "step": 90780 }, { "grad_norm": 0.2092052847146988, "learning_rate": 2.3016834611997773e-06, "loss": 0.0323, "step": 90790 }, { "grad_norm": 0.36980101466178894, "learning_rate": 2.296727086550665e-06, "loss": 0.0349, "step": 90800 }, { "grad_norm": 0.35999685525894165, "learning_rate": 2.2917759286597507e-06, "loss": 0.0343, "step": 90810 }, { "grad_norm": 0.31798255443573, "learning_rate": 2.2868299880685006e-06, "loss": 0.0298, "step": 90820 }, { "grad_norm": 0.4310190677642822, "learning_rate": 2.2818892653177827e-06, "loss": 0.0257, "step": 90830 }, { "grad_norm": 0.7721925377845764, "learning_rate": 2.2769537609479196e-06, "loss": 0.0263, "step": 90840 }, { "grad_norm": 0.49583888053894043, "learning_rate": 2.2720234754986357e-06, "loss": 0.0316, "step": 90850 }, { "grad_norm": 0.17080505192279816, "learning_rate": 2.2670984095091108e-06, "loss": 0.0259, "step": 90860 }, { "grad_norm": 0.3112761378288269, "learning_rate": 2.262178563517936e-06, "loss": 0.0332, "step": 90870 }, { "grad_norm": 0.3029824495315552, "learning_rate": 2.257263938063137e-06, "loss": 0.0325, "step": 90880 }, { "grad_norm": 0.9735933542251587, "learning_rate": 2.2523545336821715e-06, "loss": 0.0398, "step": 90890 }, { "grad_norm": 0.7781155109405518, "learning_rate": 2.247450350911939e-06, "loss": 0.0395, "step": 90900 }, { "grad_norm": 0.27507245540618896, "learning_rate": 2.2425513902887328e-06, "loss": 0.0232, "step": 90910 }, { "grad_norm": 0.40555045008659363, "learning_rate": 2.2376576523483015e-06, "loss": 0.0219, "step": 90920 }, { "grad_norm": 0.5369182229042053, "learning_rate": 2.232769137625829e-06, "loss": 0.0288, "step": 90930 }, { "grad_norm": 0.782779335975647, "learning_rate": 2.227885846655903e-06, "loss": 0.0377, "step": 90940 }, { "grad_norm": 0.5580698251724243, "learning_rate": 2.2230077799725523e-06, "loss": 0.0288, "step": 90950 }, { "grad_norm": 0.24632057547569275, "learning_rate": 2.2181349381092396e-06, "loss": 0.0319, "step": 90960 }, { "grad_norm": 0.29090365767478943, "learning_rate": 2.2132673215988552e-06, "loss": 0.0266, "step": 90970 }, { "grad_norm": 0.29262715578079224, "learning_rate": 2.2084049309737065e-06, "loss": 0.0278, "step": 90980 }, { "grad_norm": 0.43494173884391785, "learning_rate": 2.2035477667655513e-06, "loss": 0.0248, "step": 90990 }, { "grad_norm": 0.6024148464202881, "learning_rate": 2.1986958295055315e-06, "loss": 0.031, "step": 91000 }, { "grad_norm": 0.6072601675987244, "learning_rate": 2.193849119724284e-06, "loss": 0.034, "step": 91010 }, { "grad_norm": 0.4433021545410156, "learning_rate": 2.1890076379518175e-06, "loss": 0.0238, "step": 91020 }, { "grad_norm": 0.5097008943557739, "learning_rate": 2.1841713847175925e-06, "loss": 0.0284, "step": 91030 }, { "grad_norm": 0.3487078547477722, "learning_rate": 2.1793403605504968e-06, "loss": 0.0334, "step": 91040 }, { "grad_norm": 0.45103028416633606, "learning_rate": 2.1745145659788413e-06, "loss": 0.033, "step": 91050 }, { "grad_norm": 0.764049768447876, "learning_rate": 2.1696940015303645e-06, "loss": 0.0256, "step": 91060 }, { "grad_norm": 1.2231370210647583, "learning_rate": 2.1648786677322397e-06, "loss": 0.0406, "step": 91070 }, { "grad_norm": 0.3247630000114441, "learning_rate": 2.1600685651110673e-06, "loss": 0.0279, "step": 91080 }, { "grad_norm": 0.3473272919654846, "learning_rate": 2.1552636941928717e-06, "loss": 0.0289, "step": 91090 }, { "grad_norm": 0.44263696670532227, "learning_rate": 2.1504640555031043e-06, "loss": 0.0229, "step": 91100 }, { "grad_norm": 0.45834726095199585, "learning_rate": 2.1456696495666506e-06, "loss": 0.0247, "step": 91110 }, { "grad_norm": 0.3566795885562897, "learning_rate": 2.1408804769078084e-06, "loss": 0.0394, "step": 91120 }, { "grad_norm": 1.2509217262268066, "learning_rate": 2.136096538050325e-06, "loss": 0.0276, "step": 91130 }, { "grad_norm": 0.6570138931274414, "learning_rate": 2.13131783351736e-06, "loss": 0.0355, "step": 91140 }, { "grad_norm": 0.2661312222480774, "learning_rate": 2.126544363831512e-06, "loss": 0.0325, "step": 91150 }, { "grad_norm": 0.3698805272579193, "learning_rate": 2.121776129514791e-06, "loss": 0.0244, "step": 91160 }, { "grad_norm": 0.18731601536273956, "learning_rate": 2.1170131310886523e-06, "loss": 0.0227, "step": 91170 }, { "grad_norm": 0.5808999538421631, "learning_rate": 2.112255369073968e-06, "loss": 0.0346, "step": 91180 }, { "grad_norm": 0.49217867851257324, "learning_rate": 2.1075028439910328e-06, "loss": 0.0241, "step": 91190 }, { "grad_norm": 0.6284044981002808, "learning_rate": 2.1027555563595813e-06, "loss": 0.0314, "step": 91200 }, { "grad_norm": 0.3177367150783539, "learning_rate": 2.0980135066987706e-06, "loss": 0.0243, "step": 91210 }, { "grad_norm": 0.3301873803138733, "learning_rate": 2.093276695527191e-06, "loss": 0.041, "step": 91220 }, { "grad_norm": 0.28532373905181885, "learning_rate": 2.0885451233628396e-06, "loss": 0.0323, "step": 91230 }, { "grad_norm": 0.9396454691886902, "learning_rate": 2.0838187907231633e-06, "loss": 0.037, "step": 91240 }, { "grad_norm": 0.237830251455307, "learning_rate": 2.0790976981250153e-06, "loss": 0.0282, "step": 91250 }, { "grad_norm": 0.4947231113910675, "learning_rate": 2.0743818460847e-06, "loss": 0.0278, "step": 91260 }, { "grad_norm": 0.4906105101108551, "learning_rate": 2.069671235117926e-06, "loss": 0.0368, "step": 91270 }, { "grad_norm": 0.47871893644332886, "learning_rate": 2.0649658657398484e-06, "loss": 0.0298, "step": 91280 }, { "grad_norm": 0.546374499797821, "learning_rate": 2.060265738465034e-06, "loss": 0.0255, "step": 91290 }, { "grad_norm": 0.7169852256774902, "learning_rate": 2.055570853807487e-06, "loss": 0.0217, "step": 91300 }, { "grad_norm": 0.418161004781723, "learning_rate": 2.0508812122806196e-06, "loss": 0.0265, "step": 91310 }, { "grad_norm": 0.38362592458724976, "learning_rate": 2.046196814397294e-06, "loss": 0.0263, "step": 91320 }, { "grad_norm": 0.5506232976913452, "learning_rate": 2.0415176606697896e-06, "loss": 0.0313, "step": 91330 }, { "grad_norm": 0.450033038854599, "learning_rate": 2.0368437516098126e-06, "loss": 0.0291, "step": 91340 }, { "grad_norm": 0.5787060856819153, "learning_rate": 2.032175087728483e-06, "loss": 0.0284, "step": 91350 }, { "grad_norm": 0.38644906878471375, "learning_rate": 2.027511669536375e-06, "loss": 0.0328, "step": 91360 }, { "grad_norm": 0.2939862310886383, "learning_rate": 2.0228534975434643e-06, "loss": 0.026, "step": 91370 }, { "grad_norm": 0.24797135591506958, "learning_rate": 2.0182005722591548e-06, "loss": 0.0318, "step": 91380 }, { "grad_norm": 0.3491809070110321, "learning_rate": 2.0135528941922944e-06, "loss": 0.028, "step": 91390 }, { "grad_norm": 0.29304853081703186, "learning_rate": 2.008910463851149e-06, "loss": 0.024, "step": 91400 }, { "grad_norm": 0.3805902898311615, "learning_rate": 2.0042732817433963e-06, "loss": 0.0263, "step": 91410 }, { "grad_norm": 0.4812715947628021, "learning_rate": 1.999641348376158e-06, "loss": 0.0224, "step": 91420 }, { "grad_norm": 0.30363374948501587, "learning_rate": 1.9950146642559787e-06, "loss": 0.0337, "step": 91430 }, { "grad_norm": 0.9210590124130249, "learning_rate": 1.990393229888815e-06, "loss": 0.0282, "step": 91440 }, { "grad_norm": 0.3315799832344055, "learning_rate": 1.9857770457800684e-06, "loss": 0.0273, "step": 91450 }, { "grad_norm": 0.3193003833293915, "learning_rate": 1.9811661124345517e-06, "loss": 0.0273, "step": 91460 }, { "grad_norm": 0.3873213827610016, "learning_rate": 1.976560430356522e-06, "loss": 0.024, "step": 91470 }, { "grad_norm": 0.6359875202178955, "learning_rate": 1.9719600000496383e-06, "loss": 0.0379, "step": 91480 }, { "grad_norm": 0.4158249795436859, "learning_rate": 1.967364822017004e-06, "loss": 0.0341, "step": 91490 }, { "grad_norm": 0.2418302744626999, "learning_rate": 1.962774896761127e-06, "loss": 0.0254, "step": 91500 }, { "grad_norm": 0.29848194122314453, "learning_rate": 1.958190224783968e-06, "loss": 0.0202, "step": 91510 }, { "grad_norm": 0.3504873812198639, "learning_rate": 1.9536108065868975e-06, "loss": 0.0311, "step": 91520 }, { "grad_norm": 0.7369195222854614, "learning_rate": 1.949036642670715e-06, "loss": 0.0335, "step": 91530 }, { "grad_norm": 0.3702034652233124, "learning_rate": 1.9444677335356377e-06, "loss": 0.0308, "step": 91540 }, { "grad_norm": 0.7045227289199829, "learning_rate": 1.939904079681326e-06, "loss": 0.0379, "step": 91550 }, { "grad_norm": 0.24937143921852112, "learning_rate": 1.935345681606837e-06, "loss": 0.0242, "step": 91560 }, { "grad_norm": 0.4144960045814514, "learning_rate": 1.930792539810694e-06, "loss": 0.0251, "step": 91570 }, { "grad_norm": 0.33566799759864807, "learning_rate": 1.9262446547907877e-06, "loss": 0.0277, "step": 91580 }, { "grad_norm": 0.47730445861816406, "learning_rate": 1.921702027044503e-06, "loss": 0.0319, "step": 91590 }, { "grad_norm": 0.4929884076118469, "learning_rate": 1.917164657068593e-06, "loss": 0.0373, "step": 91600 }, { "grad_norm": 0.6728560328483582, "learning_rate": 1.912632545359272e-06, "loss": 0.0309, "step": 91610 }, { "grad_norm": 0.1835206001996994, "learning_rate": 1.9081056924121488e-06, "loss": 0.0255, "step": 91620 }, { "grad_norm": 0.23693184554576874, "learning_rate": 1.9035840987222887e-06, "loss": 0.0292, "step": 91630 }, { "grad_norm": 0.1686592847108841, "learning_rate": 1.8990677647841515e-06, "loss": 0.0286, "step": 91640 }, { "grad_norm": 0.4275412857532501, "learning_rate": 1.8945566910916534e-06, "loss": 0.0333, "step": 91650 }, { "grad_norm": 0.40649911761283875, "learning_rate": 1.8900508781381054e-06, "loss": 0.0261, "step": 91660 }, { "grad_norm": 0.20158132910728455, "learning_rate": 1.8855503264162633e-06, "loss": 0.0195, "step": 91670 }, { "grad_norm": 0.43878114223480225, "learning_rate": 1.8810550364182998e-06, "loss": 0.027, "step": 91680 }, { "grad_norm": 0.5180642008781433, "learning_rate": 1.8765650086358112e-06, "loss": 0.0289, "step": 91690 }, { "grad_norm": 0.7243894934654236, "learning_rate": 1.8720802435598151e-06, "loss": 0.0352, "step": 91700 }, { "grad_norm": 0.3749637007713318, "learning_rate": 1.8676007416807694e-06, "loss": 0.0307, "step": 91710 }, { "grad_norm": 0.6274694800376892, "learning_rate": 1.8631265034885437e-06, "loss": 0.035, "step": 91720 }, { "grad_norm": 0.4093412756919861, "learning_rate": 1.858657529472424e-06, "loss": 0.024, "step": 91730 }, { "grad_norm": 0.3269785940647125, "learning_rate": 1.8541938201211473e-06, "loss": 0.0232, "step": 91740 }, { "grad_norm": 0.5871074199676514, "learning_rate": 1.8497353759228398e-06, "loss": 0.0232, "step": 91750 }, { "grad_norm": 0.4611594080924988, "learning_rate": 1.8452821973650835e-06, "loss": 0.0242, "step": 91760 }, { "grad_norm": 0.20484836399555206, "learning_rate": 1.840834284934856e-06, "loss": 0.0259, "step": 91770 }, { "grad_norm": 0.2795460522174835, "learning_rate": 1.8363916391185954e-06, "loss": 0.0277, "step": 91780 }, { "grad_norm": 0.2840421795845032, "learning_rate": 1.8319542604021245e-06, "loss": 0.0356, "step": 91790 }, { "grad_norm": 0.41457894444465637, "learning_rate": 1.8275221492707273e-06, "loss": 0.0252, "step": 91800 }, { "grad_norm": 0.30611348152160645, "learning_rate": 1.8230953062090717e-06, "loss": 0.0355, "step": 91810 }, { "grad_norm": 0.3278142511844635, "learning_rate": 1.8186737317012926e-06, "loss": 0.0213, "step": 91820 }, { "grad_norm": 0.29676905274391174, "learning_rate": 1.8142574262308977e-06, "loss": 0.03, "step": 91830 }, { "grad_norm": 0.36127981543540955, "learning_rate": 1.8098463902808727e-06, "loss": 0.0314, "step": 91840 }, { "grad_norm": 0.91846764087677, "learning_rate": 1.805440624333593e-06, "loss": 0.0209, "step": 91850 }, { "grad_norm": 0.7816702723503113, "learning_rate": 1.8010401288708734e-06, "loss": 0.0344, "step": 91860 }, { "grad_norm": 0.4010694622993469, "learning_rate": 1.7966449043739343e-06, "loss": 0.0325, "step": 91870 }, { "grad_norm": 0.8078552484512329, "learning_rate": 1.79225495132343e-06, "loss": 0.0264, "step": 91880 }, { "grad_norm": 0.5646470189094543, "learning_rate": 1.7878702701994487e-06, "loss": 0.0348, "step": 91890 }, { "grad_norm": 0.32369741797447205, "learning_rate": 1.783490861481496e-06, "loss": 0.0217, "step": 91900 }, { "grad_norm": 0.2223655730485916, "learning_rate": 1.7791167256484832e-06, "loss": 0.024, "step": 91910 }, { "grad_norm": 1.031468391418457, "learning_rate": 1.774747863178766e-06, "loss": 0.0406, "step": 91920 }, { "grad_norm": 0.3177528977394104, "learning_rate": 1.7703842745501186e-06, "loss": 0.0266, "step": 91930 }, { "grad_norm": 0.4491862952709198, "learning_rate": 1.7660259602397366e-06, "loss": 0.0337, "step": 91940 }, { "grad_norm": 0.34255412220954895, "learning_rate": 1.7616729207242388e-06, "loss": 0.029, "step": 91950 }, { "grad_norm": 0.23611217737197876, "learning_rate": 1.7573251564796556e-06, "loss": 0.0327, "step": 91960 }, { "grad_norm": 0.501683235168457, "learning_rate": 1.7529826679814676e-06, "loss": 0.0365, "step": 91970 }, { "grad_norm": 0.3923279345035553, "learning_rate": 1.7486454557045562e-06, "loss": 0.033, "step": 91980 }, { "grad_norm": 0.3338385224342346, "learning_rate": 1.7443135201232364e-06, "loss": 0.0248, "step": 91990 }, { "grad_norm": 0.5196227431297302, "learning_rate": 1.7399868617112347e-06, "loss": 0.0333, "step": 92000 }, { "grad_norm": 0.7376461029052734, "learning_rate": 1.7356654809417171e-06, "loss": 0.0257, "step": 92010 }, { "grad_norm": 0.28941991925239563, "learning_rate": 1.7313493782872448e-06, "loss": 0.0284, "step": 92020 }, { "grad_norm": 0.5811251997947693, "learning_rate": 1.7270385542198453e-06, "loss": 0.0266, "step": 92030 }, { "grad_norm": 0.39431872963905334, "learning_rate": 1.7227330092109307e-06, "loss": 0.0292, "step": 92040 }, { "grad_norm": 0.27642345428466797, "learning_rate": 1.718432743731352e-06, "loss": 0.0241, "step": 92050 }, { "grad_norm": 0.3619566559791565, "learning_rate": 1.7141377582513718e-06, "loss": 0.0275, "step": 92060 }, { "grad_norm": 0.27322667837142944, "learning_rate": 1.709848053240698e-06, "loss": 0.0335, "step": 92070 }, { "grad_norm": 0.420322984457016, "learning_rate": 1.7055636291684328e-06, "loss": 0.0196, "step": 92080 }, { "grad_norm": 0.3023239076137543, "learning_rate": 1.7012844865031185e-06, "loss": 0.028, "step": 92090 }, { "grad_norm": 0.31464526057243347, "learning_rate": 1.6970106257127138e-06, "loss": 0.0368, "step": 92100 }, { "grad_norm": 0.3447871208190918, "learning_rate": 1.6927420472646117e-06, "loss": 0.0239, "step": 92110 }, { "grad_norm": 0.3081853687763214, "learning_rate": 1.6884787516256106e-06, "loss": 0.0319, "step": 92120 }, { "grad_norm": 0.5300484895706177, "learning_rate": 1.6842207392619325e-06, "loss": 0.0336, "step": 92130 }, { "grad_norm": 0.374571830034256, "learning_rate": 1.679968010639238e-06, "loss": 0.0332, "step": 92140 }, { "grad_norm": 0.3300679326057434, "learning_rate": 1.6757205662225829e-06, "loss": 0.0372, "step": 92150 }, { "grad_norm": 0.35235196352005005, "learning_rate": 1.6714784064764844e-06, "loss": 0.0297, "step": 92160 }, { "grad_norm": 0.33736979961395264, "learning_rate": 1.6672415318648383e-06, "loss": 0.0264, "step": 92170 }, { "grad_norm": 0.5149563550949097, "learning_rate": 1.6630099428510016e-06, "loss": 0.0346, "step": 92180 }, { "grad_norm": 0.2995433211326599, "learning_rate": 1.6587836398977153e-06, "loss": 0.0383, "step": 92190 }, { "grad_norm": 0.24179694056510925, "learning_rate": 1.6545626234671762e-06, "loss": 0.0277, "step": 92200 }, { "grad_norm": 0.4149988293647766, "learning_rate": 1.650346894020971e-06, "loss": 0.0284, "step": 92210 }, { "grad_norm": 0.2345319390296936, "learning_rate": 1.6461364520201472e-06, "loss": 0.0284, "step": 92220 }, { "grad_norm": 0.5749078392982483, "learning_rate": 1.6419312979251366e-06, "loss": 0.0396, "step": 92230 }, { "grad_norm": 0.31902769207954407, "learning_rate": 1.6377314321958214e-06, "loss": 0.038, "step": 92240 }, { "grad_norm": 0.6037968993186951, "learning_rate": 1.6335368552914787e-06, "loss": 0.0308, "step": 92250 }, { "grad_norm": 0.40219786763191223, "learning_rate": 1.6293475676708358e-06, "loss": 0.0275, "step": 92260 }, { "grad_norm": 0.8483873605728149, "learning_rate": 1.6251635697920097e-06, "loss": 0.0326, "step": 92270 }, { "grad_norm": 0.5573690533638, "learning_rate": 1.6209848621125624e-06, "loss": 0.0343, "step": 92280 }, { "grad_norm": 0.2262510359287262, "learning_rate": 1.6168114450894778e-06, "loss": 0.031, "step": 92290 }, { "grad_norm": 0.4828763008117676, "learning_rate": 1.612643319179158e-06, "loss": 0.0311, "step": 92300 }, { "grad_norm": 0.6215964555740356, "learning_rate": 1.6084804848374046e-06, "loss": 0.0306, "step": 92310 }, { "grad_norm": 0.254756897687912, "learning_rate": 1.6043229425194817e-06, "loss": 0.0262, "step": 92320 }, { "grad_norm": 0.3121913969516754, "learning_rate": 1.6001706926800308e-06, "loss": 0.0232, "step": 92330 }, { "grad_norm": 0.5032939910888672, "learning_rate": 1.5960237357731444e-06, "loss": 0.0285, "step": 92340 }, { "grad_norm": 0.5404907464981079, "learning_rate": 1.5918820722523264e-06, "loss": 0.0332, "step": 92350 }, { "grad_norm": 0.2942652404308319, "learning_rate": 1.587745702570509e-06, "loss": 0.0366, "step": 92360 }, { "grad_norm": 0.3260817229747772, "learning_rate": 1.5836146271800356e-06, "loss": 0.0283, "step": 92370 }, { "grad_norm": 0.2614692747592926, "learning_rate": 1.5794888465326728e-06, "loss": 0.0255, "step": 92380 }, { "grad_norm": 0.18823876976966858, "learning_rate": 1.5753683610796155e-06, "loss": 0.0314, "step": 92390 }, { "grad_norm": 0.3685716688632965, "learning_rate": 1.5712531712714528e-06, "loss": 0.031, "step": 92400 }, { "grad_norm": 0.2918468117713928, "learning_rate": 1.5671432775582474e-06, "loss": 0.0354, "step": 92410 }, { "grad_norm": 0.4621537923812866, "learning_rate": 1.563038680389428e-06, "loss": 0.0286, "step": 92420 }, { "grad_norm": 0.47738802433013916, "learning_rate": 1.5589393802138808e-06, "loss": 0.0274, "step": 92430 }, { "grad_norm": 0.34253478050231934, "learning_rate": 1.554845377479891e-06, "loss": 0.0316, "step": 92440 }, { "grad_norm": 0.47498220205307007, "learning_rate": 1.550756672635173e-06, "loss": 0.0343, "step": 92450 }, { "grad_norm": 0.4584974944591522, "learning_rate": 1.5466732661268635e-06, "loss": 0.0256, "step": 92460 }, { "grad_norm": 0.6200719475746155, "learning_rate": 1.5425951584015163e-06, "loss": 0.0249, "step": 92470 }, { "grad_norm": 0.21699944138526917, "learning_rate": 1.5385223499051026e-06, "loss": 0.0287, "step": 92480 }, { "grad_norm": 0.7088163495063782, "learning_rate": 1.5344548410830328e-06, "loss": 0.0255, "step": 92490 }, { "grad_norm": 0.3228924572467804, "learning_rate": 1.5303926323801121e-06, "loss": 0.0256, "step": 92500 }, { "grad_norm": 0.4514947235584259, "learning_rate": 1.5263357242405795e-06, "loss": 0.0213, "step": 92510 }, { "grad_norm": 0.38484784960746765, "learning_rate": 1.5222841171080915e-06, "loss": 0.0303, "step": 92520 }, { "grad_norm": 0.2830701172351837, "learning_rate": 1.5182378114257268e-06, "loss": 0.0288, "step": 92530 }, { "grad_norm": 0.2815258502960205, "learning_rate": 1.5141968076359869e-06, "loss": 0.026, "step": 92540 }, { "grad_norm": 0.32158929109573364, "learning_rate": 1.5101611061807908e-06, "loss": 0.028, "step": 92550 }, { "grad_norm": 0.7300203442573547, "learning_rate": 1.506130707501463e-06, "loss": 0.0224, "step": 92560 }, { "grad_norm": 0.5226947665214539, "learning_rate": 1.5021056120387734e-06, "loss": 0.038, "step": 92570 }, { "grad_norm": 0.3669759929180145, "learning_rate": 1.4980858202329084e-06, "loss": 0.0246, "step": 92580 }, { "grad_norm": 0.27016186714172363, "learning_rate": 1.4940713325234445e-06, "loss": 0.0269, "step": 92590 }, { "grad_norm": 0.2083607017993927, "learning_rate": 1.4900621493494137e-06, "loss": 0.0215, "step": 92600 }, { "grad_norm": 0.8984419703483582, "learning_rate": 1.4860582711492544e-06, "loss": 0.0363, "step": 92610 }, { "grad_norm": 0.4373045861721039, "learning_rate": 1.4820596983608325e-06, "loss": 0.031, "step": 92620 }, { "grad_norm": 0.3361067771911621, "learning_rate": 1.4780664314214044e-06, "loss": 0.024, "step": 92630 }, { "grad_norm": 0.3000401556491852, "learning_rate": 1.474078470767687e-06, "loss": 0.0362, "step": 92640 }, { "grad_norm": 0.2811749279499054, "learning_rate": 1.470095816835787e-06, "loss": 0.0266, "step": 92650 }, { "grad_norm": 0.4510931372642517, "learning_rate": 1.4661184700612395e-06, "loss": 0.0324, "step": 92660 }, { "grad_norm": 0.41665855050086975, "learning_rate": 1.4621464308790133e-06, "loss": 0.0278, "step": 92670 }, { "grad_norm": 0.5228087902069092, "learning_rate": 1.4581796997234775e-06, "loss": 0.0344, "step": 92680 }, { "grad_norm": 0.18997548520565033, "learning_rate": 1.4542182770284297e-06, "loss": 0.0244, "step": 92690 }, { "grad_norm": 0.26437607407569885, "learning_rate": 1.450262163227084e-06, "loss": 0.0243, "step": 92700 }, { "grad_norm": 0.39913633465766907, "learning_rate": 1.4463113587520671e-06, "loss": 0.0388, "step": 92710 }, { "grad_norm": 0.28041836619377136, "learning_rate": 1.4423658640354443e-06, "loss": 0.0286, "step": 92720 }, { "grad_norm": 0.46296176314353943, "learning_rate": 1.438425679508687e-06, "loss": 0.0285, "step": 92730 }, { "grad_norm": 0.5659652352333069, "learning_rate": 1.4344908056026895e-06, "loss": 0.0275, "step": 92740 }, { "grad_norm": 0.23999400436878204, "learning_rate": 1.4305612427477521e-06, "loss": 0.0219, "step": 92750 }, { "grad_norm": 0.5538867712020874, "learning_rate": 1.42663699137362e-06, "loss": 0.0219, "step": 92760 }, { "grad_norm": 0.7537627220153809, "learning_rate": 1.4227180519094386e-06, "loss": 0.0387, "step": 92770 }, { "grad_norm": 0.446774959564209, "learning_rate": 1.4188044247837706e-06, "loss": 0.0261, "step": 92780 }, { "grad_norm": 0.3463050425052643, "learning_rate": 1.4148961104246128e-06, "loss": 0.0254, "step": 92790 }, { "grad_norm": 0.44980406761169434, "learning_rate": 1.4109931092593731e-06, "loss": 0.0235, "step": 92800 }, { "grad_norm": 0.6376141309738159, "learning_rate": 1.4070954217148658e-06, "loss": 0.0338, "step": 92810 }, { "grad_norm": 0.4552609920501709, "learning_rate": 1.4032030482173497e-06, "loss": 0.031, "step": 92820 }, { "grad_norm": 0.5824984908103943, "learning_rate": 1.3993159891924901e-06, "loss": 0.032, "step": 92830 }, { "grad_norm": 0.9481059908866882, "learning_rate": 1.3954342450653524e-06, "loss": 0.0379, "step": 92840 }, { "grad_norm": 0.4607483148574829, "learning_rate": 1.3915578162604582e-06, "loss": 0.0299, "step": 92850 }, { "grad_norm": 0.32069671154022217, "learning_rate": 1.3876867032017126e-06, "loss": 0.0356, "step": 92860 }, { "grad_norm": 1.0299298763275146, "learning_rate": 1.383820906312472e-06, "loss": 0.0268, "step": 92870 }, { "grad_norm": 0.319036602973938, "learning_rate": 1.3799604260154752e-06, "loss": 0.0293, "step": 92880 }, { "grad_norm": 0.537522554397583, "learning_rate": 1.376105262732913e-06, "loss": 0.0377, "step": 92890 }, { "grad_norm": 0.4749935567378998, "learning_rate": 1.37225541688637e-06, "loss": 0.0368, "step": 92900 }, { "grad_norm": 0.2597234845161438, "learning_rate": 1.3684108888968594e-06, "loss": 0.0362, "step": 92910 }, { "grad_norm": 0.5687507390975952, "learning_rate": 1.364571679184823e-06, "loss": 0.0293, "step": 92920 }, { "grad_norm": 0.4576641023159027, "learning_rate": 1.3607377881701077e-06, "loss": 0.04, "step": 92930 }, { "grad_norm": 0.25979694724082947, "learning_rate": 1.3569092162719787e-06, "loss": 0.033, "step": 92940 }, { "grad_norm": 0.311862975358963, "learning_rate": 1.353085963909123e-06, "loss": 0.0255, "step": 92950 }, { "grad_norm": 0.3361283242702484, "learning_rate": 1.3492680314996454e-06, "loss": 0.0297, "step": 92960 }, { "grad_norm": 1.0037510395050049, "learning_rate": 1.345455419461078e-06, "loss": 0.0347, "step": 92970 }, { "grad_norm": 0.20163094997406006, "learning_rate": 1.3416481282103431e-06, "loss": 0.0276, "step": 92980 }, { "grad_norm": 0.37075528502464294, "learning_rate": 1.337846158163819e-06, "loss": 0.0363, "step": 92990 }, { "grad_norm": 0.2479042410850525, "learning_rate": 1.334049509737273e-06, "loss": 0.0332, "step": 93000 }, { "grad_norm": 0.37651824951171875, "learning_rate": 1.3302581833459116e-06, "loss": 0.0324, "step": 93010 }, { "grad_norm": 0.4499039351940155, "learning_rate": 1.3264721794043366e-06, "loss": 0.0298, "step": 93020 }, { "grad_norm": 0.47370103001594543, "learning_rate": 1.3226914983265836e-06, "loss": 0.0439, "step": 93030 }, { "grad_norm": 0.4780227839946747, "learning_rate": 1.3189161405260997e-06, "loss": 0.0345, "step": 93040 }, { "grad_norm": 0.40179866552352905, "learning_rate": 1.3151461064157656e-06, "loss": 0.0229, "step": 93050 }, { "grad_norm": 0.6655704379081726, "learning_rate": 1.3113813964078459e-06, "loss": 0.0331, "step": 93060 }, { "grad_norm": 0.2655336260795593, "learning_rate": 1.3076220109140614e-06, "loss": 0.0278, "step": 93070 }, { "grad_norm": 0.40368252992630005, "learning_rate": 1.3038679503455275e-06, "loss": 0.032, "step": 93080 }, { "grad_norm": 0.35085928440093994, "learning_rate": 1.3001192151127773e-06, "loss": 0.0262, "step": 93090 }, { "grad_norm": 0.2953014075756073, "learning_rate": 1.2963758056257657e-06, "loss": 0.029, "step": 93100 }, { "grad_norm": 0.27138379216194153, "learning_rate": 1.2926377222938768e-06, "loss": 0.0196, "step": 93110 }, { "grad_norm": 0.4995114505290985, "learning_rate": 1.2889049655259e-06, "loss": 0.035, "step": 93120 }, { "grad_norm": 0.4520469009876251, "learning_rate": 1.2851775357300312e-06, "loss": 0.0411, "step": 93130 }, { "grad_norm": 0.6452812552452087, "learning_rate": 1.281455433313916e-06, "loss": 0.0301, "step": 93140 }, { "grad_norm": 0.4772818684577942, "learning_rate": 1.2777386586845796e-06, "loss": 0.0288, "step": 93150 }, { "grad_norm": 0.5630384683609009, "learning_rate": 1.2740272122484964e-06, "loss": 0.0362, "step": 93160 }, { "grad_norm": 0.40280571579933167, "learning_rate": 1.2703210944115308e-06, "loss": 0.0266, "step": 93170 }, { "grad_norm": 0.4661126136779785, "learning_rate": 1.2666203055789916e-06, "loss": 0.0235, "step": 93180 }, { "grad_norm": 0.3452010750770569, "learning_rate": 1.2629248461555832e-06, "loss": 0.0271, "step": 93190 }, { "grad_norm": 0.4473426342010498, "learning_rate": 1.2592347165454431e-06, "loss": 0.024, "step": 93200 }, { "grad_norm": 0.34324684739112854, "learning_rate": 1.2555499171521102e-06, "loss": 0.0417, "step": 93210 }, { "grad_norm": 0.2649018466472626, "learning_rate": 1.251870448378556e-06, "loss": 0.0238, "step": 93220 }, { "grad_norm": 0.602903425693512, "learning_rate": 1.2481963106271478e-06, "loss": 0.0332, "step": 93230 }, { "grad_norm": 0.6299163699150085, "learning_rate": 1.2445275042997084e-06, "loss": 0.0372, "step": 93240 }, { "grad_norm": 0.35899946093559265, "learning_rate": 1.2408640297974283e-06, "loss": 0.0323, "step": 93250 }, { "grad_norm": 0.3293415606021881, "learning_rate": 1.2372058875209535e-06, "loss": 0.0271, "step": 93260 }, { "grad_norm": 0.5870851874351501, "learning_rate": 1.2335530778703252e-06, "loss": 0.024, "step": 93270 }, { "grad_norm": 0.4183669090270996, "learning_rate": 1.2299056012450182e-06, "loss": 0.0291, "step": 93280 }, { "grad_norm": 0.6820093989372253, "learning_rate": 1.2262634580439026e-06, "loss": 0.0351, "step": 93290 }, { "grad_norm": 0.4619016647338867, "learning_rate": 1.2226266486652926e-06, "loss": 0.0429, "step": 93300 }, { "grad_norm": 0.3888511061668396, "learning_rate": 1.2189951735068927e-06, "loss": 0.0292, "step": 93310 }, { "grad_norm": 0.9825748205184937, "learning_rate": 1.2153690329658407e-06, "loss": 0.0279, "step": 93320 }, { "grad_norm": 0.43768763542175293, "learning_rate": 1.2117482274386859e-06, "loss": 0.0347, "step": 93330 }, { "grad_norm": 0.372864693403244, "learning_rate": 1.2081327573213953e-06, "loss": 0.027, "step": 93340 }, { "grad_norm": 0.57514888048172, "learning_rate": 1.2045226230093465e-06, "loss": 0.0312, "step": 93350 }, { "grad_norm": 0.29124730825424194, "learning_rate": 1.2009178248973296e-06, "loss": 0.0317, "step": 93360 }, { "grad_norm": 0.4287497401237488, "learning_rate": 1.1973183633795849e-06, "loss": 0.026, "step": 93370 }, { "grad_norm": 0.14012965559959412, "learning_rate": 1.193724238849725e-06, "loss": 0.033, "step": 93380 }, { "grad_norm": 0.5317187309265137, "learning_rate": 1.190135451700808e-06, "loss": 0.0251, "step": 93390 }, { "grad_norm": 0.7809480428695679, "learning_rate": 1.186552002325292e-06, "loss": 0.0342, "step": 93400 }, { "grad_norm": 0.3772403299808502, "learning_rate": 1.1829738911150578e-06, "loss": 0.0265, "step": 93410 }, { "grad_norm": 0.7896714210510254, "learning_rate": 1.1794011184613985e-06, "loss": 0.0272, "step": 93420 }, { "grad_norm": 0.8676866888999939, "learning_rate": 1.17583368475504e-06, "loss": 0.0385, "step": 93430 }, { "grad_norm": 0.8968162536621094, "learning_rate": 1.1722715903860982e-06, "loss": 0.03, "step": 93440 }, { "grad_norm": 0.5070536732673645, "learning_rate": 1.168714835744128e-06, "loss": 0.0311, "step": 93450 }, { "grad_norm": 0.25255087018013, "learning_rate": 1.165163421218085e-06, "loss": 0.0273, "step": 93460 }, { "grad_norm": 0.5495425462722778, "learning_rate": 1.1616173471963477e-06, "loss": 0.0424, "step": 93470 }, { "grad_norm": 0.3655751943588257, "learning_rate": 1.1580766140667109e-06, "loss": 0.0223, "step": 93480 }, { "grad_norm": 0.29576772451400757, "learning_rate": 1.1545412222163764e-06, "loss": 0.0328, "step": 93490 }, { "grad_norm": 0.35408663749694824, "learning_rate": 1.1510111720319795e-06, "loss": 0.0333, "step": 93500 }, { "grad_norm": 0.2715378403663635, "learning_rate": 1.1474864638995663e-06, "loss": 0.0286, "step": 93510 }, { "grad_norm": 0.46463656425476074, "learning_rate": 1.1439670982045735e-06, "loss": 0.0321, "step": 93520 }, { "grad_norm": 0.47796598076820374, "learning_rate": 1.1404530753318877e-06, "loss": 0.0351, "step": 93530 }, { "grad_norm": 0.3265046775341034, "learning_rate": 1.136944395665801e-06, "loss": 0.0315, "step": 93540 }, { "grad_norm": 0.4833305776119232, "learning_rate": 1.1334410595900014e-06, "loss": 0.0309, "step": 93550 }, { "grad_norm": 0.3352941870689392, "learning_rate": 1.129943067487621e-06, "loss": 0.0344, "step": 93560 }, { "grad_norm": 0.41073769330978394, "learning_rate": 1.126450419741193e-06, "loss": 0.0247, "step": 93570 }, { "grad_norm": 0.3860410749912262, "learning_rate": 1.1229631167326671e-06, "loss": 0.028, "step": 93580 }, { "grad_norm": 0.44925224781036377, "learning_rate": 1.1194811588434052e-06, "loss": 0.0294, "step": 93590 }, { "grad_norm": 0.5048693418502808, "learning_rate": 1.1160045464542023e-06, "loss": 0.0292, "step": 93600 }, { "grad_norm": 0.3923693597316742, "learning_rate": 1.1125332799452325e-06, "loss": 0.0298, "step": 93610 }, { "grad_norm": 0.5825259685516357, "learning_rate": 1.109067359696131e-06, "loss": 0.0268, "step": 93620 }, { "grad_norm": 0.2527141571044922, "learning_rate": 1.1056067860859166e-06, "loss": 0.0243, "step": 93630 }, { "grad_norm": 1.8754199743270874, "learning_rate": 1.102151559493031e-06, "loss": 0.0269, "step": 93640 }, { "grad_norm": 0.48993372917175293, "learning_rate": 1.0987016802953332e-06, "loss": 0.0291, "step": 93650 }, { "grad_norm": 0.8626105189323425, "learning_rate": 1.0952571488701048e-06, "loss": 0.0269, "step": 93660 }, { "grad_norm": 0.537861704826355, "learning_rate": 1.0918179655940163e-06, "loss": 0.0323, "step": 93670 }, { "grad_norm": 0.6410015225410461, "learning_rate": 1.088384130843184e-06, "loss": 0.0278, "step": 93680 }, { "grad_norm": 0.47142431139945984, "learning_rate": 1.0849556449931296e-06, "loss": 0.0337, "step": 93690 }, { "grad_norm": 0.29536283016204834, "learning_rate": 1.0815325084187866e-06, "loss": 0.0254, "step": 93700 }, { "grad_norm": 0.41603320837020874, "learning_rate": 1.0781147214944943e-06, "loss": 0.0255, "step": 93710 }, { "grad_norm": 0.24816913902759552, "learning_rate": 1.0747022845940313e-06, "loss": 0.0248, "step": 93720 }, { "grad_norm": 0.3660184144973755, "learning_rate": 1.0712951980905605e-06, "loss": 0.0246, "step": 93730 }, { "grad_norm": 0.39212700724601746, "learning_rate": 1.0678934623566839e-06, "loss": 0.0351, "step": 93740 }, { "grad_norm": 0.4277152717113495, "learning_rate": 1.0644970777644091e-06, "loss": 0.0259, "step": 93750 }, { "grad_norm": 0.4615762233734131, "learning_rate": 1.0611060446851673e-06, "loss": 0.0265, "step": 93760 }, { "grad_norm": 0.4253540635108948, "learning_rate": 1.0577203634897836e-06, "loss": 0.0269, "step": 93770 }, { "grad_norm": 0.49294525384902954, "learning_rate": 1.0543400345485178e-06, "loss": 0.0252, "step": 93780 }, { "grad_norm": 1.1079449653625488, "learning_rate": 1.0509650582310404e-06, "loss": 0.0331, "step": 93790 }, { "grad_norm": 0.8126813173294067, "learning_rate": 1.0475954349064287e-06, "loss": 0.034, "step": 93800 }, { "grad_norm": 0.19791053235530853, "learning_rate": 1.0442311649431825e-06, "loss": 0.0268, "step": 93810 }, { "grad_norm": 0.22548454999923706, "learning_rate": 1.0408722487092071e-06, "loss": 0.0403, "step": 93820 }, { "grad_norm": 0.28278473019599915, "learning_rate": 1.0375186865718424e-06, "loss": 0.0205, "step": 93830 }, { "grad_norm": 0.3354560136795044, "learning_rate": 1.0341704788978113e-06, "loss": 0.0273, "step": 93840 }, { "grad_norm": 0.28871917724609375, "learning_rate": 1.0308276260532879e-06, "loss": 0.0312, "step": 93850 }, { "grad_norm": 0.5363390445709229, "learning_rate": 1.0274901284038241e-06, "loss": 0.0308, "step": 93860 }, { "grad_norm": 0.21667124330997467, "learning_rate": 1.0241579863144114e-06, "loss": 0.0247, "step": 93870 }, { "grad_norm": 0.42918306589126587, "learning_rate": 1.0208312001494413e-06, "loss": 0.0291, "step": 93880 }, { "grad_norm": 0.49044910073280334, "learning_rate": 1.0175097702727453e-06, "loss": 0.0362, "step": 93890 }, { "grad_norm": 0.24896396696567535, "learning_rate": 1.0141936970475275e-06, "loss": 0.0295, "step": 93900 }, { "grad_norm": 0.279239296913147, "learning_rate": 1.0108829808364417e-06, "loss": 0.0321, "step": 93910 }, { "grad_norm": 0.3406013250350952, "learning_rate": 1.0075776220015376e-06, "loss": 0.0327, "step": 93920 }, { "grad_norm": 0.4403347671031952, "learning_rate": 1.0042776209042816e-06, "loss": 0.0254, "step": 93930 }, { "grad_norm": 0.4041624963283539, "learning_rate": 1.000982977905568e-06, "loss": 0.0312, "step": 93940 }, { "grad_norm": 0.6164730191230774, "learning_rate": 9.976936933656867e-07, "loss": 0.0307, "step": 93950 }, { "grad_norm": 0.45642393827438354, "learning_rate": 9.944097676443442e-07, "loss": 0.0301, "step": 93960 }, { "grad_norm": 0.2781229019165039, "learning_rate": 9.911312011006757e-07, "loss": 0.0322, "step": 93970 }, { "grad_norm": 0.4874911606311798, "learning_rate": 9.878579940932109e-07, "loss": 0.0297, "step": 93980 }, { "grad_norm": 0.35493457317352295, "learning_rate": 9.84590146979908e-07, "loss": 0.0333, "step": 93990 }, { "grad_norm": 0.29703226685523987, "learning_rate": 9.813276601181365e-07, "loss": 0.028, "step": 94000 }, { "grad_norm": 0.44623100757598877, "learning_rate": 9.780705338646724e-07, "loss": 0.0228, "step": 94010 }, { "grad_norm": 0.4747156798839569, "learning_rate": 9.748187685757083e-07, "loss": 0.0307, "step": 94020 }, { "grad_norm": 0.5405715703964233, "learning_rate": 9.715723646068543e-07, "loss": 0.0267, "step": 94030 }, { "grad_norm": 0.4976470470428467, "learning_rate": 9.68331322313143e-07, "loss": 0.0263, "step": 94040 }, { "grad_norm": 0.43208691477775574, "learning_rate": 9.650956420489909e-07, "loss": 0.0326, "step": 94050 }, { "grad_norm": 0.1920875608921051, "learning_rate": 9.618653241682595e-07, "loss": 0.0313, "step": 94060 }, { "grad_norm": 1.1736540794372559, "learning_rate": 9.586403690242053e-07, "loss": 0.0366, "step": 94070 }, { "grad_norm": 0.543373167514801, "learning_rate": 9.554207769695122e-07, "loss": 0.0253, "step": 94080 }, { "grad_norm": 0.34329870343208313, "learning_rate": 9.522065483562659e-07, "loss": 0.0311, "step": 94090 }, { "grad_norm": 0.40698108077049255, "learning_rate": 9.489976835359681e-07, "loss": 0.0269, "step": 94100 }, { "grad_norm": 0.5566681623458862, "learning_rate": 9.457941828595329e-07, "loss": 0.0306, "step": 94110 }, { "grad_norm": 0.5827808380126953, "learning_rate": 9.425960466772965e-07, "loss": 0.0281, "step": 94120 }, { "grad_norm": 0.452602356672287, "learning_rate": 9.394032753390014e-07, "loss": 0.0323, "step": 94130 }, { "grad_norm": 0.24115483462810516, "learning_rate": 9.362158691938017e-07, "loss": 0.0218, "step": 94140 }, { "grad_norm": 0.8071067929267883, "learning_rate": 9.330338285902685e-07, "loss": 0.0269, "step": 94150 }, { "grad_norm": 0.3605019748210907, "learning_rate": 9.298571538763845e-07, "loss": 0.0266, "step": 94160 }, { "grad_norm": 0.2642477750778198, "learning_rate": 9.266858453995441e-07, "loss": 0.0274, "step": 94170 }, { "grad_norm": 0.5210911631584167, "learning_rate": 9.235199035065645e-07, "loss": 0.0298, "step": 94180 }, { "grad_norm": 0.28206124901771545, "learning_rate": 9.203593285436574e-07, "loss": 0.0265, "step": 94190 }, { "grad_norm": 0.321016788482666, "learning_rate": 9.172041208564686e-07, "loss": 0.0294, "step": 94200 }, { "grad_norm": 0.3221161961555481, "learning_rate": 9.140542807900332e-07, "loss": 0.0253, "step": 94210 }, { "grad_norm": 0.34753939509391785, "learning_rate": 9.109098086888313e-07, "loss": 0.0287, "step": 94220 }, { "grad_norm": 0.24690589308738708, "learning_rate": 9.077707048967265e-07, "loss": 0.036, "step": 94230 }, { "grad_norm": 0.3673691749572754, "learning_rate": 9.046369697570057e-07, "loss": 0.0246, "step": 94240 }, { "grad_norm": 0.3337953984737396, "learning_rate": 9.015086036123721e-07, "loss": 0.0214, "step": 94250 }, { "grad_norm": 0.3485789895057678, "learning_rate": 8.983856068049413e-07, "loss": 0.0277, "step": 94260 }, { "grad_norm": 0.5100995302200317, "learning_rate": 8.952679796762398e-07, "loss": 0.0337, "step": 94270 }, { "grad_norm": 0.3219805359840393, "learning_rate": 8.921557225671951e-07, "loss": 0.0257, "step": 94280 }, { "grad_norm": 0.45656365156173706, "learning_rate": 8.890488358181792e-07, "loss": 0.031, "step": 94290 }, { "grad_norm": 0.4024224877357483, "learning_rate": 8.859473197689372e-07, "loss": 0.0285, "step": 94300 }, { "grad_norm": 0.532908022403717, "learning_rate": 8.828511747586588e-07, "loss": 0.025, "step": 94310 }, { "grad_norm": 0.40956953167915344, "learning_rate": 8.797604011259287e-07, "loss": 0.0306, "step": 94320 }, { "grad_norm": 0.39551833271980286, "learning_rate": 8.766749992087487e-07, "loss": 0.0287, "step": 94330 }, { "grad_norm": 0.6235025525093079, "learning_rate": 8.735949693445378e-07, "loss": 0.0266, "step": 94340 }, { "grad_norm": 0.510109007358551, "learning_rate": 8.705203118701211e-07, "loss": 0.0225, "step": 94350 }, { "grad_norm": 0.4887590706348419, "learning_rate": 8.674510271217295e-07, "loss": 0.0365, "step": 94360 }, { "grad_norm": 0.4643779397010803, "learning_rate": 8.643871154350336e-07, "loss": 0.0309, "step": 94370 }, { "grad_norm": 0.30978408455848694, "learning_rate": 8.613285771450818e-07, "loss": 0.0339, "step": 94380 }, { "grad_norm": 0.25427117943763733, "learning_rate": 8.582754125863623e-07, "loss": 0.0237, "step": 94390 }, { "grad_norm": 0.42202168703079224, "learning_rate": 8.55227622092758e-07, "loss": 0.0242, "step": 94400 }, { "grad_norm": 0.4654254615306854, "learning_rate": 8.521852059975743e-07, "loss": 0.0328, "step": 94410 }, { "grad_norm": 0.35099729895591736, "learning_rate": 8.491481646335231e-07, "loss": 0.0313, "step": 94420 }, { "grad_norm": 0.8509178161621094, "learning_rate": 8.46116498332733e-07, "loss": 0.0283, "step": 94430 }, { "grad_norm": 0.475974440574646, "learning_rate": 8.430902074267333e-07, "loss": 0.0324, "step": 94440 }, { "grad_norm": 1.3044742345809937, "learning_rate": 8.400692922464925e-07, "loss": 0.0377, "step": 94450 }, { "grad_norm": 0.40348950028419495, "learning_rate": 8.370537531223577e-07, "loss": 0.0268, "step": 94460 }, { "grad_norm": 0.2963985502719879, "learning_rate": 8.340435903841092e-07, "loss": 0.0215, "step": 94470 }, { "grad_norm": 0.364822655916214, "learning_rate": 8.310388043609341e-07, "loss": 0.03, "step": 94480 }, { "grad_norm": 0.3291003406047821, "learning_rate": 8.280393953814303e-07, "loss": 0.0346, "step": 94490 }, { "grad_norm": 0.5878297090530396, "learning_rate": 8.250453637736077e-07, "loss": 0.0285, "step": 94500 }, { "grad_norm": 0.35444971919059753, "learning_rate": 8.22056709864899e-07, "loss": 0.0271, "step": 94510 }, { "grad_norm": 0.20927023887634277, "learning_rate": 8.190734339821204e-07, "loss": 0.0244, "step": 94520 }, { "grad_norm": 0.7656623721122742, "learning_rate": 8.160955364515333e-07, "loss": 0.0282, "step": 94530 }, { "grad_norm": 0.4145977199077606, "learning_rate": 8.131230175987936e-07, "loss": 0.0239, "step": 94540 }, { "grad_norm": 0.5496434569358826, "learning_rate": 8.101558777489693e-07, "loss": 0.0407, "step": 94550 }, { "grad_norm": 0.8658741116523743, "learning_rate": 8.071941172265396e-07, "loss": 0.0329, "step": 94560 }, { "grad_norm": 0.3265942335128784, "learning_rate": 8.042377363553955e-07, "loss": 0.0293, "step": 94570 }, { "grad_norm": 0.25856003165245056, "learning_rate": 8.01286735458856e-07, "loss": 0.031, "step": 94580 }, { "grad_norm": 0.6366099119186401, "learning_rate": 7.983411148596298e-07, "loss": 0.0347, "step": 94590 }, { "grad_norm": 0.25291621685028076, "learning_rate": 7.954008748798425e-07, "loss": 0.0272, "step": 94600 }, { "grad_norm": 0.2769468426704407, "learning_rate": 7.924660158410369e-07, "loss": 0.028, "step": 94610 }, { "grad_norm": 0.43209221959114075, "learning_rate": 7.895365380641728e-07, "loss": 0.0347, "step": 94620 }, { "grad_norm": 0.5112416744232178, "learning_rate": 7.866124418695941e-07, "loss": 0.0308, "step": 94630 }, { "grad_norm": 0.7263442277908325, "learning_rate": 7.83693727577095e-07, "loss": 0.032, "step": 94640 }, { "grad_norm": 0.405943900346756, "learning_rate": 7.807803955058535e-07, "loss": 0.0263, "step": 94650 }, { "grad_norm": 0.26122304797172546, "learning_rate": 7.778724459744702e-07, "loss": 0.023, "step": 94660 }, { "grad_norm": 0.3897717595100403, "learning_rate": 7.749698793009464e-07, "loss": 0.0331, "step": 94670 }, { "grad_norm": 0.4014924466609955, "learning_rate": 7.720726958027169e-07, "loss": 0.0286, "step": 94680 }, { "grad_norm": 0.4601767659187317, "learning_rate": 7.691808957965951e-07, "loss": 0.0272, "step": 94690 }, { "grad_norm": 0.2530076503753662, "learning_rate": 7.662944795988336e-07, "loss": 0.0204, "step": 94700 }, { "grad_norm": 0.2743914723396301, "learning_rate": 7.634134475250854e-07, "loss": 0.0298, "step": 94710 }, { "grad_norm": 0.28017863631248474, "learning_rate": 7.605377998904262e-07, "loss": 0.029, "step": 94720 }, { "grad_norm": 0.5044038891792297, "learning_rate": 7.5766753700931e-07, "loss": 0.0257, "step": 94730 }, { "grad_norm": 0.3733886778354645, "learning_rate": 7.54802659195647e-07, "loss": 0.0295, "step": 94740 }, { "grad_norm": 0.2230607271194458, "learning_rate": 7.519431667627253e-07, "loss": 0.0229, "step": 94750 }, { "grad_norm": 0.8461614847183228, "learning_rate": 7.490890600232503e-07, "loss": 0.0432, "step": 94760 }, { "grad_norm": 0.5946139097213745, "learning_rate": 7.462403392893502e-07, "loss": 0.03, "step": 94770 }, { "grad_norm": 0.4035591781139374, "learning_rate": 7.433970048725536e-07, "loss": 0.0289, "step": 94780 }, { "grad_norm": 0.2231171727180481, "learning_rate": 7.405590570838062e-07, "loss": 0.0254, "step": 94790 }, { "grad_norm": 0.20130683481693268, "learning_rate": 7.377264962334596e-07, "loss": 0.0231, "step": 94800 }, { "grad_norm": 0.2920379638671875, "learning_rate": 7.348993226312773e-07, "loss": 0.0265, "step": 94810 }, { "grad_norm": 0.5673903226852417, "learning_rate": 7.32077536586434e-07, "loss": 0.027, "step": 94820 }, { "grad_norm": 0.5699689984321594, "learning_rate": 7.292611384075221e-07, "loss": 0.0253, "step": 94830 }, { "grad_norm": 0.3704364001750946, "learning_rate": 7.264501284025338e-07, "loss": 0.0329, "step": 94840 }, { "grad_norm": 0.8015066981315613, "learning_rate": 7.236445068788789e-07, "loss": 0.0294, "step": 94850 }, { "grad_norm": 0.28013452887535095, "learning_rate": 7.20844274143373e-07, "loss": 0.0284, "step": 94860 }, { "grad_norm": 0.5514815449714661, "learning_rate": 7.180494305022489e-07, "loss": 0.0295, "step": 94870 }, { "grad_norm": 0.4216873049736023, "learning_rate": 7.152599762611456e-07, "loss": 0.0289, "step": 94880 }, { "grad_norm": 0.3406769037246704, "learning_rate": 7.124759117251079e-07, "loss": 0.0464, "step": 94890 }, { "grad_norm": 0.62451171875, "learning_rate": 7.096972371986033e-07, "loss": 0.0347, "step": 94900 }, { "grad_norm": 0.8732365965843201, "learning_rate": 7.069239529855054e-07, "loss": 0.0348, "step": 94910 }, { "grad_norm": 0.4930236339569092, "learning_rate": 7.041560593890939e-07, "loss": 0.0233, "step": 94920 }, { "grad_norm": 0.4664950668811798, "learning_rate": 7.013935567120599e-07, "loss": 0.0278, "step": 94930 }, { "grad_norm": 0.34083887934684753, "learning_rate": 6.986364452565009e-07, "loss": 0.0213, "step": 94940 }, { "grad_norm": 0.6937850713729858, "learning_rate": 6.95884725323942e-07, "loss": 0.0427, "step": 94950 }, { "grad_norm": 0.17755365371704102, "learning_rate": 6.931383972152982e-07, "loss": 0.0287, "step": 94960 }, { "grad_norm": 0.3291989862918854, "learning_rate": 6.903974612309127e-07, "loss": 0.0236, "step": 94970 }, { "grad_norm": 0.25951823592185974, "learning_rate": 6.876619176705179e-07, "loss": 0.0246, "step": 94980 }, { "grad_norm": 0.2968331575393677, "learning_rate": 6.8493176683328e-07, "loss": 0.0302, "step": 94990 }, { "grad_norm": 0.6617844104766846, "learning_rate": 6.822070090177657e-07, "loss": 0.0313, "step": 95000 }, { "grad_norm": 0.31990230083465576, "learning_rate": 6.794876445219367e-07, "loss": 0.0347, "step": 95010 }, { "grad_norm": 0.38578492403030396, "learning_rate": 6.767736736431885e-07, "loss": 0.0257, "step": 95020 }, { "grad_norm": 0.6751106381416321, "learning_rate": 6.740650966783113e-07, "loss": 0.0311, "step": 95030 }, { "grad_norm": 0.27033141255378723, "learning_rate": 6.713619139235239e-07, "loss": 0.0222, "step": 95040 }, { "grad_norm": 0.426960289478302, "learning_rate": 6.686641256744286e-07, "loss": 0.0252, "step": 95050 }, { "grad_norm": 0.23619084060192108, "learning_rate": 6.659717322260562e-07, "loss": 0.0294, "step": 95060 }, { "grad_norm": 0.23342953622341156, "learning_rate": 6.632847338728377e-07, "loss": 0.0245, "step": 95070 }, { "grad_norm": 0.3913153409957886, "learning_rate": 6.60603130908627e-07, "loss": 0.0291, "step": 95080 }, { "grad_norm": 0.36026081442832947, "learning_rate": 6.579269236266783e-07, "loss": 0.0309, "step": 95090 }, { "grad_norm": 0.2514401078224182, "learning_rate": 6.552561123196632e-07, "loss": 0.0221, "step": 95100 }, { "grad_norm": 0.4548299312591553, "learning_rate": 6.525906972796425e-07, "loss": 0.0311, "step": 95110 }, { "grad_norm": 0.4204667806625366, "learning_rate": 6.499306787981108e-07, "loss": 0.0386, "step": 95120 }, { "grad_norm": 0.36716538667678833, "learning_rate": 6.472760571659631e-07, "loss": 0.0245, "step": 95130 }, { "grad_norm": 0.453693151473999, "learning_rate": 6.446268326735061e-07, "loss": 0.0342, "step": 95140 }, { "grad_norm": 0.27113139629364014, "learning_rate": 6.41983005610458e-07, "loss": 0.0231, "step": 95150 }, { "grad_norm": 0.25146883726119995, "learning_rate": 6.393445762659378e-07, "loss": 0.0348, "step": 95160 }, { "grad_norm": 0.3601292371749878, "learning_rate": 6.36711544928481e-07, "loss": 0.0205, "step": 95170 }, { "grad_norm": 0.6899257898330688, "learning_rate": 6.340839118860298e-07, "loss": 0.0406, "step": 95180 }, { "grad_norm": 0.511247456073761, "learning_rate": 6.314616774259485e-07, "loss": 0.0341, "step": 95190 }, { "grad_norm": 0.2992532551288605, "learning_rate": 6.288448418349858e-07, "loss": 0.024, "step": 95200 }, { "grad_norm": 0.33540093898773193, "learning_rate": 6.262334053993235e-07, "loss": 0.0376, "step": 95210 }, { "grad_norm": 0.32790881395339966, "learning_rate": 6.236273684045501e-07, "loss": 0.0233, "step": 95220 }, { "grad_norm": 0.35536855459213257, "learning_rate": 6.210267311356488e-07, "loss": 0.0307, "step": 95230 }, { "grad_norm": 0.4140656292438507, "learning_rate": 6.184314938770197e-07, "loss": 0.0355, "step": 95240 }, { "grad_norm": 0.3034851849079132, "learning_rate": 6.158416569124858e-07, "loss": 0.0251, "step": 95250 }, { "grad_norm": 0.6061075925827026, "learning_rate": 6.132572205252541e-07, "loss": 0.0181, "step": 95260 }, { "grad_norm": 0.5722630023956299, "learning_rate": 6.106781849979648e-07, "loss": 0.031, "step": 95270 }, { "grad_norm": 0.4023876488208771, "learning_rate": 6.081045506126482e-07, "loss": 0.0235, "step": 95280 }, { "grad_norm": 0.42342206835746765, "learning_rate": 6.055363176507678e-07, "loss": 0.042, "step": 95290 }, { "grad_norm": 0.6990505456924438, "learning_rate": 6.029734863931602e-07, "loss": 0.0307, "step": 95300 }, { "grad_norm": 0.2677459418773651, "learning_rate": 6.004160571201179e-07, "loss": 0.0285, "step": 95310 }, { "grad_norm": 0.4266037940979004, "learning_rate": 5.978640301112947e-07, "loss": 0.0278, "step": 95320 }, { "grad_norm": 0.830818235874176, "learning_rate": 5.9531740564579e-07, "loss": 0.0289, "step": 95330 }, { "grad_norm": 0.8072030544281006, "learning_rate": 5.927761840020918e-07, "loss": 0.0273, "step": 95340 }, { "grad_norm": 0.26156023144721985, "learning_rate": 5.902403654581113e-07, "loss": 0.0254, "step": 95350 }, { "grad_norm": 0.5199830532073975, "learning_rate": 5.877099502911542e-07, "loss": 0.034, "step": 95360 }, { "grad_norm": 0.4947775602340698, "learning_rate": 5.851849387779496e-07, "loss": 0.0391, "step": 95370 }, { "grad_norm": 0.3709614872932434, "learning_rate": 5.826653311946262e-07, "loss": 0.0292, "step": 95380 }, { "grad_norm": 0.4213026762008667, "learning_rate": 5.801511278167249e-07, "loss": 0.0255, "step": 95390 }, { "grad_norm": 0.3419293761253357, "learning_rate": 5.776423289191868e-07, "loss": 0.0287, "step": 95400 }, { "grad_norm": 0.29880571365356445, "learning_rate": 5.751389347763869e-07, "loss": 0.0373, "step": 95410 }, { "grad_norm": 0.45861002802848816, "learning_rate": 5.726409456620841e-07, "loss": 0.0263, "step": 95420 }, { "grad_norm": 0.3678516149520874, "learning_rate": 5.701483618494541e-07, "loss": 0.0246, "step": 95430 }, { "grad_norm": 0.29542258381843567, "learning_rate": 5.67661183611079e-07, "loss": 0.0301, "step": 95440 }, { "grad_norm": 0.33400869369506836, "learning_rate": 5.651794112189579e-07, "loss": 0.0212, "step": 95450 }, { "grad_norm": 0.4595295190811157, "learning_rate": 5.627030449444903e-07, "loss": 0.0322, "step": 95460 }, { "grad_norm": 0.37888965010643005, "learning_rate": 5.60232085058493e-07, "loss": 0.0356, "step": 95470 }, { "grad_norm": 0.65305095911026, "learning_rate": 5.577665318311776e-07, "loss": 0.0361, "step": 95480 }, { "grad_norm": 0.4258967339992523, "learning_rate": 5.553063855321838e-07, "loss": 0.0309, "step": 95490 }, { "grad_norm": 0.3921591341495514, "learning_rate": 5.528516464305467e-07, "loss": 0.0363, "step": 95500 }, { "grad_norm": 0.3085099756717682, "learning_rate": 5.504023147947068e-07, "loss": 0.0253, "step": 95510 }, { "grad_norm": 0.5210075378417969, "learning_rate": 5.479583908925279e-07, "loss": 0.0368, "step": 95520 }, { "grad_norm": 0.24565963447093964, "learning_rate": 5.455198749912626e-07, "loss": 0.0241, "step": 95530 }, { "grad_norm": 0.4711940288543701, "learning_rate": 5.430867673575979e-07, "loss": 0.0377, "step": 95540 }, { "grad_norm": 0.27709054946899414, "learning_rate": 5.406590682575985e-07, "loss": 0.0227, "step": 95550 }, { "grad_norm": 0.2700817883014679, "learning_rate": 5.382367779567687e-07, "loss": 0.0282, "step": 95560 }, { "grad_norm": 0.5005359649658203, "learning_rate": 5.358198967200023e-07, "loss": 0.0222, "step": 95570 }, { "grad_norm": 0.5281249284744263, "learning_rate": 5.334084248115989e-07, "loss": 0.0359, "step": 95580 }, { "grad_norm": 0.3362497389316559, "learning_rate": 5.310023624952754e-07, "loss": 0.0225, "step": 95590 }, { "grad_norm": 0.6246123313903809, "learning_rate": 5.286017100341656e-07, "loss": 0.0288, "step": 95600 }, { "grad_norm": 0.3649922013282776, "learning_rate": 5.262064676907874e-07, "loss": 0.0363, "step": 95610 }, { "grad_norm": 0.36049339175224304, "learning_rate": 5.238166357270924e-07, "loss": 0.0262, "step": 95620 }, { "grad_norm": 0.4249517321586609, "learning_rate": 5.214322144044215e-07, "loss": 0.0192, "step": 95630 }, { "grad_norm": 0.347210168838501, "learning_rate": 5.190532039835327e-07, "loss": 0.0366, "step": 95640 }, { "grad_norm": 0.444747656583786, "learning_rate": 5.166796047245903e-07, "loss": 0.0333, "step": 95650 }, { "grad_norm": 0.9428853392601013, "learning_rate": 5.143114168871699e-07, "loss": 0.0322, "step": 95660 }, { "grad_norm": 0.5820509791374207, "learning_rate": 5.119486407302532e-07, "loss": 0.0294, "step": 95670 }, { "grad_norm": 0.4654204547405243, "learning_rate": 5.095912765122335e-07, "loss": 0.035, "step": 95680 }, { "grad_norm": 0.8554480075836182, "learning_rate": 5.07239324490899e-07, "loss": 0.0394, "step": 95690 }, { "grad_norm": 0.34959661960601807, "learning_rate": 5.048927849234608e-07, "loss": 0.0354, "step": 95700 }, { "grad_norm": 0.5458112955093384, "learning_rate": 5.025516580665358e-07, "loss": 0.0366, "step": 95710 }, { "grad_norm": 0.49960756301879883, "learning_rate": 5.002159441761412e-07, "loss": 0.0234, "step": 95720 }, { "grad_norm": 0.24810996651649475, "learning_rate": 4.978856435077062e-07, "loss": 0.0281, "step": 95730 }, { "grad_norm": 0.4536961317062378, "learning_rate": 4.955607563160769e-07, "loss": 0.0329, "step": 95740 }, { "grad_norm": 0.44250982999801636, "learning_rate": 4.932412828554944e-07, "loss": 0.0299, "step": 95750 }, { "grad_norm": 0.3492237627506256, "learning_rate": 4.909272233796114e-07, "loss": 0.026, "step": 95760 }, { "grad_norm": 0.3001393675804138, "learning_rate": 4.88618578141492e-07, "loss": 0.0358, "step": 95770 }, { "grad_norm": 0.3572632372379303, "learning_rate": 4.863153473936011e-07, "loss": 0.0306, "step": 95780 }, { "grad_norm": 0.2690746784210205, "learning_rate": 4.840175313878259e-07, "loss": 0.0258, "step": 95790 }, { "grad_norm": 0.2984279990196228, "learning_rate": 4.817251303754433e-07, "loss": 0.0245, "step": 95800 }, { "grad_norm": 0.4044957756996155, "learning_rate": 4.794381446071583e-07, "loss": 0.0282, "step": 95810 }, { "grad_norm": 0.46947771310806274, "learning_rate": 4.771565743330597e-07, "loss": 0.0341, "step": 95820 }, { "grad_norm": 0.3544527590274811, "learning_rate": 4.7488041980266463e-07, "loss": 0.0308, "step": 95830 }, { "grad_norm": 0.3706792891025543, "learning_rate": 4.726096812648795e-07, "loss": 0.0328, "step": 95840 }, { "grad_norm": 0.29055002331733704, "learning_rate": 4.703443589680445e-07, "loss": 0.028, "step": 95850 }, { "grad_norm": 0.753374457359314, "learning_rate": 4.6808445315987826e-07, "loss": 0.0324, "step": 95860 }, { "grad_norm": 0.3821948170661926, "learning_rate": 4.658299640875274e-07, "loss": 0.0246, "step": 95870 }, { "grad_norm": 0.26138192415237427, "learning_rate": 4.635808919975393e-07, "loss": 0.0227, "step": 95880 }, { "grad_norm": 0.5397437214851379, "learning_rate": 4.6133723713586706e-07, "loss": 0.0294, "step": 95890 }, { "grad_norm": 0.37718087434768677, "learning_rate": 4.590989997478756e-07, "loss": 0.0247, "step": 95900 }, { "grad_norm": 0.29089298844337463, "learning_rate": 4.5686618007833027e-07, "loss": 0.0278, "step": 95910 }, { "grad_norm": 0.471791535615921, "learning_rate": 4.54638778371419e-07, "loss": 0.0312, "step": 95920 }, { "grad_norm": 0.439505398273468, "learning_rate": 4.524167948707192e-07, "loss": 0.0364, "step": 95930 }, { "grad_norm": 0.437393456697464, "learning_rate": 4.5020022981921986e-07, "loss": 0.0298, "step": 95940 }, { "grad_norm": 0.3292682468891144, "learning_rate": 4.47989083459327e-07, "loss": 0.0412, "step": 95950 }, { "grad_norm": 0.7061328291893005, "learning_rate": 4.457833560328528e-07, "loss": 0.0272, "step": 95960 }, { "grad_norm": 0.2917473316192627, "learning_rate": 4.4358304778100434e-07, "loss": 0.0312, "step": 95970 }, { "grad_norm": 0.2722873389720917, "learning_rate": 4.413881589444058e-07, "loss": 0.0308, "step": 95980 }, { "grad_norm": 0.3423134982585907, "learning_rate": 4.391986897630873e-07, "loss": 0.0283, "step": 95990 }, { "grad_norm": 0.374024897813797, "learning_rate": 4.370146404764908e-07, "loss": 0.0461, "step": 96000 }, { "grad_norm": 0.3461434543132782, "learning_rate": 4.3483601132345285e-07, "loss": 0.0337, "step": 96010 }, { "grad_norm": 1.8360874652862549, "learning_rate": 4.32662802542233e-07, "loss": 0.0364, "step": 96020 }, { "grad_norm": 0.29665714502334595, "learning_rate": 4.3049501437047444e-07, "loss": 0.0273, "step": 96030 }, { "grad_norm": 0.38454732298851013, "learning_rate": 4.283326470452653e-07, "loss": 0.0308, "step": 96040 }, { "grad_norm": 0.2698609232902527, "learning_rate": 4.2617570080306647e-07, "loss": 0.0208, "step": 96050 }, { "grad_norm": 0.7815397381782532, "learning_rate": 4.2402417587975585e-07, "loss": 0.034, "step": 96060 }, { "grad_norm": 0.6239776015281677, "learning_rate": 4.218780725106286e-07, "loss": 0.0273, "step": 96070 }, { "grad_norm": 0.6340154409408569, "learning_rate": 4.197373909303748e-07, "loss": 0.0317, "step": 96080 }, { "grad_norm": 0.44858306646347046, "learning_rate": 4.17602131373096e-07, "loss": 0.0232, "step": 96090 }, { "grad_norm": 1.1209425926208496, "learning_rate": 4.154722940723055e-07, "loss": 0.04, "step": 96100 }, { "grad_norm": 0.5655556917190552, "learning_rate": 4.1334787926091136e-07, "loss": 0.0257, "step": 96110 }, { "grad_norm": 0.9336406588554382, "learning_rate": 4.1122888717124444e-07, "loss": 0.0294, "step": 96120 }, { "grad_norm": 0.28580692410469055, "learning_rate": 4.091153180350249e-07, "loss": 0.0322, "step": 96130 }, { "grad_norm": 0.3397689461708069, "learning_rate": 4.0700717208340125e-07, "loss": 0.0288, "step": 96140 }, { "grad_norm": 0.3892999589443207, "learning_rate": 4.049044495469112e-07, "loss": 0.0282, "step": 96150 }, { "grad_norm": 0.4816236197948456, "learning_rate": 4.0280715065549866e-07, "loss": 0.0292, "step": 96160 }, { "grad_norm": 0.3978753387928009, "learning_rate": 4.0071527563853575e-07, "loss": 0.0276, "step": 96170 }, { "grad_norm": 0.47009673714637756, "learning_rate": 3.9862882472477823e-07, "loss": 0.021, "step": 96180 }, { "grad_norm": 0.6822125315666199, "learning_rate": 3.9654779814239375e-07, "loss": 0.0295, "step": 96190 }, { "grad_norm": 0.24996022880077362, "learning_rate": 3.944721961189668e-07, "loss": 0.0267, "step": 96200 }, { "grad_norm": 0.3568686842918396, "learning_rate": 3.9240201888147697e-07, "loss": 0.0314, "step": 96210 }, { "grad_norm": 0.48248210549354553, "learning_rate": 3.9033726665632096e-07, "loss": 0.0342, "step": 96220 }, { "grad_norm": 0.5114561319351196, "learning_rate": 3.882779396692959e-07, "loss": 0.0254, "step": 96230 }, { "grad_norm": 0.3525954484939575, "learning_rate": 3.8622403814559947e-07, "loss": 0.0328, "step": 96240 }, { "grad_norm": 0.3199070394039154, "learning_rate": 3.841755623098575e-07, "loss": 0.0232, "step": 96250 }, { "grad_norm": 0.3434975743293762, "learning_rate": 3.821325123860742e-07, "loss": 0.0264, "step": 96260 }, { "grad_norm": 0.7486192584037781, "learning_rate": 3.800948885976818e-07, "loss": 0.0271, "step": 96270 }, { "grad_norm": 0.26702526211738586, "learning_rate": 3.780626911675078e-07, "loss": 0.0361, "step": 96280 }, { "grad_norm": 0.2769666016101837, "learning_rate": 3.76035920317791e-07, "loss": 0.039, "step": 96290 }, { "grad_norm": 0.37299221754074097, "learning_rate": 3.740145762701819e-07, "loss": 0.0258, "step": 96300 }, { "grad_norm": 0.5112379789352417, "learning_rate": 3.7199865924572585e-07, "loss": 0.0261, "step": 96310 }, { "grad_norm": 0.28608500957489014, "learning_rate": 3.699881694648799e-07, "loss": 0.0219, "step": 96320 }, { "grad_norm": 0.5610876083374023, "learning_rate": 3.679831071475126e-07, "loss": 0.0329, "step": 96330 }, { "grad_norm": 0.42278122901916504, "learning_rate": 3.65983472512893e-07, "loss": 0.0281, "step": 96340 }, { "grad_norm": 0.3192477524280548, "learning_rate": 3.6398926577969616e-07, "loss": 0.0242, "step": 96350 }, { "grad_norm": 0.373055636882782, "learning_rate": 3.620004871660032e-07, "loss": 0.0236, "step": 96360 }, { "grad_norm": 0.3039243221282959, "learning_rate": 3.600171368893124e-07, "loss": 0.0213, "step": 96370 }, { "grad_norm": 0.5097336769104004, "learning_rate": 3.580392151665113e-07, "loss": 0.0339, "step": 96380 }, { "grad_norm": 0.20185653865337372, "learning_rate": 3.5606672221391024e-07, "loss": 0.0185, "step": 96390 }, { "grad_norm": 0.45532169938087463, "learning_rate": 3.540996582472089e-07, "loss": 0.0245, "step": 96400 }, { "grad_norm": 0.3184249699115753, "learning_rate": 3.521380234815297e-07, "loss": 0.0279, "step": 96410 }, { "grad_norm": 0.43509483337402344, "learning_rate": 3.5018181813138985e-07, "loss": 0.0267, "step": 96420 }, { "grad_norm": 0.42369818687438965, "learning_rate": 3.482310424107238e-07, "loss": 0.0284, "step": 96430 }, { "grad_norm": 0.7124198079109192, "learning_rate": 3.4628569653285537e-07, "loss": 0.0252, "step": 96440 }, { "grad_norm": 0.26961126923561096, "learning_rate": 3.44345780710531e-07, "loss": 0.0313, "step": 96450 }, { "grad_norm": 0.6138803362846375, "learning_rate": 3.424112951558978e-07, "loss": 0.0454, "step": 96460 }, { "grad_norm": 0.39172157645225525, "learning_rate": 3.40482240080503e-07, "loss": 0.025, "step": 96470 }, { "grad_norm": 0.3590300977230072, "learning_rate": 3.3855861569531133e-07, "loss": 0.0255, "step": 96480 }, { "grad_norm": 0.3703309893608093, "learning_rate": 3.3664042221068226e-07, "loss": 0.0257, "step": 96490 }, { "grad_norm": 0.3891972303390503, "learning_rate": 3.347276598363924e-07, "loss": 0.0265, "step": 96500 }, { "grad_norm": 0.5988304615020752, "learning_rate": 3.328203287816134e-07, "loss": 0.0276, "step": 96510 }, { "grad_norm": 0.5059558153152466, "learning_rate": 3.3091842925493386e-07, "loss": 0.0333, "step": 96520 }, { "grad_norm": 0.5460865497589111, "learning_rate": 3.2902196146433196e-07, "loss": 0.0286, "step": 96530 }, { "grad_norm": 0.31563693284988403, "learning_rate": 3.2713092561720836e-07, "loss": 0.0272, "step": 96540 }, { "grad_norm": 0.4674050807952881, "learning_rate": 3.2524532192036996e-07, "loss": 0.0295, "step": 96550 }, { "grad_norm": 0.16243474185466766, "learning_rate": 3.2336515058001836e-07, "loss": 0.0252, "step": 96560 }, { "grad_norm": 0.4978247284889221, "learning_rate": 3.2149041180176143e-07, "loss": 0.0227, "step": 96570 }, { "grad_norm": 0.5086909532546997, "learning_rate": 3.1962110579062954e-07, "loss": 0.0192, "step": 96580 }, { "grad_norm": 0.3479872941970825, "learning_rate": 3.177572327510314e-07, "loss": 0.0232, "step": 96590 }, { "grad_norm": 0.38550323247909546, "learning_rate": 3.158987928868151e-07, "loss": 0.0329, "step": 96600 }, { "grad_norm": 0.4265536367893219, "learning_rate": 3.1404578640120144e-07, "loss": 0.0365, "step": 96610 }, { "grad_norm": 0.37760475277900696, "learning_rate": 3.1219821349683933e-07, "loss": 0.0221, "step": 96620 }, { "grad_norm": 0.4073769748210907, "learning_rate": 3.103560743757783e-07, "loss": 0.0274, "step": 96630 }, { "grad_norm": 0.33404406905174255, "learning_rate": 3.0851936923946834e-07, "loss": 0.0227, "step": 96640 }, { "grad_norm": 0.31410133838653564, "learning_rate": 3.06688098288771e-07, "loss": 0.0228, "step": 96650 }, { "grad_norm": 0.33674731850624084, "learning_rate": 3.0486226172394825e-07, "loss": 0.0234, "step": 96660 }, { "grad_norm": 0.1788473278284073, "learning_rate": 3.030418597446738e-07, "loss": 0.0216, "step": 96670 }, { "grad_norm": 0.25939425826072693, "learning_rate": 3.012268925500272e-07, "loss": 0.0282, "step": 96680 }, { "grad_norm": 0.40094783902168274, "learning_rate": 2.994173603384831e-07, "loss": 0.0224, "step": 96690 }, { "grad_norm": 0.47297441959381104, "learning_rate": 2.9761326330793317e-07, "loss": 0.034, "step": 96700 }, { "grad_norm": 0.7306125164031982, "learning_rate": 2.958146016556751e-07, "loss": 0.0268, "step": 96710 }, { "grad_norm": 0.21459956467151642, "learning_rate": 2.94021375578396e-07, "loss": 0.0364, "step": 96720 }, { "grad_norm": 0.3400217890739441, "learning_rate": 2.9223358527221135e-07, "loss": 0.0302, "step": 96730 }, { "grad_norm": 0.3921263813972473, "learning_rate": 2.904512309326257e-07, "loss": 0.0309, "step": 96740 }, { "grad_norm": 0.3725506067276001, "learning_rate": 2.8867431275455545e-07, "loss": 0.0344, "step": 96750 }, { "grad_norm": 0.44003263115882874, "learning_rate": 2.86902830932323e-07, "loss": 0.0413, "step": 96760 }, { "grad_norm": 0.3762577772140503, "learning_rate": 2.851367856596565e-07, "loss": 0.0354, "step": 96770 }, { "grad_norm": 0.4440850615501404, "learning_rate": 2.83376177129685e-07, "loss": 0.0299, "step": 96780 }, { "grad_norm": 0.3641113042831421, "learning_rate": 2.8162100553494884e-07, "loss": 0.0343, "step": 96790 }, { "grad_norm": 0.3779636025428772, "learning_rate": 2.798712710673834e-07, "loss": 0.0244, "step": 96800 }, { "grad_norm": 0.35797545313835144, "learning_rate": 2.7812697391834676e-07, "loss": 0.0293, "step": 96810 }, { "grad_norm": 0.48154982924461365, "learning_rate": 2.7638811427858627e-07, "loss": 0.0292, "step": 96820 }, { "grad_norm": 0.6706401705741882, "learning_rate": 2.746546923382665e-07, "loss": 0.0276, "step": 96830 }, { "grad_norm": 0.3933550715446472, "learning_rate": 2.729267082869469e-07, "loss": 0.029, "step": 96840 }, { "grad_norm": 0.36301901936531067, "learning_rate": 2.712041623135986e-07, "loss": 0.0294, "step": 96850 }, { "grad_norm": 0.38354411721229553, "learning_rate": 2.694870546065931e-07, "loss": 0.0365, "step": 96860 }, { "grad_norm": 0.5563824772834778, "learning_rate": 2.677753853537246e-07, "loss": 0.0314, "step": 96870 }, { "grad_norm": 0.7678094506263733, "learning_rate": 2.660691547421601e-07, "loss": 0.032, "step": 96880 }, { "grad_norm": 0.5204772353172302, "learning_rate": 2.6436836295850033e-07, "loss": 0.0279, "step": 96890 }, { "grad_norm": 0.6881804466247559, "learning_rate": 2.6267301018874093e-07, "loss": 0.0278, "step": 96900 }, { "grad_norm": 0.1926296353340149, "learning_rate": 2.6098309661828355e-07, "loss": 0.0221, "step": 96910 }, { "grad_norm": 0.3908654451370239, "learning_rate": 2.592986224319305e-07, "loss": 0.0268, "step": 96920 }, { "grad_norm": 0.4738430082798004, "learning_rate": 2.576195878139009e-07, "loss": 0.0292, "step": 96930 }, { "grad_norm": 0.41228821873664856, "learning_rate": 2.5594599294780364e-07, "loss": 0.0258, "step": 96940 }, { "grad_norm": 0.22409643232822418, "learning_rate": 2.542778380166644e-07, "loss": 0.0346, "step": 96950 }, { "grad_norm": 0.21078357100486755, "learning_rate": 2.526151232029095e-07, "loss": 0.0326, "step": 96960 }, { "grad_norm": 0.3401053249835968, "learning_rate": 2.5095784868836567e-07, "loss": 0.0228, "step": 96970 }, { "grad_norm": 0.4769571125507355, "learning_rate": 2.493060146542825e-07, "loss": 0.0369, "step": 96980 }, { "grad_norm": 0.25492382049560547, "learning_rate": 2.476596212812876e-07, "loss": 0.0234, "step": 96990 }, { "grad_norm": 0.35298314690589905, "learning_rate": 2.4601866874943703e-07, "loss": 0.0294, "step": 97000 }, { "grad_norm": 0.38299399614334106, "learning_rate": 2.4438315723818163e-07, "loss": 0.0266, "step": 97010 }, { "grad_norm": 0.24271468818187714, "learning_rate": 2.4275308692637833e-07, "loss": 0.0241, "step": 97020 }, { "grad_norm": 0.34801778197288513, "learning_rate": 2.411284579922901e-07, "loss": 0.0256, "step": 97030 }, { "grad_norm": 0.33960041403770447, "learning_rate": 2.395092706135804e-07, "loss": 0.0267, "step": 97040 }, { "grad_norm": 0.23037147521972656, "learning_rate": 2.3789552496731872e-07, "loss": 0.0308, "step": 97050 }, { "grad_norm": 0.3913806676864624, "learning_rate": 2.3628722122999158e-07, "loss": 0.0352, "step": 97060 }, { "grad_norm": 0.5519087314605713, "learning_rate": 2.3468435957747503e-07, "loss": 0.0312, "step": 97070 }, { "grad_norm": 0.17679114639759064, "learning_rate": 2.330869401850566e-07, "loss": 0.0266, "step": 97080 }, { "grad_norm": 0.36541929841041565, "learning_rate": 2.3149496322742437e-07, "loss": 0.0212, "step": 97090 }, { "grad_norm": 0.6251248717308044, "learning_rate": 2.2990842887868346e-07, "loss": 0.0263, "step": 97100 }, { "grad_norm": 0.21162430942058563, "learning_rate": 2.2832733731232292e-07, "loss": 0.0305, "step": 97110 }, { "grad_norm": 0.4296448826789856, "learning_rate": 2.2675168870124886e-07, "loss": 0.0281, "step": 97120 }, { "grad_norm": 0.3666233718395233, "learning_rate": 2.2518148321778455e-07, "loss": 0.0278, "step": 97130 }, { "grad_norm": 0.3684118092060089, "learning_rate": 2.2361672103363706e-07, "loss": 0.0296, "step": 97140 }, { "grad_norm": 0.4554966390132904, "learning_rate": 2.220574023199251e-07, "loss": 0.0237, "step": 97150 }, { "grad_norm": 0.4572577476501465, "learning_rate": 2.2050352724717894e-07, "loss": 0.0291, "step": 97160 }, { "grad_norm": 0.394679456949234, "learning_rate": 2.1895509598532372e-07, "loss": 0.0313, "step": 97170 }, { "grad_norm": 0.42255374789237976, "learning_rate": 2.1741210870369066e-07, "loss": 0.0342, "step": 97180 }, { "grad_norm": 1.1873371601104736, "learning_rate": 2.1587456557102814e-07, "loss": 0.0301, "step": 97190 }, { "grad_norm": 0.2600274980068207, "learning_rate": 2.143424667554683e-07, "loss": 0.0422, "step": 97200 }, { "grad_norm": 0.21528522670269012, "learning_rate": 2.128158124245716e-07, "loss": 0.0308, "step": 97210 }, { "grad_norm": 0.22012034058570862, "learning_rate": 2.1129460274527668e-07, "loss": 0.0229, "step": 97220 }, { "grad_norm": 0.2108943909406662, "learning_rate": 2.0977883788395046e-07, "loss": 0.0286, "step": 97230 }, { "grad_norm": 0.3708326816558838, "learning_rate": 2.0826851800634928e-07, "loss": 0.0244, "step": 97240 }, { "grad_norm": 0.49077773094177246, "learning_rate": 2.0676364327764653e-07, "loss": 0.0399, "step": 97250 }, { "grad_norm": 0.563871443271637, "learning_rate": 2.0526421386240502e-07, "loss": 0.0314, "step": 97260 }, { "grad_norm": 0.3621498942375183, "learning_rate": 2.0377022992461025e-07, "loss": 0.0325, "step": 97270 }, { "grad_norm": 0.6081615090370178, "learning_rate": 2.0228169162763156e-07, "loss": 0.035, "step": 97280 }, { "grad_norm": 0.7095110416412354, "learning_rate": 2.0079859913425536e-07, "loss": 0.0293, "step": 97290 }, { "grad_norm": 0.4199410378932953, "learning_rate": 1.9932095260667417e-07, "loss": 0.0418, "step": 97300 }, { "grad_norm": 0.6125301718711853, "learning_rate": 1.9784875220648092e-07, "loss": 0.0273, "step": 97310 }, { "grad_norm": 0.22088828682899475, "learning_rate": 1.9638199809466907e-07, "loss": 0.0174, "step": 97320 }, { "grad_norm": 0.43390634655952454, "learning_rate": 1.9492069043164364e-07, "loss": 0.0336, "step": 97330 }, { "grad_norm": 0.6142668724060059, "learning_rate": 1.9346482937721567e-07, "loss": 0.0312, "step": 97340 }, { "grad_norm": 0.42001232504844666, "learning_rate": 1.920144150905856e-07, "loss": 0.0288, "step": 97350 }, { "grad_norm": 0.8274469375610352, "learning_rate": 1.9056944773037655e-07, "loss": 0.0397, "step": 97360 }, { "grad_norm": 0.3343828618526459, "learning_rate": 1.8912992745460102e-07, "loss": 0.0274, "step": 97370 }, { "grad_norm": 0.21722593903541565, "learning_rate": 1.876958544206886e-07, "loss": 0.0212, "step": 97380 }, { "grad_norm": 0.498104453086853, "learning_rate": 1.8626722878546942e-07, "loss": 0.035, "step": 97390 }, { "grad_norm": 0.4235454201698303, "learning_rate": 1.8484405070516851e-07, "loss": 0.0317, "step": 97400 }, { "grad_norm": 0.2672155201435089, "learning_rate": 1.8342632033542805e-07, "loss": 0.0299, "step": 97410 }, { "grad_norm": 0.32600724697113037, "learning_rate": 1.8201403783129066e-07, "loss": 0.0215, "step": 97420 }, { "grad_norm": 0.616227924823761, "learning_rate": 1.806072033471884e-07, "loss": 0.0269, "step": 97430 }, { "grad_norm": 0.3238154947757721, "learning_rate": 1.79205817036987e-07, "loss": 0.029, "step": 97440 }, { "grad_norm": 0.3054049015045166, "learning_rate": 1.7780987905393065e-07, "loss": 0.0216, "step": 97450 }, { "grad_norm": 0.3173341453075409, "learning_rate": 1.764193895506805e-07, "loss": 0.0265, "step": 97460 }, { "grad_norm": 0.32368966937065125, "learning_rate": 1.7503434867929824e-07, "loss": 0.0287, "step": 97470 }, { "grad_norm": 0.5425903797149658, "learning_rate": 1.7365475659125165e-07, "loss": 0.0239, "step": 97480 }, { "grad_norm": 0.49729904532432556, "learning_rate": 1.7228061343740332e-07, "loss": 0.0264, "step": 97490 }, { "grad_norm": 0.3257750868797302, "learning_rate": 1.7091191936803308e-07, "loss": 0.0278, "step": 97500 }, { "grad_norm": 0.23834648728370667, "learning_rate": 1.6954867453281563e-07, "loss": 0.0295, "step": 97510 }, { "grad_norm": 0.2626015543937683, "learning_rate": 1.6819087908084285e-07, "loss": 0.0298, "step": 97520 }, { "grad_norm": 0.46460142731666565, "learning_rate": 1.6683853316058485e-07, "loss": 0.0272, "step": 97530 }, { "grad_norm": 0.4528445899486542, "learning_rate": 1.6549163691995106e-07, "loss": 0.0289, "step": 97540 }, { "grad_norm": 0.24857081472873688, "learning_rate": 1.641501905062237e-07, "loss": 0.0372, "step": 97550 }, { "grad_norm": 0.3917670249938965, "learning_rate": 1.628141940661021e-07, "loss": 0.0289, "step": 97560 }, { "grad_norm": 0.3114786148071289, "learning_rate": 1.6148364774569158e-07, "loss": 0.0284, "step": 97570 }, { "grad_norm": 0.3291013538837433, "learning_rate": 1.6015855169050355e-07, "loss": 0.0468, "step": 97580 }, { "grad_norm": 0.5326487421989441, "learning_rate": 1.5883890604543873e-07, "loss": 0.0284, "step": 97590 }, { "grad_norm": 0.46264973282814026, "learning_rate": 1.5752471095482057e-07, "loss": 0.0239, "step": 97600 }, { "grad_norm": 0.4793122708797455, "learning_rate": 1.5621596656235638e-07, "loss": 0.0258, "step": 97610 }, { "grad_norm": 0.6791744232177734, "learning_rate": 1.5491267301117608e-07, "loss": 0.028, "step": 97620 }, { "grad_norm": 0.3181053400039673, "learning_rate": 1.536148304438101e-07, "loss": 0.0298, "step": 97630 }, { "grad_norm": 0.42998573184013367, "learning_rate": 1.5232243900217823e-07, "loss": 0.0266, "step": 97640 }, { "grad_norm": 0.2999787926673889, "learning_rate": 1.51035498827623e-07, "loss": 0.0284, "step": 97650 }, { "grad_norm": 0.4422798156738281, "learning_rate": 1.497540100608763e-07, "loss": 0.0439, "step": 97660 }, { "grad_norm": 0.3894735276699066, "learning_rate": 1.4847797284208153e-07, "loss": 0.0262, "step": 97670 }, { "grad_norm": 0.41519081592559814, "learning_rate": 1.472073873107882e-07, "loss": 0.0387, "step": 97680 }, { "grad_norm": 0.3357103765010834, "learning_rate": 1.459422536059407e-07, "loss": 0.0335, "step": 97690 }, { "grad_norm": 0.6136800050735474, "learning_rate": 1.4468257186589508e-07, "loss": 0.0265, "step": 97700 }, { "grad_norm": 0.37044355273246765, "learning_rate": 1.4342834222840217e-07, "loss": 0.0305, "step": 97710 }, { "grad_norm": 0.3601464033126831, "learning_rate": 1.4217956483063566e-07, "loss": 0.0207, "step": 97720 }, { "grad_norm": 0.4194265902042389, "learning_rate": 1.409362398091474e-07, "loss": 0.03, "step": 97730 }, { "grad_norm": 0.38714155554771423, "learning_rate": 1.3969836729990638e-07, "loss": 0.0303, "step": 97740 }, { "grad_norm": 0.6607015132904053, "learning_rate": 1.384659474382932e-07, "loss": 0.034, "step": 97750 }, { "grad_norm": 0.5523833632469177, "learning_rate": 1.372389803590779e-07, "loss": 0.0325, "step": 97760 }, { "grad_norm": 0.37916454672813416, "learning_rate": 1.360174661964364e-07, "loss": 0.035, "step": 97770 }, { "grad_norm": 0.7678837180137634, "learning_rate": 1.3480140508396188e-07, "loss": 0.0255, "step": 97780 }, { "grad_norm": 0.4133986234664917, "learning_rate": 1.335907971546313e-07, "loss": 0.0293, "step": 97790 }, { "grad_norm": 0.43116798996925354, "learning_rate": 1.3238564254083875e-07, "loss": 0.0349, "step": 97800 }, { "grad_norm": 0.2504480481147766, "learning_rate": 1.3118594137437323e-07, "loss": 0.0216, "step": 97810 }, { "grad_norm": 0.3262901306152344, "learning_rate": 1.2999169378644094e-07, "loss": 0.0339, "step": 97820 }, { "grad_norm": 0.6987360715866089, "learning_rate": 1.2880289990763738e-07, "loss": 0.0338, "step": 97830 }, { "grad_norm": 0.213764026761055, "learning_rate": 1.2761955986796968e-07, "loss": 0.0317, "step": 97840 }, { "grad_norm": 0.12644962966442108, "learning_rate": 1.264416737968399e-07, "loss": 0.0323, "step": 97850 }, { "grad_norm": 0.5886862277984619, "learning_rate": 1.2526924182307275e-07, "loss": 0.0252, "step": 97860 }, { "grad_norm": 0.3808954954147339, "learning_rate": 1.2410226407487124e-07, "loss": 0.0332, "step": 97870 }, { "grad_norm": 0.31886112689971924, "learning_rate": 1.229407406798555e-07, "loss": 0.0307, "step": 97880 }, { "grad_norm": 0.26376253366470337, "learning_rate": 1.2178467176505726e-07, "loss": 0.0347, "step": 97890 }, { "grad_norm": 0.31996333599090576, "learning_rate": 1.2063405745689205e-07, "loss": 0.0239, "step": 97900 }, { "grad_norm": 0.2856009304523468, "learning_rate": 1.1948889788119812e-07, "loss": 0.0257, "step": 97910 }, { "grad_norm": 0.5959815979003906, "learning_rate": 1.1834919316320303e-07, "loss": 0.0338, "step": 97920 }, { "grad_norm": 0.8975791931152344, "learning_rate": 1.1721494342754048e-07, "loss": 0.0349, "step": 97930 }, { "grad_norm": 0.3658182621002197, "learning_rate": 1.1608614879825563e-07, "loss": 0.0232, "step": 97940 }, { "grad_norm": 0.39313727617263794, "learning_rate": 1.1496280939879422e-07, "loss": 0.0285, "step": 97950 }, { "grad_norm": 0.36976829171180725, "learning_rate": 1.1384492535199687e-07, "loss": 0.0273, "step": 97960 }, { "grad_norm": 0.22264079749584198, "learning_rate": 1.1273249678011578e-07, "loss": 0.0243, "step": 97970 }, { "grad_norm": 0.43493014574050903, "learning_rate": 1.1162552380480362e-07, "loss": 0.0285, "step": 97980 }, { "grad_norm": 0.3930901288986206, "learning_rate": 1.1052400654711359e-07, "loss": 0.0247, "step": 97990 }, { "grad_norm": 0.5602573752403259, "learning_rate": 1.0942794512751597e-07, "loss": 0.0316, "step": 98000 }, { "grad_norm": 0.29628613591194153, "learning_rate": 1.0833733966587045e-07, "loss": 0.0245, "step": 98010 }, { "grad_norm": 0.348138689994812, "learning_rate": 1.0725219028143718e-07, "loss": 0.0303, "step": 98020 }, { "grad_norm": 0.4677716791629791, "learning_rate": 1.0617249709289345e-07, "loss": 0.0232, "step": 98030 }, { "grad_norm": 0.5208607316017151, "learning_rate": 1.0509826021831703e-07, "loss": 0.0323, "step": 98040 }, { "grad_norm": 0.37582629919052124, "learning_rate": 1.0402947977517508e-07, "loss": 0.0337, "step": 98050 }, { "grad_norm": 0.3277168869972229, "learning_rate": 1.0296615588035185e-07, "loss": 0.0265, "step": 98060 }, { "grad_norm": 1.0559778213500977, "learning_rate": 1.0190828865012659e-07, "loss": 0.0293, "step": 98070 }, { "grad_norm": 0.26229649782180786, "learning_rate": 1.008558782001956e-07, "loss": 0.0279, "step": 98080 }, { "grad_norm": 0.4301419258117676, "learning_rate": 9.980892464563906e-08, "loss": 0.0314, "step": 98090 }, { "grad_norm": 0.313920795917511, "learning_rate": 9.876742810095985e-08, "loss": 0.021, "step": 98100 }, { "grad_norm": 0.27667710185050964, "learning_rate": 9.77313886800446e-08, "loss": 0.0236, "step": 98110 }, { "grad_norm": 0.5210641026496887, "learning_rate": 9.670080649619717e-08, "loss": 0.0377, "step": 98120 }, { "grad_norm": 0.308727890253067, "learning_rate": 9.567568166212737e-08, "loss": 0.0365, "step": 98130 }, { "grad_norm": 0.5975943207740784, "learning_rate": 9.465601428992888e-08, "loss": 0.031, "step": 98140 }, { "grad_norm": 0.3692285716533661, "learning_rate": 9.364180449111803e-08, "loss": 0.03, "step": 98150 }, { "grad_norm": 0.3098059296607971, "learning_rate": 9.263305237661169e-08, "loss": 0.0275, "step": 98160 }, { "grad_norm": 0.5277913808822632, "learning_rate": 9.162975805671603e-08, "loss": 0.0295, "step": 98170 }, { "grad_norm": 0.371830552816391, "learning_rate": 9.063192164115442e-08, "loss": 0.0261, "step": 98180 }, { "grad_norm": 0.5599265694618225, "learning_rate": 8.963954323904511e-08, "loss": 0.0241, "step": 98190 }, { "grad_norm": 0.3841567933559418, "learning_rate": 8.865262295891796e-08, "loss": 0.0276, "step": 98200 }, { "grad_norm": 0.43049198389053345, "learning_rate": 8.767116090870331e-08, "loss": 0.0264, "step": 98210 }, { "grad_norm": 0.6887525320053101, "learning_rate": 8.669515719572086e-08, "loss": 0.023, "step": 98220 }, { "grad_norm": 0.4509129822254181, "learning_rate": 8.572461192671855e-08, "loss": 0.0346, "step": 98230 }, { "grad_norm": 0.6137838363647461, "learning_rate": 8.475952520782815e-08, "loss": 0.0286, "step": 98240 }, { "grad_norm": 0.34197595715522766, "learning_rate": 8.379989714458747e-08, "loss": 0.0201, "step": 98250 }, { "grad_norm": 0.5329421758651733, "learning_rate": 8.284572784194034e-08, "loss": 0.028, "step": 98260 }, { "grad_norm": 0.38403162360191345, "learning_rate": 8.189701740424216e-08, "loss": 0.0357, "step": 98270 }, { "grad_norm": 0.4906036853790283, "learning_rate": 8.095376593522663e-08, "loss": 0.0273, "step": 98280 }, { "grad_norm": 0.38366106152534485, "learning_rate": 8.001597353806123e-08, "loss": 0.0316, "step": 98290 }, { "grad_norm": 0.376476526260376, "learning_rate": 7.90836403152917e-08, "loss": 0.027, "step": 98300 }, { "grad_norm": 0.32109561562538147, "learning_rate": 7.815676636888092e-08, "loss": 0.0321, "step": 98310 }, { "grad_norm": 1.7581872940063477, "learning_rate": 7.723535180019226e-08, "loss": 0.03, "step": 98320 }, { "grad_norm": 0.5509597063064575, "learning_rate": 7.6319396709984e-08, "loss": 0.0293, "step": 98330 }, { "grad_norm": 0.5446537733078003, "learning_rate": 7.540890119843158e-08, "loss": 0.0307, "step": 98340 }, { "grad_norm": 0.22468246519565582, "learning_rate": 7.450386536509979e-08, "loss": 0.0232, "step": 98350 }, { "grad_norm": 0.513210117816925, "learning_rate": 7.360428930895947e-08, "loss": 0.0341, "step": 98360 }, { "grad_norm": 0.39715659618377686, "learning_rate": 7.271017312839302e-08, "loss": 0.0323, "step": 98370 }, { "grad_norm": 0.3315585255622864, "learning_rate": 7.182151692118333e-08, "loss": 0.0322, "step": 98380 }, { "grad_norm": 0.44432249665260315, "learning_rate": 7.093832078449713e-08, "loss": 0.034, "step": 98390 }, { "grad_norm": 0.33104780316352844, "learning_rate": 7.006058481493493e-08, "loss": 0.027, "step": 98400 }, { "grad_norm": 0.6221133470535278, "learning_rate": 6.918830910847552e-08, "loss": 0.0301, "step": 98410 }, { "grad_norm": 0.32031959295272827, "learning_rate": 6.832149376052033e-08, "loss": 0.0248, "step": 98420 }, { "grad_norm": 0.26048269867897034, "learning_rate": 6.746013886584912e-08, "loss": 0.0287, "step": 98430 }, { "grad_norm": 0.30909132957458496, "learning_rate": 6.660424451866431e-08, "loss": 0.0284, "step": 98440 }, { "grad_norm": 0.25919243693351746, "learning_rate": 6.575381081256327e-08, "loss": 0.0321, "step": 98450 }, { "grad_norm": 0.42633309960365295, "learning_rate": 6.49088378405549e-08, "loss": 0.031, "step": 98460 }, { "grad_norm": 0.2720020115375519, "learning_rate": 6.406932569503755e-08, "loss": 0.0232, "step": 98470 }, { "grad_norm": 0.4375864565372467, "learning_rate": 6.323527446782662e-08, "loss": 0.0295, "step": 98480 }, { "grad_norm": 0.45498114824295044, "learning_rate": 6.240668425012142e-08, "loss": 0.0278, "step": 98490 }, { "grad_norm": 0.24276509881019592, "learning_rate": 6.158355513254388e-08, "loss": 0.02, "step": 98500 }, { "grad_norm": 0.4015502631664276, "learning_rate": 6.076588720510534e-08, "loss": 0.031, "step": 98510 }, { "grad_norm": 0.6302769780158997, "learning_rate": 5.995368055722316e-08, "loss": 0.0324, "step": 98520 }, { "grad_norm": 0.5092366933822632, "learning_rate": 5.914693527773185e-08, "loss": 0.0356, "step": 98530 }, { "grad_norm": 0.8327288031578064, "learning_rate": 5.834565145483861e-08, "loss": 0.0279, "step": 98540 }, { "grad_norm": 0.5116546154022217, "learning_rate": 5.754982917618446e-08, "loss": 0.0263, "step": 98550 }, { "grad_norm": 0.30791720747947693, "learning_rate": 5.675946852878866e-08, "loss": 0.034, "step": 98560 }, { "grad_norm": 0.8501913547515869, "learning_rate": 5.59745695990932e-08, "loss": 0.027, "step": 98570 }, { "grad_norm": 0.3663977086544037, "learning_rate": 5.51951324729294e-08, "loss": 0.0302, "step": 98580 }, { "grad_norm": 0.5539485812187195, "learning_rate": 5.4421157235529093e-08, "loss": 0.0258, "step": 98590 }, { "grad_norm": 0.3953344225883484, "learning_rate": 5.365264397154679e-08, "loss": 0.0306, "step": 98600 }, { "grad_norm": 0.431816041469574, "learning_rate": 5.288959276501526e-08, "loss": 0.0331, "step": 98610 }, { "grad_norm": 0.6230515837669373, "learning_rate": 5.2132003699378876e-08, "loss": 0.0412, "step": 98620 }, { "grad_norm": 0.2340092658996582, "learning_rate": 5.137987685749912e-08, "loss": 0.0243, "step": 98630 }, { "grad_norm": 0.4916665852069855, "learning_rate": 5.0633212321610227e-08, "loss": 0.0275, "step": 98640 }, { "grad_norm": 0.45434287190437317, "learning_rate": 4.989201017338019e-08, "loss": 0.0295, "step": 98650 }, { "grad_norm": 0.29741615056991577, "learning_rate": 4.9156270493860846e-08, "loss": 0.026, "step": 98660 }, { "grad_norm": 0.5227062702178955, "learning_rate": 4.842599336351561e-08, "loss": 0.0225, "step": 98670 }, { "grad_norm": 0.2630140483379364, "learning_rate": 4.7701178862197274e-08, "loss": 0.027, "step": 98680 }, { "grad_norm": 0.36306989192962646, "learning_rate": 4.698182706918131e-08, "loss": 0.0227, "step": 98690 }, { "grad_norm": 0.3782539367675781, "learning_rate": 4.6267938063121466e-08, "loss": 0.0281, "step": 98700 }, { "grad_norm": 0.3796144127845764, "learning_rate": 4.555951192209973e-08, "loss": 0.0324, "step": 98710 }, { "grad_norm": 0.29812419414520264, "learning_rate": 4.485654872358747e-08, "loss": 0.024, "step": 98720 }, { "grad_norm": 0.4791257381439209, "learning_rate": 4.415904854445097e-08, "loss": 0.0334, "step": 98730 }, { "grad_norm": 0.4339795708656311, "learning_rate": 4.34670114609792e-08, "loss": 0.0311, "step": 98740 }, { "grad_norm": 0.4207102060317993, "learning_rate": 4.278043754884498e-08, "loss": 0.023, "step": 98750 }, { "grad_norm": 0.2643912136554718, "learning_rate": 4.209932688312712e-08, "loss": 0.0301, "step": 98760 }, { "grad_norm": 0.5164753794670105, "learning_rate": 4.14236795383216e-08, "loss": 0.0315, "step": 98770 }, { "grad_norm": 0.33489754796028137, "learning_rate": 4.075349558830821e-08, "loss": 0.0303, "step": 98780 }, { "grad_norm": 0.5233510732650757, "learning_rate": 4.008877510638387e-08, "loss": 0.0174, "step": 98790 }, { "grad_norm": 0.22098232805728912, "learning_rate": 3.942951816523488e-08, "loss": 0.0315, "step": 98800 }, { "grad_norm": 0.5349354147911072, "learning_rate": 3.8775724836959125e-08, "loss": 0.0439, "step": 98810 }, { "grad_norm": 0.5503411293029785, "learning_rate": 3.812739519305497e-08, "loss": 0.0271, "step": 98820 }, { "grad_norm": 0.3449886739253998, "learning_rate": 3.748452930442126e-08, "loss": 0.0293, "step": 98830 }, { "grad_norm": 0.5998414754867554, "learning_rate": 3.6847127241362855e-08, "loss": 0.0265, "step": 98840 }, { "grad_norm": 0.4624760150909424, "learning_rate": 3.621518907358512e-08, "loss": 0.0376, "step": 98850 }, { "grad_norm": 0.4246055781841278, "learning_rate": 3.558871487019388e-08, "loss": 0.0285, "step": 98860 }, { "grad_norm": 0.3056695759296417, "learning_rate": 3.4967704699701007e-08, "loss": 0.0276, "step": 98870 }, { "grad_norm": 0.2637990117073059, "learning_rate": 3.435215863001884e-08, "loss": 0.0243, "step": 98880 }, { "grad_norm": 0.32652047276496887, "learning_rate": 3.374207672846019e-08, "loss": 0.035, "step": 98890 }, { "grad_norm": 0.44620922207832336, "learning_rate": 3.313745906174392e-08, "loss": 0.0232, "step": 98900 }, { "grad_norm": 0.5197644829750061, "learning_rate": 3.253830569599492e-08, "loss": 0.0296, "step": 98910 }, { "grad_norm": 0.3466712534427643, "learning_rate": 3.1944616696727436e-08, "loss": 0.0249, "step": 98920 }, { "grad_norm": 0.4634566307067871, "learning_rate": 3.135639212887287e-08, "loss": 0.026, "step": 98930 }, { "grad_norm": 0.5085391402244568, "learning_rate": 3.077363205675754e-08, "loss": 0.0329, "step": 98940 }, { "grad_norm": 0.3449084758758545, "learning_rate": 3.0196336544113804e-08, "loss": 0.0249, "step": 98950 }, { "grad_norm": 0.9409031271934509, "learning_rate": 2.9624505654063383e-08, "loss": 0.0306, "step": 98960 }, { "grad_norm": 0.3263271450996399, "learning_rate": 2.90581394491507e-08, "loss": 0.0317, "step": 98970 }, { "grad_norm": 0.5158741474151611, "learning_rate": 2.8497237991309545e-08, "loss": 0.0342, "step": 98980 }, { "grad_norm": 0.5512863993644714, "learning_rate": 2.7941801341879735e-08, "loss": 0.0318, "step": 98990 }, { "grad_norm": 0.5511667132377625, "learning_rate": 2.7391829561601578e-08, "loss": 0.0352, "step": 99000 }, { "grad_norm": 1.368149995803833, "learning_rate": 2.6847322710621403e-08, "loss": 0.0231, "step": 99010 }, { "grad_norm": 0.4729670584201813, "learning_rate": 2.6308280848486022e-08, "loss": 0.0221, "step": 99020 }, { "grad_norm": 0.3431306481361389, "learning_rate": 2.5774704034137177e-08, "loss": 0.0271, "step": 99030 }, { "grad_norm": 0.27403074502944946, "learning_rate": 2.524659232593929e-08, "loss": 0.0264, "step": 99040 }, { "grad_norm": 0.22447627782821655, "learning_rate": 2.472394578163506e-08, "loss": 0.0235, "step": 99050 }, { "grad_norm": 0.37289339303970337, "learning_rate": 2.4206764458378772e-08, "loss": 0.0237, "step": 99060 }, { "grad_norm": 0.5133536458015442, "learning_rate": 2.3695048412736285e-08, "loss": 0.0276, "step": 99070 }, { "grad_norm": 0.4177955687046051, "learning_rate": 2.3188797700668395e-08, "loss": 0.0281, "step": 99080 }, { "grad_norm": 0.5595746636390686, "learning_rate": 2.268801237753082e-08, "loss": 0.0242, "step": 99090 }, { "grad_norm": 0.31497496366500854, "learning_rate": 2.2192692498090862e-08, "loss": 0.0271, "step": 99100 }, { "grad_norm": 0.43846720457077026, "learning_rate": 2.170283811652185e-08, "loss": 0.0347, "step": 99110 }, { "grad_norm": 0.5829416513442993, "learning_rate": 2.12184492863865e-08, "loss": 0.0282, "step": 99120 }, { "grad_norm": 0.3672060966491699, "learning_rate": 2.073952606066465e-08, "loss": 0.0316, "step": 99130 }, { "grad_norm": 0.24784448742866516, "learning_rate": 2.026606849171997e-08, "loss": 0.0254, "step": 99140 }, { "grad_norm": 0.4294455945491791, "learning_rate": 1.9798076631333262e-08, "loss": 0.0309, "step": 99150 }, { "grad_norm": 0.39899367094039917, "learning_rate": 1.933555053069136e-08, "loss": 0.0305, "step": 99160 }, { "grad_norm": 0.3869991600513458, "learning_rate": 1.887849024036492e-08, "loss": 0.0225, "step": 99170 }, { "grad_norm": 0.7726337909698486, "learning_rate": 1.8426895810341736e-08, "loss": 0.0319, "step": 99180 }, { "grad_norm": 0.42100462317466736, "learning_rate": 1.798076729000453e-08, "loss": 0.0232, "step": 99190 }, { "grad_norm": 0.37480610609054565, "learning_rate": 1.75401047281476e-08, "loss": 0.0236, "step": 99200 }, { "grad_norm": 0.4501909911632538, "learning_rate": 1.7104908172954626e-08, "loss": 0.028, "step": 99210 }, { "grad_norm": 0.21502362191677094, "learning_rate": 1.6675177672020868e-08, "loss": 0.0227, "step": 99220 }, { "grad_norm": 0.6343684196472168, "learning_rate": 1.6250913272342073e-08, "loss": 0.0359, "step": 99230 }, { "grad_norm": 0.5128757357597351, "learning_rate": 1.5832115020314453e-08, "loss": 0.0379, "step": 99240 }, { "grad_norm": 0.24385367333889008, "learning_rate": 1.5418782961734712e-08, "loss": 0.0292, "step": 99250 }, { "grad_norm": 0.6034804582595825, "learning_rate": 1.501091714181113e-08, "loss": 0.0307, "step": 99260 }, { "grad_norm": 0.3516325354576111, "learning_rate": 1.460851760513582e-08, "loss": 0.0253, "step": 99270 }, { "grad_norm": 0.45687946677207947, "learning_rate": 1.4211584395723566e-08, "loss": 0.034, "step": 99280 }, { "grad_norm": 1.4401127099990845, "learning_rate": 1.382011755697854e-08, "loss": 0.0246, "step": 99290 }, { "grad_norm": 0.4309174120426178, "learning_rate": 1.3434117131716494e-08, "loss": 0.0204, "step": 99300 }, { "grad_norm": 0.47082456946372986, "learning_rate": 1.3053583162142557e-08, "loss": 0.0244, "step": 99310 }, { "grad_norm": 0.27319759130477905, "learning_rate": 1.2678515689873438e-08, "loss": 0.0316, "step": 99320 }, { "grad_norm": 0.29799866676330566, "learning_rate": 1.2308914755931877e-08, "loss": 0.0284, "step": 99330 }, { "grad_norm": 0.5415503978729248, "learning_rate": 1.1944780400729993e-08, "loss": 0.0281, "step": 99340 }, { "grad_norm": 0.5887593626976013, "learning_rate": 1.1586112664085935e-08, "loss": 0.0279, "step": 99350 }, { "grad_norm": 0.3593467175960541, "learning_rate": 1.1232911585234984e-08, "loss": 0.0292, "step": 99360 }, { "grad_norm": 0.4821520745754242, "learning_rate": 1.0885177202790697e-08, "loss": 0.0316, "step": 99370 }, { "grad_norm": 0.27218660712242126, "learning_rate": 1.0542909554789315e-08, "loss": 0.0309, "step": 99380 }, { "grad_norm": 0.3116128146648407, "learning_rate": 1.0206108678656456e-08, "loss": 0.0266, "step": 99390 }, { "grad_norm": 0.7940878868103027, "learning_rate": 9.874774611223769e-09, "loss": 0.0349, "step": 99400 }, { "grad_norm": 0.5327581167221069, "learning_rate": 9.548907388728933e-09, "loss": 0.0287, "step": 99410 }, { "grad_norm": 0.6054251790046692, "learning_rate": 9.228507046804557e-09, "loss": 0.0252, "step": 99420 }, { "grad_norm": 0.2372036576271057, "learning_rate": 8.91357362048928e-09, "loss": 0.0389, "step": 99430 }, { "grad_norm": 0.5732924342155457, "learning_rate": 8.604107144227769e-09, "loss": 0.0257, "step": 99440 }, { "grad_norm": 0.3307894170284271, "learning_rate": 8.300107651859623e-09, "loss": 0.0238, "step": 99450 }, { "grad_norm": 0.6447660326957703, "learning_rate": 8.001575176630472e-09, "loss": 0.0327, "step": 99460 }, { "grad_norm": 0.6505019664764404, "learning_rate": 7.708509751186422e-09, "loss": 0.0207, "step": 99470 }, { "grad_norm": 0.2938742935657501, "learning_rate": 7.420911407579611e-09, "loss": 0.0268, "step": 99480 }, { "grad_norm": 0.4851212203502655, "learning_rate": 7.13878017725711e-09, "loss": 0.0182, "step": 99490 }, { "grad_norm": 0.3176070749759674, "learning_rate": 6.8621160910720125e-09, "loss": 0.0374, "step": 99500 }, { "grad_norm": 0.38819408416748047, "learning_rate": 6.59091917928345e-09, "loss": 0.0291, "step": 99510 }, { "grad_norm": 0.42615795135498047, "learning_rate": 6.32518947155103e-09, "loss": 0.0296, "step": 99520 }, { "grad_norm": 0.4866655766963959, "learning_rate": 6.064926996929288e-09, "loss": 0.0271, "step": 99530 }, { "grad_norm": 0.16212530434131622, "learning_rate": 5.810131783884343e-09, "loss": 0.0222, "step": 99540 }, { "grad_norm": 0.29034513235092163, "learning_rate": 5.5608038602772415e-09, "loss": 0.0333, "step": 99550 }, { "grad_norm": 0.540028989315033, "learning_rate": 5.316943253375062e-09, "loss": 0.023, "step": 99560 }, { "grad_norm": 0.41631630063056946, "learning_rate": 5.07854998984536e-09, "loss": 0.0281, "step": 99570 }, { "grad_norm": 0.2463727742433548, "learning_rate": 4.845624095756174e-09, "loss": 0.025, "step": 99580 }, { "grad_norm": 0.3936184048652649, "learning_rate": 4.618165596587121e-09, "loss": 0.03, "step": 99590 }, { "grad_norm": 0.3476037085056305, "learning_rate": 4.396174517207197e-09, "loss": 0.0196, "step": 99600 }, { "grad_norm": 0.28489214181900024, "learning_rate": 4.179650881896979e-09, "loss": 0.0282, "step": 99610 }, { "grad_norm": 0.29792702198028564, "learning_rate": 3.968594714331975e-09, "loss": 0.0256, "step": 99620 }, { "grad_norm": 0.3340865969657898, "learning_rate": 3.763006037593719e-09, "loss": 0.0265, "step": 99630 }, { "grad_norm": 0.3715929090976715, "learning_rate": 3.5628848741586786e-09, "loss": 0.0243, "step": 99640 }, { "grad_norm": 0.5649616718292236, "learning_rate": 3.368231245926001e-09, "loss": 0.0285, "step": 99650 }, { "grad_norm": 0.4138302505016327, "learning_rate": 3.179045174167561e-09, "loss": 0.0254, "step": 99660 }, { "grad_norm": 0.3795851767063141, "learning_rate": 2.9953266795834657e-09, "loss": 0.0324, "step": 99670 }, { "grad_norm": 0.3428044319152832, "learning_rate": 2.817075782263201e-09, "loss": 0.0308, "step": 99680 }, { "grad_norm": 0.30662062764167786, "learning_rate": 2.6442925016967325e-09, "loss": 0.023, "step": 99690 }, { "grad_norm": 0.43336254358291626, "learning_rate": 2.4769768567800555e-09, "loss": 0.0311, "step": 99700 }, { "grad_norm": 0.669900119304657, "learning_rate": 2.315128865809646e-09, "loss": 0.0268, "step": 99710 }, { "grad_norm": 0.36107245087623596, "learning_rate": 2.1587485464880097e-09, "loss": 0.0247, "step": 99720 }, { "grad_norm": 0.32843002676963806, "learning_rate": 2.0078359159125813e-09, "loss": 0.0331, "step": 99730 }, { "grad_norm": 0.34020909667015076, "learning_rate": 1.8623909905923776e-09, "loss": 0.0309, "step": 99740 }, { "grad_norm": 0.3054705262184143, "learning_rate": 1.7224137864257916e-09, "loss": 0.0299, "step": 99750 }, { "grad_norm": 1.06900155544281, "learning_rate": 1.5879043187283505e-09, "loss": 0.03, "step": 99760 }, { "grad_norm": 0.45993557572364807, "learning_rate": 1.458862602204958e-09, "loss": 0.0241, "step": 99770 }, { "grad_norm": 0.36098918318748474, "learning_rate": 1.3352886509720996e-09, "loss": 0.0258, "step": 99780 }, { "grad_norm": 0.46536985039711, "learning_rate": 1.2171824785356389e-09, "loss": 0.0367, "step": 99790 }, { "grad_norm": 0.412888765335083, "learning_rate": 1.1045440978185718e-09, "loss": 0.028, "step": 99800 }, { "grad_norm": 0.2629603445529938, "learning_rate": 9.973735211332714e-10, "loss": 0.029, "step": 99810 }, { "grad_norm": 0.35459503531455994, "learning_rate": 8.956707602036929e-10, "loss": 0.0263, "step": 99820 }, { "grad_norm": 0.3820144236087799, "learning_rate": 7.994358261542712e-10, "loss": 0.0239, "step": 99830 }, { "grad_norm": 0.42273038625717163, "learning_rate": 7.086687295043692e-10, "loss": 0.0195, "step": 99840 }, { "grad_norm": 0.37453585863113403, "learning_rate": 6.23369480179381e-10, "loss": 0.0229, "step": 99850 }, { "grad_norm": 0.45298224687576294, "learning_rate": 5.43538087510731e-10, "loss": 0.0337, "step": 99860 }, { "grad_norm": 0.6380533576011658, "learning_rate": 4.691745602303233e-10, "loss": 0.03, "step": 99870 }, { "grad_norm": 0.5015013813972473, "learning_rate": 4.0027890646499034e-10, "loss": 0.0402, "step": 99880 }, { "grad_norm": 0.22713497281074524, "learning_rate": 3.368511337531466e-10, "loss": 0.0203, "step": 99890 }, { "grad_norm": 0.6697403192520142, "learning_rate": 2.788912490281348e-10, "loss": 0.0314, "step": 99900 }, { "grad_norm": 0.18701978027820587, "learning_rate": 2.2639925863487954e-10, "loss": 0.0248, "step": 99910 }, { "grad_norm": 0.289548397064209, "learning_rate": 1.7937516830213163e-10, "loss": 0.0222, "step": 99920 }, { "grad_norm": 1.031265377998352, "learning_rate": 1.3781898318687702e-10, "loss": 0.03, "step": 99930 }, { "grad_norm": 0.4466700553894043, "learning_rate": 1.0173070781882566e-10, "loss": 0.0335, "step": 99940 }, { "grad_norm": 0.27739450335502625, "learning_rate": 7.111034615592261e-11, "loss": 0.0319, "step": 99950 }, { "grad_norm": 0.2702809274196625, "learning_rate": 4.595790153993918e-11, "loss": 0.0224, "step": 99960 }, { "grad_norm": 0.7068172693252563, "learning_rate": 2.6273376729779585e-11, "loss": 0.0292, "step": 99970 }, { "grad_norm": 0.8527073860168457, "learning_rate": 1.2056773868174276e-11, "loss": 0.0246, "step": 99980 }, { "grad_norm": 0.276762455701828, "learning_rate": 3.3080945149865926e-12, "loss": 0.0224, "step": 99990 }, { "grad_norm": 0.33565399050712585, "learning_rate": 2.7339625008337977e-14, "loss": 0.0246, "step": 100000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }